Economists expect unemployment benefits to lead to longer unemployment spells,
since workers will tend to be choosier about job offers the larger their income while they
are out of work. But the real magnitude of this effect has long been debated.
New research by Olympia Bover, Manuel Arellano and Samuel Bentolila reveals
that the impact of benefits on time spent in unemployment is actually quite large. Their
empirical analysis of the long-term impact of a Spanish labour market reform in 1984 -
which produced a new type of unemployed worker without any benefits at all alongside
similar workers enjoying generous benefits - has striking results. For example:
?? Three months into a spell, the chances of leaving unemployment were 22% for
those without benefits, but only 11% for those on the dole.
?? More than a half of benefit non-recipients had left unemployment at the end of the
fourth month into a spell while close to three-quarters of those on the dole were still
jobless.
?? Receipt of benefits reduced individual workers’ chances of leaving unemployment
more than a severe economic downturn (causing output growth to drop by, say, 4
percentage points).
The researchers conclude that policy-makers who wish to mitigate long-term
unemployment – a scourge of many European countries – should consider reducing the
length of statutory benefit entitlement. They estimate, for example, that in average
economic conditions, a cut in benefit entitlement from 8 to 4 months could reduce
median unemployment duration from 14 to 9 months. Of course, policy-makers should
always weigh the beneficial effect of this measure against its potential negative impact
on workers’ welfare.
In the past, efforts at estimating the magnitude of the impact of unemployment benefits
have been hampered by the difficulty in establishing whether, apart from receipt of
benefits, recipients and non-recipients were comparable workers according to other
personal characteristics. For example, as unemployment spells lengthen, the pool of
recipients may contain an increasing share of less employable individuals (the most
employable ones having left already), and this would confound the estimation of any
causal effects of benefits.
This research examines the case of Spain, where the unemployment rate reached a
staggering 24% in 1993. A 1984 labour market reform introduced fixed-term contracts
with very low severance pay. These contracts were eagerly adopted by employers so
that they soon covered a third of all employees. By producing a new type of
unemployed worker without any benefits at all, alongside otherwise similar workers
enjoying generous benefits, this reform created a long-running and countrywide ‘natural
experiment’.
ENDS
Notes for Editors: ‘Unemployment Duration, Benefit Duration and the Business Cycle’
by Olympia Bover, Manuel Arellano and Samuel Bentolila is published in the April 2002
issue of the Economic Journal.
Bover is at the Research Department of the Bank of Spain; Arellano and Bentolila are
at the Centro de Estudios Monetarios y Financieros (CEMFI) in Madrid.
For Further Information: contact Samuel Bentolila via email: bentolila@cemfi.es; or
RES Media Consultant Romesh Vaitilingam on 0117-983-9770 or 07768-661095
(email: romesh@compuserve.com).