Firing Costs Reform Boosted Italian Permanent Hiring

t s
o s
r c
o u
b SUMMARY
L a
Aimed at reducing labour market dualism and favouring job creation, in 2015 the
Italian government issued new regulations lowering firing costs for newly signed
open-ended job contracts. To foster the adoption of the new rules the government
introduced also a generous subsidy for firms hiring workers with an open-ended con-
tract. Using microdata on hiring and firing for one Italian region, Veneto, we exploit
some differences in the design of the two policies to separately identify the effects of
new firing costs on firm hiring. We find that around 8% of gross permanent hires
occurred because of the reduction of firing costs (in addition to the positive and large
effect of the hiring subsidies). The reform of firing costs contributed also to increase
the monthly rate of conversion of fixed-term jobs into permanent positions. As sug-
gested by the theory, we also find that the new firing rules made firms slightly less
reluctant to offer permanent job positions to yet untested workers.
JEL codes: J30, J23, K31

—Paolo Sestito and Eliana Viviano
Economic Policy January 2018 Printed in Great Britain

C CEPR, CESifo, Sciences Po, 2018.
V
Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

by ISTAT user
on 03 August 2018
by ISTAT user
on 03 August 2018
LABOUR MARKETS 103
Firing costs and firm hiring:

evidence from an Italian reform
Paolo Sestito and Eliana Viviano*

Bank of Italy, DG Economics, Statistics and Research
1. INTRODUCTION
Italy lost 1 million jobs over the 2008–14 period as a result of the double dip recession
stemming from the Global Financial Crisis. Because of its dualistic labour market, job
losses were concentrated among younger workers and, more broadly, among people
holding fixed-term job contracts. In 2015, the number of permanent workers has started
to rise again (by around 1% on a yearly basis), despite the still subdued GDP growth (a
bit less than 1%). In this paper, we test whether and to what extent this evolution can be
* We thank Bruno Anastasia, Gianluca Emireni, and Maurizio Gambuzza of Veneto Lavoro for provid-
ing us with the data and with valuable help in interpreting them. We also thank the Visit-INPS staff for
providing us with the aggregate data about Italy, and Samuel Bentolila, Guido de Blasio, Eugenio
Gaiotti, Andrea Ichino, Giovanni Pica, Sebastien Roux, four anonymous referees, and the discussants
and the participants at the 65th Economic Policy panel workshop, 21–22 April 2017. We also thank
the participants to the 11th ECB-CEPR Labour market workshop and to the 28th EALE conference,
Ghent 2016, for helpful comments on a previous version of the paper. We also thank Stefano Marucci
for excellent research assistance. We are solely responsible for any and all errors. The views expressed
herein are ours and do not necessarily reflect those of the Bank of Italy.
The Managing Editor in charge of this paper was Andrea Ichino.
Economic Policy January 2018 pp. 101–130 Printed in Great Britain

C CEPR, CESifo, Sciences Po, 2018.
V

by ISTAT user
on 03 August 2018
104 PAOLO SESTITO AND ELIANA VIVIANO
attributed to the reduction of firing costs for permanent contracts introduced at the
beginning of 2015.
The reform was a reshaping of the regulation about dismissals, aimed at reducing the
level and, even more, the uncertainty of firing costs for all new permanent contracts in
firms with at least 15 employees. This policy was part of a wider reform package known
as the ‘Jobs Act’ and in what follow we label it as FC (for firing costs). The new regula-
tion applied to all new permanent job contracts signed after the inception of the law, in
March 2015. To foster the diffusion of permanent contracts under the new rules and
stimulate job creation, the government introduced also a sizable temporary rebate of
non-wage labour costs which applied to all new permanent job contracts (hereafter hir-
ing subsidy, HS) offered to workers who did not hold an open-ended position in the pre-
vious semester. Besides this, the HS was not targeted to specific groups of workers and it
was not contingent upon firm-level net job creation. Rather, it applied also to conver-
sions within the same firm from a fixed-term to an open-ended position.
We separately identify the effects of the two policies by jointly exploiting their differ-
ent commencement dates (the HS came into effect in January 2015 and the FC in
March 2015) and, more importantly, the differences in the applicability of the two
schemes, as the HS applies to all firms but only to workers without a permanent job con-
tract in the previous semester, while the new FC regulation reshapes firing costs for all
new permanent contracts, but only for firms beyond the 15-employee threshold. We use
administrative microdata for the Veneto region, which allow us not only to measure
labour market flows by firm size (in the first half of 2015 and the 2 years before), but also
to reconstruct the previous status of workers in the labour market, and identify the effect
of FC separately from the effect of HS.
It has to be stressed that there are relevant aspects of the two policies that are not con-
sidered in this paper. First, since the FC reform applies to new hires, and we have data
only on a limited time span, we cannot directly test the effect of the new rules on firings.
We can test, however, the immediate reaction of firms’ hiring rates in response to a
reduction of firing costs, contributing in this way to a quite scant empirical literature.
Second, we do not discuss all the pros and cons of the two policies, like for instance the
impact of FC on job-to-job transitions, nor the general equilibrium effects that can
derive from their implementation (e.g. their effects on labour supply). However, even
with these caveats, our results clearly show that a reduction of firing costs can boost
firms’ hires and shift the employment composition towards permanent job contracts (in
addition to the effect of HS). We also shed some light on the impact of the two policies
on firms’ willingness to hire yet untested workers on a permanent basis. We find that the
reduction of firing costs slightly enhanced the willingness of firms with more than 15
employees to hire workers who never worked for the firm before, that is, not previously
screened by the firm.
Albeit relevant, the FC reform, however, was not the dominant driver of employment
trends in Italy in 2015. At the aggregate level, according to our findings around 20% of
gross permanent hires occurred because of the subsidies, whereas 8% can be attributed

by ISTAT user
on 03 August 2018
LABOUR MARKETS 105
to the new firing regulations. The doubling in the monthly conversion rate observed in
Veneto is entirely due to the two policies, the HS contributing by around 5/6.
When comparing the flows generated by FC and those due to HS one has to keep in
mind that the two measures are very different as HS is a very generous but temporary
labour costs rebate, while the FC policy is a permanent change of the rules governing
dismissals. From a theoretical point of view, in a static model, firing costs are just a tax
on firing that is economically equivalent to a component of labour costs. For a given
wage level, an increase in firing costs implies a downward shift in labour demand and
lower employment. Its relevance, however, may be limited by the fact that firing costs
may occur several years after the time of hiring and need to be discounted.1 In a
dynamic context, higher employment protection dampens employment fluctuations as
firms do not fully adjust labour input to economic shocks. Notice, however, that most of
the positive employment effects of a reduction (increase) in firing costs ought to material-
ize in an economic upswing (downswing) and in 2015 Italy was far from a cyclical
upswing so that our estimated effects, if any, are a lower bound.
As such, our paper adds to (and mostly confirms) the empirical literature focusing on
the labour demand effects of firing costs. Autor et al. (2007), using state-level data and
time variations in employment protection legislation in the United States, find that
higher employment protection reduces employment flows. Adhvaryu et al. (2013) focus
on rural India and rely on supply-side shocks (like rainfalls) and different state-level
employment protection legislation to show that employment protection reduces employ-
ment responses to shocks. In our paper, we mainly focus on a single Italian region,
Veneto. As a consequence, our results are not affected by spurious local trends as in the
above mentioned empirical papers. Our results are in line with Kugler and Pica (2008)
who use matched employer–employee data and identify the effect of firing costs by look-
ing at accessions and separation after an increase in firing costs occurred in Italy in
1990. With respect to their paper, however, we can directly estimate the impact of firing
costs on the propensity of firms to stabilize workers previously employed with a fixed-
term job contract.
Several papers have already analysed how firing costs shape firms’ propensity to
use fixed-term job contracts, not only to facilitate short-term labour adjustment,
but also to screen workers and test the suitability of a job match before offering a
permanent position (e.g., Güell and Petrongolo, 2007; Faccini, 2014). However,
to the best of our knowledge, no paper has so far shown that a reduction in firing
costs increases firms’ propensity to offer a permanent position to workers not pre-
viously screened.
1 The assumption of a constant wage is also unlikely to hold. On the one hand, part of the firing costs
may be borne by the worker, accepting a lower wage in exchange of the insurance provided by the
employment protection rules. On the other hand, firing costs may strengthen workers bargaining
power.

