Spatial Dynamics of the Livestock Sector in the United States: Do Environmental Regulations Matter?
Herath, Deepananda, Weersink, Alfons, Carpentier, Chantal Line, Journal of Agricultural and Resource Economics
This study examines the factors affecting state annual share of national inventory for each of the hog, dairy, and fed-cattle sectors using data from the 48 contiguous states for 1976 to 2000. The paper develops a state-specific, time-series environmental stringency measure and introduces instrumental variables to control for the possible endogeneity bias between livestock production decisions and regulatory stringency. The results indicate that differences in the severity of environmental regulations facing livestock producers have had a significant influence on production decisions in the dairy, and particularly the hog sector.
Key words: environmental regulation stringency, fixed-effects model, instrumental variable, livestock production, location choice, panel data analysis, pollution havens
The industrialization of the North American livestock sector has been associated with a geographic concentration of production in fewer regions and a shift in production to areas with little prior livestock experience. One of the reasons may be the increasingly important role of the processing sector and the integration of this sector back into production (Ogishi and Zilberman, 1999). Processing plants operating under economies of size are becoming larger and fewer, and scattered throughout the country with clusters of livestock farms around them (Apland and Anderson, 1996; Abdalla, Lanyon, and Hallberg, 1995). Such clusters tend to move to localities with better natural endowments, labor market conditions, and business environment due to agglomeration economies or tax policies (Roe, Irwin, and Sharp, 2002).
Changes in the spatial distribution of livestock production also may be directly affected by differences in the stringency of environmental regulations across administrative regions. A disparity in regulatory stringency among states arose in the 1980s when the federal government delegated the function of devising regulatory regimes to state authorities (Kraft and Vig, 1994; Lester, 1994; Levinson, 2000). As a consequence of the possible differences in these individual regimes, a potential emerged to foster the creation of "pollution havens," whereby lenient regulations in some regions might attract livestock producers to build their facilities in such localities. For example, Martin and Zering (1997) argue that large-scale intensive pork production has shifted to southern states such as North Carolina and Arkansas because "environmental regulations, zoning regulations, and anti-corporate farming regulations did not present insurmountable barriers to siting and building production units and processing plants in the region" (p. 49). By introducing or maintaining lax environmental regulations relative to competing regions and allowing tardy enforcement of those regulations, a region could lure "dirty" industry investments, which are important in employment creation and regional economic development (Fredriksson and Millimet, 2002; Kunce and Shogren, 2002; Levinson, 2000; Jafee et al., 1995). If regional and state governments really do engage in a race to the bottom, certain regions would have an inefficiently high number of concentrated animal feeding operations (CAFOs). Because the assimilative capacity of the environment is deliberately undervalued in a region where a race to the bottom has occurred, the heavier concentration of livestock operations in that region may pollute at a level higher than the socially optimal, and at a greater cost to the local community.
Despite the claim of its importance, the relevance of the pollution haven hypothesis in describing the relationship between environmental stringency and changes in regional livestock production has not been established. While the hypothesis has been tested for aggregated species (hog, beef cattle, dairy, and chicken) based on standard animal units (Park, Seidl, and Davies, 2002), for hog operations (Roe, Irwin, and Sharp, 2002; Metcalfe, 2001; Mo and Abdalla, 1998), and for dairy operations (Osei and Laxminarayan, 1996), the results are inconclusive. …