Taxing Democracy Understanding Tax Avoidance and Evasion 0754622436 9780754622437
Taxing Democracy Understanding Tax Avoidance and Evasion 0754622436 9780754622437
Taxing Democracy Understanding Tax Avoidance and Evasion 0754622436 9780754622437
In memory of
Leslie Whittington
(1955-2001)
Taxing Democracy
Edited by
Valerie Braithwaite
The Australian National University, Canberra
First published 2003 by Ashgate Publishing
Notice:
Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
2002036105
Index
Chapter 1
A New Approach to Tax Compliance
Valerie Braithwaite
In the late 1990s, the Australian Taxation Office (ATO) underwent a series of
reforms that set the stage for a new proactive role in building a voluntary
taxpaying culture. The evaluation of these measures is being undertaken rather
more systematically than reforms in other countries, with results that have
implications for all nations' tax regimes. This process of reform is built on the
premise that although legislation is one of the basic building blocks for
compliance, it is far from sufficient. Tax law is contestable; it is also complex; and
it is not beyond the initiative of taxpayers to avoid and evade tax in ways that are
costly, both in terms of revenue that will never be collected and enforcement that
is resource intensive. The traditional tax infrastructure of law, auditors, penalties,
debt collectors, and court cases needs to be supplemented by measures that boost
taxpayers' commitment to paying tax with or without the tax authority watching
over their shoulders.
At the heart of the reform strategies of the late 1990s was the building of a
relationship with the Australian community in which the tax office was to be (a)
professional, responsive, fair, open, and accountable in helping taxpayers comply
with their tax obligations; as well as (b) effective in bringing to account those
who intentionally avoided their obligations. Through adopting such practices, the
intent was that the tax office earn (c) the trust, support and respect of the
community (Australian Taxation Office, 1997).
The first initiative toward building this relationship was the Taxpayers'
Charter (Australian Taxation Office, 1997). The Charter articulated 12 rights of
taxpayers and committed tax officers to treating taxpayers fairly and reasonably,
to explain decisions, assist with questions, and provide reliable information, to
respect taxpayer privacy, to keep the taxpayers' compliance costs to a minimum,
and to be accountable, if necessary, through independent review. The taxpayers'
obligations, articulated also in the Charter document, were four-fold and
involved being truthful in dealings with the tax office, keeping records in
accordance with the law, taking reasonable care in preparing tax information,
and lodging tax returns and required documents by the due date.
Bringing the Charter to life was no small challenge for the Australian Taxation
Office. The traditional regulatory style of the ATO has been heavily weighted
toward command and control with the automatic application of penalties for
various forms of non-compliance (see Chapter 6). At the same time, the authority
has not always used its prosecutorial powers effectively, with a history of slap-
on-the-wrist prosecutions that rarely touch major evaders or avoiders (Grabosky
and Braithwaite, 1986). In order for the tax office to change course, it was
necessary to show fairness and reasonableness to those who were willing to
cooperate, and focus enforcement capacity on those flagrantly ignoring their tax
obligations. The companion reform that addressed this issue and enabled the
Charter to be mainstreamed in ATO operations was the ATO Compliance Model.
Originating in the Cash Economy Task Force (1998),1 the Compliance Model drew
on the work of regulatory scholars at the Australian National University as well
as on the vast research literature on tax compliance. Consistent with this
literature (see Coleman and Freeman, 1997, for example), the Task Force urged
the ATO to better understand not only the business profiles of taxpayers (which
auditors traditionally, if partially, do), but also the nature of the industry they
belong to, the economic factors that impinge on that industry and society more
broadly, and the psychological and sociological factors that frame taxpayers'
decisions or non-decisions about the actions they will take to meet their tax
obligations. In the words of the Task Force:
none of these factors stand alone as the sole reason for a taxpayer's behaviour, and equally, it is not
possible to identify which factors in combination may influence the behaviour of any one particular
person. However, it is possible to identify a combination of factors that is more likely to influence
behaviour for certain categories of taxpayers (Cash Economy Task Force, 1998, p. 20).
On the left hand side of the model are the motivational postures. These are the
stances that taxpayers openly express in their relationships with the tax
authority. These postures were identified in earlier regulatory work (Braithwaite,
Braithwaite, Gibson and Makkai, 1994; Braithwaite, 1995) to describe the way in
which taxpayers controlled the amount of social distance they placed between
themselves and the tax office. When taxpayers were open to admitting
wrongdoing, correcting their mistakes, and getting on with meeting the law's
expectations, they were likely to be displaying the postures of commitment or
capitulation (see Chapter 2 for a more detailed description of the postures). The
tax official's task is relatively straightforward in such circumstances. Their
authority will be taken seriously, and compliance will follow as long as taxpayers
know what they are supposed to do, are treated in a procedurally just manner,
and are conscious of the fact that there will be follow through by the tax
authority if they do not comply.
The tax official's task becomes increasingly harder as taxpayers put more
social distance between themselves and the authority. Capitulation describes
giving in to authority without necessarily being prepared to take the initiative to
get things right in the future; when initiative is demonstrated, commitment is the
more apt description. In contrast, the postures of resistance and disengagement
reflect a conscious holding back of cooperation. The relationship is adversarial,
and the tax official's approach to gaining compliance needs to be more strategic
than would be necessary with more cooperative taxpayers. The most difficult
stance for a tax official to deal with in the model is disengagement. Here the
taxpayer has such contempt for the system that the chances of persuasion
working are low: In such circumstances, other strategies may be equally
ineffective, leaving incapacitation as the only option (through prosecution,
imprisonment, or taking away a license to practice).
The courses of action that a tax official can take in response to compliance
problems are many and varied (see Chapters 9, 11, 12), although generally
speaking, attention seems to focus on penalties in response to law breaking. As
important as these are to an effectively functioning tax system, compliance
problems are not always black and white in the field of taxation, and in such
circumstances, it is helpful for the tax office to have a range of tools at their
disposal to manage the compliance problem well. Possible courses of action that
tax officials can take are depicted within the framework of the Compliance Model
in Figure 1.1 (middle column). They are arranged to represent different levels of
seriousness and intrusiveness on the part of the tax office, the general thesis
being that if taxpayers are prepared to meet their obligations with minimum
interference by the tax office, they should be left alone to get on with it. Needless
to say, the courses of action will change as the nature of the compliance problem
changes: What is useful for cash economy problems will not necessarily apply to
other problems (see Chapter 9).
While local areas need to develop their own compliance strategies (Sparrow,
2000), the principles that guide enforcement are more stable and are represented
on the right hand side of the Compliance Model in Figure 1.1. For those who are
willing to cooperate, the principle guiding the choice of strategy is self-
regulation. If taxpayers are committed to correcting their own mistakes, they
should be encouraged and assisted in doing so. The next level of interference
might be called enforced self-regulation. Taxpayers have responsibility for
correcting their own mistakes, but a mechanism is in place to ensure they do so,
and to provide feedback to indicate whether or not the taxpayers' compliance
plan is sound.
Above these levels are the traditional principles of tax office enforcement that
have a command and control quality. A soft version of command and control
regulation is to wave a big stick, but to exercise discretion around using
punishment to improve compliance. The hard version is non-discretionary
punishment such that a sanction automatically follows when non-compliance is
detected, regardless of the circumstances. The principles of enforcement, from
the bottom to the top of the pyramid, involve a transfer of power from the
taxpayer to the tax office, and a concomitant loss of freedom on the part of the
taxpayer.
At the heart of the Compliance Model are the concepts of responsive
regulation and regulatory pyramids to guide an authority's response to non-
compliance (Ayres and Braithwaite, 1992). Responsive regulation steps away from
a command and control approach to regulation and moves regulators beyond a
mentality that if they go strictly by the book in dealing with non-compliance,
their problems will be over. A considerable research literature supports the
failings of command and control regulation when applied indiscriminately in
areas where compliance and non-compliance are multi-faceted and complex
phenomena (Bardach and Kagan, 1982; Gunningham and Grabosky, 1998;
Sparrow, 2000). It is very easy for a regulatory agency relying on a simplistic
conception of enforcement to fall foul of accusations of unreasonableness and
unfairness (Bardach and Kagan, 1982). Tax administrations across the world have
received their share of criticism of this kind (Report of the National Commission
on Restructuring the Internal Revenue Service, 1997; The Report of the
Committee of Experts on Tax Compliance, 1999; Senate Economics References
Committee, 2001).
Regulatory pyramids are designed to promote self-regulation and they
advocate only as much enforcement by the authority as the situation requires to
gain compliance. Two basic assumptions underlie their effective use. First, most
of the population are assumed to be located at the base of the pyramid. In other
words, costly enforcement resources are not wasted on those who are willing to
comply, but are reserved for the smaller proportion of the population not willing
to cooperate with the authority, clustering around the higher levels of the
pyramid. Second, regulatory pyramids demand of an authority the capacity and
readiness to follow through on non-compliance, escalating the costs to the
taxpayer to the point of incapacitation if necessary.
The essential compliance generating dynamic of the pyramid relies on
knowing that it is less costly to resolve a problem at the bottom of the pyramid
than to allow it to escalate to the top of the pyramid. This applies to both
regulator and regulatee, and incorporates costs of a material, social or
psychological kind. The knowledge that drives behaviour comes into play at two
levels. First, the cooperative approach at the bottom of the pyramid involves
persuasion in both directions: Taxpayers have the opportunity to persuade the tax
office at the same time as the tax office is trying to persuade the taxpayer. As the
conflict moves up the pyramid, taxpayers lose power to persuade as the tax office
moves into command and control mode. Second, the regulatory pyramid
communicates consequences of non-compliance, and most importantly, it signals
that the delivery of the consequences is contingent upon the next move of the
regulatee. If the regulatee chooses a cooperative response, the regulator
cooperates. If the regulatee's choice is uncooperative, the regulator moves to a
higher level of enforcement that imposes higher costs on the non-complier. This
treatment sits comfortably alongside the Taxpayers' Charter with its in-built
concept of procedural justice (see Chapter 3).
The implementation of the ATO Compliance Model mainstreamed the
Taxpayers' Charter, but at the same time challenged traditional ways of working
beyond the human management system. The administrative, technical and legal
systems were implicated in the change process as well (see Chapter 8). Foremost
in everyone's mind was the fact that the ATO Compliance Model came about
through transplanting ideas developed in other regulatory contexts to the field of
taxation (see Chapter 6). Adaptation across fields was based on intuitive
judgement, and to a considerable extent faith on the part of ATO senior
management that it might work (see Chapter 7). In order to progress the use of
the Model judiciously, there was a need to monitor and evaluate its effectiveness,
making adjustments where required (see Chapter 8). With this purpose in mind, a
six-year research partnership was set up between the Australian Taxation Office
and the Australian National University in August, 1999 with the establishment of
the Centre for Tax System Integrity.
The research program for the Centre for Tax System Integrity revolves around
three questions that are relevant to tax administrations around the world: (a)
What options do democratic states have for cultivating a voluntary taxpaying
culture? (b) How practicable and desirable is the ATO Compliance Model for
managing tax compliance and non-compliance? (c) Can evidence-based tax
administration be built around a continuing program of experimentation that
builds an increasingly rich tool-kit of cost-effective strategies (represented in the
middle column of the model) for improving compliance? This volume presents
findings from the first phase of our research. Together these chapters present
evidence supporting the relevance of the Model to the taxation context, as well as
stories of success and frustration as the ATO undertook the systematic process of
implementation. Experiences in developing and using the ATO Compliance
Model have been shared with a number of other tax authorities including those
in Britain, New Zealand, Canada, Bulgaria, Thailand, and East Timor and
through training programs for tax administrators in Commonwealth countries in
Asia, Africa and the South Pacific. This volume provides an opportunity to share
the ideas behind the Model and our current knowledge of its usefulness more
broadly with all who are interested in tax administration in the global
community of the 21st century.
Part I (The Relationship between the Tax Office and the Community) of this
volume examines a set of issues that will be most familiar to readers as arguable
causes of non-compliance. In the present context, however, their importance
stems from their centrality in establishing a cooperative relationship between a
tax authority and the community.
The first question addressed in Chapter 2 is whether or not the concept of
motivational postures has relevance in depicting the quality of the relationship of
a tax authority with the community. Having established benchmarks for each
posture in the community, attention is turned to self-reported compliance. Are
individual taxpayers as compliant as their motivational postures suggest? The
answer is a resounding no. Non-compliant actions are found among those
committed to the system, those who have capitulated to tax authority, those who
resist it, and those who have disengaged from it. The readiness of non-compliers
to cooperate is varied, as postulated in the ATO Compliance Model.
Chapter 3 addresses one of the most important issues in any relationship
between a democratically elected government and its people, perceptions of
justice. Michael Wenzel provides a framework for justice research in the taxation
context through drawing a distinction between different kinds of justice
(distributive, procedural, retributive) and pointing out that justice takes on quite
different meanings at different levels of analysis. Justice can be adjudicated at an
individual level (am I being treated fairly?) or at a group level (is my group (e.g.,
taxi drivers) being treated fairly?) or at a societal level (is our tax system fair for
all?). Knowing what kind of justice is under consideration is not always clear in
tax research that considers justice as a cause of non-compliance. It is particularly
important for future research to learn whether or not the kinds of justice that
shape non-compliance are different from the kinds of justice that enable tax
authorities to manage compliance and claim legitimacy within the democracy.
The question of tax office legitimacy is of central concern in Natalie Taylor's
chapter on social identity (Chapter 4). Taylor takes up Michael Wenzel's
distinction between whether one thinks of oneself as a member of a group within
society or whether one identifies with a more inclusive group such as Australian
taxpayer. Using the written responses of 155 Australians to an open-ended
question about the tax system, Taylor demonstrates that it is the superordinate
inclusive identities that are associated with the granting of legitimacy to the tax
office and the readiness to cooperate with the tax system. Taylor concludes by
issuing a challenge to tax regimes that continue to adhere to a narrow
individualistic conception of self-interest as the fundamental motivation of
taxpayers. What such authorities may be giving away, according to Taylor, is the
key to their legitimacy.
The cash economy - the compliance problem that gave rise to the ATO
Compliance Model - has always been hidden from view, its size being estimated
by economists through a variety of indirect methods. In Chapter 5, Friedrich
Schneider 'sizes' Australia's cash (shadow) economy in comparison to other
countries. Then follows a microanalysis of individual taxpayers who answered
survey questions on shadow economy participation in 2000 and 2002. Using these
data, Valerie Braithwaite, Friedrich Schneider, Monika Reinhart, and Kristina
Murphy examine the importance of deterrence, justice, identity, and motivational
postures in determining who moves into the shadow economy, who moves out,
who stays involved, and who remains apart from shadow economy activities.
Part II examines the ATO Compliance Model as Change Agent. The section
begins with the story of how the Compliance Model was received by operative
staff in the Australian Taxation Office (Chapter 6). Jenny Job and David Honaker
present a warts and all account of the early stages of implementation of the
Model based on interviews conducted by the senior author. While there were
enthusiasts, there were also resisters. From both camps there were some who
correctly foresaw how far reaching the changes could be for the tax office and for
the way it conducted its business in the future. Along with insightfulness, were
feelings of threat and loss. Some complained of being pushed into something that
was untried and untested, and there were misunderstandings and myths, not
uncommon when innovation is in the air.
While the views from below reflected both excitement and cynicism about the
prospects of implementing the ATO Compliance Model, those higher up in the
organisation showed no reluctance in owning the Model. Kersty Hobson uses
transcripts of interviews with champions of the ATO Compliance Model to
analyse the ways in which leaders understood the model and presented it to staff
(Chapter 7). Hobson draws an interesting distinction in terms of how the ATO
Compliance Model was taken on board by senior staff. Some, in the words of one
of Hobson's interviewees, worked 'inside the model', while others stood outside,
trying to determine where they and their group should be located within the
framework. For the first group, the model was a dynamic tool to be played with
and pushed to its limits in analysing and managing compliance. For the second, it
was a static entity, that was to be used at worst, as another rulebook, at best a
cookbook.
In Chapter 8, the action of the ATO Compliance Model moves out of the office
and into the field. Neal Shover collected data from ATO field officers in the Cash
Economy Building and Construction Project and from owners of small building
and construction firms. These data suggest some progress in the direction of
effective implementation and the building of better relationships with taxpayers.
Shover warns, however, that it would be premature to claim success for the
model at this stage, and recognises some of the real world problems that highlight
the need for continuing evaluation. Organisational capacity for reform is
fundamental to the introduction of responsive regulation. The ATO provided
resources in the early stages, but the organisation was forced to redirect much of
its attention to the introduction of a goods and services tax in July 2000. As
Shover explains, the ATO lacked 'a calm environment [that] lends itself to the
deliberate and self-reflective decision-making that can nurture and sustain
[change]'. Consequently, the opportunity to evaluate ATO Compliance Model
implementation in a rigorous and systematic way was lost. At the same time,
there was a failure in organisational capacity to make the changes in the
administrative and technical system to allow responsive regulation to operate
fully.
Taking the ATO Compliance Model out of cash economy and recommending
its use in other ATO business lines did not occur without considerable scepticism,
most notably from the Large Business and International line. In Chapter 9, John
Braithwaite provides a review of the relevance of the ATO Compliance Model to
large business, arguing that basic ideas translate across contexts, although there
may be need to package the model differently. Braithwaite's chapter underlines
the point that the principles of responsive regulation travel widely, but that
regulatory pyramids cannot and should not be treated as cookbooks. Each
compliance group needs to find its own strategies that suit the problem, the
context, and the available resources. And they need to consult widely with the
community to find these strategies. Braithwaite illustrates this point with the
proposal of a compliance-tax-rate-spiral for reducing the incentives for game
playing among the very large corporates. The idea is that when the large
corporates as a group reach a series of benchmarks in extra dollars collected in
tax, they be rewarded through a lowering of company tax rates. The idea of the
compliance-tax-rate-spiral is not something that Braithwaite envisages as
anything other than a point for debate at this stage: instead it 'signals the kind of
world that might one day be possible if only we can learn how to forge a more
meaningful business-community-government partnership toward a decent tax
system'.
The purpose of responsive regulation and the ATO Compliance Model is to
develop, in conjunction with the community, a sophisticated plan that can
effectively manage non-compliant taxpayers, while being supportive of those
who are compliant. Up to this point, we have assumed that we know what
compliance is, and if there is doubt, that the tax office has the authority to clarify
things for us. In Part III (Beyond the Compliance Model), the limitations of this
worldview are exposed.
In Chapter 10, John Braithwaite, Yvonne Pittelkow and Rob Williams focus on
the difficulties of detecting suspected non-compliance in the tax affairs of
wealthy individuals in Australia. These authors provide a series of statistical
analyses of risk data from 235 high wealth individuals to show how important
the expert analyst's hunch is in deciding where the greatest risks to tax revenue
lie: When aggressive tax avoidance is obvious, it has long past its use-by date for
those with the money to pay for state-of-the-art financial advice.
The game of tax avoidance is addressed by Doreen McBarnet in Chapter 11.
McBarnet points out that compliance or non-compliance is a binary classification
that falls apart once tax avoidance enters the scene. Tax avoidance creates what
McBarnet calls creative compliance whereby taxpayers adhere to the strict letter
of the law, but find loopholes and caveats to minimise their tax without regard
for the spirit of the law. The ATO Compliance Model, as it stands, offers little
help to tax officers dealing with creative compliance because if no law has been
broken, there can be no top to the regulatory pyramid to encourage cooperation
at lower levels. In such circumstances, management of the human system cannot
meaningfully take place in the absence of revision of the legal system. McBarnet
describes the practice of introducing principles that span the law in a bid to
ensure that the intention of the legislature provides a backstop for legal
interpretation of law. McBarnet, however, is pessimistic that the cat and mouse
game of creative compliance can be changed by law itself. The change that is
required is more fundamental and attitudinal: Law must be seen as something to
be 'respected' rather than 'material to be worked on' to one's advantage.
The industry of tax avoidance rests on the talents of financial advisors. In
Chapter 12, John Braithwaite reports findings based on interviews with 27
advisors whose clients include the wealthiest people in Australia. Advisors were
invited to comment on the performance of the ATO's High Wealth Individuals
Taskforce that was set up in 1996, and were also drawn into discussions of better
ways of improving compliance among high wealth individuals and identifying
deficiencies in the law. The interviews themselves were evidence of the ATO
Compliance Model at work, a willingness by all parties to engage in dialogue and
share information, allowing persuasion to work in both directions. At the same
time, the interviews yielded valuable policy insights about where enforcement
capacity should be focused and ways in which the ATO Compliance Model can
be misleading if the object of attention is solely the taxpayer. Braithwaite argues
that, in the case of high wealth individuals, the dangers of 'enforcement
swamping' are so great that the kind of 'simplistic purism' that focuses on the
taxpayer as the object of enforcement has to be abandoned. Instead, the High
Wealth Individual Taskforce needs to continue along its current path 'of targeting
nodes of control over decisions of major import for tax compliance', be they
wealthy individuals and the suite of entities they control, tax managers of large
corporations, influential advisors, or promoters of aggressive tax avoidance
schemes.
In the final chapter, questions of compliance and what it means to comply are
embedded within a model that extends across the human, administrative and
legal systems that comprise a tax authority. Local compliance solutions do not
always sit comfortably alongside other local solutions. The principle introduced
to reconcile tensions within the compliance plan of a tax office is integrity
defined as unity and soundness of purpose. The argument presented in Chapter
13 is that effective compliance management and institutional integrity are
interconnected. Both can be optimised when a tax authority understands and
works with the community, bringing interested parties into discussions about the
tax system, how it should be designed, and what purpose it should serve. The
responsiveness to the outside environment, however, must be matched by
responsiveness internally. Tax offices need to have the organisational capacity to
change their administrative and legal structures, and to ensure that information
flows freely up and down the organisation. Through the quality of
responsiveness, internally and externally, tax administrations in the 21st century
may lose certainty, but gain an assurance that they greatly need: The
acknowledgment of the community that they are indeed taxing in the interests of
the democracy as a whole.
Taxing Democracy brings together the contributions of researchers from three
continents, ali of whom have spent time together in the Centre for Tax System
Integrity over its first three years, as full-time research staff, visiting fellows, or
research affiliates. Our thanks to the Australian Taxation Office and the Research
School of Social Sciences at the Australian National University for funding the
Centre and its research. Without their support and their cooperation, this
initiative would not have been possible.
Like all enterprises of this kind, there are people, whose names do not appear
in the pages that follow, who make it all possible. For her generosity of spirit and
infinite wisdom, my thanks to our administrator, Linda Gosnell, who keeps the
Centre running smoothly, and seemingly, effortlessly. To Sophie Cartwright who
meticulously and patiently prepared this manuscript for publication, thank you
from us all. And to the postgraduate students, university colleagues and
colleagues from the Australian Taxation Office who have provided critical
comments, have participated in conferences and seminars with us, and who have
opened doors so that this research could be conducted, our most sincere thanks.
Among this very large support team is Andrew Stout who has been our master
engineer, building bridges between the academic and bureaucratic worlds and
generating a continuing dialogue between the Centre and the Australian Taxation
Office. We are indebted to him for contributing in such an important way to the
richness of the research that we are able to share in this volume.
The name of one visiting fellow that sadly does not appear in this volume is
that of our late colleague, Leslie Whittington. Leslie was a tax economist from
Georgetown University, who was to spend study leave at the Centre, with her
husband, Charles, and their young daughters, Zoe and Dana. Leslie and her
family started their journey on September 11, 2001. All were on board the plane
that was hijacked by terrorists and flown into the Pentagon. We dedicate this
volume to her memory.
Note
1 The Cash Economy Task Force comprised 13 members spanning non-government (industry,
accountancy and commerce, welfare, and university) and government sectors. The Task Force was
chaired by Mr. David Butler, now Commissioner of Inland Revenue New Zealand, and Mr. Neil
Mann, now Deputy Commissioner, Australian Taxation Office.
References
Theoretical Background
Motivational postures have been used in past research to capture the way
regulatees position themselves in relation to regulatory authority (Braithwaite et
al., 1994; Braithwaite, 1995). Authorities may have legal legitimacy, but this does
not guarantee them psychological legitimacy. Individuals and groups evaluate
authorities in terms of what they stand for and how they perform. As evaluations
are made, revised, shared and accumulated over time, individuals and groups
develop positions in relation to the authority. A psychological concept that is
central to positioning is social distance (Bogardus, 1928). Bogardus used this term
to refer to the degree to which individuals (or groups) had positive feelings for
other ethnic groups and ascribed status to other ethnic groups. In the regulatory
context, social distance indicates liking and the ascription of status to the
regulatory authority. When individuals and groups decide how much they want
to associate or be aligned with an authority, and how much they want to be out
of reach of and out of contact with the authority, they are indicating the social
distance they wish to place between themselves and the authority.3
The distance placed between regulatee and regulator may be intuitive at first,
but it does not remain that way for long. Individuals and groups articulate their
beliefs, develop rationalisations for their feelings, and use values and ideologies
to justify the ways they position themselves in relation to legally sanctioned
authorities (Sykes and Matza, 1957; Rokeach, 1973; Thurman, St. John and Riggs,
1984; Griffin and Buehler, 1993; Bersoff, 1999). These interconnected sets of beliefs
and attitudes are shared, borrowed, challenged, and elaborated upon even further
as part of the social life of a community. The interconnected sets of beliefs and
attitudes that are consciously held and openly shared with others are called
motivational postures. Five motivational postures have been identified as
important in the context of taxation compliance, (a) commitment, (b)
capitulation, (c) resistance, (d) disengagement, and (e) game playing.
The two postures that reflect an overall positive orientation to authority are
commitment and capitulation. The kinds of beliefs and attitudes that comprise
these postures are represented in Table 2.1. Commitment reflects beliefs about the
desirability of tax systems and feelings of moral obligation to act in the interest of
the collective and pay one's tax with good will. Capitulation reflects acceptance
of the tax office as the legitimate authority and the feeling that the tax office is a
benign power as long as one acts properly and defers to its authority.
In contrast to these postures of deference, are three postures of defiance. The
first is the familiar posture of resistance. Resistance reflects doubts about the
intentions of the tax office to behave cooperatively and benignly towards those it
dominates and provides the rhetoric for calling on taxpayers to be watchful, to
fight for their rights, and to curb tax office power. Disengagement is also a
motivational posture that communicates resistance, but here the disenchantment
is more widespread, and individuals and groups have moved beyond seeing any
point in challenging the authorities. The tax office and the tax system are beyond
redemption for the disengaged citizen, the main objective being to keep both
socially distant and blocked from view.
The fifth posture is game playing. Unlike the previous postures, game playing
has not been examined in other regulatory contexts, emerging instead from
discussions about posturing with tax officials and taxpayers. The behaviours
previously have been described by social scientists working in fields involving
economic regulation (McBarnet, 1992; McBarnet and Whelan, 1999). McBarnet
(Chapter 11, this volume) sees game playing as a particular kind of attitude to
law: Law is seen as something to be moulded to suit one's purposes rather than as
something to be respected as defining the limits of acceptable activity. Game
playing was included as a motivational posture for the purposes of testing
whether or not players consciously adopted this style of engagement with the tax
system and the tax office.
Between June and December, 2000, a national survey was conducted by the
Centre for Tax System Integrity at the Australian National University (for details
see Braithwaite, 2001; Braithwaite, Reinhart, Mearns and Graham, 2001). A
stratified random sample of 7754 persons was selected from the publicly available
electoral rolls. A lengthy questionnaire on tax matters was sent to each person
who had been randomly selected, together with a letter explaining the intent of
the study and a stamped addressed envelope for the return of the completed
questionnaire. Two reminder cards were sent at two to three week intervals.
After 5 weeks, a second questionnaire was posted to non-respondents, again
followed by two reminder cards. (Details of the methodology of the survey are
available in Mearns and Braithwaite, 2001.)
Completed returns of the survey were obtained from 29 per cent of the sample,
providing 2040 cases for further analysis. This response rate, while low in
absolute terms, compares favourably with rates reported for other tax surveys
(Pope, Fayle and Chen, 1993; Wallschutzky, 1996; Kirchler, 1999; Webley, Adams
and Elffers, 2002). Interestingly, the sample provided a relatively representative
cross-section of the population with regard to sex, ethnicity, education, age,
occupation, and marital status (see Mearns and Braithwaite, 2001). The biases
that were detected were an over-representation of those in scribing occupations
who would have been more comfortable with a detailed response-intense
questionnaire, and an under-representation of younger age groups (18 to 25 years)
who traditionally are difficult to recruit for self-completion surveys.
Commitment
Paying tax is the right thing to do.
Paying tax is a responsibility that should be willingly accepted by all Australians.
I feel a moral obligation to pay my tax.
Paying my tax ultimately advantages everyone.
1 think of tax paying as helping the government do worthwhile things.
Overall, 1 pay my tax with good will.
I resent paying tax. (reversed)
1 accept responsibility for paying my fair share of tax.
Capitulation
If you cooperate with the Tax Office, they are likely to be cooperative with you.
Even if the Tax Office finds that I am doing something wrong, they will respect
me in the long run as long as 1 admit my mistakes.
The Tax Office is encouraging to those who have difficulty meeting their
obligations through no fault of their own.
The tax system may not be perfect, but it works well enough for most of us.
No matter how cooperative or uncooperative the Tax Office is, the best policy is
to always be cooperative with them.
Resistance
If you don't cooperate with the Tax Office, they will get tough with you.
The Tax Office is more interested in catching you for doing the wrong thing,
than helping you do the right thing.
It's important not to let the Tax Office push you around.
It's impossible to satisfy the Tax Office completely.
Once the Tax Office has you branded as a non-compliant taxpayer, they will
never change their mind.
As a society, we need more people willing to take a stand against the Tax Office.
Disengagement
If 1 find out that 1 am not doing what the Tax Office wants, I'm not going to lose
any sleep over it.
I personally don't think that there is much the Tax Office can do to me to make
me pay tax if I don't want to.
I don't care if I am not doing the right thing by the Tax Office.
If the Tax Office gets tough with me, 1 will become uncooperative with them.
I don't really know what the Tax Office expects of me and I'm not about to ask.
Game playing
I enjoy spending time working out how changes in the tax system will affect me.
I enjoy talking to friends about loopholes in the tax system.
1 like the game of finding the grey area of tax law
I enjoy the challenge of minimising the tax 1 have to pay.
The Tax Office respects taxpayers who can give them a run for their money.
Are Motivational Postures Discernible Among Taxpayers and
Citizens?
Figure 2.1 Mean scores for motivational postures of commitment, capitulation, resistance,
disengagement, and game playing
Lodgment
In the Community Hopes, Fears and Actions Survey, two questions were asked
about lodgment of a 1998-99 tax return. Respondents were asked 'Should you
have filed an income tax return in 1998-99' and immediately afterward, 'Did you
file an income tax return for 1998-99?' 81.8 per cent of respondents said that they
should have lodged a return or that they did not know if they should have lodged
a return for the 98-99 financial year.4 Of this group, 4.7 per cent had not yet
lodged their return. When those who did not need to lodge were included with
the compliant group, the per cent non-compliant on lodgment in the sample was
3.8 per cent. In other words, 3.8 per cent of the sample acknowledged having the
capacity to defy the ATO's request to lodge, and did so.
In addition, respondents were asked: 'Have you any income tax returns not yet
completed from previous years?'. Of the total sample, 4.5 per cent said that they
were in this situation.
Not having lodged a 98-99 income tax return that respondents said should have
been lodged and not having lodged returns for earlier financial years were
actions that were significantly and notably correlated (r = .37, p < .001). Those
who had not lodged a 98-99 return were also likely to have not lodged previous
returns. Using these data, a non-compliance index called non-lodgment was
constructed. The index was calculated by adding together compliance scores for
the two time periods. A score of 2, meaning that neither the 98-99 return nor all
earlier returns had been filed, was obtained by 1.6 per cent of the sample. A score
of 1, meaning that either the 98-99 return or an earlier return had not been
lodged, was obtained by 5.1 per cent of the sample. A score of 0, meaning that all
tax returns that should have been lodged had been lodged, characterised 93.3 per
cent of the sample.
Non-payment of Tax Debt
One question was used to assess having a tax debt: 'Do you have an outstanding
debt with the Tax Office?'. Those who said they had an outstanding debt
constituted 3.4 per cent of the sample. It should be noted that those with a debt
are not necessarily behaving this way without tax office permission: They may
have arranged a payment plan with the ATO. It is common knowledge, however,
that the law requires taxpayers to pay the money they owe on time, and
therefore, carrying an outstanding debt remains an example of not behaving in
accordance with the law, even if one is granted an extension of time to pay.
Respondents were presented with income from a variety of sources and were
asked if they 'did not declare it', 'declared some', 'declared most', or 'declared all'
in their 98-99 return. They were also given the option of indicating that they
received nothing from this income source (scored the same as 'declared all' for the
purposes of analysis). The sources of income were: (a) salary, wages; (b)
honorariums, allowances, tips, bonuses, director's fees; (c) eligible termination
payments; (d) Australian government allowances like Youth Allowance, Austudy,
Newstart; (e) Australian government pensions, superannuation pensions, and
other pensions or annuities; (f) interest; and (g) dividends. The percentage of the
sample failing to declare each type of income is given in Table 2.3. Because of the
small percentages in the failure to declare 'some', 'most' and 'all' categories,
responses were combined into one non-compliant category (see far right column
in Table 2.3).
Table 2.3 Percentage of sample not declaring all income: 'Think about each of the sources of income
listed below and select the response that best describes your 1998-99 income tax return'
Income source
did not declared declared failure to declare
declare it some most (total)
salary, wages .4 .7 2.1 3.2
honorariums, tips,
1.9 1.3 1.3 4.5
allowances, bonuses
eligible termination
.5 .7 .5 1.7
payments
Australian government
.5 .8 .4 1.7
allowances
Australian government
1.4 .8 .5 2.7
pensions
interest 2.0 1.3 1.8 5.1
dividends .9 .7 .6 2.2
Respondents were asked about whether, in the last 12 months, they had been a
provider of services in the cash economy or a purchaser of such services: (a)
'Have you worked for cash-in-hand payments in the last 12 months? By cash-in-
hand we mean cash money that tax is not paid on', and (b) 'Have you paid
anyone cash-in-hand payments in the last 12 months for work or services they
provided to you? By cash-in-hand we mean cash money that tax is not paid on'.
In response to the first question, 6.1 per cent said that they had worked for
cash-in-hand payments, and 14.5 per cent said that they had paid for cash-in-
hand services. When responses were combined for providers and purchasers, 19.1
per cent of respondents were participating in the cash economy, with 1.5 per cent
being both a purchaser and provider.
Over-claiming Deductions
Two questions were asked regarding deductions claimed on the 98-99 income tax
return. The first was: 'As far as you know, did you exaggerate the amount of
deductions or rebates in your 1998-99 income tax return?'. The majority of
respondents (89.8%) answered 'not at all', but 7.1 per cent admitted to
exaggerating 'a little', 2.1 per cent 'somewhat', .3 per cent 'quite a lot', and .7 per
cent 'a lot'.
The second question on over-claiming was: 'Think of the deductions and
rebates you claimed in your 1998-99 income tax return. Would you say you were
(a) absolutely confident that they were all legitimate, (b) a bit unsure about some
of them, (c) pretty unsure about quite a lot, or (d) haven't a clue, someone else did
it'. Most of the sample claimed to be absolutely confident about the legitimacy of
the claims (84.8%), 7.9 per cent were a bit unsure about some, .7 per cent were
unsure about a lot, and 6.6 per cent did not have a clue because someone else had
completed the tax return for them. For this latter group, signing an income tax
return appears to be a ritualised activity with a third party being assigned
responsibility for its accuracy.
These data show that 10 per cent are willing to admit to some over-claiming
and 15 per cent are prepared to express some uncertainty about whether their
claims for deductions and rebates are all legitimate. Responses to these two
questions were used to form an over-claiming deductions index. Before this could
be done, the second question was dichotomised in terms of whether respondents
were absolutely confident (84.8%) or not (15.2%). Responses to the two questions
were then correlated (r = .34, p < .001), showing that the more one exaggerates
deductions, the less confident one admits to being about the correctness of the
claim. Responses to these two items were transformed into standardised scores
(mean of 0, standard deviation of 1) and then summed to form an over-claiming
deductions index (alpha reliability coefficient = .51). The percentage expressing
doubts or admitting to over-claiming deductions was 19.8 per cent of the sample.
Respondents were presented with a list of 8 strategies that are known to provide
for tax minimisation. Respondents were asked if they were able to minimise their
tax through these strategies in the 1998-99 financial year. Respondents were also
asked to circle a special category if they did not know what the particular
strategy was. The findings are reported in Table 2.4 under two headings. First, the
percentage of the sample using the strategy is recorded. The second column
represents the percentage of the sample with no knowledge of the strategy as a
method of minimising tax. Of particular note is that although tax minimisation is
a popular topic for media attention, knowledge about the specific methods of tax
minimisation have not penetrated into the community as extensively as had been
assumed when the survey was conducted. This apparent lack of understanding of
specific methods of tax minimisation among some segments of the population
may be comforting to a tax administration that feels in control of its taxpayers
and able to shield its constituency from temptation. However, in a world where
mass marketed schemes are aggressively promoted and individual taxpayers need
to be alert to the dangers, a poor understanding of tax avoidance measures and
their consequences leads to vulnerability in the system (Commonwealth
Ombudsman, 1999; Murphy, 2002).
The strategies listed in Table 2.4 differ enormously in the degree to which they
meet the expectations of the tax office, or to put it another way, the degree to
which they risk being defined as avoidance measures by the tax office. Paying
into superannuation schemes to minimise tax is explicitly encouraged, for
example, whereas using off-shore tax havens is explicitly discouraged. There is no
way in which Part IVA of the Income Tax Assessment Act 1936 can be
operationalised in relation to survey responses to divide strategies into those that
are legal and those that are illegal. It is possible, however, to divide the strategies
in terms of the degree to which they push the limits of legality.5 With this in
mind, the list of 8 was divided into those that are on the more cautious side
(negative gearing, employee share arrangements, salary packaging,
superannuation planning, and warrants or leveraged investments) and those that
are on the more aggressive side (schemes, tax shelters, and off-shore tax havens).
Table 2.4 Percentage of sample using tax minimisation strategies and lacking knowledge of these
strategies
Evidence to suggest that people who used one strategy were more likely to use
others was not strong. The correlations among the 8 strategies were all positive,
but they were also relatively low ranging from .02 to .34 (median = .09).6 It seems
most likely that having found one strategy, most individuals really did not need
to look for another. As a result, a tax minimisation index was not formed from
these measures, but it was still possible to count the number of strategies that
were being used by each respondent, of either a cautious or aggressive kind. A
count on the use of the five cautious strategies revealed that 69.1 per cent were
using none, 22.6 per cent were using one, 7.1 per cent were using two, 1.1 per cent
three, .1 per cent four and .1 per cent five. On the three more aggressive
strategies, 96.9 per cent were using none, 2.7 per cent one, .3 per cent two, and .1
per cent three.
From a legal perspective, purpose or intent is critically important for
ascertaining the acceptability of a tax minimisation scheme to the ATO (Part IVA
of the Income Tax Assessment Act 1936). In the context of a general population
survey in which individuals are asked to self-report on their activities, intent was
defined in terms of how much effort the taxpayer dedicated to finding ways to
minimise tax. A scale to measure effort to minimise tax was constructed from
two items. The first question was: 'Some people put in a lot of effort to plan their
financial affairs in order to legally pay as little tax as possible. How much effort
did you or your family devote to this objective in preparing for your 1998-99
income tax return?'. In response, 6.7 per cent circled 'a lot', 8 per cent 'quite a bit',
17.2 per cent 'some', 21.2 per cent 'a little', and 46.9 per cent 'none'. The second
question asked respondents: 'In preparing for your 1998-99 income tax return, did
you look at several ways of arranging your finances to minimise your tax?'.
Respondents replied 'yes' (22.2%) or 'no' (77.8%).
These two items correlated positively with each other (r = .49, p < .001),
showing that those who put a lot of effort into legal tax minimisation were also
likely to have looked at several different ways of arranging their finances to
minimise tax. When the responses to these questions were considered conjointly,
45.2 per cent of the sample reported having put no effort into minimising their
tax. Scores on the two items were standardised and were summed to produce the
tax minimisation effort index (alpha reliability coefficient = .66).
Table 2.5 Pearson product-moment correlation coefficients among indicators of tax non-compliance
and minimisation
Table 2.6 Point-biserial correlation coefficients between motivational postures and compliance related
actions
Motivational postures Evasion related actions Avoidance related actions
Figure 2.3 Motivational posture mean scores for avoiders and non-avoiders
Conclusion
This chapter theoretically and empirically differentiates the compliance related
actions of taxpayers from their evaluation of the tax system and the tax office,
expressed in terms of motivational postures. Motivational postures reflect the
social distance that individuals wish to place between themselves and the tax
authority. Increasing social distance indicates increasing dislike for the authority
and a lowering of the status ascribed to that authority. It does not, necessarily,
signal disobedience. In some ways, motivational postures can be thought of as an
indicator of the degree to which an individual is giving consent to the authority:
Consent to consider that individual as a participant in the tax system and consent
to being regulated by the authority. Giving consent is a different phenomenon
from obeying a request from a legally designated authority.
When framed in terms of consent and compliance related action, the two-
dimensional conception of responsiveness offered in this chapter is less puzzling.
Non-compliant actions may be initiated for any number of reasons, only some of
which are attitudinal. Once non-compliance has occurred, it requires a response
by the authority. All too often, authorities make the assumption of consistency
between attitude and behaviour: People who do the wrong thing are bound to be
nasty pieces of work, and need to be treated like the villains they are (Braithwaite
et al., 1994). This is not always the case, as this chapter demonstrates. The
important question then is does this simplistic formulation of 'bad guys do bad
things' cause harm? From the perspective of an authority, part of dealing with an
individual's non-compliance is to ensure that it will not happen again, and part is
to show the community that compliance standards are high and will be
maintained. Neither of these goals is served well through management strategies
that provoke non-compliant individuals into revoking their consent to be a
participant in the system. No-one is going to like being sanctioned for non-
compliance, but few benefit when discontent of this kind is fuelled by
disrespectful treatment from the authorities, leaving individuals with a life long
passion for resistance and defiance. Furthermore, community confidence in
standards is unlikely to be boosted when the story of detected non-compliance is
trumped by a horror story of unfair treatment at the hands of the authority.
When individuals withdraw their consent to being part of the tax system, the
legitimacy of the system itself is vulnerable.
Thus, non-compliant actions on the part of taxpayers must be met by a
responsiveness from authority that recognises and deals with the wrongful act,
but at the same time works to bring the more cooperative motivational postures
to the fore (Braithwaite, 2002). Resentment and anger may be present, but the
findings presented in this chapter suggest that there is also likely to be goodwill
and acceptance of the rules of the game, if they can be brought into the
discussion and the resolution of the problem. The challenge for tax administrators
is to play a two-handed game: To deal with the wrongdoing today, while
nurturing consent for tomorrow.
Notes
1 These two aspects of community responsiveness were recognised by Schmolders (1970) in his early
work on tax systems in different cultures.
2 The poor relationship between attitude and behaviour has been widely recognised and has given rise
to research that has tried to build theory that provides a better fit between attitudes and behaviour in
line with consistency theorists' general expectations (see, for example, Lewis, 1982; Hessing, Elffers
and Weigel, 1988; McGraw and Scholz, 1991; Scholz, McGraw and Steenbergen, 1992; Taylor, Chapter
4, this volume). This chapter, however, pursues a different direction.
3 Black (1976) has used the term, relational distance, to describe social distance from the regulator's
perspective. In the present context, social distance is used to represent the regulatee's perspective.
4 Of the respondents, 32 (1.6%) said that they did not know if they should have filed a return. Given that
citizens have a responsibility to find out if they are exempt, this group was included with those who
agreed that they should have filed a return.
5 Tax researchers use the term, avoision, to capture the problem associated with differentiating legal and
illegal tax effective schemes (Seldon, 1979; Sawyer, 1996; see also James and Alley, 1999).
6 For the correlational analyses, respondents who had never heard of the strategy were included with
those who had not used the strategy.
7 In order to exhaust all possibilities, the motivational postures were correlated with the 9 specific
compliance related measures. The results did not change substantively: Significant correlations with
specific compliance-related actions were reflected in the amalgamated action measures presented in
Table 2.6.
References
'In this world nothing can be said to be certain, except death and taxes.'
Traditional economic models of tax evasion (see Allingham and Sandmo, 1972)
regard tax compliance as an outcome-maximising decision between the
alternatives of: (a) truthfully paying tax which results in a certain loss, and (b)
evading tax which results, with some uncertainty, in either a reduced loss (in
terms of taxes not paid) or an even greater loss (due to the fines imposed if the
evasion is detected and penalised). These early models, however, neglect the fact
that taxpayers also have their share in government services and public goods that
are funded by the tax revenue. Taxpayers receive some gains in exchange for the
taxes they pay, and not only may they consider these outcomes in their rational
equation, but they may also evaluate whether the exchange is an equitable one or
not (Cowell, 1992; Falkinger, 1995).
Further, the issue of public goods in return for taxes adds a social dimension to
the problem, because the amount of revenue available, and the quantity and
quality of public goods provided, is not solely dependent on the single taxpayer's
choice to pay or evade taxes. Rather, taxpayers are interdependent, as their
outcomes in terms of the public goods they share are a function of their combined
behavioural choices. The issue of tax compliance can therefore be considered a
social dilemma (see Dawes, 1980). Individual taxpayers may choose to evade tax
in order to maximise their personal outcomes and still enjoy their share of the
public good, which is not affected by single defective choices. However, if many
taxpayers chose to do so, revenue would fall to a level where certain public goods
would be no longer affordable and everyone's outcomes would be reduced
(Weigel, Hessing and Elffers, 1987; Elffers, 2000). Models of tax compliance would
have to incorporate taxpayers' awareness of their mutual interdependence in
their rational decision of how to maximise individual outcomes. Furthermore, as
research on social dilemmas informs us (see Pepitone, 1971; Wit, Wilke and
Oppewal, 1992; van Dijk and Wilke, 1993), taxpayers may also evaluate what
would be fair for them to contribute to the public good, considering their relative
resources and relative share of the public good as well as others' level of evasion
(Bordignon, 1993).
