Dr. Martin Zwick

Date of Award


Document Type


Degree Name

Doctor of Philosophy (Ph.D.) in Systems Science


Systems Science

Physical Description

4, xiii, 200 leaves: ill. 28 cm.


Diversification in industry -- Oregon, Diversification in industry -- United States, Unemployment -- Oregon




The purpose of this study is to investigate some widely-held assumptions regarding the value of diversification as an economic strategy. It has ofien been suggested that economic diversity enhances economic performance, either by promoting higher levels of economic well-being or by improving the ability of regions to cushion the adverse effects of economic cycles. This is the conventional wisdom, but it has not been adequately tested, although some attempts have been made to relate measures of diversity to other economic indicators (e.g., Rodgers, MacLaughlin, Conkling). The current study explores this particular issue, and the results obtained should be of interest to economists, regional scientists, and development planners and policymakers. Shannon's entropy function, applied to the distribution of employment in different economic sectors, was used as an index of diversity. This measure allows not only comparison of changes in diversity over time, but also, through its decomposition properties, a means of analyzing the nature of such changes. Economic performance was assessed in terms of unemployment and per capita income, considered in four ways: the level of the variable, its rate of change over time, the degree of instability of the level, and the degree of instability of its rate of change. Eight hypotheses were formulated and tested with data from the counties of Oregon for the ten-year period from 1972 to 1981. To provide a comparative perspective for the Oregon investigation, a U.S. study was also conducted for the same period. Calculations of both studies revealed diversity to be negatively but very weakly correlated with unemployment; the Oregon finding, however, did not quite satisfy the 5% significance standard used throughout this research. While a weak positive association was found between diversity and per capita income of Oregon counties, a larger negative association was observed between the two variables in the U.S. study. These results can be explained either as an effect of differing levels of geographic aggregation or in terms of differences among the particular specializations of low diversity counties and states. For Oregon, relations between the variables for nonrecession years were stronger than for recession years. The study further showed that diversified counties of Oregon were more stable in unemployment and per capita income and showed lower rates of growth of unemployment and higher rates of growth of per capita income than the more specialized counties. None of these associations, however, was particularly strong. For the U.S. study, no evidence was found for any relation between diversity and either growth rates or stability. In general, correlations between diversity and income-based measures were larger than between diversity and unemployment-based measures; also, percentage changes associated with differences in diversity were considerably greater for the income-based measures. Although expected patterns of relationship were thus found to hold, if weakly, for the counties of Oregon, comparison with the national study suggests that results may not be generalizable to other, especially larger, geographic units. Whether diversification is useful for regional development depends at least partially on the specific character of the industries in the region's economy.


Portland State University. Systems Science Ph. D. Program.

Persistent Identifier