Empirical Analysis of the Severance Pay
Non-Performance in Slovenia
Milan Vodopivec
Lilijana Madjar
Primož Dolenc
Combining information from the Firm Survey of Labor Costs with the
information about claims filed with the Guarantee Fund by workers
whose employers defaulted on their severance pay obligations, the pa-
per analyzes the so-called non-performance problem of severance pay
- the fact that coverage, and thus legal entitlement, does not guarantee
the actual receipt of the benefit - as experienced in Slovenia in 2000.
The findings are threefold: (i) one-third of total obligations incurred by
firms failed to be honored and only a small portion of defaulted sever-
ance pay claims was reimbursed by the Guarantee Fund; (ii) while both
men and women seem to be equally affected, workers older than 40
were disproportionally represented among those whose severance pay
claims failed to be honored; and, (iii) among firms that incurred sever-
ance pay liabilities, larger and more productive firms were more likely
to observe their fiduciary obligations and pay them out. These findings
corroborate the weaknesses of severance pay as an income protection
program, pointing to the large scale of the non-performance problem
and the inequities created by it.
Key Words: severance pay, severance pay non-performance,
Guarantee Fund, Slovenia
jEL Classification: J65, J32
Introduction
Being widely used in both the developed and developing world, sever-
ance pay is the most prevalent program offering income compensation
in the case of job loss.1 Despite being so widespread, evaluations show
that severance pay not only creates important inefficiencies but also often
fails to provide adequate protection. On the efficiency front, severance
Dr Milan Vodopivec is a Lead Economist at the World Bank, usa.
Lilijana Madjar is Director of the Public Guarantee Fund of Slovenia.
Dr Primož Dolenc is an Assistant Professor at the Faculty
of Management Koper, University of Primorska, Slovenia.
Managing Global Transitions 7 (4): 333-348
pay reduces employment and labor market flows, hinders technologi-
cal progress and innovations, pushes workers into the informal sector,
and creates significant litigation costs (see Addison and Teixeira (2001)
for a review of both theoretical and empirical effects). Its scorecard on
the income protection front is also rather negative. First, generous sev-
erance pay hinders access to formal sector jobs by disadvantaged groups
(oecd 1999). Second, the same amount of severance is paid regardless
of the duration of the unemployment spell following the separation, re-
sulting in over-payments to workers with short and under-payments to
workers with long unemployment spells. And third, severance pay suf-
fers from the so-called non-performance problem - the fact that cover-
age, and thus legal entitlement, does not guarantee the actual receipt of
the benefit.
The non-performance of severance pay is largely an 'uncharted terri-
tory,' as only a handful of studies provide hard empirical evidence about
this aspect of severance pay. Because severance is not administered by a
public authority, information about the incidence of severance pay obli-
gations as well as about how frequently firms actually honor such obli-
gations is rarely accessible. While ad-hoc evidence exists (for example,
from litigation cases where workers are suing their employers for the
non-payment of severance pay), we are familiar with only two studies
that report evidence on non-performance-based on micro-data. One is
MacIsaac and Rama (2000), who estimate that in the early 1990s only
about half of Peruvian workers legally entitled to severance pay received
the benefit (MacIsaac and Rama report that the payment was more likely
if workers had a written contract and if they worked in a large, unionized
firm that paid social security contributions). The other is Mansor et al.
(2001), who report that Malaysian workers who were laid off in 1998 re-
ceived 83 percent of the total amount of severance pay that they claimed
from their employers.
This paper is an attempt to provide further insights into the non-
performance problem of the severance pay. It focuses on the Slovenian
severance pay program and addresses the following questions:
•	How severe is the non-performance problem in Slovenia? That is, of
total liabilities arising from the payment of severance, what are (i)
the share of severance pay paid out by firms, (ii) the share paid out
by the Guarantee Fund of Slovenia, and (iii) the share that failed to
be paid out?
•	In particular, are any groups strongly affected? That is, what is the
age and sex composition of workers whose severance pay claims are
not honored?
• What are the characteristics of the firms that honor their obligations
arising from severance pay in comparison to those that fail to do so?