Social psychological research also informs us that social dilemmas assume a
quality of greater competitiveness and less cooperativeness when the involved
parties define themselves as members of different groups rather than as
individuals (see Brewer and Schneider, 1990; Schopler and Insko, 1992). A crucial
factor contributing to this effect may be the perceived sharedness of and
consensus on distinctive group interests that give social-normative support for
one's tendency to act in the interests of one's group (Zander, 1971; Haslam, 2001).
Importantly, once one identifies with a group, norms and perceived consensus
within the group should make members act in the interests of their own group,
even if they personally would not profit from their own behaviour. Most models
of tax evasion assume that taxpayers are motivated to maximise their individual
outcomes; they do not consider the possibility that taxpayers define themselves as
members of social groups and act in terms of the interests and norms of their
group and fellow group members (Sigala, Burgoyne and Webley, 1999). Again, it
also holds for this level of analysis that group members are not only concerned
about maximising their group's outcomes but also about their group receiving the
outcomes it is perceived to be entitled to, according to ingroup norms of justice
and fairness. Research on relative deprivation has repeatedly shown that group
deprivation is a stronger predictor for social protest and resistance than feelings
of personal deprivation (see Dubé and Guimond, 1986; Walker and Mann, 1987;
Hafer and Olson, 1993). The possibility that tax evasion is an act of social protest
against a tax system perceived to be unfair to their ingroup needs to be
considered.
Finally, research shows that persons or groups involved in a social dilemma are
more cooperative and show more concern for collective outcomes when they
identify themselves in terms of the same inclusive group (Brewer and Schneider,
1990); for instance, the society including all individuals and all different groups of
taxpayers. Defining themselves more inclusively, their selves and correspondingly
their self-interests become so transformed that they include the interests of others
and of the collective as a whole (Brewer, 1991; Morrison, 1997). A concern for the
interests, goals and values of the inclusive category has rarely been taken into
account in research and models on tax evasion (see Taylor, Chapter 4, this volume
for a departure from general practice). Again, based on a collective self-
identification, taxpayers might feel committed to maximising their collective
outcomes as well as committed to socially shared and normative representations
of their collective identity that prescribe certain distributions of burdens and
goods as appropriate and just. A concern for a fair society rather than for one's
personal or group interests would then motivate taxpayers. Tax evasion would
depend on taxpayers' (socially mediated) perceptions of whether the current tax
practice is either conducive to maximising the collective's welfare or consistent
with a representation of how society should look.
To sum up, research on tax evasion has been dominated so far by
individualistic approaches that focus exclusively on the motivation to maximise
personal material outcomes. The analysis needs to be extended to also include the
possibility of taxpayers defining themselves in more inclusive ways, either as
members of different social groups or, most inclusively, as members of the society
as a whole, implying a concern for outcomes of their ingroup or the welfare of
their whole nation, respectively. Furthermore, models of tax evasion need to take
into account that taxpayers may not only want to maximise their interests,
however defined, but also desire to see justice and fairness realised (see Kinsey,
Grasmick and Smith, 1991). In the remainder of this chapter, I will look more
closely at different justice considerations that may play a role in tax compliance
and differentiate between an individual, group and inclusive (societal) level of
analysis.
Empirical research on tax compliance often refers to issues of justice and fairness
in a rather undifferentiated and/or selective manner. Given the dominance of the
rational actor approach, authors interested in issues of justice and fairness seem
to have found it their main challenge to demonstrate the general importance of
these perceptions for tax compliance. Justice perceptions have often been either
operationalised through global measures, or selected aspects (e.g., one's relative
tax burden) have been treated as representative indicators of generic concepts
such as 'fiscal fairness' (see De Juan et al., 1994). Other research has looked at
single justice issues in more depth, but in isolation. Although some authors have
demanded that attention be 'directed toward what forms of inequity are likely to
affect tax evasion behavior' (Spicer and Becker, 1980, p. 174) and 'that one has to
specify fairly carefully what one means by the inequity or injustice that is often
cited as a motive for evading taxes' (Cowell, 1992, p. 540), so far these questions
have not been systematically addressed.
The most common differentiation in research on tax compliance refers to the
concepts of exchange equity, vertical equity and horizontal equity (e.g., Kinsey
and Grasmick, 1993). Exchange equity concerns the perceived value of tax-funded
government benefits and services received relative to one's tax contribution.
Vertical equity concerns the burden of taxes for certain social strata relative to
other strata. Horizontal equity concerns the burden of taxes for members relative
to others within a given social stratum. Other studies have focussed on the
structure of tax rates (e.g., Roberts, Hite and Bradley, 1994) or on procedural
fairness in audits (e.g., Stalans and Lind, 1997). A few studies have tried to specify
empirically the dimensionality of the fairness concept. Based on factor analysis,
Gerbing (1988, cited in Roberts and Hite, 1994) found four dimensions of tax
fairness, namely general fairness and distribution of tax burden, exchange with
government, taxes of the wealthy and progressivity of tax rates. Christensen,
Weihrich and Newman (1994) also used factor analysis and found five factors:
personal payment level, exchange with government, tax rate structure, special
provisions and overall fairness.
However, outcomes of factor analyses depend on the measures fed into the
analyses; the dimensionality of a concept cannot be decided on purely empirical
grounds. We require a theoretical framework that guides the formulation of
measures and that aids us in systematically investigating the role of fairness for
taxpaying attitudes and behaviour. Such a framework should also help us
integrate diverse research findings, point to areas that have been neglected so far
and identify areas where the evidence is either inconsistent or conclusive. In the
following, I will suggest a taxonomy for this purpose and review the literature on
this basis.
Similar to the priority that social psychology has historically given to distributive
justice (Adams, 1965; Walster et al., 1973; Leventhal, 1976; Lerner, 1977), tax
compliance research has so far mainly focussed on distributive aspects of
taxation. It has done so on all three levels of analysis, as defined above, however
without differentiating them explicitly in most cases. Some research has used
measures that did not specify the level of analysis and so these cases cannot be
categorised unambiguously into the present schema (see Table 3.1).
Table 3.1 Distributive justice in taxation: examples for three levels of analysis
Table 3.2 Procedural justice in taxation: examples for three levels of analysis
Taxpayers can also evaluate the amount of voice or control they have in
decision processes; that is, the degree of consultation and representation in tax
matters. In their study on audits, Stalans and Lind (1997) found that taxpayers
who mentioned that auditors were unresponsive to their views and comments
thought auditors tried less hard to be fair. Aim et al.'s (1993) finding falls in the
same category of fairness effects. As mentioned earlier, participants who could
decide over the use of tax revenue by vote and majority rule, rather than the
spending purpose being imposed on them, were more compliant (see Tyler,
Rasinski and Spodick, 1985).
Another potential subject of fairness evaluations is the extent and quality of
information provided by the ATO. While access to and provision of information
may also impact on compliance costs (see below), they are first of all understood
here in their intrinsic value of providing transparency, justifications and
explanations for decisions (informational justice) (Greenberg, 1993). Magner,
Johnson, Sobery and Welker (2000) found that perceived attempts to justify a
revenue spending decision did not contribute uniquely, that is, beyond the effects
of other fairness criteria, to perceived procedural justice. However, Wartick's
(1994) research, referred to earlier, showed that the provision of explanations for a
tax law change can increase perceptions of fairness. However, the outcomes of
her two studies were inconsistent in that the justification effect occurred either
only for those less well off due to the change (study 1) or only for those
unaffected by the change (study 2). As Wartick argues, the results might reflect
the fact that the justification needs to be found adequate in order to be effective
(Greenberg, 1993).
Moreover, procedural justice was defined earlier as implying a concept of
entitlement. Hence, it could be argued that taxpayers would regard a certain
treatment as particularly fair when it satisfies their perceived entitlements. In a
study with a student sample in Australia, I asked respondents to rate their
perceptions of and response to a reminder letter that they hypothetically received
from the ATO (Wenzel, 2001c). The letter either contained respectful treatment
(interpersonal justice), provided explanations and justifications (informational
justice), or was a usual, rather concise, letter from the ATO (control).
Furthermore, each letter highlighted one of three rights from the Taxpayers'
Charter (Australian Taxation Office, 1997), namely, rights to respectful treatment,
explanation of decisions, or minimisation of costs (control). In line with the
prediction, the interpersonal justice letter was considered fairer than the other
two letters, when the interpersonal right (respect) was salient. The informational
justice letter was considered fairer than the other two letters when the
information right (explanation) was salient. Feelings of entitlement seem to play
a role for perceptions of procedural fairness. The results suggest that tax
authorities could indeed profit from alerting attention to and granting taxpayer
rights, but they would need to assure that the rights are indeed fulfilled.
Finally, individuals may also be concerned about the efficiency of their
interactions with the ATO or, conversely, the costs of their attempts to be
compliant. Possible issues could be the promptness of correspondence, provision
of assistance and clear instructions, waiting time on the phone and length of
queues at information desks (see Smith and Stalans, 1991). For instance, in Stalans
and Lind's (1997) study, respondents' satisfaction with their audit treatment was
affected by the perceived time the auditor used to gather information and make a
decision. Likewise, respondents were more dissatisfied, the longer it objectively
took to reach a decision. Wallschutzky (1984) found that convicted evaders were
less satisfied than the control group with the efficiency and speed of the ATO's
handling of their tax returns.
The same issues of procedural justice could be analysed at a group level, which
research has so far neglected. For instance, taxpayers could feel they are treated
disrespectfully as members of a certain group (rather than as individuals) (see
Hobson, 2002; Murphy, 2002a). They could believe that the ATO is not neutral
and treats their group differently from another group (see Hobson, 2002; Murphy,
2002b). Likewise, they could feel that the ATO allows voice, and listens, to certain
societal groups rather than others. They could think that the ATO goes to great
lengths to explain decisions to, and try to receive consent from, some groups but
not when it comes to their group. Or, they could believe that their dealings with
the ATO are less efficient and the costs of complying with the laws are higher for
their group than for others. A good example of this latter issue is the strong
dissatisfaction small business owners expressed after the introduction of Tax
Reform in Australia in 2000. With the introduction of the Goods and Services Tax
(GST), and the (initial) requirement to lodge quarterly tax statements, they felt
small business was carrying the greatest burden of the tax changes. They also felt
disadvantaged compared to larger business who were in a better position to make
one-off investments in structural changes to cope with the new system.
At a societal level, the same four issues and how they apply to the ATO's
treatment of all taxpayers could be of concern. For instance, the ATO's formal
granting of respective rights and service standards could be seen as a measure of
procedural fairness at a societal level. Based on a survey with Australian citizens,
Braithwaite and Reinhart (2000) found that respondents had greater trust in, and
more favourable attitudes towards, the ATO when they believed that the ATO
met their obligations set out in the Taxpayers' Charter. The Australian Taxpayers'
Charter not only grants fair and respectful treatment and confidentiality, but also
makes explicit taxpayers' options for making complaints and appealing decisions.
The latter would be an instance of voice and process control at a societal level.
Perceptions of large-scale voice and control about taxation issues, however, are
also based on views about the presence of participatory and democratic political
structures (as simulated in the study by Aim et al., 1993 or referred to in the
vignette study by Tyler et al., 1985). The third issue of informational justice is
played out at the societal level in terms of the ATO's transparency towards the
public, its presentation in the media and the extent of communication with
taxpayers.
Overall, these societal-level issues have rarely been investigated in their
potential impact on tax compliance. In contrast, the fourth issue, namely the
efficiency or costliness of the tax system, has attracted quite a lot of research
attention. This issue has been dealt with in particular under the labels of
administration and compliance costs (e.g., Sandford, Godwin and Hardwick,
1989) and the complexity of the tax system (e.g., Milliron, 1985). Both topics are
related (e.g., Blumenthal, 2000), but only the latter has been discussed explicitly in
relation to perceptions of fairness, with authors disagreeing about how, or
whether at all, complexity is related to fairness) (see Cames and Cuccia, 1996).
According to Cames and Cuccia (1996), complexity overall is negatively related to
perceived equity; however, taxpayers can regard specific complexities as justified
and thus these would contribute less to perceptions of unfairness. Yet, Smith
asked respondents to rate the probable effectiveness of eight possible 'ways that
might help the IRS [Internal Revenue Service] do a better job' (1992, p. 237).
Respondents most strongly recommended a simplification of the tax system; and
the more unfair they thought the current tax system was, the more they
recommended simplification. It may be added here that respondents' second
strongest recommendation was to increase the likelihood of tax offenders being
caught. This leads us to the third area ofjustice.
Table 3.3 Retributive justice in taxation: examples for three levels of analysis
Taxpayers could question the fairness of the treatment that they (or another
target individual) receive as a response to a suspected or established act of tax
evasion. They could find a penalty too severe (or not severe enough) relative to
the offence, given the degree of blame attributed to oneself (or the other person)
or compared to other cases. They could find the audit process too rigid, ignorant
of possible harm to the person or their business, or a mere harassment. At a
group level, the same issues could be considered for the treatment of one's own
group relative to other groups. At a societal level, some possible issues could be
the perceived fairness of the general severity of penalties for tax offences, the
relative severity of penalties for small versus serious offences (i.e., a version of
vertical justice), or the perceived quality of the punishments such as their
punitive versus reintegrative character (see Braithwaite, 1989).
As stated above, there is not much research available on these issues. Vogel
(1974) asked participants to select appropriate penalties (no penalty, fine,
imprisonment) for acts of tax evasion differing in seriousness. However, he
compared the responses with the penalties suggested for other offences, but not
with the perceived actual penalties for tax evasion. Thus, the questions were used
as an indirect measure of people's attitudes towards tax evasion as an offence
compared to other offences (see Song and Yarbrough, 1978), but not people's
perceived fairness of the current system of penalties for tax evasion. There is,
however, much anecdotal evidence for taxpayers' resentment over incidences
where individuals defied the ATO, evaded or avoided tax, and yet escaped
prosecution. In fact, the resentment is often better understood as an intergroup
phenomenon in that certain groups of people (e.g., the rich, big business) are
considered to be able to dodge their taxpaying responsibilities and the ATO is
perceived as being soft on 'them' (see Shover, Job and Carroll, Chapter 8, this
volume; Braithwaite, Chapter 12, this volume).
Likewise, researchers have argued that taxpayers subject to investigation or
punishment could resent their treatment, find penalties unfair or audits
unreasonably intrusive, and as a consequence develop more negative attitudes
towards the ATO (Strumpel, 1969; Spicer and Lundstedt, 1976; Sheflrin and Triest,
1992; Murphy, 2002a, 2002b). At a group level, taxpayers could also question the
consistency of the ATO's enforcement procedures and penalty regimes across
different societal groups. For instance, in the 1990s, tax miminisation schemes
gained popularity, and were strongly promoted, among groups of middle and
working-class Australians (see Murphy and Byng, 2002). While the ATO tolerated
the practices for several years, in early 1998 it decided to crack down on the
schemes, declared them illegal and amended previous tax assessments. Taxpayers
involved in the schemes faced large tax bills, penalties and interest charges
(Murphy, 2002a, 2002b). There was not merely a perceived inconsistency over
time; more crucially people resented the decision because they thought it
reflected discriminatory treatment of them, the normal middle or working-class
people, versus the rich who always go unpunished. As one tax scheme promoter
put it: 'You see, it was OK while it was the top end of town, but when it was the
mums and dads of Australia starting to take advantage of the same tax breaks,
then, of course, it was time to call a halt' (Australian Broadcasting Commission,
Four Corners, 2001) (see also Senate Economics References Committee, 2001;
Hobson, 2002; Murphy, 2002a, 2002b, for work on perceptions of ATO fairness by
scheme investors).
Issues of retributive justice may also play a role in the success or failure of tax
amnesties (Hasseldine, 1998); however, tax amnesties have been analysed so far
mainly from an economic expected utility perspective (e.g., Aim and Beck, 1990;
Stella, 1991). Under these programs, previously non-compliant taxpayers are
asked to come forward and disclose their tax deficiencies, normally with the
incentive of penalties being waived before a certain deadline. The aims of such
measures are twofold: to get hold of revenue that would otherwise be lost
because detection of the evasion would be either unlikely or costly, and to bring
taxpayers who dropped out of the system back into the system. Taxpayers who
previously evaded tax but have changed their attitude would no longer feel
forced to repeat their previous behaviour out of fear of receiving high penalties.
They might therefore view the amnesty as a way of giving them a fair chance. In
contrast, taxpayers who are already honest may consider a tax amnesty unfair,
because it does not acknowledge their integrity, disadvantages them materially
and rewards tax evasion (Hasseldine, 1998). The perceived unfairness of tax
amnesties, in addition to their potential impact on expected utilities from evasion
behaviour, may undermine tax morality in the long run.
Conclusion
The evidence for the role of justice perceptions for tax compliance, as reviewed in
the previous sections, supports my introductory argument that a pure self-interest
account is insufficient for a proper understanding of taxpaying behaviour.
Furthermore, the review demonstrates the complexity of questions of fairness in
the area of taxation and supports my view that we need to differentiate more
precisely the aspects of justice we are talking about. The distinction between
distributive, procedural and retributive justice is an established one in
psychological research on justice (Tyler and Smith, 1998). As shown, all three
areas seem to be potentially relevant to the phenomenon of tax compliance.
Likewise, the distinction between different levels of analysis, even though often
overlooked, is considered valuable in psychological research on justice (Tyler and
Smith, 1998). fo some extent, this distinction was already implicit in the fairness
concepts applied to tax compliance (e.g., horizontal versus vertical justice), but an
explicit and systematic application of different levels of analysis proved possible
and instructive. In particular, an intermediate group level analysis has been
largely neglected in tax compliance research, whereas studies on justice and
relative deprivation have demonstrated the distinctive dynamics when people do
not consider themselves as individuals but rather as members of social groups
(e.g., Smith et al., 1994; Platow, O'Connell, Shave and Hanning, 1995; Wenzel,
2002). An analysis of tax non-compliance as a group level response that is based
on an interpretation of the situation shared within one's relevant ingroup, and
that is justified as a response against the perceived unfair treatment of one's
ingroup (relative to other groups), is more than plausible. Unfortunately, however,
there is limited research that speaks directly to such an account.
Indeed, the main value of the present review is less likely to consist in its
statement about what we know about the role of justice perceptions for tax
compliance, but rather what we do not know. For several issues, the review
revealed inconsistent findings that could not be resolved here, because a more
thorough analysis of the different methodologies and operationalisations would
be required. More importantly, however, the argument is that we cannot expect
all findings to be consistent as they refer to various forms of fairness perceptions.
The taxonomy suggests which results should be comparable and therefore which
inconsistencies need to be resolved. Moreover, the review showed that a number
of issues identified on the basis of the present taxonomy have rarely been
investigated at all. Specifically, questions of retributive justice at all three levels of
analysis have been largely ignored. Likewise, only a few studies have looked at
the role of procedural justice for tax compliance. Further, even in the area of
distributive justice, which has been the main focus of research, a group level
analysis has been clearly neglected so far. Future research needs to address these
issues.
Future research would also need to test for independent and unique effects of
the various justice considerations, as they are likely to be empirically correlated
with each other. Research could try to establish which of the more finely
differentiated justice considerations are most strongly related to, and most
predictive of, tax compliance (Wenzel, 2001d). The results could be valuable for
tax legislators and tax authorities, not only by providing a more differentiated
diagnosis of the underlying problem, but also by suggesting priorities for
addressing the issues.
However, contextual conditions should furthermore moderate the impact of the
various justice concerns on taxpaying behaviour. We need to apply and develop
more refined theoretical accounts for the relevance of justice concerns. Empirical
inconsistencies revealed in the present chapter, where consistency was expected,
may be partly due to methodological differences between the studies; however,
they also encourage us to look for moderating factors and gain a better
understanding of when and why people are concerned about fairness at all.
In a recent study on tax compliance (Wenzel, 2001b), I tested the prediction
that concerns for procedural and distributive justice would depend on the level of
identification with the inclusive category (i.e., one's nation) within which
procedures were applied and resources distributed (Tyler, 1997; Tyler and Smith,
1999; Wenzel, 2000, 2001a, 2002). Based on a survey with Australian citizens
(Braithwaite, 2001), the results confirmed the predictions for two forms of tax
compliance (i.e., under reporting of non-pay income, exaggerations of
deductions). Self-interest variables were more influential when respondents were
less identified as Australians, while perceptions of procedural and distributive
justice were more positively related to compliance when respondents identified
strongly as Australians. Two other forms of compliance were not related to
fairness concerns but to self-interest considerations. These two forms, namely
cash under reporting and tax minimisation, could be considered more legitimate
behaviours, at least for certain reference groups, and might therefore constitute
rational choices rather than a protest against perceived injustices. Thus, not only
could social identity influence whether or not people are concerned about justice,
but qualities of the specific behaviours could moderate whether or not these
would be used as a response to perceived unfairness.
Certainly, further research in this direction is necessary. The present chapter
was not meant to offer substantial theoretical propositions about underlying
psychological processes. Rather, it provides a map to the field, using a larger scale
and covering familiar and unfamiliar territory that will hopefully act as a tool for
future expeditions.
Notes
1 I will not distinguish here between entitlement and deservingness, but see Feather (1999) and Steil
(1994).
2 In fact, they could also compare their share in tax-funded benefits with their share at future times or
under counterfactual conditions; for instance, when the opposition party suggests a different use of
tax revenue (see Folger, 1986).
References
Understanding why taxpayers do or do not comply with tax regulation has been
the focus of much research over the past three decades. Quite naturally, it is
assumed that if we were able to acquire a good understanding of why taxpayers
do or do not comply, we would be in a much better position to obtain greater
levels of compliance with tax reporting, and bridge the tax-gap (Internal Revenue
Service, 1997) more effectively. However, despite the wealth of research
conducted on this question, a satisfactory explanation remains elusive. The task is
not helped by the fact that defining what constitutes compliance is in itself
difficult, with the subtleties surrounding evasion, avoidance, mere error, and
intention often making them indistinguishable (see Long and Swingen, 1991;
Roth, Scholz and Witte, 1989). Even so, our ability to pinpoint underlying
motivations for compliant or non-compliant behaviour, however defined, remains
weak.
Researchers from various disciplines have investigated the effects of variables
such as legal sanctions (Grasmick and Scott, 1982), stigmatisation (e.g., Porcano
and Price, 1993), probability of audit (Webley, 1987), conscience appeals (Schwartz
and Orleans, 1967), self-interest (McGraw and Scholz, 1991), opportunity (Klepper
and Nagin, 1989), omission and commission (e.g., Christensen and Hite, 1997),
message framing (Schisler, 1994), perceptions of fairness (e.g., Roberts and Hite,
1994) and, of course, demographic features (e.g., Baldry, 1987). Investigation of
these variables has made use of (a) self-reports of actual behaviour, attitudes, or
intentions (see Kinsey, 1984), (b) analytical models of tax evasion which rely
heavily on economic deterrence theory (e.g., Allingham and Sandmo, 1972), (c)
actual data obtained from tax agencies (e.g., McGraw and Scholz, 1991), and (d)
experimental studies which explicitly manipulate variables of interest (e.g., Beck,
Davis and Jung, 1991).
While each of these variables appears to have some causal or correlational
connection with compliance, there is often difficulty in replicating findings, and
links between variables have not been established (see Andreoni, Erard and
Feinstein, 1998). Further, it has become clear that techniques to improve
compliance based solely on surveillance and sanctions are inadequate (Kirchler,
1998; James and Nobes, 1998; Tyler, 1998) or counterproductive (Schwartz and
Orleans, 1967; Blumenthal, Christian and Slemrod, 1998), and do not explain the
voluntary compliance that occurs in the absence of surveillance (Alm, 1991), or
changes in attitude as a result of an appeal to conscience (e.g., McGraw and
Scholz, 1991; Mason and Mason, 1992). For tax systems that rely on voluntary
self-reporting of tax obligations, the need to understand the processes underlying
compliance is increasingly urgent.
The purpose of this chapter is to argue for a different approach to
understanding the motivations underlying taxpaying behaviour. This approach is
social-psychological in nature and aims to show that attitudes towards paying tax
are not stable, but are fluid and are an outcome of how one defines oneself in
relation to the tax system, the tax authorities and other groups of taxpayers. It
will be argued that the variables that are important in motivating behaviour will
be an outcome of this self-definition.
The chapter begins with a brief explanation of why current approaches are
inadequate for explaining taxpayer motivations and why a self-definitional
approach is useful. It will be shown, by analysing spontaneous taxpayer
comments written at the end of a tax questionnaire, that how one defines oneself
in relation to tax authorities and other groups of taxpayers affects attitudes to
paying tax, the strength of objection to or acceptance of paying tax, the perceived
fairness of tax, and the degree to which self-interest versus civic duty is likely to
be a motivating factor. It is concluded that understanding how taxpayers perceive
themselves in the tax system is fundamental to understanding the motivations
that underlie taxpaying behaviour.
Self-interest Models of Non-compliance
Perhaps one of the distinguishing features of the Australian tax system is that it is
purportedly about achieving justice and fairness, and turning inequality into
equality. Everyone who earns above a certain amount of income is required to
pay a certain amount of tax. Further, those who earn more are required to pay a
higher level of tax than those who earn less. This is deemed to be a fair process as
it essentially relies on the ability to pay. Apart from providing essential services
from which everyone benefits, tax revenue is also used to provide a safety net for
those in need.
Justice, fairness and equality, however, are not objective standards (Tyler,
Boeckmann, Smith and Huo, 1997). Perceptions of these standards vary with self-
categorisations (Wenzel, 2000). What might be perceived as being fair at the
superordinate level (e.g., paying more tax than others) could be perceived as
being highly unfair at the subgroup level. As Smith and Tyler note:
Distinguishing between levels of inclusiveness suggests that when a superordinate category is more
important to people, inequities between different groups represent an intragroup situation with
implications for collective cooperation and harmony. In contrast, if a particular group is more important
to people, inequities between different groups represent an intergroup situation with different groups
competing for resources and power (1996. p. 175).
This suggests that when a subgroup identity is salient, what is perceived as fair or
just is determined with reference to what other groups on similar dimensions
have. This is similar to the notion of horizontal inequity (Moser, Evans and Kim,
1995), except that the perceived inequity relates to comparison with other groups
rather than individuals. If other groups of taxpayers are perceived to be doing
better than one's own, and this is deemed to be illegitimate and unfair (no
obvious reason why 'they' should be getting a better deal than 'us'), collective
relative deprivation can result (see Walker and Mann, 1987); that is, a subjective
sense of collective, group-based injustice ('we have been treated unfairly'),
generating anger and resentment and a strong desire to remedy the situation.
However, perceptions of injustice are not simply related to inequality in
outcomes (distributive justice), but can also be related to the perceived unfairness
of the methods and procedures used to determine the outcomes (procedural
justice). If the methods by which outcomes are distributed are perceived to be
fair, then discrepancies in outcomes may also be judged to be fair (Tyler, 2001).
Song and Yarbrough (1978) noted that the taxpayers' complaint is not that too
many citizens cheat the government and get away with it, but that the
government provides unequal opportunities to different income groups.
Evaluations of procedural justice have been linked to voluntary acceptance of
decisions made by authorities, obedience to laws and legitimacy of authorities
(Tyler, 1990; Tyler and Lind, 1992). This implies that if inequities in outcome
(paying more tax than others) are perceived to result from unfair procedures in
the tax system, perceptions of group deprivation should increase, subgroup
identities should become stronger, intergroup hostility should increase, and
attitudes towards paying tax should become negative. In particular, Smith and
Tyler (1996) have argued that procedural justice concerns should be dominant at
the superordinate level for two reasons. First, distributive injustice concerns are
associated with subgroup differentiation that is not found at a superordinate
level. Second, being treated in a procedurally fair manner conveys that one is
valued and respected by other group members; a message that promotes self-
esteem and shapes behaviour toward other group members. At the subgroup
level, however, the distinctions between subgroups, which are not so apparent at
the superordinate level, are highly apparent, focusing concerns on distributive
outcomes ('what they are getting in relation to what we are getting'). Hence, in
the study being presented, it was predicted that taxpayer comments would relate
both to distributive and procedural justice concerns (Hypothesis 3). Further, in
line with Smith and Tyler (1996), it was predicted that concerns about distributive
justice would be greater at a subgroup than at a superordinate level of identity
(Hypothesis 4).
Research Design
Measuring Self-categorisation
Self-categorisation was investigated by analysing comments written
spontaneously at the end of the questionnaire. At the top of a blank page were
the words 'If you have any comments which you would like to add, please write
them below'. Comments written on this page were therefore unsolicited, allowing
spontaneously generated self-categorisations to emerge. As the questions of
interest in this chapter revolve around self-definition, only those who wrote
comments at the end of the questionnaire in relation to tax were included in the
analyses. The number of respondents who wrote comments relating to tax was
155 (15% of the pre-July sample). Ages ranged from 19 to 80, with a median age of
46. Fifty-two per cent of these respondents were male, while 48 per cent were
female. Thirty-seven per cent of respondents wrote half a page, 21 per cent wrote
a whole page, and 8 per cent wrote two pages.
To investigate whether social identities could be detected in the comments
respondents wrote, three different coders (two of whom were familiar with the
concept of self-categorisation) classified each respondent as reflecting personal,
social or unclear identities. A social identity was coded as existing if the
respondent referred to him/herself in a way that clearly indicated social group
membership. When a social identity was detected, the coders were required to
identify which social ingroup the respondent belonged to. The groups identified
were: average Australian; pay-as-you-earn; homemaker; single parents; single
income families; families with children; welfare recipients; retired; working
singles or couples; rural Australians; Australian; low income; hardworking
Australian; or small business owners. A consensus code (cf. Lupfer, Weeks, Doan
and Houston, 2000) was established if two or three of the coders agreed in their
codings. Consensus codes were established for 96 per cent of the coding
decisions. Of the 155 respondents, 40 were unable to be classified.
To establish level of inclusiveness, these social groups were collapsed into two
categories: superordinate (i.e., Australian; average Australian; hardworking
Australian) and subgroup (i.e., the remainder of the groups). These categories
reflected the degree to which others were included in one's self-concept, and
hence provided the means for investigating how self-categorisation affects
attitudes.
Findings on Self-categorisation
Taxpayers' comments relating to tax at the end of the questionnaire (n = 155)
were coded by two independent raters into categories of issues raised (Cohen's
kappa coefficient = .67, which Fleiss (1981) regards as an acceptable level of
intercoder reliability). The issues raised in the comments, and the proportion of
respondents who referred to each issue, are shown in Table 4.1, Almost all the
comments referred to perceptions of injustice. The most common complaint was
that the tax system was inequitable (51%).
It was hypothesised that the social identities of taxpayers would be discernible
from their comments (Hypothesis 1). This was so for the majority of those who
wrote comments (74%). Further, it was possible to determine the level of
inclusiveness of social identities. Sixty-seven per cent comprised a superordinate
level of identity, 48 per cent a subgroup level.
Issue referred to %
Attitudes to Tax
Implications of Findings
It seems clear that attitudes towards tax are not simply driven by personal self-
interest variables but are also affected by how taxpayers perceive themselves,
other taxpayers and tax authorities. Clearly, taxpayers can see themselves as
relatively interchangeable with other taxpayers, and interchangeable with a
subgroup of taxpayers in contrast to another group, or interchangeable with the
group of taxpayers as a whole. Self-interest, then, which might drive attitudes
and behaviour at a personal level of identity, is transcended by group values and
interests. The question of 'what is good/right for me' becomes 'what is good/right
for us'. Judgments about fairness reflect the degree to which other taxpayers are
perceived as similar to self or not. At a subgroup level, the focus is on distributive
injustice because the distinction between subgroups and unfair outcomes is more
clearly defined than at the superordinate level ('what are they getting that we
aren't?').2
The role of authorities, however, is also very important in affecting tax-related
attitudes. The less representative of taxpayers authorities are perceived to be, the
more resistant taxpayers are in their motivational postures and the more negative
their attitudes are towards paying tax. This, it is argued, stems from the fact that
unrepresentative authorities are perceived as illegitimate because they do not
represent 'us' appropriately (Haslam, 2001). Decisions made by an illegitimate
authority, then, are seen as invalid, removing the obligation to accept or obey
them (Tyler, 1998). The importance of this point in relation to social identity
should perhaps be spelt out a little more clearly.
In their argument, Smith and Tyler (1996) noted that procedural justice
concerns should be greater at a superordinate level because fair treatment by
those included within one's self-concept confers a sense of respect and pride in
oneself, which is important for self-worth. This sense of self-respect and pride
then leads to acceptance of rules and decisions made by ingroup others, even if
the outcomes do not personally benefit oneself. However, these authors also point
out that in order for authorities to gain obedience to rules and decisions, those
authorities must be included within the superordinate category. If they are
excluded, their influence is reduced because their treatment of group members
does not affect perceptions of self-worth (Tyler and Smith, 1999). While it might
be expected that at a superordinate level of identity (e.g., Australian) authorities
might normally be included within the self-concept, it seems clear that this was
not so in the present study. The degree to which authorities (both the government
and ATO) were perceived as representative was not greater at the superordinate
level than at the subgroup level. Given that the comments expressed by taxpayers
were in the form of complaints (i.e., negative), it seems likely that those who
perceived themselves as similar to other Australians were defining themselves in
contrast to authorities, while those at the subgroup level were defining
themselves in contrast to other subgroups as well as in contrast to authorities.
After all, if one's subgroup is perceived to be unfairly disadvantaged in
comparison to another subgroup, it is presumably natural to also focus on those
who have allowed the injustice to occur.
Two issues, however, render perceptions of authorities more problematic when
social rather than personal identity is salient. Firstly, perceptions of group-based
injustice become stronger when social rather than personal identity is salient
(Walker and Mann, 1987). Second, authorities not included within one's own
category membership run the risk of being perceived as condoning or
representing outgroup interests, making the potential for intergroup hostility
even stronger.
It is also interesting to note that while procedural waste, procedural violation
and procedural illegitimacy were referred to equally often at the superordinate
and subgroup levels, procedural inequity was referred to significantly more often
at the subgroup level. When distributive outcomes were perceived as unfair,
taxpayers focused on the unfair procedures that were directly relevant to
producing those outcomes. This, of course, makes sense because judgments about
unfair outcomes are not made in a vacuum. To judge an outcome as unfair must
mean that the procedures that produced the outcome are also judged as unfair. It
is difficult to see that an outcome could be judged as unfair if the procedures that
produced the outcome are judged as fair. This result also shows the
interdependence between distributive and procedural justice, a link that is
sometimes lacking in justice research (see Tyler, 2001).
Conclusion
The purpose of this chapter has been (a) to argue that an understanding of social
identity processes is fundamental to understanding why particular factors might
motivate behaviour in some situations and not others, (b) that to treat taxpayers
as rational, self-interested, utility-maximising actors in all situations is to limit
our ability to understand the processes involved in taxpaying behaviour, and (c)
to show that taxpayer social identities exist and vary according to the situational
and psychological contexts in which taxpayers find themselves. Resulting
differences in attitudes and perceptions of fairness are natural outcomes of this
self-categorisation process.
The attention given to the role of unrepresentative authorities in this chapter is
not to downplay the importance of a superordinate identity which incorporates
the ATO, the government and other taxpayers. As hypothesised earlier in this
chapter, self-categorisation at this level is where most compliance is to be
expected, since all subgroups become incorporated into the self-concept, making
better outcomes for everyone an important goal. The focus on self-interest is
replaced by a desire to ensure that all category members get a better deal, even if
this is not to one's personal advantage. This, as mentioned earlier, is what
researchers who focus on trying to induce 'civic virtue' essentially do. However,
the requirement is that the organisations in question (here the tax authorities and
government) are included within that superordinate category (Smith and Tyler,
1996; Haslam, 2001). Clearly, this is not an easy task. Perceptions of the ATO are
affected by perceptions of the government. Illegitimacy of one implies some
illegitimacy of the other. Further, it is possible that the context in which
taxpayers find themselves may be too well-defined to allow an easy transition to
a context in which category membership is more inclusive. While these issues are
obviously problematic, acknowledgement of them is a step towards
understanding taxpayer attitudes. A next step is to identify ways in which it is
possible and feasible for tax authorities to be seen as representing, as far as
possible, the interests of all taxpayers, rather than just a few.
Notes
1 Kerry Packer is Australia's wealthiest individual whose taxpaying activities were subject to public
scrutiny in 1991 (House of Representatives Select Committee on the Print Media, 1992) and in 2000
(The Australian, 'Packer Sues over Internet Tax Ads', 5 September, A. McGilvray and A. McKenzie, p.
1),
2 See also Wenzel, Chapter 3, this volume for a further discussion of level of identity and justice.
References
In this chapter, we examine the issue of movement into and out of the cash
economy. All indications are that the cash economy is on the increase worldwide
and is posing a challenge for sovereign states aiming to control its growth. An
increasing economic literature is addressing the factors that drive its development
and map its effects on the official economy, and society at large. Less widely
researched is the behaviour of those moving into and out of the cash economy,
what motivates them to take part, and what keeps them at a distance from this
kind of activity (see Wiegand (1994) as a notable exception). The literature on
individual tax evasion points to a number of possible explanatory variables,
among them perceptions of justice, civic virtue, moral obligation, social identity,
and social norms, to say nothing of the classic self-interest variables of human
greed as an instigator, and fear of punishment as an inhibitor (see Webley,
Robben, Elffers and Hessing, 1991; Andreoni, Erard and Feinstein, 1998;
Richardson and Sawyer, 2001). In this chapter, some of these ideas are drawn
together through the concept of motivational postures. Motivational postures are
central to the operation of the social rift model of regulatee responsiveness,
which comprises a set of explanatory propositions for how and why people
distance themselves from authority and find the psychological freedom to act
outside the constraints imposed by that authority. While social distancing is not a
sufficient condition for taking part in the cash economy (people must find the
prospect attractive or advantageous to them), it is a necessary condition for
throwing off the constraints imposed by an authority through laws and
regulatory practices, persuasion and punishment, obligations and moral pressure.
Background
The non-observed economy, variously referred to as the underground, cash,
hidden, black, or shadow economy, is monitored by governments worldwide with
a view to ascertaining its size and curtailing its growth. The concern that
governments express over the non-observed economy is based not so much on
the fact that it exists, but on the percentage of Gross Domestic Product (GDP) it
represents (Schneider and Enste, 2000; Bajada, 2002; Ott, 2002; Schneider, 2002). If
the non-observed economy increases substantially as a percentage of official
GDP, governments are likely to see an erosion of their tax base, and a reduction
in the revenue that they can expect to collect for the purposes of governance. Of
further concern is the fact that the government's ability to plan future economic
policy depends on reliable and accurate estimates of economic activity. Distortion
in the estimates weakens the accuracy of statistical modelling of future economic
performance.
Apart from making life difficult for government administrators, a thriving
non-observed economy can signal systemic problems of governance. For instance,
low public confidence in political and social institutions of governance has been
linked with a shift in productivity from the official to the underground economy
(Ott, 2002). Furthermore, a thriving underground economy provides something of
a haven for serious criminal activity (Organisation for Economic Cooperation and
Development, 2002). Within its fold lie a raft of activities including the avoidance
of tax (indirect or direct) and government regulation, defrauding government
social security and health care systems, and the illegal production of goods and
services forbidden by law (such as terrorism, drug trafficking or people
smuggling).
Little consensus exists about how the non-observed economy should be
defined or measured. The handbook of the Organisation for Economic
Cooperation and Development (OECD) (2002) on Measuring the Non-observed
Economy identifies four components: (a) underground production (activities that
are legal and productive, but concealed from public authorities); (b) illegal
production; (c) informal sector production; and (d) production of households for
their own final use. The size and impact of the components will differ across
countries, and not surprisingly, different countries (and researchers) adopt
different definitions of what components should be measured, and what the best
method of measurement is likely to be given the context (see, for example,
Schneider and Enste, 2000; Bajda, 2002; Ott, 2002; Schneider, 2002).
In spite of these difficulties, some progress has been made toward 'sizing' the
non-observed economy worldwide. Table 5.1 presents figures from the work of
Friedrich Schneider, tracing the growth in what is referred to as the shadow
economy in 21 OECD countries over a ten year period. While the average for
OECD countries is estimated as having reached the 16 per cent mark, the best
estimates available for developing countries and countries in transition are
substantially higher (41% and 38% respectively).
Table 5.1 Size of shadow economy in per cent of GDP for 21 OECD countries using currency demand
method
Research Goals
The purpose of this chapter is to focus on the behaviour of individuals and to ask
why some people in a country like Australia find their way into the cash
economy, while others do not. The data were obtained from the Community
Hopes, Fears and Actions Survey (Braithwaite, 2001; Braithwaite, Reinhart,
Mearns and Graham, 2001) conducted in 2000 and the Australian Tax System:
Fair or Not Survey (Braithwaite and Reinhart, forthcoming) conducted in 2001-
2002. The surveys were based on a stratified random sample of Australians
selected from publically available electoral rolls. In 2000, the response rate was 29
per cent (n = 2040) (see Braithwaite, Chapter 2, this volume for further details). In
2002, 1160 of these respondents were successfully contacted and completed a
further survey.
The data on cash economy activity were collected through a self-report
methodology in which respondents were asked about their actions, both as a
purchaser of services on which tax is not paid and as a supplier of services on
which tax is not paid. While the method is not without its problems or
limitations, self-reports have proven to be surprisingly useful across a range of
criminological research contexts (Junger-Tas and Marshall, 1999).
The survey findings are presented here in two parts. First, we examine the
social demographic profile of those who do not participate in the cash economy
(neither as a supplier nor purchaser) and those who do, and examine the
likelihood of people moving into and out of such activities over an 18 month
period. Next, we test five hypotheses about cash economy participation. All are
derived from a social rift analysis of regulatee responsiveness (Braithwaite,
Braithwaite, Gibson and Makkai, 1994; Braithwaite, 1995). The central idea of the
social rift model is that individuals control the distance they place between
themselves and an authority in order to psychologically minimise the threat the
authority poses and the influence the authority can have over their behaviour
(see Braithwaite, Chapter 2, for a description of motivational postures). As
individuals distance themselves from an authority, they may also perceive
weaknesses in the integrity of the authority. A perceived lack of integrity may
make it easier to devalue the authority. This process has been discussed by others
in terms of self-justification or rationalisation of the ways in which one departs
from or fails to satisfy social expectations, laws and rules (Sykes and Matza, 1957;
Thurman, St John and Riggs, 1984). Part of the process of distancing and
devaluing may involve finding social support for the position one is taking.
Through finding like-minded others, a shared social identity may emerge that
gives members credibility and confidence in their opposition to the authority (see
Taylor, Chapter 4, this volume). In certain regulatory contexts, however, the
distancing and the devaluing may be rather private undertakings, given that
individuals may be reluctant to expose supposed misdemeanours to public gaze.
With social distance and justification for the distance in hand, the individual is
psychologically free to behave in ways that are not necessarily to the authority's
liking. In these circumstances, the individual will not necessarily engage in non-
compliant activities: The opportunity and the incentive need to be present. But
given opportunity and incentive, non-compliant activity becomes an option, once
psychological freedom from authority influence has been achieved.
Hypothesis Development
Motivational postures reflect the degree to which individuals are accepting of a
tax authority in terms of its goals and ways of operating and the degree to which
they are sympathetic to the enterprise and open to its influence. Two postures,
capitulation and commitment, are sympathetic postures, the former because
resistance to authority seems useless, and the latter because paying tax is seen to
be a noble action. Three other postures represent the placing of greater distance
between the tax office and taxpayers. Resistance reflects the posture of those who
are within the system but object strongly to the way it is operating,
disengagement reflects the posture of those who have cut themselves off
completely from the system and want nothing more to do with it, and game
playing reflects detachment with effective defiance. Those who adopt the game
playing posture relax the social distance constraints to the point where they can
obtain the information they need to beat the tax office at its own game. Game
playing is about tax avoidance, that is, finding ways of legally using law against
the tax authority and sidestepping the obligation to pay tax (see McBarnet,
Chapter 11, this volume).
Involvement in the cash economy, at least as a supplier, does not involve
playing with or misunderstanding law, but rather it involves knowingly engaging
in illegal activity. For this reason, one might expect to encounter disengagement
more often than normal in this group, with commitment being displayed a little
less often. Thus, we hypothesise that the motivational postures most likely to be
related to being involved or not involved in the cash economy are disengagement
and commitment, with disengagement increasing the likelihood of participation,
and commitment decreasing participation (Hypothesis 1).
Those who break the law through cash economy activity are likely to be able
to justify their behaviour. Justification can take many forms and may vary from
individual to individual depending on experience and circumstances. A
justification, however, that is likely to be systematically expressed by those
involved in the cash economy is the rather broadly based one that the tax office
lacks integrity. By this we mean that the tax authority lacks soundness of purpose
in the goals it pursues, and the way it pursues them. Thus, we hypothesise that
cash economy activity will be higher among those who believe that the tax office
displays low integrity (Hypothesis 2).
The psychological purpose of distancing from and dismissing the integrity of
an authority is to gain a sense of well-being. In theory, personal well-being
involves creating one's own sense of security outside the reach of the authority.
This may be with like-minded others in one's social network or it may be through
one's own assemblage of beliefs and observations about how the world operates
and one's place in it. We thus propose three hypotheses associated with self-
protection from sanction, shame or guilt: (a) Cash economy activity will be
higher among those who do not anticipate being caught or punished by the tax
office (Hypothesis 3); (b) cash economy activity will be higher among those who
anticipate feeling no shame or guilt in the event of being caught by the tax office
(Hypothesis 4); and (c) cash economy activity will be higher among those who do
not identify with the group called 'honest taxpayers' (Hypothesis 5).1 Once
constraints on behaviour have been relaxed through removing concerns about
being caught and/or fined, through eliminating feelings of shame and guilt, and
through re-interpreting so called wrongdoing to make it acceptable, the path has
been cleared for participation in illegal activity.
