ARC Centre of Excellence in Population Ageing

Research

Working Paper 2020/06

Transitioning from JobKeeper

Bruce Chapman and John Piggott

This paper can be downloaded without charge from the ARC Centre of
Excellence in Population Ageing Research Working Paper Series available at
www.cepar.edu.au
 Transitioning from JobKeeper*

Bruce Chapman1 and John Piggott2

1 Bruce Chapman
College of Business and Economics
Australian National University
Canberra, ACT, Australia
Bruce.chapman@anu.edu.au

2 John Piggott
ARC Centre of Excellence in Population Ageing Research (CEPAR)
University of New South Wales
Sydney 2052
Australia
j.piggott@unsw.edu.au

* This work was supported by the College of Business and Economics at the Australian National
University, and the ARC Centre of Excellence in Population Ageing Research (CEPAR) under ARC
grant CE17010005. We would like to thank Madeline Dunk for research support.

1
 Transitioning from JobKeeper
Bruce Chapman and John Piggott

Background

The JobKeeper scheme aims to provide financial assistance to those in danger of being laid
off, and to assist business to stay afloat until more familiar economic and employment activity
resumes. In financial terms, it is very significant: relative to population and GDP, it is the
largest wage-subsidy scheme in the world. A critical issue, not so far addressed in public
discussion, is how the initiative might be phased out when the economy recovers, in a way that
facilitates the survival of business while at the same time minimising further fiscal outlays, and
avoiding continuing to subsidise businesses that may not in normal times be viable.

This short paper addresses the question of how the JobKeeper program might be wound down
at the end of its full subsidy term.

Proposal

Specifically, we are proposing a specific application of the contingent debt model, whose
salient characteristic is that a repayment schedule is triggered only when the debtor can afford
the repayments. The best known policy structure of this kind is the Higher Education
Contribution Scheme (HECS), which originated in Australia in 1989 and has now been adopted
in various forms in 10 other countries. Under HECS, an example of an income-contingent loan
(ICL), students are provided with a benefit in the form of university tuition, which is paid for
if and when recipients can afford it. The great advantage of such borrowing arrangements is
that if future financial circumstances, which are unpredictable, turn out to be adverse there are
no, or only very low, repayments required.

A variant of this arrangement, customised to business, is a revenue-contingent loan (RCL),

which operates in a similar way to an ICL, but is contingent on revenue reaching some
threshold level. This application has been explored over a long period by Botterill and
Chapman (2006) and Chapman and Lindenmayer (2019), and entered the Australian
contemporary economic crisis debate through Botterill, Chapman, McKibbin and Withers
(2020). We note we are not alone in advocating a critical role for government organized
contingent debt for business in the current economic malaise (Bonardi, Brülhart, Danthine,

2
Jondeau, Rohner, 2020). It is particularly suited as a means of transitioning out of JobKeeper
because that policy relies on revenue as a criterion for eligibility.

JobKeeper is specifically directed towards the maintenance of a business infrastructure, and is

widely anticipated to be effective in that objective. But much of the good that is achieved might
be undone if it is suddenly withdrawn with no alternative mechanism in place for paying
employees. It is likely that firms facing financial adversity at that time will have difficulties
procuring traditional commercial lending assistance, with private lenders having limited
precedent for risk assessment in these circumstances.

One way to phase out JobKeeper is through the provision of RCLs to help smooth the
transition. There is an infinite range of settings around this structure. For example, if the
government cut the JobKeeper allowance to $500 per fortnight, RCLs could be offered to
qualifying firms (that is, firms now accepting JobKeeper payments) at a rate of $1000 a
fortnight per employee for the ensuing three months, meaning that the short-run financial
assistance to companies would be unchanged; this would provide more time for demand to pick
up and economic normalcy to be at least in sight. But the extra burden on future generations of
the wage subsidy program would be only one third of what it would be were JobKeeper to be
maintained in its current form.

As well, with the loan being made available, firms would be making borrowing decisions in
light of their own financial, employment and market circumstances. It will be difficult to design
a direct subsidy policy which accommodates a wind-down of the scheme, while at the same
time addressing the heterogeneity of circumstances that businesses will face over this period.
Even within particular sectors firms will find themselves confronting a wide range of financial,
demand and employment situations.

For many firms, without a buffer of this type, the withdrawal of JobKeeper could be very harsh,
and might mean in the aggregate increased job shedding, further demand reductions, and
heightened uncertainty at a time when insecurity is already at a historic high.

