The purpose of this study is to review the previous studies on the concept and measurement of manufacturing diversification, to revise A. Rodgers' method of measurement, and verify it with the manufacturing employment data of Korea.
(1) In broad sens...
The purpose of this study is to review the previous studies on the concept and measurement of manufacturing diversification, to revise A. Rodgers' method of measurement, and verify it with the manufacturing employment data of Korea.
(1) In broad sense, the term "absolute or maximum diversification" has been used to indicate equal employment in all major categories of manufacturing industries, while the term "specialization" refers to the concentration of manufacturing employment in one or a few industries. We have to understand them as a counter-tendency or inversely correlated terms regardless of which is favorable structure, and consequently maximum diversification is equal to minimum specialization and minimum diversification to maximum specialization. Nevertheless, some scholars have denied such self evident concepts and dealt with them as different facets of local economy which does not occupy positions on a single continum of economic structure. For example, J. B. Parr maintains that maximum diversification is not equal to minimum specialization though minimum diversification is synonymous with maximum specialization, because the former is concernd with equal shares of employment while the later corresponds to national structure.
This mistaken usage of the terms arises from the difficulty of defining what is "balanced" industrial structure in any realistic sense. It is insisted that there should be no final answer as far as balanced, ideal or normal industrial structure is concerned; there is the only structure for any case any every time.
(2) Previous studies may be classified into one of three groups, depending upon ther definition of "normal" employment, i.e., absolute diversification ---equal employment shares--, national everage or the others including minimum requirements. As we may not always consider the above mentioned as balanced, ideal or normal state of employment we have to treat such definitition not as the norm but as the benchmark for measuring certain employment structure. As far as industrial structure is concerned, the term "norm" which implies generally a optimum state of reference is the object itself we try to define, but the term "benchmark" is a point of reference from which measurement of any sort may be made. Consequently, the latter is not and object but a means to an end.
(3) Both P.S. Florence and A. Rodgers thought "national average" as a benchmark for measuring employment structure, though the latter considered simultaneously "maximum specialization" for compromise. On the contrary, both R.C. Tress and E.C. Conkling treat "absolute diversification" as a benchmark, though the latter used Lorenz curve for indexing employment diversity. No doubt, their computing processes are different respectively except for A. Rodgers and R. C. Tress. The computed results gained from the four measures applying to 32 cities in Korea show that Tress and Rodgers' indices are not perfectly identical in their magnitudes but in the rank order of the diversities, although they used different benchmarks for measuring.
This means that the rank order of diversification indices in employment structure depends on what process of computing is performed rather than on what benchmark is dealt with. Rodgers and Tress used same formula but different benchma가 considering simultaneously "Maximum specialization" for compromise.
(4) Although Rodgers' measure has bee widely used in defining regional economic structure, there are two weak points in terms of arithmetic context; one is concerned with more diversified structure than national average, and the other with the expression of specialized or diversified facet in employment. A. Rodgers disregarded negative quantity of the indices in computing process when the employment structure of certain city is more diversified than national average. Moreover, his index represents more conveniently the facet of specialized employment rather than that of diversified one. Accordingly, the more employment specializes the more the index increases. This is very unconvenient for one who intends to review the diversification facet of employment structure in certain region because of the inverse tendency rodgers' index represents.
Considering the above mentioned weak points, the writer proposes that Rodgers' refined diversification index of which the above weak points are excluded should be designated as "Adjusted RDI".