Research Findings
The first task was to find out the social demographic characteristics of those who
were involved in cash economy activity in both 2000 and 2002, those involved in
2000 but not 2002, those not involved in 2000 but involved in 2002, and those not
involved at either time. The analysis was carried out for suppliers and purchasers
of labour. In order to find out who was a supplier, respondents were asked: 'Have
you worked for cash-in-hand payments in the last 12 months? By cash-in-hand
we mean cash money that tax is not paid on' (response categories were 0
indicating no and 1 indicating yes). Purchasers were classified according to their
answer to the following question: 'Have you paid anyone cash-in-hand payments
in the last 12 months for work or services they provided to you? By cash-in-hand
we mean cash money that tax is not paid on' (response categories were 0
indicating no and 1 indicating yes).
Of the 1119 people who provided answers to the questions about supplying cash-
in-hand work or services in the 2000 and 2002 surveys, we found that the vast
majority claimed not to be involved at either time (90.4%). The remainder
comprised 2.0 per cent (N = 22) who were supplying labour in 2000 and 2002, 3.9
per cent (N = 44) who were supplying labour in 2000 but not 2002, and 3.7 per
cent (N = 41) who were not supplying labour in 2000 but were in 2002. These
figures suggest that among those who are involved in the supply of labour, there
is considerable movement in and out of the shadow economy (79% are involved
at one time, but not the other). Such a pattern implies that the drivers of cash
economy participation are not stable but change as the individual's life space
changes. Within this frame, it is worth noting that between the 2000 and 2002
surveys, a goods and services tax was introduced in Australia, allegedly making it
harder for small businesses to operate successfully within the cash economy.
When attention is turned to those who purchase cash economy work or
services, again we find the majority of survey respondents uninvolved at both
times, although the percentage claiming non-involvement in purchasing is
substantially lower than the percentage claiming non-involvement in supplying
(75.4%). The remaining 25 per cent are distributed evenly across the remaining
three options for involvement: 7.5 per cent were involved in both 2000 and 2002,
8.0 per cent were involved in 2000 but not 2002, and 9.2 per cent were involved in
2002 but not 2000. Again, purchasers tended to move in and out of cash economy
activity across the 18 month to two year period, with 70 per cent involved at one
time, but not the other.
The numbers involved in the cash economy from this community survey are
on the smallish side for purposes of statistical analysis, but given the rarity of
data of this kind, some insights may be gleaned from examining the
characteristics of those who are repeat players, transitional players and non-
players.
We considered it likely that type of involvement as a supplier or purchaser
would be related to a range of social demographic characteristics such as age, sex,
marital status, number of children, family income, education, work status,
occupation, work sector, and country of birth (Australian versus non-Australian).
On the supply side, few of these variables were relevant to cash economy
involvement. The significant relationships2 to emerge were as follows. Younger
people aged 18 to 30 were more likely to be supplying labour in the cash economy
(and were also the age group most likely to move out of this kind of work from
2000 to 2002). Those owning their own business had disproportionately high
involvement in supplying labour in the cash economy. The relationship between
education and cash economy labour supply was most notable among those with
diplomas. A diploma, and to a lesser extent a tertiary qualification, indicating the
possession of skills of a specialised kind, increased the likelihood of cash
economy involvement. Being a supplier of labour in the cash economy at some
time was also more likely among those with a personal income, or a family
income, in the middle range of AS 19 000 to A$47 000.
Being a purchaser of cash economy labour had several links to social
demographic characteristics. Purchasers were more likely to be in the middle age
group (31 to 55 years), married, to have a university degree and a professional or
managerial occupation, to work full-time, and to be among the highest personal
income and family income group (that is, a personal or family income of A$48
000 to A$250 000). Purchasers were also more likely to work in their own
businesses.
Table 5.2 Mean scores and significant differences on motivational postures for four groups defined by
cash economy activity in 2000 and 2002
The extent to which the collection of taxes is perceived as being fairly shared
across social groups (the avoidance of a free-rider effect) was measured using an
index developed by Kinsey and Grasmick (1993). These authors argued that an
individual's perceptions of unfairness in contributions to the tax system could be
inferred from the degree to which citizens rated groups of taxpayers differently
on the question, were they paying their fair share of tax. Kinsey and Grasmick's
method is a useful one, but their measure is reliant on a small number of
taxpayer groups. After all, the amount of dispersion or unfairness detected will be
totally dependent on the nature of the groups rated by respondents. In order to
resolve this problem, we decided to focus on groups that would give us maximum
variation in occupational status from the very rich to the ordinary worker.
Differences in the treatment of the rich and powerful and the average working
person have caused considerable dissatisfaction with the tax system in Australia
(see Taylor, Chapter 4, this volume).
Respondents in the present study were given sets of occupations that fell into
the following categories: (a) top earning professionals and CEOs of corporations;
(b) small business owners and farm owners; and (c) trades people and low income
wage and salary earners. Ratings of fairness of tax contributions were averaged
for each group and the standard deviation of these scores was used as an index of
unfairness in contributions. The greater the standard deviation, the more the
respondent perceived differences in the fairness of tax contributions from high to
low status groups.
The use of tax for the public good was measured in terms of citizen satisfaction
with government. First, we asked: 'Overall, how dissatisfied or satisfied are you
with the way the government spends taxpayers' money?'. We then asked two
questions that formed the fair deal index: (a) Do you think that the tax you pay is
fair given the goods and services you get from the government? and (b) Would
you prefer to pay less tax even if it means receiving a more restricted range of
goods and services? (reversed scored).
Consultation was a special purpose scale made up of the following items: (a)
The Tax Office consults widely ^bout how they might change things to make it
easier for taxpayers to meet their obligations; (b) The Tax Office goes to great
lengths to consult with the community over changes to their system; (c) The Tax
Office is more concerned about making their own job easier than making it easier
for taxpayers (reverse scored); and (d) The Tax Office listens to powerful interest
groups, not to ordinary Australians (reverse scored).
Treating taxpayers with respect was based on the work of Tom Tyler (1997)
and asked respondents to indicate level of agreement with: (a) The Tax Office
respects the individual's rights as a citizen, and (b) The Tax Office is concerned
about protecting the average citizen's rights.
The scale to measure treating taxpayers as if they are trustworthy was taken
from the work of Braithwaite and Makkai (1994): (a) The Tax Office treats people
as if they can be trusted to do the right thing, and (b) The Tax Office treats people
as if they will only do the right thing when forced to (reverse scored).
Finally, the Australian Taxation Office (ATO) has a Taxpayers' Charter that
delineates 12 standards for how taxpayers should be treated (see Braithwaite,
Chapter 1, this volume for a description). Respondents rated the ATO on their
performance on each standard (Braithwaite and Reinhart, 2000). These responses
were summed to give a Taxpayers' Charter score which corresponds well with
the concept of procedural justice, that is, the degree to which the tax office
employs fair and reasonable processes in its dealings with taxpayers (see Wenzel,
Chapter 3, this volume).
Together, these measures represent tax office integrity. In Table 5.3, the
different kinds of involvement in the cash economy are related to each of the
integrity measures, first for suppliers, then for purchasers. With the exception of
the first measure (rich not paying) which is a dispersion measure, the integrity
scale scores range from 1 to 5, with a higher score indicating more of the attribute
being measured.
Table 5.3 Mean scores and significant differences on integrity measures for four groups defined by
cash economy activity in 2000 and 2002
These findings provide limited support for the integrity hypothesis when
applied to suppliers, and broader support for the hypothesis when applied to
purchasers. The major differences between the groups involved those who had
never been involved in the cash economy and those who were involved on both
occasions in 2000 and 2002.
Suppliers who were repeat players in the cash economy were less likely to
perceive the tax office as having respect for taxpayers and as enacting the
principles of the Taxpayers' Charter. These were the only integrity measures
related to being a supplier of labour in the cash economy.
Purchasers who were repeat players had more doubts about the integrity of the
tax office in terms of its adherence to the Charter, its consultation record and its
respect for taxpayers. They also complained more about disparities among groups
in the payment of a fair share of tax. Those who had never been involved in
purchasing in the cash economy were significantly more satisfied about the way
the government spent taxpayers' money than were those involved in purchasing
in the cash economy on a transient basis.
Table 5.4 Mean scores and significant differences on sanctioning measures for four groups defined by
cash economy activity in 2000 and 2002
Conclusion
Several concepts central to the social rift model of regulation have been shown to
predict cash economy activity. The model proved particularly useful in
differentiating those who were repeat players in the cash economy over a two
year period and those who had never been involved.
Suppliers of cash economy labour were more disengaged from the tax system
and expressed a relative lack of commitment to the system. They downplayed the
prospects of getting caught more than others, and expressed less concern, shame
and remorse for tax evasion than non-participants. Furthermore, they were more
likely to believe that the tax office was lacking in procedural justice, through
failing to meet their obligations under the Taxpayers' Charter and failing to
respect taxpayers. In social demographic terms, suppliers tended to be younger, to
have been educated along a vocational training path, to have family incomes
falling into the middle income range, and to be running their own businesses.
Purchasers who were repeat players displayed a different style of engagement
with the tax system and the cash economy. Like suppliers, when asked to imagine
a scenario where they were caught evading tax, they expressed less fear of being
caught and of the consequences. Interestingly, they did not distance themselves
from the honest taxpayer identity nor did they express a lack of commitment to
the system. As purchasers they apparently felt they could simultaneously be
supporters of the tax system and the cash economy. Nevertheless, they were
critical of tax office performance in terms of treating taxpayers with respect,
consulting with taxpayers and abiding by the Charter. Purchasers also perceived
greater disparity among occupational groups in the extent to which they paid
their fair share of tax. Demographically, purchasers were among the more
privileged in our society. They belonged to high income families, they were well-
educated, employed, had professional or managerial occupations, were married
and were in the middle age range. Purchasing cash economy labour was also
more common among those with their own businesses.
Thus, purchasers and suppliers in the cash economy are not too much alike,
apart from the fact that both are more likely to be found among the self-
employed. Suppliers are the less successful group according to standard
indicators: They are not as rich as purchasers, they are younger, less settled and
less well educated. Moreover, they have placed themselves at a greater distance
from the tax authority, expressing doubts about the office's procedural fairness
and little concern about the impact of sanctions. Suppliers show signs of placing
themselves beyond the reach of regulators, more so than appears to be the case
for purchasers.
These findings are only the beginning to understanding how individuals
become caught up in cash economy activity. Even so, there is evidence here to
support some of the main ideas that economists have been touting in their
analyses of growing cash economies at the macro level. For instance, sanctioning
and the perceived chances of being caught have emerged as a strong and
consistent predictor of cash economy activity for both suppliers and purchasers.
Future work by our research group will explore the experience of sanctioning and
whether or not this intervention discourages future cash economy activity. While
deterrence theory is clearly relevant to the management of individual
participation in the cash economy, so too are notions of tax system integrity and
procedural justice. As Ott (2000) and her colleagues have observed in transitional
economies, the cash economy provides an escape for people who have lost
confidence in government and what it purports to represent within democratic
society.
The surprising, and very important outcome of our research to date is that
issues of integrity, in the sense of adherence to the Taxpayers' Charter, social
fairness and citizen inclusiveness, are more important in predicting the future
behaviour of purchasers than suppliers, and purchasers do not, it appears, exist
on the margins of our society. They are more often than not found among the
privileged ranks of our social order: High income earners, the well educated,
professionals, managers and business owners, married and in the prime years of
their lives. They have the credentials for being 'masters of the universe' to borrow
a phrase from Tom Wolfe's (1987) The Bonfire of the Vanities. The question is:
Are they also driving tax evasion in the cash economy?
Notes
1 While it would have been interesting and more directly relevant to ask about identification with the
category, 'dishonest taxpayer', we were concerned about the reliability and validity of responses to
this question in a general population survey. Such a question is better suited to populations with a
known and proud history of resistance to taxation.
2 The four types of involvement in the cash economy were cross-tabulated against the social
demographic variables (each expressed as a categorical variable) and each relationship was tested for
statistical significance using the chi-square test of independence.
References
In the closing decades of the 20th century, public administrators in many Western
nations faced growing demands to restructure their traditional approaches to
service provision and law enforcement. In some areas of public administration
this was coupled with challenges to fundamental legitimacy. The new demands
and attacks on the legitimacy of public administration emerged in a context of
considerable intellectual debate over the efficacy of alternative approaches to
state regulation. The net result of controversy and debate were fundamental
changes in how public agencies do business. Nowhere was this more apparent
than in the area of taxation administration, and developments in Australia are
illustrative. The Australian Taxation Office (ATO) moved from its long
established style of command and control administration and tax enforcement to
a program of responsive regulation. It did so even as it faced the added challenge
of implementing an unprecedented and sweeping program of taxation reform.
This chapter examines staff perceptions of and early experiences with the ATO
program of responsive regulation. It begins with brief comments on the changing
political, popular and intellectual environment of public administration at the
dawn of the new millennium. This is followed by a review of how the ATO
responded to the increasingly challenging environment of taxation
administration and a brief explanation of the objectives of this investigation. An
outline of the methods employed in our research is then presented. Drawing from
interviews with ATO personnel, the bulk of the chapter is a discussion of what
was learned about the implementation process and its reception by field level
personnel. The chapter concludes with summary conclusions that may be useful
for both the ATO and other public agencies contemplating or facing a shift to
responsive regulation.
Background
Commencing in the 1980s, public administration came under growing state and
public demands to become more market-focused, service-oriented, open and
efficient (Hughes, 1994; Sparrow, 2000). Many of the management strategies and
practices espoused by critics were borrowed from private enterprise. They
focused on 'service, customers, quality, and process improvement' instead of
'compliance management, risk control, or structuring the application of
enforcement discretion' (Sparrow, 2000, p. 2). They required a shift from 'rigidity'
with its focus on 'equity and due process' to 'flexibility' with a focus on 'results'
(Hughes, 1994, p. 259; Gregory, 1999, p. 63). Change of this magnitude can be
difficult, particularly for public agencies whose principal role is not service
delivery but regulation or enforcement. Taxation agencies, for example, faced the
challenge of implementing these reforms while simultaneously delivering
obligations for compliance with taxation regulations (Sparrow, 2000).
Coupled with new demands on public administration generally were
challenges to the legitimacy of tax systems. In some countries this began with
reports of citizen or business outrage at abuses of power by tax administrations.
In the United States, citizens charged that personnel of the Internal Revenue
Service (IRS) were 'rude, abusive, or unhelpful', and also that 'the IRS retaliates
against those who criticise it' (Report of the National Commission on
Restructuring the Internal Revenue Service, 1997, pp. 1-2). There were complaints,
moreover, that the IRS was 'reactive' rather than 'proactive', that customer
satisfaction was low, and that the public believed the quality of IRS service was
'deteriorating' (1997, p. 11). As a result of these attacks, the National Commission
on Restructuring the Internal Revenue Service was created and charged with
recommending 'changes to the IRS that will help restore the public's faith in the
American tax system' (1997, p. v).
The IRS did not stand alone as a target of citizen criticism, however. In the
1990s the Australian media increasingly drew attention to poor ATO practices
and excessive use of power. There were charges of ATO harassment of 'small fry'
to pay small amounts of money they owed while 'big debtors' were let off the
hook (Sydney Moming Herald, 1996a, 1996b; The Age, 1996a, 1996b). ATO staff
were accused of being 'bloodhounds' (Dobbie, 1993), who 'monstered' and
'tormented' taxpayers, and who launched 'covert attack' on them (Australian
Financial Review, 1997; Anonymous, 1997a, 1997b, 1997c; Hunt, 1998).
Accusations of excessive and unfair use of power (Gumley and Wyatt, 1996) were
balanced with claims that the ATO was 'out of touch' and 'lacked understanding'
of'commercial reality' (Anonymous, 1997a, 1997b, 1997c). There were suggestions
that the ATO's actions were 'morally wrong' (Sydney Morning Herald, 1996a,
1996b) and that poor ATO use of penalties 'threatened the integrity of the tax
system' (The Age, 1996a, 1996b). Newspapers highlighted unacceptable internal
ATO practice and sarcastically mocked ATO wastage of taxpayers' money (Sun
Herald, 1996). Criticism and demands for change came from the community and
government alike. These demands challenged the 'traditional' social order within
the ATO and taxation administrations that operated in similar fashion.
Challenges to the practices and the legitimacy of public administration
occurred against a backdrop of significant policy developments in the movement
to enhance citizen and corporate compliance with regulatory rules (Parker, 2000).
In the preceding decades, the dominant approach to compliance assurance was
one referred to as 'enforced compliance' (Shover, Clelland and Lynxwiler, 1986) or
'command and control regulation' (Aalders and Wilthagen, 1997; Dodd and
Hutter, 2000). As seen by its supporters, this approach to regulation parallels the
way criminal justice agencies and personnel typically approach their work.
Drawing from the assumptions and principles of deterrence theory, advocates of
command and control regulation call for precise and narrowly drawn rules,
threatened penalties for non-compliance, and punishment for violators (Reiss,
1984; Grabosky, 1995; Dodd and Hutter, 2000). Critics, however, charged that
command and control regulation builds on mistaken notions about business firms
and the meaning of non-compliance, and these flaws make it unreasonable and
ineffective in practice (Bardach and Kagan, 1982; Grabosky, 1995; Paternoster and
Simpson, 1996; Aalders and Wilthagen, 1997; Vaughan, 1998). Deterrence based
regulation, they said, could not accommodate what was learned about the nature
of non-compliance; its 'one size fits all' conception of threats and punishment
seemed poorly suited to the empirical realities of the matter.
Controversy over the merits of enforced compliance and other approaches to
regulation prompted a synthesis known as 'responsive regulation' (Braithwaite,
2002). This is regulation that is 'responsive to industry structure in that different
structures will be conducive to different degrees and forms of regulation' and also
one that is 'attuned to the differing motivations of regulated actors' (Ayres and
Braithwaite, 1992, p. 4). This style of administration is knowledgeable about and
takes into account the problems, motivations, and conditions behind
noncompliance. Strong emphasis is placed on educating firms about rules and
assisting them in efforts to comply, while programs that rely principally on
threats and the mechanical imposition of penalties are de-emphasised. The ATO
adapted and extended the Ayres and Braithwaite approach to responsive
regulation to include individual taxpayers and other entities, as well as the
owners of firms. The ATO assumption was that different types of taxpayer have
differing motivations and taxation structures and that regulating taxpayers
responsively may encourage voluntary compliance.
Programs of responsive regulation legitimise and make available to officials a
range of options when responding to compliance problems (Gunningham and
Grabosky, 1998). The fundamental assumption is that a substantial proportion of
taxpayers will self-regulate with minimal external monitoring so long as they are
treated fairly by regulatory officials and are met with understanding and
assistance should they encounter problems doing so (Tyler, 1990; Makkai and
Braithwaite, 1996). For citizens and businesses that fail to comply despite appeals
and cooperative actions, officials may escalate their responses and sanctions in
proportional fashion. As the seriousness of infringements and the wilfulness they
represent increases, officials may employ less conciliatory responses.
Research Findings
Perceived Legitimacy
The legitimacy of the Compliance Model for some respondents also evolved
from the way it reduced their uncertainty and fear when they creatively adapted
agency resources to address difficult cases. They enjoyed the new latitude and
'breathing space' afforded them by the increased capacity to customise responses
to the case at hand, that is, to use their discretion as Black (2001) describes. Some
staff, moreover, spoke of the new feeling of security and validation they gained
from regulating responsively:
[The Compliance Model] justifies and protects ATO staff. Staff get comfort from knowing they have
other options [than audit and prosecution].
They recounted how they benefited from the reinforcing qualities of an 'action
plan' that structures and guides the resourcefulness called for by the Compliance
Model:
In the past after a review was done, officers were left with no path to follow. The Compliance Model
shows a complete compliance program. It gives staff a path to follow.
The Compliance Model gives the basis for making cultural change. The desire to make this change is not
just the rhetoric of the leaders. The Model gives us a different framework for dealing with any activity in
the future.
Such comments suggest that staff felt empowered and personally reinforced by
the Compliance Model. During the period in which the transition was made from
a formal approach based on definitive prescriptions for forcing compliance to a
less formal approach based on creative use of resources to encourage compliance,
feeling empowered can be critical for agency success. With the command and
control model, staff felt powerless to make change when policies were not
effective. Much of the time, they had no avenues by which to gain compliance
other than through the use of threats. The Compliance Model, in a very real
sense, gave them permission to 'think on their feet', which helped to develop their
capacity for making judgments on a case-by-case basis.
For some, then, the legitimacy of the Compliance Model was bolstered by
belief that it is more rewarding to staff personally to encourage user-friendly
compliance than to impose mechanically the penalties prescribed by command
and control administration. One manager told us that his team members were
'happy working this way, because it is a better, nicer way and they are seeing a
changed response from the public'. He also said that the relationship between the
ATO and the industry group he was working with had improved.
Communication was enhanced because 'working this way builds a greater
rapport with the taxpayer'. Consequently, he felt 'much happier working this
way'. Respondents, moreover, appeared to realise that when both parties to a
transaction are happy with the results, future transactions become much easier to
negotiate.
Adoption and use of the Compliance Model also contributed to increased job
satisfaction for some staff. Respondents suggested, for example, that it improved
the quality of communication when compliance issues were raised with
management. Respondents reported that staff were talking about the Compliance
Model and its principles during meetings with management and were pointing to
how it 'said' certain factors must be considered. One team leader made the point
that 'staff are trying hard' to embrace the Model and change the way they work
for the better. He also said that he had 'seen an improvement in sharing best
practice, sharing stories in meetings, and an increase in discussion [among staff]
about cases'.
This officer had noticed a decrease of the old 'audit culture' that had supported
the use of punishment driven compliance and an increase in the ability of ATO
staff to develop new strategies. His team had regularly discussed the Compliance
Model and had used case studies to practice using its principles. These are
examples of learning and confidence building via practice and staff discussions,
which are tools that have been found to be very helpful in learning to regulate
responsively (see Braithwaite, Chapter 9, this volume; Shearing and Ericson, 1991;
Sparrow, 2000).
By improving staff communication, the Compliance Model also improved
inter-staff relationships. One person commented that the team she worked in was
thinking not only about how the Compliance Model fosters better
communication between ATO staff and taxpayers but also between staff
members themselves. She believed this change in thinking had occurred because
of the 'support' and 'encouragement' of their manager, and that the team had
received effective training in the use of the Compliance Model. She added that
the team enjoyed doing the training because it was 'interactive' and that there
was talking afterwards within the team about the training and what it meant for
their work. As a result of the increased quality of staff interactions, many
interviewed staff reported being happier with the quality of their work and with
the response they were getting from taxpayers.
Overall, the interviewed staff perceived implementation of the Compliance
Model as a legitimate goal for the ATO. The strong points of the Compliance
Model were reported to be that it encouraged staff creativity and sensitivity to
the needs of individual taxpayers. It provided guidance for staff as they
developed taxpayer specific plans of action, it developed the capacity of
individual staff for making judgments or 'thinking on their feet', it increased job
satisfaction among the staff, it improved communication between staff members
and between staff and management, and it improved ATO/taxpayer relationships
in many of the more difficult cases.
This is not to deny, however, that difficulties inevitably emerge which must be
faced when putting in place new policies. This may be true particularly when
efforts are made to do so in large organisations, where both organisational and
individual level barriers or constraints may prove difficult to surmount. Our
interviews suggest this was the case at the ATO.
Organisational Barriers
When implementing far reaching programs like the Compliance Model in a large
bureaucracy like the ATO, research has shown there can be resistance to change.
Kanter (1989) found that 'in the traditional bureaucratic corporation...narrowly
defined jobs constricted by rules and procedures also tended to stifle initiative
and creativity' (p. 280). Some of our respondents noted how this tendency to play
by the rules or maintain the status quo in the organisation manifested itself in
various structural barriers to full implementation of the Compliance Model.
An important barrier arose from the fact that communication, networking, and
mutual support can prove difficult in a large, segmented organisation that also
serves a vast geographic region. At the time the interviews were conducted, the
ATO employed 19,000 staff split into 12 divisions, with 34 separate offices spread
out over a country of 7,692,030 square kilometres (Commissioner of Taxation,
2000; The Europa World Year Book, 2001). The result can be a sense of operational
and individual isolation. The impact of this on some offices in this structure is
that 'staff memory' tends to be more local than organisational. In other words,
issues of concern to the bureaucracy as a whole may not be salient for offices and
staff in far flung parts of the ATO's jurisdiction. The implementation of an
innovative idea like the Compliance Model in a hierarchical and segmented
organisation like the ATO required the institution of procedures at all locations
that ensured effective leadership, good communication between management and
staff when problems arise, and creation of staff opportunities for mutual support
and knowledge sharing sessions when needed.
Another barrier to implementation stems from staff being presented with
unfamiliar tasks caused by adoption of the Compliance Model. Interviewed staff
who had focused for many years on the same type of work faced new challenges
when learning to regulate responsively. These challenges varied from case to case,
from person to person, and from office to office, making differing demands on
each staff member's particular skill set. In some cases there appeared to the
respondents to be a single skill, single focus mind set that served to limit
acceptance of the new approach:
There is a definite understanding that the Compliance Model is applicable to the ATO generally, but not
necessarily to [a staff member's] work in particular.
Staff are not looking at the connections between the Compliance Model and what they do in their
work...there is a lack of feeling that it is a 'whole thing'. People are just focusing on bits of it.
Another respondent told us that 'too many people have stayed in the one job for
too long and now find it difficult to change'.
Our interviews suggest that acceptance of and the ability to make use of the
Compliance Model depended somewhat on the type of work staff performed;
thus, specialisation and having a defined skill set constrained the ability of some
staff to take a broad, organisational view of compliance. The ways in which
activities like audit and help and education could be related was not clear to
them, and this caused confusion. Some respondents held the view that staff with
a 'rulebook mentality' were most likely to resist taking on the responsibility of
thinking for themselves. For example, those whose work traditionally focused on
long-term investigation of past taxpayer activities were less accepting of the
future-oriented activities of the Compliance Model. One interviewee showed his
lack of interest in the model when he stated:
The Compliance Model is irrelevant in ninety five per cent of dealings with clients. We should start
again...go back to audit and refine the way we do audits. We could have refined the audit process.
[Instead] we just said it wasn't working.
By July 1999, a feeling had grown in some ATO branch offices that:
The Compliance Model is on everyone's mind, but there are different perceptions of what it means. There
is an element of it not getting through...the assimilation of help and advice as part of what we do. Staff
still compartmentalise activities. The reason is that different skills are required and many staff do not
have the skills to do the help and education work.
Acceptance of and ability to use the Compliance Model seemed to come more
easily in those parts of the organisation that did client contact work aimed at
helping taxpayers. One manager of a team with a client contact focus commented
that his team thought the Compliance Model was simple and easy to grasp. He
and his staff viewed this as a positive thing because it helped them readily
develop their own thoughts on help and education. On the other hand, staff
whose work had a long-term and in-depth investigative focus reportedly were
unaccustomed to 'thinking on their feet'. Such skills are required by an eclectic
administrative style like the Compliance Model, which integrates several
formerly separate approaches.
One barrier noted by the interviewed staff was the existence of an 'entitlement
culture' at the ATO in which some managers and staff do not routinely question
the relevance or feasibility of a new idea that administration puts before them. In
this situation, staff did not do things for themselves or initiate action on a
problem but instead wait for others to make the needed changes. In addition,
management can buy into this status quo mode of thinking and influence staff
members to accept the way things are. For example, one staff member
commented that using the Compliance Model was not easy because a senior
manager in his branch office had 'views' about staff using personal judgment
rather than following policy and established guidelines. This example highlights
Black's (2001, p. 14) comments on the role of leadership in 'defining'
organisational culture. A manager's world view has a strong influence on the
approach staff take to their work, and the acceptance or otherwise of a change in
the way they work.
In sum, organisational barriers to implementation of the Compliance Model at
the ATO that were noted by our respondents include inherent constraints on
communication in a large, segmented organisation spread out over a vast area,
the unfamiliarity or newness of the tasks presented to staff by the adoption of the
Compliance Model, the lack of experience of some divisions with people-oriented
activities, and the existence of a culture at the ATO in which staff do not attempt
innovation, but rather wait for others to make changes to established routines.
Our respondents noted that for some, fear of change manifested itself in
criticism of academics and their theories. Some ignored the Compliance Model,
saying it was just theory that had not been tested in the taxation arena. One
person suggested that there was 'an anti-intellectual bias in the organisation and
anything intellectual is mockingly dismissed'. This cynicism seems to have
disguised a fear of change, highlighting what Kanter (1983, p. 92) called the
reinforcement of a 'culture of inferiority'. In a very real sense, 'outsiders' who
come into an organisation, suggest new ideas and achieve success, can serve to
make 'insiders' depressed about their abilities. In such a situation, there are few
opportunities for change for the better.
Some respondents had noticed that fear of change seemed to manifest itself in
a lack of confidence for some staff members, which was accompanied by
discomfort when asked to work with the Model. This problem seemed to be one
of inexperience, as staff were not used to applying the higher levels of personal
judgment asked of them by the Compliance Model. Two respondents talked of
this discomfort among the staff:
[After having] developed a friendly relationship with a taxpayer, they can't now get tough. It's like
dobbing in a mate. Before the Compliance Model, staff came from a strong position or base when going
to a taxpayer's premises to do an audit. There was no personal relationship...staff need greater clarity
about what the lower level of the Compliance Model is all about.
It is hard to be soft one time and tough the next... [It is] hard to be all things.
For some, to be both 'soft' and 'tough' was to exhibit 'questionable behaviour'.
Rather than appear to be clumsy implementers, some staff appear to have used a
'corporate myth' to rationalise protecting command and control administration at
the ATO and the way they had always worked. Such myths have been found to
warn group members away from 'unacceptable' behaviour (Douglas, 1986). In this
case, the myths helped some staff to rationalise that it was best to uphold the
clear cut rules of the 'enforced compliance culture'.
We were told that some staff feared the undesired consequence of a loss of
status because their specialised role perhaps would no longer be valued, due to
the new focus in the ATO. Hollander (1964) described this individual barrier as
feelings of lack of security. Much of this concern about loss of status appeared to
some respondents to be due to mistaken interpretations of the Model. Many
auditors were found to have the mistaken impression that the Compliance Model
meant they and other auditors should be doing 'help and education' work, as
audits had become a thing of the past.1 Still other auditors believed there were
simply no other options for them to be anything but tough and impersonal. Some
perceived no need to try and understand the taxpayers they were dealing with.
Typical were the two respondents who said:
By the time a taxpayer gets to the auditors, they are all crooks. [Auditors] have to be tough; [auditors]
can't be nice or soft.
Staff can't do both help and audit work...that's changing hats midstream. Why do I want to distribute
pamphlets? I have tertiary qualifications. Half of the help and educate staff don't have tertiary
qualifications.
Thus, the ATO's adoption of the Compliance Model was perceived by some
staff as undermining revered positions that had the 'superior' skills of auditing
and legislative drafting. Other staff reportedly avoided becoming victims of this
all or none mentality and, as a result, were able to develop an understanding of
the holistic nature of responsive regulation (see Hobson, Chapter 7, this volume
for a description of working within the model).
A related personal barrier some staff had to surmount was belief that the
Model required use of skills they did not have and did not value, such as the
'people' skills required to educate and provide assistance to taxpayers. The issue
for them was not the 'specialist versus the generalist' argument discussed above.
Rather it was whether they had the ability to 'mesh' what they do with the
prevailing desired organisational goals or outcomes. Additionally, some auditors
believed that having to do other types of work would contribute to the loss of
their specialised skills and the loss of ability to maintain the quality of their
auditing work.
Other individual level barriers to implementation noted by the respondents
revolved around the degree to which leaders endorsed and visibly practiced the
Compliance Model. Successful implementation of a new approach requires not
only effective management but also leadership. 'Leadership and management
differ in that management is designed to promote stability or to make the
organisation run smoothly, while the role of leadership is to promote adaptive
change' (Wood, Wallace, Zeffane, Schermerhorn, Hunt and Osborn, 1998, p. 524).
Leaders can serve as role models for those less sure of how to proceed. On the
other hand, lack of leadership can prove to be a powerful constraint on
implementation efforts. Some of our respondents believed this occurred at the
ATO, as lower level staff looked to senior staff for 'permission' to adopt the new
approach. One told us that 'those at the lower levels might be aware and
understand, but they will not do it. They will wait until told by their managers to
do it'.
A recurrent theme in our conversations with ATO staff members who had
received training and who had experience using the Compliance Model was
belief that staff do not buy into a new approach as readily when management
does not meaningfully involve them in decision-making. Researchers have
consistently found that one of the basic requirements for effective
implementation is communication and consultation with those affected
(Hollander, 1964; Argyte, 1972; Wood et al., 1998). Staff participation in
management decision-making has been linked with positive worker attitudes,
organisational loyalty, and feelings of security and job interest among workers
(Hollander, 1964). The Compliance Model training sessions were a positive
contribution to implementation for some staff in that they signalled that
management was giving them an opportunity to understand where the
organisation was going. For others, the training sessions had a negative effect on
their attitudes toward the Model and their feelings of security within the
organisation. This may have been partly because their participation was not
active and they perceived the sessions as simply a 'token', sometimes 'contrived'
to get staff to do what management wanted (Hollander, 1964, p. 41). Indeed, some
respondents believed they had little or no real input in the implementation
process:
ATO staff are tired of talking about it. They put ideas up to management and never hear of any action, so
they lose enthusiasm.
Staff just get it done to them... there is reduction by stealth. Staff had no opportunity to provide input to
these things. They were just told it was happening.
The respondents frequently expressed the view that before staff would fully
adopt the Compliance Model their managers had to demonstrate their acceptance
of and enthusiasm for its principles. This need for managers to lead by example
was also characterised as the need for them to 'walk the talk'. The respondents
believed a manager's behaviour should visibly model the principles of the
Compliance Model. If it did not, efforts to implement a new administrative
approach can be severely compromised:
Much depends on team leaders. They need to talk about the [Compliance Model with staff].
If a manager is negative, staff will be negative. They won't speak up and will drift with the others.
There was the perception of some staff that 'middle management is one of the
main blockers' and that some managers had not bought into the Compliance
Model to the point they were actively pursuing it in their divisions:
Managers and directors are not fully responding despite all the work that has gone into them.
Senior people know and understand the Compliance Model but are just paying lip service to it. They can
get away by just referring to it in documents without having to do anything else.
Some staff commented that their managers were 'too busy' to be actively
thinking about implementing the Compliance Model. Other staff noticed
evidence of passive resistance by leaders who represented special areas of
interest. Commenting about one group of specialist staff who were viewed as
leaders, one respondent said:
If we are about change and doing things differently, they will come along for the ride, [but they will] be
passive, [they will be] blockers. They filter information, and they have a position of power, and they can
use it wrongly [because] they want to be comfortable. They create the environment for lower level staff.
Several staff commented that it was the Compliance Model itself, which was
providing leadership to the ATO by setting direction and setting up avenues of
support for both management and field staff. Some believed that if they used the
principles in the Compliance Model, managers would offer more support for how
they did their job. If a manager is only paying lip service to or is passively
resisting the Compliance Model, then the inspiration and direction it can provide
for the organisation cannot be utilised. One person stated that:
There has been some lack of direction from the top in the past. Staff felt that management were just
going around in circles and doing the same things over and over because they didn't know where [else]
to go. No one knew where they were. Staff did not know where they stood. [We are] starting to get a
feeling amongst staff now that leaders are getting it a bit more consistent. Everyone needs to know
where they stand at all times.
A final barrier to implementation that was noted by our respondents was that
some managers might lack the interpersonal skills needed to stimulate, inspire,
and encourage staff in innovative initiatives. The groups with managers that have
such 'people skills' may be the most likely to experience success with the
innovative Compliance Model. In the words of one respondent:
A manager makes the section. [They make) a big contribution, [so] the ATO should look at managers'
personalities and see if they are suitable to be managers who will take on change and encourage others to
do so...[It] needs to identify the personalities and plant them in the right place...spread them around.
Some staff expressed the belief that the ATO should actively recruit these social
skills necessary for the implementation of the Compliance Model:
Give them a problem to solve and look at how they do it...the way they think...the processes they use to
solve the problem...See if their attitudes reflect the type we want in the ATO.
Staff comments suggest that a manager with technical skills aplenty but few
social skills or a manager that is too busy to actively pursue innovation may not
be the best candidate to lead, aid, and inspire implementation of the Compliance
Model. For many of those interviewed, the implementation of a user-friendly,
people-oriented program like the Compliance Model requires a communicative
leader with good social and verbal skills. Leaders who lack these attributes
impede organisational efforts to reform or change.
In sum, it appears that many individual level constraints acted to circumscribe
the effectiveness of efforts to implement the Compliance Model. They include: (a)
passive resistance to the model by some individuals in positions of power; (b)
staff member's fear of making mistakes when changing long established rules and
procedures; (c) lack of trust or faith in academic theories; (d) a staff member's
discomfort with the idea of using higher levels of personal judgment in dealings
with taxpayers; (e) a staff member's fear of loss of status or job security due to
specialised roles no longer being valued; (f) lack of effective, innovative
leadership by some managers; (g) lack of management inclusion of staff
participation in decision-making during the early stages of implementation; (h)
lack of management acceptance of, and enthusiasm for the Compliance Model;
and (i) management not possessing the social skills required to inspire and
encourage innovation.
ATO staff welcome acquisition of new skills of report writing and the
interpersonal skills they will need to work with and negotiate effectively with
taxpayers. They recognise, moreover, that change is a long term process.
Note
1 Audit work 'traditionally' has been regarded [by auditors within the Australian Taxation Office] as the
"real' work of a taxation administration. Auditors had a particular status within a tax office and were
seen as members of an elite club. In the days when much of a taxation administration's work was
transactional, audit staff possessed specialised skill sets, often developed formally as well as on the
job. Auditors competed amongst themselves to raise revenue. The auditor raising the highest amount
of revenue during the year received organisational recognition, and a 'god-like' reputation.
References
A 'champion' has status or clout and advocates on behalf of others. Champions, in fact, help legitimise
the idea originator, serving as a bridge or translator between the sometimes unconventional originator
(ideator or inventor) and the more traditional organization (Rosenfetd and Servo, 1990, p. 54).
When you're pioneers, it's always difficult and you feel as if you're up against it or you're walking on
quicksand. But when it all happens, it all just clicks into place, and my belief is it will click into place. It's
just a matter of finding that switch that will make it happen (quote from ATO champion interview).
This paper examines change 'champions' in the Australian Taxation Office (ATO).
Specifically, it explores why certain ATO staff members have become the
champions or advocates of the ATO Compliance Model, a tax compliance
innovation introduced into the ATO through collaboration with researchers at the
Australian National University. It seeks to understand why ATO champions chose
to become the 'translators' of the Compliance Model, as the above Rosenfeld
quote suggests, and why, despite many barriers, they had faith that change would
prevail, as the second quote suggests. Specifically, this paper seeks to examine
how these individuals understood and made use of the Compliance Model; and
what this can tell us about how the ATO might take the Model forward.
Why focus on the ATO Compliance Model champions? First, the presence of
motivated individuals who advocate and advance new ideas is believed central to
operationalising positive change within an organisation (Bouwen and Fry, 1991;
Rogers, 1995; see Job and Honaker, Chapter 6, this volume; Shover, Job and
Carroll, Chapter 8, this volume). Much has been written about the personalities
and practices of these innovators. Yet most research has taken place in the
corporate sector where innovation and profit margins are key operational and
behavioural drivers (Damapour and Gopalakrishnan, 2001). The public service
has remained a relatively uncharted area of inquiry, yet employs a significant
proportion of the workforce (see Australian Bureau of Statistics, 2001), who are
charged with delivering timely, effective and efficient public policy through
large-scale organisational structures.1 Therefore, studying champions in the ATO
offers a rare glimpse of how individuals innovate in a public service department.
Second, research often focuses on the adoption of new ideas that have been
generated within an organisation (West and Farr, 1990). By contrast, the
Compliance Model was informed by independent academic research, that was
developed within the ATO context (see Chapter 1 for details of the history and
development of the Compliance Model). Studying the ATO Compliance Model
offers an opportunity to examine how such an externally generated model is
adopted and translated into meaningful practice by individuals within an
organisation - a process that may become important to understand further as
partnerships between universities and public organisations grow in the face of
changing knowledge climates.
Finally, there is the type of innovation that the ATO Compliance Model
represents. Innovations normally pertain to either changes in products or changes
in processes (Damapour and Gopalakrishnan, 2001). A product is goods or
services offered to a client, whilst a process is the mode of production or delivery
of goods or services. The majority of literatures about innovation and change in
organisations focus around the introduction of new technologies or systems of
production in industry (see Rogers, 1995). By contrast, the Compliance Model is
both a product and a process - a 'soft' (ways of thinking, ATO philosophy) and
'hard' (routine practices, services, publications) form of change (Harrisson,
Laplante and St-Cyr, 2001). The rationale of the model is to encourage ATO staff
to treat taxpayers as individuals who are embedded in specific contexts. This
perspective affects processes, whilst also making ATO staff think pre-emptively
about how to write and target their literatures, affecting products. As a result, it
is an 'outward-looking' innovation that cuts across organisational boundaries,
and has ethical and value implications for its practitioners and for taxpayers. This
makes it an unusual form of organisational innovation and one worthy of further
investigation.
Studying ATO Compliance Model Champions:
Epistemological Approaches to Change
How then is change or the effects of the introduction of the ATO Compliance
Model to be studied? The scope of conceptual frameworks to choose from is vast.
Possible perspectives include looking at the ways in which the types of
organisational structures and procedures in the ATO promote or stifle innovation
(Kanter, 1983; Bouwen and Fry, 1991; Harrisson et al., 2001); or examining the
salient characteristics of the ATO Compliance Model, to assess its 'adoptability'
(Rogers, 1995; Damapour and Gopalakrishnan, 2001). Also possible is studying
how factors external to the ATO promote innovation within the organisation
(Hollingsworth, 2000); the key practices or characteristics of champions (Rogers,
1995; Hooper and Potter, 1999); or different stages in the process of creating and
adopting the Compliance Model (King, 1990; Rogers, 1995). Thus, the innovation,
the organisation, the context, or the personalities involved are all potential
subjects of study, either in isolation or as cause-and-effect variables.
Yet taking a contrasting epistemological approach could potentially be highly
beneficial when asking questions about individual understanding and translation
of an abstract concept such as the ATO Compliance Model. Welford (1998) has
argued that organisational researchers are too quick to adopt the prevailing
epistemologies of contemporary management theory, thus limiting the questions
that can be asked about organisational change. Change processes and outcomes
can be studied using the tools of mathematical modelling (Visser, 2000),
psychology, and systems (King, 1990), to name but a few. Yet, it is important to
understand that underpinning the above body of research are implicit
assumptions about the functioning and structuring of organisations and
individuals. For example, organisations are normatively mapped out as structural
entities, which are:
Rational collectivities of social actors engaged in purposive goal-oriented activities within the domain of
a clearly defined boundary (Boden, 1994, p. 15).
This positivist approach could tell the following story of ATO Compliance
Model champions. The ATO Compliance Model appeals to certain ATO staffs
values and attitudes. These values are either changed or evoked, and are then
translated into altered behaviour, as long as personal or institutional barriers do
not stand in the way (see Fishbein, 1993; Terry, Gallois and McCamish, 1993).
Although this is a gross simplification of the positivist story, it is worthwhile
asking if there is another story to tell and whether the ATO Compliance Model,
being the type of innovation it is, benefits from the adoption of a different
epistemological approach. This paper takes the position that indeed there is
another approach worth taking as the Compliance Model is not a discrete or
bounded innovation whose effects can be tracked through indicators such as
increased output or product quality alone (although these factors could offer
some information on the impact of the Compliance Model on the ATO). Instead it
is an idea about ways of acting and thinking, whose translation requires turning
it into meaningful practices, set within the everyday contexts of the ATO. In this
sense, it is 'below the line' of most organisational change studies. This means it
requires treating champion's ideas, thoughts and experiences of the ATO
Compliance Model as instrumental to its enactment, not merely as incidental or
anecdotal to the 'real' changes taking place. It therefore warrants an approach
that places individual social practices as a central part of a conceptual framework.
Despite a slow uptake of this approach in the academy, critical social sciences can
be highly informative in unravelling organisational practices and change (see
Stubbs, 2000; Harrisson et al., 2001). Rather than casting organisations as static,
structural entities within which workers slot into immutable roles, organisations
are instead viewed as on-going processes, constantly re-created through norms,
values and expectations (Kanter, 1983; Hollingsworth, 2000). This position comes
under the broad framing of social constructionist or anti-reductionist approaches
(Sibeon, 1999), which posit that investigating how people think and behave
requires addressing values, beliefs and cultural assumptions, and how these all
relate to cultural meanings (Reser, 1995). These relate not only to the
organisational context of investigation but also to the broader social contexts of
the individual actor (see Edwards, Ashmore and Potter, 1995; Burningham and
Cooper, 1999).
One theory that is very useful in understanding this approach is Anthony
Giddens' structuration theory (1984). Structuration theory focuses on the on-
going processes of social life and their role in constituting patterns of ordering at
all levels of society. Giddens basically argues that recursive and routine actions (a
form of knowledge called practical consciousness) and awareness or knowledge
that is verbally or cognitively expressed (called discursive consciousness) form
the bedrock of social life. What we all know about how to 'go on' in everyday life
and what we can give discursive vent to are not incidental to how society is
structured and functions. Rather, they are the very processes that shape our 'here
and now', and future practices. This approach places human knowledge at the
centre of social processes. The theory postulates that through on-going practices,
the consequences of actions feed back into social reproduction across periods of
time, creating sets of social conditions that are beyond the intentions and remit of
particular actions.
