Should We Insure Workers or Jobs during Recessions? (Journal of Economic Perspectives, 2022)
Summary
Journal of Economic Perspectives article, Volume 36, Number 2, Spring 2022, pages 29-54, Should We Insure Workers or Jobs During Recessions?, by Giulia Giupponi, Camille Landais, and Alice Lapeyre. The article states that after the COVID crisis the US response focused on insuring workers' income through more generous unemployment insurance, while Europe emphasized preserving the relationship between workers and firms through short-time work. It states that in the United States the fraction of the working-age population on unemployment insurance benefits surged from about 2 to 12 percent in April 2020, while more than 16 percent of the working-age population in the weighted European series was enrolled in short-time work. It explains how short-time work and unemployment insurance differ.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
- Issuer
- Source documents
- Document type
- Should We Insure Workers Or Jobs During Recessions
- Case
- Should We Insure Workers Or Jobs During Recessions
Cite this document
“Should We Insure Workers or Jobs during Recessions? (Journal of Economic Perspectives, 2022).”
Original at researchonline.lse.ac.uk. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/should-we-insure-workers-or-jobs-during-recessions/
, SHA-256 2b0914e8e4113fd34c2c8b03bded838fd50acc1213285048ed83739ed8534916.Full text
Giupponi, Giulia, Landais, Camille & Lapeyre, Alice (2022) Should we insure
workers or jobs during recessions? Journal of Economic Perspectives, 36(2),
29-54. https://doi.org/10.1257/jep.36.2.29
https://researchonline.lse.ac.uk/id/eprint/115371/
Version: Published Version
LSE Research Online is the repository for research produced by the London School
of Economics and Political Science. For more information, please refer to our Policies
page or contact lseresearchonline@lse.ac.uk
Journal of Economic Perspectives—Volume 36, Number 2—Spring 2022—Pages 29–54
Should We Insure Workers or Jobs
During Recessions?
Giulia Giupponi, Camille Landais, and
Alice Lapeyre
I
n the wake of the COVID crisis, labor market policy responses on both sides of
the Atlantic have been immediate, absolutely unprecedented in scope—and
also diametrically opposed in nature. To put it simply, the focus of the US labor
market policy response was on insuring the income of workers against the cost of
job losses. This was done by aggressively increasing the generosity of unemployment
insurance. In Europe, the emphasis was on preserving the relationship between
workers and firms, which translated into generous subsidies for hours reductions
and temporary layoffs through short-time work or related schemes.
Panel A of Figure 1 gives a visual representation of these polar strategies. In
the United States, the fraction of the working-age population on unemployment
insurance benefits surged from about 2 to 12 percent in April 2020, and, although
it declined very quickly after that, at the end of 2020 it was still higher than at the
peak of the Great Recession of 2007–09. If we look at the weighted sum of the
four largest European economies—Germany, the United Kingdom, France, and
Italy—the increase in the number of unemployment insurance recipients was very
limited, but take-up of short-time work skyrocketed, with more than 16 percent of
the working-age population enrolled in this type of scheme in April 2020. There was
no such increase in short-time work take-up in the US economy, although about 25
US states have operational work-sharing schemes similar to short-time work.
■ Giulia Giupponi is Assistant Professor of Economics, Bocconi University, Milan, Italy.
Camille Landais is Professor of Economics, London School of Economics, London, United
Kingdom. Alice Lapeyre is a PhD candidate in Economics, CREST-ENSAE (Center for
Research in Economics and Statistics, École Nationale de la Statistique et de l’Administration
Économique), Paris, France. Their email addresses are giulia.giupponi@unibocconi.it,
c.landais@lse.ac.uk, alice.lapeyre@ensae.fr.
For supplementary materials such as appendices, datasets, and author disclosure statements, see the
article page at https://doi.org/10.1257/jep.36.2.29.
30 Journal of Economic Perspectives
Figure 1
Labor Market Policy Responses to Recessions and Non-Employment Rates in the
United States and Europe
A: Unemployment insurance and short-time work take-up
20
STW/UI take-up (percent working-age pop.)
Europe STW United States STW
Europe UI United States UI
16
12
8
4
0
n 2
Ja 00
n 5
2
Ja 00
n 6
2
Ja 00
n 7
2
Ja 00
n 8
Ja 020
n 9
2
Ja 01
n 0
Ja 201
n 1
2
Ja 01
n 2
2
Ja 01
n 3
2
Ja 01
n 4
2
Ja 01
n 5
2
Ja 01
n 6
2
Ja 01
n 7
2
Ja 01
n 8
2
Ja Ja 01
n 9
Ja 220
n 0
2
Ja 02
n 1
20 22
B: Non-employment rate
45
Europe United States
41
Non-employment rate (percent)
37
33
29
25
n 2
Ja 00
n 5
2
Ja 00
n 6
Ja 020
n 7
2
Ja 00
n 8
2
Ja 00
n 9
2
Ja 01
n 0
Ja 201
n 1
2
Ja 01
n 2
2
Ja 01
n 3
Ja 120
n 4
2
Ja 01
n 5
2
Ja 01
n 6
2
Ja 01
n 7
Ja 120
n 8
2
Ja Ja 01
n 9
2
Ja 02
n 0
Ja 220
n 1
20 22
Note: Panel A reports the evolution of short-time work (dashed lines) and unemployment insurance
(solid lines) take-up in Europe (red lines) and the United States (blue lines), each computed as the ratio
of the number of individuals in the program in a given month, as a percent of the quarterly working age
population. The series for Europe is a weighted average of the series for Germany, France, Italy, and the
United Kingdom, weighted by the working-age population. Panel B reports the evolution of the non-
employment rate—that is, one minus the employment rate (and thus including both the unemployed
and those out of the labor force). In both panels, the plotted series are moving averages of the raw series
over the period up to June 2021. The moving average is based on twelve lagged terms, one forward
term, and uniform weights. Data on employment come from OECD. Data on short-time work and
unemployment insurance take-up come from the OECD and national statistics. See Online Appendix C
for details on data sources and the construction of short-time work/unemployment insurance take-up.
Giulia Giupponi, Camille Landais, and Alice Lapeyre 31
Some consequences of these opposite labor market strategies on non-
employment rates are laid bare in Panel B. While the US economy experienced
a spike in non-employment, and continued to see high rates of non-employment
in late 2020, employment rates did not bulge in Europe despite the severity of the
shock. Interestingly, the much larger cyclicality of the US labor market relative to that
of European countries was already visible in past recessions, during which Europe
already experimented, although to a much lower degree, with short-time work usage.
Do US policymakers get it right by focusing their labor market policy response to
recessions on insuring workers through unemployment insurance? Or should they
use more short-time work and focus more on preserving jobs, as in Europe?
Addressing these questions is complicated by the remarkably small attention
devoted to short-time work relative to the sprawling literature on unemployment
insurance—an imbalance we hope to remedy.
Short-Time Work and Unemployment Insurance
Some Institutional Features
While most people are familiar with unemployment insurance policies, short-
time work schemes are not as well-known. How do they work in practice? What
countries use them and how long have they been in place? How do they compare
with unemployment insurance in terms of generosity, coverage, or eligibility? Let us
start by clarifying a bit the institutional background.
Short-time work—also known as short-time compensation, work sharing, or
shared-work programs—is a subsidy for temporary reductions in the number of hours
worked in firms experiencing temporary drops in demand or production. Short-time
work programs allow employers facing temporary shocks to reduce their employees’
hours instead of laying them off. The program provides a subsidy to employees put
on reduced hours (that is, put on short-time work) equivalent to a fraction of their
lost earnings. Short-time work cushions the adverse effect of reductions in business
activity on both firms and workers, averting the risk of layoffs and insuring workers
against the cost of drops in hours worked. Unemployment insurance programs,
instead, provide a temporary subsidy to laid-off workers who lost their job through
no fault of their own. Hence, while both programs provide insurance against labor
market shocks, short-time work programs insure job matches against temporary
hours reductions, while unemployment insurance insures workers against job loss.
We now provide a general description of the features of both programs.
Several European countries and US states have short-time work schemes in place.1
Whilst different schemes are characterized by different requirements and generosity,
they all share a common feature in that the application process must be initiated by
1
At the time of writing, 25 states are operating active programs in the United States: Arizona, Arkansas,
California, Colorado, Connecticut, Florida, Iowa, Kansas, Maine, Maryland, Massachusetts, Michigan,
Minnesota, Missouri, Nebraska, New Hampshire, New Jersey, New York, Ohio, Oregon, Pennsylvania,
Rhode Island, Texas, Washington, and Wisconsin. Vermont ceased to operate its short-time work
program on July 1, 2020.
32 Journal of Economic Perspectives
the employer, who must submit an application to the relevant administrative agency.
If successful, the application grants a number of subsidized hours of short-time work
that can be used by the firm. Workers’ subsidies are computed as a percentage of the
earnings lost due to hours not worked, typically up to a cap. The disbursement of the
subsidy is usually advanced to the worker by the firm, who is subsequently reimbursed
by a social insurance administration via lower contribution payments. In exceptional
cases, the social insurance agency pays the subsidy directly to the worker. Short-time
work schemes are usually funded through a combination of social insurance contri-
butions—paid by eligible firms and workers—and an experience rating component,
paid only by those firms and workers that benefit from the program.
Short-time work regulations can include work-sharing requirements, specifying a
minimum and/or maximum hours reduction, a minimum number of employees
or share of the workforce to be involved in the program, and how hours reductions
should be distributed across the workforce. In the majority of countries, there is
no maximum hours reduction per worker, meaning that short-time work can cover
both partial and full hours reductions (that is, down to zero hours worked). In the
United States, instead, work weeks must be reduced by at least 10 percent and no
more than 60 percent to benefit from the program.
