The Emerald Research Register for this journal is available at The current issue and full text archive

ve of this journal is available at

http://www.emeraldinsight.com/researchregister http://www.emeraldinsight.com/0142-5455.htm

HIWPs and
High-involvement work bargaining
practices and employee power

bargaining power 453

Frederick Guy
School of Management and Organisational Psychology, Birkbeck College, Received December 2002
Revised December 2002
University of London, London, UK Accepted January 2003

Keywords Empowerment, Negotiating, Working practices, Retailing,

Customer service management
Abstract High involvement work practices (HIWPs) may empower employees to do their jobs
better, and also empower them at the bargaining table. This paper considers whether
non-universal adoption of productivity-enhancing work practices may, at least in part, be explained
by this dual nature of empowerment. It examines the case of a customer service programme in the
Northern California division of Safeway stores, its affect on the outcome of a strike against
Safeway, and the subsequent pattern of adoption (and non-adoption) of similar programmes
among Safeway’s competitors. It concludes that the dual nature of empowerment can help explain
the apparent paradox posed by empirical studies; that although HIWPs improve the performance
of all sorts of organisations, most organisations do not adopt HIWPs.

Introduction
A range of work practices, which have been labelled both “high-involvement
work practices” (HIWPs) and high-performance work practices (HPWPs), are
seen by many as ways in which the active involvement of ordinary employees
can improve the performance of firms (hence the labels; I will use the HIWP
label below). Two controversies surround HIWPs. One controversy is whether
such practices will improve the performance of organisations as a general rule,
or only that of organisations with certain strategic needs; the former position
has been called “universalist”, the latter “strategic contingency”. Empirical
studies mostly find that HIWPs improve performance for organisations
regardless of strategic need, which lends support to the universalist position.
On the other hand, these and other studies find that HIWPs are nowhere near
universally adopted; if the universalist theory is correct, we are left with the
need to explain why these universal benefits are so often left on the table.
The second controversy concerns whether, and in what respect, HIWPs
empower employees. The view of HIWP advocates is that the practices in

The author wishes to thank: Carolina Otero for research assistance; Richard Benson, Jim Liggins
Employee Relations
and Gary Smith of UFCW Local 870, Safeway employees Jay Klein, Brenda Grissom and Vol. 25 No. 5, 2003
Diane Poe, and Karl Kruger, for their time and cooperation; Jonathan Michie, Elaine Marriole and pp. 453-469
q MCB UP Limited
two referees for comments on an earlier draft; and the School of Management and Organisational 0142-5455
Psychology, Birkbeck College, for financial support. DOI 10.1108/01425450310490165
ER question work because they allow employees to make decisions that make a
25,5 difference. Decision making by lower-level employees may improve
performance for various reasons: for instance, because it allows employees to
use tacit knowledge which is not available to higher level managers, or because
the freedom to make decisions (alone or associated with performance pay)
provides motivation for greater effort. On the other hand, the enthusiasm for
454 HIWPs coincides with a tremendous improvement in monitoring and control,
through the application of new information and communications technologies
(ICTs). Some observers reckon the restrictions imposed by the latter more than
outweigh the freedoms granted in name of the former, with HIWPs acted out in
an electronic panopticon.
Considering the empowerment controversy side by side with the
universalist/contingency controversy raises the problem of defining
“performance”. Most of the universalist/contingency literature ignores the
possibility that employers and employees may have conflicting objectives, and
measures organisational performance from the perspective of a
profit-maximising employer. If there is no conflict between employees and
employer, this choice of performance measure does not matter; if there is no
conflict, then profit maximisation will occur when productivity and value
added are maximised, which is to say when the pie to be divided among
employer and employees is as large as it can be. The empowerment question,
however, raises the issue of conflicting interests. If employees are empowered
or disempowered, which is to say if changes occur in the extent to which they
can make decisions about their jobs, this may be reflected in changes in the
relative bargaining power of employees and employers. This can drive a wedge
between productivity and profitability.
A wedge between profitability and productivity in turn re-casts the
universalism/contingency question. It is possible, for instance, that HIWPs
improve productivity as a general rule, but improve profitability only for firms
with particular strategic needs. Moreover if, following the adoption of a HIWP
programme, changes in bargaining power are slower to take effect than
changes in productivity, then the HIWPs may have a positive effect on profits
in the short run but a negative effect in the long run; the reverse may be true if
the changes in bargaining power take place more quickly than those in
productivity.
In this paper, I examine the case of a particular HIWP in one division of one
company: a customer service programme adopted by Safeway Stores in its
Northern California division. The programme aimed at empowering employees
to meet customer needs and at establishing a closer relationship between
customers and employees. The programme was judged a success by Safeway
management. However, by strengthening the relationships between employees
and customers, the programme also appears to have empowered employees in a
way not intended by management, strengthening the hand of the employees’
union, the United Food and Commercial Workers (UFCW) when bargaining for HIWPs and
wages and benefits. We consider the strategic factors that make the bargaining
programme important enough for Safeway that it maintains the programme power
despite its contribution to employee bargaining power. We also consider the
market positions of Safeway’s principal competitors in relation to their
decisions to adopt, or not adopt, similar programmes.
The next section of the paper presents the methodology. The following
455
section reviews theories of HIWP in relation to both performance and
empowerment, together with both theoretical and empirical literature on these
questions. The following section examines the Safeway/UFCW case. The
penultimate section considers implications of this case for theories of HIWP.
Conclusions are presented in the final section.

