Using 1980 state Census data, the Gini ratio, Theil index, coefficient of variation, Atkinson measure (four values), and Nelson index as separate measurements of income inequality are regressed with nineteen SES variables traditionally found significant with Gini in past research. The standard deviation of educational attainment is dominant among all independent variables as the strongest predictor of inequality. Mean family income, dominant in prior research, is a distant second in predictiveness. More diversity than commonality is evident among the nineteen SES variables as they predict income inequality measured by the eight different techniques. Copyright 1988 by MIT Press.

Multiple Measurements of U.S. Income Inequality

A recent paper in this REVIEW (LongRasmussen-Haworth, 1977) challenges the view that urban size is a source of income equality. L-R-H develop an econometric model that explains variation in the Gini coefficient of male incomes and family incomes. They conclude that urban size creates a more unequal distribution of urban income (as indicated by a significant negative sign on the SMSA population variable). This paper makes several contributions to our knowledge of urban income distribution. First, unlike previous papers, it develops a theoretical framework for separating the effects of urban size and level of development. Second, it provides a theoretical basis, as well as empirical justification, for the hypothesis that level of development (as measured by urban income level) bears a non-linear relationship to urban income inequality, becoming negative at high levels of income. Third, all authors except L-R-H anticipate inequality to decline while moving up the urban hierarchy (size distribution of cities). This paper makes the point that the urban hierarchyequality hypothesis is partly supported but that the conclusion hinges on the choice of inequality measure. Fourth, virtually all previous research on urban income distribution uses the Gini coefficient as the measure of income inequality. This paper uses three measures to test the sensitivity of conclusions to the choice of inequality measure: the Gini coefficient, the incidence of poverty, and the percentile index. Despite the widespread belief by most economists that efficiency and equity do not go hand-in-hand, the urban size-equality view suggests there is no trade-off between the two. Richardson, for example (1973, p. 53), argues that "inequality decreases as we move up the urban hierarchy."1 Most of the empirical work on state, county, and urban income distribution supports this view (Mattila and Thompson, 1968; Hoch, 1972; Farbman, 1974; Aigner and Heins, 1967; AlSumarrie and Miller, 1967; Conlisk, 1967; Murray, 1969; Frech and Burns, 1971; Burns, 1975; Danziger, 1975). Income inequality, as measured by the Gini coefficient, consistently declines with income level of the state, metropolitan area, or county. The incidence of poverty also appears to decline with urban income level (Omrati, 1968; Richardson, 1973). However, only three studies include both population and income level in their testing (Burns, 1975; Danziger, 1975; and Long-Rasmussen-Haworth, 1977), and only the L-R-H research finds both population and income to be significant variables. Both the theoretical argument and the empirical evidence presented in this paper point to the existence of an efficiency-equity trade-off. Previous research on urban income distribution seems to miss the trade-off because the various issues involved have not been clearly separated. Thus, several basic questions surrounding urban income distribution are investigated in this paper: 1. Does urban size, as measured by SMSA population, have a separate effect from level of development? A theoretical basis for a separate inequality-creating effect of urban size is an expected productivity-agglomeration effect which increases the productivity of skilled labor more rapidly than the productivity of unskilled labor. 2. Can we expect level of development, as measured by urban income level, to have a consistent equalizing effect on urban income distribution? A theoretical argument based on an expected amenity-compensation effect leads us

https://www.jstor.org/stable/info/1926067

Urban Income Distribution and the Urban Hierarchy-Equality Hypothesis

The relationship between city size, urban growth, and the metropolitan income distribution is discussed in this paper. Drawing on basic economic theory we develop the 'monopoly' hypothesis which su...

Income Distribution, City Size, and Urban Growth

A N important and growing interest among scholars is the examination of the relationship of the size distribution of income to city size. Such information is critical to the development of urban growth policies and to our understanding of living costs, poverty, and public expenditures in urban areas. Previous studies of this relationship have been motivated by one of the following distinct, although not mutually exclusive, hypotheses: (1) the occupational and wagestructure of the local labor market will change with increasing city size to cause income inequality either to decrease as a result of rising average incomes (Duncan and Reiss, 1956; Murray, 1969; and Richardson, 1973) or increase via a widening distribution of labor skills (Mathur, 1970; Farbman, 1975);' (2) the functioning of capital markets will improve as city size increases so that investment in human capital rises and the average rate of return is depressed to reduce inequality (Frech and Burns, 1971; and Burns 1976); and (3) the principal beneficiaries of increasing city size and urban growth will be those individuals who possess "monopoly" advantages in the marketplaces, such as landlords and individuals owning enterprises with scale economies or holding important non-duplicative executive and bureaucratic positions, so that the benefits from increasing city size will be unequally distributed and cause the level of income inequality to rise (Haworth, Long, and Rasmussen, 1978). To our knowledge none of these hypotheses have been examined for cities that are partially or totally removed from the influence of SMSA economic regions. However, each argument possesses implications for smaller cities which, when examined, produce results relating to the validity and overall understanding of the hypothesis being presented. For example, the monopoly hypothesis may be relevant to SMSAs but we do not know the city size at which monopoly advantages in the marketplaces become significant in worsening income inequality. The relatively smaller demand for rental properties, more competitive business environment, and lesser importance of executives and bureaucrats will substantially reduce the effect of monopoly advantages in smaller cities. It is therefore quite possible that growth in smaller cities may not exert an independent inequality increasing effect on the distribution of income. Arguments can also be made for modificationis in the human capital hypothesis. While an individual's investment in human capital may substantially define his earnings' potential, imperfections in factor markets and discriminatory employment barriers will influence the earnings that are realized. Furthermore, disequilibrium effects in both capital and factor markets resulting from urban growth (in addition to possible changes in attitudes affecting employment barriers) are likely to vary among cities. Consequently, these additioial considerations can influence the equalizing role which capital markets are presumed to play with increasing city size. Moreover, businesses and labor will be attracted or repelled by existing agglomeration economies that are related to city size. Therefore, changes in the occupational and wage structure are not independent of city size. Increases in population in smaller cities may yield gains from specialization and diversification that permit lower income groups to increase their Received for publication May 21, 1979. Revision accepted for publication December 10, 1979. * Northern Illinois University. The author wishes to thank Barry Field and two anonymous referees of this REVIEW for their helpful comments on earlier drafts of this paper. 1 Citing the labor-supply-oriented model by Newhouse (1971) and the labor-demand-oriented model by Thurow (1975), which' both assign industrial mix a major role in inequality, Danziger (1976) finds that 7 out of the 11 major Census occupational groups are significant in explaining inter-city variations in inequality but was unable to relate city size to inequality. Long et al. (1977), on the other hand, found population size and change in population variables to be positive and significantly related to inequality but failed to relate their finding to any particular hvyothesis.

Barbara B. Murray

Papers

Metropolitan Interpersonal Income Inequality