Eligibility requirements set the conditions under which employers or employees
can take part in the scheme. For employers, the main requirement is proof of
economic need, such as a reduction in business activity above a certain threshold.
In some countries, access to the program is restricted to firms operating in certain
sectors of the economy or with size above given thresholds. An agreement with
union representatives can also be required. To be eligible for short-time work,
workers are typically required to have been employed and contributing to social
insurance for a minimum amount of time. Some short-time work schemes do not
cover workers on temporary contracts.
Conditionality requirements for employers can include the prohibition of
dismissals and the development of a recovery plan, while for workers, either training
or job search requirements could be in place.2
Unemployment insurance provides temporary financial assistance to eligible
unemployed workers who become unemployed through no fault of their own. To
receive unemployment insurance, laid-off workers must file a claim with the relevant
administrative agency. To be eligible, a worker is typically required to have worked
for a minimum amount of time, and in some cases to have earnings above a certain
threshold. Workers on fixed-term contracts are usually not covered by unemployment
insurance once their contract expires. If eligible, the unemployed worker receives an
unemployment benefit in cash for a given amount of time. Benefits are defined as a
fraction of recent earnings, usually up to a cap, with a replacement rate that is in most
cases lower than that of short-time work at zero hours (and the more so at partial
hours). The disbursement of unemployment benefits is made directly to the worker
by the social insurance agency, typically ensuring swift payments. To retain their
2
For a comprehensive illustration of short-time work schemes in OECD countries, see Hijzen and Venn
(2011).
Should We Insure Workers or Jobs During Recessions? 33
benefits, unemployed workers may need to fulfill job-search and availability-to-work
requirements—features aimed at ensuring that recipients do not become inactive.
Whilst in their purest form, as we have described them so far, short-time work
and unemployment insurance are polar schemes insuring jobs on the one hand
and workers on the other, in practice there exists a continuum of more nuanced
labor market policies between those two extremes. Taking the United States
as an example, workers can qualify for unemployment insurance both when on
permanent and temporary layoffs. Temporary unemployment—also called “recall”
unemployment—is a situation in which an employer lays off a worker or a group of
workers but plans to rehire them by a given date. In this circumstance, the worker is
eligible for unemployment benefits and job search requirements are usually waived.
Employers also have the option to put full-time employees who work less than full
time during a pay period due to lack of work on partial unemployment insurance.
In this case, workers are eligible for partial unemployment benefits, provided that
they do not earn more than a maximum amount of labor income per week. Apart
from this, eligibility conditions for partial unemployment insurance are usually
identical to those for full unemployment insurance.
We can therefore think of there being a spectrum of policies offering different
types of flexibility to employers: from short-time work for partial hours reductions
offering only intensive margin flexibility, to short-time work at zero hours and partial
unemployment offering both intensive and extensive margin flexibility, to recall
unemployment offering only flexibility at the extensive margin. Besides differences
in flexibility, the programs also insure different shocks and, consequently, are char-
acterized by different types of moral hazard responses. On the one hand, programs
that ensure intensive-margin adjustments tend to insure job matches: they trigger
moral hazard responses on the employer side, because employers may have incen-
tives to rely excessively on subsidized hours reductions. This possibility is especially
relevant in contexts—like European countries—in which short-time work is only
mildly experience-rated. On the other hand, unemployment insurance schemes
insure workers rather than jobs and trigger moral hazard responses in the form of
lower job search effort exerted by the unemployed.
It is worth noting that, whilst at first sight short-time work and partial or recall
unemployment might seem to serve the same insurance needs, they are funda-
mentally different in how they affect employers’ commitment to retain workers.
The possibility to make take-up conditional on the prohibition of dismissal is a key
advantage of short-time work programs compared to partial or recall unemploy-
ment. For these, it is hard to envisage a mechanism whereby firms are held to their
commitment to retain or recall workers, since such a commitment can neither be
monitored nor enforced.
Short-Time Work and Unemployment Insurance During the COVID Crisis
At the onset of the COVID crisis, the United States responded to the sudden labor
market freeze and historical surge in layoffs by aggressively extending the generosity
of unemployment insurance. In particular, the Coronavirus Aid, Relief, and Economic
Security (CARES) Act signed into law in March 2020 granted i) additional payments
34 Journal of Economic Perspectives
to everyone who qualified for unemployment benefits; ii) an extension to individuals
who would have otherwise exhausted their benefits; and iii) eligibility to self-employed
and gig workers. Specifically, the CARES Act authorized Federal Pandemic Unem-
ployment Compensation, in which unemployment benefits were increased by $600
a week from March to July 2020. Moreover, the CARES Act was complemented by
two additional stimulus packages in 2021—the Consolidated Appropriations Act
(January) and the American Rescue Plan (March)—both of which extended the
unemployment insurance measures put in place by the CARES Act.
European countries, on the contrary, responded through short-time work
or related schemes. In April 2020, the European Union announced that it would
provide financial assistance for up to €100 billion to EU countries to develop or
expand short-time work schemes.3 The majority of OECD countries had a short-
time work program in place prior to the COVID crisis: Activité Partielle in France,
Kurzarbeit in Germany, and Cassa Integrazione Guadagni in Italy. Several of those
who did not have a scheme in place introduced it (as in Hungary and the United
Kingdom), and most of those with existing short-time work schemes implemented
measures to ensure rapid access to and wide take-up of the program. Combinations
of such measures have been necessary for short-time work schemes to work swiftly
and effectively (Giupponi and Landais 2020a).4
Insurance versus Moral Hazard
Let us imagine that, in the midst of recession, we decide to increase the
generosity of social insurance by $1. Should this $1 be put into more generous
unemployment insurance or into more generous short-time work?
To approach the choice between unemployment insurance and short-time
work, our starting point is the standard trade-off that both policies have to solve
between providing insurance to workers against labor market shocks and distorting
the behaviors of firms and workers (Baily 1978; Chetty 2008). Indeed, the goal of
both policies is to provide insurance against labor market shocks. For unemploy-
ment insurance, the shock is the cost of being unemployed. For short-time work, it
is the cost of having to reduce working hours when a firm is hit by a negative shock.
Providing such insurance is likely to be socially desirable, as it transfers money
to individuals who have lower income and consumption. But transferring $1 to
these individuals will cost more than $1 because these policies also tend to distort
behaviors, a problem often called moral hazard. Individuals will search less actively
for a new job when they have unemployment insurance and will cash benefits for
longer as a result. Firms may reduce hours more than otherwise necessary if their
3
For more details on the Support to mitigate Unemployment Risks in an Emergency (SURE) program,
see https://ec.europa.eu/info/business-economy-euro/economic-and-fiscal-policy-coordination/
financial-assistance-eu/funding-mechanisms-and-facilities/sure_en.
4
Online Appendix A provides an overview of measures that have been put in place during the COVID
crisis to facilitate access to short-time work. See also Scarpetta et al. (2020) for an overview of short-time
work schemes during COVID in OECD countries.
Giulia Giupponi, Camille Landais, and Alice Lapeyre 35
workers receive short-time work benefits. From the perspective of the government,
these changes in behavior increase the cost of providing such insurance benefits.
We say that moral hazard creates a fiscal externality on the government.
What do we know about the respective moral hazard costs of unemployment
insurance and short-time work schemes? And which program is better at providing
insurance?
Willingness to Pay for Unemployment Insurance versus Short-Time Work Benefits
The value of any form of social insurance against labor market shocks depends
on how workers value insurance (that is, the extent of their risk aversion) as well on
whether workers have access to alternative means of consumption smoothing (self-
insurance). Estimating this value poses challenges. Given that both unemployment
insurance and short-time work are mandated by the government, one cannot simply
look at a market price to measure workers’ willingness-to-pay for insurance against
labor market shocks.
In general, the research literature has devoted relatively little attention to this
problem. The literature has mostly focused on measuring consumption dynamics
around labor market shocks like job loss (for example, Gruber 1997). This approach
usually finds significant but small consumption responses, which in turn translate
into a moderate value of social insurance. Recent research using alternative revealed-
preference methods (for example, Hendren 2017; Landais and Spinnewijn 2021)
suggests instead that the value of insurance against unemployment shocks is much
larger than previously thought, but is also strongly heterogeneous across individuals.
For present purposes, the key question is how the insurance value of short-time
work compares to that of unemployment insurance. The evidence on this point is
limited. However, two elements indicate that the value of unemployment insurance
may be somewhat larger than the value of short-time work.
First, recipients of short-time work and recipients of unemployment insurance
are often quite distinct populations. For example, Germany is a country where both
generous unemployment insurance and short-time work are available. We exploit
newly collected data starting in May 2020 from the High-frequency Online Personal
Panel Survey (HOPP), a longitudinal survey launched by the German Institute for
Employment Research (Haas et al. 2021). It shows unambiguously that during the
COVID crisis, short-time work tended to protect mostly insiders, individuals with
higher incomes and better self-insurance options. Unemployment insurance, to the
contrary, was mostly protecting the outsiders of the labor market, like younger indi-
viduals at the beginning of their career, or individuals with lower education and
fewer means to smooth household consumption (such as the presence of a working
partner).5 As noted by Cahuc and Carcillo (2011), because short-time work tends
to protect insiders, it is perhaps not surprising that it tends to be more prevalent in
countries with strong employment protection regulations.
Second, the value of insurance is a direct function not only of the availability
of self-insurance options, but also of the size of the consumption (or income) shock
5
For these results and additional details on the HOPP survey, see online Appendix B.