Methodology
The idea for this paper came from a conversation with Patricia English of
Menlo Park, California. Ms English told me of a conversation she had with a
clerk on the picket line at the Safeway store in Sharon Heights (an upper middle
class neighbourhood in the Silicon Valley) during the 1995 strike. The clerk had
told Ms English that very few customers were crossing the line, and attributed
this to the fact that for about a year clerks had been required to thank all
customers by name at the point of sale, and for this reason knew many of them
by name.
The empirical contribution of this paper comes from ten semi-structured
interviews with Safeway employees who are UFCW members, UFCW officers
who deal with Safeway, and former managers of stores competing with
Safeway. All of those interviewed were employed by Safeway, the UFCW or a
Safeway competitor at the time of the 1995 strike. One interview was conducted
in December 2000, the rest in April 2002.
The Safeway employees are all clerks, cross-trained between working the
front end (checkstands, or tills) and restocking shelves and coolers. One of them
had worked as a trainer in the superior customer service (SCS) programme and
then returned to being a clerk. These interviews provide us with information
about the nature of SCS, the employees’ perception of how it affects their jobs,
and their perception of its effect on their relationship.
Safeway management declined to discuss either the SCS programme or their
relations with the UFCW.
The information from these interviews is, of course, insufficient to establish
that the SCS programme has caused either an improvement in the productivity
of Safeway employees, or that it has caused an increase in their bargaining
power. It does establish that a number of people with good first-hand
knowledge say that both are so, and make a case for this. The information from
these interviews also presents a clear and plausible picture of how a HIWP
programme could increase both productivity and employee bargaining power,
ER whether or not these things happened in this case. In light of the scarcity of
25,5 other evidence on the question of how HIWPs and employee bargaining power,
we believe that the case is worth reporting, and relating to the empirical and
theoretical literature on HIWPs, performance, and power.

456 HIWPs, performance, and empowerment

Adoption and non-adoption of HIWPs
There is little agreement as to precisely what HIWPs are (Becker and Gerhard,
1996; Edwards and Wright, 2001). I will adopt a minimal definition, namely
that these are practices which attempt to improve organisational performance
by increasing the range of actions, decisions or relationships for which
employees are intended to be responsible. This definition covers the use of
self-managed teams, but is not limited to situations with teamwork; it is more
specific than the flexible deployment of labour, or multi-skilling, in that it
requires that employees be responsible, recognising the need to move from one
task to another (though it is hard to draw a clear line between the two, since the
“responsible” employee may be deciding change tasks in accord with some
clearly prescribed contingency guidelines); it ignores many of the HR practices
often associated with HIWPs, such as selection, training, and evaluation; it also
ignores the vexed (Guest, 1998) issue of commitment. In adopting such a
minimal definition, I am also failing to deal with the possibility that minimal
HIWPs are not effective HIWPs, raised by several studies in manufacturing
industries which have found that certain bundles of practices are necessary for
effectiveness (Arthur, 1994; Ichniowski et al., 1997; MacDuffie, 1995). The
definition here is more in the spirit of Drago (1996), who finds employee
involvement in both “transformed” and “disposable” workplaces (but not in
“traditional” ones); the first of these is a high commitment and the second a low
commitment type.
A problematic aspect of our definition, and an issue central to this paper, is
the meaning of “performance”. Shareholders, managers, writers on corporate
strategy, and scholars with no other data, tend to define performance as
profitability. We define performance as labour productivity, where the
productive factor is the disutility of labour to the worker (so that sweated
labour is not necessarily more productive, even if it produces more per hour,
because the subjective input from the workers is greater). The productivity
definition is preferable both from the standpoint of economic theory, and from
its ability to accommodate a stakeholder or a bargaining analysis. We will,
however, occasionally make reference to the profitability criterion, for two
reasons: first, because we assume that it is the criterion used by management
when deciding whether to implement, and whether to continue, a HIWP;
second, because many empirical studies employ it.
Empirical research has favoured the universalist view that HIWPs offer
substantial benefits to the owners of firms, independent of strategy or external
environment (Appelbaum et al., 2000; Delery and Doty, 1996; Guest, 1997; HIWPs and
Huselid, 1995). Edwards and Wright (2001) argue that this conclusion is bargaining
weakened somewhat by the disagreement, among studies, about what HIWPs power
are and varying measures of organisational performance; it might equally said
that the conclusion holds up remarkably well despite different definitions of the
key variables. If the universalist claim is true, the question of why HIWPs are
not more widely adopted is a pressing one. 457
Strategic contingency theories do provide an explanation for non-universal
adoption, holding that the adoption of HIWPs will bring a net benefit to an
organisation only if it also adopts certain strategies. Porter (1985) links HIWPs
to strategies of focus and differentiation, but not to low cost strategies. Miles
and Snow (1984) take a similar line, but link HIWPs to innovation rather than
differentiation/focus. Studies in manufacturing have found that HIWPs are
matched with flexible and high quality production systems (Arthur, 1994;
MacDuffie, 1995; Youndt et al., 1996), which supports the strategic contingency
view. Although statistical studies have found little support for these theories,
we should note that data collected to study the HIWP-performance relationship
may not be well suited to addressing the strategy-HIWP relationship: when
studying the HIWP-performance relationship, the reduction of omitted variable
bias favours focussing on similar firms, or even (as in the studies by
Appelbaum et al. (2000) and Ichniowski et al. (1997)) on similar plants or
processes rather than entire firms; unfortunately, this may either remove
information about strategy (which tends to be firm-level rather than plant- or
process-level), or reduce variance in the strategies observed.

