Heterogeneous Preferences Regarding Global Climate Change

Abstract

We examine the structure of preferences for mitigating impacts of global climate change that will not occur during the lifetimes of most who are alive today. Because no market data exist for such distant markets, a statedpreference approach is used. The analysis is based on the random-parameters logit model, and the results indicate substantial heterogeneity in respondent preferences, that mean willingness to pay is a significant and increasing function of the scope of the impact, and, provocatively, that respondents have the same preferences over the two very different time horizons that we consider. E examine the structure of preferences for mitigating impacts of global climate change that will not occur during the lifetimes of most who are alive today. Studies using current market data have estimated the potential costs of climate change on agriculture (Mendelsohn, Nordhaus, & Shaw, 1994) and on other aspects of measured production (Nordhaus, 1994), but market data can not reveal the willingness of those currently alive to pay for measures to mitigate impacts that may not happen for 100 years or more. There is serious public discussion, involving diplomats at the highest levels, about measures to mitigate emissions that would cost the United States alone 100 to 200 billion dollars annually. Therefore, estimating the willingness of those currently alive to pay for mitigation measures is not a fanciful exercise, but rather a crucial effort to determine how much mitigation should be undertaken. It is the preferences of those alive today regarding acceptable levels of climate impacts—and not the unknowable preferences of those who do not yet exist—that will ultimately determine how much mitigation we pursue in the next quarter-century. As this question cannot be answered with market data, we use the stated preference (SP) approach and elicit preferences for preventing climate change induced impacts. A host of climate change impacts have been enumerated. This study focuses on one type of impact that may be important but has received little attention from economists. Scientists have predicted that widespread ecosystem shifts may accompany changes in climate (VEMAP, 1995). Ecosys- tem shifts are likely to evolve slowly, indicating that lost use values are likely to be small for those alive today, yet the public may still be willing and asked to pay for mitigation measures that are designed to slow the transformation of ecosystems. The current debate over climate change demon- strates two things. The first is that some people do care about long-term changes that will never affect them or perhaps even their children. The second is that preferences regarding the impacts of climate change are very heterogeneous as other members of the public appear much less concerned. Long-term horizons combined with heterogeneous prefer- ences makes ecosystem shifts a useful place to begin analyzing preferences for mitigating climate change im- pacts. A stated preference survey was designed that asked respondents to choose their most preferred program to reduce ecological impacts from a set (menu) of five or six different programs. The programs varied in terms of costs, the degree of ecosystem change, and in the mitigation method. Four different menus provided to each respondent yielded four distinct valuation responses. Section II de- scribes the random-parameters logit model for repeated choices that allows for a flexible correlation structure based on heterogeneous preferences. The ecosystem change and the stated preference survey are discussed in section III. In section IV, we estimate the random-parameters model and estimate the willingness to pay to prevent various levels of ecosystem change. Section V concludes with a discussion of how the scope of the impact, preference heterogeneity, and the time horizon affect willingness to pay.