36 Journal of Economic Perspectives
experienced upon transitioning to the program. As shown in Figure 2, short-time
work clearly insures smaller shocks. The figure builds on administrative data from
Germany (Tilly and Niedermayer 2016) and Italy during the Great Recession of
2007–09 (Giupponi and Landais 2020b), and compares, using an event study design,
the evolution of total earnings plus transfers around the onset of an unemployment
spell and a spell of short-time work. In both panels, we see that the drop in earn-
ings and transfers is much more severe and persistent for the unemployed than for
workers on short-time work. But we also see an interesting difference. In Germany,
the earnings of workers who experienced a short-time work spell had fully recov-
ered after three years (Panel A). In Italy, to the contrary, they were still 30 percent
lower than the year before entering short-time work, and they were converging to
the level of earnings of workers having experienced an unemployment shock instead
(Panel B). The main explanation for this discrepancy is that the Italian recession was
much more protracted, and the shock to firms was therefore much more persistent.
This, in turn, reminds us that short-time work tends to insure against temporary
shocks, but is less effective at insuring against persistent or permanent shocks: if the
shock persists, a firm will not hold onto its workers and will eventually lay them off.
The Relative Moral Hazard Costs of Unemployment Insurance and Short-Time Work
The literature on the moral hazard costs of unemployment insurance is
abundant (for example, see Schmieder, von Wachter, and Bender 2016). A main
conclusion is that the duration of unemployment spells is strongly responsive to
the generosity of unemployment insurance. A smaller literature also investigates
the impact of unemployment insurance generosity on the probability of entering
an unemployment insurance program, and finds moderate responses. In general,
there is less scope for moral hazard along the extensive margin of becoming unem-
ployed (as opposed to the intensive margin of being unemployed for a longer time),
as layoffs are well-defined and well-monitored events, and those who quit are quite
restricted in their ability to access unemployment insurance in many countries.6
Overall, the consensus is that the fiscal externality of increasing the generosity of
unemployment insurance is relatively large: the cost to the government of an addi-
tional $1 of unemployment insurance ranges from $1.50 to $2.50.
Evidence on the moral hazard costs of short-time work is much more limited,
but two elements suggest that these costs might be significant. First, while access to
short-time work is generally made conditional on firms experiencing economic or
financial distress, the definition of distress is not always very precise and can prove
hard to enforce, leaving some room for manipulation. Second, short-time work
subsidizes hours reductions, which requires an effective monitoring of hours actually
worked by employees, a notoriously difficult task for government administrators. As
a result, the massive extension across Europe during the COVID crisis of access to
6
Recent papers hint towards layoff date being responsive to discontinuities in eligibility or generosity
of unemployment insurance, one potential explanation being workers and firms bargaining over the
timing of the layoff: for example, see Khoury (2021) for evidence on France, and Albanese, Picchio, and
Ghirelli (2020) for evidence on Italy.
Should We Insure Workers or Jobs During Recessions? 37
Figure 2
Evolution of Earnings and Transfers Around the Events of Job Loss and Short-
Time Work During the Great Recession
A: Germany
0
Evolution of net earnings + transfers
−0.1
−0.2
−0.3
−0.4
−0.5 Short-time work
Unemployed
−0.6
−4 −3 −2 −1 0 1 2 3 4 5
Time (years)
B: Italy
0
Evolution of net earnings + transfers
−0.1
−0.2
−0.3
−0.4
−0.5 Short-time work
Unemployed
−0.6
−4 −3 −2 −1 0 1 2 3 4 5
Time (years)
Note: The figure reports the evolution of earnings and transfers around job loss (in grey) or around the
start of a short-time work spell (in blue). It shows that job loss is associated with a much larger and much
more persistent drop in resources than short-time work, implying that the marginal insurance value is
likely greater for unemployment insurance than for short-time work. Panel A reproduces estimates from
Tilly and Niedermayer (2016) which uses German administrative data from the Institute for Employment
Research (IAB). It corresponds to a weighted average of the effect of short-time work and unemployment
insurance on income by tenure category using as weights their share in the population. Panel B reproduces
estimates from Giupponi and Landais (2020b) and uses administrative data from INPS on the universe of
employer-employee matches and social security payments in the private sector in Italy.
short-time work for small businesses, where the difficulty of monitoring hours can
be even more acute, has generated fears of a surge in moral hazard.
The existing evidence on the moral hazard costs of short-time work comes
almost exclusively from the Great Recession, but it suggests, interestingly, that these
38 Journal of Economic Perspectives
costs are smaller than anticipated. In the context of Italy, for instance, Giupponi
and Landais (2020b) identify behavioral responses to short-time work using varia-
tion in eligibility rules across firms, and find that for every €1 transferred to a worker
on short-time work during the Great Recession, the total cost to the government
implied by behavioral responses was around €1.4. In the context of Switzerland,
Kopp and Siegenthaler (2021) compared firms who were successful to firms who
were unsuccessful in their short-time work application during the Great Recession
and find a negative mark-up: in other words, short-time work paid for itself.7
What can explain these relatively small (or even negative) fiscal externalities of
short-time work, in contrast with the relatively large moral hazard cost of unemploy-
ment insurance? First, it seems that, at least during the Great Recession, there was
not much manipulation in the reporting of hours worked.8 Second, it appears that
the probability of an individual worker being put on short-time work did not respond
significantly to the generosity of short-time work subsidies, at least not during the Great
Recession. Figure 3 illustrates this point using a large discontinuity in the short-time
work subsidy amount available to workers in Italy at a particular wage threshold. Panel A
shows, using Italian administrative data, that the average short-time work subsidy
increases by 12 percent at the wage threshold. Yet, there is no sign of discontinuity at
the threshold in the probability that a worker is put on short-time work, conditional
on being employed by a firm using it (Panel C), nor in the share of potential working
hours spent on short-time work conditional on being on reduced hours (Panel E).9
Figure 3 also shows evidence of large moral hazard responses to unemployment
insurance benefits in the same Italian context (Scrutinio 2018). In Italy, potential
unemployment benefit duration is entirely based on age at layoff: workers fired before
turning 50 are eligible for eight months of unemployment benefits, while workers
fired after turning 50 can receive up to twelve months of benefits (Panel B).10 Panel D
plots the density of layoffs by age in months and shows a sharp response to the increase
in benefit duration at age 50, suggesting that workers manipulate their layoff date to
obtain more generous benefits. Panel F shows that the average number of weeks on
unemployment insurance—that is, the intensity of utilization—increases sharply from
23 to 31 weeks at the threshold (based on evidence from Scrutinio 2018).
7
Evidence from Germany shows that, during the Great Recession, the utilization of short-time work was
concentrated among large firms hit by the trade collapse (Boeri et al. 2011). The extension of short-time
work schemes to smaller firms during the COVID crisis, as well as the much larger scale of utilization of the
program, may limit the applicability of evidence on moral hazard from the Great Recession to the COVID
one.
8
Using data on firm balance sheets, Giupponi and Landais (2020b) show that in firms taking up short-
time work, value-added per worker fell significantly, and by about the same magnitude as hours per
worker. This indicates that reduction in hours upon take-up of short-time work is in large part a real
response rather than a reporting response.
9
In Italy, the amount of short-time work subsidy received by the worker is equivalent to 80 percent of
forgone earnings due to hours not worked, up to a cap established by law each year. For example, in
2021, workers with contractual monthly earnings above €2,159.48 can receive benefits up to €998.19. For
those with contractual earnings above €2,159.48, the cap is €1,119.72.
10
The benefit amount is proportional to average wages earned in the three months before layoff up
to a cap. Workers receive 60 percent of their average wage for the first six months, 50 percent for the
following two months, and 40 percent for the remaining four months, if still eligible.
Giulia Giupponi, Camille Landais, and Alice Lapeyre 39
Figure 3
Short-Time Work versus Unemployment Insurance Generosity and Take-Up
A. Short-time work benefit schedule B. Unemployment insurance duration schedule
1,300 14
Statutory short-time Statutory unemployment
1,200 12
1,100 10
benefit duration in months
work benefit amount 1,000 8
900 6
−4
−480
−320
−360
−200
−140
−180 46
46–0
47–6
2
−60
6
120
180
2400
0 47–0
48–6
48–0
49–6
49–0
50–6
50–0
300
360
420
4800 51–6
51–0
52–6
52–0
53–6
53–0
54–6
–0
Contractual earnings relative to threshold Age at layoff
C. Extensive margin response: probability of D. Extensive margin response: density of age
worker receiving short-time work at unemployment insurance benefit start
16 0.014
Probability of
14
12 0.012
Density
10
short-time work receipt
0.01
8
6
0.008
4
−4
−480
−320
−360
−200
−140
−180 46
46–0
47–6
47–0
48–6
48–0
2
−60
60
12
180
2400
0 49–6
49–0
50–6
50–0
51–6
51–0
52–6
300
360
420
480
0 52–0
53–6
53–0
54–6
–0
Contractual earnings relative to threshold Age at layoff
E. Intensive margin response: hours reduction F. Intensive margin response: duration of
(conditional on short-time work take-up) unemployment insurance receipt
(conditional on take-up)
78 7.4
Hours reduction (as percent Paid unemployment
6.9
69
6.4
60
duration in months
of potential working hours)
6
51
5.5
41 5.1
−4
−4 80
−320
−360
−200
−140
−1 80 46
46–0
47–6
47–0
48–6
2
−60
60
12
180
2400
0 48–0
49–6
49–0
50–6
50–0
51–6
51–0
300
360
420
4800 52–6
52–0
53–6
53–0
54–6
–0
Contractual earnings relative to threshold Age at layoff
Note: The figure reports a set of regression-discontinuity graphs to illustrate the relationship between
short-time work/unemployment insurance generosity and take-up. Left-hand side panels report evidence
on short-time work, right-hand side ones on unemployment insurance and are based on Scrutinio
(2018). Panel A shows the short-time work benefit schedule as a function of a wage threshold. The
benefit amount is based on a 70 percent hour reduction and 2021 short-time work cap parameters. Panel
C reports the probability for a worker to be put on short-time work in a given year (pooling years 2011–
2014), conditional on being employed in a firm that uses short-time work in that year. Panel E reports the
average number of hours a worker spends on short-time work per month (as a percent of total potential
working hours), conditional on being on short-time work in that month. Panel B plots unemployment
insurance potential benefit duration as a function of a worker’s age at layoff. Panel D reports the density
of layoffs by age at layoff. Panel F shows the average duration of unemployment insurance in months,
conditional on unemployment insurance take-up. Panels C–F are based on Italian Social Security data.