Dynamics of HIWP adoption and implementation

Several explanations for non-universal adoption have to do with what we
might call the dynamics of adoption and implementation of HIWPs. Ichniowski
et al. (1997), in their study of American steel finishing lines, find that HIWPs
are nearly universal in both greenfield plants and plants which had been
recently shut down reopened under new management; they suggest that
universal adoption is slowed by prior investment in relationships not
compatible with HIWPs, and by entrenched mistrust between management and
labour. On the other hand, many studies find that HIWPs often do not last, that
they are “fragile” or that they “decay” (Doeringer et al., 1998; Osterman, 1994;
Pfeffer, 1998). Helper et al. (2002) find that the distribution of effects of HIWPs
on profitability is bimodal, likely to be either somewhat positive or sharply
negative: the positive effects on profitability come from increased productivity,
the negative ones from the costs, when demand is down, of having invested in
relationships with employees. This accords with those explanations which nest
the problem of strategic contingency in a larger institutional problematic,
following Wilkinson’s (1983) work on “productive systems”. Thus the
instability of HIWPs in the USA (Konzelmann and Forrant, 2002) and the UK
(Michie and Quinn, 2002) is explained in terms of institutional environments
ER favouring short-termism by investors and customers, and facilitating
25,5 hire-and-fire employment practices.
This paper advances a bargaining power explanation for non-universal
adoption of HIWPs. The bargaining power story also provides another
explanation for change over time in the fortunes of a HIWP: if a HIWP results
in increased employee bargaining power, and that increase was not anticipated
458 by management at the time an HIWP was adopted, then employee bargaining
power can result in decay of the HIWP.
Although some are about slow take-offs and others are about decay, what
these dynamic stories have in common is the ability to complicate econometric
estimation of the relationship between HIWPs and company performance. In
simple cross-sections, the dynamic effects may be missed altogether. In
principle, longitudinal (panel) data offers a solution to this, but most panels are
short and, as Huselid and Becker (1996) remind us, standard panel estimation
techniques aggravate errors-in-variables bias (Griliches and Hausman, 1986),
with the result that estimated coefficients (whether positive or negative) are
underestimated (biased toward zero). Huselid and Becker (1996) relate this to a
data set in which the performance benefits of HIWP appear to be delayed; in
this case, the positive effects of HIWP on performance are underestimated. But
the same problem would apply: the data deals a decaying relationship between
an HIWP and performance, where decay is correlated with certain company
characteristics (strategy, for instance). In the latter case the coefficients on
variables explaining poor performance would be biased toward zero, providing
erroneous confirmation of the universalist view.

