Determinants of the Level and Distribution of Family

A recursive model of the level and distribution of family income in Standard Metropolitan Statistical Areas is formulated and estimated. The model is related to both human capital and job competition models. Then, the impacts of changes in racial, educational, and industrial composition on the level and distribution of income are analyzed. DETERMINANTS OF THE' LEVEL AND DISTRIBUTION OF FAMILY INCOME IN METROPOLITAN AREAS, 1969 The level and distribution of family income in 1969 varied widely 1 across Standard Metropolitan Statistical Areas (SMSAs) in the U.S. Stamford, Conn., had the highest median income, $15,862; Bristol, Conn., had the most equally distributed incomes, a Gini coefficient of .272. Of all the metropolitan areas, McAllen, Texas, was the poorest, with a median income of $4776, as well as the most unequal, with a Gini coefficient of .469. In this paper a model of the determinants of this variation in income levels and size distributions is formulated and estimated. The model emphasizes the role of the industrial structure