The economist’s box of tools for dealing with environmental harms is extensive. It begins with assessing why private arrangements fail to account for environmental benefits. If this assessment concludes that markets fail, the next step is benefit-cost analysis, grounded in methods to estimate willingness to pay for environmental improvements. This identifies the difference between where markets land and where we ought to be—where “ought” refers to the point where the cost of additional environmental protection just begins to exceed the value of additional environmental benefits. Once that point is identified, economists can then offer programs such as environmental taxes and marketable permits—and sometimes, with disparagement, technology-specific “command-and-control” requirements, to get us there.
Largely omitted from the toolbox is changing preferences. For economists, this omission is natural, as a fundamental economic principle is to take preferences as the given raw data from which valuations in environmental policy, and virtually every other policy setting, are estimated. My RFF colleagues Jim Boyd and Carolyn Kousky, in a series of blog posts, have discussed potential virtues and implications of preference changes as a policy instrument. Some puzzles deserve emphasis.
One issue is where to stop. Suppose that at current preferences, 100 tons of a pollutant would be emitted. But after verifying a market failure and undertaking a benefit-cost analysis, the optimal quantity of emissions, taking externalities into account, would be 60 tons. Rather than impose a tax, issue permits to emit only 60 tons, or employ some other policy approach, suppose that policymakers were able to get to 60 tons by manipulating preferences—a huge assumption, of course, as Jim and Carolyn have discussed. For example, suppose policymakers could strengthen desires to improve the environment, leading to a greater the willingness to pay for greener options, perhaps electric cars compared to those with gasoline engines.
So far, perhaps so good. But at the new preferences, a benefit-cost analysis using the greater willingness to pay for environmental benefits would lead the optimal level to be below 60, say 40. Should one stay at 60 because that was the original target, even if it is inefficient with the new preferences? Should one change preferences again to get to 40? But what if at the second round of changed preferences the optimum becomes 25—does the cycle continue indefinitely? If this is a foreseen development at the outset, might one decide then that if the net benefits of ending up below 60, at 25, say, are lower than staying above 60, at 100—one should not have bothered changing preferences at all?
One could avoid these puzzles if one could change preferences with (likely unrealistic) precision. If the demand for environmental benefits at the new preferences is perfectly inelastic at some level, leading to no more than 60 tons of emissions, then at those new preferences the optimal amount of environmental benefit would still be 60.
While that would, if possible, avoid unending cycles, it creates a different and more fundamental puzzle. If policymakers had the ability to tune preferences to match any target, so that the new target is consistent with the changed preferences because they are inelastic at that target, then the choice of target becomes completely open. One could “solve” the market failure here by making preferences less green, so that the “optimal” level of emissions becomes 100, not 60. One need do nothing more than change preferences to adapt to the no-policy situation.
If all targets could be optimal with malleable preferences, how do we decide which target answer is the right one? Because the preferences that conventionally define “better off” are being manipulated, that standard solution begs the question. Moreover, some environmental policies may require making some worse off, for example, when mitigating climate change requires those in the present to sacrifice to prevent disaster for future generations.
Anything goes is not a solution, but it’s crucial to reflect on why. It is hard to avoid the conclusion that changing preference to influence policy depends on having access to the “right” answer apart from what people’s preferences happen to be. This may be appropriate, in that the validity of any moral claim regarding rights to a healthy, just, and sustained environment may be no more subject to people’s preferences as whether astrophysics should depend on desires for the earth to be the center of the universe.
This difficulty is shared in other settings, where policymakers believe they know the “right” answer to what people should want and conclude that people’s choices do not reflect their “true” preferences. In either case, benefit-cost analysis, based on observing consumer’s choices, loses its foundation.
The best alternative may be to use a fair and open democratic process to choose those who would change (or ignore) revealed preferences. This sounds more radical than it is; we do it all the time. We elect officials who directly or indirectly make choices according to noneconomic values based on ethical rights or distributive justice. Moreover, we also cannot forget that though economics takes preferences as given, they have to come from somewhere. Manipulating preferences is already part of public policy, most notably using education to inculcate preferences to be good citizens as well as to acquire useful skills. Although the puzzles mentioned here are real, we may not have the choice to ignore them, much as we might prefer to do so.
Read more on these points:
- Brennan, T. 2006. “Green Preferences as Regulatory Policy Instrument.” Ecological Economics 56(1): 144–154.
- Brennan, T. 2014. “Behavioral Economics and Policy Evaluation.” Journal of Benefit-Cost Analysis 5(1): 89–109.
I thank Jim Boyd and Carolyn Kousky for their comments on this post and for all of the blog posts on this topic they have written in their series, Are We Becoming More Environmental? Reflections on Trends in Environmental Desire. All errors here—which they can and should point out—remain my full responsibility.