Empowerment, conflict, and the extent and nature of HIWPs

By empowerment I mean the employee’s ability to make choices which make a
difference for the employer. One element in such an ability is the potential
impact of the choices open to an employee at a given point in time. Another is
the likelihood that, after making a certain choice, the employee will be able to
continue making choices. Here we must consider the relationship between
employee choices and the employer’s information and control systems. If every
new choice which the employer regards as correct is soon captured by the
employer’s information system and made part of an algorithm directing future
action, and if choices the employer views as incorrect are promptly detected
and lead to disciplinary measures, there is little empowerment in choice.
If HIWPs improve productivity or profitability through employee
involvement, where involvement means individual and/or collective decision
making by lower-level employees, then HIWPs entail some sort of
empowerment for the employees. But, for the reasons just alluded to, there is
considerable debate about whether, and to what extent, HIWPs result in a net
change in employee power. Consider two aspects of this: first, does the explicit
empowerment offered by the HIWP replace a pre-existing (and possibly less
restrictive) tacit empowerment of the older system, in which rules while rigid
may have been imprecise, or honoured more in the breach? Systems that can be HIWPs and
brought to a standstill by working to rule may not be flexible in concept, but bargaining
evidently allow employees significant discretion. Second, is empowerment with power
regard to certain decisions cancelled out by closer monitoring and control,
either over the execution and outcomes of those decisions, or in other aspects of
the job?
While early assessments of HIWPs tended to see employee involvement
459
either as a substitute for monitoring and control, or as nothing but a
sophisticated form of monitoring and control, a growing body of evidence
shows the two to be complements (see, for instance, Edwards et al., 1998;
Ezzamel and Willmott, 1998; Frenkel et al., 1998; Sewell, 1998). There is
particularly strong evidence for this in jobs involving customer service (Kinnie
et al., 2000). Indeed, there is reason to believe that tight behavioural controls are
particularly prevalent in customer service functions, independent of the
presence of HIWPs (Ng and Dastmalchian, 1998).
An absence of net empowerment need not imply that the HIWP is somehow
bogus. In a world where employers design organisations to achieve certain tasks
through process design, coordination of activities, and control of employees who
have their own interests, we should not be surprised to see the devolution of
decision making going hand in hand with the development of monitoring and
control. Consider an example of this from outside of the present HIWP question:
the rise of the multi-divisional corporation from the 1920s onward represented a
substantial decentralisation of decision making within large corporations, and
was made possible by vastly improved central accounting controls (Chandler,
1962). The new financial control systems at such companies as General Motors
and DuPont codified, and subjected to measurement and central auditing, large
amounts of previously tacit knowledge, and at the same time created a
framework which empowered divisional managers to respond flexibly to
customers in ways they previously could not do. Decentralisation and control are
two sides of the same coin, more often than not.
In saying this, I do not mean to suggest that the balance of employee
decision making and employer control is some kind of natural constant.
Different work systems may of course involve different degrees of employee
power, as may particular work practices.
The power to make a difference on the job also means bargaining power for
the employee. This may, as in the case considered in this paper, show up in the
outcome of a formal collective bargaining process. Standard efficiency wage
models, however, predict that an increase in an employee’s power to affect
employer outcomes raises the employee’s equilibrium rate of pay, even in the
absence of collective bargaining (Akerlof and Yellen, 1986). In light of this, the
best evidence that HIWPs do, on balance, empower employees, is that they
appear to have a positive effect on pay (Appelbaum et al., 2000; Helper et al.,
2002).
ER That aspects of a HIWP system have the potential to enhance collective
25,5 bargaining power is clear in the case of self-directed work teams, and Fairris
(1997) shows how formal systems of consultation may lay a basis for kinds of
collective action not intended by management. In this paper we consider the
apparently more innocuous practices associated with “getting close to the
customer”. The use of customer relationships by the Teamsters Union was a
460 widely publicised feature of their strike against UPS in the USA in 1999.
Efficient parcel pickup and delivery requires that drivers have special
knowledge of customers, and so UPS keeps drivers on regular routes. There is
little question that these practices contribute to the productivity of UPS
operations. In the run-up to the 1999 strike, the union encouraged drivers to
brief customers on the union’s side of the dispute. What effect this had on the
strike settlement, and hence UPS profitability, is difficult to judge. The case
discussed below has similar features, but interpretation of the consequences is
more clear cut.

Superior customer service and the 1995 strike at Safeway

Safeway’s market position
Safeway Stores is one of the leading food retailers in North America, with
approximately 2,500 supermarkets. 225 of these are in its Northern California
division. (The Safeway chain in the UK was once part of the same organisation,
but is now a separate company.)
Safeway is based in Northern California, and has long been the market
leader there. While that much has remained the same, the marketplace has
changed rapidly: 25 years ago Safeway faced two or three major competitors in
most markets in the region, and these competitors all ran supermarkets which,
by today’s standards of differentiation, looked very much the same. Now
Safeway is in the middle of a much more widely differentiated market. On the
one hand it faces large discount operators (Walmart, Costco and so on), which
operate at lower margins than Safeway or similar operators can achieve; on the
other hand, many of its smaller competitors have been converted into
up-market gourmet and/or natural foods emporia, operating at high margins
but offering extremely high quality and service. Along with increased
differentiation there has been consolidation at the middle of the market, so that
in most communities Safeway faces only one major competitor it its own niche.
The largest of these is Albertson’s, a chain that in national terms is even larger
than Safeway.
The central strategic problem for Safeway, as for any retailer in a similar
mid-market position, is to find a way to make money in this middle ground.
This requires providing service, quality and variety sufficient to justify an
adequate profit margin, while at the same time maintaining high volume. One
element in Safeway’s strategy has been programme called “superior customer
service” (SCS), to which we return below.
Representation of safeway’s workforce HIWPs and
Most non-management employees in Safeway’s Northern California stores, bargaining
about 20,000 employees, are represented by the United Food and Commercial power
Workers (UFCW). Northern California includes nine UFCW locals, but one
contract is negotiated for the entire division. Similar contracts cover employees
at Albertsons; at Cala and Ralphs (both owned by Kroger, another large
national chain); and at some units of Raley’s, a regional chain. These are
461
Safeway’s principal competitors in the mid-market. Some smaller mid-market
supermarkets are also covered by similar contracts. Overall, the UFCW
represents somewhere between 30,000 and 35,000 in this market.
The similarity in contracts has been maintained for some decades through
pattern bargaining, a model contract being negotiated between the UFCW and
one of the major mid-market employers. Raley’s, in its non-union stores, offers
terms nearly identical to those of the UFCW contract. Some up-market stores
are also represented by the UFCW, usually a legacy of these stores’ history as
small mid-market operators. None of the employees in any of the discount
chains, however, is represented by the UFCW. The workforce in most of these
stores is not unionised, and where it is it is represented by the Teamsters, on a
contract which is substantially inferior from the employee’s point of view.
As with the British supermarket clerks studied by Grimshaw et al. (2002),
the overall experience of these employees in the flexible, responsive markets of
the late twentieth century has not been a happy one. In the early 1980s, before
the period we are studying, these jobs were altered substantially by the
introduction of bar codes, optical scanners, and the associated information
systems. These systems enabled retailers to serve customers more flexibly, for
instance by analysing sales data for individual stores and tailoring product
lines more closely to the various communities served, and also by more closely
integrating the supply chain. They also reduced costs in a number of ways;
among these was the reduction of shrinkage (theft) both by pinpointing
products subject to shrinkage, and by preventing clerks from under-ringing
(giving spontaneous discounts to friends and relatives). The prevention of
shrinkage was a marked disempowerment of clerks, removing from them the
moral choice of whether or not to engage in theft. Operation of scanners also
requires less skill than entering prices on a keypad. Whether due to
disempowerment or deskilling or both, the clerks in question saw their real
wages fall substantially between 1975 and 1995, reversing decades of growth,
in spite of continued representation by the same union and in the midst of the
unrivalled prosperity of a region centred on the Silicon Valley (Pelletier, 2001).
Increases both in responsiveness to customer wants and in logistical agility,
wrought by bar codes and associated technologies, fostered changes in the
retail food market in the USA, making it both more segmented and more
competitive. It was as part of their response to this new competitive situation
that Safeway Stores adopted, in its California divisions, the SCS programme
ER discussed below. Thus, the very supply chain technologies that have deskilled
25,5 the jobs of supermarket clerks, and made them subject to much closer
monitoring, have also created conditions which, at least in this company, call
for new customer service responsibilities from those same clerks.