by ISTAT user
on 03 August 2018
Our results are also a contribution to the literature about the effects of a reduction in
both the uncertainty and the expected level of firing costs. Actually, the FC reform
mostly tempered the high uncertainty stemming from the possibility that in the previous
regime judges could decide in favour of the dismissed worker, mandating not only mon-
etary reimbursement, but also the worker’s reinstatement in the firm. The employment
effects of FC we uncover in the data are likely to be driven mostly by their uncertainty
component.
The paper is organized as follows. In Section 2, we briefly describe the FC and HS
policies introduced by the Italian government in 2015. In Section 3, we describe our
datasets. Section 4 deals with our estimation strategies and presents the results of the
empirical analysis. Section 5 concludes.
2. THE ITALIAN LABOUR MARKET AND THE POLICY UNDER SCRUTINY
The Italian labour market is heavily segmented in permanent and fixed-term workers.
The dualism arose since the end of the nineties when the government progressively
introduced different types of fixed-term contracts to increase flexibility in the use of
labour. Higher flexibility, however, was not accompanied by changes in firing costs for
permanent job contracts.2
During the 2000s, the share of fixed-term workers increased rapidly to around 13%.
More than 60% of new hires were fixed-term. Fixed-term contracts were used not only
to face labour demand uncertainty, but also as a cheap screening device before hiring
workers with an open-ended job contract. People employed temporarily suffered the
most for the consequences of the Global Financial Crisis, when firms, facing a sudden
drop in their activity, used all the available margins to adjust labour input.
Before 2015, Italy was characterized by firing costs whose main feature was their
uncertain and (potentially) high amount.3 Dismissals were costless for firms if they were
motivated by a just cause, that is, worker misbehaviour or the firm need to reduce or
reorganize its workforce. However, whenever a worker objected to the dismissal and the
courts deemed the dismissal to be unfair, the costs could be rather high, particularly in
the case of firms with at least 15 employees. Firms with less than 15 employees could
(according to law no. 108/1990) choose between reinstating the worker or paying a pre-
set severance payment (tied to the worker’s seniority and varying from 2.5 to 6 times
their monthly pay). For firms with at least 15 employees, the general rule was the rein-
statement of the worker. Firms could side-step reinstatement by reaching a private
agreement with the worker, at a potentially high cost.
2 See Sestito (2002) and Pirrone and Sestito (2006).

3 We here refer to individual dismissals.

by ISTAT user
on 03 August 2018
LABOUR MARKETS 107
The most critical aspect of this regime was the uncertainty about the timing and con-
tents of the judges’ decisions, which varied greatly.4 The widespread delays in the Italian
civil justice system are also likely to have compounded these inefficiencies as workers
were lacking any income support (and pay) pending the litigation, whereas firms could
end up having to pay high wage arrears if finally the court decided in favour of the
worker.
Most of the analysis made in Italy focused on the local effect of the regime’s disconti-
nuity around the 15-employee threshold. There is evidence that the potentially higher
costs affecting firms with 15 or more employees may have somehow limited the growth
of smaller firms as they approached the 15þ threshold (see Schivardi and Torrini,
2008). The effects, while statistically significant, were however not very economically sig-
nificant, as the discontinuity in the size distribution of firms around the threshold is not
as marked as the sharp discontinuity taking place for instance in France around the 50-
employee threshold, relevant for the rules concerning the role of unions at the firm level
(see e.g., Garicano et al., 2013; Gourio and Roys, 2014). Such a limited impact may
however be due to measurement errors in measuring firm size, as there were ways to cir-
cumvent the threshold (for instance by employing apprentices whose presence in the
firm’s workforce, by law, does not concur to determine the 15-employee threshold).
The reform of 2015, the ‘Jobs Act’ (Decree No. 183/2014)5 was mainly aimed at
reducing the uncertainty about firing costs. The Jobs Act limited the possibility of rein-
statement, allowing it for discriminatory dismissals and for a few specific cases of discipli-
nary dismissals. As a general rule, it established that unfair dismissals must be
compensated by an amount of money strictly predetermined by law and proportional to
job tenure (from a minimum of four times the monthly pay to a maximum of 24 times,
i.e., 2-month pay for every year of seniority). This monetary compensation may be
reduced if the worker agrees to end any pending litigation about the nature of the dis-
missal, and the worker is exempted from paying taxes on the compensation received.
The new rules however only apply to permanent contracts signed after March 7,
2015, when the new law came into force, by firms with more than 15 employees (or by
those firms which reach the 15þ threshold with their new hires). Existing employees, as
well as the newly signed permanent contracts in firms staying below the 15-employee
threshold, remain covered by the previous legal framework.6
All in all, the new regime reduces both the expected firing costs and, most signifi-
cantly, the uncertainty surrounding it for firms above the 15-employee threshold, with
4 See Ichino (1996).

5 The Jobs Act is a wider reform dealing with the strengthening of unemployment benefits and the
launch of a national Agency in charge of active labour market policies. In this paper, we refer only to
the new rules on firing costs, issued in the final days of 2014 and in force since March 2015.
6 The Jobs Act states that for firms with less than 15 employees firing costs cannot exceed 6-months’ pay
with a minimum of 2-months’ pay (one per year of seniority). Before the Jobs Act, firing costs varied
from 2.5 to 6-months’ pay. Differently from Fana et al. (2015) we believe that this change is negligible.

by ISTAT user
on 03 August 2018
no change for those below the threshold.7 It is not easy, however, to quantify the reduc-
tion in uncertainty in monetary terms; even using risk-neutral firms as a benchmark, it
may depend on the judgments issued by local courts and the length of trials. The quanti-
fication of the reduction in firing costs may be further complicated by the presence of
the above mentioned regime offering tax-free treatment of severance pay.8
To foster job creation and the adoption of open-ended contracts under the new FC
rules the government also introduced a very generous non-conditional HS. The subsidy,
established by the Financial Stability Law for 2015 (but already announced at the end of
October 2014) covered all new permanent workers hired by any firm from January to
December 2015, provided the worker did not have a permanent contract in the previous
6 months. The subsidy was a 3-year exemption from social security contributions up to a
threshold, which was quite high compared with the average contributions typically paid
by firms for workers (according to the government’s estimates the subsidy should fully
cover the social security contributions for almost 80% of new hires). Conversions from
fixed-term to permanent job contracts within a given firm were also subsidized (conver-
sions from apprenticeships were instead excluded as they benefit from an ordinary sub-
sidy which was not modified by the HS policy). Last, in October 2015 the government
announced a reduction in hiring subsidies for permanent workers hired from January to
December 2016, and equal to a 2-year exemption for no more than 40% of social secur-
ity contributions.
The two policy measures undertaken in 2015 and analysed in this paper almost over-
lap, because both target permanent hires and job contract conversions from fixed-term
into open-ended contracts. There is, however, a small difference in their timing: from
January 2015 for the subsidy, from March 7, 2015 for the new firing costs. Moreover,
there are some differences in the population targeted by the two policies that can be
used to separately identify their effects. Subsidies were paid to firms of any size, while
the new FC regulation applies to firms with at least 15 employees; the subsidy applied
only to workers without a permanent job contract in the previous 6 months, while the
previous status of the worker is irrelevant for the application of the new FC. Thus, infor-
mation about firm size and workers’ past work histories, together with the precise date
of the new contract allow for the separate identification of their effects.
Many important issues, however, must be taken into account. First, firms’ hiring strat-
egies in 2015 were probably influenced by the announcement of HS in October 2014
(3 months before its implementation) and the announcement of its reduction in October
7 For those approaching the 15-employee threshold there is a reduction in the costs of overcoming the
threshold as they are exempted from the consequences of overpassing it. For this reason, differently
from Schivardi and Torrini (2008), in some estimates exclude firms close to the threshold.
8 This may have two different possible implications: on the one hand, it directly reduces the cost for a
firm of a dismissal deemed unfair by a judge; on the other hand, it may stimulate the worker to chal-
lenge a dismissal that is fair, because the challenge might lead to an agreement somehow more accept-
able for the firm given its reduced costs.

by ISTAT user
on 03 August 2018
LABOUR MARKETS 109
June 15
End of the
sample
Oct 14 Dec 14 Jan 15 Mar 15 Oct 15 Jan 16

HS2015 FC announced HS2015 FC implemented HS2016 HS2016
announced implemented announced implemented
Figure 1. Timeline of the FC and HS policies. Dates of announcement and

implementation.
2015 (for hires taking place in 2016). Similarly, also the reform of FC was announced at
the end of December 2014, with the policy being actually implemented only in March
2015. In Figure 1, we report the timeline of the interventions undertaken in 2015, as
well as the time of their announcement (at the end of 2014). As better explained in
Section 4 our estimates take into account the possible reactions of firms to the announce-
ment of the FC reform and of HS in 2015. We will not consider the potential impact of
the subsidy reduction in 2016, announced in October 2015, as our period on analysis is
until June 2015 (i.e., it ends before the announcement of the HS reduction).
Second, it is important to notice that changes in the evolution over time of permanent
versus temporary hires (across firms’ size classes and workers types) may not help to iden-
tify the effects of the two policies as these may have affected also temporary job con-
tracts. Firms often hire fixed-term workers, test their skills, and then convert their
contracts into open-ended ones. Such a two-step strategy is convenient for firms because
it allows them to verify the goodness of the job match; such a convenience is increasing
in firing costs and in the uncertainty about the job candidate’s quality, the only restrain-
ing factor coming from the risk that the worker, when hired on a temporary basis, is
more likely to search for an alternative job offer. The FC reform might have then
reduced the propensity to adopt such a two-step strategy, as it could have reduced the
reluctance of 15þ firms to offer open-ended positions to unknown workers, defined as
those who were never employed in the firm in the past (with an fixed-term contract).
The HS policy might have had the opposite effect, as firms were enticed to test as many
unknown eligible workers as possible in order to select the ones to be eventually con-
verted into subsidized permanent contracts later on (within December 2015).
3. GENERAL FEATURES OF THE DATA
In this paper, we mainly use microdata about the so-called Comunicazioni Obbligatorie
(CO), collected in the Veneto region. For some robustness checks we also rely on admin-
istrative data, drawn from the Social Security database, managed by the Italian

