Thanks for the link.

The author brings a very good point which is that modern economies have shipped their dirty manufacturing jobs overseas and that Kuznets curves need to be corrected for trade (the consumption-CO2 curve above is taking into account CO2 related to imports in goods and the other one is not). What I suspect, is that most of the published Kuznet curves are biased production-Kuznet curves which are not trade corrected.

FMagyar,

Thanks for the link. I read through some of the paper (rather quickly) and found a few concerns with his regression technique, but nothing major.

FMagyar and Sam,

The paper takes into account inter-state commerce, but what about international commerce? It was my understanding that the U.S. showed EKCs for some air pollutants because the industrial processes associated with those pollutants moved to other countries, not states, e.g. to Mexico.

David, Sam,

First a disclaimer, I claim no expertise in any of this and though I did read the paper, I'm only an interested layman. However this comment by the author did catch my attention.

By characterizing the income-CO2 relationship for the U.S. states, this analysis can also help inform our understanding of greenhouse gas emissions in an international context. First, illustrating the economic dynamics of per capita CO2 for high-income states provide evidence of what may occur for countries as they achieve advanced stages of development. Second, the potential role of trade in emissions-intensive goods (e.g., electricity) in the income-CO2 relationship may be valuable for other regions of the world that may share similar characteristics to the U.S. states. For example, the European Union, with converging incomes, policies, and institutions and substantial cross-border trade in emissions-intensive goods and energy may follow similar production- and consumption-based CO2 trends as the U.S. states. Third, if trade in emissions-intensive goods is as important in the international context as for the U.S. states, then this work suggests that studies that attempt to forecast carbon dioxide emissions may produce biased results without attempts to correct for trade.The next section briefly reviews several key hypotheses of the environmental Kuznets curve literature as it relates to greenhouse gas emissions. The third section provides an overview of the data used in this paper, including the novel state-level carbon dioxide emissions dataset constructed by the author.