May 09, 2007
Dissertation
Essays on Environmental and Natural Resource Economics: Introductory Chapter
Informed policy circles no longer question whether to use economic instruments to solve environmental problems. Domestic European debates on global climate change focus on whether to use carbon markets or taxes. Even calls for banning incandescent light bulbs are wrapped in arguments of induced technological change and similar economic language. That is part of the problem. Economics is used to justify virtually everything, making it an extremely powerful tool – and increasing the stakes for economists producing environment-related research.
The United States imports around 10 million barrels of oil per day. That number is known and of much concern in the energy and national security communities. It also forms the basis of many studies concerned with macroeconomic effects of foreign energy dependency. Other important research areas in economics are the “pollution haven hypothesis,” based on the fear that dirty industries may migrate to poorer countries in search of lax environmental standards, and the “environmental Kuznets curve,” the conjecture that pollution follows an inverse-U-shaped pattern in relation to income. Demonstrated anecdotally, large-scale studies generally find no support for the pollution haven hypothesis and empirical results for the environmental Kuznets curve are by and large fickle. Research leading to my first paper shows that all three areas may have overlooked an important aspect of the data.
In addition to direct oil imports, the United States imports 0.8 million barrels embedded in manufacturing goods. That number has important national security implications. It also enables a direct way of studying both the pollution haven hypothesis and the environmental Kuznets curve. Relatively richer countries consume more energy-intensive products than they produce. Trade accounts for the difference. Future research needs to show whether that translates into a large-scale pollution haven effect. But at the very least, it shows the importance of choosing the correct model and data, especially since the data are not complicated. Careful accounting can produce the necessary figures and highlight important yet previously overlooked details.
Conventional wisdom says that negative environmental externalities dampen growth. Sick workers are less productive. That is certainly true and an important problem to study. Numerous economic papers have made seminal contributions in this area. The general conclusion seems clear: pollution hampers growth. The second chapter of my dissertation indicates that the opposite could be the case as well. Measured output might grow because of environmental externalities: pollution decreases the availability of previously free environmental services such as clean air. Households satisfy their demand by purchasing air purifiers. This transaction increases economic activity: GDP goes up. But households need to work more to pay for previously free environmental services: welfare goes down. In many ways, this is a problem of mismeasured welfare. A proper welfare indicator, such as comprehensive or “green” Net National Income (NNI) would decrease in this economy, but I do not argue for scrapping GDP altogether. Economic activity goes up, as it should. Welfare in this model does not keep pace with output because people value the environment and their leisure. Green NNI would go down.
One hopes that this issue has not been studied before because broad empirical investigations deemed the effect negligible, but I am afraid one reason is mathematical tractability. The model is extremely complicated, even by growth theory standards, and leaves much room for future theoretical work. But that is not the highest research priority. Instead, further work ought to investigate this phenomenon empirically. An important approach might be to look at natural experiments or case studies where pollution has sparked the adoption of new products or services to defend against negative environmental impacts. While important as a first pass, it will always be possible to find particular cases where consumer products replaced free environmental services. It would be particularly valuable to demonstrate the relative importance of negative externalities in economic growth by using cross-country or cross-regional growth regression, the tools of empirical growth economists, or engage in a large-scale accounting exercise in an attempt to account for defensive expenditures in national income accounts.
My third paper, joint work with C.-Y. Cynthia Lin, underscores some of the same points of theoretical modeling and empirical investigation. In many ways, this paper has the typical format of an economic paper. (It is also the only one of the three papers that has already been accepted for publication in the Journal of Environmental Economics and Management, the top field journal.) The paper revisits the basic Hotelling model of natural resource extraction, which uses the arbitrage condition to show that the value of natural resources in the ground rises at the rate of interest. The basic model assumes no technological progress, no stock effects, no new discoveries, no demand changes, and no market imperfections. None of these assumptions is realistic. We relax the first two and calibrate the model to fit data from 1970 through 2004. I feel fortunate to have co-authored this paper but am fearful that it taps into the same potential fallacies I highlight in my first two chapters.
The modeling choice was largely driven by mathematical feasibility. The choice of time periods was entirely driven by data availability. Our model – like most others looking at the same phenomenon – is silent as to whether the two theoretical extensions are indeed the most important ones in the list. Putting full faith in the model, one can conclude that technological progress had been a powerful force, keeping resource prices constant over long periods of time. Another equally potent force might have been new discoveries. We acknowledge as much throughout the paper, but by virtue of our model’s focus on technological progress are not able to say much more. The same goes for the empirical analysis. Since 2004, prices for most minerals have increased dramatically, defying our observation that prices have remained constant over long periods of time. Is it because technological progress has suddenly stopped? That is what our model seems to suggest. Or does it relate to a host of other factors including rapid demand increases in China and India, which lay outside of our model?
Economic work on the environment wields enormous influence. My brief experience writing for the editorial board of the Financial Times has shown me how decisions with significant public policy consequences are often based on a single economic study. Striving to have one’s work be as inclusive as possible is not merely an academic exercise. Being able to draw conclusions unbound by data availability and modeling choices should be the starting point, not the end result of good economic research in general. It is all the more important in an area where economists answer questions linked to a system much larger than the one we are used to studying. After all, in the words of Senator Tim Wirth, “the economy is a wholly-owned subsidiary of our environment.”
Full dissertation.
Posted by Gernot Wagner on Wednesday, May 09, 2007. 