by ISTAT user
on 03 August 2018
National Social Security Institute (INPS), available only at the aggregate level (see
Appendix).
In Italy, all private firms (and some in the public sector) have to electronically notify
to the local public labour market agency all events concerning hiring, firing, conversions
of contracts from fixed-term to open-ended, and fixed-term contract time extensions.
The CO database is managed by the Italian regions, under the coordination of the
Ministry of Labour. Because of the decentralization of the data-collection process, the
quality of the microdata (especially information concerning contract conversions) differs
across Italian regions. Data issued by the Veneto region are characterized by high
quality.
Veneto is located in the north-eastern part of the country. The weight of the manu-
facturing industry in the region’s economy is among the highest in the country. It repre-
sents slightly less than 10% of the overall Italian labour market (in terms of employees).
Albeit limited to a single region, microdata allow us to take account of compositional
factors and to identify more precisely the effect of the two different policies.
For each event recorded in our dataset we can identify both the firm and the worker
involved and the contract type, that is, whether permanent or fixed term. The latter
group includes: (i) standard fixed-term dependent contracts, (ii) agency workers, (iii)
apprentices,9 (iv) consultants (so-called parasubordinati, i.e., a group of workers with no fir-
ing protection and whose contractual arrangements blend elements of both dependent
and independent employment); and (v) intern (so-called tirocinanti).10
On top of the relevant anonymized identifiers (firm and worker), we know firms’
size11 (by size class), sector of activity, and some workers’ demographic characteristics.
We exclude firms in the agricultural sector and domestic workers hired by households
(for which FC and/or HS did not apply) and focus on industry, construction, and pri-
vate services.12
Since we have microdata on both firms and workers, we construct two datasets, one
for firms and one for workers. Because of limits in data availability the dataset spans
from January 2013 to June 2015 (we also have data on the second semester of 2012 to
determine workers’ eligibility condition at the beginning of 2013). Even if this time span
is quite short, it allows us to avoid two possible confounding events. The first is the intro-
duction, since July 2015, of new limits for the use of consultants, which might have
9 Legally, apprenticeships are classified as open-ended contracts. As there is no firing cost in case the
contract is terminated before or at the end of the apprenticeship period we classify them as fixed
term.
10 We also exclude job transitions in the tourism sector, as the very high seasonality of employment in
this sector cannot be fully captured by the 3-year sample used in this paper.
11 The size of the firm is not directly communicated at the time of hiring/firing/conversion, but can be
ascertained via other compulsory reporting requirements such as those under Law 68/1999 on legally
protected workers.
12 The CO excludes most of the public sector but a few contractual categories referring to temporary
hires.

by ISTAT user
on 03 August 2018
LABOUR MARKETS 111
indirect effects on the use of other types of job contracts. The second, as mentioned in
the previous section, is the government’s announcement of its intention to extend HS to
2016 but for a smaller amount and duration (up to 40% of total social security for
2 years). This announcement might have affected flows in the last part of the year, induc-
ing firms to anticipate at the end of 2015 the hires originally planned for 2016 (and cash
the full HS).
We measure firm size at the end of the previous year, to avoid the potential endoge-
neity of size due to firms’ response to the FC reduction. In almost all the estimates, from
the original sample, we exclude newly born firms (i.e., those which are in the sample at
time t, but have no record before t). This selection strategy is particularly important for
firms entering in the market (and in the dataset) in 2015, as it ensures that the composi-
tion of our sample does not depend on the policy itself. For each month we know how
many workers have been hired or fired (as well as contract conversions), whether the
workers had/had not a permanent job position in the previous 6 months (which deter-
mines the eligibility condition relevant for the HS since January 2015) and firm size
(15þ or less). For simplicity, we label all workers with no permanent job contract in the
previous 6 months as ‘eligible’, even if the observations refer to the years before 2015,
when the HS did not apply.
The final dataset is composed of monthly data for around 120,000 firms which
roughly corresponds to the universe of firms13 whose main branch is located in
Veneto.14
Concerning individuals, we exclude individuals entering into the labour force in 2015
as their entrance is affected by the policies (as made for firms). We then focus only on
individuals with at least one fixed-term job contract from January 2013 to October
2014, the time of the announcement of the HS policy. We also focus on workers aged
no more than 55 years, to exclude those who retire. After this sample selection our sam-
ple includes roughly 185,000 individuals. Notice that this group is quite close to the
average stock of fixed-term workers recorded by the Labour Force Survey in Veneto.
In this paper, we mainly look at two outcomes: (i) gross hiring (and the related individ-
ual probability to find a job) and (ii) net job creation. Net job creation is defined as the
difference between hires and job separations, the latter including not only fires, but also
voluntary job separations and retirement. Net job creation of open-ended contracts
includes also conversions. In the case of job separations, used to calculate net flows, the
13 In Italy, firms undertaking collective layoffs (which involve permanent workers and occur in case of at
least 5 layoffs within 12 months) are forbidden to hire new workers within a certain time period from
the layoff. Unfortunately, our dataset lacks information to precisely identify collective layoffs. Their
inclusion in our sample would then bias our estimates downwards.
14 To draw this conclusion we have compared our dataset with external sources like the INPS database,
available at the firm level, and records all firms with employees by region. After the same sample
selection used for the Comunicazioni Obbligatorie, we end up with a discrepancy of less than 1% with
respect to our sample probably due to the use of slightly different sectoral classification criteria in the
CO database.

by ISTAT user
on 03 August 2018
eligibility status is set equal to the one observed at the time the worker is separated from
the job.
Last, since our estimation strategy is mainly based on a diff-in-diff model, we rely on
aggregate INPS data for the period between January 2006 and December 2015. This
longer time dimension is used to check the validity of the parallel trend assumption (see
Appendix).
Table 1 reports the average monthly flows by firm size, type of contract in 2015, and
the previous period. On average, gross monthly flows of permanent workers are quite
modest as they amount to 0.1 worker (around 1 permanent worker per year) in small
firms and 1 per year in larger firms. Net flows of open-ended contracts are instead close
to zero, for both small and large firms, especially before the implementation of the poli-
cies. Flows of temporary workers are instead larger. Interestingly, gross and net flows of
permanent workers increase after the implementation of the policies, whereas the net
creation of temporary jobs declines.
Table 2 refers to conversions from a fixed term to an open-ended position. The
monthly probability of a contract conversion is around 1% before the inception of the
policies. Afterwards it increases by around 1% among eligible workers.
Figures 2 and 3 allow us to appreciate the evolution of open-ended contracts and con-
versions over time. First, Figure 2 shows that before 2015 the flows of permanent
Table 1. Sample characteristics
Average 2013–4 2015
Flows of permanent contractsa

Gross hiring
Firms <15 0.008 0.013
Firms 15þ 0.047 0.082
Flows of permanent contractsa
Net hiringb
Firms <15 0.0001 0.004
Firms 15þ 0.002 0.026
Flows of fixed-term contractsa
Gross hiring
Firms <15 0.021 0.023
Firms 15þ 0.250 0.312
Flows of fixed-term contractsa
Net hiringc
Firms <15 0.000 0.005
Firms 15þ 0.0003 0.058
Share of firms in the sample
Firms <15 (%) 8.6 8.5
Firms 15 þ (%) 91.4 91.5
Note: Flows by contract type and firm size (monthly averages).

a
Average number of contracts per month.
b
Net job creation of permanent contracts is equal to the sum of jobs created plus conversions minus job
separations.
c
Net job creation of fixed-term job contracts is equal to the sum of jobs created minus conversions minus job
separations.
Sources: Authors’ calculations on Veneto CO data.

by ISTAT user
on 03 August 2018
LABOUR MARKETS 113
Table 2. Sample characteristics
Average 2013–4 (%) 2015 (%)
Probability of conversions by firm size

Firms <15 0.9 1.8
Firms 15þ 0.9 1.8
Probability of conversions by eligibility status of the worker
Non-eligiblea 0.2 0.1
Eligibleb 1.1 2.4
Note: Conversion from a temporary to a permanent job position by firm size and eligibility status of the worker
(monthly averages).
a
With a permanent job contract in the previous 6 months, or apprentices.
b
Without a permanent job contract in the previous 6 months.
1500
1000
500
0
-500
13
14
im )
.)
.)
n.
S n.
pl
pl
n-
n-
an
(H an
im
Ja
Ja
15 C
S
C
H
(F
n- 4(
4(
5
Ja -1
-1
-1
ec
ct
ar
O
time
Firms <15 Firms 15+
Figure 2. Veneto – net job creation of permanent contracts by firm size from
January 2013 to June 2015. Vertical lines indicate the time of the announcement
and the implementation of the FC and HS policies.
Notes: Sample of firms already operating before the implementation of the policies. Net job creation of permanent
contracts is equal to the sum of jobs created plus conversions minus job destructed. Data are regressed on month
dummies, interacted by firm size, type of contract, and eligibility. The figure plots the residuals plus the uncondi-
tional average.
Source: Authors’ calculations on Veneto CO data.
contracts were around zero in both small and 15þ firms. In November and December
2014, at the time of the announcement of the HS reform they declined sharply in both
groups of firms, and increased again at the time of the implementation of the HS policy.
This evidence confirms that it is important to take into account the strategic behav-
iour of firms at the announcement of the policies, as they probably changed their hiring
schedule to benefit from the policies. In the next section, we discuss how we take account
of these timing issues in our estimates.

