نیروی کار و یا مصرف مالیات؟ یک برنامه با یک مدل تعادل عمومی پویا با عوامل ناهمگن

In response to the long-run structural challenges and deficits accumulated by the current crisis, many western governments are intending to raise consumption taxes but trying to avoid higher labor taxes at the same time. In fact, the tendency is to lower labor taxes if the government's fiscal situation allows it. Also the Finnish government has many times highlighted the urgent need for this kind of tax reform, i.e. the reform that raises consumption taxes but decreases labor taxes. In addition to a change in the source of taxation, this kind of tax policy switch also contains another aspect: replacing a progressive tax with a flat tax. From the theoretical point of view, changing the structure of taxes can be seen as part of a larger issue, the design of optimal tax system. The theoretical underpinnings of the topic can be found e.g. in Mirrlees, 2006, Salanié, 2003 and Kaplow, 2008. In macro context, tax structure changes have been analyzed using a variety of approaches. The important work has been done by Auerbach and Kotlikoff (1987) who consider changes in taxes in an overlapping generations setting with exogenous growth. Jones et al. (1993) study the issue in an infinite-horizon representative-agent framework with endogenous growth and Coleman (2000) in the context of optimal Ramsey tax policy. Regardless of many theoretical articles concerning the topic, studies with a more empirical approach are harder to find. Auerbach (1996) estimates that various proposals to replace the current income tax system in the U.S. with a consumption tax would produce long-run output gains of 3.2% to 9.7%. Heer and Trede (2003) study the output and distribution effects of tax reforms in a general equilibrium model calibrated to fit the stylized facts of the German economy. In their study income taxes are replaced with a flat-rate tax or consumption taxes. Their results show a significant rise in output, negligible effects on labor income distribution, but quite considerable (negative) effects on wealth distribution. Nishiyama and Smetters (2005) also study a similar kind of tax reform, i.e. a reform in which a progressive income tax is replaced by a flat consumption tax. They use an overlapping-generations model in which agents face idiosyncratic wage shocks and longevity uncertainty. They find that the effects of the tax reform crucially depend on the insurability of the wage shocks. In a pure empirical study based on the cross sectional data of 22 OECD countries Kneller et al. (1999) find that by raising consumption taxes and declining labor and other distortionary taxes, considerable output and employment gains would be reached. Bleaney et al. (2001) use the same data and end up with the same conclusions. Unlike the previous investigation, they also try to eschew biases associated with incomplete specification of the government budget constraint and endogeneity of fiscal or investment variables. Tervala and Ganelli (2008) study the effects of a tax structure reform with an open economy DSGE (dynamic stochastic general equilibrium) model. They find modestly positive effects on growth in the long run when labor taxes are replaced with consumption taxes. However, the model they use does not include capital, and its calibration does not represent any particular country.1 While we have some international evidence of the effects of tax reforms, there are almost no empirical macro studies of the tax structure changes that use Finnish data. Only Kilponen and Vilmunen (2007) make an exception for this. They find that shifting taxes from labor towards consumption produces a significantly positive employment and GDP effect. Their study uses DSGE macromodel that also tries to capture the behavior of the pensioners. For this reason, the results are very sensitive to the assumptions made for labor supply. Hence we still know very little how a tax structure reform would affect the output and employment in Finland. And we know almost nothing about the distributional effects of the reform. To understand the effects of labor and consumption taxes, I first discuss the theoretical aspects of direct and indirect taxation. Then, in order to assess these effects quantitatively, I apply a general equilibrium model with heterogeneous agents to compare three fiscal regimes: i) progressive labor taxes that correspond to the Finnish system, ii) flat-rate labor tax, and iii) only a consumption tax. That said, I utilize the aspects of the framework presented in Heer and Trede, 2003 and Heer and Maussner, 2009. Nevertheless, the model presented in this paper has many unique characteristics. Unlike these previous studies in which income taxes are levied similarly on capital and labor, my framework is the Finnish dual income tax system that treats capital and labor income separately. This allows me to focus purely on the comparison of labor taxes and consumption taxes. Also, I change the theoretical assumptions concerning the risk of unemployment and calibrate the model to fit the stylized facts of the Finnish economy. The results show that replacing progressive labor taxes with a flat-rate labor tax produces a slightly larger economy with fractionally more inequality. The output effect is almost totally due to the increase in capital stock. In the second and main experiments I find that the tax reform that replaces progressive labor taxes with a flat consumption tax has only minor effects on labor supply and gross labor income distribution, a positive effect on capital stock, but a negative effect on wealth distribution (i.e. wealth concentration increases). The sensitivity analysis shows that with less risk averse agents, the contribution of capital to output effect decreases but wealth concentration increases more when compared to the benchmark results. The paper is organized as follows. Section 2 discusses the theoretical aspects of labor and consumption taxes. Section 3 introduces the model I use for simulation, and in Section 4 the model parameters are calibrated. In Section 5 I discuss the results from different tax policies. Final section concludes.