While this specific policy has not so far been adopted, in the unprecedented health and
economic crisis now being experienced the suggestion of a Covid19 RCL would be expected
to have more appeal for decision-makers.

3
Why revenue?

We begin with the question, why revenue? A RCL provides the same sort of insurance and
income-smoothing properties as HECS, but if income-contingency was the basis of collection
of a loan for business the system would not work. Profits might seem to be a good basis of
collection, but these are open to manipulation to facilitate extensive repayment avoidance.

This is why the suggested loan collection basis would use quarterly statements of firm revenue,
which are legally required through each enterprise’s Business Activity Statement (used among
other things to collect the GST). A critical point is that business revenue is not subject to any
form of deductions or other potential avoidance behaviours, and it is this simplicity that makes
revenue ideal as the collection contingency (Botterill, Chapman and Egan 2006).

How would this work?

The following issues need to be addressed:

(i) The government would have to decide on the level of withdrawal of the $1500 per
fortnight per employee direct payment. It is unlikely that this would be done in one
step; rather, a progressive withdrawal would need to be specified. The precise
parameters of the withdrawal schedule would be determined in light of prevailing
macroeconomic circumstances and outlook, but we have chosen not to model the
gradual transition (which would be straightforward, as would be a plethora of other
possibilities).

(ii) The limits to the loan need to be specified. These could be determined by the
revenue required to maintain initially a $1500 a fortnight support per employee,
taking account of the remaining direct government support. For example, in a period
where JobKeeper is still paying $500 per employee per fortnight, the loan limit
would be set at $1000 per employee per fortnight, and this is what we have modelled
and report below. Of course, a plethora of alternative choices are open to the
government, including a complete withdrawal of all direct wage subsidies.

(iii) The rate of collection of repayments from revenue need to be specified. In the
examples provided in Botterill and Chapman (2017) and Chapman and

4
 Lindenmayer (2019) the proportion of annual revenue used to repay the debt varies
between 2 and 8 per cent, and we have chosen the mid-point of 5 per cent.

(iv) A repayment holiday could be part of the policy. For example, there could be an
initial payment postponement until the beginning of the 2021/22 financial year (July
2021), which is what we model.

(v) A rate of interest on the debt needs to be specified. In what follows and for
simplicity we have set this at zero in real terms, which is the arrangement for
HECS. We did some sensitivity analysis in our exercises which shows that this
assumption is unimportant, essentially because the times taken for full RCL
repayments are so short

(vi) The population of businesses covered needs to be clear. We have confined the
JobKeeper RCL transition to firms currently eligible for the first wave of the wage
subsidy.

Modelling and presentational assumptions

First, because the debt level is incurred on the basis of $1000 per employee per fortnight, but
collected on the basis of the firm’s future annual revenue, for the empirical analysis an
assumption is needed concerning the relationship between the firm’s annual revenue and
JobKeeper employment. We assume that the firm’s wage bill is 60 per cent of revenue which,
with the further assumption that employees receive a wage of $50,000 pa, allows us to calibrate
the number of employees covered by JobKeeper on the basis of firms’ assumed annual revenue
streams.

Second, myriad possible cases could be presented and we have had to choose between a
complex reality and the need for presentational parsimony. In our trade-off there are six
categories of debt, each associated with annual revenue estimates. The revenue categories,
assumed number of employees, RCL debt levels, and annual revenue estimates are shown in
Table 1.

5
While the mid-points of the annual revenue categories can be used to estimate future scenarios
very simply, it is more likely that there will be relatively low revenue in the first year of
repayment with a resumption of more typical revenue experiences beyond this. Accordingly,
we assume that in the first and second calendar years of repayment of the RCL annual revenues
are respectively 50 and 75 per cent of the mid-point of the annual revenue categories, after
which firms experience annual revenues of the mid-point of their category. No repayments are
required until June 2021.

Table 1
RCL Debt and Revenue Categories

Min and max annual Assumed annual Assumed number of Assumed RCL debt
revenue category revenue (midpoint) JobKeeper per firm ($000s):
($000s) ($000s) employees per firm:

75 – 225 150 2 13

225 – 1,000 612.5 7 45.5

1,000 – 10,000 5,500 66 429

10,000 – 100,000 55,000 660 4,290

100,000 – 1,000,000 550,000 6,600 42,900

1,000,000+ 2,000,000 24,000 156,000

The aggregate results and some illustrative examples

On the basis of our assumptions and loan collection parameters, we are able to provide some
illustrations. Figure 1 shows the RCL cumulative loan amounts still owing in the next few
calendar years and Figure 2 shows the proportions of the total RCL debt repaid for each of the
six categories.