by ISTAT user
on 03 August 2018
.025
.02
.015
.01
.005
0
3
im )
.)
.)
n.
S n.
1
pl
pl
n-
n-
an
(H an
im
Ja
Ja
15 FC
S
C
H
(F
n- 4(
4(
5
Ja -1
-1
-1
ec
ct
ar
O
M
time
Firm <15 Firm 15+
Figure 3. Veneto – individual probability of conversion from a temporary into a

permanent job position by firm size from January 2013 to June 2015. Vertical
lines indicate the time of the announcement and the implementation of the FC
and HS policies.
Notes: Sample of individuals with a temporary job contract before October 2014, that is, before the announcement
of the policies. To partially control the effect of seasonality, data are regressed on month dummies. The figure
plots the residuals plus the unconditional average probability.
Source: Authors’ calculations on Veneto CO data.
Figure 2 reports the monthly probability that a fixed-term contract is converted into
an open-ended one. Once again the figure confirms a sharp increase after January
2015, for both 15þ and smaller firms. All in all, the charts presented in this section
show that there are discontinuities in the hiring trends along some of the dimensions
affected by the two policies here considered.
4. EVIDENCE FROM MICRODATA
In this section, we carry out several exercises to estimate the effects of the reduction of
FC from the perspective of both firms and workers, while controlling for the impact of
HS. First, we focus on permanent contracts and we identify the effect of FC by compar-
ing the change in monthly hiring in small versus large firms, before and after the FC
reform. Second, we look at contract conversions from fixed-term to open-ended by firm
size and we estimate the separate effects of FC and HS, by comparing the conversion
rate of eligible versus non-eligible workers in firms of different size classes. The third
exercise is a simple decomposition of aggregate flows which allows us to get a rough esti-
mate of the impact of the two policies on flows, in a setting which controls for firms’
unobserved heterogeneity.
To take into account the possible anticipation effects described in Section 2, in all the
exercises the time of the relevant policy change (the beginning of the so-called ‘post’

by ISTAT user
on 03 August 2018
LABOUR MARKETS 115
period) is the announcement of the policy, that is, October 2014 for HS and December
2014 for FC (see Figure 1). In this way, the estimated effects are net of any postpone-
ment of firms’ hiring behaviour, because they are the average of the reduction of flows
at the time of the announcement of the policies and the subsequent increase after their
implementation. A separate analysis at the end of this section provides an assessment of
these temporary substitution effects.
4.1. Firm hiring after the reform of FC
The identification of the effects of the FC reform may be obtained by comparing open-
ended hires and net job creation of small and large firms before and after the FC
reform. Independently on the effect of HS, the reform of firing costs should increase the
propensity of firms with at least 15 employees to offer open-ended positions to workers
by more than what observed in firms below the threshold. (We cannot carry out analysis
at the 15-employee threshold because in our dataset firms’ size is collected by size
classes.)
As already mentioned, we measure firms’ size at the end of the previous year. In
such a way we avoid the bias due to the possible manipulation of the treatment thresh-
old induced by the policy itself. Importantly, we exclude firms between 10 and 14
employees as they are more likely to be affected by the Jobs Act, which allows them to
maintain lower firing costs even in the case they reach the 15þ employee threshold. We
estimate:
niym ¼ ci þ cy þ cm þ c1 Xiðy1Þ þ d1 D FC ðsize i¼15þ y;m Dec 14Þ þ iym ; (1)
where niym is the flow of firm i, in year y, and month m, Xiðy1Þ is firms’ size at the end
of year y1, ci , cy , cm , are, respectively, firm-fixed effects and year and month dummies,
D FC ðsize i¼15þ; ym Dec 14Þ is a dummy for observations in firms 15þ after the announce-
ment of the reform of firing costs (December 2014) and iym is an error term. The coeffi-
cient d1 measures the effect of the reform of FC on permanent employment, identified
with respect to firms with less than 15 employees. In estimating Equation (1) we include
also the local unemployment rate, interacted with firm size to capture differences in
business cycle among different types of firms (in any).
The first two columns of Table 3 refer to gross and net open-ended hires, respectively,
and do not include firm-fixed effects. Columns 3 and 4 replicate the analysis with fixed
effects. We find that the FC reform has a positive, and highly significant effect on both
gross and net hires in all the specifications. Since the coefficient d1 represents the abso-
lute increase in monthly flows after the inception of the FC reform it is straightforward
to calculate its aggregate effect of FC: it increases gross hires by roughly 1 worker per
year (0.085 * 12 months) and net hires by around 0.2 workers (0.016 * 12), that is,
roughly 1 worker per year every five 15þ firms.

by ISTAT user
on 03 August 2018
Table 3. Firms 151 versus others (excluding firms 10–14): gross and net hiring
of permanent workersa,b
Gross hiring Net hiring Gross hiring Net hiring Ratio known workersc
d1 (effect of FC) 0.102 0.032 0.085 0.016 0.028

[0.000]*** [0.000]*** [0.000]*** [0.005]*** [0.001]***
Fixed effects No No Yes Yes Yes
Observationsd 3,448,116 3,448,116 3,448,116 3,448,116 118,563
Firms 117,298 117,298 117,298 117,298 51,517
Average-dependent 0.0505 0.0034 0.0505 0.0034 0.558
variable
a
Robust standard errors; p-values in brackets. *, significant at 10%; **, significant at 5%; ***, significant at 1%.
b
All models include a ‘post’ dummy, month dummies, and the province-level unemployment rate interacted with
firm class-size.
c
Sample of firms with positive gross hiring. Share of workers who had past work experiences within the firm in
total hires of permanent workers. Conversions excluded.
d
Firm * number of months in the sample.
In the last column, we carry out a different exercise. As mentioned in Section 2

one prediction of the theory is that lower firing costs may reduce the reluctance of
firms to offer a permanent job position to workers who were not screened by the firm
in the past, for example, by the use of a temporary job contract. Thus, lower FC
should increase the propensity of firms to hire directly unknown workers, because it
could be easier to dismiss them in case of a bad match. We then select the pool of
firms which record a permanent job contract in a given month and we compute the
share of known workers in the open-ended gross hires taking place in that month.
The effect of the FC policy is captured by a term interacting a dummy for the intro-
duction of the new FC regulation and the firm’s class size. As predicted by the theory
we find a negative sign. While statistically significant, the effect is however rather
small. On average the share of known workers in total open-ended hires decreased
by around 3 p.p., on an average equal to 56%.
To control for the fact that larger firms have naturally larger flows, in Table 4 we
consider the probability of hiring (i.e., to have at least one permanent contract hire
in a given month) and the probability to increase the size (positive net hire; Columns
1 and 2). In Columns 3 and 4, we report the results of regressions where the FC effect
is split by detailed firms’ size class. We find that the probability of both gross and net
hiring increases, confirming the relevance of firing costs for firms’ flows and the rela-
tionship between FC and firm (in-)action found for instance by Kugler and Pica
(2008). The effect is smaller close to the 15þ threshold; it is rather uniform for the
other size classes.
4.2. Probability of a conversion from fixed-term to open ended
We use the individual-level panel dimension and we analyse the probability that a
fixed-term worker has a contract conversion within the same firm (conditional on

by ISTAT user
on 03 August 2018
LABOUR MARKETS 117
Table 4. Firms 151 versus others (excluding firms 10–14): probability of positive
gross hiring and positive net hiring of permanent workersa,b
Gross hiring Net hiring Gross hiring Net hiring
d1 (effect of FC) 0.031 0.023

[0.000]*** [0.000]***
d1 (effect of FC) * Size 15–19 0.016 0.013
[0.000]*** [0.000]***
[0.000]*** [0.000]***
[0.000]*** [0.000]***
[0.000]*** [0.000]***
d1 (effect of FC) * Size 100þ 0.047 0.013
[0.000]*** [0.000]***
Fixed effects Yes Yes Yes Yes
Observationsc 3,448,116 3,448,116 3,448,116 3,448,116
Firms 117,298 117,298 117,298 117,298
Average-dependent variable (%) 3.43 3.05 3.43 3.05
Note: Effect by detailed firm size.

a
b
Linear probability models. All models include a ‘post’ dummy, month dummies, and the province-level unem-
ployment rate interacted with firm class-size.
c
Firm * number of months in the sample.
being in the labour market in 2013 or in 2014). The effect of FC is identified by the
size of the firm in which the contract conversion takes place. In this setting, the prob-
ability of conversion can be affected by HS. Therefore, in some regressions we con-
trol for the eligibility status of workers, determined by her past working history. Our
sample then includes all temporary workers ‘at risk’ of finding a permanent position,
that is, both standard temporary workers and apprentices, not subject to HS in case
of conversion. In its more general formulation, we define a dummy variable p equal
to 1 if the fixed-term position is converted into an open-ended position and we
estimate:
pweym ¼ cw þ ce þ cy þ cm þ d1 D FC ðsize w works¼15þÞðy; m Dec 14Þ