A tax reform that puts more weight on consumption taxes but reduces labor taxes has been vividly discussed in the Western countries during recent years. Economists and politicians generally see consumption taxes as the least distortionary way to increase tax revenues collected by the government. However, the positive effects of consumption tax are debatable while its offsetting effect on labor supply makes the net output change rather ambiguous. Consumption tax may be justified on the grounds that it is also a tax on existing assets, which does not affect labor supply decision of households. Still, the significance of this effect is uncertain, at least in the long run. The efficiency of the consumption tax is also much dependent on whether a price level change due to the consumption tax increase is compensated to pensioners and other non-working groups. Another, even a bigger issue than the output effects, is the reform's effect on income and wealth distribution, i.e. on inequality. In this study, I use a dynamic general equilibrium model with heterogeneous agents to assess the output and distributional effects of tax policy reforms. The agents in the model differ with regard to their productivity and employment status that are subject to idiosyncratic shocks; hence the agents are mobile and their productivity and employment status may change between periods (Table 4). Table 4. Effects of tax policies. Tax policy K N View the MathML sourcen¯ r Ginil Giniw View the MathML sourceσn/n¯ Progressive labor taxes 3.48 0.317 0.304 7.77 0.218 0.403 0.320 Consumption tax 3.96 0.321 0.308 7.21 0.214 0.420 0.319 Table options In order to differentiate the effects of progressivity from the source of taxation, I begin by simulating a switch from progressive to a flat-rate labor tax. This results in an economy with some degree larger capital stock, negligibly more employment but slightly more inequality. The main results concern the tax policy reform that replaces progressive labor taxes with a proportional consumption tax. According to the simulations this reform results in a significant rise in capital accumulation, a negligible change in labor supply and gross labor income distribution, but a relatively considerable increase in wealth concentration. To summarize, the tax system that replaces labor taxes with consumption taxes produces a more capital intensive economy with somewhat more wealth inequality. Even if the model simulations prove to be relatively robust on the basis of the sensitivity analysis, there are also reasons why the results should be interpreted carefully. The reasons are discussed in Heer and Trede (2003) who use a modeling technique similar to mine (Table 5). This includes the possible transition effect after the tax policy change and the strong assumption about the exogeneity of workers' productivities, which is also independent of the tax policy regime. It is also worth noting that this study analyzes labor supply only along the intensive margin; tax structure change may naturally have an effect along the extensive margin, i.e. whether people attend in the labor markets. From an empirical point of view, one has to also remember that the model economy only consists of labor supplying agents: pensioners and students are not taken into account in this paper. Hence it is probable that the model experiment underestimates the distributional effects of the tax reforms. Despite these reservations, I argue that the results prove that replacing labor taxes with consumption taxes only slightly improves employment, and albeit the reform increases capital accumulation significantly, it contributes negatively on wealth inequality. Table 5. Sensitivity analysis of σ. σ Tax policy K N View the MathML sourcen¯ r Ginil Giniw View the MathML sourceσn/n¯ 1 Progressive labor taxes 2.91 0.268 0.254 7.85 0.225 0.281 0.315 1 Flat-rate labor tax 2.98 0.270 0.254 7.71 0.232 0.349 0.320 1 Consumption tax 2.93 0.276 0.261 7.51 0.228 0.486 0.318 3 Progressive labor taxes 4.00 0.352 0.339 7.59 0.212 0.379 0.323 3 Flat-rate labor tax 4.31 0.352 0.339 7.25 0.213 0.386 0.326 3 Consumption tax 4.73 0.354 0.342 6.85 0.206 0.384 0.30