40 Journal of Economic Perspectives
Rigidities or frictions to individual level bargaining within the firm may explain
why short-time work take-up does not respond much to variation in the generosity
of the subsidy at the individual level. These rigidities in turn can also rationalize why
firms’ behavior is generally much more responsive to a variation in the firm’s rather
than the worker’s side of the job surplus (Jäger, Schoefer, and Zweimüller 2019).
Firms appear to be responsive to how social insurance programs like unemployment
insurance or short-time work are financed: for instance, if unemployment insurance is
funded by experience-rated premia on firms (that is, firms where more workers claim
unemployment insurance pay somewhat higher premia), this tends to reduce layoffs
significantly. Time-series evidence also suggests that the take-up of short-time work by
firms declines significantly with the tightening of short-time work’s experience rating.
The last reason why the fiscal cost of short-time work appears limited, and prob-
ably the main reason for it, is that the fiscal cost of short-time work is of course
endogenous to the generosity of unemployment insurance. If more generous short-
time work prevents layoffs, these positive employment effects mechanically reduce
the fiscal cost to the unemployment insurance system, as fewer workers end up
collecting unemployment insurance. In turn, the more generous the unemploy-
ment insurance system relative to short-time work, the larger will such savings be
for the government. This leads us to the central question: does short-time work
effectively save jobs? If so, what are the welfare consequences?
Short-Time Work and Job Destruction
While both unemployment insurance and short-time work offer insurance
against labor market shocks, they differ in a fundamental way. Short-time work seeks
to preserve labor market matches by subsidizing jobs rather than job-seekers. In other
words, short-time work, contrary to unemployment insurance, aims at reducing job
separations. It is therefore critical to establish to what extent short-time work effec-
tively saves jobs. Moreover, the social welfare impact of saving jobs will depend on
whether separations are inefficiently high in recessions to begin with. So let us first
review the evidence on the employment effects of short-time work, before delving into
the reasons why layoffs may be inefficient and determining whether subsidizing labor
hoarding may be socially desirable.
The Employment Effects of Short-Time Work
To determine whether short-time work saves jobs, a natural place to start
is to leverage the large variation in short-time work usage during the COVID
crisis across countries. Figure 4 provides such a macro perspective and shows
the presence of a very robust negative correlation between the fraction of the
working-age population that took up short-time work and the evolution of the non-
employment rate during the crisis. One additional worker enrolling in short-time
work is correlated with .27 fewer workers being non-employed. This strong corre-
lation between employment and short-time work usage echoes time-series and
cross-country evidence from previous recessions (for example, Van Audenrode
Should We Insure Workers or Jobs During Recessions? 41
Figure 4
Short-Time Work Usage and Non-Employment During the COVID Crisis: Cross-
Country Evidence
4 USA
Resid. change in non-employment rate (y-o-y)
GBR
2
USA
SWE USA
USA
ESP ESP IRL
ESP SWEBEL
SWE
0 IRL
BELESP BEL
ESP
SWEBEL
BEL GBR SWE IRL
FRAGBR
GBRIRL
FRA
FRA FRA
FRA
−2 USA GBR
IRL
βSTW = −0.27 (0.066)
−4
−12 −10 −8 −6 −4 −2 0 2 4 6 8 10 12
Resid. short-time work take-up (percent working-age pop.)
2020:II 2020:III 2020:IV 2021:I 2021:II Fitted line
Note: The figure reports a scatter plot of the relationship between the year-on-year change in the quarterly
non-employment rate and in the rate of short-time work take-up at the country level. Data are not
seasonally adjusted. To remove the seasonal component, we take the year-on-year change—i.e., for a given
year-quarter YYQX, we apply the following transformation to the data: x ̃YYQX = xYYQX
– x(YY–1)QX
. Short-time
work take-up is computed as the ratio of the number of individuals in the program over the working-age
population. For short-time work take-up in 2020 we take the variable in level as take-up was close to 0 in
2019. Outcomes are residualized against year-quarter fixed effects, the year-on-year change in the number of
COVID cases (linear and quadratic), and in unemployment insurance take-up. The red line represents the
linear fit. The figure reports the slope coefficient and associated standard error (in parentheses), clustered
at the country level. Data on employment come from OECD. Data on short-time work and unemployment
insurance take-up come from the OECD and national statistics (Scarpetta, Carcillo, and Hijzen 2022). Data
on COVID cases come from the Johns Hopkins Coronavirus Resource Center. See online Appendix C for
details on data sources and the construction of short-time work/unemployment insurance take-up.
1994; Abraham and Houseman 1993; Boeri et al. 2011; Hijzen and Venn 2011;
Cahuc and Carcillo 2011).
Going beyond such correlations is complicated, and direct causal evidence
on the employment effects of short-time work is scant.11 The issue lies in the lack
of credibly exogenous sources of variation in short-time work treatment across
firms—an issue that will become even more acute for the current recession, as most
countries have purposefully extended short-time work access to every firm. This
situation severely complicates identification, with no obvious method to control for
the selection of firms into short-time work.
11
Using firm survey and administrative data in Denmark at the onset of the COVID pandemic, Bennedsen
et al. (2020) compare actual furlough decisions and counterfactual decisions in the absence of aid. They
find that the policy was effective in preserving job matches at the beginning of the pandemic.
42 Journal of Economic Perspectives
Three recent papers focusing on the Great Recession do seek to address these
selection problems and provide credible evidence of a positive, strong, and causal
relationship between short-time work and employment. Looking at Swiss firms,
Kopp and Siegenthaler (2021) compare firms whose short-time work application
was granted to similar firms whose application was rejected. The unsuccessful estab-
lishments provide a valid counterfactual for the successful ones because approval
practices across Swiss cantons are partly idiosyncratic. They find that short-time
work prevented a large number of dismissals and significantly reduced the inci-
dence of long-term unemployment.
Looking at French firms, Cahuc, Kramarz, and Nevoux (2021) use the prox-
imity of a firm to other firms that used short-time work in the past as an instrumental
variable for short-time work take-up during the Great Recession. As an alternative
instrument, they use response-time variation in the administrative treatment of
short-time work applications across French departments. They find large and signif-
icant employment effects of short-time work treatment.
Finally, looking at data from Italy, Giupponi and Landais (2020b) exploit
plausibly exogenous variation in short-time work eligibility rules based on the
interaction between industry and firm size. Their approach and main results are
presented in Figure 5. The figure shows the evolution of the difference in short-
time work take-up between eligible firms and similar counterfactual firms without
access to short-time work, around the time of the Great Recession (blue diamonds).
The evidence confirms that after the onset of the crisis, the take-up of short-time
work among eligible firms surged quickly. The chart further reports the evolution
of hours (red triangles) and total employment (purple circles) in eligible firms
relative to counterfactual non-eligible firms. It demonstrates that short-time work
had large and significant effects on firms’ employment at both the intensive and
extensive margin. Compared to counterfactual firms, firms treated by short-time
work experienced a 40 percent reduction in hours worked per employee, which was
met by an increase of similar magnitude in the number of headcount employees.
Consistent with the findings of Kopp and Siegenthaler (2021) using Swiss firms,
further results show that the employment effects are mostly driven by a reduction in
dismissals among firms that would otherwise experience mass layoffs. Interestingly,
Giupponi and Landais (2020b) also find no effect of short-time work on the wages
of incumbents nor on the wages of new hires.
The Welfare Value of Labor Hoarding Subsidies
Overall, short-time work does seem to preserve jobs. But why is that valuable?
In other words, why are employment adjustments at the intensive margin (hours
reduction) versus extensive margin (layoffs) not equivalent in terms of welfare?
Preserving job matches is valuable for at least three reasons. First, frictions
in the labor market, as well as hiring and training costs, make it costly for firms
to replace workers and for workers to change jobs. Second, workers may accumu-
late human capital that is specific to their job, and separations risk destroying this
valuable source of idiosyncratic productivity. Finally, unemployment often entails
long-run scarring effects for workers (as discussed, for example, by Sullivan and
Giulia Giupponi, Camille Landais, and Alice Lapeyre 43
Figure 5
Employment Effects of Short-Time Work in Italy: Evidence from Quasi-random
Eligibility Variation across Firms
8 STW take-up (left axis) 0.08
log hours worked per employee
Estimated effect on hours/employment
log headcount employment
Probability to receive short-time work
6 0.06
4 0.04
2 0.02
0 0
−2 −0.02
−4 −0.04
Estimated hours effect βIV = −0.514 (0.034)
Estimated employment effect βIV = 0.382 (0.036)
−6 −0.06
20
20
20
20
20 00
01
02
03
04
05
20
20 06
07
20 20
20
20
20
20
20
20
08
09
10
11
12
13
14
Note: The figure reproduces estimates from Giupponi and Landais (2020b) on the effect of short-time
work eligibility on the probability of short-time work take-up, the log of hours worked per employee, and
the log of employment headcount at the firm level. The graph reports the estimated coefficients and
associated confidence intervals (capped vertical bars) from a reduced-form regression of the outcome of
interest on an indicator of short-time work-eligibility at the firm level interacted with year dummies. All
results are relative to 2007. The graph also reports the instrumental variable coefficient (and standard
errors in parentheses) of the effect of short-time work take-up on log hours worked per employee and
log employment headcount.
von Wachter 2009). As a consequence, we should observe significant labor hoarding:
firms and workers should be willing to preserve matches when hit by negative shocks.
