Garry F. Barrett

CONSISTENT TESTS FOR STOCHASTIC DOMINANCE

Methods are proposed for testing stochastic dominance of any pre--specified order, with primary interest in the distributions of income. We consider consistent tests, that are similar to Kolmogorov--Smirnov tests, of the complete set of restrictions that relate to the various forms of stochastic dominance. For such tests, in the case of tests for stochastic dominance beyond first order, we propose and justify a variety of approaches to inference based on simulation and the bootstrap. We compare these approaches to one another and to alternative approaches based on multiple comparisons in the context of a Monte Carlo experiment and an empirical example. Copyright The Econometric Society 2003.

Adverse Selection and the Decline in Private Health Insurance Coverage in Australia: 1989-95

The decline in private health insurance coverage over the period 1989–95 is analysed using the ABS National Health Surveys. Individuals’ health status and health risk behaviours are found to be significant determinants of their decision to purchase private health insurance. At a point in time, the pool of the insured is very heterogeneous, with a mix of both good and bad health risks. It is found that the decline in insurance coverage over the period 1989–95 coincided with an increase in the degree of ‘adverse selection’ within the insured population.

Inequality in Male and Female Earnings: The Role of Hours and Wages

The authors decompose annual earnings into hours of work and hourly earnings and analyze male-female differences in earnings inequality using Canadian data. Their results indicate that the larger female inequality in earnings is due to a greater inequality in the distribution of hours of work. The distributions of wages for men and women are either statistically indistinguishable or more equal for women. The authors compare two data points, 1988 and 1981, and find the same structure in the gender comparisons. Also, changes between 1988 and 1981 in earnings inequality are generated from movements in the hours distributions. Copyright 1996 by MIT Press.

Working part time: by choice or by constraint

The selection of workers into part-time jobs and the wages they earn are analysed using the 1989 Labour Market Activity Survey. We focus on the distinction between voluntary and involuntary part-time workers, since involuntary part-time workers earn substantially lower wages than other workers. We find that the selection of individuals into involuntary part-time work is more closely tied to job characteristics than personal or human capital characteristics. The lower wage earned by involuntary part-time workers is largely due to differences in returns and selection effects, rather than differences in endowments.

Statistical Inference with Generalized Gini Indices of Inequality, Poverty, and Welfare

This article considers statistical inference for consistent estimators of generalized Gini indices of inequality, poverty, and welfare. Our method does not require grouping the population into a fixed number of quantiles. The empirical indices are shown to be asymptotically normally distributed using functional limit theory. Easily computed asymptotic variance expressions are obtained using influence functions. Inference based on first-order asymptotics is then compared with the grouped method and various bootstrap methods in simulations and with U.S. income data. The bootstrap-t method based on our asymptotic theory is found to have superior size and power properties in small samples.

The Dynamics of Participation in the Sole Parent Pension

The length of time families spent on the Sole Parent Pension (SPP) is analysed using the FaCS Longitudinal Data Set covering the period from June 1995 to June 1999. It is found that lone mothers, relatively younger and older lone parents, and those with younger children had lower exit rates from SPP. There is significant regional variation in the SPP exit rate, and there is evidence this is related to the implicit subsidy of public housing. Lone parents with greater job attachment had substantially shorter stays on SPP. Program payment levels had a relatively small effect on the length of time on SPP.

The effect of educational attainment on welfare dependence: Evidence from Canada

Abstract The impact of educational attainment on the duration of welfare spells is analysed using a unique data set derived from the administration of the social assistance program in Canada over the period 1986–1993. The empirical analysis includes controls for demographic characteristics, program parameters, labour market conditions and unobserved individual characteristics. It is found that that educational attainment has a much greater impact on the welfare exit rate for women than for men. Additionally, the welfare exit rate for women is more sensitive to family status and program benefits, but less sensitive to the unemployment rate, relative to that for men.

The Asymptotic Distribution of the Generalized Gini Indices of Inequality

This paper presents the asymptotic distributions of the S-Gini and E-Gini relative and absolute indices of inequality. The generalized Gini indices are defined in terms of Lorenz ordinates and results on the asymptotic covariance matrix of Lorenz ordinates are used to derive the large sample properties of the indices. The methods are illustrated by examining changes in the distribution of nondurable consumption in Canada over the 1980s. The authors find that inequality was lower in 1972 and 1990 than in 1984. Over the entire data period, relative inequality was either static or increased slightly, while absolute inequality declined.

Incorporating Covariates in the Measurement of Welfare and Inequality: Methods and Applications

Methods for comparing social welfare and inequality across populations typically involve the entire distribution of economic wellbeing. Conditional analysis requires an estimate of the entire distribution conditional on a large set of covariates. In this paper, we present methods for estimating conditional distributions including flexible parametric, semiparametric and non‐parametric approaches. We demonstrate how to use the statistical properties of the estimators to conduct inference for welfare and inequality comparisons conditional on covariates. Further, we consider how to use the results to perform counterfactual analysis.

Using Engel Curves to Estimate the Bias in the Australian CPI

The Australian consumer price index (CPI) is a Laspeyres price index which is subject to a number of well-known biases. In this article we evaluate the performance of the Australian CPI as a true cost of living index by comparing CPI-deflated food Engel curves estimated with the Australian Bureau of Statistics Household Expenditure Surveys. Our findings indicate that the Australian CPI overstated the change in the general cost of living between 1975/1976 and 2003/2004 by 34 per cent. The CPI was especially inaccurate for single adults and lone mother families. In contrast, the CPI was relatively accurate in measuring the change in the cost of living for two-adult and working families.