462 SCS
In 1994, after two years of more limited experimentation, Safeway had
launched a programme called “superior customer service” (SCS) in its two
California divisions. Under SCS, employees are trained to shift from what the
company calls a “task orientation” to a customer service orientation. The task
orientation entailed completing assigned tasks in a prescribed manner, e.g.
stocking the dairy case in the morning in a certain way within a certain amount
of time. Under SCS, a Safeway employee doing this same task is also expected
to anticipate customer needs, and if noticing that a customer looks a bit lost, to
make eye contact and ask if they need help; if the customer then asks where to
find a particular product, to escort the customer to the place where that product
should be found; if the product is out of stock, to suggest a substitute; and if the
out of stock product had been advertised at a reduced price, to offer the choice
of a raincheck (the same product at the reduced price at a later date), or a price
reduction on a substitute product. Thus, an employee detailed to stock the
dairy case, on noticing a befuddled customer, is expected to volunteer for what
may become a fairly complicated errand.
The SCS programme exhibits the dual devolution/control features of HIWPs
discussed above. On the one hand, employees are charged with departing from
their assigned tasks in order to make customers happy; this represents a
considerable broadening of job descriptions, and requires new judgements and
customer interactions. On the other hand, the procedures for making customers
happy under different contingencies are spelled out in some detail. In many
respects they are, in fact, scripted.
Scripting is particularly evident in the interface at checkout (the till). As part
of SCS, cashiers are required thank each customer by her/his surname at the
time of purchase. The cashier knows the customer’s name if the customer has
used a Safeway loyalty card (or having forgotten it, entered their telephone
number into the keypad at checkout), because the customer’s name is printed
on their receipt. Almost all customers use a loyalty card, because the savings
are substantial and the cards are issued without delay; on two occasions during
visits to the area in recent years, I have found myself in a Safeway store
without a loyalty card, obtained one on the way in, and was in the database to
be thanked by name when I paid for my groceries. Clerks do have some latitude
in how they deliver these thanks: I’ve heard “Thank you, Mr Guy”, “Have a nice
day, Mr Guy” and “Have a good one, Mr Guy”; my informants assure me that
all of these are in keeping with this genuinely Californian procedure. What is
required is that title and surname be used (though the first name may be used if
the employee knows the customer well enough.) In addition to this standard HIWPs and
procedure, further scripts are given with regard to special offers. bargaining
The employees I interviewed all agreed that Safeway customers appreciate power
SCS. “Customers love it”, “they eat the program up”, “[SCS] creates a more
down home feel” were among their comments. They were divided on the
question of whether SCS actually strengthened their relationships with
customers. All agreed, however, that comparing with before the SCS
463
programme, they now knew more than twice as many customers by name.
The employees also believed that the customers felt that they now had a
stronger relationship with Safeway employees.
Service offered in a Safeway store, and adherence to SCS guidelines, is
monitored through a system of mystery shoppers. According to the employees
I spoke with, each store receives one mystery shopper visit per week, after
which the store manager receives a report detailing the mystery shopper’s
experience, and giving the store a rating on a scale of 1 to 10. Employees expect
store managers are expected to arrange some sort of celebration when their
store receives a rating of 10.
Together with SCS, Safeway adopted a gainsharing programme, under
which most store-level employees (but not members of the night crew, who
restock stores when closed) are eligible for bonuses on the basis of their store’s
financial performance.
The Safeway employees I interviewed all felt that SCS had made their jobs
more demanding: the employee stocking the dairy case is on the one hand
expected to do that task expeditiously; on the other, the employee may be
reprimanded (or, if they respond to the criticism by saying they had not
understand what was required of them, the employee will be sent on a
re-training course called the Employee Improvement Program), and told they
have a “task orientation” if, in order not to be interrupted, they avoid eye
contact with a customer who is trying to find something. For this reason, the
attitude towards both mystery shoppers, and towards being held to account for
following SCS procedures, was an adversarial one. At the same time, they
believe that Safeway does offer far better customer service than its competitors
in the same niche (this accords with my experience in the stores of Safeway and
other national mid-market chains in both California and Arizona). They also
believe that better customer service is responsible both for Safeway’s strong
financial performance, and for their job security: at the time of these interviews,
there were no Safeway employees on layoff in UFCW local 890, while as of
April 2002 the list of Albertson’s employees awaiting recall went back to 1998.
In other words they accepted that, from the standpoint of the various
stakeholders in the Safeway organisation, that SCS offered a win-win solution.
Union officers told me that Albertsons had been attempting, for several
years, to implement a similar programme. One of Albertsons difficulties had
been getting employees to address by name customers who they didn’t know.
ER Discount competitors in the same market – Walmart, Costco, Trader Joe’s –
25,5 have not, according to UFCW sources (and verified by my experience as a
shopper) attempted to adopt any similar customer service programme. Three of
the Safeway employees interviewed – who, between them, had 79 years
experience in grocery retailing – had an exchange about customer service and
Walmart’s strategic options:
464
Grissom (former SCS trainer): If Walmart was smart, they’d do a service program
Poe: Walmart doesn’t have the quality of employee. [Klein concurs]
Grissom: They could do what Safeway does – do it or get out.
Poe: It’s easier to find a minimum wage job.
Grissom: So if I’m Walmart, don’t pay as much as Safeway, but pay as much as you can
afford, and improve service.
Klein: Walmart is about one thing, price.
Grissom: Implement a service program, raise wages, give service. Customers get great prices,
service, don’t come back to Safeway.

