A Keynesian reduced-form quarterly model of regional income determination is constructed and tested for the eight regions of the United States. The results support the view that national economic policy exerts a significant direct influence on a region's export earnings. The findings are in conflict with some monetarist models, in that fiscal policy, as well as monetary policy, strongly influences regional basic economic activity. Also, the findings suggest the desirability of incorporating national policy variables into regional econometric models.

Shift analysis and the technique of entropy indicate an increase in the strength of rural and small-town counties competing with urban areas for manufacturing employment during the study period. Increase in the strength of the most rural counties apparently was confined to labor-oriented industries, but small towns and small cities were successful in attracting industries of varying orientations. Despite the rapid rate of industrialization, rural counties experienced slow or negative population growth; this is attributed to the absence of sizeable employment multiplier effects and to offsetting employment declines in extractive industries. The implication is that considerable potential exists for further rural industrialization.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.. Clark University is collaborating with JSTOR to digitize, preserve and extend access to Economic Geography.The ideas of information theory, including the key concept of entropy, have recently been shown to have potential in the investigation of geographic phenomena [1, 8, 10]. Medvedkov [8] has described some efforts in this direction and has suggested that there may be two aims associated with the use of entropy. The more ambitious aim is to search for a methodology for mathematized geography, and the more modest aim is to assess specific solutions based on entropy measurement as compared with conventional probability approaches. Our goal here is of the more modest level in that we shall demonstrate that entropy is a useful index of the geographic concentration of economic activity. In particular, we shall use entropy and its counterpart, the numbers-equivalent, to test the hypothesis that in the 194-county Tennessee Valley Region there was a significant decline in the geographic concentration of manufacturing activity during the period 1959 to 1968. We shall examine the hypothesis that geographic concentration is less for those industries we have defined as "labor-intensive" than for other types of industries. Further, we shall contend that differences in concentrations in different types of industries are explained, at least in part, by different sets of location factors. Our approach involves the use of asymptotic distribution theory to provide a statistical means for determining whether a particular index of concentration differs significantly from another. ENTROPY AND CONCENTRATIONIn communication theory entropy quantifies the degree of disorder or uncertainty in a system. Consider a set of k categories C1, C2, ..., C17 and a ran-

No abstract