6
 Figure 1
RCL amounts still owing over time, different firm revenue categories

What is striking about the data from Figures 1 and 2 is that for the RCLs described and
modelled, and on the basis of quite different loan levels and assumed revenue streams, the RCL
for all six of these different situations are repaid in full by 2024.

These results imply strongly that a sensibly-based RCL design, in combination with a reduction
of the grants-based component of JobKeeper to just less than 17 per cent of the original policy
(the level of direct wage subsidy assistance is reduced by two thirds, and the original duration
of six months is cut in half), is able to deliver the same level of initial financial support to firms
in a three-month transition policy. This analysis indicates economic transition from JobKeeper
can be achieved without substantial financial disruption to business.

7
 Figure 2
Cumulative proportions of the RCL repaid over time, different firm revenue categories

To provide further insight into what this might mean for firms in specific circumstances, we
provide three hypothetical case studies, based on our understanding of the types of situation
likely to be faced. These hypothetical case studies are real world illustrative examples of what
might happen with respect to RCL repayments for the policy rules and scenarios considered.
They are complements to the aggregate modelling, which essentially illustrates that the RCL
system considered is likely to be viable for a large range of business experiences. We reiterate
that the case studies and the broader modelling are small subsets of the plethora of potential
scenarios; clearly, myriad alternative specifications are possible.

8
 Case Study 1: Francesco’s Hairdressing and Beauty Salon, Cosmopolitan

Cosmopolitan was established in 2001 in a Melbourne suburb, initially with the equivalent of
5 full-time staff. The business expanded consistently and by 2020 was receiving about $1
million per annum in revenue with 10 equivalent full-time staff. The impact of the virus in
February/March 2020 was very significant, and while the proprietor, Francesco, chose to
remain open for significantly diminished business, the enterprise suffered around a 70 per cent
fall in revenue. The business took advantage of the JobKeeper scheme which covered the
wages of 10 employees, although Cosmopolitan still struggled to meet other major costs, the
main one being rent.

At the conclusion of the JobKeeper period (the end of October 2020), business had picked up
but was still only about half what it had been. Many clients had not returned and because there
was such low turnover for most of 2020 almost no new clients had been found to replace the
normal levels of customer attrition.

Francesco chose to take advantage of the RCL offered to him, which was the scenario modelled
and reported above, of $1000 per fortnight per employee for the 3-month period, from the
beginning of November 2020 to the end of January 2021. Determining the total RCL borrowing
needs information on the loan amount per person per fortnight, the number of employees
covered, and the number of fortnights involved. For the suggested RCL policy in the context
of Cosmopolitan’s borrowing needs when the current JobKeeper stops, this amounts to a total
RCL repayment obligation of $(1000*10*6.5) = $65,000.

In terms of repayment of the debt, this of course depends on Cosmopolitan’s annual revenue,
which had historically been about $1,000,000. Given the slowdown in its business,
Cosmopolitan’s first quarterly Business Activity Statement in 2021 reports revenue on an
annual basis of only $125,000. In terms of repayment of the RCL, in the 2021 calendar year
the business is obligated to repay on the basis of post-June revenue only, so that the RCL
repayment is $125,000*.05*.5 = $5,000 (meaning that $60,000 of the RCL is still owed).
Cosmopolitan’s annual revenue then recovers significantly to reach $750,000 in the 2022
calendar year, which results is a RCL repayment in that year of $750,000*.05 = $37,500. There
thus remains an outstanding RCL obligation of $27,500 which is easily repaid at the beginning
of 2023 because Cosmopolitan’s revenues are then back to the historical norm.

9
 Case Study 2: Everytime Fitness Club

The Everytime Fitness Club Australia is one of the 3 largest fitness enterprises in the country
and before Covid19 had a typical annual revenue of around $350 million. As a result of the
necessary imposition of restrictive health regulations the firm effectively shut down in March
2020 and is not expected to re-open for several months. The firm has around 1250 employees
all of whom qualified for JobKeeper.

By November 2020 the fitness industry is allowed to operate again, but through a change in
exercise lifestyles, demand for its services is initially well below the historical norm. However,
to maintain staff and to prepare for a brighter future Everytime chooses to borrow through the
RCL and takes a debt to cover the costs of 900 of its employees, receiving the $1000 per
employee per fortnight for the 3-month period. This amounts to a total RCL borrowing of $5.85
million.