But in practice, the socially efficient level of labor hoarding may not be achieved.
Liquidity constraints are probably the most obvious and prevalent reason: during a
shock, a firm may lack the funds necessary to pay wages and retain its workers. Giroud
and Mueller (2017) document that, during the Great Recession, US firms facing
higher liquidity constraints, as proxied by pre-crisis levels of leverage, were (all else
equal) more likely to reduce employment in response to a consumer demand shock.
Of course, employers could try to negotiate temporary wage or hours adjustments
with their employees to deal with such liquidity constraints. But bargaining costs and
commitment issues may often make such renegotiation impractical. Therefore, wage
and hours rigidities may interact with liquidity constraints to amplify the employ-
ment response to negative shocks (Schoefer 2016; Jäger, Schoefer, and Zweimüller
2019). Finally, note that generous and imperfectly experience-rated unemployment
insurance may also already distort workers’ and firms’ choices in favor of (socially inef-
ficient) dismissals. As pointed out by Braun and Brügemann (2014), this interaction
44 Journal of Economic Perspectives
between short-time work and the pre-existing distortions caused by unemployment
insurance is critical to the welfare analysis of short-time work.
If separations are indeed inefficiently high during recessions because of
liquidity constraints and other bargaining frictions, subsidizing labor hoarding
can be efficient. Indeed, evidence from Giupponi and Landais (2020b) regarding
Italian firms strongly supports this idea. It shows that liquidity-constrained firms,
identified using various indicators from balance-sheet data, were much more likely
to take up short-time work. Moreover, the treatment effects of short-time work were
much more positive for these firms. The number of jobs saved per subsidized hour
was significantly larger for them, and so was the effect of short-time work on the
probability that the firm survived.
In sum, the liquidity constraint channel seems critical in explaining the excess
sensitivity of employment adjustments to productivity shocks and supports the idea
of having job match subsidies to correct for inefficiently high separations. Yet two
important questions remain.
First, what is the welfare value of saving these jobs? The answer depends on the
value of the surplus of the marginal job match saved: the larger the value of a match,
the larger the positive welfare effect of preserving it. Unfortunately, this value is an
object that is hard to fathom, let alone to measure precisely, and on which there is
little consensus in the literature.
Second, why should short-time work be the only way to implement such
subsidies? What about other policy instruments? A natural alternative instrument
would be “recall” unemployment insurance, under which workers can return to
their former employer after a spell on unemployment insurance, thus preserving
the job match. But in practice, recall unemployment insurance entails much less
commitment to preserving the job match. Furthermore, it lacks the flexibility to
insure against partial reductions in hours, a flexibility which can prove effective in
addressing financial constraints and in preserving employment. What about direct
wage subsidies, or direct provision of liquidity with temporary loans, such as the
Paycheck Protection Program in the United States? If liquidity constraints are in
fact the main underlying source of inefficiency, tools addressing these financial
constraints directly may be more appropriate than short-time work.
But even with these alternatives in mind, two arguments tend to support
short-time work. The option of short-time work offers expediency: it can almost
immediately provide the funds necessary to cover a firm’s payroll. In contrast,
dedicated loan programs, as demonstrated by the experience of the US Paycheck
Protection Program, can take more time to be activated and for funds to actually
reach firms. Short-time work also may offer superior targeting, because it channels
liquidity to firms that are willing to reduce their hours, which tends to be an effec-
tive screening mechanism. In practice, short-time work selects firms effectively hit
by negative shocks, as measured by revenues, labor productivity, or the predicted
probability to engage in mass layoffs (Giupponi and Landais 2020b; Kopp and
Siegenthaler 2021). This screening property makes short-time work more effective
than non-targeted wage subsidies, which can end up subsidizing a lot of non-
marginal matches.
Should We Insure Workers or Jobs During Recessions? 45
Reallocation in the Labor Market
Recessions are times of intense reallocation between workers and firms (Foster,
Grim, and Haltiwanger 2016). They are also usually characterized by slackness in
the labor market: many workers are searching for jobs while firms demand less labor
and post fewer vacancies.
Both short-time work and unemployment insurance affect workers’ search
effort and firms’ labor demand. To assess the social desirability of both programs,
we must therefore also factor in how they impact labor market reallocation. We start
by focusing on the effect of both social insurance schemes on the overall tightness
of the labor market. We then delve into the impact of both policies on the optimal
sorting of workers into firms in the labor market.
Social Insurance and the Tightness of the Labor Market
Reallocation in the labor market occurs as workers search for new jobs and
firms hire new workers. When there are a lot of workers searching for jobs, it is easy
for firms to hire new workers: the labor market is “slack,” which is good for firms,
but bad for workers. When, to the contrary, there are a lot of vacancies but few
workers searching for jobs, the labor market is “tight.”
Unemployment insurance and short-time work affect at the same time the
aggregate search effort of workers and the incentives for firms to hire new workers:
their overall effect on the tightness of the labor market is therefore a priori ambig-
uous. If generous unemployment insurance increases wages or if short-time work
strongly reduces the need for new hires, more generous social insurance might
strongly reduce the number of vacancies posted by firms and make the labor market
even more slack in recessions, delaying recovery. To the contrary, if labor demand
is rigid during recessions and the labor market exhibits job rationing, workers
searching for jobs find themselves in a rat race. In such contexts, reducing search
effort through more generous unemployment insurance or short-time work can
benefit workers by increasing the tightness of the labor market (Landais, Michaillat,
and Saez 2018b).
What do the data tell us about the overall effect of unemployment insurance
and short-time work on equilibrium tightness? As a starter, we can again exploit
the large variation in short-time work and unemployment insurance usage across
countries and over time during the recent crisis. For this purpose, we built consis-
tent measures of job-filling probabilities, using the ratio of hires to vacancies. These
measures are direct proxies of the slackness of the market: the tighter the market,
the harder it is for firms to hire workers, and the lower the job-filling probability as
a result. We then correlate the change in job-filling probabilities with the change in
short-time work and in unemployment insurance take-up across countries and across
quarters during the current recession. Our results, reported in Figure 6, show that
short-time work (Panel A) and unemployment insurance (Panel B) both seem nega-
tively correlated with job-filling probabilities (or equivalently positively correlated
with labor market tightness) in a recession, which is consistent with the presence
of significant job rationing in downturns. The evidence suggests that increasing the
46 Journal of Economic Perspectives
Figure 6
Cross-Country Correlation between Job-Filling Probability and Take-Up of Short-
Time Work and Unemployment Insurance
Panel A. Job-filling probability and short-time work take-up
4
3
Resid. change in q(θ) (q-o-q)
ESP
2
FRA ESP USA
BEL
1 USA IRL
SWE ESP
FRA
SWE SWE
0 IRLUSA USA
BEL
BEL FRA IRL
SWE BEL
IRL
−1 BEL ESP FRA
SWE
−2
USA
−3 ESP
βSTW = −0.125 (0.081)
−4
−10 −8 −6 −4 −2 0 2 4 6 8 10
Resid. short-time work take-up (q-o-q, percent working-age pop.)
Panel B. Job-filling probability and unemployment insurance take-up
4
3 ESP
Resid. change in q(θ) (q-o-q)
2
SWE USA
BEL
ESP
1 ESP
BEL FRA
SWE IRL
IRL USA FRA
0 USA
BEL
USA
BEL
FRA SWE
SWE IRL
FRA IRL
−1
SWE ESPBEL
USA
−2
−3 ESP
βUI = −0.128 (0.066)
−4
−10 −8 −6 −4 −2 0 2 4 6 8 10
Resid. unemployment insurance take-up (q-o-q, percent working-age pop.)
2020:II 2020:III 2020:IV 2021:I 2021:II Fitted line
Note: The figure shows how short-time work and unemployment insurance take-up during the COVID
crisis correlate with tightness in the labor market. We use the vacancy-filling probability q(θ) as a proxy for
labor market tightness. The higher the vacancy-filling probability, the easier it is for firms to hire workers
when opening a vacancy, and the slacker the labor market as a result. Both panels report scatter plots
of the relationship between the quarter-on-quarter change in q(θ) and in the rate of short-time work/
unemployment insurance take-up at the country level. Data are not seasonally adjusted. To remove the
seasonal component, we take the quarter-on-quarter change—that is, for a given year-quarter YYQX, we
apply the following transformation to the data: x ̃YYQX = xYYQX – xYYQ(X–1). Short-time work and unemployment
insurance take-up are computed as the ratio of the number of individuals in the program over the working-
age population. Outcomes are residualized against year-quarter fixed effects, the quarter-on-quarter
change in the number of COVID cases (linear and quadratic), and in the take-up of the other policy
instrument. The red line represents the linear fit. The figure reports the slope coefficient and associated
standard error (in parentheses), clustered at the country level. Data for European countries come from the
Job Vacancy Statistics and Labor Force Survey, and for the United States from the Job Openings and Labor
Turnover Survey. For European countries, hires are proxied by recent job starters—i.e., individuals who
reported having started their employment in the last three months before the interview. Job openings are
restricted to the private sector. Data on short-time work and unemployment insurance take-up come from
the OECD and national statistics (Scarpetta, Carcillo, and Hijzen 2022). Data on COVID cases come from
the Johns Hopkins Coronavirus Resource Center. See online Appendix C for details on data sources and
the construction of short-time work/unemployment insurance take-up.
Giulia Giupponi, Camille Landais, and Alice Lapeyre 47
generosity of short-time work or unemployment insurance in a recession can be an
effective way of alleviating the search inefficiencies created by rat-race externalities.