The strike of 1995

In the year following the implementation of SCS, the UFCW contracts with
Safeway and its competitors expired. Safeway was picked to set the pattern.
Contracts in this market have long been renegotiated at three-year intervals, and
the last strike against Safeway in Northern California had been in 1980. In 1995,
the UFCW and Safeway had trouble reaching an agreement. Two major points of
contention were Safeway’s proposals to eliminate both health insurance for
employee’s spouses and dependants, and also an hours guarantee without which
made a large number of part-time workers would no longer be eligible for group
health insurance. In the context of the US health care system, such a change
would have been equivalent to a substantial pay cut for many employees.
Safeway UFCW members struck, setting up picket lines around the stores.
To support Safeway’s bargaining position its biggest direct competitor,
Albertson’s, locked UFCW members out. Both chains continued to operate with
management employees and strike breakers.
The UFCW had an unprecedented degree of success shutting down Safeway
operations during the strike. Both union officials and employees who
participated in the strike attribute this to two factors: public sensitivity to the
health insurance issue at the time of the strike, and the fact employees on the
picket line were able to address more customers by name than they had in the
past. Jim Liggins, the UFCW local 870 officer responsible for Safeway at the
time of the strike (and, unusually for a union official, a former supermarket
manager himself) offers the following as a typical exchange:
Mrs Smith approaches the Safeway parking lot, and is greeted by a picket, Kay. Kay says,
“Good afternoon, Mrs. Smith. We’re in a dispute with Safeway. We’re just trying to save our
health plan. We’d appreciate it if you didn’t shop here today”.
“At that point”, says Liggins, “Most people have the basic decency to turn
around and drive away”.
 The SCS scripting had been turned on its head. The strike lasted nine days. HIWPs and
The hours guarantee and eligibility for spouses and dependants were kept, and bargaining
overall the union officers maintain that they got a better settlement than they power
had expected.

The aftermath of the strike 465

UFCW officers and the employees we interviewed believe that their success in
the strike was due in large part to SCS, because it enabled the pickets to address
a greater number of regular customers by name. The UFCW officers report
that, on the basis of informal conversations with various Safeway managers,
have concluded that this belief is shared by the top management of Safeway.
For this reason, the union expected Safeway to modify or abandon SCS.
Safeway not only kept the programme without modification, however, but has
since extended it beyond its California divisions. Safeway’s competitor and
strike ally Albertson’s is attempting to implement a similar programme. No
similar programme is evident, however, in Safeway and Albertson’s
competitors at the discount end of the market.
According to union officers, past Safeway practice had been to begin
bargaining for a new contract a few months before expiration of the old, and go
down to the wire (or past the wire, to a strike). When the contract negotiated in
1995 still had one year to run, however, Safeway proposed a three-year
extension, and reached a quick agreement with the UFCW. When the extended
contract was coming to an end in 2001 settlement was not so easy, and Safeway
took steps to convince UFCW members to vote against a strike authorisation,
showing a video that stressed the competitive threat from the largely non-union
discount chains, and giving supportive UFCW members rides to the voting
place. In the event, the strike vote took place shortly after 11 September 2001, in
which circumstances most of the UFCW locals voted against strike
authorisation, and the union was forced to make significant concessions in
the contract. Despite these very different outcomes, what 1998 and 2001 have in
common is that Safeway took steps to avoid a strike, as it had not done in 1995.
One reading of this is that the success of the SCS programme has, by
strengthening the relationship between employees and customers, made strikes
costlier for Safeway.