Because of information availability, Slovenia is particularly suitable for
studying the above questions. First, in 2000 the Statistical Office of Slove-
nia carried out a survey of labor costs incurred by firms, the survey that
among others provides information about the amount of severance paid
out by each firm. Second, Slovenia is one of few transition countries that
has introduced the Guarantee Fund to help the laid-off workers with par-
tial reimbursement of their outstanding severance pay claims, and the
information collected by this fund is a valuable source for studying sev-
erance pay non-performance.2
These two sources - the Firm Survey of Labor Costs and the Guaran-
tee Fund - are thus complementary, one providing information about
severance pay obligations paid out by firms and the other about obliga-
tions that failed to be paid out by firms. Taken together, the data establish
a composite rendering of fulfilled and unfulfilled severance pay obliga-
tions in Slovenia.
The main findings of the paper are as follows. First, the nonperfor-
mance of severance pay has been a significant problem in Slovenia, with
one-third of total obligations incurred by firms failing to be honored
(in 2000, the year focused upon by the study) and only a small portion
of non-paid severance pay claims being reimbursed by the Guarantee
Fund. Second, while both men and women seem to be equally affected,
workers older than 40 were disproportionally represented among those
whose severance pay claims failed to be honored. And third, among firms
that incurred severance pay liabilities, larger and more productive firms
were more likely to pay them out.
These findings corroborate the weaknesses of severance pay as an
income protection program, pointing to the large scale of the non-
performance problem and inequities created by it.
The paper is organized as follows. The second section describes the
legal framework of severance pay in Slovenia. The third section provides
a comparison of severance pay programs in transition countries. The
fourth describes the data and methodology, and the fifth section presents
the results of the empirical analysis of the non-performance problem of
the Slovenian severance pay program. The last section concludes with a
summary and policy implications.
Description of the Legal System of Slovenia's Severance Pay
In Slovenia, severance pay is regulated by the Labor Code, the Law on
the Public Guarantee Fund, and the Law on Bankruptcy and Liquida-
tion, and is further guided by Collective bargaining agreements, as well
as individual contracts (on the managerial level). Mandated severance
pay is paid to laid-off workers and workers who retire, with the level of
pay proportional to the work tenure of the worker with his or her former
employer. To address the non-performance problem of severance pay, a
Guarantee Fund was introduced in 1997, with the Fund partly reimburs-
ing the unpaid severance pay claims of workers.
Slovenia introduced a Labor Code in 1990 and a new one in 2003. The
1990 Labor Code mandated severance for early retirees as well as for re-
dundant workers. While for early retirees the law did not prescribe the
amount of severance pay, it did so for redundant workers. For each year
of service, workers with at least two years of service were entitled to half
of their monthly average wage for every year of service, with the wage
determined on the basis of the wage paid in the last three months of em-
ployment. Other cases for severance pay were not legally binding.
The 2003 Labor Code significantly differs from the previous one, by
defining more precisely the obligations on the part of employers and the
rights of workers. Workers are entitled to severance pay if they retire or
they are dismissed (either because of business reasons or bankruptcy or
even in the case of his/her incompetence). Retired workers are entitled to
the severance pay of two average wages, calculated from three-months'
average wage in Slovenia, or (if more favorable to the employee) two av-
erage wages, calculated from his/her three-months' average wage before
retirement. In contrast, the basis for the calculation of the severance pay
for dismissed workers is the average monthly wage which was received
by the employee, or which would have been received by the worker if
working, in the last three months before the termination is taken. The
employee is entitled to severance pay amounting to:
•	1/5 of the basis for each year of employment with the employer, if
the employee has been employed with the employer for more than
one and up to five years;
•	1/4 of the basis for each year of employment with the employer, if
the employee has been employed with the employer for period from
five to fifteen years;
•	1/3 of the basis for each year of employment with the employer, if
the employee has been employed with the employer for period ex-
ceeding fifteen years.
It is worthwhile to stress that also under the 2003 law, the severance pay
program remains unrelated to the unemployment insurance program.
That is, qualifying workers receive severance pay and, in addition, they
also qualify for unemployment insurance benefits (which can be received
for up to two years, see van Ours and Vodopivec 2006).