To give an example, a tax officer may carry out sets of routine practices on a
day-to-day basis that are based on a great deal of 'hidden' knowledge (practical
consciousness). This may include sending out letters to taxpayers or entering data
onto a computer system. Taken-for-granted knowledge is expressed in these
actions, some of which may be verbally shared and recreated with other ATO
staff (discursive consciousness) through informal conversations and team
meetings. These practices have set intentions (such as to collect revenue and
promote compliance) but they also have other outcomes that are not directly
intended, perhaps affecting taxpayer's views of the ATO and institutionalising
certain practices through their constant use. The combination of intended and
unintended outcomes sets the scene for, and both constrains and enables, the
future conditions of action.
This framework is pertinent to asking questions about the role of the ATO
Compliance Model champions. How Giddens (1984) constructs social ordering
and the issues that the ATO Compliance Model aims to address are closely
related. They both place as central the need to understand what individuals know
and how they act, and the effects of these actions on others. This framework, or
epistemological outlook, will be used here to examine the champions'
understandings and rationality, to unravel the forms of knowledge they bring to
their work with the Compliance Model, and also to examine the implications of
the champions' actions for the ATO.
Research Methods
The contradictions this threw up in the interviews were interesting. In one, the
individual gave a thorough and multi-faceted explanation of their understanding
of the ATO Compliance Model:
So I see the model gives quite clear levels and responses and it lets you take into account individual's
personal circumstances, so people at the end of the day feel they are being properly treated. They have
been listened to, they have had an opportunity to debate the issue. At the end of the day they might feel
that they don't like the answer but they think they had a proper hearing... But on the other side too, I
think it's about understanding industries and individuals and the drivers within an industry to help us
know how to target our approaches and educational material, or enforcement activity or the balance of
those sorts of activities. So it's really about...an individual interacting with the system - treat them in a
professional, personal way, but also enable the ATO to understand the various business sectors.
Yet, later the same interviewee said, when asked what the CTSI can do to help the
ATO:
Well, I would think just helping understand it more deeply... maybe a refresher course or a refresher
discussion that would actually help. Because everyone gets very busy focused on outcomes and things
This contradiction could be interpreted as the champions being those who have
paid the model the loudest lip service. Or, maybe the ATO Compliance Model had
become such a part of their thinking, straddling the boundary of practical and
discursive consciousness, that the central tenets of the Compliance Model were
now part of their taken-for-granted knowledge. This latter reasoning is a feasible
conclusion to draw, and is discernible by taking a closer look at what the ATO
Compliance Model represented to champions.
Many interviewees stated that their initial reaction to the Compliance Model was
that it made intuitive common sense:
It all fell into place.
I just thought it was common sense. I liked the fact that it brought a lot of things together I suppose. That
here is a model, and you need to be able to place your client on that pyramid, but before you did that you
needed to understand what he's on about in that context of the BISEP.3 I thought it was sort of tidy, sort
of academic but at the same time it looked very practical to me.
Thus, on one level the Compliance Model appealed, as it brought together the
ideas and ideals that seem to have been part of these champions' practical
consciousness for a long time. It not only brought them to the awareness of
individuals, but also enabled connections to be made between often-scattered
ideas and bits of knowledge. This is an important process as:
Pointing out connections between previously isolated bits of people's assumptions can create both greater
awareness of those bits and new cognitive links among them (Strauss and Quinn, 1997, p. 40).
For some, there was almost a 'Eureka!' moment, where a sense of clarity was
gained about their own knowledge. However, the ATO Compliance Model is
more than just an idea that appealed to champions because it legitimised their
worldview. It also provided them with a new discursive tool that they hoped
would help bring about changes in the ATO.
Over a quarter of interviewees directly stated they saw the ATO Compliance
Model as a form of philosophy, a framework or a touchstone, which they could
use to inform their outlook and guide their practices:
At the end of the day the ATO Compliance Model is really a qualified philosophy, principles as to how
you want to behave.
The introduction of the Compliance Model into the ATO was timely as it fitted
neatly into ongoing discourses about reshaping the tax system and how it is
administered. This discourse, which is commonly defined as a form of argument
or 'group of utterances which seem to have a coherence and a force to them in
common' (Mills, 1998, p. 7), is directly related to the introduction of the
Taxpayers' Charter. The Compliance Model also resonated with the recently
released Public Service Code of Conduct (see Public Service and Merit Protection
Commission, 2000) and other tax reforms that had taken place within the ATO
over the previous few years (see Shover et al., Chapter 8, this volume). In line
with these initiatives, the Compliance Model embodied ideals that the ATO
should do more than collect revenue by concerning itself with building
community confidence and partnerships. Thus, the ATO Compliance Model was
very much:
In full alignment with the Taxpayers' Charter. We are clear that we are dealing with the individual
taxpayer, and the circumstances of taxpayers need to be examined and not just put all in one big group -
and say "everything is the same, everything applies to that group'.
The last two years the Compliance Model has been specifically mentioned in our corporate plan in
different contexts, and it's particularly compatible with the Taxpayers' Charter goals. In the early days
when we were communicating the Compliance Model to staff we had a lot of success linking it with the
Taxpayers' Charter and using it to support its goals.
Thus, the Compliance Model fitted the general direction in which ATO
managers and business planners wished to head. However, the discourse of the
Compliance Model was not only in-keeping with conceptual currents in the ATO.
It also provided new ways for the champions to think about and create 'space'
within the ATO, in which they could do things differently.
The ATO Compliance Model was a new way of talking and communicating
with others. For example, one interviewee called the Compliance Model a 'tool to
talk'. Another interviewee stated how it had given space, or 'permission' to those
who wanted to air views about how to do their jobs differently. Another saw it as
giving staff a 'bit of paper' to overtly show how they were taking the community
into account, in adopting an ethical stance in interactions with taxpayers. Thus, it
was not only what the model represented that was important, but also the
institutional space it afforded champions to rethink their practices and future
direction. Where once staffs cries of 'we should do things differently' may have
fallen on deaf ears, champions now had a hook to hang their beliefs on - 'the
Compliance Model says...' - which legitimised their positions and gave them the
green light to try different approaches:
We were certainly open to trying new things and looking for new ways of operating because we had
been doing the same thing for so long. It was really along some of the lines we've been heading or trying
to head anyway, but it probably just put some structure around that.
For those who had worked hard at trying to 'sell' the principles of the
Taxpayers' Charter within the ATO, the Compliance Model was a new bow to
their arrow:
And what the model did was gave me a way of telling stories.
It also presented the opportunity, if not the outcome, for the ATO to reach a
common 'lexicon of language', in terms of understanding and sharing common
principles and practices, in this way, the ATO Compliance Model can be framed
as a 'social representation', which is a form of shared knowledge that is produced
and communicated through social processes, moving from the abstract to
common sense or common dialect, through interaction (Kruse and Schwarz, 1992;
Billig, 1993; Moscovici, 2001). Although there is not the space to develop this
argument further here, Lalli's (1999) assertion that social representations
reconstruct common sense in local (everyday) situations and can therefore be
seen as the 'local operators of common sense' is an interesting one that merits
future consideration in relation to organisational innovation.
The Compliance Model also positioned champions and the ATO in broader
business discourses. Senior staff were only too aware of the trends occurring in
international management practices, and felt that the ATO Compliance Model
allowed them to take their place in this arena. Again, this is a sense of
legitimation, of being an international 'player', shown by the many references
made to how other countries were working with these issues and how using the
Compliance Model put the ATO 'ahead' in terms of being pro-active and forward-
thinking:
I can recall we had some visitors from the UK last year and I remember at the end of the week and a half
he said it was one of the best and most exciting things he'd seen, the Compliance Model. So he was
taking it back to the UK.
The fact that the Compliance Model is based on research from outside the ATO
made it exciting to some champions, and enabled them to feel they are part of
'something' beyond the boundaries of the ATO:
When I spoke to some staff and to others, you were able to say This is not just the ATO saying this, this
has come from worldwide research of regulatory practices!'.
This process appeared to allow staff to feel more comfortable with themselves
and their social role. Instead of being the dreaded tax collector, commonly
perceived as making unreasonable demands on honest people, they instead felt
that the ATO Compliance Model allowed them to help taxpayers and some have
experienced members of the community warm to them as they do. In sum, they
felt more liked and happier in their jobs:
I must say that it is far more pleasant work to do than actually go through someone like a dose of salts.
Examples were given in the interviews of how changing simple procedures had
made their jobs more tolerable, such as altering the tone of lodgment letters, as
referred to in the following quote:
Not only did we get better lodgement outcomes but, more significant when I think what we got, was that
when people rang up they were nice. They didn't ring up and abuse you for sending them a nasty letter,
they rang up and they were nice.
This enabled ATO staff to adopt discourses that may previously have seemed
incongruous coming from a tax officer. For example, interviewees talked about
bringing to their workplaces practices based on the concepts of human dignity,
openness and respect, in keeping with the tenets of the Compliance Model.
Adopting the Compliance Model also helped to reshape the public face of the
ATO, according to the champions. One interviewee talked of how working with
the Compliance Model required staff with different personality traits and
outlooks than the old command and control style of auditing. As a result:
I know there's a certain number of jobs that we have advertised in the last 12 months or so in the
preparation for the GST workload that have explicitly included within the selection requirements those
sorts of traits.
The above argument is not intended to portray the ATO Compliance Model as a
way of talking that was adopted by champions in a distanced, self-interested
sense, to simply get what they want. Rather, the adoption of the discourse of the
Compliance Model, for all its attendant benefits, was part of the process of
internalization. That is, there was a subtle but important distinction in the
Compliance Model interviews between those who thought about the Compliance
Model and those who thought with it.
As already discussed above, it appeared that those who had a deep
understanding of the ATO Compliance Model could not always remember much
about its abstract form. This did not denote a lack of comprehension, as
interviewees displayed an understanding of the Compliance Model that went
beyond merely repeating its contents, to grasping its complexity, ambiguity and
potential ramifications. This was because they viewed the model dynamically.
Indeed, it was applied dynamically and was adapted into various useable guises.
For example, one interviewee talked about having used the Compliance Model as
an educational tool to talk to taxpayers:
When I was doing real lime reviews, as part of an explanation to the taxpayer, I used to explain the
Compliance Model. I'd just draw a little triangle, fairly basic and I explained it from a soft angle to them.
Arid a lot more people after that explanation would accept the fact that even if they were audited there's
a reason behind it.
Most champions did not use the model in such an overt way, as they had
internalised it as 'another way of thinking'. For example, one interviewee
described how they used the ideas of the model on ATO staff, such as to inform
approaches to internal training. Indeed, some champions believed that it was only
when they had stopped making overt reference to the model, when it had been
digested and had become second nature, that they had truly understood it:
It's not a model that you can pick off the shelf sometimes and use, it's more just a part of the way we do
things. So I think we would consider it to be a success if we had all the elements of the model embedded
in our work practices without calling it a model any more.
This is because:
It's not a set of rules, it's a philosophy.
Thus, these champions were suggesting that the Compliance Model had become
embedded in their practices, making it 'a cultural thing', as one interviewee
referred to it. Or:
It's not so much the model or anything but it's just been accepted as a way we actually go about doing
things now.
This differed from those who thought about the Compliance Model. From this
second perspective, the Compliance Model triangle was used to locate individual
positions in the ATO, to understand where their work practices fitted in and
where taxpayers sat. This, however, caused no end of confusion, as it suggested a
static interpretation of the model. As the following quote suggests, when the
model was presented to some staff:
The understanding of the model was not always good and we did get a lot of people who thought what it
meant was you didn't respond appropriately to non-compliance. People tended to think statements like
"we treat people as honest unless we know otherwise' meant that even though the overwhelming
evidence was that they'd been diddling and telling fibs, we would have to accept them as honest. That, I
think, is an area where the subtleties of the model are not always well understood.
This distinction was well made by one interviewee. He compared staff who
had to use the model to plan business (and therefore had come to understand the
Model through 'playing' with it) with ground staff who had 'mapped' themselves
onto the model and used it to locate where they were in reference to others in the
ATO:
So at the rank and file level people map themselves on a model - which I suppose is a human tendency to
find where I fit. They don't see it as sort of dynamic - that if you fit here, that's where I live and that's
where i am staying and that there isn't a role upwards or downwards in it.
As the Compliance Model is shaped like a pyramid, some ATO staff appeared
to think it suggested all ATO compliance work should be located at the bottom, at
the 'soft' end of the Model, forsaking prosecution for education. In this way, they
saw the Compliance Model as providing the ATO with a:
Visual sense of proportion as to what the ATO wants and where we'd like to go.
In one sense this is true. The Compliance Model does encourage ATO staff to use
the tools of education and persuasion with taxpayers. However, this reading also
meant that many ATO staff did not see the model as applicable to them. Others
saw it as a threat to the work they did:
A lot of people see it as the Compliance Model moving from an audit environment to a help and
education environment and a lovey-dovey rose coloured glasses situation. But again, if you step back, it's
not that at all, if you really understand what it's all about.
This 'thinking about' reading of the model created a whole host of problems.
Because interviewees had tried to apply the principles of the Compliance Model
in a rigid way, not adapting it but sticking to its original abstract format, they had
great difficulty matching it up with their everyday realities within the ATO. In
this sense, these individuals (a few of the champions, but mostly reportage from
champions about others) talked about using the Compliance Model as if they
were laying siege to the ATO. The interviewees with this outlook talked as if they
were pushing an idea against great resistance, continually 'belting it over the
head' to try and get the message across. This kind of 'pioneer spirit', adventurous
as it may appear on the surface, was a rather impoverished translation of the
Compliance Model in comparison to those who had made it actionable through
the internalisation processes. The distinction between these two ways of seeing
the ATO Compliance Model is sketched out in Figure 7.1, which presents this
argument mapped onto a simplified version of the ATO Compliance Model (see
Chapter 1 for a more complete diagram of the Compliance Model).
The first figure on the left shows an interpretation of the ATO Compliance
Model that is one-dimensional, the 'I am here' or the 'taxpayer is here' reading of
the model and its behavioural implications. The second figure on the right shows
how champions who have internalised the ATO Compliance Model as a way of
thinking saw it as multi-dimensional. This was not about using the model to
position oneself in the ATO (that is, I am at the soft/hard end) but rather thinking
'inside' the model about how their practices affect outcomes, in reference to both
temporal and contextual factors.
To explain the right-hand side of Figure 7.2 farther, the implications of the
Compliance Model were seen as spanning a wide array of contexts, which affect
each other incrementally. This applied both to ATO staff practices, in terms of
actions and material, as well as to a wide array of social relations, both personal
and professional. One interviewee talked about how ideas embodied in the ATO
Compliance Model were used by some staff to think about how they might alter
senior manager behaviours in relation to their acceptance of the model. This
resulted in the formulation of a strategy for dealing with managers' outlooks,
which was based on the principles of the model. As another interviewee aptly
stated:
Figure 7.2 'Thinking about' and 'thinking with' the Compliance Model
We work inside the model. Now, if we interpret the model as one dimension, we could have a self-
fulfilling prophecy. We prosecute them, they get angry, they don't lodge. It's a catch-22, it's a circular
argument. So by looking at the compliance model a little bit more strategically, you can add that extra
dimension where at the moment it doesn't work.
This was about seeing how contexts or spaces are interconnected across time.
Interviewees who had been thinking with the Compliance Model saw it as
encouraging a middle-term perspective, in terms of the ramifications of ATO
practices. For example, one interviewee suggested that rather than interpreting
the model as suggesting the ATO should not prosecute as much, it was rather a
way of seeing the effects of prosecutions on future actions, not just on individual
taxpayers but also on Australian society as a whole. With this perspective, the
aim was not to reach a general consensus on what to do or how to enact the ATO
Compliance Model by being given step-by-step instructions. Rather it was about
being part of an on-going dialogue about where to go now. As one interviewee
commented:
I don't think you'll ever get consensus. I think genuine commitment is what I think is essential.
Where did these differences in outlook stem from, between those who thought
with and those who thought about the Model? One of the key factors creating
these differences were individual experiences of the model, both in terms of
attempts to implement it, as well as how their position in the ATO affected their
personal outlook.
Although the management and business lines of the champions in the ATO vary
in the way they saw the Compliance Model, they appeared to share some sense of
'vision' in terms of future directions of the ATO. In this way, the model appealed
as 'intuitive business common sense', as one interviewee put it, as a way to
increase compliance and thus increase revenue. Yet, this stance was not only
about roles in the ATO, but also about how work, and experiences in work,
affected lifeworlds. 'Lifeworld' here refers to the idea that understanding the
Compliance Model was not solely contingent upon narrow work-defined or
isolated cognitive processes. Rather, it was related to significant life-experiences,
which are given meanings through new forms of learning and experience (Finger,
1994). Interviewees' positions in the ATO did not just entail performing sets of
tasks and adopting a work persona in a detached manner (in comparison to
Goffman, 1969; Pruzan, 2001). Instead, their heterogeneous knowledge and
experiences (see Latour, 1993; Murdoch and Clark, 1994) contributed to how they
saw the Compliance Model. For example:
I think there's certainly a better appreciation of what the model is at senior management level than there
is further down. So, I suppose, the level of understanding is still cascading through the organisation. I
think there's been more education at the senior level because they are people who are planning our
business so they're getting to see the concepts of the model. Like education will teach you about the
model, but they actually have to do it and I find models hard until I play with them. I think a lot of the
senior management might be the same. When you start looking at business strategy through concepts
that are consistent with a model like that then you actually get to see how it can operate and what it's
about.
In contrast to this:
I mean we have an area called Special Investigations. We work on taskforces with the National Crime
Authority and the National Federal Police, basically auditing drug dealers, money launderers, people
involved in illegal occupations. So there you see the worst of the world. So that's more painful. With
some of our staff who have grown up doing tax audits - that's what they did for years, got rewarded for
doing tax audits more quickly and producing more revenue - they've found it quite hard.
Therefore, it was not only their experiences of the tax system but also experiences
of the ATO Compliance Model itself that made a difference to how it was
internalised and operationalised. As one interviewee aptly observed:
People have some resistance to the concept until such time as they personally experience it or see the
results.
The interviews supported this assertion time and time again. Some champions
may have intuitively reacted 'we already do that' when first seeing the model, but
others suggested that only through 'getting in there' had they come to understand
it:
I guess it's one of those things that intuitively you have difficulty disputing and we have had that
response right through when we have been publicising the model in the ATO. People are saying 'yes, we
already do that' and I suppose that demonstrates our superficial understanding of what it was really
about, and I suppose my initial views were that it wasn't that much different to what we were doing, but
the more I explored it and the more I sought to understand it, the more I realised how we weren't really
doing those things at all. We were imposing our own will on the market segments that we had created
rather than understanding their needs. And instead of looking at what they were doing we were looking
at what they should be doing.
Even those who were sceptical when first seeing the model had their opinions
altered when they had worked with the model in their everyday practices:
My first reaction was 'what's this shit?', not in terms of what the model was trying to say, it was just the
language, it was in academia. It wasn't until we started using it at the first conference, when we started
playing with some of the ideas in it, that I started to understand it. So I guess for a person like me the
worst thing to do was to get me to read something about it. The first thing to do is not even tell me that
there's a model. Just start doing some things.
Thus, how the Compliance Model was viewed was dependent upon how life
experiences in the ATO had shaped individual perceptions of both tax issues and
human nature. For example, champions often stated that tax systems are about
'doing the right thing', not only from the taxpayers' perspective but also from the
ATO's:
My philosophy is that a big percentage of the taxpayers will do the right thing if you just help them do
the right thing.
When asked about the role of the ATO in general, two-thirds of interviewees
stated that they believed it was about collecting revenue in a fair and reasonable
way, and enabling the government to carry out equitable practices of
redistribution:
Yes, we do have a huge social obligation and we have to make a social contribution, so the way we raise
revenue has to be a legitimate form - and we fit within social justice, redistribution of wealth.
Half of these respondents also added that the tax system encompassed issues of
community ownership, and confidence that the ATO would be fair, just, and
listen to an individual's concerns. Indeed, when asked to summarise the strengths
of the Compliance Model, just under half said it enabled them to better
understand the taxpayers' lives and contexts, enabling them to treat taxpayers
more personally, not only with the aim of increasing compliance but also because
that is 'how it should be'. This, therefore, was not a neutral or detached position
to take on the responsibilities of ATO officers. It was instead one bound up with
moral and ethical issues beyond the ambit of simply collecting revenue. The ideas
represented by the Compliance Model, being abstract, moral and instantly
recognisable, were parts of the champions' worldview. As one interviewee stated:
It just made sense of the world where I live.
Some interviewees even reported that they, or other staff members, had applied
the Compliance Model to their personal relationships with great success!
However, this is not to give the impression that all champions were in
agreement about the place that morals and values have in the ATO. Some felt that
it was essential that staff now bring their values to work with them, and think
beyond the office door. Others felt this was a dangerous move, as who judges
what is a worthy or right value? Some ATO staff felt justified in trying to 'get'
those rich and famous Australians who, in the public eye too, were not paying
their fair share of tax, although they were not doing anything illegal per se. This
was not a tax judgment but a moral one, seemingly based on principles of 'fair
play' that are believed to be enshrined in the tax system (see Braithwaite,
Reinhart, Mearns and Graham, 2001 for data on public perceptions of tax in
Australia). Ultimately it appeared that champions at least hoped the ATO
Compliance Model could lead the way through this ethical minefield. That is, by
bringing tax administration and procedural fairness together, through
encouraging staff to think about the implications of their practices not just from a
revenue perspective but also from a contextual one, the ethics of 'fair play' would
become a natural part of ATO behaviour. In short, they could persuade members
of the public to pay their tax, rather than having to force them to pay.
However, these changes were often reported by champions as not working 'on
the ground' or 'in the field'. Those working more closely with field staff suggested
that changes were not really happening on a day-to-day basis because:
It doesn't work within the public service delegated authority type regime very well.
Interviewees were clear that to get over these barriers and have the Compliance
Model fully accepted there needed to be a 'culture' change in the ATO. This was
not an easy task to undertake:
It takes time. You're stepping over the boundaries a little bit that currently exist and that is a real mind
shift, for that to occur.
You do see general support. There's probably reluctance though, in stretching the boundaries more so
than we've got at the moment. We need to break down some of those paradigms and say 'Well, what
does it mean for us in the new environment?', and there needs to be space for that thinking and there
also needs to be a level of support to take in that work.
Indeed, daunting it is. A parallel example can be made from the environmental
change literature. Some environmental researchers suggest that changing
behaviour without changing values is adequate (this is called Applied Behaviour
Analysis: see Geller, 1989). A second view is that technology and the processes of
widespread ecological modernisation have the answers (see Cohen, 1998). A third
approach argues that values of 'deep ecology' have to be nurtured before wide-
scale environmental improvements can be seen (see Pepper, 1996). The ATO faces
a similar issue but on a smaller scale. That is, how to create the widespread
acceptance of a new way of thinking and acting that is fundamentally moral? If
Giddens is used constructively, it can be suggested that changing practices is the
best way forward. As one interviewee suggested:
I think it's just about keeping it fairly simple to start with. Just starting with some fairly simple sorts of
changes. We ran some fairly small pilots that were able to actually demonstrate some changes that it
made. I guess it's about actually making it practical for staff. It just clicks with people. That one message.
If you started with that you've gone somewhere, you've gotten somewhere with it - and then if you come
back again and you talk about 'let's do something else', you just get that gradual change I suppose.
Thus, altering practices so that they embody the ATO Compliance Model ideas
seems to be the most practicable way forward. This assertion can be supported by
research the author carried out into individuals taking part in a sustainable
consumption behaviour change program in the UK, called Action at Home
(Hobson, 2001a, 2001b). Over a six month period, it was found that many
participants of Action at Home only changed a few small, seemingly mundane
practices, such as turning off the taps when brushing their teeth or switching
lights off when leaving the room. However, what was significant were the
changes in personal outlooks that resulted from participants thinking about the
social and environmental implications of their previous habits, and their
reconsiderations of the types of attitudes these previous practices embodied.
Issues such as water and energy use were brought from the practical to discursive
consciousness, making individuals ask themselves why they carried out certain
actions the way they did, and what the ramifications of their actions might be.
The resultant magnitude of behaviour change may not appear significant.
However, it seemed to leave behind a 'lens of difference'. This means that
participants were left with a residual awareness of the issues that Action at Home
discussed. They had not become converted to being 'green' but had internalised a
consideration for the environment that touched and subtly affected the way they
viewed the world and their own behaviour.
The same could be argued for the ATO. By altering small practices, ATO staff
can experience the Compliance Model in action, and can carry out the principles
of the Compliance Model without having it overtly enforced. This is a small but
important step forward in getting ATO staff to rethink their practices. It is not
about trying to create new values, but rather put into action the ideas of the ATO
Compliance Model through small changes to practices.
Examining briefly the processes through which the Compliance Model appears
to have made an impact in the ATO supports this assertion. For one, there are the
'above the line' examples, such as embedding it into ATO recruitment strategies.
Making the Compliance Model part of staffs everyday thinking is partially a
result of these official stances. For example, the need for strong leadership in the
ATO was stressed time and time again in reference to making the Compliance
Model work (see Bouwen and Fry, 1991; Hooper and Potter, 1999; Job and
Honaker, Chapter 6, this volume; Shover et al., Chapter 8, this volume). Yet, there
appear to be other processes involved - ones 'below the line' of official managerial
practice - that are equally important and hinge upon the ATO Compliance Model
being translated into everyday behaviours for staff, either by champions involved
in training or through the informal 'mentoring' of staff. As champions were able
to think with the model, they clearly saw that the translation process of the
model from the abstract into the everyday requires tapping into the knowledge
and language of ATO staff, not simply expecting them to jump 'cultures' as the
result of one presentation or meeting.
For example, one interviewee talked about how the language of the
Compliance Model had been changed to enable it to resonate with ATO staff:
Originally we used to have some delineations within the compliance pyramid that showed education
advice here and then audit here and people love to box things so they kind of saw it was a static kind of
model. So, when I introduced it to people the first thing you had to do is — when it had that language on
it — was take it all off. Let people use their own words and let them write their own words on the model
so that then they could be engaged.
Another champion told of how she used stories to translate the model into
reality. Through detailing to other staff her own personal experiences of the ATO,
in trying to correctly lodge her personal tax returns and feeling that she was
treated as guilty until proven innocent by the ATO, she was able to bring the
model to life for staff. In training sessions:
I read them the letter5 and they said 'Oh'. So practically I could say 'see, this is what's happening'. So 1
was able to use that story. That story worked wonders.
The same interviewee later explained why telling stories of actual experiences,
using ATO language and processes, works:
But when you actually say in their language 'well you know when you do blah blah blah procedure?
Dah, dah, dah happens'. As soon as you say that, you've got them hooked because they think 'She's one
of us!'.
Being able to make this initial connection then enabled staff to understand and
work with the Compliance Model, to get into the model rather than feeling they
had this rigid entity into which they must squeeze their work. As one interviewee
put it:
Instead of chipping away, these things could just be tike a snow-ball running down the hill, it would
really feed on itself.
There were signs that a snow-ball effect was happening, as many champions
were positive that the Compliance Model was making an impact on the ATO,
although there are still many barriers to overcome. The key point to make is that
changes were happening through staff using the modeL,, talking to others about
it, and seeing its positive benefits for themselves:
When I first went around the country to talk to my teams, I got the same response - 'It's just a waste of
time. All taxpayers are cheats'. But fortunately when team members actually have gone out and applied
the concept or used this strategy, they personally see the result that comes of it. Therefore, there's a
change in mind-set, to be more accepting of the concept.
These can be seen as processes of social learning and social diffusion. Rogers
(1995) argues that an individual's adoption of innovations stems not just from
reading about an innovation but from listening to and witnessing how other
individuals in their social network have evaluated the innovation (see Flinn, 1997
for further discussion of social learning). Talk between ATO staff and others, be it
informal or formal, is important as change is achieved through communication
and negotiation, as well as through simply watching others. This is social
learning. The adoption of new ideas can rarely be imposed from the outside
(Harrisson et al., 2001) but instead takes place along 'learning-action' networks
(Clarke and Roome, 1999), which cut across formal organisational hierarchies and
boundaries. This is social diffusion. Thus, bringing the ATO Compliance Model to
life is not about individuals singularly adopting the Compliance Model internally
and then putting it into action in tandem with others. It is instead a process of
'joint action' (Shotter, 1993a, 1993b), facilitated through different processes of
learning.
De Young's (1993) typology of behaviour change processes makes this point. He
suggests that both tangible and intangible processes of behaviour change exist.
Examples of tangible forms are feedback, material incentives and legal mandates,
whilst intangible forms can be direct experience, personal insight, self-monitored
feedback, and a sense of confidence and commitment. The point to be taken from
this is that adoption of the Compliance Model in the ATO should not be about
employing just one of these forms of change. Instead, it requires a mixture of
leadership, support, information, policy reworking, personal experience,
commitment to change, and broad thinking. The Compliance Model champions
appear to embody and encompass this broad array of behaviour change
processes, enabling them to think with the model, and offering interesting
insights into how change might be forwarded in organisations such as the ATO.
Conclusion
This paper has sought to address why certain individuals in the ATO can be
considered Compliance Model champions. It has attempted to offer an alternate
story from positivist approaches, by examining how these individuals talked
about the ATO Compliance Model and its place in their lifeworld. Through this
'below the line' approach the themes of experience, social learning and
interaction have emerged as positive aspects of change. Yet the departing message
of this research should not be read as advocating all tax staff adopt 'touchy-feely'
approaches, as several interviewees put it. The ATO has to administer tax laws
across a large and diverse country, a task that requires procedures and systems
able to cope with vast quantities of information and assessments. The messages
are rather that the lifeworld of many tax officers is such that the Compliance
Model would not have immediate resonance in relation to their everyday
experiences. Thus, the process of translation is important. Often this translation is
believed to involve either changing the wholesale 'culture' of the ATO or turning
the Compliance Model into a step-by-step rulebook. This research advocates a
more middle-ground approach from the lessons the champions have to offer, by
examining how small practices embody the central tenets of the ATO Compliance
Model. As a result, it encourages the taking of a gradual, iterative approach to
change.
This research can also reflect on implications for how taxpayers are treated. In
the spirit of reflexivity, it is hoped that organisations like the ATO might
conclude that their own staffs rejection or disengagement from the abstract form
of the Compliance Model might give some insight into how the taxpayer
struggles with tax returns. Do tax officers speak to the 'world where they
[taxpayers] live'? Why should the taxpayer comply with a set of regulations or
practices that do not take into account the constraints and lack of adequate
infrastructure in their everyday lives, when some tax officers feel exactly the
same way about the Compliance Model? Funnily enough, this argument brings
this research full circle, back to the Model itself. The fact that this research
supports a 'below the line' approach to encouraging changes in practice can be
seen as offering support to the central tenets of the Compliance Model. That is,
by understanding the consequences of one's actions and the messages embodied
in practices, we can begin to understand and positively affect others' practices
and bring them into a more supportive and compliant environment, whether we
are talking about tax office staff or the taxpaying community at large.
Notes
1 For example, the ATO seeks to maximise revenue through efficient operational structures, similar to
private industry. It differs in its role of administering tax law and being directly implicated with
issues of procedural and distributive justice.
2 Interviews were carried out by members of the CTSI and the Regulatory Institutions Network,
Australian National University. Interviewers were Brenda Morrison, Kristina Murphy and Declan
Roche.
3 BISEP represented the idea that to understand the compliance of an individual or group, one needed to
understand the lifespace in which they operated. The initials of this acronym stand for B = Business
profile, I = Industry factors, S = Sociological factors, E = Economic factors and P = Psychological
factors.
4 For example, several interviewees made reference to the way current information technology systems
did not allow them to access information required for a 'complete picture' of individual taxpayer's
circumstances and history. Yet understanding history and circumstances is at the heart of
implementation of the ATO Compliance Model.
5 This letter was received by the interviewee from the ATO after she had failed to lodge correctly. The
interviewee felt the tone of the letter made the assumption that she was trying to cheat on her tax
payments.
References
This research commenced some eighteen months after the Compliance Model
was adopted by the ATO leadership and was added to ongoing initiatives aimed
at the cash economy. It was premised on awareness that reformers cannot assume
that their intentions and plans will meet with committed and faithful
implementation. The reasons for this are several, but it is well recognised that
bureaucracies charged with implementing reform policies can show highly varied
capacity for the job at hand:
Policy ideas that sound great in theory often fail under conditions of actual field implementation. The
implementation process has a life of its own. It is acted out through large and inflexible administrative
systems and is distorted by bureaucratic interests. Policies that emerge in practice can diverge, even
substantially, from policies as designed and adopted (Bardach, 2000, p. 25).
Doubtless it was recognition of this that led the ATO's Cash Economy Task Force
to emphasise as constraints on the implementation and impact of tax reform
'attributes of the tax regime,' including both 'its administration, and ATO cultural
factors' (Cash Economy Task Force, 1997, p. 6).
Because the Compliance Model posits complex and slow-developing long-term
relationships between the behaviour of enforcement officials and the responses of
taxpayers, it was recognised at the outset that an adequate evaluation of the
CE/B&C project would require collection of data over several years. At inception,
our objective was a retrospective process analysis. Process analysis is used in
evaluation research as a way of understanding how reforms are implemented and
brought to bear in the day-to-day work routines of managers and employees
(Weiss, 1998). It can lay bare how organisational realities and dynamics constrain
reforms adopted in other circumstances for reasons often unknown to or
unappreciated by those charged with implementing the policy. Typically, process
analysis includes use of interviews and observation as a way of interpreting the
statistically observable consequences of a new policy or program.
We planned to collect a variety of data on the ATO's implementation of the
Compliance Model as well as evidence of some of its short-term impacts. We
assumed that by using a variety of methods and collecting a variety of data,
findings gained from multiple methods or sources to some extent would
compensate for their individual shortcomings. Twenty months after its
inauguration, however, the impending arrival of fundamental tax reform (the
Goods and Services Tax, or GST) caused a substantial reduction in personnel and
resources committed to the CE/B&C project. This also caused us to scale back our
original research plans and to shift our focus somewhat to examination of the
ATO's capacity for effective implementation of the Compliance Model.
Data were collected over a ten month period (August 1999 through May 2000).
We began by interviewing fifteen ATO managers or employees, some face-to-face
and others by long-distance telephone. These conversations, which averaged
approximately 25 minutes in length, helped familiarise the senior author with the
ATO and provided an opportunity to talk with several managers who played
important roles in the field-level implementation process. This was followed in
November-December 1999 by visits to five ATO field offices where significant
CE/B&C project initiatives were underway (Sydney CBD, Parramatta, Penrith,
Box Hill, and Dandenong). We interviewed 26 field-level managers and
operatives about their experiences with and opinions of the Compliance Model in
building and construction. The semi-structured interviews followed a topical
guide and averaged approximately 35 minutes each. All were tape recorded and
transcribed for subsequent analysis. Following these interviews, in May-August
2000, a survey instrument was emailed to all ATO personnel who at any time
were assigned to the CE/B&C project. After two follow-up reminders to the 260
persons who were sent surveys, 32 per cent were returned to us.2
During May-June 2000, we interviewed the owners of 27 small building and
construction firms in Melbourne and Adelaide. The semi-structured interviews,
which were tape recorded for subsequent transcription and analysis, averaged
approximately 30 minutes and explored respondents' experiences with and beliefs
about the ATO and its operations. The interviews focused primarily on
respondents' contacts with the ATO in the years immediately before and after
introduction of the Compliance Model in 1997. Although we had planned to
interview equal numbers of firms/owners with varying degrees of contact with
the ATO in the preceding two years, a variety of difficulties left us unable to do
so. Among them, we were unsuccessful in obtaining telephone numbers for many
firms listed in ATO files, and a substantial majority of those we did talk with
declined our request for an interview. We estimate that not more than five per
cent of all firms that were contacted agreed to and completed an interview. As a
result, the sample of businesses almost certainly is tilted toward those who
comply with tax requirements and have no particular reason, therefore, to be
wary of sharing their experiences with and opinions of the ATO. As one of our
subjects told us:
I think if a company is fair dinkum and up-front, they shouldn't be worried about the ATO, you know. If
you're doing the right thing you've got nothing to worry about.
In the language and logic of the Compliance Model (see Figure 1.1 in Chapter 1),
these firms and their owners are located near the bottom of the enforcement
pyramid, and our findings are limited to the population they represent.
In addition to interviews with ATO personnel and the owners of small building
and construction firms, we also interviewed an executive of the Department of
Public Prosecutions and a representative of the Master Builders Association of
Victoria. These interviews also were tape recorded. We recognise that the short
follow-up period we employed and the limitations of our samples substantially
diminish confidence in our research findings and the degree to which they can be
generalised to other times and locales.
Enforcement Technology
Research Findings
To what extent were our assumptions about organisational capacity for change
matched by conditions in the CE/B&C project during the first few years of its
operation? Did it have or acquire the resources to suggest that over a longer
period of time the expected results of responsive regulation would be realised?
Doubtless few would dispute the assumption that committed and supportive
executives are critical to an organisation's capacity for faithful implementation of
reform policies. One way this can be demonstrated is by providing the level of
fiscal and staffing resources required. Ample evidence suggests that strong and
ongoing support was given by senior ATO managers. This support came from the
highest levels of the ATO, particularly from the Commissioner, the Deputy
Commissioner of the Small Business area, the Assistant Commissioner
responsible for managing the cash economy initiative, and the Project Leader for
Building and Construction. They sponsored and provided resources for training
in the Compliance Model for all staff on the CE/B&C project, and they visited
and provided encouragement. When the initial training program was begun,
senior ATO managers signalled their support by attending and participating in
the training course. They discussed the direction the ATO was moving in and
how the Compliance Model complemented it. They emphasised that ATO
practices and procedures were important issues in building and maintaining
community confidence in the ATO and its ability to administer the taxation
system. None of the personnel we interviewed faulted ATO executives for failure
to support or to provide sufficient resources for the CE/B&C project.
Survey findings were not strongly one sided, however. Thirty-five per cent of
our survey respondents indicated that at some time they were responsible for
supervising other ATO personnel as part of the CE/B&C project. They ranged
from team leaders to mid-level managers. All survey respondents, former
supervisors and operatives alike, were asked to indicate their agreement or
disagreement with the statement: 'The leadership of the Australian Taxation
Office failed to devote sufficient resources to implementation of the Compliance
Model'. For supervisors, 33.3 per cent of respondents agreed or agreed strongly
whilst 40.7 per cent disagreed or disagreed strongly. For respondents who were
not supervisors, the results are 34.6 per cent and 36.5 per cent respectively. Taken
together, our interview and survey findings suggest that proportionately few
supervisors and operatives perceive a failure by ATO executives to provide
adequate fiscal resources and personnel for the CE/B&C project.
Here and in other ways as well, support for the Compliance Model by ATO
executives was not lost on field-level staff. When asked for his opinion about the
Compliance Model, one ATO operative spoke with apparent conviction about the
importance of treating taxpayers with 'understanding', 'respect' and 'compassion'.
As the interview progressed, however, and he was asked to reconcile seemingly
contradictory statements, he said that his earlier comments reflected belief that
ATO executives are committed to the Compliance Model, that they expect field-
level staff to get on board, and that they would be watchful for signs of weak
commitment and follow through. Only then did he acknowledge that his earlier
characterisation of the model masked considerable doubt about its efficacy at the
coalface.
This is not to say the Compliance Model was welcomed by all or even by most
staff. Some saw it as a hopelessly abstract approach that was developed by
academics and then adopted by ATO executives:
Let's face it: the Compliance Model was an academic model adapted to the tax situation. It was
developed by,... I think, people outside of [ATO]. ATO saw it as the way we should be moving,...and I'm
not disagreeing. I think they're right. But it was done at the more senior level by people who are inclined
that way [toward consultative approaches] anyway.
For most field-level personnel of the ATO, adoption of the Compliance Model for
the small building and construction industry apparently came out of the blue.4
For those we interviewed, there was limited awareness of how the decision was
made combined with some disappointment and cynicism:
I'm an employee of the Taxation Office, and this is an instruction that's put to me. I've no input into its
adoption whatsoever. I'm not asked whether, you know, this is a good idea. Should we go this way, or
whatever? You're an experienced taxation officer. You've been in this place for [more than 25] years,
what have you found in that period of time about, you know, compliance in taxation?
This was imposed certainly. No two ways about it. We were given two days to provide feedback... It was
like, okay, 'we've made a decision, but we should get something from the people in the office to make
them think that they're having some feedback into it'. [W]e didn't feel we were part of it at all...[T]he
decision was made before we even were given a chance to offer feedback. I think they ended up - so
many people complained they gave us another week to give feedback, but it was still basically we felt
that they just threw it away or put it in a corner and basically came out with the model they wanted.
For these reasons, perhaps, a mid-level manager who helped implement the
Compliance Model told us that:
It wasn't universally welcomed in the ATO, or in the project [CE/B&C project] anyway, which is
probably representative of the ATO. I think the majority of people in the ATO thought straight off 'oh,
yeah, now we're going soft on everyone', you know. I think that was the quick interpretation of it, that
'oh, now we just pat them on the head and tell them to be good in the future and walk away', you know.
The respondent noted further that apparently little thought was given to 'the
ability of people' (staff) to shift to the new approach, adding that 'it's a huge jump
for some'.
Whatever their initial reactions, however, successful implementation of a
reform is enhanced if staff take on as their own its fundamental assumptions.
Consequently, in our survey we asked current and former CE/B&C project
personnel to comment retrospectively on their expectations about the ATO
Compliance Model at introduction and after they gained experience using it.
Responses to this series of questions suggest that the subjects fall into three
groups. Nearly ten per cent (9.8%) indicated they were very sceptical about the
Compliance Model when they were introduced to it whilst 4.7 per cent were
equally enthusiastic about it. The remaining 85.6 per cent of respondents ranged
from somewhat sceptical to somewhat enthusiastic. Although few of those who
held strong opinions about the Compliance Model at introduction changed their
assessment with experience, for those in this large group with less extreme
opinions a majority (54.1%) became more positive about it. Our interview data are
consistent with this pattern.
Q: So, you have had two years experience now with the Compliance Model?
A: Yes.
A: I like it. I like it. It has a greater impact, because previously you were only dealing, you know, [one-
on-one] and had no impact on the industry. But now. bringing in the various associations and what have
you and getting the publicity, it has a wider impact. So it is more one-to-many, which I like. A lot more
flexibility, a lot more choice, which I like.
[O]riginally... it was seen as 'oh, this is going to be Mickey-Mouse type of activity that they're going into.
When will they get back to doing actual work?'. And things like that. After a while of going through it,
they could see the advantages of doing some of the alternative type of work as well as the audit
activities. But I think it was only when they could see that there was room for the two.
Growing field-level support for the Compliance Model does not extend to an
unqualified endorsement of it; staff beliefs about and experience with tax
enforcement have left many ambivalent about the long-term consequences of
responsive regulation. On one hand, many prefer real time reviews to audits.
Generally the latter are triggered by suspicion or evidence that something is
amiss with the tax behaviour of a business. The resulting detailed examination of
financial records can be both adversarial and confrontational. They can be
difficult both for operatives and for the targets of their efforts. Several of the ATO
operatives we interviewed commented spontaneously on their dislike for the
audit process, which puts a premium on being able to act assertively. For
example:
While I was an auditor I hated it. When I was doing my degree I always wanted to be an auditor, and it's
like, you know, you get there and you think, 'this is horrible'. So I didn't like it. I didn't like the aggression
that came along with auditing. It was more than assertive, it was probably aggressive...And it was like,
let's see what we can get, how we can intimidate this person, that sort of thing...So I really learned to
hate it a lot. When the Building and Construction Project came along, and we saw the methodology was
real time reviews, we didn't know what it meant. But when we saw that it was a lot softer than an audit,
a lot faster, and less intimidatory, and probably more helpful, I liked it. I preferred that way of doing
things.
On the other hand, former CE/B&C project staff members tend to be rather
cynical about the sources of tax compliance. Asked, for example, whether 'fear of
penalties is more important than conscience and the obligations of citizenship' as
reasons for tax compliance, 71.8 per cent of our survey respondents agreed or
agreed strongly. Clearly, they believe that a strong threat is fundamental to the
compliance assurance process. Not surprising, therefore, one of their most
frequently mentioned concerns was whether the ATO would maintain the level
of prosecutions needed to make it work. Put differently, they wonder whether the
'benign big gun' will be employed sufficiently often to maintain an acceptable
level of tax compliance.
The enforcement technology at the heart of the Compliance Model does not
end with real time reviews; responsive regulation highlights the importance of
involving community and industry groups in efforts to maintain or improve
overall tax compliance. This means that ATO staff now are expected also to
contact and work with these stakeholders in order to improve awareness of and
support for the tax system. This requires work routines and personal skills
different from audits and penalties, the traditional mainstays of tax enforcement.
Early on in the CE/B&C project, the Project Leader visited the senior
management of the Master Builders' Association and the Housing Industry
Association in Canberra. He explained that the project would be using the
principles of the Compliance Model, and outlined the benefits its use would bring
to the ATO and to the building and construction industry. Similar meetings were
held with the management of comparable associations around Australia. The new
efforts bore fruit; the Master Builders' Association announced its support of the
work the ATO was doing in the industry and its willingness to work with the
ATO to ensure cash economy practices were ended (Australian Taxation Office,
1999c).5 It is premature to examine the effects of these complex and indirect
effects of responsive regulation, but field-level ATO personnel who responded to
our survey were notably positive in their assessments of how well this process
proceeded and the payoffs from it. Specifically, they were asked to indicate their
degree of agreement or disagreement with the statement: 'During the time I
worked in the Cash Economy Building and Construction Project, I witnessed
substantially improved trust and cooperation between trade associations and the
Australian Taxation Office'. A majority of respondents (60%) agreed or agreed
strongly whilst 26.5 per cent were undecided, and the remainder disagreed or
disagreed strongly.