Discussion and conclusion

Under SCS, Safeway employees’ job descriptions are broadened; employees are
given responsibility for making certain decisions to further the objective of
making customers happy, and for building personal rapport with customers. In
other respects, SCS is limited as a HIWP: for instance, neither teams nor
employee involvement in decision making have been introduced, and employee
decisions are meant to follow a quite prescriptive set of rules.
ER While the intended empowerment apparent in the design of this system is
25,5 quite limited, the SCS programme appears to have strengthened employee
bargaining power, as evidenced in the outcome of the 1995 strike, and
Safeway’s subsequent efforts to avoid a strike. And, if we are to judge from
Safeway’s continuation and extension of SCS after the settlement, and
Albertson’s attempt to implement a similar programme, productivity increases
466 due to the programme were greater than increases in labour costs.
On the other hand, Safeway’s large discount competitors have not adopted
programmes involving the same sort of customer-employee interaction. This is
consistent with the strategic contingency view, namely that organisations
competing on price rather than differentiation or focus will not benefit from
HIWPs. What must be emphasised is that we lack the information necessary to
distinguish between two different explanations for this strategic choice. One is
that a programme such as SCS would not raise labour productivity in the large
discount stores. This might be so because customers in discount stores attach
less importance to relationships and more to price, or because repetitions of the
same customer-employee pairings are less frequent in discount stores, making
the relationships weaker. The other possibility is that an HIWP such as SCS
would increase productivity, but that the owners and managers of these stores
anticipate that increased employee bargaining power would ultimately
increase labour costs by more than the gain in productivity.
Where the second explanation is true, employers are sacrificing efficiency in
order to claim a larger share of value added for profits rather than wages. This
is consistent with the Marglin (1974)/Stone (1975) view of the development of
the factory system. It also squares with a growing body of research finding that
employee ownership can improve efficiency, especially when combined with
HIWPs: if the employees are the owners, then the conflict between profitability
and productivity disappears (see, for instance, Bowles et al., 1993; Kruse, 1992).
But is the second explanation true? We don’t have the information to say in
the present case. We should take it seriously, however, because it makes better
sense of the empirical results on HIWPs and performance. The second
explanation is consistent with findings of a universal positive HIWP effect on
productivity alongside non-universal adoption. The first explanation is not.
Moreover, if increased employee bargaining power due to HIWPs becomes
effective only after a lag, then the second explanation is consistent with
econometric findings of a universal positive HIWP effect on profitability
alongside non-universal adoption. In the case of Safeway Stores in Northern
California, there was a lag of a year; that was a case with a long established and
recognised union. Given the barriers to organising workplaces in the USA, the
lag before employees in an unorganised workplace would be able to exercise
their new, latent, bargaining power (whether through organising or through a
credible threat of organising) would probably be much longer. For this reason
an overall positive effect of HIWPs on profitability would be found in cross
sectional regressions. In principle, under this hypothesis, panel data should HIWPs and
make it possible to sort out those companies for which HIWPs provide long run bargaining
net improvements to productivity from those which do not; but long panels are power
scarce and, for reasons discussed above, short panels do not provide much (if
any) relief here. Nonetheless, the second explanation offered above – universal
productivity gains, lagged improvements in employee bargaining power, and
long run mixed results on profitability – is consistent with empirical findings
467
of universal HIWP benefits for employers, together with non-universal
adoption of HIWPs.