However, because of the continuing restrictions to business activity in the fitness sector from
Covid19, Everytime’s revenue in calendar year 2021 is only $60 million, implying a RCL
repayment in 2021 of $(60*.05*.5) million = $1.5 million, leaving a remaining repayment
obligation of $4.35 million. Demand grows at a strong rate and in calendar year 2022 revenue
reaches $250 million, meaning the RCL is fully repaid in that year.

Case Study 3: Dyers Department Store

Dyers department store is Australia’s second biggest chain of this type, with around 26,000
employees and pre-Covid19 annual revenue of around $1.5 billion. In March 2020 the company
experiences a huge fall in demand and is forced to close temporarily, accepting JobKeeper
payments for 20,000 of its laid off workers. At the end of the JobKeeper period Dyers chooses
to take the RCL for all of these workers and incurs a debt of $130 million.

By the end of the JobKeeper period, Dyers demand has had only a small recovery and, in part
as a result of the significant move to on-line shopping, annual revenue for calendar year 2021
is $500 million, implying RCL repayments of $500*0.5*0.05 = $12.5 million, which means
that $117.5 million is still owed.

Revenue recovers to reach $1 billion in calendar year 2022, and this is the annual level it
remains at for several years. With this annual revenue RCL repayments in calendar year 2022
of $(1*0.05) billion = $50 million, meaning that at the end of 2022 $67.5 million is still owed

10
by Dyers. With the on-going maintenance of Dyer’s revenue at $1 billion a year, a further $50
million is repaid in calendar year 2023, and the debt is completely repaid during 2024.

Conclusion

This short paper has outlined how a financially very large government support package, the
JobKeeper scheme, might be phased down as the economy re-opens and recovery takes shape.
Our suggested transition package facilitates continued financial support and stability for firms
through the provision of government-controlled RCLs. Repayment mechanisms have been
tested with appropriate professionals, and we note that there is support for such an approach
from overseas researchers.

In both concept and with well-administered practice, contingent debt of the form we have
examined for business has substantial potential to ease the Australian economy back into
normalcy over time, and without the financial shocks and uncertainties implicit in a sudden
withdrawal of support when JobKeeper formally ends. Very importantly, the transition
approach suggested minimises the very concerning future requirements of ever-increasing
public sector debt.

We stress that there is a plethora of modelling scenarios that could be used; our assumptions,
modelling and case studies are offered to illustrate the potential and not the details of a
contingent debt approach to the transition process from JobKeeper. The analysis can be seen
to be a beginning for the policy exploration of the role of RCL.

This analysis has confined itself to the phase-down of the JobKeeper scheme. Broader
application of the RCL through the recovery phase may well be appropriate. For example,
while many enterprises are facing current dramatic revenue shortfalls, others are still in receipt
of strong revenues, but anticipate a fall-off in demand in the next several months. This is true,
for example, in the construction industry, where current projects are being completed, but new
initiatives are being put on hold. RCLs may well be a useful policy tool in these circumstances.

It is possible, of course, that the Australian economy and business recovers quickly in the next
few months and if this regeneration is sufficiently healthy there may not need to be a transition
policy in place. As citizens we hope that this happens, but as economists we believe that having
available an operational and sensible plan is a balanced and useful way to think about economic
policy.

11
References

Jean-Philippe Bonardi, Marius Brülhart, Jean-Pierre Danthine, Eric Jondeau, Dominic

Rohner (2020), The economics of wage compensation and corona loans: Why and how
the state should bear most of the economic cost of the COVID lockdown, VOX Policy
Posts, 06 April.

Linda Botterill, Bruce Chapman and Michael Egan (2006), ‘Income Contingent Loans for
Drought Relief’ (2006), Farm Policy Journal, Vol. 3, No. 2: 59-67.

Linda Botterill, Bruce Chapman and Simon Kelly (2017), ‘Revisiting Revenue Contingent
Loans for Drought Relief: Government as Risk Manager’ (2017), , Australian Journal
of Agriculture and Resource Economics, Vol. 61(3): 367-384.

Linda Botterill, Bruce Chapman, Warwick McKibbin and Glenn Withers (2020), “Give people
and businesses money now they can pay back later (if and when they can)”, the
Conversation March 30.

Bruce Chapman and David Lindenmayer (2019), ‘A novel approach to the sustainable
financing of the global restoration of degraded agricultural land’ (2019), Environmental
Research Letters, Vol. 14.

12