To protect worker's rights in the case of a firm's insolvency, in 1997
Slovenia - following the 1980 eu directive 80/987 - introduced the Pub-
lic Guarantee Fund. Workers, legally entitled to severance pay but un-
successful in its exaction, can claim partial reimbursement of their sev-
erance pay claims from the Fund, with the ceiling on such reimburse-
ments being a monthly minimum wage.3 Moreover, under the 1993 Law
on Bankruptcy and Liquidation, workers can sue their former employers
that undergo a liquidation or bankruptcy process, with workers' sever-
ance pay claims having a priority over other claims (up to a limit - for
details, see Kresal Šoltes 1997).
Review of Severance Pay Program in Transition Countries
In putting Slovenia in an international context, we draw heavily on the
Schwab (2003) analysis of 21 transition countries.4 While all these coun-
tries mandate severance pay, the countries differ in important details.
These include the extent of coverage, eligibility conditions, generosity of
benefits and whether benefits should vary with seniority, and what to
do when bankruptcy prevents the employer from making severance pay-
ments.
ELIGIBILITY
Transition countries mandate severance pay for economic dismissals
such as the employer's liquidation, bankruptcy, or reduction of staff due
to economic, technological, structural, or similar changes.5 Many coun-
tries require severance pay only for economic dismissals. These countries
include the Czech Republic, Georgia, Hungary, Macedonia, Poland, Ser-
bia, Slovak Republic, Slovenia, and Vietnam. In some other countries,
though, workers are also entitled to severance pay for a variety of other
dismissals. These other dismissals are generally for individual reasons,
such as when the worker proves incompetent for the position or is dis-
abled by health reasons.
Not all dismissed workers, even among those dismissed for economic
reasons, are entitled to severance payments. Countries differ in eligibility
conditions. One-third of the countries included in our analysis require a
minimum length of employment with the firm before a dismissed worker
is entitled to severance pay. The required seniority ranges from one to
three years for economic dismissals, and up to five years for other dis-
missals. Slovenia and Vietnam require one year of employment before a
worker is entitled to severance pay. Bosnia and Herzegovina, Croatia, and
Macedonia require two years of seniority. Hungary requires three years
of seniority. Bulgaria requires five year of seniority before a worker is en-
titled to severance pay for dismissals due to illness (but has no seniority
requirements for economic dismissals).
LEVEL OF BENEFITS
Of the 21 countries included in our analysis, 13 use a sliding scale con-
nected to years of employment - Bulgaria, China, Croatia, Estonia, Hun-
gary, Latvia, Lithuania, Macedonia, North Korea, Poland, Serbia and
Montenegro, Slovenia, and Vietnam; severance pay in the rest of the
countries included in the study does not vary with seniority. By design,
in sliding-scale countries more senior workers are entitled to more gen-
erous severance pay. In general, the level of benefits in sliding-scale coun-
tries exceeds those in fixed-benefit countries.
DEALING WITH THE NON-PERFORMANCE PROBLEM
A major issue connected with severance pay is the inability of insol-
vent employers to make severance payments. Fifteen countries have rat-
ified ilo Convention 173, including four transition countries: Latvia,
Lithuania, Slovakia, and Slovenia. According to this convention, coun-
tries can choose between giving priority to severance pay claims in the
employer's bankruptcy proceedings or creating a Guarantee Fund to pro-
tect severance-pay claims (together with unpaid wages) - with Slovenia,
as mentioned above, opting for both.
Data and Methodology
Below we describe the micro-level data sources and methodology used
in the empirical analysis of Slovenia's severance pay.
DATA SOURCES
The following data sources are used:
•	Firm-level data were obtained from the 2000 Labor Costs Survey
in Slovenia, administered by the Statistical Office of Slovenia. The
sample comprised 3,021 enterprises, selected among those with 10 or
more workers. Information included severance pay that firms paid
both to laid-off and retired workers.
•	Individual-level data were gathered from worker requests to the
Guarantee Fund of Slovenia in the period from 1994 to 2003. For
each individual, data included unpaid severance pay obligations,
requested amount from the Fund, amount paid by the Fund, the
gender and age of the applicant, and previous employer.