þ b1 D HS ðe¼1Þðy; m Oct 2014Þ þ weym ; (2)
where cw are worker-fixed effects, aimed at capturing workers’ unobserved heterogeneity,

ce , cy , and cm are fixed effects for workers not having a permanent job in the previous
semester (e), for year (y), and month (m). The dummy D FC ðsize w works¼15þÞðy; m Dec 2014Þ
is equal to 1 if the worker w is in a 15þ firm and observations refer to the period following
the announcement of the FC reform. Year and month dummies account for time trends.
Conversions in small firms after December 2014 represent the control group used to iden-
tify the effect of the FC. The variable D HS ðe¼1Þðy; m Oct 2014Þ is a dummy equal to 1 if
the observation refers to the period from October 2014 onwards and involves workers

by ISTAT user
on 03 August 2018
eligible for the HS. The comparison group is composed of workers who had a permanent
job positions in the previous 6 months and workers with an apprenticeship job contract,
whose contract conversion cannot receive the HS.
We also include the local-level unemployment rate interacted with firm size to cap-
ture group-specific business cycle effects. Moreover, since we have a panel of individuals
and the variable pweym typically does not change after a person has found a permanent
job, we drop the worker from the sample after (s)he finds a permanent job (unless after
some time she re-enters the pool of job seekers because of a job destruction).
To strengthen identification and eliminate possible source of remaining endogeneity,
in some estimates we focus upon those who had a fixed term job contract in September
of 2014, that is, before the announcement of the first of the two policies and we look at
the probability that their contract is converted into a permanent one in the period
October 2014–June 2015. In this case, the comparison group is composed of fixed-term
workers in September 2013.
Table 5 displays the results. Columns 1 and 2 refer to the entire period. The first
includes only the effect of FC, Column 2 controls also for the eligibility status. Columns
3 and 4 refer to the subsample of individuals employed in September of each year with
a fixed-term job contract. Confirming the results for firms, the reduction in firing costs
increased the probability for worker with a temporary position to have a contract con-
version. The increase is 0.4 p.p., which is a large impact, as the average probability is
equal to 1%. The effect of FC slightly diminishes if we include HS, whose impact is very
high, as it doubles the probability of a contract conversion. If we select our sample to
temporary workers already employed before the announcement of the policies, the sign
and the size of the coefficients remain unchanged.
Table 5. Veneto – probability of contract conversiona,b
Full sample Being employed in September 2013 or 2014,

with no open-ended job position
d1 (effect of FC) 0.0042 0.0028 0.0041 0.0024

[0.000]*** [0.000]*** [0.000]*** [0.000]***
b1 (effect of HS) 0.015 0.0177
[0.000]*** [0.000]***
Individuals 185,544 185,544 140,413 140,413
Observationsc 1,525,787 1,525,787 1,067,641 1,067,641
Average-dependent 1.21 1.21 1.37 1.37
variable (%)
a
Linear probability models. Robust standard errors; p-values in brackets. *, significant at 10%; **, significant at
5%; ***, significant at 1%. Workers who find an open-ended position at time t are excluded from the sample
from t þ 1 onwards.
b
firm class-size.
c
Individuals * number of months in the sample.

by ISTAT user
on 03 August 2018
LABOUR MARKETS 119
4.3. Decomposing aggregate flows
After the estimation of the effect of the FC reform on firm hiring we assess the impact of
both FC and HS on the observed total flows, by decomposing them according to firms’
and workers’ characteristics. It is important to stress, however, that any comparison
between the contribution of HS and FC must consider that it is extremely difficult to
impute a monetary value to the reduction of FC, and that HS are only a temporary 3-
year measure, whose impact of firms’ labour costs depends on the expected duration of
job matches made under the HS.
The possibility to identify the effect of the two policies stems mainly on the differences in
their design (HS apply to workers not employed with a permanent job contract in the pre-
vious 6 months; FC apply to 15þ firms) and the timing of their implementation. Moreover,
we cannot simply compare the trends of hires by type of contract, because the policies may
affect the convenience of firms to hire also on a temporary basis. This may occur, for
instance, because of the two-step strategy ‘hiring as temporary and then convert’ described
in Section 2. In order to identify a flow not affected by the two policies, to be used as a sort
of ‘control’ group, in this section we focus on events in small firms (as such unaffected by
changes in FC) involving workers who were not eligible for the HS (i.e., workers with a per-
manent position in the previous semester, or apprentices). Job-to-job transitions of non-
affected workers in small firms are then used as a proxy of what would happen to flows as a
consequence of the business cycle, as more buoyant economic conditions may favour work-
ers attempting to pursue their idiosyncratic preferences with respect to the firm where they
work.
We further split firm flows: the flow of those finding a permanent job position and the
flow of those who find a fixed-term job; each flow is further decomposed according to
the eligibility of workers involved. Thus, for each firm in our sample we have four cells,
given by the intersection of (i) the type of contract c (open-ended vs. fixed term) and (ii)
whether the workers involved are ‘eligible’, e. For each firm we also know whether it is
15þ or smaller.
Since this is a simple decomposition of aggregate flows, here we do not drop firms
born in 2015, and we estimate the impact of HS also on newly born firms. We end up
with around 130,000 firms per month and since we have observations for 30 months
our estimates are based on more than 15,000,000 cells. We then carry out the following
decomposition:
niecym ¼ ci þ ce þ cc þ cy þ cm þ
þ d1 D FC ðsize i ¼ 15þÞðc ¼ pÞðy; m Dec 14Þ
þd2 D FC ðsize i ¼ 15þÞðc ¼ f Þðy; m Dec 14Þþb1 DHS ðe ¼ 1Þðc ¼ pÞðy; m Oct 14Þ þb2 D HS ðe ¼ 1Þðc ¼ f Þðy; m Oct 14Þ þ iecym;
(3)
where niecym is the monthly flow, the indices i, e, c, y, and m indicate, respectively, firm i,
whether the flow is composed of eligible workers (e ¼ 1), c is the type of contract (p for

by ISTAT user
on 03 August 2018
permanent, f for fixed term), and y and m are year and month. Together with firm fixed
effects ci , the terms ce and cc are a fixed effects, respectively, for eligibility, and the type
of contract; cy and cm are time dummies.
The dummy D FC ðsize i¼15þÞðc¼pÞðy; m Dec 14Þ is a dummy equal to 1 if the cell refers to
permanent hiring in a 15þ firm occurred after the announcement of the Jobs Act
(independently of eligibility). The variable D HS ðe¼1Þðc¼pÞðy; m Post Oct 14Þ is a
dummy equal to 1 if the cell corresponds to permanent hires of eligible workers
occurred from October 2014 on (and 0 otherwise). D HS ðe¼1Þðc¼f Þðy; m Oct 14Þ and
D FC ðsize i¼15þÞðc¼f Þðy; m Dec 14Þ differ from D HS ðe¼1Þðc¼pÞðy; m Post Oct 14Þ and
D FC ðsize i¼15þÞðc¼pÞðy; m Dec 14Þ as they refer to fixed-term contracts. So, in Equation (3)
b1 and d1 identify the direct potential impact of HS and FC, respectively, on open-
ended contracts, while the terms b2 and d2 capture substitution or complementarities
induced by the policies on other types of contracts. Note that with this specification the
total effect of the two policies is the sum of both bs and ds. Last, even if not reported in
Equation (3), in all the specifications we include all the interactions between the local
unemployment rate and the type of contract, to capture the impact of the business cycle
on the different types of contracts. Since flows in the same firm are very likely to be cor-
related, standard errors are clustered at the level of firm.
The results of the estimation of Equation (3) are in Table 6. Column 1 deals with
gross hires and Column 2 with net hires. The effects of both HS and FC are positive
and highly significant. Table 6 reports also the coefficients of the impact of the policies
on fixed-term contracts. The effect of both FC and HS is positive for gross hires and
negative for net hires, signaling substitution.
To determine the size of the estimated coefficients it is important to notice that, since
we are dealing with a cell-level decomposition of monthly flows, the coefficients reported
in the table represent the (monthly) increase in the cells affected by HS or FC. We then
calculate how much of the increase in gross and net flows is due to either FC or HS by
multiplying the estimated effect by the number of cells affected by each policy. Consider
first gross hires. According to our model, in 2015 each month in Veneto there were
around 700 open-ended hires occurred because of FC and 2,000 because of HS. Since
the estimated flow of open-ended contracts is equal to 9,900, the policies, respectively,
account for 8% and 20% of the gross flow of permanent workers. These shares amount,
respectively, to 2% and 44% in the case of net open-ended hires.
To evaluate whether the two policies contributed to increase employment is impor-
tant to relate their direct impact to the increase in total net flow, which includes also
fixed term job contracts (results in the last row of Table 6).
According to our estimates in 2015 total net job creation increased by more than
1,500 job positions per month. We find that the FC policy contributed by 7%; the con-
tribution of the HS policy was even larger than the observed increase in the total flow
(144%), because of the substitution of fixed-term job contracts with open-ended ones.

by ISTAT user
on 03 August 2018
LABOUR MARKETS 121
Table 6. Veneto – firm-level decomposition of flows and the effects of the policies
on both open-ended and fixed-term workers: gross and net hiringa,b,c
Gross hiring Net hiring
d1 (effect of FC on open-ended) 0.034 0.005

