Are We Becoming More Environmental? Reflections on Trends in Environmental Desire

How to Evaluate Policies that Change our Desires

Nov 13, 2015 | James W. Boyd, Carolyn Kousky

The strength of our environmental desires is of central importance to developing efficient and effective environmental policies. Over the next several weeks, we’ll be hosting a conversation that explores whether environmental desires are changing and what that means for environmental economics and policy. This is the eighth post of eleven in the series: Are We Becoming More Environmental? Reflections on Trends in Environmental Desire. This series is a result of an RFF New Frontiers Fund award for thinking creatively about pathbreaking research to advance RFF’s mission.

Public policies are often adopted in an attempt to change behavior. This can be done by prohibiting certain activities or trying to make certain behaviors more attractive, such as by changing relative prices. If we tax carbon, for example, it would incentivize individuals and firms to use less carbon-intensive energy sources to reduce costs. Economists are quite comfortable evaluating the impact of policies that alter prices.

But can policy also change our underlying environmental desires? If so, what are the implications for policy evaluation?

All the mechanisms discussed in earlier blog posts that could drive a change in environmental desires could operate through public policy as well. Policies could provide new information, expose us to “experience goods,” alter our habits, and change social norms. As we have already discussed in this series, all these things could end up altering our environmental desires. The challenge is that evaluating the costs and benefits of a policy that has changed desires raises thorny problems for the way economists traditionally undertake policy evaluation.

Economists typically evaluate a policy by referencing people’s choices, which are assumed to reflect what makes them happy, or their underlying desires. Consider a hypothetical policy to improve the water quality of a river, in which the clean-up costs are known, but policymakers want some estimate of the benefits of this policy to know whether it should be adopted. An environmental economist would set about estimating the “willingness-to-pay” (WTP) for improving the water quality of the impacted population. There are many techniques used to do this that have been developed over the decades. All of them, however, presume that people have a certain level of desire for water quality and a WTP that reflects this. The challenge is that this analysis depends on desires that depend on the very policy we are attempting to evaluate.

To see this, consider what happens if a clean river is an experience good. After the water quality is improved, people boat on the river, picnic on the banks, and fish in it. All of these activities didn’t occur when the water quality was poor. Once individuals experience them, however, they come to desire them more than they did before. This would imply that the WTP for water quality is different after the policy is enacted than it was before, since the policy changed people’s desires.

How should the policy be evaluated in this case? Should we use the WTP from before the stream was clean or the WTP from after? In thinking about experience goods, it seems natural to us to argue that the correct measure to use is the WTP after the water quality improves. These are, in some sense, the “final” desires of the individual; the desires before the water quality improves seem to be in some sense uninformed and thus not “true.” Or it could be argued that since the policy changes desires, it should be evaluated from the position of these new, altered preferences as that is the new status quo.

Using post-policy desires to estimate WTP, however, poses numerous challenges. The first is that analysis cannot be undertaken before the policy is adopted. So if we want to evaluate a policy to determine if it indeed should be adopted, we are in the position of having to predict how desires will change in response to a policy in order to predict how WTP will change. This can be quite challenging, as we have discussed elsewhere. Problematically, it detaches evaluation from actual observation of behavior, which makes some economists uncomfortable. How can we be confident our predictions are correct? And could projections be more easily manipulated for political ends than observations of behavior? It is possible in some cases an analogous situation may be found to evaluate and used to inform projections, but not always.

Policy could also change desires through channels other than learning and experience, though, and in some of these situations we may not be as comfortable using post-policy desires to evaluate public policy. Consider the case of habituation. In some instances desires are a function of habits and current conditions. We become used to the status quo and generally prefer that it continue. Behavioral economists refer to this as status quo bias. As an example, say a community adopts a policy in response to drought that limits lawn watering. This could cause a shift to more desert landscaping in front yards. Growing accustomed to seeing such front yards may shift environmental desires toward this aesthetic, even when the water restriction is lifted. In this case, should we evaluate water restriction policies using the preferences of people before the regulation, which favored lawns, or those after the restriction, which favor desert landscaping? The former suggests much larger costs to the policy than the latter.