•	Firm-level measure of efficiency produced by production function
estimation. We used the value of the error term - Sj - for 2000,
obtained by the following oLs estimation of translog production
function for the Slovenian manufacturing firms for the 1994-2001
period:
n	1 n n
Inq1)t = a0 + ^ aklnxljkt + - ^ ßkilnx^lnx^ + s1)t, (1)
k=1	k=1 1=1
where the inputs xijkt include measures of labor, capital, and ma-
terial inputs; ak and ßkl are, respectively, first- and second-order
translog production parameters (i refers to individual firms, j to
two-digit industry categories, and t to time) - see Orazem and
Vodopivec (2008) for details of estimation and data sources used.6
Methodology for the analysis of firm-level determinants of severance
pay payout. To investigate whether firm efficiency and size affect the like-
lihood of paying severance pay given that firms incurred such costs, that
is, that they laid-off workers, we ran a multinomial logit model with the
following options for the dependent variable:
•	firm did not incur severance obligations (taken as a baseline),
•	firm incurred severance obligations and paid them, and
•	firm incurred severance obligations and did not pay them.
As explanatory variables, we used efficiency of the firm and firm
size. To capture firm size effects, we used a dummy variable indicating
whether a firm had more than 100 workers.
Empirical Results
This section presents the results of our empirical analysis of the sever-
ance pay non-performance in Slovenia. As explained above, we focus on
0-9
10-29
30-99
100-249
250+
2%
5%
39%
30%
figure 1 Paid severance pay - structure by size, 2000 (computations based on the
2000 Labor Cost Survey)
Agriculture I 0.82%
Mining 1.87%
Manufacturing
63.29%
Electricity 11.38%
Construction | 2.35%
Trade	13.67%
Hotels 2.80%
Transport 11.81%
Finance 11.77%
Real estate	8.43%
Public admin. | 0.21%
Education I 0.56%
Health 0.81%
Other services | 0.22%
figure 2 Paid severance pay - structure by industry, 2000 (computations based on
the 2000 Labor Cost Survey)
the following aspects: the severity of the non-performance problem, the
composition of workers whose severance pay claims fail to be paid out,
and the characteristics of firms that fail to pay their severance pay obli-
gations.
To put these questions in the context, however, let us first present
statistics about severance pay liabilities that were paid out. We focus on
year 2000, the year for which we have Labor Cost Survey data. First, the
overall amount of severance pay liabilities paid out in 2000 was € 17.5
million - 0.085 percent of gdp or 0.2 percent of the total wage bill. While
this is a rather modest amount, it certainly is not a negligible one. Sec-
ond, most of severance pay obligations was paid by large firms; for ex-
ample, 93 percent of severance pay was paid out by firms with more than
30 workers (see figure 1). Third, the majority (more than 60 percent) of
paid severance pay obligations was incurred in manufacturing (figure 2),
suggesting that in 2000 this sector was still undergoing an intense re-
structuring.
table 1 Severance pay payments, reimbursements, and unpaid claims, 2000
	(1)	(2)	(3)	(4)
Payments made by firms	17.5	64.3	0.085	0.162
Reimbursements made by the Guarantee Fund	0.7	2.7	0.004	0.007
Unpaid severance pay claims	9.0	33.0	0.044	0.083
Total	27.3	100.0	0.132	0.252
notes (1) amount (BE million), (2) structure (%), (3) share in gdp (%), (4) share in
worker compensation (%). Computations based on 2000 Labor Costs Survey and the
Guarantee Fund of Slovenia.
SEVERITY OF SEVERANCE PAY NON-PERFORMANCE
Our results show that in 2000, the non-performance of severance pay
posed a serious problem in Slovenia. Out of the total of € 27.3 million
severance pay obligations, € 9.0 million - 33 percent - failed to be hon-
ored (by firms that incurred these obligations or by the Guarantee Fund
- see table 1). The role of the Guarantee Fund in helping with unpaid
obligations proved to be very limited, as the Fund only reimbursed € 0.7
million or 7.2 percent of total unpaid severance pay obligations.7 Indeed,
according to its rules (see above), the Guarantee Fund reimbursed un-
paid severance pay claims only partially, and so less than 10 percent of
the average claim was actually reimbursed (see figure 3).
