The Impact of County Level Factors on Obesity in West Virginia
Anura Amarasinghe, Cheryl Brown, Gerard D'Souza, and Tatiana Borisova
Research Paper #2006-14 / view paper with Acrobat Reader (pp. 28, 88 K)
Abstract: A panel estimation of county prevalence of obesity indicates that while the percentage of the population with a completed college degree and the number of food stores available per thousand population are negatively and significantly correlated to county obesity rates, mean commuting time, average annual wage and the total number of business establishments per thousand population positively and significantly contribute to obesity. Educational attainment that raises both human and social capital, as well as changes in the built environment can play a vital role in controlling obesity in West Virginia (WV).

A Spatial Analysis of Obesity in West Virginia
Anura Amarasinghe, Gerard D'Souza, Cheryl Brown, and Tatiana Borisova
Research Paper #2006-13 / view paper with Acrobat Reader (pp. 38, 2281 K)
Abstract: A spatial panel data analysis at the county level examines how individual food consumption, recreational, and lifestyle choices ― against a backdrop of changing demographic, built environment, and policy factors ― leads to obesity. Results suggest that obesity tends to be spatially autocorrelated; in addition to hereditary factors and lifestyle choices, it is also caused by sprawl and lack of land use planning. Policy measures which stimulate educational attainment, poverty alleviation, and promotion of better land use planning and best consumption practices (BCPs) could both reduce obesity and result in sustainable development of regions where obesity is prevalent and the economy is lagging.

The Influence of Socioeconomic and Environmental Factors on Health and Obesity in Rural Appalachia
Anura Amarasinghe, Gerard D'Souza, Cheryl Brown, and Hyungna Oh
Research Paper #2006-12 / view paper with Acrobat Reader (pp. 33, 156 K)
Abstract:A recursive system of ordered self assessed health (SAH) and a binary indicator of obesity were used to investigate the impact of socioeconomic and environmental factors on health and obesity in the predominantly rural Appalachian state of West Virginia. Behavioral Risk Factor Surveillance System (BRFSS) data together with county specific socioeconomic and built environment indicators were used in estimation. Results indicate that an individual’s risk of being obese increases at a decreasing rate with per capita income and age. Marginal impacts show that as the level of education attainment increases, the probability of being obese decreases by 3%. Physical inactivity increases the risk of being obese by 9%, while smoking reduces the risk of being obese by 14%. Fruit and vegetable consumption lowers the probability of being obese by 2%, while each hour increase in commuting time raises the probability of being obese by 2.4%. In addition, individuals living in economically distressed counties are less likely to have good health. Intervention measures which stimulate human capital development and better land use planning are essential policy elements to improving health and reducing the incidence of obesity in rural Appalachia.

An Optimal Depletion CGE Model: A Systematic Framework for Energy-Economy Analysis in Resource-based Economies
Hodjat Ghadimi / Research Paper #2006-11 / view paper with Acrobat Reader (pp. 34, 105 K)
Abstract: Numerical economic models of energy fall into two general categories: models analyzing within energy sector issues and models examining the interaction between the energy sector and the rest of the economy. The first category are mostly partial equilibrium models with a very detailed and disaggregated representation of the energy sector. Although very useful for sector planning purposes this class of models essentially neglect the interdependence of the energy sector and the rest of the economy. The second category, appropriately called energy-economy interaction models, are multisectoral and general equilibrium models focusing on the relationship between the energy sector and the rest of the economy. These models offer a rich economy-wide picture but are not as detailed as the first category in their specification of the energy sector. For energy-economy interaction analysis a number of models have been employed, including input-output, macro-econometric, and computable general equilibrium (CGE), as well as hybrid of these types. With advances in computation capabilities, however, CGE models have become the standard tool and dominate the mainstream of the economic discipline. The model presented in this paper belongs to the optimal depletion category of computable general equilibrium models. It is an optimization model that solves the inter-temporal depletion problem subject to workings of a multi-sector market economy, where relative prices play a crucial role. Such a formulation establishes general equilibrium linkages between the optimal depletion of the resource and the rest of the economy and thus it provides a systematic framework to analyze energy-economy interactions in resource-based economies.

Identifying Spatial Clusters within U.S. Organic Agriculture
Cheryl Brown and Daniel Eades / Research Paper #2006-10 / view paper with Acrobat Reader (pp. 51, 1,963 K)
Abstract:The market for organically produced products has experienced rapid growth in recent decades; however, this growth has not been distributed evenly across the country instead concentrating in certain regions. Employing measures of spatial concentration and association we identify those counties in which organic production is clustered or represents a proportion of the agricultural economy greater than what would be expected by national trends. Results show that spatial clustering of organic agriculture does exist based on data from the U.S. Census of Agriculture on organic farms, acreage, and value of sales. Counties with the largest location quotients for organic production were most often located in the western U.S., especially California, Washington, and Oregon, the Great Plains states, New England, and in some cases, select counties within Mid-Atlantic States. Organic production clusters as measured by the local Moran’s I statistic followed a similar pattern, clustering primarily in the western U.S. with additional High/high clusters found in the Great Plains, upper Midwest, and areas of New England. When these values were adjusted to represent organic agriculture’s share of a county’s total agriculture, central cluster counties were most likely to be found in New England. Results describing the correlation between organic support establishments and production within identified clusters suggest that organic operations in California and New England may be following different marketing strategies that promote or reduce the likelihood of identifying input-output relationships within these clusters.

County-Level Determinants of Local Public Services in Appalachia: A Multivariate Spatial Autoregressive Model Approach
Gebremeskel H. Gebremariam, Tesfa G. Gebremedhin, and Peter Schaeffer
Research Paper #2006-9 / view paper with Acrobat Reader (pp. 29, 236 K)
Abstract: A multivariate spatial autoregressive model of local public expenditure determination with autoregressive disturbance is developed and estimated in this paper. The empirical model is developed on the principles of utility maximization of a strictly quasi concave community utility function. The existence of spatial interdependence is tested using Moran’s I statistic and Lagrange Multiplier test statistics for both the spatial error and spatial lag models. The full model is estimated by efficient GMM following Kelejian and Prucha’s (1998) approach using county-level data from 418 Appalachian counties. The results indicate the existence of significant spillover effects among local governments with respect to spending in local public services. The OLS estimates of the conventional (non spatial) model of local public expenditure determination and the corresponding maximum likelihood estimates of the spatial lag and the spatial error models are also presented for comparison purposes. The GMM estimates are found to be more efficient.

next page. . .