Hypothesis tests for dominance in income distributions has received considerable attention in recent literature. See, for example, Barrett and Donald (2003a, b), Davidson…
Hypothesis tests for dominance in income distributions has received considerable attention in recent literature. See, for example, Barrett and Donald (2003a, b), Davidson and Duclos (2000) and references therein. Such tests are useful for assessing progress towards eliminating poverty and for evaluating the effectiveness of various policy initiatives directed towards welfare improvement. To date the focus in the literature has been on sampling theory tests. Such tests can be set up in various ways, with dominance as the null or alternative hypothesis, and with dominance in either direction (X dominates Y or Y dominates X). The result of a test is expressed as rejection of, or failure to reject, a null hypothesis. In this paper, we develop and apply Bayesian methods of inference to problems of Lorenz and stochastic dominance. The result from a comparison of two income distributions is reported in terms of the posterior probabilities for each of the three possible outcomes: (a) X dominates Y, (b) Y dominates X, and (c) neither X nor Y is dominant. Reporting results about uncertain outcomes in terms of probabilities has the advantage of being more informative than a simple reject/do-not-reject outcome. Whether a probability is sufficiently high or low for a policy maker to take a particular action is then a decision for that policy maker.
The methodology is applied to data for Canada from the Family Expenditure Survey for the years 1978 and 1986. We assess the likelihood of dominance from one time period to the next. Two alternative assumptions are made about the income distributions – Dagum and Singh-Maddala – and in each case the posterior probability of dominance is given by the proportion of times a relevant parameter inequality is satisfied by the posterior observations generated by Markov chain Monte Carlo.
The tax evasion phenomenon affects the economic systems of European countries in different ways. The literature shows that individuals provide biased information both to…
The tax evasion phenomenon affects the economic systems of European countries in different ways. The literature shows that individuals provide biased information both to administrative agencies and household surveys. The effects of tax evasion could thus influence the income inequality computed in official statistics.
In this paper, I investigate whether tax evasion generates a bias when inequality indices are computed using household survey data. To achieve this, I apply a parametric model of the Dagum type (three parameters) on the gross personal income of 27 European countries, distinguishing between the self-employed and employees. Subsequently, the parameters computed in the model are used as dependent variables in seemingly unrelated regressions.
I find that for the self-employed, tax evasion tends to reduce inequality as measured by regular wage statistics. Thus, the results reveal that tax evasion distorts inequality indices, generating an underground inequality.
This paper examines the gender differences of expenditure distribution within the last decade in Spain. In particular, the Lorenz dominance is tested using expenditure…
This paper examines the gender differences of expenditure distribution within the last decade in Spain. In particular, the Lorenz dominance is tested using expenditure distributions as approximated by the Dagum model. The sensitivity of the results to some conceptual choices, such as the equivalence scale or the gender reference, is also analysed.
This paper aims to estimate the global income distribution during the nineties using limited information. In a first stage, we obtain national income distributions…
This paper aims to estimate the global income distribution during the nineties using limited information. In a first stage, we obtain national income distributions considering a model with two parameters. In particular, we propose to use the so-called Lamé distributions, which are curved versions of the Sigh-Maddala and Dagum distributions. The main feature of this family is that they represent parsimonious models which can fit income data adequately with just two parameters and whose Lorenz curves are characterized by only one parameter. In a second stage, global and regional distributions are derived from a finite mixture of these families using population shares. We test the validity of the model, comparing it with other two-parameter families. Our estimates of different inequality measures suggest that global inequality presents a decreasing pattern mainly driven by the fall of the differences across countries during the course of the study period that offsets the increase in disparities within countries.
This paper considers a parametric model for the joint distribution of income and wealth. The model is used to analyze income and wealth inequality in five OECD countries…
This paper considers a parametric model for the joint distribution of income and wealth. The model is used to analyze income and wealth inequality in five OECD countries using comparable household-level survey data. We focus on the dependence parameter between the two variables and study whether accounting for wealth and income jointly reveals a different pattern of social inequality than the traditional “income only” approach. We find that cross-country variations in the dependence parameter effectively account only for a small fraction of cross-country differences in a bivariate measure of inequality. The index appears primarily driven by differences in inequality in the wealth distribution.
There has been growing interest in recent years in modelling various poverty‐related issues. However, there have not been many attempts at empirical estimation of best‐fit…
There has been growing interest in recent years in modelling various poverty‐related issues. However, there have not been many attempts at empirical estimation of best‐fit income distribution functions with an objective of subsequent use in poverty focused models. The purpose of this paper is to fill this gap by empirically estimating best‐fit income distribution functions for different household income groups and computing poverty and inequality indices for Sri Lanka.
The authors empirically estimated a number of popular distribution functions found in the income distribution literature to find the best‐fit income distribution using household income and expenditure survey data for Sri Lanka and subsequently estimated various poverty and inequality measures.
The results show that the income distributions of all low‐income household groups follow the beta general probability distribution. The poverty measures derived using these distributions show that among the different income groups, the estate low‐income group has the highest incidence of poverty, followed by the rural low‐income group.
According to the best of the authors' knowledge, empirical estimation of income distribution functions for South Asia has never been attempted. The results of this study, even though based on Sri Lankan data, will be relevant to most developing countries in South Asia and will be very useful in developing poverty alleviation strategies.
This paper deals with poverty decompositions into subgroups defined with respect to intervals of income and the robustness of comparisons of the absolute contribution of…
This paper deals with poverty decompositions into subgroups defined with respect to intervals of income and the robustness of comparisons of the absolute contribution of such groups to poverty. For instance, world poverty estimates by the World Bank often distinguish between the extreme poor whose incomes are lower than $1.25 a day (in PPP terms) and the other poor with incomes between $1.25 and $2.5 a day. Existing dominance conditions can tell whether overall poverty and extreme poverty have declined in a robust manner when comparing countries at two points of time, but they cannot say anything for the contribution of the non-extreme poor to overall poverty. In the present paper we propose stochastic generalized dominance criteria to perform robust poverty ordering when the focus is placed on some interval of the poverty domain. Using generated data based on grouped data from World Bank’s PovcalNet tool, the paper finally investigates whether the robust decline of extreme poverty around the world during the last decades was also accompanied by a decline of the contribution of non-extreme poverty.