[0.000]*** [0.090]*
b1 (effect of HS on open-ended) 0.015 0.017
[0.000]*** [0.000]***
d2 (effect of FC on fixed-term) 0.07 0.029
[0.000]*** [0.000]***
b2 (effect of HS on fixed-term) 0.002 0.009
[0.177] [0.004]***
Fixed effects Yes Yes
Firms 128,395 128,395
Observations 15,407,400 15,407,400
Number of firms each month
Number of cells FC 21,894 21,894
Number of cells HS 128,395 128,395
Estim. increase in the flow of open-ended contracts
FC (d1 * number of cells affected by FC)e 744 109
HS (b1 * number of cells affected by HS)f 1,926 2,183
Estim. increase as share of flow of open-ended contracts
FC (d1 * number of cells affected by FC) (%) 8 2
HS (b1 * number of cells affected by HS) (%) 20 44
Estim. increase as share of the increase in total flow
FC (d1 * number of cells affected by FC) (%) 14 7
HS (b1 * number of cells affected by HS) (%) 35 144
a
The data are organized in firm-level cells, obtained by the intersection of the following criteria: eligibility status
of workers involved in the flow, contract type (open-ended vs. fixed term).
b
c
type of contract.
d
Firms * four cells per firm *number of months in the sample.
e
Number of cells subject to FC each month.
f
Number of cells subject to HS each month.
Our estimates imply that in the first semester of 2015 both policies increased the number
of people employed with a permanent job contract by 0.7%.
Last, from the weight of HS in the flow of gross hires one can assess the deadweight
loss of the policy itself, that is, the amount of open-ended gross hires which were
financed by the HS but would have happened even without the subsidy. Taking into
account that approximately 70% of permanent hires involve workers eligible for the
HS, three out of five of the hires of eligible workers would have happened even in the
absence of the HS and so, the effective per capita cost of subsidized open-ended contracts
has to be multiplied by approximately 2.5 (each subsidy was worth around 6,500 euro
per year according to INPS data).

by ISTAT user
on 03 August 2018
4.4. Timing effects and robustness checks
In the previous sections, we analyse changes in flows occurred after the announce-
ment of the two polices. Our estimates are average effects, net of the temporary sub-
stitution effects related to firms’ strategic choice of postponing some hires to
benefit from the HS and FC policies. Here, we carry out some exercises aimed at
quantifying the timing effects of FC (a similar strategy can be used for the timing
effects of HS).
As shown in Figure 1 firms may have reduced permanent hiring from October 2014 to
January 2015, waiting for the implementation of the HS. On top of that, large firms may
have further reduced permanent hiring in December 2014, at the announcement of the FC
reform. We then use the same specification in Equation (1) and data until December 2014
to get an estimate of the postponement effect due to the announcement of the FC reform,
additional to the one caused by the announcement of HS. In this exercise dates before
December 2014 are the ‘pre’ period, and December 2014 is the ‘post’ period. For estimat-
ing the effect of postponement, in principle we could also use data on flows occurred in
January and February 2015. However, as HS was implemented in January 2015, the inclu-
sion of these months may affect the estimation of the ‘timing effect’ of the FC only. Our
results (see Table 7) confirm the existence of a postponement effect, which is quite small for
gross hires and larger for net hires (suggesting that firms preferred to stop normal turnover).
According to our estimates firms reduced net permanent hiring by around one out of seven
workers from December 2014 to February 2015 and started hiring again after March 2015
(the total effect of this behaviour is the one described in Section 4.1 and Table 3).
We have carried out several robustness checks. First, it is still possible that flows of
firms operating in the same sector are correlated because of unobserved sector-level
shocks. To check for this possibility we have carried out some estimates with standard
errors clustered by firm and sector (45 sectors). The results remain unchanged. Second,
Table 7. Firms 151 versus others: postponement effects due to the announce-
ment of the FC reform on gross and net hiring of permanent workers and proba-
bilities of positive flows (data until December 2014)a,b
Gross hiring Net hiring Prob. Gross hiringc Prob. Net hiringc
d1 (effect of FC) 0.009 0.141 0.012 0.016

[0.252] [0.000]*** [0.000]*** [0.000]***
Observationsc 2,744,328 2,744,328 2,744,328 2,744,328
Firms 117,298 117,298 117,298 117,298
Average-dependent 0.046 0.0001 0.0317 0.0281
variable
a
b
firm class-size.
c
Linear probability models.
d
Firms * number of months in the sample.

by ISTAT user
on 03 August 2018
LABOUR MARKETS 123
a large fraction of gross monthly hires is concentrated at zero. Using the same specifica-
tion as in Equation (1) and a tobit model, we test whether our main results are affected
by the choice of a simple linear model. Our results are robust also with respect to this
further check (available upon request).
5. CONCLUSIONS
In this paper, using a diff-in-diff approach, we analyse the reaction of firms to the reform
carried out in March 2015 by the Italian government and aimed at reducing firing costs
for firms with at least 15 employees (not only a monetary reduction, but also a decrease in
the uncertainty about the consequences of unfair dismissals). Our econometric exercises
are intended to uncover the direct effect of a firing cost reduction on firms’ hiring, in addi-
tion to the effect of another important policy undertaken by the government almost at the
same time, that is a generous HS to firms offering open-ended job contracts.
We find that the two policies were successful in reducing dualism and stimulating
labour demand, even during a recession period characterized by very high macroeco-
nomic uncertainty. Even if it cannot be disputed that the most important driver of the
recovery of employment in Italy was the HS, our results show that the reform of FC con-
tributed to increase firms’ hiring rates.
Another relevant result is that the FC reform impacted upon the willingness of firms
to hire, on a permanent basis, yet unscreened workers, slightly reducing the reliance
upon a widespread two-step strategy through which firms first hire on a temporary basis
and then eventually convert the contract into a permanent one.
Our estimates do not consider all the aspects relevant in judging the appropriateness
of the two measures [see Brown et al. (2011) for a wider theoretical discussion on hiring
subsidies]. In particular, we do not discuss the pros and cons of the current temporary
and rather unselective hiring subsidies vis-à-vis permanent subsidies targeting specific
groups of workers supposed to be weaker and less employable (e.g., youths and long-term
job seekers). Furthermore, we do not deal with the merits and pitfalls of subsidies for gross
hiring which subsidizes also part of the normal turnover taking place in a firm. Our esti-
mates also fall short of an overall evaluation of the new firing rules introduced by the
Jobs Act. As a matter of fact, the Jobs Act does not modify the general principle that only
dismissals opposed by the worker and considered unfair by a judge must be compensated
by firms. Differently from the previous regime, the reinstatement of the worker, still possi-
ble, applies only to few and better specified cases (the ones deemed to be discriminatory).
Also, the uncertainty concerning the amount of the financial compensation possibly stem-
ming from a judiciary intervention has been considerably lowered as its amount has been
capped and pre-specified by the law, being now an increasing function of worker’s senior-
ity. It is uncontroversial that the new regime will provide more certainty to workers and
firms. Since it applies only to new contracts, however, it could impact negatively on job-
to-job transitions as workers might prefer to do not change employment to avoid to lose
the higher protection ensured by the previous regime. This effect, which could negatively

by ISTAT user
on 03 August 2018
affect workers’ reallocation, cannot be evaluated at this stage, because of the lack of data
on a sufficiently long ‘post-reform’ period. Our estimates do not allow to consider all
these effects, whose relevance is likely to increase over time as the stock of employees will
be increasingly made up by people hired according to the new rules.
Furthermore, our estimates do not consider the overall general equilibrium implica-
tions. Albeit small, the effects of the FC reform might have significant general equili-
brium effects, for instance if its allocative effects cumulate to other market imperfections
(e.g., capital market imperfections) in shifting the whole firm size distribution. We how-
ever believe that our empirical exercises are a step forward for the comprehension of the
reaction of firms to changes in employment protection, and the impact of a reduction in
FC on firms’ hiring behaviour.
Discussion
Uwe Sunde
University of Munich
This paper provides a very timely evaluation of a reform of the Italian labour market in
the aftermath of the Great Recession, during which more than a million jobs were lost
over the period 2008–14. The overall aim of the reform was to promote (permanent)
employment. One part of the reform was the Jobs Act, enacted in March 2015. It consti-
tuted the main labour market policy with the aim to relax the FC legislation for firms
with 15þ employees and introduce codified severance pay to increase transparency and
reduce uncertainty regarding the size of severance pay. A second reform (denoted HS)
was the introduction of an unconditional hiring subsidy that was paid after January
2015 to convert the contracts of workers with non-permanent contracts to permanent
and that would cover a 3-year exemption from social security contributions.
The paper mainly uses microdata from administrative records in the Veneto region. In
addition, administrative data from the Social Security database (managed by the Italian
National Social Security Institute, INPS) are used. The identification strategy uses a differ-
ence-in-difference strategy that exploits the different dates at which the policies were
announced, in interaction with the applicability of the policies to different types of firms
and workers, respectively. In a first step, the analysis investigates the effect of the firing cost
reform by comparing hiring in small and large firms. In a second step, the effect of hiring
subsidies is investigated by comparing conversions from fixed to permanent positions of
eligible and non-eligible workers. In a third step, a simple decomposition analysis of aggre-
gate flows gives insights about the overall effects of the reform on job flows.
The empirical findings show that the firing cost reform led to a significant increase in
both gross and net hiring rates and reduced the ratio of hired workers who were

