مالیات بهینه از نیروی کار غیر ماهر با جستجوی کار و کمک های اجتماعی

In order to explore the optimal taxation of low-skilled labor, we extend the standard model of optimal nonlinear income taxation in the presence of quasi-linear preferences in leisure by allowing for involuntary unemployment, job search and an exogenous welfare benefit. In trading off low-skilled employment against work effort of higher skilled workers, the government balances distortions on the search margin with those on work effort. Higher welfare benefits typically reduce taxes paid by low-skilled workers and raise marginal tax rates throughout the skill distribution.

Widening wage dispersion raises the question how public policy should protect the living standards of unskilled workers, as policy makers are increasingly concerned about the adverse incentive effects of generous income support. In response to these concerns, many countries have cut taxes on unskilled work in order to combat poverty while at the same time encouraging unskilled workers to look for work. Both the United States and several European countries have already introduced or are considering in-work tax benefits for unskilled work in the form of an Earned Income Tax Credit (EITC). These tax policies are part of active labor-market policies and ‘welfare-to-work’ programs, where governments fight poverty by raising employment of unskilled workers. To investigate the optimal response of tax policy to income support provided through the welfare system and to declining relative wages of unskilled labor, we extend the standard model of optimal nonlinear income taxation developed by Mirrlees (1971). In particular, we incorporate labor-market imperfections that induce governments to provide income support, namely search costs and involuntary unemployment. In the presence of a search margin, the government has to account for not only the standard incentive compatibility constraint on work effort, but also a participation constraint on the willingness of low-skilled agents to look for work. Indeed, from an analytical point of view, our main contribution is to add a participation constraint to the optimal tax problem, including the decision regarding which types should optimally participate in job search. In doing so, we extend both the optimal tax literature, which typically abstracts from participation constraints, and the literature on optimal nonlinear monopoly pricing, which generally assumes that the lowest participating type is exogenously given. Within a nonlinear pricing framework, Rochet and Stole (2002) recently added an endogenous participation constraint by allowing agents to differ in both outside options and preferences for quality. Our analysis deviates from that of Rochet and Stole (2002) in two respects. First, in the nonlinear pricing problem explored in Rochet and Stole (2002), the monopolist cares only about profits earned on the participating agents. In our optimal tax problem, in contrast, also agents who do not participate appear in the objective function because the government is interested in the utilities of both participating and nonparticipating types. The second difference with Rochet and Stole (2002) is that we allow agents to differ in only one dimension; agents feature different skill levels but exhibit the same search costs. Within the context of our labor-market application, this is a reasonable assumption, which is in fact employed by most of the labor-market literature on search (see, for instance, Mortensen and Pissarides, 1999). This assumption implies that the participation constraint is binding only at the bottom of the skill distribution. Rochet and Stole (2002), in contrast, derive a binding participation constraint for each type. The literature on optimal income taxation has modelled unemployment of unskilled agents as these agents reducing the hours they work in their jobs when they face low gross wages and rapidly rising marginal tax rates. Accordingly, low productivity workers are bunched in low- or zero-production jobs. By introducing a participation margin and positive search costs, we model another type of bunching at the bottom of the skill distribution: unskilled agents do not search for work and thus drop out of the labor force. Heckman (1993), for instance, stressed that ‘a crucial theoretical distinction with important empirical pay off is that between labor supply choices at the extensive margin (…) and choices at the intensive margin’. Empirical work does in fact reveal that unskilled workers adjust their labor supply in response to tax and benefit programs on mainly the extensive margin (i.e. leaving the labor force altogether, for example through early retirement) rather than the intensive margin (i.e. reducing the hours they work in their jobs) (see, e.g. Eissa and Liebman, 1996, Kimmel and Kniesner, 1998, Blundell, 2001, Meyer and Rosenbaum, 2001 and Meyer, 2002. This explains the policy concern about welfare programs and high taxes on unskilled work discouraging low-income earners from looking for work. Indeed, our model is consistent with the stylized fact that low-skilled agents feature the highest long-term unemployment rates (see OECD, 2001). Also Saez (2002) incorporates the two labor-supply margins of not only hours worked but also labor-force participation in an optimal income tax model.2 Our approach differs from that of Saez in three important respects. First of all, whereas Saez assumes that all unemployed have voluntarily left the labor force, we account also for involuntary unemployment. Whereas allowing for unemployment risk does not complicate the analysis, it provides an implicit justification for substantial welfare benefits. Agents thus face two risks: being born with low ability and being involuntarily unemployed. More generally, we are more explicit than Saez (2002) about the labor-market imperfections affecting the costs and effectiveness of labor-market search, including the welfare implications of these imperfections. Our analysis differs from Saez (2002) also in that the government takes the welfare benefit as exogenously given when optimizing the tax system. Hence, the government can employ only the nonlinear income tax to optimize social welfare. Indeed, in practice, taxes and social assistance are often set by distinct agencies based on rather different interests and considerations. In some federal countries, for example, local governments determine social assistance benefits, while the central government is mainly responsible for the tax system. One can also interpret the minimum income floor set by social assistance benefits as being determined by considerations outside our model. Alternatively, one can view our analysis as exploring how the tax system can be employed to address the possibly suboptimal aspects of social assistance. A final difference is that Saez (2002) allows for more general preferences that are not necessarily quasi-linear in leisure. Whereas his results are thus more general than ours, our specific assumptions on preferences enable us to derive more analytical results on comparative statics with respect to public spending, labor-market imperfections (such as the costs and effectiveness of search) and institutional features of the welfare system. This sheds additional light on the determinants of the optimal tax schedule. Indeed, a substantial literature (see, e.g. Boadway et al., 2000, Ebert, 1992, Weymark, 1986, Weymark, 1987 and Lollivier and Rochet, 1983 has turned to quasi-linear preferences in leisure in order to obtain more intuition for the determinants of the optimal nonlinear income tax, as these preferences allow for closed-form solutions of the standard optimal nonlinear income tax problem. Our quasi-linear preferences also imply that a utilitarian government cares about the distribution of consumption rather than the distribution of work effort. Indeed, policy debates typically focus on raising consumption rather than reducing work effort of the poor. We extend the literature on optimal nonlinear income taxation with quasi-linear preferences in three ways; we allow for involuntary unemployment, a participation (or search) constraint, and an exogenous welfare benefit. These three extensions make this literature more relevant for addressing the timely policy question of how the optimal tax system should treat low-skilled employment in the face of income support. The participation margin also eliminates some of the unrealistic implications of a model with quasi-linear preferences for aggregate labor supply elasticities. Hence, the introduction of a participation margin is particularly important in the context of this particular model. The rest of this paper is structured as follows. After Section 2 introduces the model, Section 3 sets out the optimal tax problem. In exploring the consequences of a binding participation margin, Section 4 discusses, among other things, how labor-market imperfections and the features of the welfare system impact the optimal income tax. Section 5 concludes. The main results are proved in Appendix A. The proofs of some auxiliary results can be found in Boone and Bovenberg (2002), which is the working paper version of this article.

