Estimating Wal-Mart's Impacts in Maryland: A Test of Identification Strategies and Endogeneity Tests

摘要

In this paper, I estimate the impact of Wal-Mart on labor markets in Maryland. My goal is to compare estimation techniques that incorporate corrections for endogeneity of Wal-Mart's entrance and those that test for and fail to reject exogeneity in Wal-Mart's entrance decision. The instrumental variable approaches I test include those offered by Basker and Neumark et al., and a new test introduced in this paper. I also explain why differences in choice of sample time and location may lead to different findings as to the impact of Wal-Mart. I find that there is weak evidence of endogeneity in Wal-Mart's entrance decision, but that the bias is opposite of that anticipated by most researchers. Most interestingly, there are no significant differences across instrumental variable techniques employed in the literature. As for Wal-Mart in Maryland, that exposure to Wal-Mart leads to lower retail employment (from zero to 414 retail jobs), but significantly higher retail wages (as much as $1.95 per hour) at the county level. I also find that exposure to Wal-Mart has no effect on aggregate labor markets, though aggregate wages are positively affected, although the impact is very small (roughly $160 per year). I recommend that Maryland adopt a policy neutral approach to Wal-Mart.