by ISTAT user
on 03 August 2018
LABOUR MARKETS 125
previously known to the firm. These effects are economically significant, with gross hires
increasing by about 1 worker per year for firms with 15 or more employees, and net hires
increasing by around 0.2 workers. Moreover, the effect is larger for larger firms. When
considering the effect on the probability of contract conversion from fixed term to perma-
nent, the reduction in firing cost as well as the implementation of hiring subsidies had sig-
nificant positive effects. When considering gross hires, the effects of both reforms are
positive and significant. The aggregate flow analysis reveals that the policies account for
substantial shares (8% and 20%, respectively) of the gross hires with permanent contracts.
The paper thereby provides relevant information about the effectiveness of two recent
reforms of the Italian labour market. A particularly interesting insight refers to the effects
of a policy (the firing cost reform) that mainly reduced uncertainty of firms about the
actual costs of labour adjustment, on the hiring margin. This contribution notwithstand-
ing, several issues are noteworthy.
The identification strategy uses the announcement dates of the reforms as defining
criterion for the treatment period, instead of the implementation dates. This by itself is
somewhat unusual and justified by the anticipation of firms and the close timing of the
two reforms. In fact, the HS reform was announced at the end of October 2014 and
implemented in January 2015, whereas the FC reform was announced at the end of
December 2014 and implemented on 7 March 2015. A cursory look at the raw data of
net job creation, shown in Figure 2, reveals that net job creation went up sharply in early
2015 for firms with <15 employees and firms with >15 employees, but the increase for
the larger firms came somewhat earlier, between December 2014 and January 2015,
whereas the increase came about 1 month later for the smaller firms that were not
affected by the firing cost legislation. Likewise, the raw data on conversion rates shown
in Figure 3 show a change in December 2014 and January 2015, with larger firms react-
ing somewhat slower. While from a theoretical point of view one might expect the
announcement of a policy to affect decisions, justifying the identification strategy, it is
nevertheless not quite clear why, in this particular case, firms should implement employ-
ment changes before the respective policies (in particular hiring subsidies) are in place.
The discussion in Section 4.4 refers to this point, but mainly focuses on strategic post-
ponement of adjustments in hiring. However, looking at Figures 2 and 3, one gets the
impression that most of the action occurs between December 2014 and January 2015,
around the announcement of the FC reform and the implementation of the HS reform.
This makes it difficult to infer much from the raw data, but it also puts a question on the
identification using announcement dates. An interesting next step could therefore be to
investigate the effectiveness by systematically varying the treatment period indicator
between announcement date and implementation date, or the estimation of a more flex-
ible model that allows disentangling the precise timing of the effects.
The raw data shown in Figure 2 also reveal that net job creation appeared to fall after
March 2015 for both types of firms. This might have to do with the fact that both poli-
cies were initially planned to be effective for 2015 only. The present analysis relies on
data that end in June 2015. Potentially, the decline in the raw data is an artefact of the

by ISTAT user
on 03 August 2018
graph. In any case, a longer panel of data that can be used to also evaluate the effect of
the withdrawal of the policies would clearly be desirable. This provides an interesting
avenue for future research.
From a conceptual point of view, both policies target different dimensions of hiring.
Moreover, due to the eligibility criteria, some firms are affected by none of the policies,
others by one, yet others by both. Hence, one is tempted to think about the total effect
of both policies and whether there are any complementarities. An earlier version of the
paper (presented in the panel) included empirical specifications of the gross and net hir-
ing regressions with an interaction term between HS and FC policy and delivered a sig-
nificant and positive coefficient for the interaction term, but a weaker effect of the FC
reform. This suggests that the results presented in the final version might be a conserva-
tive estimate, and also conceal a considerable amount of heterogeneity (which, in some
sense, shows up in the results for different firm sizes in Table 4). On a related but more
technical note, this also raises the question about censoring. In a setting with firm-level
observations on a monthly frequency, there is a lot of censoring with zero hiring or con-
version, particularly for the smaller firms, as can be inferred from the descriptives shown
in Table 1. This might explain part of the heterogeneity in the effects. The authors
report that the main results are robust to the estimation of Tobit models that account
for this censoring, but the quantitative differences are not reported. Moreover, the inclu-
sion of province-level unemployment rates (multiplied by firm size bracket) might call
for inference using clustered standard errors. Future work might be needed to address
these issues and investigate their relevance for the findings.
In sum, this is an interesting study of a very recent reform of the Italian labour mar-
ket. The results suggest that the reforms achieved their goals of stimulating hiring and
the conversion of temporary contracts to permanent contracts. Interestingly, the quanti-
tative results of costs for the creation of an open-ended contract for 1 year of approxi-
mately 16,500 Euro are in the range of results of what other studies find for the effect of
fiscal stimulus packages on employment.15
Panel discussion
Roberto Galbiati asked how much is the average severance payment before the shock
compared with the expected severance payment after, and whether the authors did
some analysis on firm size. Giacomo Calzolari inquired if the reduction in firing costs
was translated into more (or less) firing. On this point, Andrea Ichino clarified that only
workers hired after 2015 were subject to the reduction in firing costs reform.
15 For instance, Buchheim and Watzinger (2017) report costs of 25,000 Euros for creating one job for 1
year when evaluating the effects of a fiscal stimulus focused on infrastructure investments.

by ISTAT user
on 03 August 2018
LABOUR MARKETS 127
Tommaso Monacelli questioned why one should compare the two policies given that
hiring subsidies are temporary, while a change in regulation to reduce firing costs has
permanent nature. He also suggested examining how many jobs would be created if we
could project the results on a national scale. On the role of political economy, Moritz
Schularick highlighted that there is a limited amount of political capital that govern-
ments can use to push policies through and hiring subsidies are easier for politicians to
implement.
Nicola Fuchs-Schündeln recommended exploiting the dynamics of the effects by ana-
lysing the period between the announcement and implementation of the policies.
Finally, Andrea Ichino argued that the impact of the hiring subsidies is what one would
expect and suggested to focus on the firing costs reform.
Replying to comments and questions, E.V. first acknowledged that the paper indeed
covers too many aspects that need to be described more clearly. She also clarified that
they deal with anticipation effects (e.g., by looking at the time of the announcement),
and that their estimates are not driven by the spike observed in December 2015, i.e.,
before the reduction in hiring subsidies. E.V. also stated that job transitions were so far
used in the analysis as driven by individuals’ preferences to change jobs, and clarified
that they have no reliable data on severance payments. Finally, she observed that their
estimates suggest that around half of the 700,000 permanent jobs created in 2015 were
due to the combination of the two policies net of any anticipation effects.
APPENDIX: EVIDENCE FROM AGGREGATE DATA: ITALY 2006–15
In this Appendix, we carry out some additional analysis based on aggregate data
from the INPS database. The aim of this Appendix is two-fold. First, we show general
trends observed in Italy in 2015. Second, thanks to the availability of a longer time
series, we propose some robustness checks for our identification strategy.
The INPS data cover the period between January 2006 and December 2015.
Unfortunately, from the INPS database we have only aggregate flows for four groups
of workers/firms, identified by the intersection of the following characteristics: (1)
whether the workers is eligible for the HS or not; (2) whether the firm was 15þ or
smaller. We the use INPS data to analyse overall trends in monthly hiring and firing,
for each month between January 2006 and December 2015.
Figure A1 reports net job creation in Italy from January 2006 to December 2015
and the composition of newly created jobs during this period, measured by the share
of new hires with a permanent job contract in total new (gross) hires. Two main facts
emerge clearly. First, the share of open-ended contracts in total contracts declines
from 2009 to 2014, and starts increasing again at the beginning of 2015. It peaks at
the end of 2015, when firms have probably anticipated hiring originally planned for
2016 to benefit for the generous HS for permanent workers, expiring at the end of
December 2015. Second, in 2015 the total number of people employed with any type
of contract expands considerably. The peak observed at the end of 2015 indirectly

by ISTAT user
on 03 August 2018
.7
200
150
.6
% open-ended
Total net flow
100
.5
50
.4
0
-50
.3
-05 an-06 an-07 an-08 an-09 an-10 an-11 an-12 an-13 an-14 5
ar1 ec-1
5
Jan J J J J J J J J J -M D
Jan
Total net flow