This paper has explored how the income tax system should optimally respond to an exogenously given welfare benefit in the presence of costly labor-market search and nonverifiable skills. We showed that optimal unemployment is determined by the requirement that distortions on the extensive margin balance those on the intensive margin. On the one hand, generous in-work benefits help to alleviate distortions on the participation margin by encouraging more low-skilled workers to actively look for work. On the other hand, such benefits make it more attractive for high-ability agents to mimic lower ability agents, thereby distorting work effort. The government thus faces a trade-off between boosting (low productive) employment and raising work effort of higher skilled workers. If the government lacks information on individual skills, a distorted participation margin in the form of a positive implicit tax rate between nonemployment and employment is therefore the price for combatting poverty while at the same time protecting labor supply of higher skilled workers. A similar trade-off appears in determining optimal retirement schemes. In particular, rather than linking public retirement benefits to the retirement age in an actuarially fair way, the government may want to favor early retirement to aid low productivity individuals suffering from poor health (see Cremer et al., 2002). In the presence of low search costs, low welfare benefits, a concentrated skill distribution and large public spending, the government may find it optimal to employ a progressive tax system providing generous in-work benefits to low-skilled workers in order to induce all agents to search. Such a progressive tax system features a positive marginal tax rate at the bottom. This contrasts with the familiar result from the optimal tax literature that, in the absence of bunching at the bottom, those with the lowest skills should face a zero marginal tax rate (see Seade, 1977). This new result shows that the welfare system and the participation margin may importantly affect the optimal tax system. Social assistance and positive search costs may overturn also the well-known result from the optimal tax literature that marginal taxes should be positive in the interior of the income distribution. We showed that the government may optimally increase in-work benefits with gross income (implying a negative marginal tax on work effort) in order to limit excessive entry into the labor market. In particular, whereas low tax levels combined with positive marginal tax rates at the bottom help to encourage search if labor-force participation of low-skilled workers is taxed on a net basis, negative marginal taxes rates for low-skilled workers help to discourage excessive entry of low-skilled workers if this entry is subsidized. This latter case provides a rationale for minimum wages and hiring and firing costs. The incorporation of labor-market imperfections and the welfare system into a model of optimal nonlinear income taxation enabled us to investigate how these new elements impact the optimal income tax. We showed, for example, that more generous welfare benefits tend to raise marginal tax rates, as the government cuts the average tax burden on low-skilled workers in order to encourage these workers to continue to look for a job. Accordingly, skilled workers finance not only more generous social assistance, but also a lower tax burden of low-skilled workers. In future research, we would like to investigate optimal tax policy if the government can simultaneously set welfare benefits, search obligations, and other categorical social insurance benefits (such as disability benefits based on a signal of skill type). A study of these issues would need to allow for imperfect information on search behavior, household structure, and skill types. In exploring the optimal trade-off between passive welfare benefits and active labor-market policies, we also intend to account for negative external effects of unemployment.