This cross-country evidence is corroborated by a stream of recent papers that
identify the impact of social insurance on search externalities and equilibrium
tightness using quasi-experimental designs. Lalive, Landais, and Zweimüller (2015)
exploit a massive expansion in the generosity of unemployment insurance to a large
subgroup of workers in Austria and show that non-eligible workers have significantly
higher job finding rates, lower unemployment durations, and lower risk of long-term
unemployment as a result. Marinescu (2017) uses job board data and exploits quasi-
random variation in unemployment insurance expansions across US states during
the Great Recession: she finds that unemployment insurance reduced search effort
significantly but did not affect job vacancies, so that tightness went up significantly
as a result. Marinescu, Skandalis, and Zhao (2020, 2021) exploit variation in unem-
ployment insurance across US labor markets stemming from the CARES Act and
Federal Pandemic Unemployment Compensation (FPUC). Using granular data
from the online job platform Glassdoor, they show in both cases that increases in
unemployment insurance generosity significantly increased labor market tightness.
Finally, using exogenous variation across local labor markets in Italy in their expo-
sure to short-time work, Giupponi and Landais (2020b) find that greater access to
short-time work decreases the job finding probability in the labor market, but that
the magnitude of the effect is small. Overall, these results confirm that both unem-
ployment insurance and short-time work increase tightness during downturns, and
the effect seems to be more pronounced for unemployment insurance.
Of course, the welfare consequences of increasing tightness depend on whether
tightness is inefficiently low or high in recessions. Historically, labor markets tend to
be very slack during downturns. Michaillat and Saez (2021) offer a general charac-
terization as well as a measure of the efficient level of tightness in the United States.
They find that the labor market has been particularly inefficiently slack during past
recessions. The intuition is that the social cost of unemployment is very large relative
to firms’ recruiting costs during downturns. Pushing tightness up and increasing the
job-finding probability of workers is then socially desirable: the reduction in the social
cost of unemployment greatly outweighs the increased costs of recruiting for firms.
However, evidence from the current crisis suggests that this time is different.
Looking at the long-run evolution of the average vacancy-filling probability in the
United States, it is striking to see that it has remained at a historic low during the crisis.
Overall, this recession seems unique: it is a tight recession in the labor market.12
Can this sustained level of tightness actually be explained by the large expan-
sion of unemployment insurance generosity and coverage in the United States at
the onset of the COVID crisis? Would the situation be different if US policymakers
had resorted more to short-time work, which seems to put less upward pressure
12
For more information on labor market tightness during the pandemic, including figures showing
the hires-to-vacancies ratio for the US and various European economies, see online Appendix D. In the
US data, the brief surge in the second quarter of 2020 can be entirely explained by early recalls from
unemployment.
48 Journal of Economic Perspectives
on tightness? When we consider comparable data for European countries, we find
that they have also experienced what appears to be tight labor markets during the
pandemic recession, which suggests that the mix of social insurance policies used
during the COVID crisis is probably not responsible. But this pattern also implies
that there is probably no need to push tightness further up going forward. Exploring
the factors behind this uniquely high level of tightness during a slump is important
to guide the policy response during the recovery.
Slowing Down Reallocation across Firms and Sectors
Recessions trigger shocks that are asymmetric across firms and sectors. As a
result, significant reallocation usually follows in the labor market: workers move away
from firms persistently hit by bad shocks and toward more productive job matches,
a movement which enhances aggregate efficiency. In recent months, concerns have
emerged again on the impact that higher social insurance might have on the pace
of this sectoral and firm reallocation (for example, Barrero et al. 2021).
Both unemployment insurance and short-time work have the potential to
hinder reallocation, although the mechanism by which they do so differs. In theory,
unemployment insurance is a general brake to aggregate reallocation: by lowering
the search effort of the unemployed, it can slow the pace at which workers who
have been dismissed from lower productivity jobs may move to more productive
matches. Short-time work is a specific brake to sectoral/firm reallocation: it discour-
ages workers in firms/sectors that are hit by productivity shocks from reallocating
to other firms/sectors by keeping them in their jobs. The extent to which this is
problematic for aggregate productivity depends on whether the shock is temporary
or permanent: if the shock is permanent, then short-time work may subsidize persis-
tently unproductive matches and hinder reallocation towards more productive job
matches.
How serious are these negative reallocation effects of unemployment insur-
ance and short-time work in practice? Regarding unemployment insurance, we know
surprisingly little on its overall impact on reallocation and aggregate efficiency in the
labor market. For short-time work, evidence from Giupponi and Landais (2020b)
on Italian firms sheds some interesting light on its impact on reallocation. It finds
that short-time work tends to subsidize persistently low productivity matches, as low
productivity firms tend to over-select into short-time work. Italian firms that were
already below the median of labor productivity before the onset of the recession
were twice as likely to select into short-time work during the Great Recession, and
the employment effects of short-time work are also significantly lower for these low
productivity firms. Furthermore, exploiting variation across local labor markets, they
show that (exogenously) higher exposure to short-time work is significantly and nega-
tively correlated with the employment growth of high productivity firms. In other
words, high productivity firms have a harder time growing in a local labor market
where low productivity firms have more access to short-time work. While this clearly
supports the idea that short-time work slows down reallocation, the magnitude of
the estimated effects remains small. However, the level of take-up of short-time work
was also much smaller during the Great Recession than in the current crisis, and one
Should We Insure Workers or Jobs During Recessions? 49
cannot exclude that short-time work may have much stronger negative effects on real-
location in the current recovery.
Further Externalities
Besides inefficient separations and reallocation frictions, it is worth pointing
to a few further externalities with which short-time work and unemployment insur-
ance may interact: aggregate demand externalities, fairness, and health.
A usual argument in favor of generous social insurance during recessions
relates to their fiscal multiplier effects: unemployment insurance and short-time
work transfer money to individuals who tend to have higher-than-average marginal
propensities to consume. These high marginal propensities to consume, in turn,
may help trigger positive aggregate demand externalities in a slump. A small litera-
ture has tried to embed social insurance into New Keynesian models to quantify
the size of these multiplier effects (for example, McKay and Reis 2016; Michaillat
and Saez 2019; Guerrieri et al. 2020; Kekre forthcoming). How large are these fiscal
multiplier effects? Which program commands the larger fiscal multipliers: short-
time work or unemployment insurance?
Unemployment insurance, as explained above, tends to insure individuals
experiencing larger shocks with lower means to smooth consumption: this suggests
that unemployment insurance recipients are likely to have larger marginal propen-
sities to consume. But short-time work, by preserving employment and improving
expectations regarding future employment and income, may reduce the need for
precautionary savings and thus raise marginal propensities to consume compared
to unemployment insurance.
The marginal propensities to consume of recipients of unemployment insur-
ance are large, and significantly larger than those of employed people. Comparing
the same individuals over time in Sweden, Landais and Spinnewijn (2021) find that
the marginal propensity to consume is around 25 percent higher when unemployed
than employed. But much less is known on the marginal propensities to consume
of individuals on short-time work. We looked at the data from the German High-
Frequency Online Personal Panel Survey (HOPP) mentioned earlier, and elicited
marginal propensities to consume following the approach of Jappelli and Pistaferri
(2014). We find that the marginal propensities to consume of German short-time
work recipients was slightly larger than that of employed workers, but smaller than
that of unemployment insurance recipients.13
However, moderate differences in marginal propensities to consume between
unemployment insurance and short-time work recipients are unlikely to translate
into sizeable differences in aggregate demand externalities between these two poli-
cies, because the fraction of the labor force receiving unemployment insurance or
short-time work is small relative to the size of the employed population. For that
13
For details and additional discussion, see online Appendix Table B1.
50 Journal of Economic Perspectives
reason, simulations such as in McKay and Reis (2016) suggest that, quantitatively, the
stabilization effects of these forms of social insurance are small and second-order.14
Fairness appears to be an important institutional tenet in European labor
markets (Saez, Matsaganis, and Tsakloglou 2012; Goldschmidt and Schmieder 2017;
Saez, Schoefer, and Seim 2019). Fairness concerns suggest that short-time work may
prove a more equitable way to insure against labor market fluctuations. If firms avoid
layoffs and instead reduce hours of work per worker, the costs of recessions are less
concentrated on a small number of workers who suffer large losses in income and
other job-related benefits. Interestingly, this argument is often mentioned in the
policy debate in countries with strong short-time work programs.
Finally, in the current pandemic, the ability granted by short-time work to flex-
ibly reduce hours of work and keep workers away from the workplace may have had
positive health externalities by reducing the spread of the virus.
Conclusion
While very little was known about short-time work schemes and their poten-
tial welfare effects, this did not prevent European policymakers from resorting to
them aggressively during the COVID crisis. The evidence gathered in this paper
and summarized in Table 1 shows they probably did the right thing. The value of
insurance provided by short-time work transfers is clearly lower than that of unem-
ployment insurance benefits, but the moral hazard they entail seems more limited
than for unemployment insurance, especially when experience-rating of social
insurance is limited. Importantly, recent evidence confirms that short-time work can
be an efficient and expedient way to attenuate the social costs created by “excess”
layoffs in recessions.
But while the policy debate has tended to discuss a choice between short-time
work and unemployment insurance, this paper has revealed two important points.
First, the frontier between unemployment insurance and partial unemployment
or short-time work policies can often be quite tenuous. In other words, in certain
circumstances, unemployment insurance schemes can be made to mimic the func-
tioning of short-time work or furlough schemes by maintaining important ties
between the unemployed and their employers.