Whilst the broad principles of the Compliance Model are endorsed by subjects,
this begs the question whether their increasing support for it in the abstract
reflects detailed understanding of its principles and whether they in fact changed
their work routines accordingly. Concern about these matters is heightened by
comments we repeatedly heard to wit: it is 'just common sense' or 'it's what we've
been doing all along'. This suggests that the more arcane aspects of the
Compliance Model may have been missed by a substantial proportion of those
assigned to the CE/B&C project. A team leader told us:
[I]t's really interesting. When we did our one-to-one session, and even afterwards, one of my team
members said to me [that] he wished he knew about this Compliance Model previously. Because it's not
just a tool you can use in your work environment, it's a tool you can use in real life, you know, in your
relationships with other people. I mean you give people a chance before you go heavy with them. It's
what we've always done but all you've done now is put a name to it. You call it a Compliance Model. It's
what we always used to do.
The subject is not alone in his perceptions and anxiety. Responding to twin
imperatives of the push for improved efficiency provided by rapid technological
change and new tax policy, the result is a marked increase in workload and
efficiency by ATO personnel (Commissioner of Taxation, 2000, pp. 13-14). Along
with outsourcing of work formerly carried out by government employees,
hoteling or 'hot desking' has changed forever the nature of operatives' work and
their interaction with supervisors and headquarters; a great deal of this now takes
place electronically. For many front-line employees, one of the principal
consequences is a sense of mistrust and anxiety to which the rapid approach of
the GST was a major contributor.
Evaluation
The ATO evaluated its cash economy projects and reported evidence that things
are working (Australian Taxation Office, 1999c, 2000). Evidence of two kinds is
cited in support of this conclusion: gains in income and reported taxes by
building and construction firms, and the results of an on-going series of surveys
in which citizens are asked their perceptions of the ATO and its operations
(Commissioner of Taxation, 2000, pp. 35-36). Both are reasonable and appropriate
measures of program impact, the former because use of the Compliance Model
clearly is meant to increase taxpaying and the latter because the new ATO, in
time, should be seen by citizens as a more helpful and less threatening arm of
government.
No data or evidence used or cited to this point as evidence of success can stand
cursory methodological scrutiny, however. Although reported income is up, this
increase must be seen against the backdrop of economic improvement and the
upward trend in tax revenues. 'For most of the past decade, total and company
tax collections have grown at a rate greater than GDP' (Commissioner of
Taxation, 2000, p. 9). Since December 1996, the ATO has conducted biannual
surveys to measure community perceptions of the organisation. Whilst citizen
perceptions of some aspects of ATO operations have improved modestly over the
past three years, 'the results to date have remained fairly constant'
(Commissioner of Taxation, 2000, p. 49). Studies of samples drawn from the
building and construction industry have been more promising, but failure to
examine the possible influence of other factors limits confidence in these
findings. The short period of time (three years) is not sufficient to warrant more
than cautious optimism.
Underlying our interviews with the owners of small building and construction
firms is the assumption that if the Compliance Model is on the right track, there
should be evidence that the word is getting out to those at whom the policy is
aimed and evidence also that they are changing their perspectives. Specifically,
we reasoned that if any short-term impacts of the CE/B&C project were to be
apparent after less than two years duration, it should be present in the
experiences and opinions of business owners who had direct contact with the
ATO during this time.
Before moving to an examination of this issue, it is noteworthy that very few
of those we interviewed expressed opinions of unfairness about the taxation
system generally. One respondent, however, told us:
[T]he thing that you'll find in the building industry is that they are particularly hostile to the taxation
department because they sit there and they watch multinational companies paying no bloody tax at all
[whilst] they are being screwed into the ground. That's where the hostility comes from. That's where a
lot of the cash economy comes from, because there...is resentment...The rates are too high, [and]
multinationals will not pay their taxes. And not only multinationals, there are others too. But generally
the wealthy don't pay their bloody taxes, and that's where the resentment comes from.
Q: What ...could ATO do to soften this perception of unfairness in the taxation system?
A: Charge multinational companies exactly the same rate of tax that we pay, and don't let them write off
losses that everybody knows are fake. Everybody knows that these losses are paper losses. The money's
being shipped out overseas. We know that, the Government knows that...But they never touch them. It is
time - all over the world if the multinationals could be pulled into line because if they pay their share,
everyone else will have to - we'll all of a sudden find that everything's down to a reasonable rate. I mean,
these people deal in billions of dollars not in, you know, $50,000 a year or something like. They deal in
billions of dollars.
A: Grossly unfair. It's to the point of nearly being highway robbery. The middle and lower income
earners are paying for everything. The rich at the top they don't. They write it off on schemes. They can
afford fancy accountants and all the rest of it. The rest of us can't... It's ridiculous.
This subject was virtually alone in his charge that the taxation system is
distributively unfair.
Most business owners were less critical of the taxation system overall and
directed their criticism at the ATO specifically. They were mistrustful and cynical
about it and its policies. A substantial proportion believe that little good and
potentially a great deal of trouble and lost production time can come from
contacts with the ATO. One remarked.
Q: How often does ATO come up - I realise these are unusual times because of GST coming in, right - but
if you can go back to six months ago, before everyone started to worry about GST, how often does ATO
come up in conversation on the job?
A: Hardly ever...
Q: When it does come up, what is the general tenor of comments about the ATO?
A: Fear. ...Probably it's the fear of the unknown, you know. We all reckon that public servants are
arrogant pricks anyway that really don't know what they're doing, and the only reason they're there is
because they couldn't make it in the real world. . .That's the sort of attitude because when - if you talk to
anybody that's met them, that's how they come over. It's like a policeman standing there...and everybody
fears a policeman, don't they? Doesn't matter whether you're not doing anything or whether you were,
you know...[Y]ou think, 'what have I done wrong?'. And you get the same sort of thing when they
mention the ATO or the tax department. 'What in the hell have I done wrong now?'. And that's the
reason they're not there to help you, they're there to nail you.
Another and unusually clear representation of this was given by a plumber who
was interviewed along with his business partner and mate on their job site:
I mean, it's all too hard: that something that should be so simple is made so hard. On top of going to
work and making a living, we're expected to work all our tax out and everything else out, and it's that
bloody complicated. Chartered accountants can't work it out who are trained to do that. And yet they
want us to do it. I mean, like, how many of those guys can come and put up gutters and dig trenches and
lay plumbing? Fucking none of them! But they expect us to be able to do it - to keep our books in - in
absolute perfect order.
The Tax Office is very inflexible when it comes to listening to any reasons why - it's not like dealing
with anybody else. I mean, it's basically, "You were due to pay that on the 21st'. And this one young
lady,.. I spoke to her and I said, 'Look, you know, can I have, you know, two weeks [to meet tax
obligations]?'. (And I had never defaulted on my taxes.) She said, 'No, I've got a record here that you paid
late the last two months'. And I said, 'Yes, but I have paid, and I am finding it really hard at the moment'.
She said, 'No, you will pay like immediately or I'll start fining you tomorrow, and you're going to get' - I
can't remember what she said - some sort of record as a problem payer...And I thought she was a very
hard woman...So...yes, I lost it. I couldn't even talk to her. I hung up. Anyway, she rang me back a couple
of days later, and she was a little bit nicer. But...- it had got to the point where [our business] was really
struggling.
Q: What is your impression of how small business owners like yourself generally see the ATO?
Q: Yes. Does this ever come up in conversation between you and your mates?
A: Not really, no. The consensus is everybody hates it, you know. The ATO is there to rob us of any
chance of making any profit for anything, and that's it, and I think that's the consensus with most people.
Another subject commented simply that 'the public perceives them to be just a
bully'.
Wariness and cynicism were not universal, however; a minority of the
respondents noted that the ATO merely carries out policies enacted and modified
periodically by 'politicians' (i.e., Parliament). Consequently, a subject told us: '[I]f
you need any advice, then who better to go to than the ATO - apart from your
own accountant. [If] there are some little things that you might want to
know...the ATO will be able to tell vou. So it's best to go there'.
One of the principal objectives of the Compliance Model is improved business
perception of the ATO and a more cooperative culture among business owners.
No one expects this will be accomplished quickly, but the Model rests on an
assumption that over the long haul, its use will produce effects of this type. Our
interviews focussed principally on subjects' experiences with and perceptions of
the ATO over the past five years. Few of the building and construction owners
were aware that the ATO had launched the Compliance Model, but a much larger
number reported they had personal contacts with the ATO during this time. One
of the most striking characteristics of the responses we received was the large
proportion who spoke positively of the way they were treated in these contacts. A
man whose business records were audited by ATO staff some ten years ago
commented on the contrast between that experience and a real time review
conducted by ATO approximately one year before our conversation:
Q: You mentioned a few minutes ago that, historically, the building and construction industry is one
where - you didn't say it this way, but I think this is what you meant - people are suspicious of, a bit
wary, of the Australian Taxation Office?
A: Years ago, yes...Up until the last few years, you know. I must admit the first time I ever got audited
that they treated me like a common criminal, you know, until you're proven innocent. And consequently
I had a shocking attitude to the taxation department because I just thought all they were trying to do is
put you out of business...I must admit I had a shocking attitude towards them...But I think that the people
that have contact with them now understand that they've got their job to do and that they are trying to
be user-friendly as much as possible... I must admit, my attitude's been changed completely.
As impressive as these comments are, they and others like them are subject to a
host of interpretations other than the tempting one that business owners are
changing their perceptions as a result of ATO's use of the Compliance Model. The
respondent was asked, for example:
Q: Do you think getting a little bit older also has anything to do with this?
Q: How would you interpret that? A; I just think that when you're younger you're a little bit more gung
ho sort of thing... but 1 think as you get older you start to become a little bit wiser and start to
understand, you know, they've [ATO] got their job to do and, consequently, we need to pay taxes to have
all the benefits that everybody needs. You do have that [attitude]...when you're young. And that's when
most companies are starting - they just want to go full on trying to get their businesses up and running,
and if anybody comes in - it doesn't matter if it's the taxation department or anybody else - to interrupt
what they're trying to achieve then they would try to dismiss them.. And I think most people that run
their businesses are inclined that way.. They just don't like to be pulled up . .They just want to go for it.
Our cautious interpretation of what we were told, however, should not obscure
the fact that this respondent was not alone in noting that he has seen a change in
the ATO. Responding to a question, another business owner commented:
Q: Has this [most recent] contact [with ATO] in any way changed the way you think about them?
A: Absolutely. Absolutely. Like I said...it was, first of all it was nice to see that it was actually a human
being that worked for the ATO. Because as far as I always thought, they weren't humans anyway. Yes, it
was nice to have somebody you could look at and actually speak to, and, yes, he was a nice guy.
Q: But when we began talking, you used some of the words that [others] have used to
A: That's right. Yes. And that was from speaking over the phone [with them]. And.. . when I had this
meeting with them some time ago, that was exactly the way that I'd always thought of them. Especially
from that initial time [several years previously]. Yes, my perception definitely changed after having this
guy here a couple of weeks ago. Like I said, he was nice, he was understanding. He explained, you know,
what we wanted to know... It was good to be one-on-one and see one of these guys.
It is well to recall again that the sample of firm owners we interviewed is made
up almost entirely of those with a record of compliance. This should temper our
readiness to interpret these findings as indicative of success for the ATO
Compliance Model. They are strongly suggestive, however.
Lessons
This research began with two overarching questions: (a) to what extent did the
ATO have or acquire during the initial two years of operation of its CE/B&C
project capacity to implement the ATO Compliance Model with a reasonable
chance of success; and (b) what can be learned from the CE/B&C project
initiative that could be useful now that the Compliance Model has been extended
throughout the Australian Taxation Office?
Regarding the former, there is strong evidence that ATO executives supported
the Compliance Model and provided adequate resources for its trial
implementation. There is evidence also, albeit considerably weaker, that field-
level staff gradually warmed to the approach although they continue to harbour
ambivalence about it and reservations about its long-term impact. The
Compliance Model was adopted by the ATO during a time of considerable
external change and uncertainty, and it can be extremely difficult to isolate and
determine the effects of new policies under conditions such as these. The ATO
was powerless to maintain for the CE/B&C project an environment in which it
could function smoothly and reflectively. The problems, however, are not
confined to the absence of stable environmental conditions; the Compliance
Model was adopted at a time of change and uncertainty within the ATO as well.
Management initiatives to increase employee productivity coupled with changes
in employee work routines contributed to pervasive uncertainty and stress.
For reasons presumably diverse and complex, some conditions important for
achievement of reform initiatives did not materialise, however. Plans for
evaluating the CE/B&C project were not built in and were hampered in any case
by shortcomings in routine data collection and management capabilities. The
ATO's Case Management Reports and Analysis system (CaMRA) is a case in
point. CaMRA was designed for reporting in the area of withholding tax, and
only later was modified for use in other areas of taxation. The modifications,
however, were stop gap in nature, and the finished product was unsatisfactory.
When applied to the CE/B&C project, the resulting system of data and case
management was of limited utility both for ATO personnel and for those who
would use it to examine the implementation and effectiveness of the ATO
Compliance Model. It proved unreliable as a guide for drawing our sample of
firms with variable intensity of contact with the ATO during the preceding two
years, and even efforts to determine how many RTRs were conducted during the
same period yielded inconsistent numbers. Despite indications of considerable
organisational capacity for responsive regulation, the ATO probably will not
know with any degree of confidence how well the program operated.
What lessons can be drawn from the CE/B&C project that may be of value
now that the ATO has extended the Compliance Model to all areas of compliance
assurance? Clearly, new policies and procedures for evaluating them should be
launched at one and the same time. This was not done when the ATO moved to
responsive regulation of the small building and construction industry. One
possible consequence of this is limited gain in knowledge of how to implement
and make work soundly new policies. An interview subject told us:
I think the biggest problem in the office is that we always go - we don't have balance. We go all or
nothing. So, we go Compliance Model.. .'This is the be all and the end all, and it's happening today, and
everybody will live and breathe it'. And that was the message that was put out: 'You will live the
Compliance Model'...It should be - and I am not sure how to achieve it - a more subtle thing; it should be
built, and it shouldn't just be bang, thrown on the table and 'here, take it away. This is what you have got
to do'. We do a lot of things that way. The Office seems to always work that way; if it is going to start
something it doesn't pilot it, it doesn't test it, it goes for it.
Although this subject may be unaware of it, his comments on the importance of
testing new programs carefully and rigorously is disputed by few policy analysts.
Not only should they be tested, but procedures for doing so and a clear
understanding of the specific kinds of data to be collected should be built into the
program from the outset. Decisions about the kinds of evidence that will be taken
as a sign of success or failure should be reflected in routine record keeping.
Whether or not the turbulent environment conditions in which the ATO has
operated in recent years and its own unsettled internal worlds would have made
possible anything more than was done is unknown, but the ATO's evaluation
studies and the results of this research are cause for modest optimism.
Notes
1 Several months before it adopted the ATO Compliance Model, the ATO issued the Taxpayers' Charter
(Australian Taxation Office. 1997). Seen as a covenant with taxpayers, it spells out their rights and
obligations and the options available to them should they be dissatisfied with their treatment by the
ATO
2 This low response rate leaves us unable to assume confidently that those who returned the survey are
representative of all who took part in the CE/B&C project. Although we can only speculate as to the
reasons for it, the timing of the survey was unfortunate; by the time it was mailed to ATO personnel,
the CE/B&C project had been scaled back substantially. The Compliance Model no longer was a
salient issue or focus of work routines, and many ATO staff may have stopped caring about their
experiences using it. In addition, the volume of e-mail field-level personnel receive daily is both
sizeable and increasing, and this may make it easy to overlook or ignore messages which clearly do
not require a response. The average age of the 84 survey respondents was 42.6 years, and the average
length of time they were assigned to the CE/B&C project was 22.6 months. Women comprise 21 per
cent of the respondents. When asked if they had worked as a supervisor at any time during the
project, 34.6 per cent of the respondents answered affirmatively.
3 BISEP represented the idea that to understand the compliance of an individual or group, one needed to
understand the environment in which they operated. The initials of this acronym stand for B =
Business profile, I = Industry factors, S = Sociological factors, E = Economic factors and P =
Psychological factors.
4 Some may see irony in the fact that the ATO Compliance Model, which highlights the importance of
and the payoff from consultation, education and cooperative relationships, was adopted and
implemented in a top-down, command and control fashion within the Australian Taxation Office.
5 The one building and construction industry representative we interviewed was quite positive in his
comments about responsive regulation as he had experienced it. He was impressed particularly by
the willingness of ATO personnel to attend meetings of their organisation, to do presentations as part
of the meeting program, and afterwards to answer questions posed by builders in attendance.
6 In 30 years research, the senior author rarely has encountered the level of unease and anxiety
expressed in interviews with ATO field-level personnel. Whether this results from unfamiliarity with
academic investigators or from the pace of change in their work day, many of those we interviewed
seemed unusually jittery.
References
In April 1998 the Cash Economy Task Force (1998) produced an influential
Australian report entitled Improving Tax Compliance in the Cash Economy.
Improving Tax Compliance in the Cash Economy proposed a new Compliance
Model for the Australian Taxation Office (ATO) with four elements: (a)
understanding taxpayer behaviour; (b) building community partnerships; (c)
increased flexibility in ATO operations to encourage and support compliance; and
(d) more and escalating regulatory options to enforce compliance. During late
1998 and early 1999, the question was asked whether or not the Cash Economy
Compliance Model was relevant to tax compliance for the large business market
segment. At the time there were doubts.
This chapter opens up some issues for consideration about applying the model
to the ATO's compliance work on large business. Preliminary examination
revealed that the four elements of the mode] mentioned above all have relevance
to large business, though a different kind of relevance than in the case of the cash
economy. This chapter is structured around considering each of these elements in
turn. The spirit of the chapter is to identify a number of policy ideas that might
give meaningful content to the model in the large business context.
Although this chapter will argue that the ATO Compliance Model provides a
valuable framework for large business compliance work, the concepts and
enforcement strategies through which one escalates in a large corporate
compliance pyramid will be found to be quite different from those in a cash
economy compliance pyramid. For example, at the peak of the ATO's Compliance
Model enforcement pyramid, the ATO confronts disengagers from the tax system
using a prosecutorial strategy. Many individual taxpayers and small businesses
do opt out of the tax system altogether. Even some larger businesses that we call
organised crime (but that in reality is rather disorganised) do this. They do not
play the game, sometimes because they are so cynical or disenchanted about it,
sometimes to evade it in a calculative way. But some large businesses, on the
other hand, are quintessential game players. At the peak of the large business
enforcement challenge are the most entrepreneurial of the players of the tax
planning game. Disengagement is not a major source of non-compliance here.
There is no problem with adapting the ATO Compliance Model to this reality
of large corporate compliance. The idea of a compliance pyramid is that all
compliance staff will discover a content for it that is relevant to the context in
which they work. It is not a recipe book but a model to guide strategic thinking.
The deeper question about the model for the compliance of large businesses is
that it tends to assume that the majority of taxpayers want to comply and as we
move up to more and more serious tax evasion, there are fewer and fewer
taxpayers in this category. Evidence on individual taxpayers suggests that this
assumption is fairly accurate. For example, US audit evidence has been used to
conclude that about two-thirds of individual taxpayers intend to pay the right tax
(some of them inadvertently cheat), but only a third of individual taxpayers
actually set out to cheat in a significant way (and then to varying degrees)
(Andreoni, Erard and Feinstein, 1998, p. 820).
With large corporations in Australia, it is clear that many more than two-
thirds of taxpayers intend to comply with the letter of the law, or at least to have
a 'reasonably arguable position' that they have done so most of the time.
However, it seems equally clear that there are many who do not intend to comply
with what the ATO regards as the policy purposes of the parliament's tax laws.
The evidence for this is that more than half of them pay no tax, some for many
years, which is certainly not parliament's intention.2 Australia collects more than
the Organisation for Economic Cooperation and Development (OECD) average in
company tax as a proportion of all tax revenue and as a proportion of Gross
Domestic Product (GDP) and is one of the few countries in the late 1990s that saw
corporate tax revenues increase at a faster rate than the increase in company
profits. The corporate tax collection crisis is not distinctively an Australian one.
From these known patterns of taxpaying, the fact that compliance policy must
confront squarely is that while the pattern of individual compliance is broadly
pyramidal, the pattern of large coiporate compliance is egg-shaped, with most
taxpayers playing for the grey (see Figure 9.1). Large corporate compliance
strategy might be conceived as pushing the large grey bulge in the middle of the
egg downwards into the white zone - so that the egg becomes a pyramid.
It is hard to make compliance strategies work when compliance behaviour is
egg-shaped. As corporate compliance behaviour takes on more of a pyramidal
shape, the compliance policies discussed in this chapter will progressively become
more effective. The task is more daunting than for tax collection with small
business and individuals because the reality of the challenge is majority
noncompliance with the policy purposes of the law (at least as they are conceived
by many officers of the ATO). The argument to be made here is that law reform
is the first of three key circuit breakers to move that grey bulge down into the
white. Two further circuit-breakers are proposed in this chapter: One is for the
compliance debate to take a more democratic turn, the other is for it to take a
more international turn. This chapter seeks to show that once these three circuits
are broken, there are a number of more specific policy options for improving
compliance.
Figure 9.1 Patterns of individual and corporate tax compliance
The Australian people increasingly believe that large corporations and high
wealth individuals do not pay the share of tax that the law says they should
(Braithwaite, Reinhart, Mearns and Graham, 2001). A June 1998 community
survey of 1000 Australians, commissioned by the ATO, found that only 32 per
cent believed that 'Tax laws are effective in making sure large companies pay
their share of tax' declining to 27 per cent for 'very wealthy people'. Only 20 per
cent believed 'The ATO does a good job of stopping tax avoidance by large
companies', falling to 15 per cent believing 'The ATO does a good job stopping tax
avoidance by very wealthy people' (see also Wirth, 1998).
Such perceptions are a major risk to voluntary compliance by individual
taxpayers. The reason is that there is evidence that individuals are more willing
to pay their taxes honestly when they perceive most others to be honest; this
evidence suggests that the direction of causality here is that when citizens
perceive most others to be cheating, they are more likely to cheat themselves
(Levi, 1988; Scholz, 1998; see also Cowell, 1990).
Perceptions are so entrenched that it is past the point of government trying to
pretend to the Australian people that all large corporations are paying a share of
the tax burden commensurate with the share of Pay-As-You-Go (PAYG)
taxpayers. The better way to manage this risk to the revenue is to be open with
the Australian people about the depth of the problem. This means explaining to
them that it is not that companies are cheating in large numbers. It is that some
large corporations are aggressively planning their way around Australian law and
moving profits and losses around the globe to limit tax liability.
Of equal importance, a full and frank public debate about the nature of the
problem is the way to begin to bring the Australian judiciary to an understanding
of their central importance to solving the problem. The behaviour of the judiciary
in the aftermath of tax reform is ultimately the most critical risk in the new ATO
environment. History teaches us that when judges are literalist in the way they
interpret tax law, legal entrepreneurs will open up more and more loopholes in
the law. Yet because judges are properly independent of the executive
government, it is not appropriate for the ATO to 'tackle' the judiciary when they
fail to interpret tax law in a more purposive way.
So another path is needed. Most Australian judges believe in the sovereignty of
parliament or the sovereignty of the Australian people or both. A tax compliance
debate that engages the parliament with the desire of its people, openly
expressed, that the law be interpreted to require large corporations to pay the
share of tax the law intends is the best way to seek to move judges who believe in
a sovereign parliament responding to a sovereign people. The judges understand
that a literalist approach to tax adjudication can survive while tax law remains
arcane to the people; equally, they understand that judicial repute suffers in the
democracy when this approach is later discredited in the eyes of the people and
their parliament, as happened to the Australian judiciary in the 1980s (Levi, 1988).
In the long run, court cases like the recent Packer saga, which led to Australia's
richest person being accused in the press of paying virtually no tax, are corrosive
of respect of the people for the law and the courts.3 The more open the public
debate on tax law, the more clear it will be to the judiciary why this is so, and the
more pointedly will they grasp the necessity and propriety of their being more
purposive and less literalist readers of tax law.
Margaret Levi's (1988) analysis of the 1983-85 tax reform debate is that it
engaged the Australian people in a way that increased their preparedness to
comply with the tax laws. Part of this accomplishment was that they did see the
Barwick (literalist) era of judicial interpretation in tax matters, of which they
came to disapprove, come to an end.
The ATO generally has done a good job of the first four stages of this process,
particularly at a strategic and organisational level, but is only beginning to do a
good job on the fifth stage (Treat the Risk), and is still doing a poor job on the
final stage (Monitor and Review). In establishing the context, identifying the risks
and analysing the risks, the ATO has the advantage of being an unusually
knowledge-rich organisation. It takes research seriously. While there are gaps in
this knowledge which need to be filled, the bigger challenge is to find existing
knowledge and synthesise its implications.
While ATO staff are increasingly on top of risk analysis and assessment,
mostly they do not operate by searching and seizing opportunities to leverage
those risks. Ask in a large business industry segment in the ATO whether they
have assessed and ranked risks and they will pull out a document that shows
where they have done that. This is genuinely impressive. But ask fieldwork or
management staff to tell you the best stories of how they have leveraged each of
those risks and only some will enthuse with their triumphs.
Risk leveraging is a creative activity. It requires creative staff. It is a bad idea to
provide a formula for how to do it because advisers will soon learn that formula.
Continuous reinvention of risk leveraging is what will keep them guessing and
therefore complying. A culture of continuous reinvention of risk leveraging
seems to us to require taking storytelling more seriously within a tax authority.
The ATO has begun to move away from being a business run according to a
procedures manual (see Job and Honaker, Chapter 6, this volume). At the level of
informal staff interaction, ATO culture is no longer just a rulebook, it is also a
storybook.
Some staff see ATO management culture, in contrast, as a thicket of models (an
observation also made by Job and Honaker, Chapter 6, this volume). The
challenge is to value stories that make sense of models and models that provide a
conceptual scheme for generating better stories. One staff response to the roll-out
of the ATO Compliance Model was: 'Oh no, not another model. Now a [ATO]
Compliance Model to add to Risk Assessment,...System Analysis and Planning,
Strategic Direction, Performance Management, Business Systems models,
Accountability and Governance, and on and on'. Our suggestion is that the
approach to moving the ATO risk management model beyond risk assessment to
risk leveraging is to create a framework for storytelling about compliance
successes. Some leading corporations, such as 3-M, have come to the conclusion
that an excess of abstraction is their problem and have taken the remedy so far as
to write their strategic plan in storytelling fashion (Shaw, Brown and Bromiley,
1998). The rationale for the storytelling approach lies in optimising human
capacity for not only digesting information but also acting on it.
Stories are central to human intelligence and memory. Cognitive scientist William Calvin describes how
we gradually acquire the ability to formulate plans through the stories we hear in childhood. From
stories, a child learns to 'imagine a course of action, imagine its effects on others, and decide whether or
not to do it'...Cognitive scientists have established that lists, in contrast, are remarkably hard to
remember...(Shaw, Brown and Bromiley, 1998, p. 42).
It was not until 1914 that in the average encounter between a doctor and a patient
the patient was more likely to come away better off than worse off.
Unfortunately with law enforcement, we have probably not yet reached 1914. If
we put more police into a neighbourhood, we are just as likely to increase the
crime rate as to reduce it. The reason is that, like doctors with leeches, police do a
lot of things that make crime worse (as well as a lot that makes it better).
Law enforcement needs to learn two things from medicine. One big step
medicine took toward making us healthier early this century was a research
investment in randomised controlled trials of its risk leveraging strategies. If it
believed a particular therapy or a particular pill would outperform existing
therapies, it would randomly assign a large group of patients to the new
treatment versus the old. In effect, this meant tossing a coin to decide whether
patients got the currently popular therapy or the new one. With a large enough
sample, the miracle of randomisation delivered two groups that were exactly
comparable in every way other than the treatment. If fewer patients died under
the new therapy, then we could be fairly sure that the therapy was the reason.
This was the big advance over the old science that allowed quacks to delude
themselves into believing that when patients got better after being treated with
leeches that it was the leeches that cured them. Or when they died that the
leeches had not been applied early and often enough.
As more and more therapeutic advances increased our life expectancy,
medicine's new problem was of too much knowledge. Most doctors were not up-
to-date with what science was finding. So a new evidence-based medicine
movement to get the results of science through to doctors in a digestible form
began: A science of diffusing science through soft networks of revered peers. The
nation that ventures into an evidence-based tax administration comparable to
evidence-based medicine is likely to become the cutting edge. Evidence-based tax
administration will be much cheaper than evidence-based medicine and much
less fraught with ethical dilemmas.
Imagine a tax authority comes up with an idea for a new auditing product,
which it believes will improve compliance at lower cost. Some people think they
are wrong, others that they are right. A new evidence-based Commissioner sees
that there are some good arguments on both sides. Moreover, they both have
equally good arguments as to why their approach would be procedurally fairer
and cheaper for taxpayers, so there are no ethical arguments against
experimentation. There are resources available for 200 audits where this kind of
audit is relevant. So the Commissioner requires that a risk analysis select the 200
most suitable targets for audit. A random number generator then assigns 100 of
these cases to the new audit product and 100 to the old audit.
Where there are one or two extremely large or atypical corporations these
might be excluded from the experiment on grounds that if both ended up in the
same group, this would skew the results. Or the randomisation can block on
them, so that it is guaranteed that one will go into the experimental group, the
other into the control group. Absent such extraordinary lumpiness, the laws of
probability prove that randomly assigning two hundred companies or two
hundred individuals is almost certain to produce two groups with almost
identical breakdowns on average income, age, sex, in fact anything we can
measure and check (and everything that we cannot measure and check). The only
reason for a difference in tax paid by the two groups is that one gets the old style
audit rather than the new style audit. There is no need for complicated multiple
regression analysis. Tax officers just count how much money comes through the
door from the new audit group compared to the old audit group.
An even better research design will require a risk assessment to identify the
300 best targets for auditing. Then the computer will assign 100 to the new audit,
100 to the old audit, 100 to no audit. It might be that even though the new audit
brings in more tax than the old audit, over a three year follow up, neither group
outperforms the 'no audit' group. The reason might be that the 'no audit' group
are afraid of an audit in years two and three, while the audit groups believe they
get a free kick for those two years.
Under both research designs it is possible to test which is the cheaper process
to run without having to contend with doubters who say the new process was
cheaper only because it was tried out on less complicated audits. Randomisation
will have assured that on average the new and old groups are of equal
complication.
In few areas of problem-solving could randomised controlled trials be more
viable than in tax administration. It has been necessary to the advancement of
medicine to randomly assign sick patients to a placebo (an inert pill) or to no
treatment when they might have been helped. While many of us are alive today
because this was done, it raised awful ethical dilemmas. There is not this level of
ethical dilemma in randomly assigning a high risk taxpayer to miss out on an
audit. There would be in a world where the community regarded it as just that all
high risk taxpayers should be audited. The world of tax is not such a world - the
supply of compliance work is always less than the need for it; a high risk
company that is randomly assigned to no audit this year can always be
purposively assigned to audit next year or the year after or both. However high
their risk, a tax authority cannot afford to audit them every year.
Many other kinds of compliance initiatives have the same character as audit in
these regards. Consider, for example, an interesting initiative of simply writing to
companies in the manufacturing sector that have paid no tax for three years to
say that the ATO is concerned about this situation and requests them to write
with an indication of whether they think their circumstances will change such
that they will be paying some tax next year. The hypothesis is that the fear that
they are being monitored in some special way will cause them to pay more tax in
the next year. 100 companies could be randomly assigned to no letter and no
audits; 100 companies to the letter and X audits for those who fail to reply with
an indication that they expect to pay tax next year; 100 to no letter and X audits;
100 to the letter and no audits. This design would enable an assessment of
whether the (practically costless) strategy of getting the computer to send a
hundred letters increases tax receipts even without backing them up with audits
for those who ignore them. Or whether this would only work when coupled with
a commitment to do something if tax did not come in.
A major advantage that tax has in enabling randomised controlled trials in
comparison with medicine is that it has the entire population of taxpayers on the
computer, so randomisation with high external validity can be done at low cost.
Most people who do not have experience of randomisation doubt that with
corporations you would get an experimental group and a control group that
would be identical on relevant characteristics. A needed confidence-building
measure is to ask the doubters to name the variables on which the experimental
group might end up different from the control group. The data would be collected
to show that the two groups had the same value on average for these measures.
The approach of the ATO Compliance Model values flexibility, showing that
there are many ways of delivering on compliance and objectives of the
Taxpayers' Charter. Institutionalising a storytelling culture is one way of
fostering flexibility. Success stories that grab people's interest will be stories of
innovation, of more flexible responses than have been attempted in the past.
A growing source of flexibility is to take problems to international forums. The
Advance Pricing Arrangement (APA, called Advanced Pricing Agreements in
many countries) is one approach to locking in higher tax receipts from
transnational corporations (Killaly, 1996). APAs are negotiated arrangements
between the ATO and corporations on transfer pricing methodologies negotiated
at the OECD, which result in an appropriate allocation of income and expenses
between related parties that are selling goods or services between different
countries. Negotiating APAs is painstaking work, but because they lock in higher
returns than audits do and because they shift the rules of the game to more
cooperative ones with business, the investment is well justified.
On the other hand, the ATO needs to monitor the cost of keeping APAs up to
date in the face of company-, product- and time-specific changes that make the
parameters of the APA obsolete. In addition, there was a worry that only
'squeaky clean' companies would ask for APAs, hence skewing ATO activity to
areas of low risk. Initially, companies were reluctant to enter into APA
negotiations because they feared this may have revealed tax liabilities going back
over many years. However since then APAs have become widely used. The kind
of flexibility of ATO response that could be examined is for the ATO to consider
granting an amnesty on tax liabilities going back more than two years as part of
the incentive for entering into negotiations. International tax competition can
cause compliance problems that can only be addressed through international
cooperation via forums such as the OECD and multi-lateral APAs. The same
international imperatives apply with tax havens, E-commerce and a variety of
other questions.
The real power to increase flexibility does not lie with senior managers so
much as with fieldworkers who are daily at the coalface (Sparrow, 2000).
Fieldworkers must be the key risk identifiers. Their performance reviews should
give high priority to evidence of identifying wider risks in the course of their
fieldwork, documenting them and following through to ensure that the risks are
analysed and treated. In the view of some, this follow through is not their job.
Their job is to report the risks they identify to their immediate boss so that
management can take care of it. This is neither our view nor that of senior ATO
management involved in large business compliance work. The ATO needs more
leadership from below. Fieldworkers can often report risks to Segment Managers
who can be too busy to follow through on them. It is often more efficient for the
fieldworker with the direct experience of the risk to chase it up through the
bureaucracies, to participate in the senior discussions there on what can be done
to treat it and then to offer to be a 'test pilot' for the treatment - to report back
whether the treatment is working with the initial kind of risk they identified.
Better still, the fieldworker might recruit one of the clients they work with to be
an external test pilot as well, and also participate in the discussion at senior levels
concerning the problem they have decided to own.
If tax professionals at all levels of the ATO from fieldworkers up are
encouraged to be leaders (see Job and Honaker, Chapter 6, this volume), to own
risks with a commitment to follow through all the stages of risk management,
then loop closing will improve. Most critically, the law reform loop will respond
faster, and fieldworker morale will improve compared to a situation where they
sit on their hands and grumble about management dropping the ball. What
follows is that the tax authority storytelling culture needs to be about stories of
junior employees being rewarded in performance reviews for following risks
right through to treatment and evaluation. In few areas is this more important
than law improvement.
In the midst of the wave of tax law reform currently being experienced in
Australia we must remember that it is naïve to believe that we can ever reform
tax law and get it right in a way that will remain right for long. Nothing is more
important to improving corporate tax compliance than law reform. But the
reality is that as soon as the new law is in place, there will be legal entrepreneurs
who will be at work to open up loopholes in it, and globalisation will over time
deliver changes that will make it progressively obsolete. So law reform is not the
task of an historic moment, but a continuous process.
The ship of tax reform needs to be continually rebuilt at sea as legal
entrepreneurs open up one leak after another. Every decade or so the
accumulation of extra pieces added to the ship to plug these holes will start to
weigh it down with an excessive burden of complexity. The ship then has to be
taken into dock and rebuilt from the ground up in a more simplified way. The
most crucial assessment of the health of the system must involve a sophisticated
group looking periodically at the ship to assess whether the overall pattern of its
complexity justifies such a systemic legal refit. Plugging holes at sea and periodic
simplifications of the whole structure are both recurrently necessary because an
accumulation of new rules to plug old loopholes can be a resource for opening up
new loopholes. Complexity favours the legal entrepreneurs. The objective is to
give the judges only as much detail as they need to apply the policy purposes of
the law with as much certainty as is possible for the contrived uniqueness of the
circumstances with which they are continually presented (see McBarnet, Chapter
11, this volume).
The job of fieldworkers is to be the antennae that detect those risky new
contrivances as early as possible. A good auditor has the ability to see that an
issue that has come up in an audit will be a risk to the revenue in many other
cases unless the law is clarified or improved. Provision exists for auditors to
identify risks on case files that are available for computer analysis to group
problems that are looking for common solutions. The culture change needed is
one where writing down the risk is just the beginning of a process for the
fieldworker of following the risk through to ensure there is a ruling or litigation
to clarify the case law, an actual change in the law, or a principled decision to let
the risk sit there until the next comprehensive reform of the law. At every stage
the fieldworker has a role as a reality tester for the lawyers, as an antenna to
detect the latest manoeuvre of the other side and as a test pilot for the proposed
legal remedy.
Tax law will become more certain and effective when fieldworkers are engaged
with the daunting challenge of continuous improvement of a living law rather
than the narrow challenge of enforcing a static law. It will become even more
certain and effective if fieldworkers engage the public-regarding side of large
business at the coalface, where the problem in the law breaks out, to help repair
the law in the public interest. Law reform in the past has been too top-down and
insufficiently continuous. Also, as argued at the beginning of the chapter, it has
not engaged the people of Australia in public discussion and understanding often
enough. The ATO needs to persuade the people of Australia that more equitable
and certain enforcement of the tax law is something they can reasonably demand
of their institutions - the parliament, the judiciary and the ATO. By articulating
forcefully to the Australian people that the ATO does not want to be let off the
hook for tax integrity, it makes it harder for the parliament and judiciary to be
allowed off the hook (which to a significant extent they have been in the past).
It is doubtful that any amount of synoptic brilliance in re-configuring tax law
at one point of history can deliver Australia a high integrity tax system. Rather,
what is required is for us to be more pre-emptive through a continuous
improvement process of simplification-clarification-resimplification-clarification.
This iterates endlessly. Commitment to excellence in the pre-emptiveness and
responsiveness required for continuous improvement of tax law is the key.
Australia can have a high integrity tax system, but only if the ATO, the
parliament and the judiciary jointly end the buck-passing. My plea is that the
ATO publicly put up its hand and force the hands of the other institutions.
Improved disputing can improve compliance considerably. The reason is the now
substantial evidence that when people and companies believe they have
experienced fair procedures, they are more likely to comply with the law (Lind
and Tyler, 1988; Tyler, 1990, 1998; Tyler and Dawes, 1993; Makkai and
Braithwaite, 1996). The Large Corporate volume of the ATO Professionalism
Survey does show procedural fairness to be a concern of large corporates with
ATO staff (Donovan Research, 1998).
Adversarialism arises often in tax fieldwork. Procrastination as an alternative
to resolving disputes commonly uses delaying tactics such as manipulating the
administrative privilege of accountants, Freedom of Information requests,
administrative law hearings, holding back on assistance with providing requested
records, providing only parts of the information requested, and failing to attend
meetings. Good tax office practice is to refuse to tolerate failure by either party to
resolve disputes. Otherwise the agency hands victory to the people who practice
'Defer, Delay, Defeat'. Headbutting or delay that is obstructing resolution can be
dealt with by widening the circle involved in the dispute. Some large businesses
are alert to this option themselves as they regularly go over the head of
fieldworkers to their superiors. Experience demonstrates it to be a good option for
the large business fieldworker as well (and so does theory, Braithwaite, 1997). If a
matter is important enough, the fieldworker's senior manager can write to the
CEO saying, in effect, 'We need you to supply this information in a timely
fashion so we can settle this matter. We can use our powers to compel you to do
so and stand ready to go to court to enforce this. But this is not the way we like
to do business. Can we meet and exchange the information we need to get this
dispute over with?'. At lower levels, an auditor having difficulty with the tax
manager of a company can ask for a meeting with the tax manager and his boss
together, and then with the boss's boss if that meeting accomplishes no
reconciliation.
The rationale for this path to flexible dispute resolution is that large businesses
are full of many adversarial people and many cooperative people. The trick is to
move up the organisation through various layers of adversarial managers until
the tax officer reaches a cooperative manager who insists that the matter be
settled rather than take up more of everyone's time. Actually, the grounds for
optimism that this works are even stronger than suggested so far. Even the most
adversarial of executives have a cooperative, socially responsible self as well as a
combative self (see Braithwaite, Chapter 2, this volume). The gifted tax officer
has the ability to treat clients with a respect that persuades even the most
combative of them to put forward their socially responsible self. If she has a bad
day where she fails to pull this off, she knows how to retreat, widen the circle for
another day on which she hopes to catch the new player when his socially
responsible self is to the fore. Moving up the organisation until a more senior
cooperative person is found who will instruct the junior obstructionist to
cooperate can be time-consuming. But it is less time-consuming than escalating
prematurely to setting up an arbitration or litigation, or leaving the problem to
fester.
The first step up a large business compliance pyramid might be a Tax Strategy
Review that checked directors' minutes, working papers, accounting and tax
manuals, audit files, financial planning files, budgets/forecasts and the like. This
is a small step up the pyramid for a large corporation. It is hardly an enforcement
step at all, merely a risk assessment exercise. Nevertheless, it is a good first step to
take when perhaps all that is required is a signal that the tax authority is
reviewing the affairs of the business.
The second step might be Real Time Enquiries to clarify a concern; the third a
Special Purpose Audit or other low-intensity cooperative audit; the fourth a Full
Cooperative Audit (see Figure 9.4). When escalated cooperative audits meet with
repeated obstructionism, escalation to what might be called 'Assertive Use of
ATO Powers' could be deployed with little hesitation. This means use of formal
powers to demand documents and answers to questions, and surprise visits to
premises to demand access to documents where this is necessary. Escalation from
cooperative audit to assertive use of ATO powers should not normally occur until
there has been opportunity for an ATO senior manager and the CEO to discuss
cooperative dispute resolution options.
When both obstructionism and reasonable suspicion of serious non-compliance
exists, a Litigation Task Force might be put on the case (see Figure 9.4). The
litigation task force would be multidisciplinary, including a lawyer who would
liaise with the Director of Public Prosecutions. The obligation of the head of the
litigation task force would be to ensure that the necessary evidence was collected
to pursue prosecution and other penalties. This is not to say that maximum legal
action would be pursued; rather, the objective would be to make it clear that this
was inevitable unless the taxpayer moved the relationship down the pyramid to
more cooperative problem resolution. There has been a view that the evidence
rarely exists to warrant higher penalties or for prosecution to be successfully
pursued. Doubts exist about this assumption. Rather, an alternative view is that a
large corporation which has escalated this far up the pyramid will be vulnerable
somewhere in its complex of activities to false and misleading statements having
been made, among other vulnerabilities. In part, this view on tough enforcement
may itself stem from a perceived lack of support from ATO managers, the ATO
itself and even the Director of Public Prosecutions. The use of a litigation task
force could help crystallise this support and more tangibly signal to the staff and
the taxpayer involved that swift and sharp enforcement will be taken.
The next rung in the compliance pyramid will be assessments/penalties that
trigger litigation by the taxpayer and/or prosecution by the Director of Public
Prosecutions (see Figure 9.4).
The final rung of the compliance pyramid could be referral to the National
Crime Authority (NCA).6 This occurs rarely at the moment, a situation that
should continue. The NCA's superior powers and criminal enforcement
capabilities should not be employed even on the most hardened criminals, but
only on uncooperative hardened criminals who are believed to be major tax
evaders.
The pyramid in Figure 9.4 is an illustrative pyramid that may be useful in
navigating the compliance options available in large business compliance at this
point in time. No compliance pyramid can be effective without changing over
time, with context, and in response to research findings from evidence-based risk
leveraging. Nor should this compliance pyramid, or any other, be prescriptive.
They are no more than a guide to compliance decisions that should never be
allowed to trump contextual wisdom. It should never be a criticism of an officer
that they are not following the ATO pyramid. It should be cause for praise when
the idea of the pyramid is used flexibly to display an utterly original array of
innovative risk leveraging strategies through which the ATO might escalate in
dealing with a tough case (as Job and Honaker, Chapter 6, and Hobson, Chapter 7,
argue in this volume). Rather than slavish implementation of 'the Model' we
should want success stories of innovation in pyramidal thinking about leveraging
compliance.
By thinking pyramidally about the enforcement options available, the tax
officer engenders confidence and motivates cooperative compliance through
showing a willingness to escalate up the pyramid. A regulator who believes that
her pyramid is credible and who is willing to escalate up it if necessary rarely
actually finds it necessary to escalate far up the pyramid. Confidence about the
authority of the pyramid of enforcement capabilities projects an image of
invincibility to the corporation that is the subject of tough enforcement. This
rarely fails to engender cooperative compliance at low levels of the pyramid,
enabling de-escalation to even lower levels. The effective regulator is cooperative
and trusting at first, tough if that trust is abused, tougher and tougher if it is still
abused, but forgiving if trust and cooperation is finally restored (Ayres and
Braithwaite, 1992, Chapter 2). It does not matter that escalation at lower levels of
the pyramid has very little bite with large corporations; at that level the objective
is to signal escalation rather than cutting deeply into the corporate's interest. The
crunch is whether the tax authority is willing to escalate until it does reach
something that bites.
