Optimal taxation with gradual learning of types

An important feature of life-cycle models is the presence of uncertainty regarding one’s labor income. Yet this issue, long recognized in different areas, has not received enough attention in the optimal taxation literature. This paper is an attempt to fill this gap. We write a simple 3 period model where agents gradually learn their productivities. In a framework akin to Mirrlees’ (1971) static one, we derive properties of optimal tax schedules and show that: i) if preferences are (weakly) separable, uniform taxation of goods is optimal, ii) if they are (strongly) separable capital income is to rate than others forms of investiment.

Commodity Taxation and Social Insurance

We investigate optimal commodity taxation in a social insurance framework based on Varian (1980). We show that the tax prescriptions in this moral hazard framework are notably similar to those derivetaxation are valid in this setup. We incorporate pre-committed goods - those whose consumption must be decided before the resolution of uncertainty - and show that tax prescriptions are also analogous to the existing literature. The robustness of tax rules across these setups is explained by the relaxation of incentive compatibility constraints.

On the optimality of the Friedman Rule with heterogeneous agents and non-linear income taxation

We study the optimal “inflation tax” in an environment with heterogeneous agents and non-linear income taxes. We first derive the general conditions needed for the optimality of the Friedman rule in this setup. These general conditions are distinct in nature and more easily interpretable than those obtained in the literature with a representative agent and linear taxation. We then study two standard monetary specifications and derive their implications for the optimality of the Friedman rule. For the shopping-time model the Friedman rule is optimal with essentially no restrictions on preferences or transaction technologies. For the cash-credit model the Friedman rule is optimal if preferences are separable between the consumption goods and leisure, or if leisure shifts consumption towards the credit good. We also study a generalized model which nests both models as special cases.

Taxation of couples: a mirrleesian approach to collective households

Esta dissertação apresenta uma abordagem alternativa para o problema de taxação de famílias. Mais especificamente modelamos a decisão familiar com um modelo de barganha de Nash em que os o governo determina de forma ótima as utilidades de discórdia. Demonstramos um Princípio da Revelação para esse modelo de forma a reduzir a classe de mecanismos possíveis, além disso calculamos os ganhos do mecanismo ótimo em relação a outros mecanismos razoáveis por meio de exemplos. Discutimos algumas implicações associadas ao mecanismo ótimo.

The risk-properties of human capital and the design of government policies

Whether human capital increases or decreases wage uncertainty is an open question from an empirical standpoint. Yet, most policy prescriptions regarding human capital formation are based on models that impose riskiness on this type of investment. In a two period and finite type optimal income taxation problem we derive prescriptions that are robust to the risk characteristics of human capital: savings should be discouraged, human capital investments encouraged and both types of investment driven to an efficient level from an aggregate perspective. These prescriptions are also robust to the assumptions regarding what choices are observed, despite policy instruments being not.

Yet another reason to tax goods

The optimal taxation of goods, labor and capital income is considered in a two period model where: i) private information changes through time; ii) savings are not observed, and; iii) savings a§ect preferences conditional on the realization of types. The simultaneous appearance of these three elements cause optimal commodity taxes to depend on o§-equilibrium savings. As a consequence, separability no longer su¢ ces for the uniform taxation prescription of Atkinson and Stiglitz (AS) to obtain. If preferences are homothetic AS is partially restored: taxes are uniform within periods, however, future consumption is taxed at a higher rate than current consumption.

Banning information as a redistributive device

In this paper we analyze the optimality of allowing firms to observe signals of workers’ characteristics in an optimal taxation framework. We show that it is always optimal to prohibit signals that disclose information about differences in the intrinsic productivities of workers like mandatory health exams and IQ tests, for example. On the other hand, it is never optimal to forbid signals that reveal information about the comparative advantages of workers like their specialization and profession. When signals are mixed (they disclose both types of information), there is a trade-off between efficiency and equity. It is optimal to prohibit signals with sufficiently low comparative advantage content.

Redistribution with ex-ante unobserved choice

Introducing dynamics to Mirrlees' (1971) optimal taxation model creates a whole new set of issues that are only starting to be investigated in the literature. When choices are made before one's realizing her productivity incentive constraints ought to be defined as a function of more complex strategies than in the static case. 80 far, all work has assumed that these choices are observable and can be contracted upon by the government. Here we investigate choices that : i) are not observed, andj ii) affect preferences conditional Oil the realization of types. In the simplest possible model where a non-trivial filtration is incorporated we show how these two characteristics make it necessary for IC constraints to be defined in terms of strategies rather than pure announcements. Tax prescriptions are derived, and it is shown that they bear some resemblance to classic optimal taxation results. We are able to show that in the most 'natural' cases return on capital ought to be taxed . However, we also show that the uniform taxation prescription of Atkinson and Stiglitz fails to hold, in general.

