The Impact of Taxation on Unemployment in OECD Countries

There have been numerous studies of the impact of labor market rigidities on unemployment rates. The general conclusion for both OECD as well as other countries is that rigidities explain a significant part of the observed international variation in unemployment rates (Layard, Nickell, and Jackman 1991). The primary interest of the present paper is the impact of taxation on unemployment in OECD countries. The relation between unemployment and taxation has recently attracted special attention. For example, Nickell (1997) has found that taxation is a significant factor in explaining differences in unemployment rates across countries (see also Scarpetta 1996; Nickell and Layard 1997; Heitger 1998; and Elmeskov, Martin, and Scarpetta 1998). Since high unemployment rates may lead to higher government expenditures and taxes, the question is whether the impact of the tax burden on unemployment has been estimated correctly--that is, whether the estimates are consistent and unbiased. To evaluate the true impact of taxation on unemployment, Hausman specification tests can be carried out (Hausman 1978). With the help of these tests it is possible to investigate whether the impact of the tax burden on unemployment is exogenous. If the outcome is that the null hypothesis (that taxes are exogenous) has to be rejected, a two-stage least squares estimation procedure can provide unbiased and consistent estimates of the tax burden's impact on unemployment and thus correct for the simultaneity bias. The empirical investigation will focus on total unemployment but also examine long-term and short-term unemployment. The reason is that an increase in long-term unemployment will eventually create a larger effect on government expenditures (and taxes) than will an increase in short-term unemployment. Since the tax burden is assumed to be only one of the structural features of the labor market in OECD countries, the impact of the other determinants also has to be estimated. The investigations are based on Nickell's (1997) data set, which contains the relevant data for two periods, namely 1983-88 and 1989-94. Unemployment Experience in OECD Countries Unemployment rates in OECD countries differ widely (Table 1). The rate of total unemployment in 1983-88 in Spain was 19.6 percent while the rate in Switzerland was only 0.8 percent. In 1989-94 these two countries again reported the highest (18.9 percent) and the lowest (2.3 percent) unemployment rate. The OECD average rate of total unemployment increased only slightly from 7.8 percent in 1983-88 to 8.0 percent in 1989-94. Total unemployment in Europe was higher than in non-European countries in both periods. At the same time the variation of total unemployment rates increased. Short-term unemployment (i.e., a duration of unemployment less than one year) also varied widely in OECD countries. Again in Switzerland in both periods the short-term unemployment rate was lowest. In contrast, Canada in 1983-89 and Spain in 1989-94 reported the highest rate (about 9 percent). The mean rate of short-term unemployment in OECD countries accounted for more than half of the rate of total unemployment--with the non-European countries showing a much higher share. The variation in short-term unemployment in Europe and non-Europe increased in the course of time, but in both periods the rate was higher in the European countries. Long-term unemployment in non-European countries seems to have been a problem of only minor importance--l.0 percent and 1.4 percent in 1983-88 and 1989-94, respectively--compared with Europe where this type of unemployment accounted for 4.2 percent and 3.7 percent. Characteristics of OECD Labor Markets The theoretical framework to explain unemployment rates in OECD countries is given by the familiar model of Layard, Nickell, and Jackman (1991). The model is characterized by an upward sloping wage-setting schedule which follows from the assumption that real wages are the result of a bargaining process between employers and employees--and a downward sloping labor demand schedule. …

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.