Social security, income taxation and poverty alleviation.

• Author: Maldonado, Darío
• Pages: 39(21)

Resumen. En este artículo considero los argumentos que justifican el uso del sistema de seguridad social como parte de un sistema para la redistribución. Asumiendo que el gobierno se preocupa por la desigualdad del ingreso y por la pobreza, estudio el diseño del sistema de seguridad social conjuntamente con el del impuesto sobre el ingreso laboral. El objetivo principal del artículo es mostrar que la seguridad social adquiere mayor importancia como instrumento para la redistribución cuando el planificador se preocupa por la utilidad individual y la pobreza, que cuando sólo se preocupa por la utilidad.

Palabras clave: bienestar, pobreza, impuestos a la renta, seguridad social.

Clasificación JEL: H21, H23, H51, H55.

Abstract. In this paper I consider the normative arguments that justify a public social security system as a redistributive device when government is concerned with individual utility and poverty. Redistribution can be done using social security, income taxation or both. The main objective of this paper is to show how the consideration of a planner that cares about poverty and utility increases the desirability of social security with respect to the case when the planner only cares about utility.

Key words: welfarism, poverty, income taxation, social security.

JEL classification: H21, H23, H51, H55.

1. Introduction

   The problem of redistribution can be seen from three different domains: positive, normative, and political. From the positive domain the concern is to provide potential explanations of inequality. From the normative domain one looks for explanations about why redistribution should occur and which are the best instruments for redistribution. Lastly, the political domain attempts to understand how the decisions regarding redistribution are taken in a society. This paper focuses on the normative domain of redistribution; naturally, the positive domain must also be considered since it is needed as a framework to study the reaction of workers to policies.

   The positive domain is taken as given. Individuals are assumed to differ in wage and probability of getting sick, both of which affect consumption. Since I will assume that individuals' welfare is fulfilled through consumption, heterogeneity generates inequalities that may justify government intervention. Individuals with a low wage will have a low labor income and hence a lower consumption level. Heterogeneity in the probability of sickness affects welfare since it also affects consumption. If individuals are risk averse, and insurance markets are complete, they will buy full insurance. This means high-risk individuals (those with high probability of illness) will face a higher cost of insurance and will have a lower consumption than low-risk individuals.

   The setting in this paper is one in which the social planner has a plurality of objectives. In economics, it is traditionally assumed that the central planner determines the optimal allocations in a way that only considers the utility level of the agents in the economy. Generally welfare is measured through some aggregation of the utility of the population. This approach has been criticized for being limited in terms of information; indeed, Atkinson (1995) and Sen (1999) have argued that welfarism (as they call the traditional approach in Economics) leaves aside information that may be important for policy making. In this paper I will deal with the case in which the planner cares about individual utility levels but also about the poverty level.

   My objective in this paper is to examine under which conditions social insurance is a useful mechanism to alleviate poverty when the central planner is not only concerned about income distribution but also about poverty. The distinction between income distribution and poverty is important. Usually, the introduction of concerns about income distribution is done respecting individuals' own perception of wellbeing; this can be regarded as a subjective measure of wellbeing. However, wellbeing may also have objective dimensions. The introduction of poverty concerns aims at considering these alternative dimensions. Since the economic literature recognizes taxation as a powerful instrument to redistribute, attention is centered in what can be done by social security in terms of redistribution (besides what it is already done by optimal taxation).

   To the best of my knowledge, this problem has not been treated extensively in the literature. Blomqvist and Horn (1984), Rochet (1991), Cremer and Pestieau (1996) and Henriet and Rochet (1998) have looked for the justification of social security in an optimal taxation context. All these authors work on welfarist grounds and in their works individuals differ in two aspects: wage and the probability of falling ill.

   Earlier works dealing with problems of taxation in non-welfarist settings include Kanbur, Keen and Tuomala (1994) and Wayne (2001). The first solve the non-linear income taxation problem for a planner that is concerned only about the poverty level. Wayne (2001) uses the approach suggested by Atkinson (1995), in which the social planner is concerned about both types of objectives: welfarist and non-welfarist. This is the approach followed in this paper.

   I will proceed as follows: the second section establishes the framework for the analysis of the normative problem; particular attention is given here to how poverty is understood and why a central planner should care about it. The third section provides a general discussion about redistributive policy based on a general overview of the related literature. The fourth section works along the lines of Cremer and Pestieau (1996) and Wayne (2001) to show that social security is justified (in the presence of an optimal income taxation system) when the planner has a plurality of objectives. Two types of settings are established for the social security system. In the first one, social security is universally provided by a planner who chooses the rate of coverage of the system that is equal for all the population; in the second one the planner has the possibility to offer different coverage to poor and non-poor. The fifth section concludes.

2. Poverty and welfare issues

   The standard approach in the economic literature is welfarist. According to this approach, the central planner aims to maximize a welfare function that depends upon individuals' utility levels. Generally, social welfare is measured through the Bergson-Samuelson welfare function that aggregates individual's utility functions into one measure of social welfare. This function is assumed to satisfy non-paternalism, Paretian property and aversion to inequality. The non-paternalism features of this function are introduced through the fact that the function only takes into account utility levels, in particular no other information that may be important from the social planner's viewpoint is taken into account. The Paretian property states that if every individual prefers one of two alternatives, the social planner should also prefer that alternative. This is recognized by using social welfare functions that are increasing in all the arguments. The aversion to inequality is introduced by using concave social welfare functions, where the degree of concavity reflects the degree of aversion to inequality.

   The central planner is thus worried only with the perception that individuals have about their own welfare. In a society consisting of I individuals, the central planner's objective function is of the type W([U.sub.1], [U.sub.2], ..., [U.sub.I]), where [U.sub.i] is the utility function of individual i. A particularly important social welfare function that has been extensively used in the economic literature is the utilitarian social welfare function, which is the sum of individual utilities. This last type of social welfare function serves to illustrate an important point. The planner can be concerned about inequality of different variables. If the planner is utilitarian, it is not concerned with inequality of utility (or welfare) but it may still be concerned with inequality of consumption if utility is concave.

   Several criticisms can be made to this welfarist approach. Perhaps, the most relevant here is that of Sen (1999) and Atkinson (1995). These authors say that the welfarist approach has informational problems, as the concentration on individual's perception leaves aside information that may be relevant for policy making. One example of the type of information that could be important for policy decision-making and is not taken into account by the welfarist literature is that of freedom. Another example, the one on which attention is concentrated here, is the limitations of the welfarist case when the existence of poverty is recognized.

   Another criticism usually made to the welfarist approach is that it neglects inequality in other dimensions or variables (Sen, 1980 and 1983). Well known examples about the consequences of welfarism when individuals differ in marginal utility of consumption illustrate this type of concerns. The welfarist approach may end up justifying repugnant conclusions like justifying very unequal societies just because there is a group of individuals that has a higher marginal utility of consumption. Regarding this line of critique, I must say that viewing the problem as that of the correct information for social evaluation can lead to clearer conclusions than regarding it as that of the right dimension for the concern about inequality.

   There are several arguments that provide a justification of why a central planner should worry about poverty. The two most important are those that see poverty as a problem of subsistence as well as an externality over the society as a whole. Seeing poverty as the lack of subsistence means is the most common approach. The subsistence concept has to be understood in a wider sense than the biological one. This is something that has been...