% open-ended
Figure A1. Italy; total net job creation of open-ended contracts (left-hand axis,
thousand) and share of open-ended contracts in total new contracts (right-hand axis).
Source: Authors’ calculations on INPS data, private sector excluding agriculture. To partially control the effect of
seasonality, raw data have been organized in cells, obtained by the intersection of firm size, and eligibility of work-
ers. Then, raw data are regressed on month dummies, interacted by firm size and eligibility. The figure plots the
residuals plus the unconditional average.
confirms that our estimates based on microdata until June 2015 have the advantage
of being unaffected by the need of firms to anticipate hiring in December 2015. It is
very likely, actually, that a firm trying to anticipate in 2015 hires originally planned
for 2016 would find it more convenient to hire in December 2015 than in June 2015.
As in Section 4.3 we use flows of non-eligible workers in small firms as a ‘control’
group for other treated groups. More in detail, we consider only two types of flows:
(1) the one composed of non-eligible workers in small firms (non-affected by the poli-
cies, labelled as NA); and (2) all the other flows [affected (A), i.e., composed of eligible
workers and/or flows generated by 15þ firms]. We label the two groups as
g ¼ NA; A, respectively, and we look at differences in trends before and after the
implementation of the two policies. Within this simple diff-in-diff framework, this
identification strategy allows us to uncover the total effect of FC and HS, and it is
based on the assumption of parallel trends before the inception of the policies. We
test this hypothesis with INPS data because they have a longer time dimension than
CO data on Veneto (INPS data are monthly data spanning from 2006 to 2015).
The first column of Table A1 reports the coefficients corresponding to a dummy
equal to 0 when the net flow corresponds to the control group and 1 otherwise. This
dummy is further interacted with year dummies from 2011 onwards, corresponding
to the bust of the sovereign debt crisis (for simplicity, the period 2006–10 is the omit-
ted category; regressions with dummies for each year give similar results). The results
are clear-cut: no significant difference in trends can be found before 2015, being the

by ISTAT user
on 03 August 2018
LABOUR MARKETS 129
Table A1. Italy – net job creation for treated and controlsa
Whole periodb Period May-June of each yearc
d1 & b1 (effect of FC and HS) * 2011 5,197 9,251

[0.419] [0.496]
d1 & b1 (effect of FC and HS) * 2012 1,640 5,680
[0.860] [0.781]
d1 & b1 (effect of FC and HS) * 2013 919 4,990
[0.874] [0.748]
d1 & b1 (effect of FC and HS) * 2014 172 9,940
[0.984] [0.258]
d1 & b1 (effect of FC and HS) * 2015 48,461*** 42,424***
Observations 240 40
Average monthly flow 11,127 75,309
Note: Difference in trends before and after the inception of the two 2015 policies.
a
b
Cell-level data: two groups (affected by the policies vs. non-affected); 2 month; 10 years.
Sources: Authors’ calculations on INPS aggregate data.
difference very small compared with the average size of the monthly flow (last row of
the table). Instead, in 2015, the flow affected by the policies increases considerably.
The results of the second column refer only to May and June of each year to avoid
that the results are driven by (i) the large spike observed in December 2015 and/or
(ii) the implementation of other parts of the Jobs Act reform (since July 2015), (iii) the
possible anticipation effects of the two policies in October 2014 and December 2014
(and hiring postponement at the date of the implementation of the two policies in
January and March 2015).16 The results are fully consistent with the ones of the first
column. Before 2015, the difference was small and not significantly different from
zero, while the opposite holds true in 2015.
These longer time series can be used also for another exercise, which allows us to
control for possible autocorrelation in diff-in-diff estimates. We calculate, for each
month, the difference between the flows of treated and non-treated workers and we
check whether it significantly increased because of the two policies. To ensure that
our results are not driven by the very high spike observed in December 2015, we limit
our analysis to the period January 2006–June 2015. Further, to control for possible
timing effects since the announcement of the first policy in October 2014, we look at
changes in the difference since that date, as in the main estimates of the paper. The
estimated monthly variation is then the average effect of the policies in the period
October 2014–September 2015. In doing so, we apply also Newey–West estimator,
to adjust estimates for possible auto-correlation in residuals.
16 Other time intervals give similar results; they are available upon request.

by ISTAT user
on 03 August 2018
Table A2. Italy – difference between the flows of treated and non-treated individ-
uals (respectively, eligible workers and workers in firms 151, and non-eligible
workers in small firms)a
Gross hiring Net hiring

OLS Newey OLS Newey
d1 & b1 (effect of FC and HS) 49,787 49,787 29,962 29,962

[0.000]*** [0.001]*** [0.001]*** [0.092]*
Observations 114 114 114 114
Average difference in flowsb 296,240 296,240 83,921 83,921
Note: The effect of autocorrelation.

a
Data from January 2006 to June 2015. Robust standard errors; p-values in brackets. *, significant at 10%; **, sig-
nificant at 5%; ***, significant at 1%.
b
For a comparison of the estimated effect net of seasonality, the row reports the average flow in the first semester
of each year.
Sources: Authors’ calculations on INPS aggregate data.
The results of this exercise are in Table A2. The first column reports the OLS esti-
mate. The second is based on the Newey–West estimator. The results confirm the
existence of differences in trends during the period under scrutiny and, more impor-
tantly, that autocorrelation has a small impact on the precision of our estimates.
REFERENCES
Adhvaryu, A., A.V. Chari and S. Sharma (2013). ‘Firing costs and flexibility: evidence from
firms’ employment responses to shocks in India’, The Review of Economics and Statistics, 95,
725–40.
Autor, D., W. Kerr and A. Kugler (2007). ‘Does employment protection reduce productivity?
Evidence from US states’, The Economic Journal, 117, 189–217.
Brown, A.J.G., C. Merkl and D.J. Snower (2011). ‘Comparing the effectiveness of employment
subsidies’, Labour Economics, 18, 168–79.
Buchheim, L. and M. Watzinger (2017). ‘The employment effects of countercyclical infrastruc-
ture investments’, Munich Economics Discussion Paper 2017–6.
Faccini, R. (2014). ‘Reassessing labour market reforms: temporary contracts as a screening
device’, The Economic Journal, 124, 167–200.
Fana, M., D. Guarascio and V. Cirillo (2015). ‘Labour market reforms in Italy: evaluating the
effects of the Jobs Act’, Working Paper No. 5/2015. Available at http://www.isigrowth.eu/
2015/12/08/labour-market-reforms-in-italy-evaluating-the-effects-of-the-jobs-act/.
Garicano, L., C. Lelarge and J. Van Reenen (2013). ‘Firm size distortions and the productivity
distribution: Evidence from France’, NBER Working Paper No. 18841.
Gourio, F. and N. Roys (2014). ‘Size-dependent regulations, firm size distribution, and realloca-
tion’, Quantitative Economics, 5, 377–416.
Güell, M. and B. Petrongolo (2007). ‘How binding are legal limits? Transitions from temporary
to permanent work in Spain’, Labour Economics, 14, 153–83.
Ichino, P. (1996). Il lavoro e il mercato. Mondadori.
Kugler, A. and G. Pica (2008). ‘Effects of employment protection on worker and job flows: evi-
dence from the 1990 Italian reform’, Labour Economics, 5, 78–95.
Pirrone, S. and P. Sestito (2006). Disoccupati in Italia. il Mulino.
Schivardi, F. and R. Torrini (2008). ‘Identifying the effects of firing restrictions through
size-contingent differences in regulation’, Labour Economics, 15, 482–511.
Sestito, P. (2002). Il mercato del lavoro in Italia. Laterza.

by ISTAT user
on 03 August 2018

Firing Costs Reform Boosted Italian Permanent Hiring

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Firing Costs Reform Boosted Italian Permanent Hiring

Uploaded by

Copyright:

Available Formats

t s

JEL codes: J30, J23, K31

Economic Policy January 2018 Printed in Great Britain

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Firing costs and firm hiring:

Paolo Sestito and Eliana Viviano*

Economic Policy January 2018 pp. 101–130 Printed in Great Britain

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

2. THE ITALIAN LABOUR MARKET AND THE POLICY UNDER SCRUTINY

2 See Sestito (2002) and Pirrone and Sestito (2006).

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

4 See Ichino (1996).

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Oct 14 Dec 14 Jan 15 Mar 15 Oct 15 Jan 16

Figure 1. Timeline of the FC and HS policies. Dates of announcement and

3. GENERAL FEATURES OF THE DATA

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Table 1. Sample characteristics

Average 2013–4 2015

Flows of permanent contractsa

Note: Flows by contract type and ﬁrm size (monthly averages).

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Table 2. Sample characteristics

Average 2013–4 (%) 2015 (%)

Probability of conversions by ﬁrm size

Firms <15 Firms 15+

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Firm <15 Firm 15+

Figure 3. Veneto – individual probability of conversion from a temporary into a

4. EVIDENCE FROM MICRODATA

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

4.1. Firm hiring after the reform of FC

niym ¼ ci þ cy þ cm þ c1 Xiðy1Þ þ d1 D FC ðsize i¼15þ y;m Dec 14Þ þ iym ; (1)

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

d1 (effect of FC) 0.102 0.032 0.085 0.016 0.028

In the last column, we carry out a different exercise. As mentioned in Section 2

4.2. Probability of a conversion from fixed-term to open ended

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Gross hiring Net hiring Gross hiring Net hiring

d1 (effect of FC) 0.031 0.023

Note: Effect by detailed ﬁrm size.

pweym ¼ cw þ ce þ cy þ cm þ d1 D FC ðsize w works¼15þÞðy; m Dec 14Þ

where cw are worker-fixed effects, aimed at capturing workers’ unobserved heterogeneity,

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Table 5. Veneto – probability of contract conversiona,b

Full sample Being employed in September 2013 or 2014,

d1 (effect of FC) 0.0042 0.0028 0.0041 0.0024

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

4.3. Decomposing aggregate flows

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

Gross hiring Net hiring

d1 (effect of FC on open-ended) 0.034 0.005

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995

4.4. Timing effects and robustness checks

d1 (effect of FC) 0.009 0.141 0.012 0.016

Downloaded from https://academic.oup.com/economicpolicy/article-abstract/33/93/101/4833995