Empirical evidence that firms respond to differences in environmental compliance costs among jurisdictions - the industrial flight hypothesis - has been mixed. In this paper we propose a new explanation for this ambivalence which also sheds light on an emerging paradox in overseas investment: U.S. capital outflows appear to be sensitive to rising abatement costs, yet the investment recipients are increasingly countries with comparably stringent environmental standards. Our explanation for this paradox is based on the role of regulatory process costs, usually left out of total compliance costs. Most empirical work has looked for capital movement in response to differential abatement costs, predicting that investment would flow from environmentally stringent regimes to those with more lax environmental regulations, for example, from the U.S. to some developing countries. But little evidence has been found of these pollution havens. If however, at the margin, it is the liability risks, uncertainties and other expenses generated by the legislative and adversarial approach to regulation in the U.S that constitute the most important compliance cost differentials, then we have been looking for U.S. capital outflows in the wrong places. We begin by providing some background on adversarial legalism and transaction costs. Since data are unavailable for directly testing our hypothesis, in Section IV we present evidence in an effort to refute, or at least cast doubt, on the competing industrial flight hypotheses. We find that the share of U.S. direct investment abroad (DIA) in dirty industries, relative to clean industries, is increasing within several OECD countries with relatively strict environmental regulations, but cooperative legal and regulatory regimes. Additionally, Europe has been increasing their worldwide share of dirty DIA from the U.S. At the same time, European shares of clean DIA are falling, suggesting that the increase in dirty DIA is not simply a response to Europe becoming an increasingly favorable investment location for reasons unrelated to environmental regulations. Though our empirical results are necessarily tentative and our analysis limited by available data, we conclude that the policy debate over harmonizing environmental regulations to discourage industrial flight is, at least, incomplete without considering the regulatory and legal institutions that create, monitor, and enforce those regulations.