Income and Inequality: The Role of the Service Sector in the Changing Distribution of Income.

Whaples, Robert

For about two decades now, income in the United States has slowly become less equally distributed. The extent and causes of this trend have been hotly debated among scholars, in the press, and, recently, on the campaign stump. At the December, 1992 Economic Summit, the changing income distribution was given almost as much attention as economic growth and health care spending. Leading Clinton administration officials, such as Secretary of Labor Robert Reich, are weighing policies designed to reverse the trend.

Crafting successful policies requires a basic understanding of the underlying causes of this complex phenomenon. The leading explanations focus on shifts in the demand for skilled and unskilled labor in conjunction with changing technology, changing product demand and the globalization of the economy. There are important subplots about demographic shifts (for example, increases in female labor force participation and the percent of single-mother headed households) and measurement errors. Less noticed has been the widening disparity in the distribution of income across different sized communities, as, for example, higher-paying producer services have tended to locate in metropolitan areas, while lower-paying consumer services have been more evenly distributed across space.

These issues are a Gordian knot. It is virtually impossible to remove one thread for examination, since each is securely fastened to the others. Alexander's sword won't suffice, a surgeon's scalpel is required to carefully unravel the interacting causes.

Cathy Kassab's study is an effort to focus on one strand in this knot. Her study, which grew out of a dissertation completed at Pennsylvania State University, examines the impact of changing levels of employment in the service and manufacturing sectors. It considers how growth in the service sector affects the community income level and its distribution, how changes in the industrial mix of communities affect the distribution of income across places, and how the impact of the service sector on income compares with that of manufacturing.

The bulk of the empirical work examines a sample of communities in the Mid-atlantic region. This region was chosen because the transition from a manufacturing-based to a service-based economy occurred there with great clarity during the 1970s, and because the Mid-Atlantic also has a relatively large rural population. Data from the United States Census of Population for 1970 and 1980 are used to construct information on aggregate family income and its distribution for the 642 communities. Data on the types of firms and number of employees present in localities are from Dun and Bradstreet's Marketing Identifiers file.

The data are used in a series of descriptive tables and regressions. The aim is to determine the forces behind changes in aggregate income and its distribution between 1969 and 1978. Unfortunately, one must be wary of the coefficients from the regressions, and, especially, Kassab's interpretations of them. Because income inequality is such a complex problem almost every simplifying assumption made by Kassab obscures one of the important issues.

She assumes that each community is a spatial unit in which residents are functionally interdependent, yet she provides no evidence that her two data sets are properly linked, that individuals work and live in the same community. To simplify, she analyzes only the distribution of income among families, ignoring income earned by "unrelated individuals." But, as her data show, there was a tremendous increase in the proportion of unrelated individuals over this period, and the distribution of this group's income shifted dramatically. Thus, the identities of those in families have also changed, and the groups she compares are not consistent over time. The Dun and Bradstreet data are essential for getting information on employment by sector, but they undercount smaller firms, and lead her to ignore the roles of employment in agriculture, construction, mining, government, and the nonprofit sector. Finally, she does not control for educational levels in any of her analysis. While the community-based approach has some strengths, its major weakness is that it loses sight of the individual.

A second empirical chapter uses County Business Pattern employment data and Bureau of Economic Analysis income and population data for all the counties in the United States. The dependent variable is change in wage and salary disbursements between 1979 and 1988. This income measure is much less complete than is needed.

Using the regressions, Kassab argues that, while growth in the low-wage services is associated with a decrease in aggregate income among centrally located places, it boosts income in places remote from the interstate highway system. "These findings support statements made in the previous chapter regarding the positive impact of the low-wage service sector on the nonmetropolitan economy, and the fact that growth in this sector represents a potential source of economic development." However, this reasoning is questionable, since the regressions cannot determine whether these changes are causes or effects. Kassab implies that supply creates its own demand, that the growth in the low wage consumer service sector was exogenous to the community and boosted its income, when it was probably endogenous, reflecting increased incomes elsewhere in the community. While the purpose of this research is to provide input into policy on local economic development, the argument that isolated rural communities can use tactics such as retail development to push growth, flies in the face of a plethora of other studies and of common sense. At most this is a beggar thy neighbor strategy, with a low local income multiplier.

There are some important findings in the book, especially the regressions analyzing community-level changes in the female labor force participation rate. While Kassab's study generally bolsters mainstream policy implications, such as increased investment in education, it is not the scalpel that was required to unravel the knotty problems surrounding service sector growth, economic development, and income inequality.