How environmental regulations affect productivity of U.S. and Mexican primary iron & steel and aluminum industries

Graduation date: 2006Many economists believe that the slowdown of the productivity growth rates in\ud the U.S. economy in early 1970's was a consequence of the increased level of\ud environmental regulations. The main reason presented in support of this argument is\ud that increased environmental regulations impose an additional burden to the firms,\ud thus increasing the costs of the production of conventional outputs and decreasing\ud productivity growth rates. The opposing view, called Porter hypothesis, claims that\ud environmental regulations may have a positive effect on the performance of domestic\ud firms relative to their foreign competitors by stimulating domestic innovation.\ud Given that the environmental standards are lower in Mexico than in the U.S.\ud we expect to find relatively smaller effect of environmental regulations in Mexico.\ud During the debates of North American Free Trade Agreement (NAFTA), 1994, this\ud difference in environmental regulations levels of the two countries was hypothesized\ud to encourage the migration of U.S. dirty industries to Mexico, thus putting Mexican\ud economy into disadvantage. There are many studies of the effect of environmental\ud regulations on U.S. manufacturing performance, however there is little empirical work\ud showing the effect of environmental regulations on productivity of Mexico and\ud comparing the relative performance of industries of the two countries.\ud In this study we compare the relative performance of U.S. and Mexican\ud primary iron & steel and aluminum industries, RAMA 3411 and RAMA 3423 correspondingly (according to Mexican Economic Census 1975 industrial\ud classification). We estimate primal and dual TFP growth rates based on translog\ud restricted cost function using fixed-effects approach. We found positive effect of\ud environmental regulations on dual TFP growth in all industries. The results support\ud Porter's "win-win" hypothesis