In the 1990s, the magnitude of the non-performance of severance pay
was most likely even larger. Figure 4 shows the number of cases of sever-
ance pay reimbursements by the Guarantee Fund in the period from 1994
until 2003. This figure reflects the pattern of transition and suggests that
the problem of severance pay non-performance was probably even more
serious in the mid-1990s, when the number of workers turning to the
Guarantee Fund for reimbursements was more than double the number
in 2000. Moreover, in the period from 1994 to 2003 around 43,000 work-
ers failed to receive payment from their former employers despite their
legal entitlement.
WHICH WORKERS ARE AFFECTED BY NON-PERFORMANCE
OF SEVERANCE PAY?
To find out whether some groups of workers were disproportionally af-
fected by severance pay non-performance, we analyzed the composition
of workers reporting unpaid severance claims. We found no evidence of
differences between men and women, but workers above 40 years were
more affected by severance pay non-performance than younger workers.
342 Milan Vodopivec, Lilijana Madjar, and Primož Dolenc
i 290.50
i 285.16
1994
1995
1996
1997
1998
1999
2000
1940.31
2322.60
255.29
2972.76
240.14
2251.11
240.05
257.25
274.83
285.50
3664.83
3475.45
3583.69
3858.38
302.62
2872.17
2003
301.94
2879.94
figure 3 Average severance pay claims and reimbursements, 1994-2003 (BE, light gray
- average reimbursments, dark gray - average claims; computations based
on information provided by the Guarantee Fund of Slovenia)
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
6555
7738
9019
5725
7912
4688
3125
2989
figure 4 Number of cases of reimbursements, 1994-2003 (based on information
provided by the Guarantee Fund of Slovenia)
Figure 5 shows that severance pay non-performance has not affected
men and women differently, as the reimbursements to men and women
were rather similar in size. Over the 1994-2003 period, the Guarantee
Fund paid 50.8 percent of total severance pay reimbursements to women
and 49.2 percent to men), which correspond well to the employment
shares of these groups (in 2000, men represented 51 percent of total em-
ployment and women 49 percent).
200
figure 5 Structure of severance pay reimbursements by gender, 2000 (bE, light gray -
men, dark gray - women; based on information provided by the Guarantee
Fund of Slovenia)
Turning to the age distribution of claimants, figure 6 compares the age
distribution of severance pay claimants with the age distribution of the
active population in Slovenia in the same period (1994-2003). Clearly,
among the claimants, workers over 40 years of age are over-represented
- while their share in the population is 47 percent and 42 percent for men
and women, respectively, their share among the claimants is 64 percent
and 51 percent for men and women, respectively.
firm characteristics and non-performance
The last part of our analysis sheds light on characteristics of firms that
are paying out severance pay in comparison to those that are failing to
honor their legal entitlements. The estimation of the multinominal logit
model (see table 2) shows that the larger and more productive the firm,
the more likely it is to honor its severance pay obligations.
Concluding Remarks
Being one of the rare examples of its kind, the paper seeks to provide
insights into the non-performance problem of severance pay by analyz-
Under 20
20 to 24
25 to 29
30 to 34
35 to 39
40 to 44
45 to 49
50 to 54
55 to 59
60 to 64
Above 64
Active population
2.1%
9.9%
9.9%
12.8%
14.2%
13.8%
15.7%
14.2%
16.2%
14.3%
15.9%
13.7%
14.2%
4.
2.1%
2.1%
1.3%
i 2.4%
2.1%
Under 20
20 to 24
25 to 29
30 to 34
35 to 39
40 to 44
45 to 49
50 to 54
55 to 59
60 to 64
Above 64
Claimants
0.2%
0.1%
■	2.5%
■	3.0%
6.9%
11.0%
11.7%
16.8%
15.1%
18.2%
19.3%
21.7%
20.2%
20.7%
15.5%
i 0.9%
i 0.7%
0.0%
0.0%
0.0%
figure 6 Age distribution of claimants and active population in Slovenia, 1994-2003
(light gray - man, dark gray - women; based on the information provided
by the Guarantee Fund of Slovenia)
ing the working of this program in Slovenia. Our findings suggest that
severance pay non-performance has been a significant problem in Slove-
nia. In 2000, only two-thirds of total severance pay obligations were ac-
tually honored, a small portion of non-paid severance pay claims was
reimbursed by the Guarantee Fund, and the rest - one-third of total
obligations - was not paid at all. Moreover, we showed that while both
men and women were equally affected, workers older than 40 years were
more likely than younger ones to be confronted by severance pay non-
performance. And, finally, we also found that among firms that incurred
severance pay liabilities, larger and more productive firms were more
likely to pay them out.