Second, far from being substitutes, these two types of policies exhibit strong
complementarities. Unemployment insurance and short-time work do not insure
the same types of workers nor the same labor market risks. They do not distort
the same margins of behaviors. There are important fiscal spillovers between the
two programs, as more generous short-time work reduces the risk of layoffs. Their
effects on reallocation in the labor market are also complementary. Short-time work
14
For social insurance to have large multiplier effects, it would need to have a strong effect on the
consumption behavior of the large population of the employed as well. But in practice, the precautionary
savings channel (by which employed individuals save less when they have access to more generous social
insurance against labor market shocks) seems too small to sustain large aggregate demand externalities.
Giulia Giupponi, Camille Landais, and Alice Lapeyre 51
Table 1
The Welfare Effects of Insuring Workers versus Jobs: A Summary of the Evidence
Correction of
other inefficiencies
Value of Moral hazard/ Excess Tightness
transfer fiscal externality layoff extern. Reallocation
Short-time work + +/- ++ ? -
Unemployment insurance ++ -- ? + ?
Note: The table provides our own summary evaluation of the empirical literature evaluating features of
short-time work and unemployment insurance programs that map onto those key elements of welfare
analysis. For discussion of the underlying studies, see the text of the paper. The symbols reported in the
table refer to the magnitude of the welfare effect for each feature, as per the following legend: (+ +) large
positive, (+) positive, (+/-) both positive and negative, (-) negative, (- -) large negative, (?) no evidence.
is an efficient policy to deal with temporary shocks, while unemployment insur-
ance can take care of shocks that end up being more persistent. In countries with
already generous unemployment insurance and/or strong employment protection,
like European countries, strong cyclical short-time work programs are therefore a
valuable complement to unemployment insurance as a policy response to reces-
sions. And, in general, having both programs is a great way to handle any type of
recession in the labor market. In other words, we should insure workers and jobs
during recessions.
But this paper has also emphasized that social insurance critically interacts with
equilibrium in the labor market, with important consequences for reallocation and
efficiency. On this front, much more research needs to be done. As the current
crisis seems to be unique in maintaining high tightness in the labor market, a better
understanding of how unemployment insurance and short-time work affect reallo-
cation will be key to determine the optimal policy path for the recovery. Attention
should in particular be devoted to determining how unemployment insurance and
short-time work should be coordinated with other instruments, such as hiring subsi-
dies, in order to boost labor demand and prevent reallocation issues.
■ We thank David Autor, Tito Boeri, Xavier Jaravel, Pascal Michaillat, Daniel Reck,
Emmanuel Saez, and Johannes Spinnewijn, and seminar participants at CREST, the Institute
for Economic Analysis at Universitat Autònoma de Barcelona, the SIEP Annual Conference,
the International Monetary Fund and the Franco-German Seminar for helpful comments and
suggestions. We are grateful to Andrea Salvatori and Alexander Hijzen for sharing OECD
data on short-time work take-up, Robert Hall and Marianna Kudlyak for sharing data on
recalls, and Vincenzo Scrutinio, Jan Tilly, and Kilian Niedermayer for sharing data and
code from their papers. We also thank Francesco Armillei, Nicolas Grimprel, and Colombe
Ladreit de Lacharrière for excellent research assistance. We gratefully acknowledge support
from ERC grant #679704 -DYNAMICSS, STICERD, ESRC, and the INPS “Valeria Solesin”
Fellowship. The findings and conclusions expressed are solely those of the authors and do not
represent the views of INPS.
52 Journal of Economic Perspectives
References
Abraham, Katharine G., and Susan N. Houseman. 1993. “Does Employment Protection Inhibit Labor
Market Flexibility? Lessons from Germany, France, and Belgium.” NBER Working Paper 4390.
Albanese, Andrea, Matteo Picchio, and Corinna Ghirelli. 2020. “Timed to Say Goodbye: Does Unemploy-
ment Benefit Eligibility Affect Worker Layoffs?” Labour Economics 65: 101846.
Baily, Martin Neil. 1978. “Some Aspects of Optimal Unemployment Insurance.” Journal of Public Economics
10 (3): 379–402.
Barrero, Jose Maria, Nicholas Bloom, Steven J. Davis, and Brent H. Meyer. 2021. “COVID-19 Is a
Persistent Reallocation Shock.” University of Chicago, Becker Friedman Institute for Economics
Working Paper 2021-02.
Bennedsen, Morten, Birthe Larsen, Ian Schmutte, Daniela Scur. 2020. “Preserving Job Matches During
the COVID-19 Pandemic: Firm-Level Evidence on the Role of Government Aid.” GLO Discussion
Paper 588.
Boeri, Tito, and Herbert Bruecker. 2011. “Short-Time Work Benefits Revisited: Some Lessons from the
Great Recession.” Economic Policy 26 (68): 697–765.
Braun, Helge, and Björn Brügemann. 2014. “Welfare Effects of Short-Time Compensation.” CESifo
Working Paper Series 5063.
Bundesagentur für Arbeit Statistik. 2021. “Angezeigte und realisierte Kurzarbeit (Zeitreihe Monats und
Jahreszahlen).” Bundesagentur für Arbeit, Germany. https://statistik.arbeitsagentur.de/DE/Navi-
gation/Statistiken/Fachstatistiken/Leistungen-SGBIII/Kurzarbeitergeld/Kurzarbeitergeld-Nav.
html (accessed November 29, 2021).
Bundesagentur für Arbeit Statistik. 2021b. “Arbeitslosengeld - Deutschland, West/Ost, Länder, Kreise,
Regionaldirektionen, Agenturen für Arbeit (Zeitreihe Monats- und Jahreszahlen ab 2005).”
Bundesagentur für Arbeit, Germany. https://statistik.arbeitsagentur.de/DE/Navigation/Statis-
tiken/Fachstatistiken/Leistungen-SGBIII/Arbeitslosengeld/Arbeitslosengeld-Nav.html (accessed
November, 29 2021).
Bureau of Labor Statistics. 2000–2021. “Job Openings and Labor Turnover Survey (JOLTS) - All esti-
mates - jt.data.1.AllItems.” Washington, DC: US Department of Labor. https://download.bls.gov/
pub/time.series/jt/jt.data.1.AllItems (accessed November 27, 2021).
Cahuc, Pierre, and Stéphane Carcillo. 2011. “Is Short-Time Work a Good Method to Keep Unemploy-
ment Down?” Nordic Economic Policy Review 1 (1): 133–65.
Cahuc, Pierre, Francis Kramarz, and Sandra Nevoux. 2021. “The Heterogeneous Impact of Short-Time
Work: From Saved Jobs to Windfall Effects.” CEPR Discussion Paper 16168.
Chetty, Raj. 2008. “Moral Hazard Versus Liquidity and Optimal Unemployment Insurance.” Journal of
Political Economy 116 (2): 173–234.
Dares. 2020. “Données trimestrelles – Activité partielle au 1er trimestre 2019.” Ministère du Travail,
France. https://dares.travail-emploi.gouv.fr/donnees/le-chomage-partiel (accessed November
29, 2021).
Dares. 2021a. “Données mensuelles – Données administratives de l’activité partielle (DAP, DI, heures,
montants...).” Ministère du Travail, France. https://dares.travail-emploi.gouv.fr/donnees/
le-chomage-partiel (accessed November 29, 2021).
Dares. 2021b. “Données mensuelles CVS-CJO - Demandeurs d’emploi.” Ministère du Travail, France.
https://dares.travail-emploi.gouv.fr/donnees/les-demandeurs-demploi-inscrits-pole-emploi-
France-metro (accessed November 30, 2021).
Dong, Ensheng, Hongru Du, Lauren Gardner. 2020. “An Interactive Web-Based Dashboard to Track
COVID-19 in Real Time.” Lancet 20 (5): 533–4.
Dun & Bradstreet. 2021a. “ISO 2-Digit Alpha Country Code.” Jacksonville: Dun & Bradstreet. https://
www.dnb.com/content/dam/english/dnb-solutions/sales-and-marketing/iso_2digit_alpha_
country_codes.xls (accessed April 7, 2021).
Dun & Bradstreet. 2021b. “ISO 3-Digit Alpha Country Code.” Jacksonville: Dun & Bradstreet. https://
www.dnb.com/content/dam/english/dnb-solutions/sales-and-marketing/iso_3digit_alpha_
country_codes.xls (accessed April 7, 2021).
Employment and Training Administration. 2021. “UI Weekly Continued Claims – All programs.”
Washington, DC: Department of Labor. https://oui.doleta.gov/unemploy/docs/allprograms.xlsx
(accessed November 29, 2021).
Eurostat. 2001–2021. “Job vacancy statistics by NACE Rev. 2 activity - quarterly data (from 2001 onwards).”
Should We Insure Workers or Jobs During Recessions? 53
Luxembourg: European Union Statistical Office. https://appsso.eurostat.ec.europa.eu/nui/show.
do?dataset=jvs_q_nace2&lang=en (accessed November 27, 2021).
Eurostat. 2001–2021b. “Job vacancy statistics by NACE Rev. 2 activity - quarterly data (from 2001
onwards).” Luxembourg: European Union Statistical Office. https://appsso.eurostat.ec.europa.
eu/nui/show.do?dataset=jvs_q_nace2&lang=en (accessed November 27, 2021).
Eurostat. 2009–2021. “Recent job starters by sex and age - quarterly data, AGE From 20 to 64 years,
S_ADJ Unadjusted data (i.e. neither seasonally adjusted nor calendar adjusted data), SEX Total,
UNIT Total - lfsi_sta_q.” Luxembourg: European Union Statistical Office. https://appsso.eurostat.
ec.europa.eu/nui/show.do?dataset=lfsi_sta_q&lang=en (accessed November 27, 2021).
