Introduction à l'économétrie des mesures de pauvreté

The aim of this paper is to provide an overview of the econometric problems that are linked to the measurement of poverty. Starting from a welfare function, we first define inequality measures and then poverty measures which are based either on indices or on thenotion of stochastic dominance. We then show how the choice data definitions (income or consumption spending, equivalence scales, etc...) can modify or invert empirical conclusions when comparing countries or the time evolution of poverty. The statistical treatment of surveys data implies the use of instruments that are simple at the beginning (order statistics, density estimation), but that can become rapidly complex when one has to estimate standard deviations or even more when one wants to test stochastic dominance. Most of statistical analysis concerning inequality and poverty is descriptive. Explicative models are needed for analysing the dynamics of poverty such as dynamic factor models or quantile regressions.Finally serious empirical analysis of poverty requires the use of survey data such as the UK Family expenditure survey or the French enquête sur le budget des familles. The access to these data set for scientific researchers is very unequal, in particular in France.