وضع مالیات و بیکاری: یک رویکرد کاربردی تعادل عمومی

We present an applied general equilibrium (AGE) modelling approach to analyse employment and unemployment effects of labour tax cuts in an economy where wages are determined through firm–union bargaining at the sectoral level. In such a labour market regime, simulations for Germany show that labour tax policies can make only a small contribution to alleviating the problem of persistent unemployment.

Persistent unemployment at high levels is a central policy problem in many European countries (see OECD, 2001, for a recent overview). Among the alternative policy proposals to reduce unemployment, tax policy shifts have received much interest. If it should turn out that tax cuts on labour can produce major positive labour market effects, this would give politicians a much less controversial instrument than radical changes in the institutional labour market settings. During the last decade, the effects of taxation on unemployment have been a major research topic in public finance (see Sørensen, 1997, for an overview). The respective literature combines different theories of equilibrium unemployment with classical methods of tax incidence analysis. Although rigorous analytical models provide a number of important insights, e.g., the positive impact of labour income tax progression on employment Hoel, 1990, Lockwood and Manning, 1993, Koskela and Vilmunen, 1996 and Goerke, 1997, their contribution to practical policy making remains rather limited. The reason is that theoretical models used to investigate the relationship between taxation and (un)employment are highly stylised in order to keep analytical tractability. Accounting for a more detailed production or consumption structure and the specific institutional features of a country's labour market or tax system makes analytical solutions unfeasible, thus requiring numerical solution methods. For a quantitative assessment of the labour market impacts induced by realistic tax policy shifts, the step from stylised analytical to complex numerical models is inevitable. Such a transition has occurred since the early 1980s in the fields of applied tax and trade policy analysis using computable general equilibrium models in particular (see e.g., Shoven and Whalley, 1984 and Ballard et al., 1985). The general equilibrium approach provides a comprehensive framework for studying the effects of policy interference on all markets of an economy, based rigorously on microeconomic theory. The simultaneous explanation of income generation and spending for all economic agents allows us to address both efficiency as well as distributional effects of policy changes. This is why applied general equilibrium (AGE) models have become a standard tool for quantitative policy analysis (for surveys on the use of AGE models in different policy fields see e.g., Shoven and Whalley, 1992, Pereira and Shoven, 1992, Kehoe and Kehoe, 1994 and Fehr and Wiegard, 1996, or Weyant, 1999). To date, however, little work has been done to incorporate unemployment features within the applied general equilibrium framework, although labour market effects of policy interference have become a key interest to decision-makers. A common ad hoc modelling approach is to replace the competitive labour market of a standard general equilibrium setting with a “wage curve” (Blanchflower and Oswald, 1994). The wage curve reflects empirical evidence on the inverse relationship between the level of wages and the rate of unemployment. In such a model, the wage curve, together with labour demand, determines the level of involuntary unemployment (see e.g., Böhringer et al., 2003a and Böhringer et al., 2003b). The wage curve constitutes a convenient shortcut to incorporate unemployment, but it lacks an explicit microfoundation. This makes it impossible to analyse how specific policy measures affect the wage setting mechanism. In order to track down the causal chain from policy interference to labour market effects, one must open the “black box” of the wage curve and explicitly model the wage-setting process. Concrete examples include the efficiency–wage model provided by Hutton and Ruocco (1999) for selected EU countries and MIMIC, a detailed model of the Dutch labour market Bovenberg et al., 2000 and Graafland et al., 2001, in which wages are determined by centralised collective bargaining between firms and trade unions. In the MIMIC model, the wage-bargaining equation contains economy-wide averages of the bargaining power, output and labour demand elasticities and, therefore, largely neglects sector-specific characteristics. Here, we present an AGE modelling approach to incorporate sectoral wage bargaining which is relevant for various OECD countries including Germany, Spain, or Sweden (see EIRO, 2002): Wages are determined through firm–union bargaining at the sectoral level. Different economic conditions across sectors then produce different bargaining settings with varying bargaining power for firms and unions, making wage negotiations sensitive to the specific conditions in each sector. From a methodological point of view, our model feature of sectoral heterogeneity in wage bargaining provides an innovative contribution to the AGE literature. From the viewpoint of policy analysis, illustrative simulations for Germany indicate that cuts in labour taxes can make only a small contribution to alleviating the problem of persistent unemployment. Given the rather small impacts of tax policy changes, the ranking of the different tax instruments in terms of output and employment is rather robust and can be mainly traced back to changes in the progressivity of the wage income tax system. However, the welfare consequences for workers and capitalists can differ substantially with varying assumptions on capital mobility. The remainder of the paper is organised as follows. Section 2 describes the algebraic model structure and lays out in detail how sectoral wage bargaining can be incorporated into the general equilibrium framework. Section 3 provides illustrative simulations of labour tax policies for Germany. Section 4 concludes the paper.

High taxes on labour combined with institutional rigidities of the labour market are often held responsible for substantial structural unemployment in various OECD countries. In this vein, cuts in labour taxes are considered as a potentially important policy measure to alleviate the unemployment problem. In this paper, we developed an applied general equilibrium model that allows us to assess the quantitative effects of tax policy shifts in economies featuring decentralised wage bargaining. A distinctive and innovative feature of our model is the sectoral heterogeneity in wage bargaining between employers' associations and trade unions that represent workers of two skill groups with different bargaining power. Illustrative policy simulations for Germany show that the reduction in unemployment that can be achieved by tax reforms is rather moderate. Labour market effects can mainly be traced back to changes in the tax progression. Sensitivity analysis with respect to assumptions on the institutional setting for unemployment benefits and on capital mobility shows that the ranking of tax policy instruments in terms of output and employment remains stable. Regarding welfare, the ranking is rather robust with respect to the institutional setting for the unemployment benefits, but it is highly sensitive to the assumption on capital mobility. With internationally mobile capital, workers can substantially benefit from capital imports through higher wages, whereas capitalists no longer benefit from higher rental rates of capital (the latter being fixed at the benchmark level). International factor mobility is thus of great importance for the distributional consequences of the tax reforms we analysed. There are various aspects missing from our modelling framework that are potentially important when we assess the prominent role of tax progression in our tax policy reforms. (1) An endogenous decision on working hours may substantially reduce or even invert the employment effects of tax progression (see Sørensen, 1999 and Fuest and Huber, 2000). (2) Taxation has a strong impact on the attractiveness of the informal sector. Black market work that contributes to official unemployment becomes more attractive the higher the marginal labour income tax rates are. (3) Taxation affects endogenous decisions about human capital (Fuest and Huber, 1998), which in turn feed back to the level of unemployment. (4) As compared to skill groups in labour markets with collective wage bargaining, tax policies might work quite differently for the segment of the highest skilled workers in competitive labour markets. We plan to address these issues in future research developing the current model to the extent possible with the available data.