DECENT years have witnessed an upswing in economists' interest in various aspects of the distnibution of income. The burgeoning of the size of econometric models has also led to inquiries into the distributional aspects of macroeconomic activity. How much do the poor gain from sustained growth, and who suffers relatively during slowdown and depression? Several attempts have been made to measure just what are the impacts of cyclical economic fluctuations on the distribution of income. Studies by Metcalf (1969), Mirer (1972), Schultz (1969), and Thurow (1970) have employed quite different approaches, and have reached differing conclusions on the cyclical sensitivity of income inequality. Clearly, a more general framework of analysis is needed to evaluate these findings in some perspective. One of the most frequently used approaches in such studies has been to characterize inequality in a distribution by a small number of summary measures (such as a Gini coefficient or an income share) and simply regress these inequality indices on such factors as an unemployment rate, a participation rate, and a per capita income measure. More preferable would be an that (i) explicitly lays out a model of the various channels by which macrofluctuations affect the distribution of income and (ii) examines the effects in a disaggregative fashion on individual income levels across a distribution. Such an approach, forwarded in Beach (1976), involved first modelling the behaviour of a set of individual quantile income levels and then expressing disaggregative income inequality measures in terms of these estimated income quantiles. One can then check the reasonableness of estimated inequality behaviour by examining the underlying behaviour of the individual quantile income levels. This article extends this disaggregative to an examination of implied aggregate inequality changes and compares the results with previous findings by Metcalf (1969) and Schultz (1969). It thus attempts to evaluate and integrate a number of disparate findings by building up summary measures of inequality from individual income quantiles. The outline of the paper is as follows. The next section reviews the indirect quantile approach followed here, and presents estimation results for the cyclical behaviour of a set of income quantiles. In section III the implied behaviour of relative mean incomes and income shares is discussed, and a comparison is presented with Metcalf's results. Section IV contains an aggregation of the results and a comparison with Schultz' findings. Then section V examines the behaviour of several alternative summary inequality measures. Section VI summarizes the principal findings and draws some implications.

The paper considers the problem of statistical inference with estimated Lorenz curves and income shares. The full variance-covariance structure of the (asymptotic) normal distribution of a vector of Lorenz curve ordinates is derived and shown to depend only on conditional first and second moments that can be estimated consistently without prior specification of the population density underlying the sample data. Lorenz curves and income shares can thus be used as tools for statistical inference instead of simply as descriptive statistics.

The widely used Cochrane-Orcutt and Hildreth-Lu procedures for estimating the parameters of a linear regression model with first-order autocorrelation typically ignore the first observation. An alternative maximum likelihood procedure which incorporates the first observation and the stationarity condition of the error process is proposed in this paper. It is similar to the Cochrane-Orcutt procedure, and appears to be at least as computationally efficient. This estimator is superior to the conventional ones on theoretical grounds, and sampling experiments suggest that it may yield substantially better estimates in some circumstances.