Taken together, these findings shed a rather negative light on severance
pay as an income protection program for the unemployed. First, the pro-
gram fails to protect a significant share of those who are legally entitled
to such protection - even after the introduction of the Guarantee Fund.
table 2 Multinomial logit estimates of the likelihood of severance pay
non-performance
Multinomial logit estimates	Descriptive statistics1
Paying severance Failing to pay severance
obligations	obligations
	(1)	(2)	(1)	(2)	(3)	(4)
(a)	-0.04	0.49	-5.76**	2.17	0.07	0.21
(b)	1.38**	0.28	0.47	0.88	0.61	0.49
(c)	-286**	0.25	-5.35**	0.81		
notes The definition of dependent variable: not incurring severance obligations is
taken as a baseline, and incurring severance obligations and paying them, and incurring
severance obligations and failing to pay them, as other options. No. of observations:
816, pseudo R2: 0,056. 1 Mean value of dependent variable is 0/08, and its standard error
is 0.27. (a) Efficiency of the firm, (b) size of the firm (1 if firm's employment exceeds 100
workers, 0 otherwise), (c) constant; (1) coefficient, (2) robust standard error, (3) mean,
(4) standard deviation. Significance at 1 and 5 percent levels are indicated by ** and
*, respectively.
Second, the program is prone to creating inequities, as it disproportion-
ally affected older workers. At the same time, our findings also provide
some clues about how to make the program more effective. The fact that
less productive - and hence less profitable - firms are less likely to honor
their obligations suggests that non-performance is strongly related to the
non-funded nature and limited risk-pooling of severance pay, and thus
the recommendation of converting severance pay to a funded program.
Let us conclude with recommendations for better coordinating sever-
ance pay with other income support systems for the unemployed. First,
countries with both unemployment insurance and severance pay pro-
grams (Slovenia being one of them) can save on costs without reduc-
ing insurance by better coordinating payments under the two programs.
Namely, unemployment insurance eligibility rules could be adjusted so
that insurance benefits would only start after the severance benefits 'ex-
pire,' that is, after n months, if the individual received n monthly wages
as the severance payment (such a program is in place in some developed
countries, for example, in Canada - see Vodopivec 2004).
Another possibility - explicitly addressing the non-performance prob-
lem - is the conversion of severance pay to pre-funded unemployment
insurance savings accounts (uisas), a reform implemented by Austria
in 2002.8 Besides correcting for the non-performance problem, uisas
would improve efficiency by removing obstacles to labor market flexi-
bility and reducing litigation costs. Lastly, the most radical option is the
introduction of an integrated severance and ui system (Chilean model),
consisting of two components: uisas and a solidarity fund, with benefit
recipients first drawing benefits from their uisas and upon depletion,
reverting to the solidarity fund (for details of the reform, see Acevedo,
eskenazi, and Pages 2006, and for theoretical considerations, Parsons,
forthcoming).
Acknowledgments
The authors would like to thank the Statistical Office of Slovenia and the
Guarantee Fund of Slovenia for providing the data used in this paper.
Notes
1	See Holzmann, Iyer, and Vodopivec (2008) for a survey of the incidence of
the severance pay around the world and a review of the origin, economic
rationale, and current attempts to reform severance pay programs.
2	Guarantee Fund also exists in Estonia, Romania and Uzbekistan.
3	Under the 1993 Law on Bankruptcy and Liquidation, workers can sue their
former employers that undergo a liquidation or bankruptcy process, with
workers' severance pay claims having a priority over other claims (up to a
limit - for details, see Kresal Šoltes 1997).
4	Three socialist countries (China, North Korea, and Vietnam); eight suc-
cessor countries of ussr (Azerbaijan, Estonia, Georgia, Kazakhstan,
Latvia, Lithuania, Russia, and Ukraine); five successor countries of Yu-
goslavia (Bosnia and Herzegovina, Croatia, Macedonia, Serbia and Mon-
tenegro, and Slovenia); two successor countries of Czechoslovakia (Czech
Republic and Slovak Republic), and three former socialist European coun-
tries (Bulgaria, Hungary, and Poland).
