Economic integration and convergence: A second decomposition method

• 成果类型: Article
• 署名作者: University of Hawaii System
• 刊物名称: JOURNAL OF ECONOMIC HISTORY
• ISSN/ISSBN: 0022-0507
• DOI: 10.1017/S0022050700023573
• 发表日期: 1999
• 页码: 773-778

摘要

In this Journal, Sukkoo Kim investigates the sources of income convergence and divergence across regions in the United States.' Kim argues that there are two sources of income convergence in regional economies linked by trade: convergence in regional wage rates and convergence in regional industry mixes. Using U.S. regional data on industry mix and worker income by industry, Kim decomposes differences in regional incomes into industry wage and industry-mix components. From his decompositions, Kim concludes that changing regional industrial structures played an important role in causing the divergence and convergence of U.S. regional income per capita in both the nineteenth and twentieth centuries.(2) Kim's procedure for decomposing differences in regional incomes is, however, not unique. Decomposition procedures attempt to isolate the effect of one variable on a second variable by holding the effects of other variables constant. Like index numbers, the results of decomposition procedures often vary depending on which variables are held constant and at what levels. This study implements a second theoretically plausible procedure for decomposing regional incomes and finds that its decompositions frequently differ substantially from Kim's decompositions. Kim's central conclusion, that industry-mix played an important role in regional income convergence and divergence, can also be drawn from the results of the second decomposition, but three of Kim's six ancillary conclusions are not fully supported by the second decomposition analysis.