For the pyramid to work, fieldworkers must be confident that they will get
senior management backing if they escalate. They have that confidence because
they do not escalate or threaten escalation of a particular type unless they get
backing at a level that can support that level of escalation. Fieldworkers would
not escalate to an enforcement option and would not initiate assertive use of ATO
powers without getting the backing of their leaders in advance. Fieldworkers,
however, should not assume their leader's support when they want to escalate to
assertive use of ATO powers and beyond. They must be able to count on their
backing when the fieldworker and leader agree in advance on the circumstances
of the escalation. Once a litigation task force is established, control passes out of
the fieldworker's hands into the hands of the head of the task force (usually a
lawyer). Assurance that at some point the lawyers take over as a matter of policy
adds to the confidence and authority of the fieldworker who can say: Look,
unless we can sort this out, the conflict will escalate to the point where ATO
policy requires that it be taken over by a Litigation Task Force. Near the peak of
the pyramid, no promises of escalation would be made without the backing of the
Commissioner.
Of course, many of the strategies in Figure 9.4 will be deployed for risk
management reasons rather than out of any desire to give a signal that the
behaviour of the taxpayer is causing their case to be taken more seriously,
Sometimes when escalation up the pyramid is occurring, the circumstances will
warrant a jump up several rungs of the pyramid. The pyramid represents a
preference for resolving matters at lower levels, not a rule to do so. Starting at the
base of a pyramid and moving up progressively is a presumption that can be
overridden by compelling evidence, for example, of criminal behaviour or
aggressive tax planning, that this is a case that should go straight to the peak of
the pyramid. But we should be extremely reluctant to do that.
It is a challenge to educate people that there is no single or correct pyramid,
but that there is virtue in having the professionalism that comes from having
thought through the range of responses through which a regulator can escalate
when faced with non-cooperation. For example, one way to meet the challenge
could be to build various compliance pyramids that deal with access to
information. A workshop would get a group of fieldworkers who confront a
similar group of clients with similar access difficulties to co-design their own
pyramid of escalated response to denial of access. In the course of the workshop,
they would be shown the compliance pyramids designed by other access
workshops. Seeing their differences would both give them ideas and assurance
that they could design their own. Indeed, there would not have to be agreement
at the workshop. Different participants could go away with different pyramids
that suited their contexts. The objective would be that each participant would
develop a pyramid that she thinks would work for her and that she could get her
supervisor to back her on.
Conclusion
The four elements of the ATO Compliance Model: (a) understanding taxpayer
behaviour; (b) building community partnerships; (c) increased flexibility; and (d)
more and escalating regulatory options have been found to have relevant
meanings for large business tax policy. This is so even if it is the case that
improving compliance is a less satisfactory way of planning large corporate tax
objectives than reducing risks to the revenue and increasing legal certainty. There
is a long way to go before large business tax enforcement achieves problem-
focused flexibility informed by bottom-up storytelling and top-down demands
for evidence-based risk leveraging. The field has a long way to go before it is
genuinely a regulatory craft (Sparrow, 2000). An egg-shaped form of compliance
behaviour where most enforcement effort is directed at grey areas of avoidance
rather than evasion makes the challenge especially difficult. It also makes it a
challenge where law enforcement and law reform must be integrated instead of
separated functions. My argument is that a combination of escalated responsive
enforcement integrated with responsive law reform has some hope of shifting
much of the grey bulge of tax avoidance into tax compliance.
Notes
1 This is a revised version of a longer paper by John Braithwaite and Andrew Wirth, 'Towards a
Compliance Framework for Large Business Tax Compliance', Centre for Tax System Integrity,
Working Paper No. 24, The Australian National University, Canberra.
2 In a 29 May 1998 presentation to the NSW Annual State Seminar of the Taxation Institute of Australia,
LB&I (Large Business and International, ATO) reported that 'almost 60 per cent' of its companies
were non-taxpayers. Consolidation would doubtless reduce this proportion substantially because one
large corporate group might include a number of entities that pay no tax and others that pay large
amounts.
3 Kerry Packer's taxpaying behaviour has been subject to public scrutiny on a number of occasions. See,
for instance, House of Representatives Select Committee on the Print Media Inquiry (1992) and The
Australian (2000), 'Packer Sues over Internet Tax Ads', 5 September, A. McGilvray and A. McKenzie,
p. 1.
4 BISEP represents the idea that to understand the compliance of an individual or group, one needs to
understand the environment in which they operate. The initials of this acronym stand for B =
Business profile, I = Industry factors, S = Sociological factors, E = Economic factors and P =
Psychological factors.
5 The Large Case Program focused ATO audit activity on the largest corporate taxpayers and was
conducted during the later half of the 1980s and early 1990s.
6 This is a body akin to Crime Commissions that have special evidence-gathering powers against serious
organised crime in a number of countries. Indeed the National Crime Authority in Australia is in
transition to becoming a National Crime Commission.
References
Alm, J., Bahl, R. and Murray, M. (1993), 'Audit Selection and Income Tax
Underreporting in the Tax Compliance Game', Journal of Development
Economics, vol. 42, pp. 1-33.
Aim, J., Jackson, B. and McKee, M. (1992), 'Estimating the Determinants of
Taxpayer Compliance with Experimental Data', National Taxation Journal,
vol. 45, pp. 107-14.
Andreoni, J., Erard, B. and Feinstein, J. (1998), 'Tax Compliance', Journal of
Economic Literature, vol. 36, pp. 818-60.
Australian Taxation Office (1995), 'Staff Perceptions of the Compliance Behaviour
of the Top 100 Companies Audited Under the Large Case Program', Paper
presented to the Compliance Improvement Committee, Australian Taxation
Office, 12 December, Canberra.
Ayres, I. and Braithwaite, J. (1992), Responsive Regulation: Transcending the
Deregulation Debate, Oxford University Press, New York.
Baldry, J.C. (1987), 'Income Tax Evasion and Tax Schedule: Some Experimental
Results', Public Finance, vol. 42, pp. 357-83.
Braithwaite, J. (1997), 'On Speaking Softly and Carrying Sticks: Neglected
Dimensions of Republican Separations of Powers', University of Toronto Law
Journal, vol. 47, pp. 1-57.
Braithwaite, J. and Makkai, T. (1994), 'Trust and Compliance', Policing and
Society, vol. 4, pp. 1-12.
Braithwaite, J., Pittelkow, Y. and Williams, R. (Chapter 10, this volume). 'Tax
Compliance by the Very Wealthy: Red Flags of Risk'.
Braithwaite, V. (Chapter 2, this volume), 'Dancing with Tax Authorities:
Motivational Postures and Non-compliant Actions'.
Braithwaite, V. and Levi, M. (1998), Trust and Governance, Russell Sage
Foundation, New York.
Braithwaite, V., Reinhart, M., Mearns, M. and Graham, R. (2001), 'Preliminary
Findings from the Community Hopes, Fears and Actions Survey', Centre for
Tax System Integrity Working Paper No. 3, The Australian National
University, Canberra.
Cash Economy Task Force (1998), Improving Tax Compliance in the Cash
Economy, Australian Taxation Office, Canberra.
Clotfelter, C.T. (1983), 'Tax Evasion and Tax Rates: An Analysis of Individual
Returns', Review of Economic Statistics, vol. 65, pp. 363-73.
Cowell, F A. (1990), Cheating the Government: The Economics of Evasion, MIT
Press, Cambridge.
Donovan Research (1998), 'Australian Taxation Office Professionalism and Service
Standards Research, Volume 5: Corporates', Australian Taxation Office,
Canberra.
Feinstein, J.S. (1991), "An Econometric Analysis of Income Tax Evasion and its
Detection', Rand Journal of Economics, vol. 22, pp. 14-35.
Friedland, N., Maital, S. and Rutenberg, S.M.A. (1978), 'A Simulation Study of
Income Tax Evasion', Journal of Public Economics, vol. 10, pp. 107-16.
Grey, S. and Cooper, D. (1998), Australian Taxation Office Large Business and
International Business Line: Risk Management Process Definition, Cambridge
Broadleaf Capital International, Cambridge.
Hobson, K. (Chapter 7, this volume), 'Championing the Compliance Model: From
Common Sense to Common Action?'.
House of Representatives Select Committee on the Print Media (1992), Report
from the House of Representatives Select Committee on the Print Media News
& Fair Facts: The Australian Print Media Industry, Commonwealth of
Australia, Canberra.
Job, J. and Honaker. D (Chapter 6, this volume), 'Short-term Experiences with
Responsive Regulation in the Australian Taxation Office'.
Joulfaian, D. and Rider, M. (1996), 'Tax Evasion in the Presence of Negative
Income Tax Rates', National Taxation Journal, vol. 49, pp. 553-70.
Killaly, J. (1996), "Advance Pricing Arrangements: The Australian Position:
Practical Issues Relating to Australia's APA Program', Organisation for
Economic Cooperation and Development, Working Paper No. 6, Organisation
for Economic Cooperation and Development, Paris.
Kraakman, R.H. (1984), 'Corporate Liability Strategies and the Costs of Legal
Controls', Yale Law Journal, vol. 93, pp. 857-98.
Levi, M. (1988), Of Ride and Revenue, University of California Press, Berkeley.
Lind, A.E. and Tyler, T.R. (1988), The Social Psychology of Procedural Justice,
Plenum Press, New York.
Makkai, T. and Braithwaite, J. (1993), "Praise, Pride and Corporate Compliance',
International Journal of the Sociology of Law, vol. 21, pp. 73-91.
— (1996), 'Procedural Justice and Regulatory Compliance', Law and Human
Behaviour, vol. 20, pp. 83-98.
McBarnet, D. (1992), 'The Construction of Compliance and the Challenge of
Control: The Limits of Noncompliance Research', in J. Slemrod (ed), Why
People Pay Taxes: Tax Compliance and Enforcement, University of Michigan
Press, Ann Arbor, pp. 333-45.
—(Chapter 11. this volume), 'When Compliance is not the Solution but the
Problem: From Changes in Law to Changes in Attitude'.
Pappas, Carter, Evans and Koop (1992), Review of the Large Case Program, The
Boston Consulting Group, Boston.
Revenue Canada (1996), Revenue Canada 4024 The Audit Protocol, Revenue
Canada, Ottawa.
Scholz, J. (1998), 'Trust, Taxes and Compliance', in V. Braithwaite and M. Levi
(eds), Trust and Governance, Russell Sage Foundation, New York, pp. 135-66.
Shaw, G., Brown, R. and Bromiley, P. (1998), 'Strategic Stories: How 3M is
Rewriting Business Planning', Harvard Business Review, vol. May-June, pp.
41-54.
Shearing, C D. and Ericson, R.V. (1991), 'Culture as Figurative Action', British
Journal of Sociology, vol. 42(4), pp. 481-506.
Sparrow, M.K. (2000), The Regulatory Craft, Brookings Institution, Washington D
C.
Tyler, T.R. (1990), Why People Obey Law, Yale University Press, New Haven.
—(1998), 'Trust and Democratic Governance', in V. Braithwaite and M. Levi (eds),
Trust and Governance, Russell Sage Foundation, New York, pp. 269-94.
Tyler, T. and Dawes, R.M. (1993), 'Fairness in Groups: Comparing the Self-interest
and Social Identity Perspectives', in B.A. Mellers and J. Baron (eds),
Psychological Perspectives on Justice: Theory and Applications, Cambridge
University Press, Cambridge, pp. 87-108.
Wickerson, J. (1993), 'Measuring Taxpayer Compliance: Issues and Challenges
Facing Tax Administration', Australian Taxation Office, Canberra.
Williams, R. (1995), 'Behavioural Research Synopsis', Australian Taxation Office,
Canberra.
Wirth, A. (1998), 'Public Perceptions and Expectations of the Australian Taxation
System', Australian Taxation Office, Canberra.
Part III
Beyond the Compliance Model
Chapter 10
Tax Compliance by the Very Wealthy:
Red Flags of Risk
John Braithwaite, Yvonne Pittelkow and Robert Williams
The issue of detection of tax evasion and avoidance is no small challenge for tax
authorities. Oftentimes detection depends on extensive auditing. The ambiguity
and difficulty of the task has led tax authorities to devise risk management
strategies so that they can dedicate their limited resources to detection exercises
that are likely to uncover instances of evasion or aggressive avoidance. The
process of identifying a set of red flags that alerts tax officers to aggressive tax
planning is the focus of attention in this chapter. We illustrate how a statistical
methodology can be deployed with subjective estimates of risk made by expert
analysts and more objective data in the form of dollars at risk.1
Asking the question and getting them to focus their mind on where everything is had an impact.
(HWI advisers quoted in Braithwaite, 2001, p. 10.)
Another adviser suggested that 'the more information he's [the Commissioner]
got, the less aggressive they will be in their tax planning'. What was meant by
this comment was that through notifying the Commissioner of 'X' in 1998,
taxpayers reduced their degrees of freedom to re-configure their 1999 affairs in
such a way that not-X appeared to be the case in 1999. Changes from year to year
will be noticed, so HWIs must keep their affairs consistent with the underlying
truths of earlier declarations. Also the more holistic approach of HWI program
surveillance means there is a need to work harder at keeping the story about one
HWI entity's tax affairs consistent with that of another (Braithwaite, 2001, p. II).
The data in this chapter are based on risk ratings of 207 potential risk issues
identified from 1997 and/or 1998 tax returns for 235 individual HWIs and the
entities they control.
ATO analysts might examine as many as 100 expanded returns from all the
entities controlled by a HWI, and then in theory rate each of the 207 potential
risk issues from 0 to 10. These ratings are estimates of indications of risk; that is,
there is no guarantee that there really is a high risk. Usually an audit would be
needed, sometimes even litigation, to establish if the risk was a reality.
As with all risk ratings, the rating estimates used as the basis for this report
were checked by a supervisor who may have revised them. For the highest risks,
senior taskforce staff would also meet to discuss the analyst's assessment.
Most of the potential risk issues were not recorded or rated for each HWI. In
fact more than half the HWIs in the database had less than ten rated issues. The
maximum was fifty-one, and the minimum was one.
This chapter is concerned with high risk ratings. A rating of 1-3 was defined by
the taskforce as a low risk, 4-7 as medium, and 8-10 as high. The defining features
of a score of more than 7 in the instructions to analysts are indications of
aggressive tax planning 'like significant loss creation', followed by a list of other
types of aggressive tax planning or 'further information desirable and would
suggest audit action'. Judgments about what is aggressive tax planning are
controversial and subjective2 and there are many reasons to suspect risk ratings
as unreliable.3 In the first part of this chapter we accept these subjective
judgments at face value and use them to predict several criteria of overall risk of
the HWI and the entities they control. We then test if the 'objective' criterion of
dollars at risk adds any additional information to the more 'subjective' assessment
of risk.
The list of 207 potential risk issues used in this analysis has been until now
highly protected. While not all issues are identified specifically, it is now possible
to release results that reveal at least the issues that turned out to be important in
these analyses. The collection of risk ratings, which are the basis of these
analyses, has now been suspended and superseded, so the definition of issues
discussed in this chapter does not provide useful information to tax advisers.
Even at a more conceptual level, some of the issues, which we find to be serious
risks for 1997-98, are no longer risks as a result of tax reform.
Our objective with this research is not to revise a list of 207 risk issues that
have already been revised, but rather to seize the unique opportunity the data
provides and to explore the kinds of risk factors that predict the existence of the
highest levels of overall risk. That is, what are the issues which, when rated as an
indication of high risk, are a red flag for the existence of many other high risks to
the revenue?
Analysts rated risks in comparison with other HWIs, not general taxpayers. If
this were not the case, virtually all HWIs would have attracted a large number of
high risk ratings based simply on their wealth, income levels, and the complexity
of their business dealings. This outcome would have been unproductive for
determining where the greatest risks lay. As a result, and because not many
issues were rated by analysts, it is rare for a HWI to be given a risk rating over 7
on one of the 207 issues.
Table 10.1 shows the 13 issues most commonly given a rating greater than 7. In
addition, the ratings of the 235 HWIs on each risk have been averaged and appear
in the second column alongside the number of HWIs scoring in the extreme
categories. White these statistics suggest a somewhat different ordering of risk,
priority tends to be given to use of extreme categories rather than average rating.
Tax authorities are more interested in the extreme risks than in average risks, as
they only have the resources to deploy audits and other strategies against the
highest risks.
The most common risk issue, rated over 7 in only 12 cases, was the utilisation
of revenue losses (as opposed to capital losses) through transfers within the
HWI's group of entities - moving losses to a taxable entity to save tax for that
entity (Issue 1, Table 10.1). The second most common high risk was the use of an
offshore entity (company, trust, partnership) in an unlisted (that is, low taxed)4
country (Issue 2, Table 10.1).
There are seven instances of revenue loss creation via research and
development (R&D) deductions (Issue 5, Table 10.1). R&D investment receives a
concessional tax treatment and has traditionally been used in tax minimisation
arrangements (Issue 6, Table 10.1). There are seven instances of 'taxable
distributions to a loss entity', which is the inverse of transferring a loss to a
taxable entity. The term 'distribution' means it is limited to trusts in this
instance.5 If used unscrupulously a loss can be transferred a number of times and
can be used in each entity it passes through. This is known as 'loss cascading' and
has been considered a serious risk to the revenue.
There are eight cases of 'debt forgiveness/bad debts' (Issues 3, Table 10.1). This
technique can be used to create a revenue loss within a group of companies. If
one company fails to repay a loan, the company that made that loan can claim it
as a bad debt for tax purposes. A bad debt is tax deductible, and, when
deductions exceed income, a loss is created. An 'asset disposal' can create a
capital loss if the asset is sold for less than the purchase price. Purchase price can
be manipulated when the same person ultimately owns the company that is
selling the asset and the company that is buying the asset. In this way an
artificial, and tax deductible, capital loss will be created within the group with no
corresponding economic loss to the taxpayer. There are also eight instances of
'other significant deduction issues' (Issue 4, Table 10.1). These 'other issues' will be
discussed later in this chapter.
Table 10.1 also shows the recorded mean dollars at risk (e.g., the dollars
transferred within the HWI group) for each of the common high risk issues. The
dollars at risk are the maximum dollars at risk in either 1997 or 1998. Note this is
not the dollars at risk when the risk was rated over 7; this would be a higher
number. The mean dollars at risk number is the mean dollars at risk for this risk
across all ratings of risk from 1 to 10 (regardless of whether it was high or low).
Note also that the mean dollars at risk will include zero entries if they are
recorded. However the value of zero may mean no dollars were at risk or that the
analyst did not know the dollar amount for some HWIs or the entities they
control. The high value of $46 million for those placed in the miscellaneous 'other'
category is significant and will be interpreted later in this chapter.
Table 10.2 shows the issues with the highest mean dollars at risk across all 207
issues (as opposed to just the highest risk issues). Transfer of trading stock has an
average of $379 million at risk. This could be indicative of a transfer pricing issue,
but there are only three HWIs with recorded dollars at risk for this issue. In
general, the estimates of the mean dollars at risk should be treated with caution.
For example, if an entity is only partially controlled by a HWI there is a
possibility that the analyst has included only a portion of the actual risk.
The initial objective was to use a cluster analytic procedure on the HWI risk
factors to show which risks went together, in hope of revealing systemic risk
factors that were underpinning a variety of seemingly unrelated risks. It was
thought that risk clusters might be particularly useful in targeting types of
specific purpose audits. For example, risk cluster A should get audit product X;
risk cluster B, product Y; and so on.
Unfortunately, it eventuated that cluster analysis was inappropriate because, as
described above, high-risk ratings are rare events with joint occurrence of any
two specific high risks being even lower. In fact there are only two pairs of issues
on which four HWIs are both rated as a high risk; all the other pairwise
combinations have fewer HWIs sharing a high risk on both issues.
Thus, cluster analysis and other pattern finding methodologies are
inappropriate for this data set where the emphasis is on high risk. Empirically
this suggests that if a 'risky' taxation strategy is measured as a combination of
issues assessed as high risk, then these strategies tend to be unique to each HWI
or rather shared by very few HWIs.
Due to this finding, the analysis shifted to identifying specific risks that are the
best red flags for high overall risk. A 'high' overall risk assessment is defined with
a clear action orientation. It means the analyst and their supervisor agree that, all
things considered, the application of some kind of audit product is justified. Being
placed in the high overall risk category is still a rare event. This characterised
only thirty-three HWIs (14%). There were six HWIs who were not assessed as
being in either a high, medium or low risk category. They were treated as not a
high risk and included with medium and low risk HWIs for the purpose of the
following analysis.
Table 10.3 shows the summary from a logistic regression analysis predicting
high overall risk. The independent variables in the model are the 207 issues coded
as 1 for a high risk (that is a rating greater than 7) and zero (that is a rating of 1 to
7). Issues coded like this are referred to as flags to differentiate them from rated
issues. Because of the interest in choosing a minimum set of 'flags' we chose a
stepwise algorithm, which selects the best predictor of likelihood from amongst
the 207 flags, then the second-best after the first-best has been included, and so
on. The stopping criterion is when the addition of further flags does not
sufficiently increase predictability of membership into the correct risk group to
warrant the added complexity. A scree slope test based on the percentage
correctly predicted was used to define this stopping point. Thus we attempted to
balance statistical and practical significance.
The first predictor of high overall risk is being rated over 7 in the 'other
miscellaneous items - other' category (Issue 9, Table 10.1): The potential meaning
of this finding is discussed in the next section.6 The issue rated over 7 with the
next most predictive value in this model is 'use of an offshore entity in an
unlisted country' (Issue 2, Table 10.1). We might assume that ratings over 7 will
mostly be incurred when there is concern that the unlisted country is a tax
haven. The third most useful predictor is 'trust distributions - capital distributions
in cash (to the HWI)'7 (Issue 13, Table 10.1). Various other analyses were
undertaken to confirm the stability of these results, and this flag also turned out
to be important in these supplementary analysis. For example, 'trust distributions
- capital distributions in cash (to the HWI)' is the top predictor in a model
predicting the sum of all risk ratings (0-10) across all 207 issues. Indeed all three
flags in Table 10.3 are recurrently useful and statistically significant predictors
across different types of analyses. Column two in Table 10.3 shows the Likelihood
Chi-square statistic for the inclusion of the flag at each step of the model building
process. Beyond this set of flags, the increase in correctly predicted high risk
HWIs was small (even though other flags subsequently entered into the model
were statistically significant).
The percentage in the fifth column of Table 10.3 shows that with information
about whether or not each of these three issues is rated over 7 (and with no other
information about the case), we can correctly classify the case as High or not-
High 91.9 per cent of the time. Of the 202 HWIs in the low to medium risk group,
201 were correctly predicted. Of the 33 HWIs in the high risk group, 15 were
correctly classified into the high risk group.
Table 10.3 Predicting high overall risk with flags* using logistic regression analysis
Chi- df % Correctly
Step Flags
square Sig. Classified
1
1 24.557 88.5% Other miscellaneous items - other
0.000
1 Other miscellaneous items - other Use of an
2 21.402 90.6%
0.000 offshore entity in an unlisted country
1 Other miscellaneous items - other Use of an
3 14.894 91.9%
0.000 offshore entity in an unlisted country
Trust distributions - capital distribution in cash
(to the HWI)
While it may seem obvious that these would be crucial issues, there may be any
number of sets of completely different issues that would generate the same
reaction. The question is which of the 207 risk issues we really would have
selected as the top three predictors of overall risk. In our experience, auditors
have hunches about red flags of risk that are in no way confirmed by these data.
Some auditors believe that the use of a company controlled by a wealthy person
to own a luxury yacht, a holiday home, or a racehorse is a red flag. Although
there is some support for this view from the data, knowledge of risk on any of
these items classifies only two HWIs in the high risk group. Given the time it
takes to get on top of the complexity of cases such as these groups of HWI
entities, even very senior analysts may have a remarkably narrow scope of
experience. While their supervisors might have greater breadth of experience
built by approving the analyses of others, they lack the depth of experience of the
analyst who has pored over the case. Hence a method that takes at face value that
depth of experience, but aggregates it over the breadth of (235) cases, is valuable.
Although less obvious, use of statistical reasoning can also produce results which
are important though less intuitive.
Table 10.4 Predicting high overall risk with flags* using proxy analyses
Share trading - a group trust received substantial number of company X shares in float. Media reports
indicate HWI made substantial profits through the sale of company X shares. Trust accounts do not
reflect this amount.
While the first example involves a small amount of money we do not know
whether the analyst knows something about this financier that causes the high
risk rating. Perhaps the second example represents a kind of turning over of a
seemingly inoffensive rock where something dangerous lurks. Perhaps the whole
list of ten 'other' concerns conceals a detective's intuition about an inchoate
underlying pattern of risk. We cannot be sure either way.9 For example, the HWI
Taskforce staff suggested that in trying to make sense of losses, analysts might
well say: 'There's a big loss but I'm not sure of the source. I'll put losses down as a
risk. There is something else going on so I'll record it in the 'other' risk category.
The case looks like a high risk overall. I'll get to the bottom of it when I audit
them'. At the very least, we should read the strong predictive power of 'other
miscellaneous items - other' as a caution about relying wholly on pre-defined risk
issues for identifying aggressive tax planning. Rather, it encourages intuitive
detective work to follow risks that fall between the cracks.
Table 10.5 Predicting the total number of high risks with flags* using ordinary least squares regression
analysis
Model R2 Flags
When proxy analyses are used with this alternative definition of risk, the
conclusion is that little additional insight is afforded. 'Investments held off-shore
in a listed country' is the top predictor if 'capital loss creation via asset sales' is
removed from the potential flags. Unfortunately it may not be of practical
significance as a red flag since only one HWI was rated a risk (greater than 7) on
this issue. However not only was this HWI rated a high risk, they had the highest
number of high risks (31); just over one and a half times their nearest competitor
(19) in terms of the number of high risks!
The three best predictors when used on their own are 'capital loss creation via
asset sales', 'bank accounts and investments held in a listed country' and 'trust
distributions - taxable distributions to a loss entity', each accounting for
approximately the same proportion of variance in the dependent variable.
Table 10.6 Predicting high overall risk with dollars at risk using logistic regression analysis
Chi- df % correctly
Step Dollars at risk for issues
square Sig. classified
1 Trust beneficiary loan accounts -credit balance and
1 15.622 87.7%
0.000 no draw-downs10
1 +Trust distributions - capital distributions in cash
2 12.540 88.5%
0.000 (to the HWI) (Issue 13)
1 +Property held offshore in listed country - real
3 9.977 90.2%
0.002 estate
1
4 9.408 90.6% +Trust distributions to charitable trust
0.002
5 7.790 1 90.2% +Utilisation of revenue losses via trading activities
0.005
1 +Related party transactions with an entity in an
6 5.348 90.6%
0.021 unlisted country-royalty/ license income
1 +Use of company shareholder loan account -
7 5.037 91.1%
0.025 repayments by HWI
1 +Related party transactions with an entity in an
8 4.909 91.5%
0.027 unlisted country -provision of finance
1 +Revenue loss creation via debt forgiveness/bad
9 9.422 93.2%
0.002 debts (Issue 3)
1 +Trust distributions - capital distribution in cash
10 5.262 93.2%
0.022 (to group entity)
1 +Utilisation of revenue losses via intra group
11 11.708 94.9%
0.001 transfers (Issue 1)
Why would the dollar value of offshore real estate investment not in a tax haven
in a listed country help to predict overall risk? Firstly we need to consider the
data. There are only two HWIs with dollar amounts ever recorded for offshore
real estate investment in an unlisted country (a tax haven), so it is not surprising
that this issue was not an important predictor. There were not enough cases to
pick up any effects. Real estate in tax havens might have. However, there are
thirteen nonzero recorded dollar amounts for the issue 'property held offshore in
a listed country - real estate' (twenty-eight HWIs were rated on this issue with
two being rated as a high risk). The two largest dollar amounts were both HWIs
where overall they were rated a high risk. This issue then, manages to capture
just a few more of the high risk group over and above the first two issues in Table
10.6. On its own, or as is sometimes said, as a direct effect, it is only marginally
significant with fifty other issues being more useful. Therefore, although the
dollars at risk for this issue adds additional predictive power to the model after
the first two issues are included, it is not particularly useful in the predictive
sense on its own. How far to speculate an explanation of the predictor is not a
simple question. Usually, if the purpose of modelling is explanatory, then
theoretical considerations would be used in defining a model rather than an
algorithmic approach as used here. That said, it may be reasonable to hypothesise
a relationship of some sort between the finding that large real estate investment
in a listed country is a high risk, and the earlier finding that a high risk rating for
bank accounts and investments held in a listed country were two of the four best
independent predictors of the total number of high risk issues. Christopher Skase,
the major Australian corporate criminal who fled to Spain, is not included in
these analyses. But he is a well known case with large real estate investments and
bank accounts in a listed country.
The fourth predictor shown in Table 10.6, 'trust distributions to charitable
trust', is also not included in Tables 10.3 or 10.4. In Table 10.7, where we use
dollars at risk to predict the total number of high risk issues, another issue -
neither in Table 10.3 nor 10.4 - crops up, that is, 'revenue loss creation explained
by negative gearing'. The HWI with the largest dollars on this issue,
approximately $50 million, was also rated overall a high risk, had the second
largest number of high risk issues, and was rated high on 'other miscellaneous
items - other'. Taking all this data together, although interesting, is likely to be
indicative of the boutique strategies mentioned earlier, rather than useful for
identifying red flags of more general predictive value.
When we spoke to HWI Taskforce staff about these issues, their reaction was
the same as ours: 'Analysts would not think this was a risk in itself. However,
there is the interesting question of whether tax planners exploit this reasonable
kind of expectation. It may be that if you want to move an amount of money that
is too large to hide, it is best to make it visible through a vehicle that makes it
appear unexceptionable. Hence, it could be that having offshore real estate is not
a risk factor, but placing huge amounts of money into these vehicles should put
analysts on the alert. While negative gearing is a perfectly legitimate and normal
commercial practice, extraordinarily large losses created by negative gearing may
give reason to be watchful, though one would expect even aggressive players to
exhaust legitimate deductions like negative gearing before moving on to more
doubtful techniques.
Table 10.7 reports the predictor issues whose dollars at risk predict the number
of high risk issues. One reason for these analyses is to see how well the dollars at
risk for the different issues predict high risk, and which ones are the best
predictors. Another reason is to provide a possible corrective for the subjective
nature of the assessment of the ratings used in the first part of the paper. There is
concern that there may be consistent analyst bias in the ratings data since both
the aggregate estimates of risk (overall risk rating and number of high risk
issues), and the estimate of risk on the issues, are a subjective assessment carried
out by the same assessor and may be prone to the same bias. As a hypothetical
example, consider the analyst who rates businessman X incorrectly high on
certain issues because of bias or error; the same bias or error may result in the
businessman being rated as a high risk overall. The role of the supervisor in these
cases will eliminate some of the more obvious errors, but not all.
Table 10.7 Predicting the total number of high risks with dollars at risk for each issue using ordinary
least squares regression analysis
Adj.
Model R2 Flags
R2
1 0.503 0.501 Revenue loss utilisation via intra group transfers (Issue 1)
2 0.612 0.609 Revenue loss utilisation via intra group transfers (Issue 1)
Revenue loss creation via negative gearing
3 0.677 0.673 Revenue loss utilisation via intra group transfers (Issue 1)
Revenue loss creation via negative gearing
Trust distribution - capital distributions in non-cash (to the
HWI)
4 0.712 0.707 Revenue loss utilisation via intra group transfers (Issue 1)
Revenue loss creation via negative gearing
Trust distribution - capital distributions in non-cash (to the
HWI)
Use of shareholder loan account - repayments by HWI
5 0.732 0.726 Revenue loss utilisation via intra group transfers (Issue 1)
Revenue loss creation via negative gearing
Trust distribution - capital distributions in non-cash (to the
HWI)
Use of shareholder loan account - repayments by HWI
Trust distribution - capital distributions in cash (to the HWI
family member)
It is difficult to know where to draw the line in our interpretations since often
the number of cases is low, both in the number of issues where dollars at risk are
recorded and the number of HWIs who are assessed as being a high risk.
However the sparsity observed in this data set is very characteristic of rare events
data. If the notion of an evolutionary ecology of tax planning with the occurrence
of stable and new niches is apt, then the prediction of a small number of rare
events is important to keep abreast of new strategies.
The only issue that is consistently listed in Table 10.6 and 10.7 and the earlier
tables where risk ratings were used is 'utilisation of revenue losses via intra group
transfers'. This is actually the top predictor for the total number of high risk
issues (see Table 10.7). Since this is also recurrently predictive in the subjective
analyses, 'utilisation of revenue losses via intra group transfers' was considered a
definite contender for being a red flag issue. Of the 207 issues, it is also the one
that most commonly has a high risk rating (see Table 10.1). This makes it a
decidedly useful red flag for our final models.
In practice it appears that the two types of analyses described, those using the
estimates of risk rating and those using the dollars at risk estimates, are capturing
different aspects of the risk prediction process.
Focusing on issues from the preceding tables, we examined the relationship
between risk rating and dollar ratings. Although there is some evidence that the
higher risks had higher dollars associated with them, it was by no means
universal nor unambiguous as many issues had zero or low dollars associated
with high risk and vice versa. Thus, we need to interpret the dollars knowing that
low dollars at risk do not necessarily signify low risk in the view of the assessor.
Dollars at risk for 'revenue loss creation via negative gearing' is also a good
example of the disjuncture: Assessors rarely considered the issue a high risk but
unusually large dollar amounts recorded for this heading was predictive of high
risk.
This chapter focuses on the subjective ratings and their use in predicting high
risk. It is therefore of interest to ask if the dollars at risk for any issue adds
significantly to the prediction after the flags from Table 10.3 have been included.
This was analysed several ways and the dollars at risk for 'property held in a
listed (non-tax-haven) country - real estate' consistently added to the prediction
of high overall risk after subjective ratings were included. From our analyses it
would appear that the subjective, informed assessment of the analyst is a useful
approach to detecting risk, with the quantitative data providing further clues to
follow. This is contrary to the notion that the dollars at risk, being a more
objective measure, would have superior predictive capacity to predict: Better to
analyse subjective risk assessments first, then add some value with a diagnosis of
unusually large dollar amounts.
Our experience of the regulatory craft (Sparrow, 2000) leads us to suspect that
analysts need both the skills of a detective and those of an accountant. Nuance of
judgment is needed; things have to pass the smell test. When the good analyst
smells a rat, they are more likely to chase it down a hole than to further analyse
numbers. While quantitative analysis provides clues, systemic wisdom must then
be applied to the specific case. That case should be looked at through many
different lenses, of which the quantitative lens is only one.
These data suggest to us that there might be a case for a two-step process with
aggressive tax planning analysis. Step 1 is a qualitative diagnosis of returns and
other intelligence surrounding the case that is informed by knowledge of red
flags of systemic risk. Step 2 is a re-examination of objectively unusually large
amounts of money that occur under labels that are not viewed as red flags, but
that might become red flags when the dollars in them are extraordinary. Step 2
does not mean further interrogation of all extraordinary amounts. It means a
harder look only when exceptionally large amounts occur under non-red-flag
issues that the kind of analysis in Tables 10.6 and 10.7 reveal to be significant
predictors of overall risk.
The analyses reported in this chapter only take us as far as making some
practical suggestions for the identification of red flags for Step 1. Step 2 would
only make sense to become a standard methodology after complementing the
quantitative analyses in Tables 10.6 and 10.7 with qualitative intelligence on the
operations of aggressive tax planners.
Final Red Flags Models
Having carried out the above analyses, our final task is to convert the results
shown in Tables 10.3 to 10.7 into a more useful form for discussion and practical
application. It is clear that most of our potential red flags were consistent with
what we referred to as standard niches. Other issues appeared to be indicative of
more boutique niches. It has further been demonstrated that the number of HWIs
rated as high on any one issue was small, which means that single high risk
issues are a narrow basis for selecting cases for audit.
Could the results be made more useful for auditors by aggregating some of the
issues that rarely occur as high risks? Consider 'capital loss creation via asset
sales'. When the risk under this issue is rated high, the number of other high risk
issues (that is, the number rated over 7) averages 16 (compared to a mean of 0.86
when 'capital loss creation via asset sales' is not high). But there were only four
HWIs with a rating over 7 on 'capital loss creation via asset sales', so the level of
risk on this particular issue is sufficiently rare that it will be of limited value in
audit practice.
Consequently, we created an aggregated issue of wider scope. This issue was
'capital loss creation' instead of the narrower (but more powerfully predictive)
issue 'capital loss creation via asset sales'. The broader issue was a composite of
nine different kinds of capital loss creation. Hence, if there was a rating of more
than 7 for any single capital loss creation issue, the red flag was put up - whether
the loss was created by asset sales, property development, debt forgiveness, bad
debts, takeover/acquisition/merger, cost base manipulation, artificial loss
duplication, related party transactions or some other explanation, or even if the
capital loss creation was unexplained. Note, however, that it is not so broad as to
include revenue loss creation; it is a capital loss creation variable only.
The second recurrently predictive red flag that we broadened was trust
distributions. While 'trust distributions - capital distributions in cash (to the
HWI)' was the most predictive red flag here, twenty different kinds of
distributions from trusts were combined in the composite red flag. This included
distributions in cash and non-cash to the HWIs themselves, to HWI family
members, to group entities, to charitable trusts, to a loss entity, and from
probate/deceased estates, related entities, and so on. There were from as well as to
distributions. Finally we broadened 'other' to include all sixteen 'other' issues
including 'other significant deduction issues'.
Table 10.8 shows the summary results from a logistic regression predicting
high overall risk and an ordinary least squares regression predicting the total
number of high risk issues (i.e., with a rating over 7) with our set of red flags
which now include the broadened issues. What we are doing in Table 10.8 is a
rerun of the analyses in Tables 10.3 and 10.5, but with 'other', 'capital loss
creation' and 'group trust distributions' as broader red flags with many more high
risk cases.
Table 10.8 Summary results of the final models predicting risk using original and aggregated red flags
The adjusted R2 for the final red flags regression model predicting number of
high risks is an impressive 0.82.12 Also these red flags, when used to predict
overall high risk, yielded a slightly higher classification rate of 92.8 per cent
(compared with the three predictors in Table 10.3). It is of note, however, that the
final red flags model actually correctly predicts more HWIs into the high risk
group but at the expense of incorrectly predicting some of the other HWIs as
being high risk.
Nonetheless, we can conclude from these analyses that the final aggregated red
flags, which are arguably better for auditing purposes, do as well or better than
our previous sets of red flags. Therefore, there appears to be value for auditors in
aggregating some of the issues that rarely occur as high risks.
Of the original set of red flags, two make a contribution in these final analyses,
'utilisation of revenue losses via intra group transfers' and 'use of an offshore
entity in an unlisted country'. The former, although significant, is not particularly
useful in predicting the high overall risk group, but it does significantly increase
the amount of variance explained when predicting the total number of high risk
issues for a HWI. Hence we have identified five final red flags for overall risk: (a)
trust distributions; (b) capital loss creation; (c) use of an offshore entity in a
country that may be a tax haven; (d) utilisation of revenue losses via intra group
transfers; and (e) extraordinary risks that fall between the cracks of the other
risks.
Further analyses showed that the five subjective red flags were of more
importance in predicting high overall risk than any of the dollar amounts
associated with the red flags. The only red flag where the associated dollars at
risk added significant extra predictive information above that of the subjective
ratings greater than 7 was 'utilisation of revenue losses via intra group transfers'.
This was only the case for the model predicting number of risks greater than 7.
Conclusion
The five red flag issues are consistent with what many would regard as the
fundamentals of aggressive tax planning. Fundamentals here is used in opposition
to the idea that the best red flags would be telling symptoms of deeper problems
(e.g., converting private pleasure activities into tax deductions - pleasure craft,
horse breeding, racing deductions, for instance). The verified red flag issues are:
(a) trust distributions (especially capital distributions in cash to the HWI); (b)
capital loss creation (especially through asset sales, but not revenue loss creation);
(c) use of an offshore entity in a country that may be a tax haven; (d) utilisation
of revenue losses via intra group transfers (that is, within the group of entities
controlled by the HWI); and (e) 'other' extraordinary risks that fall between the
cracks.
The surprisingly robust results on the 'other' measures may suggest that we do
not always know what the emerging fundamentals of the future may be. These
results are not interpreted as anomalous, but rather as suggesting an evolutionary
ecology of tax planning. Tax planning strategies that everyone, particularly the
ATO, knows about will not be the most lucrative. While there will be recurrent
predation strategies, the best new strategies will be those that are not crowded
out by others who use a similar strategy. Minority strategies flourish. We
therefore caution against the idea that we can settle in advance all risk categories
for aggressive tax planning. We also highlight the importance of intuitive
detective work to follow risks that fall between the cracks. This advice follows
not only from the importance of the 'other' category, but also the result that the
estimated 'objective' dollars at risk added little explanatory power to the ability to
predict high risk above and beyond that provided by subjective risk ratings by
ATO analysts.
An evolutionary ecology of tax planning implies that some successful players
will seek new niches. Financial engineering of new derivatives never conceived
before by tax law and global capital mobility make this more possible than in the
past (Department of the Treasury, 1999). As the law adapts to close off new niches
the change in the tax law environment may also create new niches for other tax
strategies. Law makers are less and less able to control these unintended effects as
tax law changes in other countries create new niches available to local aggressive
tax planners. The aggressive tax planner benefits from both an expanding range
of niches globally, and expanding technical capabilities for local financial
engineering. The first four red flags listed here may cover the standard niches
while the last one covers the new niches.
The analyses also suggest a subset of cases that should be examined closely to
see if there is evidence of new and previously unseen strategies or niches. These
are cases where extraordinary dollar amounts are seen under issues that are not
normally red flags, especially in those instances where other red flags suggest
systemic risk.
Some aspects of the red flags we have identified should cease being indicators
of risk to the revenue in future as Australian tax reforms have been undertaken to
specifically deal with some of these problems. To this extent, the findings suggest
that tax reform was well directed. For instance, the new loss integrity measures
should prevent the 'cascading' of losses and will address some, but not all loss
creation issues. Corporate consolidation measures are intended to limit the
'trafficking' of losses from outside the company (but they will not affect the
genuine transfer of losses as this is accepted as part of normal business practice).
The proposed entity taxation regime, if adopted, will tax non-fixed trusts as
companies and treat trust distributions as dividends, which would remove much
of the tax effectiveness of trusts. This regime, however, has been put on hold at
the time of writing. While not such a recent development, the controlled foreign
corporations law was introduced to deal with tax haven use. Initially payments to
these countries decreased, but in its annual report, the ATO notes that transfers to
tax havens have increased significantly since 1996 (Commissioner of Taxation,
2000). Perhaps some aggressive tax planners have found a way of circumventing
this legislation. In any case it appears that tax haven use remains a high risk area
(Tanzi, 2000).
In a dynamic ecology of aggressive tax planning enforcement, tax
administrators must adapt to cut out old risk factors just as they must prepare for
mutations into new ones. Some of the most recurrent forms of aggressive tax
planning are variations on fundamental themes that have been with us for many
decades. Yet many of the mutations that seem so new retain the character of
those fundamental strategies. A clever new piece of financial engineering may be
clever and new at the same time as it is just another way of shifting losses, even
if it cannot be recognised at first as loss shifting.
Notes
1 The ATO acknowledges the comments made in this paper and notes that the conclusions reached are
based on aged data. The practices employed by the HWI Taskforce have significantly changed and
the conclusion reached may not be the same if this analysis were undertaken today.
2 For instance, the ATO conceives of aggressive tax planning as techniques which: (a) undermine the
policy intent of the law; (b) impact on the integrity of the tax system; and (c) erode community
confidence in the fairness and the equity of the tax system. Characteristics which mark aggressive
tax planning in the ATO's view include arrangements which: (a) are contrived and artificial in their
method of execution; (b) are uncommercial from a business or economic perspective; (c) are not
implemented as specified in contractual and other legal documentation; (d) involve round robin
finance or circular movement of funds and loans paid off by future earnings; (e) involve fraud on the
revenue; (f) involve permanent tax advantage as distinct from a timing advantage; (g) abuse a specific
concessional or anti-avoidance provision contrary to the policy underlying the law; and/or (h)
attempt to reduce the amount of tax properly payable and thus create a revenue risk.
3 For example, in some instances analysts have included a risk more than once simply because it is
feasible that the risk could fall under more than one category.
4 The Income Tax Regulations include a 'list' of countries which are considered to have tax systems
broadly similar to the Australian system. There are fifty-eight of these 'listed' countries. 'Unlisted'
countries are all those that do not appear on the list. They have taxation systems that range from
almost comparable to full blown tax havens. Hence 'use of an entity in an unlisted country' would
imply that profits were somehow derived in a low taxed country, and that a reduced amount of tax
would be paid on them.
5 Income from trusts comes in the form of a distribution: From companies it is a dividend.
6 The individual risks named in the data are loosely grouped under five broader risk categories. The
broad risk categories are International, Losses, Group Structure, Miscellaneous Income and Deduction
Items, and Other Miscellaneous Items. The broad categories in turn contain more specific risk
categories. Each of the specific risk categories contains a catch-all variable named 'other' to
accommodate the less common or less understood risks of that type. As the broad category 'other
miscellaneous items' has itself the flavour of a catch-all category, one would expect the least common
or least understood risks to appear under the label 'other miscellaneous items - other'. Oddly, this has
turned out to be the most highly predictive variable.
7 In this sense 'capital' is the money contributed to set-up a business or keep it running. Unlike profit, it
is not taxable when it is returned to the contributor (because it is assumed to have been taxed
previously). There are obvious tax benefits to the HWI if they can re-characterise profit as capital.
8 It is also a top predictor when we predict the sum of all risk rankings (0 to 10) or the number of issues
where the risk score exceeds 7.
9 One risk that was classified as 'other' that was more or less systemic was property developers treating
building costs as deductible on the revenue account while treating income from sales as capital. This
practice gives the advantage of immediate deductions for expenses but a concessional treatment of
income. It is not allowed by the legislation, but it appears that taxpayers 'try it on' in the hope the
ATO will not detect it. It has since emerged that it is widespread among property developers.
10 Loan accounts are often used by HWIs as a tax free source of cash to maintain their lifestyles. A credit
balance in a loan account would suggest that there was another source of income and analysts could
find this a cause for concern.
11 Note also the predictive power of the dollar amounts of 'trust distributions - capital distributions in
non-cash (to the HWI)' in Table 10.7.