Foster, Lucia, Cheryl Grim, and John Haltiwanger. 2016. “Reallocation in the Great Recession: Cleansing
or Not?” Journal of Labor Economics 34 (S1): 293–331.
Giroud, Xavier, and Holger M. Mueller. 2017. “Firm Leverage, Consumer Demand, and Employment
Losses During the Great Recession.” Quarterly Journal of Economics 132 (1): 271–316.
Giupponi, Giulia, and Camille Landais. 2020a. “Building Effective Short-Time Work Schemes for the
COVID-19 Crisis.” Vox EU, April 1.
Giupponi, Giulia, and Camille Landais. 2020b. “Subsidizing Labor Hoarding in Recessions: The Employ-
ment and Welfare Effects of Short Time Work.” CEPR Discussion Paper 13310.
Giupponi, Giulia, and Camille Landais. 2020c. “Replication Files for Figures 4 and 9.C: Subsidizing
Labor Hoarding in Recessions: The Employment and Welfare Effects of Short Time Work.” CEPR
Discussion Paper 13310.
Goldschmidt, Deborah, and Johannes F. Schmieder. 2017. “The Rise of Domestic Outsourcing and the
Evolution of the German Wage Structure.” Quarterly Journal of Economics 132 (3): 1165–217.
Gruber, Jonathan. 1997. “The Consumption Smoothing Benefits of Unemployment Insurance.” American
Economic Review 87 (1): 192–205.
Guerrieri, Veronica, Guido Lorenzoni, Ludwig Straub, and Iván Werning. 2020. “Macroeconomic Impli-
cations of COVID-19: Can Negative Supply Shocks Cause Demand Shortages?” NBER Working
Paper 26918.
Haas, Georg-Christoph, Bettina Müller, Christopher Osiander, Julia Schmidtke, Annette Trahms,
Marieke Volkert, and Stefan Zins. 2021. “Development of a New COVID-19 Panel Survey: The IAB
High-Frequency Online Personal Panel (HOPP).” Journal for Labour Market Research 55 (16).
Hendren, Nathaniel. 2017. “Knowledge of Future Job Loss and Implications for Unemployment Insur-
ance.” American Economic Review 107 (7): 1778–823.
Hijzen, Alexander, and Danielle Venn. 2011. “The Role of Short-Time Work Schemes During the 2008-09
Recession.” OECD Social, Employment Migration Working Paper 115.
INPS. 2021. Report Mensile Novembre 2021 – Cassa integrazione guadagni e Disoccupazione. Istituto Nazionale
di Previdenza Sociale, Italy. https://www.inps.it/osservatoristatistici/5 (accessed on December 1,
2021.
INPS. 2021b. “Visitinps Scholar Program Data.” Istituto Nazionale di Previdenza Sociale, Italy. https://
www.inps.it/dati-ricerche-e-bilanci/attivita-di-ricerca/programma-visitinps-scholars.
Jäger, Simon, Benjamin Schoefer, and Josef Zweimüller. 2019. “Marginal Jobs and Job Surplus: A Test of
the Efficiency of Separations.” NBER Working Paper 25492.
Jappelli, Tullio, and Luigi Pistaferri. 2014. “Fiscal Policy and MPC Heterogeneity.” American Economic
Journal: Macroeconomics 6 (4): 107–36.
Kekre, Rohan.Forthcoming. “Unemployment Insurance in Macroeconomic Stabilization.” Review of
Economic Studies.
Khoury, Laura. 2021. “Unemployment Benefits and the Timing of Redundancies.” Unpublished.
Kopp, Daniel, and Michael Siegenthaler. 2021. “Short-Time Work and Unemployment in and After the
Great Recession.” Journal of the European Economic Association 19 (4): 2283–321.
Lalive, Rafael, Camille Landais, and Joseph Zweimüller. 2015. “Market Externalities of Large Unemploy-
ment Insurance Extension Programs.” American Economic Review 105 (12): 3564–96.
Landais, Camille, Pascal Michaillat, and Emmanuel Saez. 2018a. “A Macroeconomic Approach to
Optimal Unemployment Insurance: Applications.” American Economic Journal: Economic Policy 10
(2): 182–216.
Landais, Camille, Pascal Michaillat, and Emmanuel Saez. 2018b. “A Macroeconomic Approach to
Optimal Unemployment Insurance: Theory.” American Economic Journal: Economic Policy 10 (2):
152–81.
Landais, Camille, and Johannes Spinnewijn. 2021. “The Value of Unemployment Insurance.” Review of
Economic Studies 88 (6): 3041–85.
54 Journal of Economic Perspectives
Marinescu, Ioana Elena. 2017. “The General Equilibrium Impacts of Unemployment Insurance: Evidence
from a Large Online Job Board.” Journal of Public Economics 150: 14–29.
Marinescu, Ioana Elena, Daphné Skandalis, and Daniel Zhao. 2020. “Job Search, Job Posting and Unem-
ployment Insurance During the COVID-19 Crisis.” Unpublished.
Marinescu, Ioana Elena, Daphné Skandalis, and Daniel Zhao. 2021. “The Impact of the Federal Pandemic
Unemployment Compensation on Job Search and Vacancy Creation.” Journal of Public Economics
200: 104471.
McKay, Alisdair, and Ricardo Reis. 2016. “The Role of Automatic Stabilizers in the U.S. Business Cycle.”
Econometrica 84 (1): 141–94.
Michaillat, Pascal, and Emmanuel Saez. 2019. “Optimal Public Expenditure with Inefficient Unemploy-
ment.” Review of Economic Studies 86 (3): 1301–31.
Michaillat, Pascal, and Emmanuel Saez. 2021. “Beveridgean Unemployment Gap.” Journal of Public
Economics Plus 2: 100009.
Nomis Official Labour Market Statistics. 2021. “Claimant count by sex and age.” Office for National
Statistics, United Kingdom. https://www.nomisweb.co.uk/query/construct/summary.asp?mode=
construct&version=0&dataset=162 (accessed on November 30, 2021).
OECD. 2021. “Social Benefit Recipients - High Frequency Database (SOCR-HF)” OECD, Paris. https://
www.oecd.org/els/soc/recipients-socr-hf.htm (accessed on November 30, 2021).
OECD. 2021b. “Labour Force Statistics – Short-Term Labour market Statistics.” OECD, Paris. https://
stats.oecd.org/ (accessed on December 29, 2021).
OECD. 2022. Riding the Waves: Adjusting Job Retention Schemes through the COVID-19 Crisis. Paris: OECD
Publishing.
Saez, Emmanuel, Manos Matsaganis, and Panos Tsakloglou. 2012. “Earnings Determination and Taxes:
Evidence from a Cohort-Based Payroll Tax Reform in Greece.” Quarterly Journal of Economics 127
(1): 493–533.
Saez, Emmanuel, Benjamin Schoefer, and David Seim. 2019. “Payroll Taxes, Firm Behavior, and Rent
Sharing: Evidence from a Young Workers’ Tax Cut in Sweden.” American Economic Review 109 (5):
1717–63.
Scarpetta, Stefano, Stéphane Carcillo, and Alexander Hijzen. 2022. Riding the Waves: Adjusting Job
Retention Schemes through the COVID-19 Crisis. Paris: OECD Publishing.
Scarpetta, Stefano, Mark Pearson, Alexander Hijzen, and Andrea Salvatori. 2020. “Job Reten-
tion Schemes During the COVID-19 Lockdown and Beyond.” OECD Policy Responses to
Coronavirus (COVID-19). October 12. https://read.oecd-ilibrary.org/view/?ref=135_135415-
6bardplc5q&title=Job-retention-schemes-during-the-COVID-19-lockdown-and-beyond.
Schmieder, Johannes F., Till von Wachter, and Stefan Bender. 2016. “The Effect of Unemployment
Benefits and Nonemployment Durations on Wages.” American Economic Review 106 (3): 739–77.
Schoefer, Benjamin. 2016. “The Financial Channel of Wage Rigidity.” Society for Economic Dynamics
2016 Meeting Papers 1605.
Scrutinio, Vincenzo. 2018. “The Medium-Term Effects of Unemployment Benefits.” Istituto Nazionale
Previdenza Sociale (INPS) Working Paper 18.
Scrutinio, Vincenzo. 2018b. “Replication Files for Figures 1 and 3: The Medium-Term Effects of Unem-
ployment Benefits.” INPS Working Paper 18.
Sullivan, Daniel G., and Till von Wachter. 2009. “Job Displacement and Mortality: An Analysis Using
Administrative Data.” Quarterly Journal of Economics 124 (3): 1265–306.
Tilly, Jan, and Kilian Niedermayer. 2016. “Employment and Welfare Effects of Short-Time Work.”
Unpublished.
Tilly, Jan, and Kilian Niedermayer. 2016b. “Replication Files for Figure 2.A: Employment and Welfare
Effects of Short-Time Work.” Available upon request (contact: jantilly@gmail.com).
Van Audenrode, Marc A. 1994. “Short-Time Compensation, Job Security, and Employment Contracts:
Evidence from Selected OECD Countries.” Journal of Political Economy 102 (1): 76–102.
Volkert, Marieke, Georg-Christoph Haas, Stefan Zins, Lisa Hellmann, Sandra Dummer, Sophie Hensgen,
Johannes Ludsteck, et al. 2021. “High-frequency Online Personal Panel (IAB-HOPP).” Research
Data Centre of the Federal Employment Agency (BA) at the Institute for Employment Research
(IAB). DOI: 10.5164/IAB.HOPP_W01-W07.de.en.v2
File and source
- File
- should-we-insure-workers-or-jobs-during-recessions.pdf
- Size
- 900,555 bytes
- SHA-256
- 2b0914e8e4113fd34c2c8b03bded838fd50acc1213285048ed83739ed8534916
- Original
- researchonline.lse.ac.uk