Wednesday, July 10, 2013
Weighing Short Term Knowns with Long Term Unknowns in Environmental Policy - a Law and Economics Perspective
Over the next two weeks I will be participating in George Mason University School of Law's Law and Economics Center Economics Institute for Law Professors. Yesterday, Joshua Wright gave a fascinating presentation that included discussion of his recent paper "Grocery Bag Bans and Foodborne Illness." In the paper, Wright and his co-author presented empirical analysis of the health effects of the many, and increasing, jurisdictions that have instituted plastic bag bans or taxes. In short, when such a ban goes into place or when bags are taxed, people shift to substitutes. These substitutes largely take the form of re-usable cloth (or other) bags. The problem, however, is that people by and large do not wash their reusable bags. These bags become contaminated with a wide variety of nasty bacteria and other contaminants, leading to an increase in both illness and death. In fact, the data demonstrates that both emergency room visits and deaths linked to these contaminants increased by 25% after plastic bag bans or taxes were instituted. This translates into, for example, between 5 and 6 extra deaths a year in San Francisco County. The purported benefits that supporters of the plastic bag ban in San Francisco put forth and which Wright's paper cites are the prevention of over 100,000 marine animal deaths per year to plastic entanglement and reducing the consumption of over 12 million barrels of oil required to produce plastic bags used in the United States annually.
The overarching point of Wright's presentation is that often our policy choices are based on conventional wisdom ("convenient wisdom") and the unintended consequences of policies should be brought to light through empirical data and economic analysis. The argument is not unlike the "Freakonomics" argument that parents are more willing to allow their children go play in the homes of children who's parents own a pool than those who own a gun, even though the data demonstrates that their children are far more likely to drown than to die by firearm accident in those homes. I fully agree that more information is always better, and that policy-makers should take into account as many variables as possible when making policy decisions.
There are a few things that bother me about relying too heavily on this approach, however. The concerns I have do not lead me to conclude that the approach should not be taken, but rather that this type of data should be passed on to policy-makers with the appropriate caveats that short term known harms may very well be outweighed by incalculable long term unknowns, and that those "known unknowns" should be detailed to the fullest extent possible.
First, I am concerned with how the benefits and harms may be presented in these types of studies: 100,000 marine animals or some number of barrels of oil versus human lives. This is not necessarily a problem of the researchers in these studies, as they may very well be merely passing along the purported harms avoided posited by the advocates of the policy (here, the supporters of the bag ban). And yet, when the results are pitched as human lives versus a limited number of other harms not directly related to immediate human life and death considerations, it is easy to see how policy-makers might choose the policy that would avoid immediate human deaths (i.e. a bag ban reversal).
Second, and relatedly, the problem is that future yearly human deaths associated with plastic bag production are largely unquantifiable. In other words, short term knowns are more easily quantifiable through this type of empirical analysis, whereas long term, aggregated harms (which may indeed result in far more average deaths per year than 5-6 in a particular jurisdiction) or avoided benefits do not lend themselves to the same type of empirical analysis. At the very least, any empirical data that is produced is based upon projections and models, not upon identifiable deaths that have already occurred. It is incredibly difficult if not impossible to calculate the aggregated contribution of plastic bag production on climate change and sea level rise, that may lead to increased death through heightened storm disaster events, expansion of disease vectors, or its contribution to increased sexual dysfunction or cancer incidence through phthalate or BPA-like chemicals that are taken up through the human body through contact with plastic bags or bioaccumulation from plastic bag pollution in the oceans. So when policy-makers have a choice between two policies, "Policy 1 (with data) vs. Policy 2 (with no or uncertain data)," there is a bias toward the policy based upon short term, quantifiable data.
These biases, of course, appear all the time in the face of uncertainty. It reminds me of the recent Louisiana Coastal Master Plan, which plans to invest $50 billion over the next few decades on projects to mitigate coastal land loss through sediment diversions, levee, dam, and other human-built construction, etc. A recent report notes that Louisiana has the highest rate of relative sea level rise in the world, because as sea levels are increasingly rising, the land in Louisiana is rapidly sinking due to the historical diversion of the Mississippi River. A full 85% of Louisianans support the investment. I am willing to bet that 85% of Louisianans would not identify themselves as ardent environmentalists (though I argue that environmentalism is not a partisan issue and we all should consider ourselves environmentalists) or as citizens who love tax expenditures on government programs. Yet the support is widespread. This is because the short term known (immediate engineering and other feats aimed at making us feel like we can forestall sea level rise) outweighs the long-term unknown (will those projects work and would it have been better to invest those funds into adapting to coastal land loss by moving out of areas likely to be lost). In fact, the science is far from certain that sediment diversions will actually undo much of the damage that has been done along the Mississippi River, meaning that we may not be able to rebuild land in order to forestall sea level rise impacts. There is a strong chance that some of these mitigation efforts will be throwing good money after bad, at least in some areas where political pressure is to "save our coast!" Yet, imagine you have two candidates running for office in Louisiana. Candidate A, knowing that 85% of the Louisiana populous wants to save the coast, says "I will support saving your coast, investing billions in coastal restoration through sediment diversions and human-built construction that will forestall sea level rise impacts." Candidate B says "there is likely not much we can do over the long term to save significant portions of the coast, considering the rate of sea level rise and the models of future climate change impacts and atmospheric carbon concentrations. Let's take that money and invest it in adapting to coastal land loss by moving you out of areas likely to be inundated." Who gets elected? My bet is on Candidate A, even though (and perhaps because of the fact that) Candidate A AND the people who voted for him/her will be dead and gone by the time it becomes clear whether the $50 billion investment failed or not.
Ultimately, I think that empirical data on short term effects is very important and should be taken into account by policy-makers - more information is never a bad thing. But, so much of economics is about using finely tuned formulas to project short-term effects, while the long-term unknowns are largely incapable of being assessed with the accuracy that economic analysis typically prefers (hence the pervasiveness of discount rates, often inappropriately calculated when assessing the availability of resources to future generations). Policy-makers need to take these analyses with a grain of salt, recognizing that much larger policy goals stretching over much longer time periods are at stake. Today's policies are not just aimed at current residents of San Francisco, but have effects on citizens the world over and in the centuries to come.
- Blake Hudson