This article appears in the following Oxford Review of Economic Policy issue: CAPITALISM AND INEQUALITY [View the issue table of contents]
The Swedish unemployment experience
* Department of Economics, Uppsala University, e-mail: bertil.holmlund{at}nek.uu.se
 |
Abstract |
|---|
By international standards, unemployment in Sweden remained remarkably low throughout the 1970s and the 1980s. In the early 1990s, however, the unemployment rate increased sharply and hit double-digit levels. The paper argues that the steep rise in unemployment was mainly the result of a series of adverse macroeconomic shocks, partly self-inflicted by bad policies, and partly caused by unfavourable international developments. The extremely contractionary monetary policy in 1992 appears to have had strong and long-lasting effects on unemployment. Institutional factors do not appear as convincing explanations of the steep rise in unemployment in the early 1990s.
Key Words: unemployment • monetary policy • labour-market institutions
The paper was prepared for the LoWER conference, ‘Institutions, Markets and European Unemployment Revisited: What Have We Learned’, Amsterdam, 18–19 April 2008. The paper draws heavily on Holmlund (2003, 2006) and Alexius and Holmlund (2008).
1 The unemployment figures that are used in this paper refer, unless stated otherwise, to labour-force survey data and national definitions. The latter differ slightly from the definitions of unemployment given by the International Labour Organization (ILO). In particular, students engaged in fulltime job search have been classified as ‘unemployed’ by the ILO but as ‘out of the labour force’ in the national definition. Since October 2007, Sweden has adopted the ILO criteria regarding students who search for work. The working-age population is generally confined to those aged 16–64.
2 The ‘latent job seeker’ category also comprises full-time students (including persons in labour-market training) who are searching for employment.
3 This measure has been used by, among others, the Bank of Canada, as an operational target of policy.
4 The main alternative is to split the sample and estimate two different models using different variables to represent monetary policy in the two regimes. However, if the sample is split in November 1992 and two different models are estimated for the two sub-periods, the recession in 1993–4 loses any possible link to the 1992 monetary policy since these two events are not included in the same sample. If we want to analyse the effects of these major Swedish monetary policy shocks we have to include data both from fixed and floating exchange-rate regimes and use a measure of monetary policy that is applicable to both regimes.
5 The Riksbank considers alternative paths of interest rates for 3–4 years. The path with ‘high’ interest rates involves an 0.7 percentage point higher interest rate by the end of 2010 relative to a baseline scenario. This path is associated with 0.6 percentage point higher unemployment by the end of 2010.
6 This refers to ‘regular’ UI benefits which are based on previous income. Individuals not qualifying for regular benefits may qualify for a basic allowance that has amounted to around 50 per cent of the maximum UI benefit level.
7 The measure of the total tax wedge is 
, where s is the payroll tax rate (levied on firms), vat is the value added tax, and t is the income tax paid by workers. See Forslund and Kolm (2004) for more information about the evolution of tax wedge.
8 Over the period 1987–2000, the flow data from the labour-force surveys indicate that the average risk of becoming unemployed was more than 10 times higher for a temporary worker than for a worker on an open-ended contract.
9 See OECD Employment Outlook 2002, ch. 5. A caveat is that the results also indicate that product-market regulations are less harmful in countries with corporatist wage-setting regimes.