Environment and Development Economics 10: 391–434C2005 Cambridge University Pressdoi:10.1017/S1355770X05002275 Printed in the United KingdomNational environmental performance:an empirical analysis of policy resultsand determinantsDANIEL C. ESTYYale Law School and Yale School of Forestry and Environmental Studies,P.O. Box 208215, New Haven, CT 06520-8215, USA. Te l: (203) 432-6256.E-mail: [email protected] E. PORTERHarvard Business School, Soldiers Field Road, Boston, MA 02163, USA.ABSTRACT. Effortsto identify the determinants of environmental policy success atthe national level have largely been anecdotal and case study based. This articleseeks to identify empirically the factors that drive environmental performance asmeasured by levels of urban particulates and sulfur dioxide and energy use per unitof GDP. Although the data are imperfect and causal linkages cannot be definitivelyestablished, the statistical analysis presented suggests that environmental results varynot only with income levels as suggested by the environmental Kuznets Curveliterature but also with both the sophistication of a nation’s regulatory regime and,perhaps more notably, its broader economic and social context. Thus, at every levelof development, countries face policy choices that determine environmental quality inimportant ways. Strong environmental performance appears to be positively correlatedwith competitiveness, putting into question the presumed trade-off between economicprogress and environmental gains. Although preliminary, these results provide evidencethat environmental decision making can be made more data driven and analyticallyrigorous.Concern for the environment exists in every country. But, despitethe emergence of significant pollution control and natural resourcemanagement programs in most nations, little rigorous analysis has beendone to identify the factors that determine whether environmental effortssucceed. Research to date has put forward theories and explored casestudies, but systematic statistical investigations in the environmental arenahave been limited. Within the discipline of economics, considerable focushas been given to the relationship between environmental outcomes andnational income (Shafik and Bandyopadhyay, 1992; Grossman and Krueger,1995; Harbaugh et al., 2000). This Environmental Kuznets Curve (EKC)literature has broadened beyond the income–environment relationship toexamine a variety of additional impacts on environmental performance,including corruption (Lopez and Mitra, 2000), freedom (Barrett and Graddy,2000), inequality (Torras and Boyce, 1998), openness to trade (Suri and392 Daniel C. Esty and Michael E. PorterChapman, 1998; Antweiler et al., 1998), energy (Suri and Chapman, 1998),and the spatial density of economic activity (Kaufmann et al., 1998).In the environmental field, a number of initiatives have been launchedto develop metrics or indicators of environmental performance.1Afewefforts have been made to use regression analysis or other advancedstatistical techniques to explore the relationship between policy choices andenvironmental performance (York et al., 2003; Dasgupta et al., 2001, 2002;Panayotou, 1997). But none of these efforts has examined a large numberof countries (including both developed and developing countries) across abroad spectrum of possible policy determinants.The resulting knowledge gap is unfortunate. Without solid evidenceon how regulatory choices and a nation’s underlying economic andlegal system affect environmental performance, policies are often basedon crude analysis, heated rhetoric, and imprecise concepts such as‘sustainable development’. The lack of systematic data and analysis fuelsthe long-standing controversies over the role of environmental outcomesin sustaining economic growth over the long term (Panayotou and Vincent,1997; Panayotou, 2000), whether environmental gains must come at theexpense of competitiveness (Porter and van der Linde, 1995; Jaffe et al.,1995; Esty and Porter, 1998), and the contours of the optimal environmentalstrategies for developing countries. More sophisticated metrics togetherwith objective ways to gauge the success of environmental policies offera constructive way forward in this highly contentious arena in whichdivergent points of view are strongly held (Esty, 2002).In this paper, we take a step towards developing a set of empiricaltools to identify the key policy options and to test their links to improved1These ‘indicator’ efforts include: the Environmental Sustainability Indexdeveloped by the Yale Center for Environmental Law and Policy and ColumbiaUniversity’s Center for International Earth Science Information Network incollaboration with the World Economic Forum (WEF, 2001; Esty et al., 2005);the Living Planet Campaign sponsored by the World Wildlife Fund (Ricketts,1999); the ‘Weight of Nations’ study conducted by the World Resources Institutewith partners in Germany, Austria, the Netherlands, and Japan (Hammondet al., 1995); initiatives by the Organization for Economic Cooperation andDevelopment (1993, 1998) and the UN Commission on Sustainable Development(1996); the International Institute for Sustainable Development (Bossel, 1999);the informal ‘Consultative Group on Sustainable Development Indicators’(International Institute for Sustainable Development, 1999); the UN EnvironmentProgramme (Bakkes et al., 1994); the European Union (Eurostat, 1999); theWorld Business Council for Sustainable Development (1998, 1999) (Verfaille andBidwell, 2000); and the Boston-based CERES group (1997) as well as a numberof national sustainable development councils and projects. A few studies havebegun to analyze the links between corporate environmental performance andcorporate profitability, most notably the corporate environmental rankings doneby INNOVEST (Dixon, 2002) and by t he Dow Jones Sustainability Index (2000).Some efforts have been made to understand partial determinants of environmentalquality (Panayotou and Vincent, 1997). The analysis presented here builds on priorefforts of Esty and Porter (2000, 2002). Earlier versions of the present analysisappeared in the 2001 Global Competitiveness Report (Esty and Porter, 2001) and(Esty and Cornelius, 2002).Environment and Development Economics 393environmental performance. In particular, using statistical methods, weexamine differences in environmental results as measured by levels of airpollution (looking at