【正文】
as “unfair” cost advantages and have suggested ways to limit or eliminate them. For instance, the labor and environmental side agreements to the North American Free Trade Agreement(NAFTA) sought to level the “standards” between the United States and Mexico. The side agreements hoped to prevent Mexico, which has lower environmental standards, from being a pollution haven for US and Canadian firms trying to avoid the costs associated with more stringent domestic environmental standards. However, the academic literature has reported very mixed results. Many studies fail to find empirical support for environmental standards to affect either trade or firms’ investment decisions, with only a select few reporting significant effects. A review of the literature is given by Jaffe et al. (1995) and Wheeler (2020), who conclude that there is hardly any empirical support for the existence of a pollution haven effect. More recently, Eskeland and Harrison (2020) find that abatement cost and pollution intensity do not affect foreign direct investment (FDI) into Morocco, C244。te d’Ivoire, Venezuela, and Mexico. Smarzynska and Wei (2020) for a sample of 24 transition countries and Dean et al. (2020) in a study of China find some, though relatively weak evidence of a pollution haven effect, the latter only for investors from Hong Kong, Taiwan, and Macao. An exception is Ederington and Minier’s (2020) study, which finds that environmental regulations significantly affect trade flows. There are a variety of factors, theoretical as well as empirical, that must be considered in a study of pollution haven effects. On the theoretical side, as a country develops and ines rise, several forces are at work simultaneously. Low environmental standards during the early stages of industrial development should attract a disproportionate amount of polluting industries, which is a position effect . Moreover, the increased scale of production should give rise to more pollution. On the other hand, as new factories replace their older and more polluting counterparts, pollution will decrease due to this “technique effect”. In particular, as multinational firms locate new plants in developing countries, they generally do so using stateoftheart technology. Thus, the effect of pollution standards 3 on investment and trade flows depends on the relative strength of position, scale and technique effects (see Copeland and Taylor 2020). On the empirical side, many studies face problems such as unobserved heterogeneity, aggregation bias or possible endogeneity of proxies for environmental stringency (Levinson and Taylor 2020). Since decisions on output, inputs, and pollution abatement are made si