Spatial Analysis of Rural Economic Development Using a Locally Weighted Regression Model

Article excerpt

This study uses locally weighted regression to identify county-level characteristics that serve as drivers of creative employment throughout the southern United States. We found that higher per capita income, greater infrastructure investments, and the rural nature of a county tended to promote creative employment density, while higher scores on a natural amenity index had the opposite effect. We were also able to identify and map clusters of rural counties where the marginal effects of these variables on creative employment density were greatest. These findings should help rural communities to promote creative employment growth as a means of furthering rural economic development.

Key Words: creative class, locally weighted regression, natural amenities, rural economic development

The post Industrial Revolution period in the United States has been a time of rapidly increasing agricultural productivity. During this time, the economy has gradually moved from one primarily based on agricultural production, to one based heavily on industrial production, and more recently to one based increasingly on service provision. These changes have led to a steady decline in the importance of agriculture to the economic base of rural America. For many geographic areas, the decline has not been offset by equivalent increases in other economic sectors, and the population has followed the jobs into other, generally more urban areas. As a result, the population of rural areas has been shrinking relative to that of urban areas and many rural areas are facing population losses. In fact, more than 25 percent of the nation's non-metro counties experienced a net loss of total population during the 1990s, and more than 85 percent of the U.S. counties that experienced net population losses during the 1990s were rural (McGranahan and Beale 2002). There is some concern that the depopulation of rural areas could gradually erode the ability of many of these communities to provide the public services necessary for their citizens (Huang, Orazem, and Wohlgemuth 2002).

While overall population and economic growth in the southern United States over the past few decades has been strong, averages obscure some disturbing trends. ' For example, a pattern of dualistic development-where rapid economic growth in and around urban areas like Atlanta and Nashville is accompanied by economic stagnation and persistent poverty in more isolated areas-has come to characterize much of the South. This dualistic development tends to increase the disparity between wealthy urban areas and poorer, more isolated rural areas. As a result, the rural South has the highest concentration of persistent poverty in the nation, with 280 of the 340 non-metro persistent poverty counties identified in the 2000 U.S. Census as being located in the South (Jolliffe 2004). Thus, stimulating rural economic development in the South remains a vitally important goal for policymakers at the local, state, and national levels.

Policymakers hoping to promote economic development have often focused on creating employment opportunities. Murdoch (2000) categorized policies designed to promote rural economic development on the basis of whether they promote "horizontal" or "vertical" networks. In Murdoch's terminology, "horizontal" networks link rural spaces to more general, non-agricultural processes of economic change. Policies based on "vertical" networks attempt to develop a rural economy by enhancing the links among various stages of the agro-food sector. The primary focus of the literature on rural economic development has been on "horizontal" networks. For example, rural economic growth has been examined with respect to transport costs (Kilkenny 1998), natural amenities, and quality of life characteristics (DeIler et al. 2001), farm subsides (Kilkenny 1993), sustainable energy for rural development (Byrne, Shen, and Wallace 1998), off-farm work decisions of husbands and wives (Huffman and Lange 1989), and more recently on "creative employment" opportunities (Rosenfeld 2004a, 2004b, 2005). …