If it appears to the reader that in this case we should also use post-policy preferences, it may be simply due to a belief—quite apart from people’s preferences—that reducing water use is a good thing. Using post-policy preferences makes the water restriction policy seem less costly. To see this possible bias in believing post-policy preferences should always be used, consider habituation to grossly unjust circumstances, such as segregation. What if people adapt to having segregated public spaces such that they desire integration less? Here, it seems obvious that post-policy preferences should not be used at all.

This may be even more clear when considering desires that are changed maliciously, such as through brainwashing or deliberate provision of misinformation by a tyrant. In this case, we would want to argue that the post-policy desires are not the “true” ones that should be used for policy evaluation—instead we should use the pre-policy desires.

It seems, then, that we are choosing whether pre- or post- policy desires be used for evaluation based on our feeling about the policy from a position independent of people’s desires. We believe in environmental protection and do not believe in segregation; these beliefs drove our decision of which attitudes we treated as “true.” Simply seeking to influence desires, such as through advertising, does not usually strike one as problematic, but altering desires by lying and oppression does, since we deem these activities as immoral from a stance quite apart from people’s desires.

This suggests the third approach (beyond using either pre- or post-policy desires) to policy evaluation when desires are endogenous to our policy choices: adopt a non-utilitarian stance. Economists tend to believe that a policy is best if it maximizes utility, or happiness, and that our choices reflect our utility. But not all scholars take this point of view or believe it is a realistic approach for policy evaluation.

There are political philosophers who argue policies should be evaluated by reference to conceptions of the greater good, public values, or common goals. One argument is that some people clearly exhibit behavior and choices, and perhaps underlying desires, that are harmful to themselves or others (such as addiction). This leads to the conclusion that revealed preferences—that is, what we observe about people’s choices—is a poor measure of how well off they are and should not be used to evaluate policy. In the same way, we could argue that behavior and desires that are informed by misinformation should not be used for evaluation by arguing that such manipulation violates higher order values we have as a society.

Scholars have also noted that citizens may have desires over policy choices pertaining to public matters that are not the same as those that govern their choices as consumers in their private life. As a simple example, individuals often may behave altruistically, preferring social allocations that do not necessarily maximize their personal allocation, such as dividing things evenly or giving more to those with less. In the case of our water quality policy above, it may be the case that an individual does not think that whether the water quality is improved should be decided based on their private desires about recreating on the river, but instead in relation to arguments about the morality of providing healthy habitat for other species. That is, whether or not my desires for enjoying a clean stream change or not, it should be cleaned for reasons outside my personal preferences.

This, however, raises its own set of difficult issues for economists. In the language of political scientists, one person’s conception of the good could vary quite a bit from another’s conception. For example, some may argue we evaluate policy against the metric of long-term sustainability, while others may argue we evaluate it against contributions to long-run economic growth. One way around this that has been explored in detail by political scientists and some economists is to instead focus on institutional design. A famous example is Rawls’ notion of everyone agreeing on a government structure before knowing their position in society. As Mark Sagoff notes: “The problem is not necessarily to determine which of these conceptions is correct, but to understand how to structure social institutions so that people with contradictory views can go about the business of trying to convert one another under terms that are congenial to all.” If we all agree on our governing institutions, then policies that arise from them are just. But, this does leave economists without nice WTP numbers to conduct efficiency and cost-effectiveness analyses.

Up next in the series—Evaluating Long-Lived Policies When Our Desires Change.

Read previous posts in this RFF blog series: Are We Becoming More Environmental? Reflections on Trends in Environmental Desire.

    1. Are We Becoming More Environmental? An Introduction

    2. Growing Environmentalism: The Difference between Desires, Behavior, and Preferences

    3. What Changes Our Environmental Desires? A Look at Taste Formation

    4. What Changes Our Environmental Desires? Experience and Learning Matter

    5. What Changes our Environmental Desires? The Impacts of Social Norms.

    6. Do Our Social Institutions Affect Our Personalities?

    7. Can We Measure—and Predict—Changing Environmental Desires?