Eliminating poverty: an empirical approach.

• Author: Faber, Adolphe J-P.
• Position: Report

1. INTRODUCTION

   1.1 At-risk-of-poverty threshold

   Eradicating extreme poverty and hunger has been declared Number One Millennium Development Goal by the United Nations. Rather than extreme poverty, in this paper we aim at an economic measure of poverty, called 'at-risk-of-poverty threshold'. It is a relative measure of the income of a typical household in its immediate environment. The European Council meeting in Laeken in December 2001 has fixed the standard 'at-risk-of-poverty threshold' at 60% of the median income in the area under review.

   Such values are usually calculated at the level of each individual EU member country. There are big differences between rich and poor member states. Luxembourg boosts one of the highest per capita incomes. However a recent study (3) comparing the values for 2008 and 2009 of a composite measure, called indicator of poverty and exclusion UE-2020, based on the above EU-SILC Enquiry concludes that the situation has deteriorated.

   In large countries there are also differences between living conditions among regions, some regions may e.g. be advantaged by a milder climate requiring less heating expenses in winter and no warm clothes. Therefore it may be necessary for large countries to calculate the at- risk-of-poverty threshold region by region.

   For the pragmatic approach, we shall start from the analysis conducted by the EU-SILC (Survey on Income and Living Conditions) on behalf of the Luxembourg Statistical Office, updated for 2008.

   1.2 The natural income distribution

   In Welfare Economics there is a classic demonstration by means of the Edgeworth Box showing that "any Pareto efficient allocation can be achieved as the outcome of competitive market processes, provided that the economy's initial endowment of resources can be redistributed, via lump sum taxes and subsidies, among agents." (4).

   Another illustration of this fact is to be found in a paper by author Brennan Scott Thompson (5). Analyzing a Flat Rate Tax (FRT) model compared with an NIT model including a social transfer called 'demogrant'. The author concludes that the FRT model including an allowance, where no tax is due, can minimize the headcount ratio--a measure of the poverty gap--but not eliminate poverty.

   The author calls such donations to the needy: 'demogrants'. We will simply assume that they are cash subsidies equivalent to the difference between the earned income and the target value of the at-risk-of-poverty threshold. Such a scheme with a fixed 'demogrant' for all earned incomes was proposed at his time by Milton Friedman6 and coined: Negative Income Tax (NIT). If NIT bears no other conditions than residence, it is called Basic Income (Wikipedia). If the 'demogrant' is targeted only at those who need them, we then have a Guaranteed Minimum Income scheme (GMI).

   Rather than granting cash, today there is consensus that this is not the most efficient way. Here are e.g. the conclusions drawn by author John Leach (7),

   "It was argued not so long ago, that the ideal social assistance program was the negative income tax. Social assistance would be delivered, not through a maze of programs and services, but by adding a line to the income tax form. People with low incomes would receive cash transfers, and people with high incomes would pay taxes. Both benefits and taxes would vary with income. This ideal is no longer sustainable. Income redistribution under such a system is limited ... (because of asymmetric information)..., further redistribution can only take place through tools like tagging and targeting. The austere provisions of the negative income tax must be replaced by a network of special cases, rules and restrictions."

   The cost of these 'demogrants' must be matched by an equivalent tax that is supposed to be levied on the higher incomes. Other taxes are possible, like lump sum taxes or taxes on consumption or financial transactions. Let us bear in mind however that not all taxes are efficient and most important, since the benefit would be limited to a minority of the population (8) in a democracy the 'paying majority' would have to be convinced to agree to their being taxed.

   Bearing in mind that the different measures that will finally be retained to overcome poverty must be thoroughly tailored to the needs, for simplification purposes however, we shall assume that a simple cash grant is provided.

   We shall also assume that the domains where either subsidy is paid or tax is raised do not overlap. If this were the case, we will consider net payments.

2. CLASSIC MODELS

   2.1 Theoretical approach to known redistribution schemes.

   Let us call x the earned income and y the net income after tax and social transfer, S the social transfer granted and T the value of the tax. For the purpose of simplification...