5	Most oECD countries also have mandatory severance pay programs, but
some - including Australia, Denmark, Finland, Germany, Japan, Nether-
lands, New Zealand, Norway, Sweden, and the United States - leave such
arrangements to collective bargaining or rely on the common law provi-
sions. For example, in the Netherlands, even though the law does not re-
quire severance pay, employers often make payments to dismissed workers
to avoid legal proceedings for an 'obviously unreasonable dismissal.' The
cantonal courts have even created a statutory-like formula for the amount
of severance payments.
6	Alternative measures of efficiency, obtained via fixed effects and random
effects estimation of the above translog production function, yielded sim-
ilar results.
7	It is possible that some workers received additional reimbursement from
bankruptcy or liquidation proceeds - we do not have any information
about such reimbursements.
8	In 2002, Austria converted its severance pay to a fully funded contribu-
tory system akin to unemployment insurance savings accounts (Koman,
Schuh, and Weber 2005). The reform extended the entitlement to work-
ers with short tenures and removed obstacles to worker mobility, granting
full portability and allowing the accumulation of benefits from the begin-
ning of an employment spell. Employers pay 1.5 percent of each worker's
salary to each individual worker, with resources held in a central account
and invested in the capital market. Laid-off workers with job tenure of
three years or more can withdraw accumulations in their accounts or keep
them and claim them upon retirement. Workers who separate voluntarily
or have tenures of less than three years are denied the right of immediate
withdrawal, a feature that may hinder worker mobility.
References
Acevedo, G., P. Eskenazi, and C. Pages. 2006. Unemployment insurance
in Chile: A new model of income support for unemployed workers.
Social Protection Discussion Paper 612, World Bank.
Addison, J., and P. Teixeira. 2001. The economics of employment protec-
tion. iza Discussion Paper 381, Institute for the Study of Labor.
Holzmann R., K. Iyer, and M. Vodopivec. 2008. Severance pay programs
around the world: Rationale, status, and reforms. In Mandated sever-
ance pay programs: An international perspective on status, concepts and
reforms, ed. R. Holzmann and M. Vodopivec, 25-87. Washington, dc:
World Bank.
Koman, R., U. Schuh, and A. Weber. 2005. The Austrian severance pay
reform: Toward a funded pension. Empirica 32 (3-4): 255-74.
Kresal Šoltes, K. 1997. Novosti pri varstvu delavskih terjatev v primeru in-
solventnosti delodajalca. Podjetje in delo 5 (23): 519-28.
MacIsaac, D., and M. Rama. 2000. Mandatory severance pay in Peru: An
assessment of its coverage and effects using panel data. Washington, dc:
World Bank.
Mansor, N., T. E. Chye, A. Boehanoeddin, F. Said, and S. M. Said. 2001.
Malaysia: Protecting workers and fostering growth. In East Asian labor
market and the economic crisis: Impacts, responses, and lessons, ed. G.
Betcherman, and R. Islam, 141-94. Washington, dc: World Bank.
Orazem, P. F., and M. Vodopivec. 2008. Do market pressures induce eco-
nomic efficiency? The case of Slovenian manufacturing, 1994-2001.
Southern Economic Journal 76 (2): 553-76.
oecd. 1999. Employment outlook. Paris: oecd.
Parsons, D. O. Forthcoming. Savings accounts as an alternative to job
separation pay: theoretical considerations. In Mandated severance pay
programs: an international perspective on status, concepts and reforms,
ed. R. Holzmann and M. Vodopivec. Washington, dc: World Bank.
Schwab, J. S., 2003. Mandating severance pay in transition economies. Pa-
per presented at the International Workshop on Severance Payments
Reform: Toward Unemployment Savings and Retirements Accounts,
Laxenburg, Austria.
Van Ours, J. C., and M. Vodopivec. 2006. How shortening the potential
duration of unemployment benefits affects the duration of unemploy-
ment: Evidence from a natural experiment. Journal of Labor Economics
24 (2): 351-78.
Vodopivec, M. 2004. Income support for the unemployed: Issues and options.
Regional and Sectoral Studies Series. Washington, dc: World Bank.