References
Braithwaite, J. (2001), Through the Eyes of the Advisers: A Fresh Look at the
Compliance of High Wealth Individuals', Centre for Tax System Integrity
Working Paper No. 22, Australian National University, Canberra.
Cohen, L.E. and Machalek, R. (1988), 'A General Theory of Expropriative Crime:
An Evolutionary Ecological Approach', American Journal of Sociology, vol.
94, pp. 465-501.
Commissioner of Taxation (2000), Annual Report 1999-2000, Commonwealth of
Australia, Canberra.
Department of the Treasury (1999), The Problem of Corporate Tax Shelters:
Discussion, Analysis and Legislative Proposals, Department of the Treasury,
Washington D.C.
Sparrow, M. (2000), The Regulatory Craft, Brookings Institution Press,
Washington D.C.
Tanzi, V. (2000), Globalization, Technological Developments, and the Work of
Fiscal Termites', International Monetary Fund Working Paper WP/00/11811,
International Monetary Fund, Washington D.C.
Chapter 11
When Compliance is not the Solution but
the Problem: From Changes in Law to
Changes in Attitude
Doreen McBarnet
Creative Compliance
Two factors contribute to the practice of creative compliance. One significant
factor is the nature and operation of law itself. The law-making process leads to
lobbying and compromise, legislators cannot address every contingency that
might arise, drafting is fallible. More fundamentally, it is in the nature of law that
it is open to different interpretations, and that its meaning and application are
arguable.
Creative compliance, however, does not arise deterministically from the nature
of law. It also requires a particular attitude to law, an attitude which, far from
seeing law as an authoritative and legitimate policy to be implemented, sees it as
a material to be worked on (McBarnet, 1984), to be tailored, regardless of the
policy behind it, to one's own or one's client's interests. And it requires active
legal work.
Law is open to alternative interpretations. Innovations in practice can leave
law behind. But grey areas, alternative interpretations and innovative legal forms
do not only arise 'naturally'. Rather, they may be motivated precisely by the
desire to outflank the law. Creative compliance involves careful scrutiny of law in
order to seek out material for and actively construct alternative and innovative
arguments and legal forms. Creative compliance involves seeking out: (a) gaps
facilitating the 'where does it say I can't do that?' argument; (b) the ex-files of
law, expressing exemptions, exceptions, exclusions, with practices then
restructured to fit within them; and (c) rules, the more prescriptive and rigid the
definitions and thresholds involved, the better, with legal forms adopted to fit
inside or outside their literal ambit (a practice of working to rule).
Regulators often express concern about uncertainties in the law being
exploited,6 but creative compliance operates particularly effectively in the context
of a rule-bound regime, where the words of the law can be treated as recipes for
avoidance.
How this works in practice can be readily demonstrated from research. When
it was proposed in the UK that value added tax (VAT) should be levied on
domestic fuel, hitherto exempt, there was extensive protest on the basis that the
old and the poor would suffer. In the event, not everyone suffered. Some
institutions, including university student residences, found a nice way to avoid
the new costs involved in heating and hot water. They would sell their boilers,
which of course remained exactly where they were in the institution's basement,
to a separate company. The company would buy the fuel to heat the water. The
company would have to pay VAT on the fuel but, as a commercial enterprise,
could reclaim it. The company then sold the water to the institution. There is as
yet no VAT on water so, hey presto, nobody pays any VAT. And the 'where does it
say I can't do that?' argument, along with a quick referral to the rules, can be
brought into play to justify it.
UK national insurance is currently being avoided in the North Sea oil industry
by transferring UK staff to overseas agencies. Transnational tax avoidance by
using tax havens or differential rules in different jurisdictions (for example, the
'Delaware Link' device (McBarnet, 1992)) is common practice. But avoidance can
also be readily accomplished within a jurisdiction simply by working to, or
playing with, the rules. For example, to avoid UK national insurance
contributions (national insurance contributions being paid by employer as well as
employee), a practice was developed in the financial sector of paying part of the
salary in the form of large bonuses in shares. No national insurance contributions
were payable on shares, though they had a clear monetary value and could be
converted immediately into cash. To deal with this, the law was changed to
impose national insurance on any payment in a 'traded commodity on a
recognised exchange'. The result was a shift to payments in fine wine, also of
clear monetary value, also amenable to immediate conversion to cash, but not
traded on a recognised exchange.
Creative compliance is pervasive in corporate practice in many areas of law
and has, over time, involved a vast array of techniques (see Griffiths, 1986, 1995).
Take an example from my research with Christopher Whelan on creative
accounting, one we have frequently cited because it demonstrates so clearly the
problems posed by creative compliance. This specific example, the 'orphan
subsidiary', was widely used in the 1980s to cosmetically enhance a corporate
group's paper profits and assets, with knock-on effects for share prices,
performance-related pay and borrowing capacity. A parent company must, under
company law, produce accounts, which include the profits/losses and
assets/liabilities of its subsidiaries. The idea is that it should provide the full
picture for all the companies it controls, and not hide away poor performances or
liabilities in a separate entity whose details it does not disclose. But what if
parent company A sets up a company B which it controls, but which is
nonetheless carefully structured to fall outside the legal definitions of a
subsidiary? In this 'orphan subsidiary' it would have the perfect vehicle to use in
ventures which involved high risk or high borrowings - such as a highly
leveraged acquisition - without the losses or liabilities appearing, detrimentally,
in its group accounts.
The statutory definition of a subsidiary under Section 736(1) of the Companies
Act 1985 involved two criteria. Company B was a subsidiary of company A if (a)
A owned more than half B's equity capital, or if (b) A controlled the composition
of B's board of directors. With a little creativity, both criteria could be readily
circumvented. There were many ways of achieving this, one of the simplest being
to set up two types of shares, ordinary shares and preference shares, 50 per cent
of each. Company A would own the ordinary shares, while its bank would own
the preference shares. Company A would not therefore own more than 50 per
cent of the equity capital. Company A and its bank would each appoint half the
directors. Company A would not therefore control the composition of the board
of directors. But the directors representing the ordinary shareholder (Company
A) would have two votes to the preference share directors' one. Company A
would not control the composition of the board of directors, but it would control
the board's votes.
These are typical examples of creative compliance at work, and they
demonstrate how the material of law is actively used to circumvent legal control.
The constructs that emerge are backed with legal arguments and the opinions of
leading counsel and, if challenged, a case can be produced (however 'bullish' - see
ahead) to claim compliance with the law. And these are merely examples.
Creative compliance is not a practice operating at the statistical margins, or at the
margins of society. On the contrary, it is pervasive, and pervasive among leading
lights in the social and corporate world. The orphan subsidiary, for example, was
just one of a vast range of creative accounting devices, used routinely in the UK
by household name companies.
Problems
Sustaining Principles
This kind of reduction can happen in a number of ways. It may occur through
demands for guidance on how principles will be applied in specific contexts.
FRS5, the Accounting Standards Board's regulation requiring the reporting of the
substance of transactions, not just their legal form, took nine years to reach the
standards book. Demands for guidance produced detailed examples which some
have certainly looked to as new rules and potential material for creative
compliance (McBarnet and Whelan, 1999). What happens in effect is that we get
rules about how principles can be used.
Lobbying may result in negotiated curtailment behind the scenes of what looks
in the books to be big powers. After the 'new approach' in UK tax, for example,
the Inland Revenue responded to lobbying and met with legal and accounting
professional bodies to negotiate the parameters within which they would apply
the new approach in practice.
Something similar can be found in the Australian context in the Ralph Report
(1999). Australia's big power is the General Anti-Avoidance Rule (GAAR), but the
Ralph Report has sought a clear statement that 'the GAAR will not apply to the
mere use in a straightforward and ordinary manner of structural features of the
law to best advantage'. A 'statement of policy should confirm the circumstances
in which the GAAR could be applied and reduce the perception that valid
business practices could unintentionally be subject to the application of the
GAAR' (p. 241). It has recommended that there should be a board to review the
application of the GAAR, including rulings on whether or not a practice is caught
by the rule. It has argued the need for producing clarity without mapping the
minefield (Ralph Report, 1999, p. 44). This unfortunately is not a solution but a
restatement of the problem - just how to produce the clarity requested without
mapping the minefield. There is a real danger that in pursuit of 'clarity', the big
power of the GAAR may be limited to narrower parameters and reduced to rules,
undermining its capacity to override rule-based avoidance.
This narrowing can also be produced through the courts, not just through the
enforcement agency losing but through the process of decisions being made
(whoever wins or loses), reasoning being set out, and new material being
provided for those bent on exploiting creative compliance. In the context of the
new approach to tax avoidance in the UK, new 'rules', and therefore material for
creative compliance, were found by scrutinising the arguments used by the
judges, even in the Ramsay case, which introduced the anti-avoidance principle
of looking through form to substance.
The Ramsay case comprised a circle of 'self-cancelling' transactions used to
create an artificial loss, not unlike the recent Australian 'investment schemes'
(Senate Economics References Committee, 2001), though more complex. The
transactions required two companies and a subsidiary controlled by the taxpayer,
two companies controlled by the scheme promoters (the Rossminster group). It
involved two loans, one share issue, the exercise of options on interest levels (the
interest rate on one loan was changed from 11 per cent to 22 per cent, on the
other from 11 per cent to zero), the sale of the 22 per cent loan to a company
controlled by the scheme promoter, which then sold the loan to the subsidiary of
taxpayer company B, two liquidations, loan repayments, and exchange of shares
for loanstock in another promoter-controlled company. Yet it was, as the judges
observed, 'all over by lunch'. What is more, these multiple deals had 'no business
purpose'. These observations were seized upon and steps taken in subsequent tax
avoidance schemes to factor in a business purpose (with 'careful minuting', as one
interviewee put it, to record it) to build in gaps and contingencies, and to change
the timescale.
Even without resort to courts, precedents build up. In the context of rulings on
the application of the General Anti-Avoidance Rule, it has been observed that 'an
important body of case law is building up' (Clayton Utz, 1999). Rulings or
clearances may be sought informally whether there is a statutory right to them or
not, and interviews with UK tax officials and accountants indicated frequent
requests to the Technical Office of the Inland Revenue for informal advice on
'hypothetical' or 'no-names' transactions. Tax officials had an ambivalent attitude
to such requests, preferring to avoid rulings on such a basis but finding the
requests a valuable source of information on the latest creative thinking. Potential
new material for creative compliance is also produced every time there is a
decision on the part of the agency on how a principle should apply to a specific
situation. Even a simple failure to challenge a practice can be treated as
endorsement by default, and built upon.
Application
Big powers are only as big as their application, and application may in practice be
curtailed by a range of factors. These can be illustrated by looking at the record
of enforcement in the new UK accounting regime.10 The body responsible for
enforcement is the Financial Reporting Review Panel, which came into being in
1991, armed with the extensive new powers listed above; and with a big stick
sanction in the background in the judicial power to make directors personally
liable for all the costs of revising and reissuing accounts which were successfully
challenged by the Panel, along with legal costs. The Panel announced that it
would use the true and fair super-principle to stop creative compliance:
Where we are firmly of the view that accounts are not true and fair, we will not be deterred from taking
action by the fact that there is room for forensic argument as to technical compliance with the particular
FRS [accounting standard] (Financial Reporting Council, 1992, p. 24).
In other words, the Panel would use the true and fair principle to override what it
saw as creative compliance.
But to date this power has never been used for this purpose. Rather, the Panel
has monitored those situations where companies have invoked the true and fair
principle to override compliance with specific rules. Indeed, it has tended to
require companies to adhere to the rules even where there is strong opinion that
following the rules does not produce true and fair accounts. In doing so the Panel
may indeed be damaging its big powers. It may be setting a precedent that
following the rules is more important than following the principle of a potentially
clashing true and fair view. The big power of FRS5, requiring companies to report
substance not just legal form, has been used only once against a small company.
No directors have had to face personal liability costs, because no cases have gone
to court.
It could be argued, of course, that all this tells us is that creative compliance
has died off in the face of the new regime. But there are still many instances of
practices that others see as 'bullish', or 'sailing close to the wind' (Griffiths, 1995;
McBarnet and Whelan, 1999), and there is some danger that by not addressing
them the Panel is in fact legitimising them.
Why then has the Panel been so circumspect in its use of its big powers? For,
although the Panel has big powers in the books, our interpretation of what was
happening was that a kind of 'self-regulation' was taking place, in the sense that
the enforcers themselves were limiting their invocation of the powers at their
disposal. Indeed, reflecting on the position of the Panel, or any agency in the
same position (the ATO with its General Anti-Avoidance Rule, for example), the
fact is that putting big powers into practice is not as straightforward as it might
seem. Indeed, there are risks in using big powers which can foster caution.
There is a risk of 'winning but losing' in court with decisions and reasoning
used in unforeseen but very damaging ways. In the accounting context, the
Argyll case in 198111 involved successful prosecution of company directors for
including in their group accounts a company they had not yet fully acquired.
They argued that they effectively controlled it, it was in substance a subsidiary,
and including it would result in true and fair accounts. The Department of Trade
followed the case with a statement, underlining the message that specific
definitions in the law had to be strictly adhered to. The unintended consequence
was the highly damaging device of the orphan subsidiary, with the case and the
statement pointed to as a powerful basis for arguing not only that the practice
was not illegal, but that directors had no legal alternative but to keep a company
that did not meet the specific definitions of a subsidiary out of their group
accounts.
More generally, going to court means losing control to the judges who may,
whether in ways favourable to the enforcement agency or not, come up with
approaches with complex implications for other instances. And of course there is
a risk of losing, not only opening the floodgates to copycat cases of the same
type, but encouraging avoidance more generally. 'A daft judge can kill a
standard', as one of our UK accounting regulators put it (McBamet and Whelan,
1999, p. 88). The ATO has had experience of this in the past in terms of judicial
treatment of the General Anti-Avoidance Rule. After the new approach cases in
UK tax, the Inland Revenue required tax inspectors not to invoke the Ramsay
ruling without getting central clearance to do so. Inland Revenue did not want
taxpayers challenging them in court on the reach of the new doctrine in the
context of just any case; the new doctrine of 'substance over form' was too
important not to keep it carefully controlled.
Small wonder then if the pragmatic approach is to avoid confrontation in
court. Indeed, big powers may be stronger for not testing them in court. The
(until recently) chairman of the Accounting Standards Board, David Tweedie,
observed of the new principle-based regime:
We're like a cross-eyed javelin thrower competing at the Olympic Games: we may not win but we'll keep
the crowd on the edge of its seats.12
Attitude
Finally, big powers, far from destroying creative compliance, may still fall prey to
it. Even big powers can be treated as 'material to work on'. We have already seen
how the judicial reasoning in the UK's new approach tax cases was mined for
'rules' which could be used to argue that new tax avoidance schemes lay beyond
the reach of the new approach.
Regulations, even regulations geared to principles or anti-avoidance doctrines,
have to be expressed in words, and even the words expressing the big powers to
curb creative compliance can be subjected to the creative and advantageous
interpretation on which creative compliance is based. Consider the orphan
subsidiary. We saw earlier how it was based on careful scrutiny of and adaptation
to very specific definitions in company law. We also saw how the law was
changed. Definitions of a subsidiary under new legislation included the 'catchall'
requirement to include an 'entity' in group accounts if the parent company had 'a
participating interest' in it (a far cry from 'more than half the equity capital') and
'actually exercised control' over it (much wider than 'controlled the composition
of the board of directors'). Legislators set out definitions of a broader, more
abstract nature, and refused to define them further, precisely (and they were quite
explicit about this) in order to stop feeding creative compliance.
Yet creative compliance continued. Even big powers have exemptions and
exceptions ('ex-file' clauses) which were sought out and used. But more
significantly still, the words in the catchall definition were themselves scrutinised
and responded to with the 'deadlocked joint venture'. This involved two
companies forming a 50-50 joint venture in which, it was claimed, neither
'actually exercised control', so that it remained off the accounts of both.
The power to override compliance with specific rules in order to require
compliance with the overarching principle of producing true and fair accounts,
has itself been scrutinised and resisted on the basis that it is complied with if
companies give the 'true and fair' version of their accounts in the notes but leave
the numbers, in the same accounts, based on literal compliance with specific rules
- even though it is accepted that compliance at this level does not give a true and
fair view.14 The numbers of course, are more important than the notes, not only
because the significance of the notes may be lost on the less expert reader, but
because the numbers are what are used in calculating market ratios such as
leverage or gearing. These ratios, in turn, are what are used by analysts in
assessing share value, by banks in lending covenants to stop excessive additional
borrowing by management, or which, when they reach a certain level, oblige
management to consult shareholders. The numbers are, in other words, the basis
of legal controls on a number of fronts.
In short, even big powers designed to counter creative compliance may
themselves be vulnerable to it if they meet with the same attitude to the law - an
attitude that treats the law as merely a material to work on. It is not only how big
powers are applied that determines their impact, but how they are received.
A Change of Attitude
Creative compliance, I suggested earlier, is the product of two factors: the nature
of law and the attitude taken to it. The application of law is problematic, given to
grey areas and alternative interpretations. But the problem posed by creative
compliance depends on those on the receiving end of law actively working on
and taking advantage of those intrinsic problems. The enhanced uncertainty
associated with principles and general anti-avoidance rules is a genuine concern.
Yet the resort to such measures is itself a response to active abuse of more specific
rules. The drift of principles to rules I have designated a 'problem', but it would
not be a problem if those rules were not likely to be actively seized on and used
to escape legal control, and if there were not a culture which treated the law as
fair game for such activity.
What this suggests is that changes in law, however sweeping, are unlikely,
alone, to eliminate creative compliance and the problems it poses for law and
enforcement. The second factor, the attitude to law, needs to change too. The
ATO, and the Compliance Model of law enforcement itself, recognise the need to
foster a cooperative taxpaying culture, and indeed recognise that it is not just law
that has to change, but attitude: 'Changing attitudes to our tax system is the
remaining element that can give a major impetus to achieving a genuine new tax
system in its fullest sense' (Commissioner of Taxation, 1999, p. xiii). The ATO is
currently working to counter non-compliance by fostering an image of tax as a
positive contribution, not as a negative imposition. It seeks to relate tax payment
to the provision of public services: 'Our tax system is important to our
community. It is about education, health treatment, support for those in need and
roads and other community assets' (Commissioner of Taxation, 2000, p. 8). The
ATO, in effect, is appealing for tax compliance as a mark of responsible
citizenship (see also Australian Taxation Office, 2000b). The UK's Inland Revenue
has announced a similar approach. It will be promoting the payment of tax as a
'badge of good citizenship', 'explicitly linked to public goods'.15
But this chapter, and the larger bodies of work on which it draws, indicates
that tackling the attitude to non-compliance, appealing for compliance, is not
enough. Whether compliance is enough depends on how people are complying.
Compliance can itself be a creative construct, and a mark of resistance to tax
policy, not cooperation with it. In aiming to construct a culture of compliance,
then, the aim must be not just a culture of compliance as opposed to non-
compliance, but a culture of compliance with the spirit of the law, rather than
creative compliance with its letter.
Creative compliance is not just a tax problem but a law problem. The ATO is
working to change the general attitude towards tax. But if a change of attitude is
required, it is not just in the attitude to tax, but in the attitude taken to law,
policy and compliance. This is true not just for taxpayers, but for their
professional advisers - the lawyers and accountants whose creative work lies at
the heart of creative compliance. What needs to be fostered is a change of
attitude to the law, in which it is seen not as a game of words, a material to be
worked on to one's own or one's client's advantage, but as an instrument of
legitimate policy to be respected, with the policy, not just the words, looked to as
the measure of compliance. That, I know, is itself problematic in many ways. But
without some shift in that direction, the concern must be that compliance will
remain not a solution but a problem for tax policy and tax enforcement, and,
indeed, for legal policy and legal control in general.
Notes
1 Or if they are lawyers, as in the celebrated recent cases of Australia's bankrupt barristers (Auditor-
General, 1999), set themselves to work.
2 The more so where taxable profits and financial reporting profits substantially overlap. There is
variability between jurisdictions as to how far this is the case. See, for example. Touche Ross (1989).
3 Funded by the Economic and Social Research Council and based on in-depth interviews with lawyers,
accountants and regulators (including the Inland Revenue. Accounting Standards Board, Financial
Reporting Review Panel. Australian Taxation Office), along with key players from business as
relevant.
4 As a Visiting Fellow at the Centre for Tax System integrity, researching the ATO Compliance Model of
enforcement and issues posed for it by legal creativity.
5 With Christopher Whelan. Funded by the Jacob Burns Fund for Socio-Legal Studies and the European
Commission. See McBarnet and Whelan (1999). on which this chapter draws.
6 See for example, Australian Commissioner of Taxation's reference to tax arrangements which seek to
exploit deficiencies or uncertainty in the law' (Commissioner of Taxation, 1999, p. xiii).
7 The Ralph Report (1999) distinguishes tax avoidance as a mis-use or abuse of law rather than a
disregard for it' (p. 243).
8 I have elsewhere described one of the functions of creative compliance as fraud insurance' (McBarnet.
1991).
10 This section draws on analysis based on joint research with Christopher Whelan (McBarnet and
Whelan. 1999).
11 Unreported magistrates court case, but very influential. Reported unofficially in Ashton (1986).
13 See McBarnet and Whelan (1991) for a deeper analysis of 'the discourse of resistance' in relation to
both tax and accounting regulation changes.
14 There was a controversial debate over this in the 1980s between accountants David Tweedie and James
Kellas, on the one hand, and Ralph Aldwinckle of the Law Society, on the other. Even after a change
of statutory wording, the debate continues. See McBarnet and Whelan (1999, Chapter 15) for a
detailed analysis.
15 Financial Times weekend 20-21 January. 2001. quoting Nick Montagu, chairman of Inland Revenue,
who also argued that a similar approach had 'worked in the Netherlands'.
References
The Problem
Australians have become more sensitive to the issue of high wealth individuals
paying their fair share of tax. A June 1998 survey of 1000 Australians
commissioned by the Australian Taxation Office (ATO) found that only 32 per
cent believe 'tax laws are effective in making sure large companies pay their
share of tax' declining to 27 per cent for 'very wealthy people'. Only 20 per cent
believed 'the ATO does a good job of stopping tax avoidance by large companies',
falling to 15 per cent who believed that 'the ATO does a good job stopping tax
avoidance by very wealthy people'. It was this kind of evidence that caused the
Keating government to make high wealth taxpayers an election issue in 1995 and
that caused the Howard government to set up the High Wealth Individuals
Taskforce after its election victory.2
The problem cannot be dismissed as a creation of political imagination: One
adviser in this study said he had a wealthy client who had not paid any tax since
1987. Certain sophisticated tax planning strategies appear to have been
widespread. These include: (a) redefining income as capital by using multiple
trust structures that conceal a common controlling mind; (b) creating artificial
losses, for example by acquiring companies or trusts with accumulated losses; (c)
disguising distributions to High Wealth Individuals (HWIs) and family members
as loans and other non-taxable benefits; (d) using offshore trusts; (e) converting
activities undertaken for private pleasure into tax losses (for example pleasure
craft, horse breeding and racing); and (f) using charitable trusts to disguise
benefits to HWIs and their families.
Method
A useful model for this research was presented with Roman Tomasic and
Brendan Pentony's 1990 study, 'Defining Acceptable Tax Conduct: The Role of
Professional Advisers in Tax Compliance'. It was decided to undertake a more
focused investigation of advisers of HWIs and to limit the range of policy issues
discussed with them. To define the topics, the author, together with Alice Dobes
(an ATO evaluator from outside the Taskforce), first conducted interviews with
most senior managers of the HWI Taskforce in Canberra, Sydney and Melbourne,
mostly meeting them one or two at a time. This led to a decision to pursue fairly
unstructured interviews with advisers of HWIs organised loosely around a set of
questions prepared in advance. The questions were designed to address three
issues: (a) the compliance problems common among HWIs; (b) to seek advisers'
criticisms of the taskforce; and (c) to harness their creativity in considering new
and better ways of improving compliance and identifying deficiencies in the law.
The objective then was to test their criticisms and ideas against the policy
thinking we had secured from the taskforce managers, focus groups that Alice
Dobes conducted with taskforce fieldwork staff subsequent to the adviser
interviews, and against what we know from the research literature on compliance
(not a lot for High Wealth Individuals).
The Sample
The sample was strategic rather than random, with 27 HWI advisers interviewed,
mostly alone, for one to two hours in 1999. All interviews were conducted face-
to-face with the exception of one which was conducted as a tele-conference.
The strategic sample included: (a) individuals who senior taskforce
management regarded as key HWI advisers; (b) advisers whose policy thinking
might be sharper because they have a larger number of HWIs and/or many more
contacts with the taskforce due to audit and information gathering activities
resulting from higher risk ratings of their HWI clients; and (c) HWI advisers who
were selected by professional bodies which were interviewed to represent the
views of their membership.
This sample ensured we secured interviews with advisers from all the (then)
Big Five accounting firms, some non-charter firms, some distinctive boutique
firms and certain prominent lawyer-advisers. In addition to the advisers I also
conducted interviews at the Taxation Institute of Australia, the Australian Society
of Certified Practicing Accountants, the Institute of Chartered Accountants in
Australia, and the Law Council of Australia. Each of these bodies had informally
canvassed their members to elicit specific concerns and suggestions.
To secure interviews, initial contact was made by letter offering prospective
participants the opportunity to speak with either myself as the author of this
chapter, or to both Alice Dobes from the ATO and myself. Confidentiality was
guaranteed and respondents were informed that while what was said would be
used, it would not be attributed to them as individuals or to their firm. They were
also assured that they were not required to divulge the names or details of their
clients.
While one adviser was overseas for the period of the study, no advisers refused
to be interviewed, and in only one case did an adviser opt for a complete absence
of the ATO during the interview.
Findings
Cost of Compliance
All but one adviser we interviewed found the cost of compliance to be high, and
higher than they believed it needed to be. The cost estimate for the professional
fees associated with completing the expanded tax return required by the program
was in most cases in the range $20 000 to $40 000 per HWI. At least this was the
estimate for the first expanded returns: views varied on how much costs fell with
subsequent returns. Adviser fees to assist with HWI audits were also reported to
be as high as 'more than $1 million'. In addition, the program was reported to
impose substantial costs on the time of the HWIs themselves and their staff. It
was a distraction when 'they want to keep their eye on the main game which is
creating wealth for the benefit of Australia'. Other advisers suggested that the
high costs of doing expanded returns for the first time were considerably reduced
on subsequent returns and could be reduced further by electronic lodgment.
The main suggestions for reducing costs related to reducing the repetitiveness
and increasing the selectivity of expanded return targeting. A HWI who controls
sixty entities has to complete sixty special returns, which demand a considerable
amount of information. Some advisers said that such burdens on HWIs created
business for them, but more were of the view that it was a kind of business they
did not want - 'drudgery' as one put it. Others said they would prefer to be
helping clients build their business rather than putting their energy into
compliance administration. Some said that they were not well geared with the
kind of staff needed to fill out hundreds of pages of returns by hand.
It's not the business we want to be in. We want to add value. We don't want to be putting numbers in
boxes. It's not satisfying. And our clients don't want to pay big dollars for that.
Because compliance costs are such a big issue, they are also a lever. Getting off
the program is seen by advisers as a major benefit they would love to be able to
secure for their HWI. The opportunity is to take them off the program only when
certain compliance outcomes have been delivered.
Many HWI staff have a different perspective from the majority analysis of the
advisers. One common view was that self-assessment in the 1980s may have gone
too far in respect of companies and trusts. It is now impossible to do a risk
assessment on these returns without seeking additional information. The HWI
expanded returns are seen as a remedy for this error of the 80s. This was also a
minority view among advisers:
I admired Trevor Boucher [former Commissioner of Taxation], but one of the worst things Boucher did
was simplify the tax return, especially for corporates. You need a complex return for complex
taxpayers...The expanded return questions are pretty good questions...Wealthy people should never be off
that cycle. Once you've done the first year, the costs are not high.
This adviser then went on to suggest, contrary to the majority view that he
would put more, rather than fewer, questions on the expanded return.
Conservatively, compliance costs of the HWI Taskforce, on the basis of the
estimates provided in our interviews, would seem to be well over $10 million a
year - that is, a more significant economic issue than the funding of the program
(which costs less than $10 million). This conservative estimate is based on
multiplying the number of expanded returns by the modal compliance cost
estimate for completing expanded returns, and adding just the estimated audit
costs in cases we were told about.
We found that some HWIs accepted these compliance costs as reasonable;
others interpreted them as persecution. How ever the ATO resolves the policy
issues here, it needs to communicate more clearly the reasons why HWIs do and
do not stay on the expanded returns. Otherwise 'the client gets a feeling of
persecution for being who they are'. Failure to communicate reasons for seeking
expanded returns risks the reaction of one HWI described by his adviser as
someone who had rejected perfectly conservative kinds of tax planning advice on
grounds that he wanted to pay his fair share of tax. His reaction to being kept on
expanded returns for another year was: 'If the ATO thinks I'm ripping the system
off, maybe I should start doing it'.
A policy option here is that after a risk assessment has been undertaken on
each expanded return, HWIs be advised that the assessment has been completed
and given a general reason if they will be required to complete an expanded
return in the following year. If they are not advised, our interviews suggest that
HWIs and advisers assume that the ATO is not doing its job and has simply not
got around to completing the risk assessment. While expanded returns should be
maintained as a routine requirement for HWIs assessed as having a high or
medium risk profile, there is a case for continuous improvement in simplifying
the process. Continuous reduction in compliance costs might be considered each
year by examining four paths to lower costs: (a) reducing the amount of
information requested in expanded returns; (b) reducing repetition of information
from year to year and form to form (for example allowing an 'unchanged since
last year' response); (c) greater discernment (possibly based on something like a
Tax Strategy Review) in assessing which individuals should complete expanded
returns in the first place and greater clarity in signalling a path for exit from the
expanded return program; and (d) further work to facilitate electronic returns.
Overall Effectiveness
We can divide interviewed advisers into three groups: (a) a group of more than a
dozen advisers who accepted the HWI Taskforce, while feeling it was conceived
in such a political way that its targeting was ineffective; (b) a group of six who
aggressively rejected the very idea of the taskforce; and (c) another group of six
who felt strongly that the taskforce was a sound initiative: 'There should have
been a program targeting high wealth individuals long ago'.
Only a few advisers thought the activities of the taskforce had made any
difference to their clients. However, a minority did say that they felt the existence
of the taskforce was making HWIs generally more cautious because they are
'under the microscope':
It's the constant surveillance of being on the program that causes compliance...Part IVA is of
indeterminate width. It might be applied more aggressively in future. So I advise clients to be careful.
Asking the question and getting them to focus their mind on where everything is had an impact.
In some cases this impact was bringing into the system business that was
offshore, sometimes openly in a way that would be brought to the attention of
the ATO, sometimes indirectly, in a way intended to make it difficult for the ATO
to notice. Several advisers mentioned instances of voluntary disclosure of large
amounts of income prompted not by audit, but simply by the fact of having to fill
out the expanded returns.
One adviser was of the view that this aspect of effectiveness would be severely
compromised if the expanded return was not universal for wealthy people with
complex affairs:
Corporate Australia would say you really should go at it hell for leather as soon as they are dropped off.
That's what happened with the Large Case Program. The Large Case Program should never have been
stopped. Corporate Australia learned the lessons from it; the ATO walked away from those lessons.
What this adviser was saying here was that notifying X in 1998 reduces the
taxpayer's degrees of freedom to reconfigure his 1999 affairs in such a way that
not-X appears to be the case in 1998. 'Changes each year will be noticed.' So
HWIs must keep their affairs consistent with the underlying truths of earlier
declarations.
Most advisers thought it would return more revenue for the ATO to shift its
resources to the cash economy; some favoured targeting corporations in high risk
industries rather than targeting individuals simply because they were wealthy.
One argued very forcefully that the best returns would be achieved with low
HWIs or sub-HWis rather than high HWIs.
Forget Kerry Packer and Rupert Murdoch. They are far too big. Their advice is too good. Go after some
middle HWIs or lower. Their accountants are small fry, many of them.
A widely held view of the HWI advisers was that a life course of HWIs could be
defined. HWIs were most aggressive when they were making their first million,
before they became HWIs. They continued to be aggressive while they were
building their empires. Then they often wanted to avoid trouble and to pay their
fair share once they had made it. Unfortunately, the HWI program is seen as
targeting a lot of people who are in this latter quiescent stage, while the real
returns are among the younger wealth-builders. One adviser classified HWIs as
'the meek, the bold and the normal'. The meek, on this analysis, would be found
disproportionately among older HWIs, the bold among wealth-builders and the
normal among mid-career HWIs.
Another view was that the very wealthy, and very aggressive, are hard targets.
The more realistic policy objective than working from the very wealthy down is
over time to push up the level of wealth where tax planning can succeed in
eliminating the need to pay tax - that is, moving from the wealthy up toward the
very wealthy.
The strongest and most widely held basis for doubt about the effectiveness of
the taskforce was the belief that it collected a lot of information and then did
little or nothing with it. Doubts were widespread that anyone had even looked at
the information in the returns they had put in, let alone subjected it to a rigorous
risk assessment and followed up with audit where appropriate: 'It's a fishing
expedition. They don't know what they're looking for'.
Risk Assessment
Another virtually universal reaction in the interviews was that the politicised
beginning of the taskforce was a source of resentment. Former Treasurer, Ralph
Willis, claimed that HWIs were avoiding $800 million a year in unpaid taxes
during the 1995 election campaign. This was a frequently cited focus of
resentment. In some cases, the Commissioner of Taxation was seen as a co-
conspirator with the government in whipping up a 'witch-hunt' against HWIs.
This then led to an attack on the way the initial targets were selected. Business
Review Weekly's list of the wealthiest 200 Australians was widely seen as the
basis of the initial targeting, and as a markedly inaccurate source. Some HWI
advisers claimed they had other clients who were wealthier than the clients who
were targeted.
Beyond the complaint about the initial basis for risk assessment, there was
little complaint about the way subsequent risk assessment was tackled. In many
cases, HWI advisers seemed to have very little knowledge of how risk assessment
was done. But if risk assessment were a black box to them, it was not a source of
complaint. Many advisers were of the view that the shift away from full audits to
risk assessment followed by a suite of audit products was a sound move by the
ATO. Others were cynical, believing that compliance was falling because the risk
of audit had fallen and that this fact was widely recognised in the marketplace.
There was general agreement, however, that the ATO was doing a more
competent job of risk assessment:
The general view in the accounting profession is that the ATO is better geared than it ever was to detect
where the leakage is. Therefore it should be possible for the ATO to keep a clamp on the most aggressive
activity. But it audits less, making that more difficult.
While there were many concerns about the fairness of the program, views about
the fairness and professionalism of the ATO staff with whom advisers had dealt
were overwhelmingly positive. A number of advisers complained that they had
little or no contact, but for those who did, only seven were critical of the
experience. The strongest criticism was lack of technical competence. Only a
handful of incidents were described where ATO staff was seen as less than fair or
professional. One repeated complaint was the recording of a 'jaundiced view of
taxpayers', their honesty or their lifestyle, in internal memoranda obtained under
Freedom of Information and even in position papers. Given the frequency with
which this kind of taxpayer will use Freedom of Information to gain more insight
into the ATO analysis of their case, especially in the context of settlement
negotiations, more consideration may be needed in expressing opinions about
taxpayers.
Some of the assessments were very positive: 'He was very good, commercial,
understood the realities of going offshore...Compared to other audits I've
experienced, it's been very professional'. In response to the question, 'Can you
think of any instances where the ATO has not honoured the Taxpayers' Charter
in its dealings with your HWI clients?', there was only one specific complaint. In
26 out of 27 cases, the answer was no, though in some cases this was qualified by
the concern that the general lack of communication in the program or aggressive
assessments to bring the taxpayer to the negotiating table may be Charter issues.
Generally, the Charter did not seem to mean a lot to these advisers. One was
brutal about it: 'The Taxpayers' Charter is a motherhood document that is really
bullshit in the marketplace'.
The generally positive results on fairness and professionalism are important
because the literature of the social psychology of procedural justice shows that
when people believe they are treated fairly, they are more likely to comply with
the law (Tyler, 1990, 1998; Tyler and Dawes, 1993; Makkai and Braithwaite, 1996;
Lind and Tyler, 1998). The fact that HWIs often have a sense of fairness that
contributes to compliance, and therefore must be sustained, is illustrated by the
following kind of request that advisers said they commonly received:
I want a tax plan that will get my tax down to X per cent, say twenty. I want to pay my fair share. But
not that much. Then others will say I want a total tax wipe-out and even pay a ridiculous amount of
money to get it.
Some advisers were critical of a failure of taskforce staff to work with other
sections of the ATO (for example Transfer Pricing, Small Business, an industry
segment of Large Business and International) to avoid turning over the same
issues with different sets of ATO people.
On the technical questions, one suggestion was that audit staff be more willing
to call in more senior technical people when the issues got thorny, and in such
cases, for the ATO technical expert to have sign-off on position papers. There are
already three levels of sign-off on HWI risk assessments (analyst, team leader,
manager/director), which are often applied to position papers.
On the other hand, sign-off rules may be too formal an approach to something
that is fundamentally a cultural challenge for the taskforce. On all aspects of
audit, communication is needed to improve quality. Getting position papers
echnically correct is just a small part of this. The taskforce must be careful in its
quality control not to end up with such a focus on avoiding errors in its written
communications that it neglects quality assurance of bigger strategic issues that
would remove the need for a written communication in the first place. One HWI
manager, for example, said that one of the most valuable pieces of advice he
regularly gives auditors is: 'I've seen this stuff before and I can tell you following
it is not productive'. His view was that the essence of professionalism was not
getting bogged down in pointless pursuits.
Training is important here as well. But perhaps formal training is the more
important path to keeping technical skills up to date (training the mind), while
informal discussion between masters and apprentices is the more important path
to improving the wisdom of strategic audit judgment (training the nose). One
HWI adviser argued that it takes 20 years for auditors in complex cases to acquire
a nose for the right lead. In light of this, he was critical of ATO early retirement
packages that were disproportionately taken up by precisely such experienced
people. Balance, an ability to extrapolate, the gift of getting an inkling on what a
transaction means, and where it might lead, without the full information, are
virtues that might be nurtured more by a retention program than by early
retirement packages for the people with these rare gifts. The learning process
perhaps can also be facilitated by the old and the wise conducting best practice
workshops for the young who do not yet know how to read the commercial signs.
When position papers involve difficult technical issues, there seems to be a
need for the level of authority for sign off to be increased. The taskforce also
needs to set expectations for hours of formal technical training that are realistic
in relation to workloads, but well in excess of what is expected of fieldworkers in
less complex areas. An ongoing identity that fosters a storytelling culture about
how to find fertile leads and how to avoid infertile ones is an issue more
systematically addressed in an earlier chapter (Braithwaite, Chapter 9, this
volume). Such a culture is one where apprentices are constantly asking masters to
relate stories of their experience on strategic decisions they are making. It is also
a culture where masters do not tell apprentices what to do so much as volunteer
stories from their experience when they see their apprentices about to repeat
mistakes they have made in the past. Best practice workshops are needed
nationally to assist with the development of such a storytelling culture that
nurtures strategic wisdom.
Communication
A bit of frank discussion goes a long way.
Auditors need to understand that it is natural for me to talk to their head. That is a cultural change
needed.
Why can't we have a relationship of trust, a process, which starts with me saying 'here's my analysis of
how things have changed since last year'.
What about the Asian and English model of tax inspectors building relationships. Almost a personal
thing in paying your tax. A more personal approach that seems to work. Then it's hard to avoid a big
issue that arises on either side (professional association interview).
Since most encounters with HWI staff were seen in a positive light by HWI
advisers, it is not surprising that most of them wanted more communication.
They particularly wanted communication about where their client was up to in
the program: 'Had he received a clean bill of health?'; 'When would she get off
the program?'. Many HWIs were reported by the advisers as philosophical, that it
was reasonable that people of their wealth might be targeted in this way. But
there was also a widely reported view that their clients felt victimised by the
program. In the view of the advisers, here was where more communication from
the ATO would help. Some older HWIs worried about their tax affairs and were
concerned to do what was necessary to get a clean bill of health quickly: 'They
say it's only money and it's better to sleep at night'. There is an opportunity lost
for both the ATO and the taxpayers to reach accommodation quickly when this is
the situation. Many advisers wanted agreed timelines for expanded returns to be
assessed, decisions made whether to take the case further or to drop their client
off the program. And they wanted communication on how this was progressing.
A majority of the advisers specifically requested that there be a taskforce
representative assigned to their case and that they meet them in advance of the
expanded return being requested. What was most commonly favoured was a
preliminary risk assessment by an analyst based on information provided by the
HWI at an initial interview. This information would include how the structure of
the entities controlled by the HWI had changed since the last assessment,
financials, extraordinary transactions, a mud-map of the group structures and the
tax reconciliation (from accounting to taxable income). Advisers repeatedly
argued that in many cases advisers were in a position to convince an analyst why
their client paid little tax even though they were very wealthy.
One adviser suggested that meetings of ATO staff with HWI advisers would
assist the professional development of both. It would improve the understanding
and clear up misunderstandings about how the AT'O works, and discussions on
technical issues would expose ATO people to the technical insights of advisers
and vice versa. An adviser gave an example. There had been enormous conflict
over a disagreement in interpretation of the law. This disagreement was fuelled
by the belief on the part of the adviser that the ATO auditor was technically
incompetent. In the wash-up, the adviser admitted that it had been his technical
analysis that had been less sophisticated. The ATO's Tax Counsel Network (TCN)
had been giving sophisticated advice to the auditor:
They were ahead of us and we didn't know. We just assumed this was another case of this auditor getting
settled law wrong. Probably, we should have been communicating directly with the Tax Counsel
Network. Certainly if there had been some communication, even from the auditor, on where he was
coming from, unnecessary confrontation would have been avoided.
Given the propensity in this game for the players to be uncharitable toward their
adversary, it is likely that in the absence of communication, the assumption that
the ATO is incompetent is exactly what the adviser will take away from the
encounter.
Similarly, how ever the ATO resolves the big policy question of whether it
should be normal or exceptional for HWIs to stay on expanded returns year in
year out, there is virtue in taking seriously the following suggestion of an adviser:
Why not a letter of commendation on clearance. Thanks for your cooperation. You've got a clear bill of
tax health.
Even if the HWI is to stay on expanded returns, presumably such a letter might
say: 'Thank you for your cooperation. A risk assessment on your return this year
has resulted in a decision not to include you in our current round of audits.
However, you will be required to complete an expanded return again next year as
our risk assessment also showed that you are likely to continue to have a high-
risk tax profile. If you have any questions about this, feel free to get in touch with
your case contact person, Mary Smith'.
Another adviser implied that competent ATO staff picked up useful risk
assessment signals from communication with advisers. This was an adviser who
said he dropped HWIs when they lied to him. He felt that through this kind of
act, the ethical adviser gives a signal that the competent ATO staff member
knows how to read.
Communication between advisers and ATO staff can help keep the ATO up to
speed with the latest arrangements in a world where 'the better an idea, the
shorter its shelf life'. Several advisers confessed to us that they had dobbed in tax
planning arrangements used by their competitors, combined with their opinion
on what its technical weaknesses were. Others explained this was widespread
because there is 'a lot of jealousy' or 'a degree of bitchiness out there'. In one
example, firm A developed a product with some bugs in it. Clients of firm B were
approached about buying into the scheme. Firm B then reverse-engineered the
scheme into an improved version and 'made noises to our contacts in the ATO'
about the bugs in firm A's version.
Some advisers and some HWI staff did not see the need for the ATO to initiate
communication. These advisers said their experience was that they could
telephone the taskforce and discuss any question that was concerning them. A
number of other advisers had not had this experience, however. They found great
difficulty in securing the communication they felt they needed, and in some
cases, any communication at all.
Another option for improving communication is suggested by the Canadian
Audit Protocols (Revenue Canada, 1996). The Canadian Audit Protocols are not
just about a move to real time audits; they are also about rewarding cooperative
relationships between the tax authority and its clients with negotiated audit
protocols that reduce compliance costs for business and increase compliance
effectiveness for the tax authority. For example, scheduling visits by different
areas of the tax authority so that disruption to business is minimised, conducting
concurrent audits, and informing business in advance of the form in which
financial records might be kept to avoid their double handling. The idea is that
Revenue Canada and participating corporations jointly produce a written
framework that establishes guidelines for the relationship and the audit process.
When a HWI is targeted for an audit, it seems quite possible and sensible to give
their adviser an audit plan, with the proviso that they might be advised at any
stage that the plan has been modified. Indeed, we were told this is generally done
with HWIs, though not always in a timely fashion.
Complexity
Half the advice I see people give is wrong...McKinsey and Co showed some years back that the ATO's
advice on the basic matters dealt with at the enquiry counter was right only thirty per cent of the time;
thirty per cent of the time it was wrong; and forty per cent of the time it was useless and beside the
point.
I know twenty per cent of it by working twelve hours a day, six days a week for twenty years.
The Commissioner would have to take two weeks off to get up to speed in what he is talking about in
that trust stuff And he should not have to. The Act should not be so complex that only those below him
have the understanding of the law to run and control agendas.
There was consensus among advisers that the complexity of the law was
keeping compliance costs high for HWIs and making effective enforcement
difficult for the ATO. There was general support for the idea of discerning general
principles in the law and disciplining specific elements of the law to be consistent
with these general principles. The idea is that all components of business tax law
should be derived from, and consistent with, a smallish set of general principles.
There was also general cynicism that the wash-up to the Ralph Inquiry (Ralph
Report, 1999) that was developing such principle-based reforms would be any
more simplifying than the results from two decades of pronouncements about
such efforts. Three years on from the interviews this cynicism has been
vindicated, as there has been little of the principle-based reform advocated by the
Ralph process.