References
Akerlof, G.A. and Yellen, J.L. (1986), “Introduction”, in Akerlof, G.A. and Yellen, J.L. (Eds),
Efficiency Wage Models of the Labor Market, Cambridge University Press, Cambridge.
Appelbaum, E., Bailey, T., Berg, P. and Kalleberg, A.L. (2000), Manufacturing Advantage: Why
High-performance Work Systems Pay Off, ILR Press, Ithaca, NY.
Arthur, J.B. (1994), “Effects of human resources systems on manufacturing performance and
turnover”, Academy of Management Journal, Vol. 37, pp. 670-87.
Becker, B. and Gerhard, B. (1996), “The impact of HRM on organizational performance: progress
and prospects”, Academy of Management Journal, Vol. 39, pp. 779-801.
Bowles, S., Gintis, H. and Gustafsson, B. (Eds) (1993), Markets and Democracy: Participation,
Accountability and Efficiency, Cambridge University Press, Cambridge.
Chandler, A.D. Jr (1962), Strategy and Structure: Chapters in the History of the American
Industrial Enterprise, MIT Press, Cambridge, MA.
Delery, J.E. and Doty, D.H. (1996), “Modes of theorizing in strategic human resource
management: tests of universalistic, contingency, and configurational performance
predictions”, Academy of Management Journal, Vol. 39 No. 4, pp. 802-35.
Doeringer, P., EvansKlock, C. and Terkla, D. (1998), “Hybrids or hodgepodges? Workplace
practices of Japanese and domestic start-ups in the United States”, Industrial and Labor
Relations Review, Vol. 51 No. 2, pp. 171-86.
Drago, R. (1996), “Workplace transformation and the disposable workplace: employee
involvement in Australia”, Industrial Relations, Vol. 35 No. 4, pp. 526-43.
Edwards, P. and Wright, M. (2001), “High-involvement work systems and performance
outcomes: the strength of variable, contingent and context-bound relationships”,
International Journal of Human Resource Management, Vol. 12 No. 4, pp. 568-85.
Edwards, P., Collinson, M. and Rees, C. (1998), “The determinants of employee responses to total
quality management: six case studies”, Organization Studies, Vol. 19 No. 3, pp. 449-75.
Ezzamel, M. and Willmott, H. (1998), “Accounting for teamwork: a critical study of group-based
systems of organizational control”, Administrative Science Quarterly, Vol. 43 No. 2.
Fairris, D. (1997), Shopfloor Matters: Labor-Management Relations in Twentieth-Century
American Manufacturing, Routledge, London.
Frenkel, S.J., Tam, M., Korczynski, M. and Shire, K. (1998), “Beyond bureaucracy? Work
organization in call centres”, International Journal of Human Resource Management, Vol. 9
No. 6, pp. 957-79.
Griliches, Z. and Hausman, J.A. (1986), “Errors in variables in panel data”, Journal of
Econometrics, Vol. 31, p. 9318.
ER Grimshaw, D., Beynon, H., Rubery, J. and Ward, K.G. (2002), “The restructuring of career paths in
large service sector organisations: ‘delayering’, upskilling and polarisation”, Sociological
25,5 Review, Vol. 50 No. 1, pp. 89-116.
Guest, D. (1997), “Human resource management and performance: a review and research
agenda”, International Journal of Human Resource Management, Vol. 8 No. 3, pp. 263-76.
Guest, D.E. (1998), “Beyond HRM”, in Sparrow, P. and Marchington, M. (Eds), Human Resource
468 Management, Pitman, London.
Helper, S., Levine, D.I. and Bendoly, E. (2002), “Employee involvement and pay at U.S. and
Canadian auto suppliers”, Journal of Economics & Management Strategy, Vol. 11 No. 2,
pp. 329-77.
Huselid, M.A. (1995), “The impact of human resource management practices on turnover,
productivity, and corporate financial performance”, Academy of Management Journal,
Vol. 38 No. 3, pp. 635-72.
Huselid, M.A. and Becker, B.E. (1996), “Methodological issues in cross-sectional and panel
estimates of the human resource-firm performance link”, Industrial Relations, Vol. 35 No. 3,
pp. 400-22.
Ichniowski, C., Shaw, K. and Prennushi, G. (1997), “The effects of human resource management
practices on productivity: a study of steel finishing lines”, American Economic Review,
Vol. 87 No. 3, pp. 291-313.
Kinnie, N., Hutchinson, S. and Purcell, J. (2000), “‘Fun and surveillance’: the paradox of high
commitment management in call centres”, International Journal of Human Resource
Management, Vol. 11 No. 5, pp. 967-85.
Konzelmann, S. and Forrant, R. (2002), “Creative work systems in destructive markets”, in
Burchell, B., Deakin, S., Michie, J. and Rubery, J. (Eds), Systems of Production: Markets,
Organisations and Performance, Routledge, London.
Kruse, D.L. (1992), “Profit sharing and productivity: microeconomic evidence from the United
States”, Economic Journal, Vol. 102, pp. 24-36.
MacDuffie, J.P. (1995), “Human resource bundles and manufacturing performance”, Industrial
and Labor Relations Review, Vol. 48, pp. 197-221.
Marglin, S.A. (1974), “What do bosses do?”, Review of Radical Political Economics, Vol. 6 No. 2,
pp. 60-112.
Michie, J. and Quinn, M.S. (2002), “Labour ‘flexibility’ – securing management’s right to manage
badly?”, in Burchell, B., Deakin, S., Michie, J. and Rubery, J. (Eds), Systems of Production:
Markets, Organisations and Performance, Routledge, London.
Miles, R.E. and Snow, C.C. (1984), “Designing strategic human resources systems”,
Organizational Dynamics, Vol. 16, pp. 36-52.
Ng, I. and Dastmalchian, A. (1998), “Organizational flexibiity in Canada: a study of control and
safeguard rules”, International Journal of Human Resource Management, Vol. 9 No. 3,
pp. 445-56.
Osterman, P. (1994), “How common is workplace transformation and who adopts it?”, Industrial
and Labor Relations Review, Vol. 47 No. 2, pp. 173-88.
Pelletier, J.J. (2001), Bay Area Blues: The High Cost of Living for UFCW Members in the Bay Area,
UFCW, San Francisco, CA.
Pfeffer, J. (1998), “Seven practices of successful organizations”, California Management Review,
Vol. 40 No. 2, pp. 96-124.
Porter, M.E. (1985), Competitive Advantage: Creating and Sustaining Superior Performance,
Academic Press, New York, NY.
Sewell, G. (1998), “The discipline of teams: the control of team-based industrial work through HIWPs and
electronic and peer surveilance”, Administrative Science Quarterly, Vol. 43 No. 2.
Stone, K. (1975), “The origins of job structures in the steel industry”, in Edwards, R.C., Reich, M.
bargaining
and Gordon, D. (Eds), Labor Market Segmentation, D.C. Heath, Lexington, MA. power
Wilkinson, F. (1983), “Productive systems”, Cambridge Journal of Economics, Vol. 7 No. 3/4,
pp. 413-29.
Youndt, M.A., Snell, S.A., Dean, J.W. and Lepak, D.P. (1996), “Human resource management, 469
manufacturing strategy, and firm performance”, Academy of Management Journal, Vol. 39
No. 4, pp. 836-66.