The Impacts Of Agglomeration Economies And Market Access On Firm Growth: An Empirical Assessment Of Food And Beverage Manufacturing In New York State
From 1998 through 2007, total manufacturing employment declined 29% in New York State, while food manufacturing employment actually grew slightly (0.96%). Policymakers wishing to retain or develop new manufacturing jobs may find that public support for agribusiness manufacturing is a useful method for creating and retaining manufacturing employment, especially in rural areas. Moreover, the establishment of a healthy food manufacturing sector can be beneficial to local agricultural producers by increasing the local demand for raw agricultural inputs. To help guide policymakers, agricultural economists have investigated the factors affecting the location decisions and growth of agribusiness manufacturers. To further the understanding of these factors, a comprehensive survey of food and beverage manufacturing establishments in New York State was conducted in 2009. The objectives were to investigate the effect of various aspects of the business environment within the state, identify the effectiveness of various firm-collaborative strategies and public policy interventions on improving firm performance, and collect plant-specific information on current and expected growth across several metrics. The unique plant-level data contains nearly 400 responses, and encompasses a range of firm sizes and industry sectors, allowing for the evaluation of important growth factors differentiated by firm size and sector. This thesis first provides a detailed analysis of the survey results to identify relationships based on firm size and industry sector. This includes a principal components analysis of business environment factor ratings to identify benefits provided by New York State's business environment, as well as impediments to growth. Then, a growth analysis is presented to estimate the effects of both firm-level and location-specific factors (including labor supply, markets, clustering, and firm characteristics) on firm-level measures of revenue growth. This work contributes importantly to the literature in the following ways: (1) The unique dataset used in this analysis provides firm-level measures of growth. While past studies have examined spatial factors affecting new plant location decisions, we are instead able to examine how these spatial factors affect the actual performance of existing plants. (2) Past studies have attempted to examine the agglomeration benefits derived from firm clustering. This study models upstream, within-stream, and downstream firm clustering simultaneously to compare the relative benefits of each. (3) To account for the heterogeneous nature of the data that includes representation from both rural and urban processing plants, within-stream firm clustering effects are allowed to vary across urban and rural areas. Previous studies did not account for differences in urban and rural clustering. We propose that differences in firm behavior and market environments between urban and rural firms may create different agglomeration effects. (4) We test for potential endogeneity of our agglomeration variable and use instrumental variables estimation to correct for potential endogeneity. While Jaenicke et al. (2009) used survey-derived variables to instrument for clustering, we instead use county-level historical data as instruments. We find evidence of: (1) positive effects from firms locating near upstream and downstream clusters; (2) different effects from within-stream clustering between urban and rural areas; (3) negative effects of within-stream clustering in rural areas; and (4) evidence of little endogeneity problems in model estimation.
dissertation or thesis