Offshore Investment
Advisers said that clients came to them requesting that they help set-up offshore
investments, accounts and credit cards. But only two said they got into assisting
HWIs with this. Two others said this was a reason why clients went elsewhere. A
number of advisers felt that few HWIs avoided taxes by using offshore strategies
such as tax havens. However three advisers felt offshore tax avoidance was the
main risk to the revenue by HWIs and some others saw it as a significant risk.
One of these made the point that a higher proportion of Australia's most wealthy
people are immigrants compared with most other countries, making offshore
trusts a bigger issue for us.
Three advisers also felt that a strategy of the aggressive HWI was to have
conservative domestic tax planning arrangements managed by a reputable
adviser from a Big Five or second-tier firm, and then to have another set of
investments offshore which are not disclosed to their conservative adviser. This
strategy implies securing a favourable risk assessment by the ATO on the basis of
the conservative side of their dual strategy, thereby protecting the offshore side of
the strategy from scrutiny.
Even those who thought offshore tax avoidance was a major problem
emphasised that not all HWIs could exploit it. Offshore strategies would be
foolish for a HWI who was a resident of Australia and who ran no actual
businesses offshore. In these circumstances, it is difficult to conceal the
movement of funds as they leave the country. For HWIs who are non-residents or
who have non-resident family members, funds can be passed to them and then
shifted offshore to a destination unknown. The ATO only has direct access to the
domestic tax affairs of non-residents. An Australian resident who is wealthy
enough to actually run a business overseas also has a formidable capability for
'bleeding some funds out' into personal accounts protected by a tax haven.
No adviser recommended targeting offshore investments as a sensible shift in
enforcement strategy for the taskforce. For some, this was because they did not
believe HWIs were using offshore accounts to avoid tax on any large scale. For
others who believed HWIs were doing so in a massive way, the belief was that
these arrangements would be impregnable or very hard to find: 'You'd end up
chasing a British Virgin Islands company where you can't find out who their
directors are under their law'.
However, three advisers recommended an amnesty on bringing home or
paying tax on offshore investments. One felt that, particularly with older HWls
who wanted peace of mind and a capacity to pass wealth on to their children
without undue complication, an amnesty based on agreement to pay back tax and
interest (but no penalty) would bring a lot of offshore income into the system. A
second adviser felt this would not generate enough incentive to bring in the huge
offshore investments which he said he knew were out there. His policy
suggestion was that as long as offshore funds were actually brought back for
investment in Australia and future taxing on-shore, all back taxes and penalties
should be waived. Only the removal of the threat of back taxes, he believed,
would see the repatriation of worthwhile amounts of tax and capital:
Half a dozen merchant bankers who were my clients would have been happy to bring money back into
Australia if they did not have to pay back taxes...If you didn't ask any questions, a huge amount of
money would come.
The large majority of HWIs, according to their advisers, want to avoid anything
illegal, any litigation, anything that will make them a test case for the
Commissioner. Consequently, advice by the adviser that a claim is illegal is
enough to stop it, a finding Tomasic and Pentony (1990) also reported. This fact
reinforces the need for good communication between the ATO and advisers, who
are the agents of much of their voluntary compliance work.
For the same reason, we were repeatedly told of the power of a simple media
release by the Commissioner to stop schemes in their tracks: 'The moment
taxpayers see a negative statement [by the Commissioner] on a scheme, it doesn't
happen. The Commissioner can close them overnight'. A number of advisers
pointed out that the ATO should be responsible enough not to issue such releases
unless they were accompanied by or followed promptly with a supporting
technical paper.
Others were critical of the under use of this effective tool of compliance: 'The
ATO sits on things it could kill off. Sometimes this is because the ATO is slow to
complete its technical analysis, sometimes because it does not yet have the
resources or political will to attack or test a scheme (and it wants to maintain a
reputation for keeping its promises). Sometimes it is slow to detect new schemes,
though most advisers felt the ATO had become much more effective in this
regard, often pointing to the potency of the strategy of Strategic Intelligence
Analysis. Most felt it was possible to keep abreast of the new schemes through a
combination of maintaining good lines of communication with their fellow
advisers: 'They love to chatter. People have big egos and love to brag. If this
fellow doesn't, the next fellow will'. Simply 'knowing what is being said from the
rostrum of Taxation Institute conferences, which covers nine schemes out of ten'
and following up the advertisements in the Australian Financial Review and
Business Review Weekly can cover much of the field.
Analysis
Most people think that compliance with regulatory laws is mainly secured by
punishments imposed by the courts if the law is violated. Hence, their crude
analysis of the HWI Taskforce would be that HWIs are almost never prosecuted
criminally and are infrequently subject to tax penalties, so this is the reason why
they pay such a low proportion of their wealth in tax. In most regulatory
domains, this view is simple-minded or mistaken, but particularly so with tax
compliance by HWIs.
The advisers told us that outright non-compliance is rarely a rational strategy
even when detection risks are extremely low. Consider the following highly
effective form of blatant evasion available to HWIs. It is loosely based on a real
case:
The HWI has a million dollars in profits. He gets a charitable deduction by donating it to a breast cancer
research foundation he sets up in Geneva. The foundation then almost immediately lends it back to the
HWI at an exorbitantly high interest rate. This interest rate enables the company that pays it to record a
loss that the HWI can then write off against profits in another entity he controls. The HWI gets his
million dollars back and two tax write-downs: A deduction on the way over to Geneva and a loss he can
use to reduce taxes on the way back from Geneva. If he disguises the transactions effectively, it almost
certainly won't be detected. If it is, the HWI has reputable people organised to testify that he always fully
intended to repay the loan. He can actually do so before the matter goes to trial. The Director of the
Breast Cancer Research Foundation will testify that they wanted to use the money well, to wait until a
research proposal came along that would really produce a medical breakthrough. But while they were
waiting they wanted to put their money to work. They knew that their benefactor, the HWI, knew how
to do that better than they. And he was generous enough to pay an above-market interest rate to ensure
that all the profits from his investment would be passed back to the Foundation. The ATO decides it does
not want a case where a judge might vilify it for persecuting a businessman dedicated to such a cause.
While this kind of case has happened, it is not a kind of scam that in the opinion
of HWI staff is widespread, not one of the major risks to the revenue among
HWIs with the level of resources needed to exploit it. Why? After all it is highly
profitable, simple enough to execute and almost guaranteed never to land the
HWI in jail. The perfect crime. The answer according to our interviews is
twofold. First, the HWI has too much to lose to risk even the remote prospect of
the perfect crime unravelling. Second, there are other perfectly legal ways she
can arrange her affairs to avoid having to pay tax on her million dollars without
losing any sleep at night. So she prefers to pay an adviser to help her execute one
of these latter strategies. The level of tax benefit in respect of her million dollars
will depend on how competent and aggressive her adviser is, both being
attributes in an adviser she can pay a premium to get.
All the government should or can do about the HWI who goes to the
extraordinarily competent, yet conservative, adviser is to close off the biggest
legal opportunities they exploit. For these conservative HWIs who pay little tax,
deterrence is utterly irrelevant as a strategy for protecting the revenue. There is
no need to deter the Geneva charity scam because this kind of option is
unthinkable to them. It would be improper to deter them from exploiting the
services of an unusually competent but law-respecting practitioner.
We now consider whether there is a place for a strategy of deterring the use of
an unusually aggressive adviser.
Two letters sent by the Big Five accounting firm, Deloitte and Touche, in late
1998 opened as follows (Novack and Saunders, 1998):
Dear
As we discussed, set forth below are the details of our proposal to recommend and implement our tax
strategy to eliminate the federal and state income taxes associated with [the company's] income for up to
five (5) years ('the Strategy').
They were sent to two medium-sized United States (US) corporations asking for a
contingency fee of 30 per cent of the tax savings from taking the tax liability to
zero.
In Australia, our interviews suggested that the then Big Five did not touch the
marketing of schemes with anything like this kind of aggression and that they did
not operate on such a contingency fee basis.3 The more aggressive shelters are
marketed by non-charter firms, some by small law firms, others by financial
advisers of a rather fly-by-night character ('they get $10 000 each from 200 people
for their scheme and then they don't care what happens'). That may be, but there
is a reality of the global market here. The fact is that in the US, the Big Five seem
to have been able to increase their profits substantially through shifts toward
such tactics. Individual staffers can secure bonuses up to US $400 000 for landing
deals such as those pursued by the Deloitte and Touche letter. Ernst and Young
and Deloitte and Touche reported a 29 per cent jump in revenues from tax
services in the US in 1997. Since 1993 tax revenues for the Big Five have grown at
twice the pace of audit revenues (Novack and Saunders, 1998). American Express
is contesting their market in a formidable way. Something is changing in the risk
environment here and it is probably global.
If it is about global change in financial engineering, Australia cannot resist it.
Novack and Saunders (1998) argue that in the US 'it has taken a while for
inhibitions to be shed and the most outlandish gimmicks to propagate'. But the
inhibitions have shed under pressure from the aggressive marketisation of
proactive as opposed to reactive advice. Today, as one adviser worried 'if you pay
enough you can find a lawyer to write an opinion supporting anything'. The
advice market, in Australia as well, decreasingly operates through advisers just
reacting with specific strategies to cope with the needs of clients as they arise.4 If
advisers do not proactively market strategies to substantially reduce the tax
liabilities of major clients, they can count on it that competitors will. Once the
large accounting firms reach the conclusion, as they clearly have in the US, that
they will lose business unless they match it with aggressively proactive promoters
of shelters, a level of global sophistication in the engineering of proactive tax
planning that is within the competence of the big firms comes into play.
Moreover, as the major accounting firms increase their proactiveness, a change
can be expected in business culture on tax matters. 'Inertia', as one HWI adviser
explained to us, is a major cause of tax compliance. If no one is pushing
aggressive tax planning at a client, they do not feel incompetent for failing to do
it. In Australia, there is a business ethos, and an ethos of the elite tax advisers,
that 'if there is not an underlying business case for an investment, but only a tax
case, don't do it'. The gradually growing clout of the general anti-avoidance
provision of Part IVA of the Income Tax Assessment Act 1936 is perhaps one
reason for this.
However, it would be naive to assume that global commercial pressures might
not change it. Thanks to the GATS (General Agreement on the Trade in Services)
agreement of the GATT (General Agreement of the Tariffs and Trade) - for which
American Express (a heavy buyer of tax lawyers and accountants) and
PricewaterhouseCoopers were the leading advocates, along with Citibank - the
financial services market is now globalising. The consequence will be that if the
Australian tax advice culture is seen by aggressive foreign firms as excessively
conservative, they may view it as a market they should attack through a sales
force that systematically works a distribution network to all HWIs. Globalised
investment banks, with all their aggression, networks, expertise and reputational
capital,5 may be the biggest threats for ushering in this change. So even if it is the
case that Australian business culture is more reactive than proactive on tax
design, and that the tax departments of large corporations are not typically seen
as profit centres in their own right (increasingly the case in American business
culture), the capacity for Australian businesses to resist a globally-driven culture
change must be in doubt. There is a kind of culture transformation and take-off
with these things, as Australia saw with the schemes of the 1970s: 'Once
companies get a taste, they become more comfortable and continue to do it'
(Novack and Saunders, 1998). A critical litmus test of the culture change is a
growing belief in the business community that paying any significant amount of
tax is a sign of weakness that might be criticised by shareholders.
So it may be that global financial services market pressures will progressively
unsettle the inertia and conservatism that delivers a lot of HWI taxpaying at the
moment. Moreover, the profits that the US accounting firms seem to be deriving
from contingency arrangements at the top end of the market suggest that there is
reason for promoters to move down the market offering proactive hustling of
more engineering-intensive tax planning.
This seems a risk to the revenue that would run down from HWIs and large
corporates to players of more intermediate levels of wealth. In addition, it would
put at risk the culture of tax compliance at lower levels of the economy which
has been reasonably strong in Australia by international standards. The worry is
that if ordinary Australians become as cynical as Italians about the tax morality
of the rich, we will be at risk of Italian levels of voluntary compliance, or even
Colombian levels. As difficult as the Commissioner's job is, we should never
assume it is incapable of getting a lot more difficult. One step that is a priority in
Australia is to persuade tax professional bodies to set self-regulatory standards
that prohibit contingency fees on reduced tax liabilities.
For these reasons, it seems necessary to guard against the risks from global
competition in tax hustling by deterring aggression in tax advice. How can that
be achieved? It is not against the law to give aggressive tax advice and nor should
it be. Advice to break the law is sufficient reason to withdraw a licence to offer
tax advice. But we have seen that this is not the real problem. Deterrence is
largely beside the point at the moment because taxpayers who breach the law on
the basis of a reasonably argued position supplied by a licensed adviser will not
be subjected to any penalty tax; they simply pay the back taxes plus interest. This
means that it can be rational to go to advisers who push their advice to the limits
of what could conceivably be accepted as a reasonably argued position.
In the US context, Novack and Saunders (1998) express this rationality as
follows: 'The IRS misses nine out of ten shelters. On the tenth, the company pays
back taxes and the government agrees to no penalties'. In the Australian context,
there appear to be two highly rational strategies if you are a HWI: One is to go to
a conservative adviser respected by the ATO who will put you only into tax
planning arrangements which have been approved by the Commissioner in a
Private Binding or Public Ruling, and who will do that with maximum
transparency to the taskforce. The taskforce is therefore likely to assess you as a
low risk. But the other rational strategy may be to conceal a highly aggressive tax
planning activity so it is most difficult for the ATO to find and then cover it with
a (weak) reasonably argued position. If you are unlucky enough to be audited and
have it detected, you roll over and pay up quickly so long as there is no penalty
in the settlement, or a minimal one. An in-between strategy of a highly visible
reasonably argued position, which would have a strong chance of success in the
courts (but is rejected by the ATO) is not attractive because you risk the
uncertainty, costs and notoriety from fighting the ATO on a test case.
The policy objective we wish to explore is deterring the second strategy so as
to encourage the first. This cannot be accomplished through penalties because the
probability of a serious penalty is too low when there is some sort of reasonably
argued position in relation to a low visibility activity that is unlikely to be
scrutinised by an auditor. It can be accomplished, however, by targeting audits
and other forms of surveillance on HWIs who use the services of aggressive
advisers or promoters. Among other things, these would be advisers and
promoters whose clients are frequently found to have activities tucked away,
which are defended by a weak reasonably argued position supplied by the
adviser. They would have a track record of promoting arrangements that have
been disallowed by the Commissioner or struck down by the courts. The idea of
this strategy is that clients who wish to avoid the risks and direct costs of an
audit, steering their business away from aggressive advisers and promoters,
would improve compliance. There are limitations to the strategy. The best tax
planning is specifically designed for a HWI and will not be noticed as a problem.
A credible litigation strategy is an imperative to show HWIs that it is not an
option to bankroll the government out of enforcement action and that the ATO
keeps its promises, even if this requires high litigation expense. Using test cases to
develop a strategic case law around Part IVA seems particularly important.
For the critics, the expanded return at the heart of the HWI Taskforce risk
assessment strategy amounts to saying: 'Give us the haystack and we'll find the
needle'. Some advisers argue that what the taskforce should be doing is 'getting
their rifle out, not their shotgun'. Advisers generally do not realise that a
systematic risk assessment is completed on the information in the expanded
return, and a team leader and manager check this assessment with sign-off.
One adviser made the point that the ATO could not do the job of targeting
audits for HWIs in a sensible way before the expanded returns were introduced.
Audits, as a result, were 'a lottery'. He therefore saw a rationale for continuing
the expanded returns on an even wider front than their current coverage:
The average innocent Australian would have less chance of being audited for the wrong reason.
The questions in the HWI returns are pretty compelling. If everyone, other than PAYE returns, had to
declare these things, there would be more compliance.
He illustrated the point with being forced to declare offshore income. He said it
was one thing to be in a situation where you can defend a failure to declare
something as an oversight. It was another to answer 'No' to that expanded return
question on overseas trusts from which you might benefit, when it can be shown
later that you knew the answer to be 'Yes'.
Theoretically, that's go to jail stuff and people want to sleep at night.
Some other advisers made similar points: 'If you have been doing the wrong
thing, some of those questions must be difficult to answer'. On this view, the
genius of the expanded return is that it puts HWIs in the position where they
have to lie outright rather than 'overlook' something. It forces into black and
white what had previously been grey. Both conscience and fear of deterrence
work better in the realm of black and white than in the realm of grey. This point
applies to both HWIs themselves and to their advisers.
The expanded returns also have preventive value. Completing an expanded
return reduces the degrees of freedom for redefining income and deductions for
the purposes of a subsequent return or audit. We agree with the adviser who
argued that this has preventive value in itself. Because HWIs do see the personal
and financial costs of being on the expanded return program as something they
want to avoid, there is also an opportunity for risk management by setting the
standards the taxpayer has to meet to get off the program. One option would be
to follow the risk assessment on an expanded return with an interview with the
adviser. At that interview it could be common for the analyst to say to the
adviser: 'On the basis of our risk assessment we are expecting your client to
continue on the expanded return program next year unless her circumstances
change so that these compliance risks no longer apply'. Each HWI risk assessment
could include a judgment of which compliance risks would have to be removed
for the risk assessment to be changed to 'low' (and therefore no expanded return).
There could then be some strategic conversations with the HWI adviser about
these risks that are keeping the HWI on expanded returns. In some cases
agreement could be reached on a compliance management plan for getting off
expanded returns.
Conclusion
Notes
1 The ATO acknowledges the comments made in this chapter and notes that they relate to past practices
employed by the HWI Taskforce. The processes employed by the HWI Taskforce have significantly
changed and the taskforce has implemented recommendations outlined in the report.
2 My gratitude to the HWI advisers, officers of professional associations and ATO staff who contributed
to this research. I must say I was impressed by their professionalism. I especially want to thank Alice
Dobes who participated in almost all the interviews conducted for the research and was a constant
source of sound advice as were Andrew Stout and Kevin Fitzpatrick.
3 The main area of contingency fee usage in the past seems to have been with firms that offer to apply
for refunds of sales tax in return for a fee of, say, 20 per cent of the refund.
4 In any case, within the Australian market, an adviser 'looking at the ball from different angles, will
often come up with an opportunity for one client. Then he applies it to a few more clients. Then the
idea might get out and some financial planner or little accountant might market it widely'.
5 The 'credibility and respectability' of investment banks with HWIs, as one adviser put it in our
interview.
References
This book analyses taxpaying and tax collecting from a relational perspective.
Together the chapters provide insight into how each side in this relationship
perceives the other, what each expects and what each gives at a time when the
tax authority is moving from a command and control mode of operation to one
that is more responsive to the environment in which it must operate. The focus
has been on compliance, that is, the extent to which taxpayers do what is
expected of them and are prepared to cooperate with the authority. But if
taxpayers offer compliance, what does the tax authority offer in return?
The usual answer to this question is government services. While not
underestimating their importance, this representation of the obligations on the
other side of the contractual arrangement risks narrowing our sensibilities to
things that we can buy, in theory, in the market place. Governments supply much
more than this:
Somewhat ironically, the continued ability of citizens to act collectively in pursuing their common social
and economic goals through democratically elected governments might rest with the ability of tax
departments to meet the challenges to tax compliance posed by the disappearing taxpayer. That is to say,
what is at stake in the challenge of tax compliance is not just tax revenues, it is democracy itself (Brooks,
1998, p. 33).
In recent years, the concept of integrity has been extended from the individual to
the collective level (Skidmore, 1995; Laufer, 1996). Organisations are said to
display integrity when they demonstrate a capacity to engage in ethical decision-
making. This capacity involves an awareness of the moral issues in play, an
openness to grapple with their complexity, resolve to embed moral
responsibilities within business plans, and the follow through to put morally
responsible decisions into practice (see Petrick and Quinn, 2000). Nowhere is the
tension between doing what is right and gaining the competitive edge more
apparent than in the private sector where profits are the bottom line for success.
A diminished capacity for integrity, however, is not unique to the private sector.
Public sector organisations are not immune from the problem, although its
expression takes a different form (Gregory, 1999; Denhardt and Denhardt, 2000).
A system operating within the public sector can be said to have integrity if it
has unity and soundness of purpose, and if it has processes in place to reflect on
and evolve that purpose in response to community needs. Unity infers neither
singleness of purpose nor institutional simplicity. Instead, it conveys
connectedness in that an organisation's many goals are pursued and its many
processes are implemented with awareness of and responsiveness to each other.
In sum, unity implies at least loose coordination among parts such that, through
reasoning and reflectiveness, an operational story can be told as to how the parts
combine to form a valuable and purposeful whole.
An absence of integrity on the unity criterion may occur when goals are
pursued at lower levels of an organisation that are antithetical to the
organisation's overarching goals. An example would be the introduction of a
collection system for a specific kind of tax that jeopardised protection of the
revenue overall. We might call this integrity strain of a vertical kind in that
actions taken at lower levels of the organisation are contrary to those expected at
higher levels if the organisation's overarching goals are to be met. At other times,
different functional groups at the same level in an organisation may fail to share
information and may act in ways that undermine the capacity of the other to
meet its objectives. Oversights of this kind describe a shortfall in integrity of a
horizontal kind.
While unity conveys some overall coordination of effort, soundness conveys
the moral appropriateness of the direction of effort. Within a democracy, the
ultimate judge of the soundness of purpose of an institution should be the people.
Integrity, in the sense of standing by goals and processes supported by the people,
involves not only commitment to the system in its current form, but also
commitment to revision of the system so that it can continue to serve those it
represents. Service entails responsiveness to government in relation to policy, as
well as listening to the community and 'helping citizens articulate and meet their
shared interests' (Denhardt and Denhardt, 2000, p. 549).
Unity, soundness of purpose and responsiveness do not in and of themselves
satisfactorily denote institutional integrity. Calhoun (1995) has observed that, at
the level of the individual, integrity involves more than developing coherent well-
reasoned positions on issues and standing by these commitments. Integrity, so
conceived, can be deemed self-indulgent. The extra element that needs to be
considered to sharpen this analysis of integrity is a social dimension. Integrity
requires a person not only 'to stand by' a position, but also 'to stand for' that
position when faced with others whose deliberations may have led them to a
different outcome (Calhoun, 1995).
When extrapolated to the level of the collective, high integrity involves
institutional engagement. But institutions cannot engage: That task must fall to
the leadership ranks of the organisation. In the public sector, those with
leadership responsibilities demonstrate the integrity of their organisation through
communicating to government and the community what they stand for and why
it is important. The often-cited public service phrase, 'to give frank and fearless
advice to government',3 exemplifies this aspect of integrity. The institution will
fall short on integrity when '[s]ocial circumstances., erect powerful deterrents to
speaking and acting on one's own best judgment' and thereby 'undermine the
possibilities for deliberating about what is worth doing' (Calhoun, 1995, p. 259).
Calhoun hastens to add that while integrity involves having a proper regard for
one's own judgment, it should not be conceived as 'just a matter of sticking to
one's guns' (p. 259). 'Arrogance', 'bullying', 'defensiveness', 'incivility', 'close-
mindedness', and 'deafness to criticism' are not qualities normally associated with
integrity. Acknowledging that others 'must themselves abide by their best
judgment seems part of, not exterior to, acting with integrity' (p. 260). integrity,
therefore, can pull members of the collective in different directions,4 creating a
diversity of positions and tensions when these differences surface. Consequently,
integrity does not imply orderliness in thought and action. On the contrary, it is a
concept that reflects capacity to navigate through messiness, showing leadership,
while allowing diverse opinions to be expressed, frankly and fearlessly, under an
umbrella of an overarching shared purpose, derived through deliberation with
citizens.
When this concept of integrity is applied to a tax system, the question arises as
to the level at which the analysis should proceed. Should integrity be applied in a
more focused way as to how taxpaying records are compiled and used for high
wealth individuals, or to the way in which the law is written in relation to
business taxation, or to the enforcement practices used by tax office auditors? An
analysis of integrity can be undertaken in any of these domains. For the purposes
of this chapter, integrity will be discussed at a broader level with reference to the
overarching objectives and principles that govern the operation of the tax office.
The purpose of the Australian Taxation Office (ATO) is 'to shape and manage
systems which support and fund services for Australians, giving effect to social
and economic policy' (Australian Taxation Office, 2000a, p. 2). Associated with
this purpose are five guiding principles: (a) to act with integrity; (b) to anticipate,
identify and manage issues in real-time; (c) to be open and accountable with each
other and clients; (d) to offer solutions that are in the community interest, while
matching the individual circumstances of clients; and (e) to meet internal and
external obligations (Australian Taxation Office, 2000a, p. 2). To effect these
principles, the ATO has 12 standards expressed in the Taxpayers' Charter to guide
tax officers in their dealings with taxpayers in the course of administering the tax
system (Australian Taxation Office, 1997): Taxpayers are to be treated fairly and
reasonably, to have their privacy respected, to be treated as honest in their tax
affairs unless the taxpayer acts otherwise, to have decisions explained to them, to
be offered assistance, advice, and information in a professional way, and to be
helped to minimize their costs in complying with tax law.
Even at the level of general principles and codes of conduct, tensions arise. For
example, is maximizing revenue collected 'to support and fund services for
Australians' compatible with abiding by the Charter so as to 'be open and
accountable'? In theory, the answer is yes, but field staff sometimes appear to be
less sure (see Job and Honaker, Chapter 6, this volume). The tension can be
illustrated through the following questions: Are authoritarian tactics justified in
cases where tax officers are chasing down taxpayers who are evading their tax?
Or to put it another way: Are Charter consistent tactics costly for a tax office and
the public when dealing with unscrupulous tax evaders?
In order to ease these challenges to system integrity, the ATO adopted their
Compliance Model (see Braithwaite, Chapter 1, this volume for a full description).
Persuasion and education are the preferred methods for eliciting compliance and
are assumed to be the most appropriate starting point for dealing with
noncompliance in the absence of information that the wrongdoing is deliberate
and likely to be repeated, in this way, the intentions behind the Taxpayers'
Charter are put into practice. While education and persuasion denote the
preferred starting point, tax officers and taxpayers know of a range of
sanctioning options that can be and will be brought into play should taxpayers
fail to cooperate. Consistent with the Charter and in keeping with the ATO's
obligation to protect the revenue, the Compliance Model puts taxpayers on notice
that tax officers will systematically increase the costs of non-compliance, while
always holding the door open for a more responsive and cooperative relationship.
In this way, the Compliance Model guides the ATO toward using its full
enforcement capacity only when taxpayers have clearly or repeatedly signalled
unwillingness to cooperate.
At the level of principles and codes of practice, the integrity story within the
ATO is quite impressive. But action does not flow directly from these principles
and codes. There is another plank that is necessary to define the way in which
tax officers and taxpayers should interact in pursuit of the ultimate goal of
supporting the Australian democracy, tax law. Tax law provides the rules that
determine taxpayer obligations and the enforcement capabilities of tax officers. In
most areas of this large body of law, it is difficult to find principles or guidelines
that provide a meaningful template for making sense of these rules, and linking
them to the other components of the system (Braithwaite, 2002). Furthermore,
there is no evidence of a shared understanding between citizens and their elected
representatives about how and why these rules evolve as they do (Ralph Report,
1999). In addition to the rules derived from tax law and ATO rulings, there are
rules entrenched in the administrative system that define the work roles of staff,
the reward structures of the organisation, record keeping and data storage
capacities, and the reasons for and methods of communicating with taxpayers.
Communication revolves around an elaborate system of automatically generated
letters to taxpayers to inform them of their obligations, to query their actions, to
deliver refunds and payments, to communicate failure to comply, and to impose
penalties, all of which have an institutional history of their own.
These legal and administrative rules, some formal and some informal, are the
centrepiece of operations for the Australian Taxation Office, and for the most part
predate overarching objectives, the Taxpayers' Charter and the ATO Compliance
Model. One might postulate that in a large bureaucracy, such as a tax authority,
practices are more likely to flow from the institutionalised formal and informal
rules, and less likely to flow from semi-detached blue prints introduced relatively
recently in the history of the organisation to give it greater legitimacy in the eyes
of the public. This means that the main challenge for a tax authority seeking
institutional integrity is to convert democratically responsive principles of action
into concrete operations and routines in the day-to-day practices of tax officers.
This is the same challenge faced by corporations seeking regulatory integrity
through turning their regulatory blueprints into meaningful practices (Parker,
2002). One of the central propositions of this chapter is that just as regulators look
for substantive compliance in the actions of those they regulate, citizens look for
substantive integrity in the authorities that seek to direct their actions.
The integrity that taxpayers observe in the tax system and its administration
may not be the same as the integrity that tax officers see from within. When
integrity is perceived to exist in the system by tax officials and those who are
experts in its operation, it may be claimed that the system has passed the test of
internal integrity. In the eyes of those who know the system well, the components
are connected in such a way that high performance in one part enhances, or at
least does not detract from performance in other parts, the overall purpose is
sound, and the organisation is responsive, able to evolve to meet community
needs. A system that is supported by a democracy, however, has to be
accountable to the electorate. External integrity is, therefore, equally important
and involves an additional level of scrutiny. Citizens must be able to observe
integrity in the way in which the tax system is designed and operates, and in the
way in which the tax authority deals with citizens and other branches of
government.
Whether the perspective is internal or external, finding holes in the integrity of
a complex system is always going to be relatively easy. Furthermore, to conceive
of social systems without a good proportion of messiness is unrealistic and
undesirable if any kind of innovation or social change is to occur. Thus, integrity
should not be reduced to a score out of 10 on a checklist. The key to assessing the
integrity of an organisation does not lie in how problem free the organisation
claims to be; but rather it lies in the organisation's capacity to acknowledge
difficulties. Where integrity is present, the organisation should be able to
demonstrate awareness of and deliberation about departures from the standard,
and how the actions taken to improve the situation were in the interests of sound
democratic governance.
With this in mind, the focal point for analysis in the remainder of this chapter
will be one particular kind of strain on integrity in tax offices worldwide; that
which occurs when overall integrity is pitted against compliance. This problem is
likely to be particularly acute for all regulatory agencies where short-term gains
threaten long-term interests. The pursuit of short-term gains puts citizens and
authorities at loggerheads. For citizens, integrity is what is expected of an
authority at all times. For tax authorities, on the other hand, the day-to-day
business is ensuring citizen compliance. At the organisational coalface, as
performance targets are being set for a workforce, compliance is more tightly
bound to the tax authority's short-term interests than integrity. Integrity comes
into focus when there is time to stand back and consider long-term interest.
Integrity also increases in salience, relative to compliance, when a tax office's
accountability is called into question, or when legitimacy is called into play to
deal with change or failure in the system.
From the perspective of the regulatee or the taxpayer, the compliance question
can be subdivided into two components: (a) receiving the request and processing
it cognitively as the regulator expects, and (b) acting upon the request in the way
the regulator expects. Failure to receive the request as intended might occur if the
taxpayer has inadvertently misinterpreted the request (e.g., misunderstanding the
difference between franked and unfranked dividends), or if the regulatee gives
the request meaning that undermines the intent of the regulator (e.g., playing for
the grey in tax law). Similarly, a failure to act upon the request, as when a tax
return is filled out incorrectly, may be due to misunderstanding or carelessness,
or it may be a case of just 'saying no'.
As we focus on each of the 16 cells in Figure 13.1, the many faces of
compliance from an administrative perspective become apparent. If we look for
the cells in which compliance is 'on track' or 'achieved', we find only four of the
16 possible outcomes meeting the criteria of success. The majority of the possible
outcomes deviate from the classic case of compliance.
Taking the 'on track' or 'achieved' compliance outcomes first, it is helpful to
ground the discussion by considering the types of tasks undertaken by a tax
authority that come closest to meeting the specified criteria of a request being
made and received as intended, and being visibly acted on as intended. One such
example would be declaring bank interest on an income tax return. Data
matching with bank records makes it possible for tax officers to keep a close eye
on interest declaration for tax purposes. For the tax officer responsible for
monitoring compliance of this kind, attention is likely to focus on detail, such as
the integrity of the data bases used to match records and tweaking the system to
catch those few who are slipping through the net. Compliance questions are
likely to take the following form: Can detection rates be improved? Can data on
non-compliance be matched with data of other kinds of non-compliance? Should
the taxpayer be relied upon to self-report on tax matters and self-assess his/her
tax contribution? It is equally valid to ask how reliable is the official data on non-
compliance that is stored by the tax authority? What kind of auditing is
necessary to ensure that the standards of compliance are maintained at high
levels and that the community's standards match those expected by the tax
authority? An equally relevant, but less often asked question for tax officers
working with 'on track' or 'achieved' compliance, is how can the public be
recognized for their cooperation?
Turning to the system failure cells, compliance now has less to do with
assessment and more to do with engaging with the taxpayer. System failure
occurs when the tax office has not made a direct request to the taxpayer
concerning her obligations. In these cases, compliance means getting people to do
something that they have not been doing 'naturally' either because it never
entered their heads or because they chose not to do it. In addressing system
failure, the focus of attention is likely to be on communication of the message.
The clarity of the request, the educational apparatus supporting the request, and
the political intent behind the message are important issues for consideration.
Whenever questions are raised about system effectiveness, reasonable doubt also
is raised over the extent to which a regulatee should be held responsible for non-
compliance.
Compliance failure, on the other hand, is likely to be addressed in different
ways from system failure. If the system appears to be working well from a
communication perspective, non-compliance can be dealt with in either of two
ways. One option is to listen to the community and understand the grounds of
resistance so that changes can be made to policy and practice. As a result,
compliance may be made easier for taxpayers, or the costs of non-compliance
may be increased through greater surveillance and penalties. Another option is to
change tack altogether. This option would not require taxpayers to understand
the message. Technology may be used to engineer compliance that has not been
forthcoming previously, and is represented in Figure 13.1 as 'bonus' compliance.
The 'bonus' cells are those in which an authority has its messages received and
acted upon by a target without the authority having to make a direct request.
Architectural strategies (Shearing and Stenning, 1985; Coglianese and Lazer, 2001)
may be used by an authority to ensure compliance through making it not only
the act of least resistance, but also the act that no longer requires a decision. For
example, the provision of software that leads a taxpayer through a series of
questions to estimate tax owing creates 'bonus' compliance for an authority in
that various options for non-compliance are removed from consideration. This is
not to deny the possibility of creative compliance while using the software
package, but rather to recognize the non-compliance that is avoided through a
process that generates compliance among taxpayers prepared to engage with a
tax authority on 'automatic pilot'. The key compliance questions in these
circumstances are big picture questions concerning legitimacy: Will segments of
the population take exception to the technological imposition of compliance and
challenge the authority of the tax office as a result? Or is architecturally guided
compliance seen as a time saving and security oriented device that serves
everyone's interests well?
Increasingly, tax officials are searching for ways of combining architectural
regulatory strategies with 'natural systems' for 'best practice'. For instance,
computer software that has the primary purpose of helping a small business
owner manage the business better may also produce information for tax
purposes, together with instructions on how to compile these data to meet tax
reporting requirements. Key questions emerging from this perspective include
how 'best practice' is learnt, how requests from regulators can be re-phrased to
reflect 'best practice', and how compliance costs can be reduced for both
regulatees and regulators.
'Bonus' compliance is an outcome that in itself has many facets. Other
strategies that can be used by a tax authority to cultivate the growth of 'bonus'
compliance involve reliance on principle versus rule based law (Braithwaite,
2002), or support for professional codes of practice for tax agents and tax lawyers
(National Review of Standards for the Tax Profession, 1994). In these
circumstances, the application of principles and codes can frame cognitive
processing, steering taxpayers and their agents away from risky schemes that are
on the border of avoidance and evasion.
The remaining cells in Figure 13.1 are called 'black holes' because regulators
cannot see what becomes of their requests for compliance. In these
circumstances, what it means to comply is the single most important question
asked by the regulator. Tax officers search for imperfect indicators that will cast
light on whether or not a response has been made to the request, and the options
must be evaluated in terms of their relative merits. These are the circumstances
where tax officials rely on informants or third parties to report wrongdoing that
comes to their attention, and scan data bases in their search for inconsistent
patterns that may flag cases of concealed tax evasion.
Among the options to be considered by administrators who oversee
compliance problems of this kind is transferring responsibility for monitoring
compliance to another party, possibly even the regulatee. Under these
circumstances, regulatees would be accountable to the regulator through having
procedures in place to regulate their own compliance, in other words,
demonstrable self-regulation. In many complex areas of regulation, regulators
require regulatees to outline their inhouse strategies for managing risk instead of
searching for compliance with externally imposed standards (Grabosky, 1995;
Gunningham and Grabosky, 1998; Braithwaite and Williams, 2001).
Different facets of compliance - assessing compliance outcomes for accuracy,
defining at least partially relevant outcomes where none are visible, evaluating
the delivery of the request, understanding the 'no' response, and capitalising on
'bonus' compliance - are of interest across the tax office, but differ in their
importance and relevance, depending on the nature of the request and the
required response. Those with responsibility for the cash economy are likely to
focus on 'black holes' and be drawn toward the re-design of the tax system to
introduce greater transparency and accountability, and the co-option of other
regulatory agencies to assist in monitoring and containment (Cash Economy Task
Force, 1998). Tax officers dealing with large business might see playing for the
grey as their major problem, an example of compliance failure that might be dealt
with through developing better intelligence systems, targeting audits more
effectively, and building capacity to elicit cooperative taxpaying behaviour
(Braithwaite, Chapter 9, this volume). Wage and salary earners who invest in
mass marketed tax avoidance schemes present another kind of challenge to a tax
authority. Tax officers struggle with system and compliance failure tangled
together so tightly that the way forward is uncertain (Senate Economics
References Committee, 2001). The priorities are likely to involve tightening laws,
building alliances with professional bodies, educating the public, and amending
tax returns. Penalties in such circumstances become a regulatory tool that can
have mixed results: A penalty indicates seriousness of an offence to the
community, but in such circumstances penalties can create a backlash of cries of
unfairness that might slow the process of reform (Hobson, 2002; Murphy, 2002a,
2002b). Ordinary taxpayers, on the other hand, who lodge their tax returns late,
fail to declare income and over-claim deductions present the most
straightforward compliance problem for tax administrators. It is not surprising
that tax officers regard compliance in this context as a question of the taxpayer
not showing due care; a problem that is most times dealt with through the
application of appropriate penalties to impress upon taxpayers the need to give
tax matters their most serious attention in the future.
The above discussion illustrates how understanding and improving compliance
requires a tax authority to entertain multiple conceptions of compliance and to be
able to change the frame of analysis for a compliance problem as attention moves
from one area to another, that is, from personal income tax, to large corporate
tax, to small business, to the cash economy, for example. For this reason, the
question of finding strategies to improve compliance is one that, in the first
instance, is best answered at the local level. Field operatives have the intelligence
to analyse compliance problems in terms of the request and action framework
provided in Figure 13.1; and furthermore, they have the experiential base for
compiling a rich array of creative and workable interventions to deal with the
problems they have observed. For tax administrations that are risk averse and
that traditionally operate within hierarchical structures, devolving problem
solving to local areas is a radical proposal (Job and Honaker, Chapter 6, this
volume). Yet, as Sparrow (2000) has pointed out, tax administrations have started
to grapple with this challenge and have produced evidence that institutional
change is possible. Through setting up special purpose task forces that draw on
the local knowledge of field staff and bring together an effective skill mix from
different sources, strategies for improving compliance in relation to specific
problems have been found.
Improvements in compliance are among the major yardsticks used by tax
authorities and their governments to assess their performance. Commonly,
compliance gains are considered against compliance costs, which refer to the
financial and opportunity costs of compliance borne by taxpayers and tax
officers. But there also may be a cost to the integrity of the tax system.
Compliance gains may mean integrity loss, at least in the view of the public.
To understand the compliance integrity trade-off, consider the following
example, one that is currently eating away at the hearts and minds of ordinary
Australians (see Braithwaite, Reinhart, Mearns and Graham, 2001). While tax
officers dealing with personal income tax scrutinize data sets to identify
irregularities in self-assessed tax returns, tax officers dealing with large
corporates and high wealth individuals work to establish cooperative
relationships with their clientele in order to collect at least some tax from those
who have the capacity to avoid paying any tax at all (Australian Taxation Office,
2000b). From a local compliance perspective, both components of the system, that
is the personal tax and large business lines, work well in achieving their goals.
For personal tax, data matching software increases the revenue collected at
minimal cost to the tax office. A letter is issued informing the taxpayer that their
tax assessment has been amended, in all likelihood with an invoice for tax owing
plus interest and a penalty. The taxpayer can contest the assessment, but in all
likelihood the taxpayer will defer to tax office authority. In the large business
line, however, the interaction between taxpayer and tax officer is likely to take a
different course. For large corporations, negotiation and regulatory conversation
triggered by lawyers' letters will often be necessary to ensure that at least some
tax is collected at regular intervals.
The important point to note here is that there is soundness in operations
within each of these functional lines in the tax office: There is a consistent
compliance goal operating in each part, that is, to maximize the revenue collected
as efficiently as possible, given the context. The contexts, of course, differ
substantially. Different types of tax are collected, different laws are applicable,
and the resources available to take advantage of legal complexity and ambiguity
are far greater among corporates than among ordinary taxpayers. At the local
level, the compliance activities of tax officers working in personal tax and those
in corporate tax seem defensible and reasonable. But there may be an unseen cost
to the tax office overall, that is, a perceived loss of institutional integrity.5
Conclusion
The above analysis reveals why finding the optimal mix of compliance and
integrity is no easy task. At the end of the day, the quality of the solution to a
compliance integrity dilemma rests on the experience and wisdom of senior
bureaucrats. What the above analysis can offer, however, are three principles that
may be useful in understanding how good decisions come to be made by senior
tax managers when a compliance integrity dilemma arises.
First and foremost, both compliance and integrity can be boosted by investing
in the human dimension of taxpayer management. Fundamentally, this means
making a concerted effort to build a shared understanding with the community
about what a tax system does and how it is best designed. Taxpayer management
extends from the general to the specific. At the general level is the task of
educating the community about the importance of a tax system and persuading
them of its value. Outlining the principles for and methods of tax collection for
the public and committing the organisation to the effective monitoring of
compliance are further steps that a tax authority can take in an effort to win
public support and establish a cooperative and responsive relationship with the
community at large. At a more specific level, taxpayer management involves not
only the pronouncement of how tax law is being interpreted, but also accessible
explanations for these decisions, along with discussions of penalties and
settlement options.
In all these ventures, responsiveness to community concerns is a key feature of
effective taxpayer management. Responsiveness means listening, and publicly
acknowledging and resolving identified problems. Taxpayer concerns should be
taken seriously and engaged with openly, inclusively and thoughtfully against the
backdrop of basic system goals and principles. Sometimes, the result may be a
change in tax office policy and practice. Other times, it may be insistence that the
tax authority's actions are consistent with principles that the community has and
continues to endorse. In no circumstances should responsiveness be interpreted to
mean appeasing special interest groups at the expense of system integrity.
Investment in the human management dimension in the long term should give
rise to a mutually reinforcing relationship between compliance and integrity.
Increased compliance should boost the external integrity of the tax system, and
the external integrity of the tax system should encourage compliance. Investment
in the human management system or taxpayer culture offers a further bonus to a
tax authority facing changes to the administrative and legal system. The
coordination of taxpayer management strategies is not only desirable, but also
practicable and relatively uncomplicated. Just as it can build cohesion in the
community, it can provide a basis for unity in a tax authority as difficult trade-
offs are made redesigning the legal and administrative system.
The second principle to emerge is that administrative and legal adaptability is
a necessary part of building internal tax system integrity, although the immediate
effects on compliance can be uncertain, as can the immediate effects on both
internal and external integrity. In such circumstances, tax authorities need to be
strategic, taking a long-term view, and putting the infrastructure in place so as to
manage unexpected outcomes. Part of being maximally responsive to the
environment is having a finely tuned intelligence system and a rich network of
support that can be part of the process of testing and redirecting change until an
appropriate balance between needs for compliance and integrity are met.
While senior managers must take responsibility for making critical decisions
about balancing compliance and integrity, knowledge and understanding of how
compliance might be best improved lies in the stories of field staff. It is only
through drawing on these stories and opening and acting on communications
from the coalface, that senior managers are able to pre-empt the trade-off costs
for their organisation. Without bottom-up feedback, top-down management of
administrative and legal change is likely to be fraught with difficulties and
setbacks. Thus, the third principle emerges, one that involves an integration of
compliance and integrity action plans. Through adopting a bottom-up approach
that carries tales from the field, tax authorities can develop a range of plans of an
administrative and legal nature for improving compliance. From a review of
compliance management plans, senior managers can choose a preferred sub-set
that will improve compliance outcomes and build tax system integrity. To choose
poorly (the top-down process) or to have poor options from which to choose (the
bottom-up process) risks long-term damage. Failure to administer the tax system
in a way that demonstrates basic respect for the democratic principles of
participation and accountability is a dangerous game. A tax authority that de-
legitimizes itself in the eyes of citizens limits its effectiveness and short-changes
citizens in terms of what they can expect from democracy.
Notes
1 An earlier version of this paper was presented at the 'Compliance Workshop: What Does it Mean to
Comply?', organised by the Regulatory Institutions Network, 17 June 2002, Australian National
University. My thanks to Barbara Nunn for broadening my perspective on integrity, and to my
colleagues, Clifford Shearing, John Braithwaite and Greg Rawlings, for their insightful and
constructive criticisms.
2 A distinction is drawn between the community's perspective, within a context where consultation is
genuine and inclusive, and the community's reaction, within a context where a change is imposed.
Community reaction provides a distorted view of the community perspective because resentment at
being excluded from the decision-making process can be mixed with genuine deliberation about the
substantive issue.
3 The importance placed on giving frank and fearless advice is reflected in the public service code of
conduct (see Australian Public Service Commission, 2000).
4 Integrity can also pull individuals in different directions internally (see Calhoun (1995) and Dobel
(1999) for discussions of why integrity is not necessarily at odds with ambivalence, inconsistent
actions over time, or value conflicts).
5 The Community Hopes, Fears and Actions Survey revealed much less community confidence in the Tax
Office's capacity to respond effectively to taxpaying defiance from high wealth individuals and large
corporates than to ordinary taxpayers and small business (Braithwaite et al., 2001).
References