Land taxes in a Latin American context

Since Henry George (1839-1897) economists have been arguing that a tax on unimproved land is an ideal tax on efficiency grounds. Output taxes, on the other hand, have distortionary effects on the economy. This paper shows that under asymmetric information output taxes might be used along with land tax in order to implement an optimal taxation scheme in a Latin American context, i.e., where land rental markets are relatively thin, land property provides non-agricultural payoffs and there is nonrevenue objectives of land taxation. Also, the model has two implications that can be tested empirically: (i) there is evasion when schemes based only on land taxes are implemented; (ii) this evasion is more severe for large landholders.

Essays in macroeconomics and public finance

Esta tese dedica-se ao estudo dos sistemas tributários. Eu investigo como um sistema tributário afeta as escolhas dos indivíduos e consequentemente os recursos do país. Eu mostro como um sistema tributário induz as escolhas das pessoas, determinado assim as alocações de trabalho, produto e consumo da economia. No primeiro e segundo capítulo eu examino a taxação sobre os indivíduos, enquanto que no terceiro e quarto capítulos analiso a incidîncia tributária sobre os diferentes agentes da sociedade. No capítulo um, eu examino o sistema tributário ótimo, seguindo Mirrlees (1971) e Saez (2001). Eu mostro como seria este sistema tributário no Brasil, país com profunda desigualdade de renda entre os indivíduos. Ademais, eu investigo o sistema tributário afim, considerado uma alternativa entre os sistemas atual e o ótimo. No segundo capítulo eu analiso o sistema tributário conhecido como sacríficio igual. Mostro como o sistema tribuária derivado por Young (1987), redesenhado por Berliant and Gouveia (1993), se comporta no teste de eficiência derivado por Werning (2007). No terceiro e quarto capítulo eu examino como propostas de reforma tribuária afetariam a economia brasileira. No capítulo três investigo como uma reforma tributária atingiria as diferentes classes socias. No capítulo quatro, eu estudo as melhores direções para uma reforma tributária no Brasil, mostrando qual arranjo de impostos é menos ineficiente para o país. Por fim, investigo os efeitos de duas propostas de reforma tributária sobre a economia brasileira. Explicito quais os ganhos de produto e bem estar de cada proposta. Dedico especial atenção aos ganhos/perdas de curto prazo, pois estes podem inviabilizar uma reforma tributária, mesmo esta gerando ganhos de longo prazo.

The long-run properties of a dynamic Mirrlees' model with aggregate shocks

Nós abordamos a existência de distribuições estacionárias de promessas de utilidade em um modelo Mirrlees dinâmico quando o governo tem record keeping imperfeito e a economia é sujeita a choques agregados. Quando esses choques são iid, provamos a existência de um estado estacionário não degenerado e caracterizamos parcialmente as alocações estacionárias. Mostramos que a proporção do consumo agregado é invariante ao estado agregado. Quando os choques agregados apresentam persistência, porém, alocações eficientes apresentam dependência da história de choques e, em geral, uma distribuição invariante não existe.

Is 'tagging' a rationale for affirmative action in education?

In this paper, we try to rationalize the existence of one of the most common affirmative action policies: educational quotas. We model a two period economy with asymmetric information and endogenous human capital formation. Individuals may be from two different groups in the population, where each group is defined by an observable and exogenous characteristic. The distribution of skills differ across groups. We introduce educational quotas into the model by letting the planner reduce the effort cost that a student from one of the groups has to endure in order to be accepted into a university. Affirmative action policies can be interpreted as a form of ``tagging" since group characteristics are used as proxies for productivity. We find that although educational quotas are usually efficient, they need not subsidize the education of the low skill group.

Yet another reason to tax goods

When, in a dynamic model, choices by an agent : i) are not observed, and; ii) affect preferences conditional on the realization of types, new and unexpected features come up in Mirrlees’ (1971) optimal taxation frame- work. In the simplest possible model where a non-trivial filtration may be incorporated, we show how these two characteristics make it neces- sary for IC constraints to be defined in terms of strategies rather than pure announcements. Tax prescriptions are derived, and we are able to show that uniform taxation prescription of Atkinson and Stiglitz fails to hold, in general. Clean results regarding capital income taxation are not easy to come about because usual assumption on preferences do not allow for determining which constraints bind at the optimum. However, in the most ’natural’ cases, we show that return on capital ought to be taxed.

The optimality of the Friedman rule when some distorting taxes are exogenous

This paper investigates the optimality of the Friedman rule in a two-sector small open economy. That policy prescription is found to be a necessary condition for Pareto efficiency. If a planner can select all conceivable distorting taxes, then, for some initial values of public debt, money balances and foreign assets, it is possible to decentralize a Pareto efficient allocation. If the planner can select only some of these tax rates, then second-best policies may also satisfy the Friedman rule. However, this last result depends on the set of tax instruments the planner can choose from.