Volume 25, No.1, Spring 2007

Politics and Environment

ECONOMIST, n. a scoundrel whose faulty vision sees things as they reall are, not as they ought to be. – after Ambrose Bierce

Do single-issue voters matter? Recent evidence suggests that, when the issue is the environment, the answer is “yes.” Examining gubernatorial election returns for the last 30 years of the twentieth century, John List and Daniel Sturm (2006) have found that electoral incentives are an important determinant of environmental policy choices. When governors are subject to electoral discipline, they enact environmental policies more closely matched to the wishes of their constituents. But when term limits make them lame ducks, those same governors are apt to ignore the wishes of the electorate and instead act according to their own preferences. Simply put, when it comes to the environment, politicians pander.

For purposes of their analysis, List and Sturm begin by characterizing states as either “green” or “brown,” depending on whether a relatively large or small proportion of residents belong to three leading environmental groups: Greenpeace, Sierra Club, and the National Wildlife Federation. The greenest states include most of those in the northeast (with Vermont the greenest of all), plus California, Oregon, and Colorado. The brownest states are clustered in the southeast, ranging from South Carolina over to Texas. Montana happens to fall on the brown side of the environmental dividing line, which is probably not unreasonable for the period under study (1970–2000).

To measure the environmental stance of a governor, the authors use state expenditure data on fish and game, forests and parks, and on other natural resources. Higher spending levels on these items are taken as evidence of a more pro-environmental policy stance. The authors also examine other potential measures of gubernatorial environmentalism (such as the costs of complying with state environmental regulations) but the empirical results do not differ appreciably, regardless of the measure used.

There are three key findings. First, environmental policy differs substantially between years when the governor is eligible to run for reelection, compared to years when term limits prevent that. Second, this policy-switching takes on a well-defined pattern. In green states, governors sharply curtail environmental spending once they are term-limited, while in brown states they substantially increase such spending when they are no longer subject to the discipline of the ballot box. In effect, regardless of the political party of the governor or the underlying preferences of the electorate, the pattern is the same: Governors shape their environmental policies to fit the preferences of voters when they are subject to voter discipline, and shape those policies to suit their own preferences when such discipline is removed.

The final key result is that the degree of policy-switching (or environmental pandering, if you prefer) is directly related to the degree of competitiveness in the electoral process. Governors who won handily in the last election are less likely to shape their policies to fit the preferences of single-issue environmental voters. In contrast, when elections are likely to be close, governors are willing to deviate the farthest from their principles to get the votes of those who feel the strongest, one way or the other, about the environment.

The spending swings that occur as governors shape their policies to fit their electoral circumstances are substantial. And not surprisingly, the biggest movements are found where elections are highly competitive, because that is where governors do the most pandering. For example, in brown states where gubernatorial elections have been close, term-limited governors increase environmental spending by almost 10 percent in their final term. In highly competitive green states, term-limited governors slash environmental spending about 20 percent during their final term. In both cases, the spending changes are almost exclusively reallocation of spending priorities: the environmental budget cuts get spent on other programs, while the environmental spending increases come out of other existing program budgets.

One striking implication of these findings has to do with the channels of political influence. It is commonplace to bemoan the excessive influence of campaign contributions on electoral outcomes, and to use this supposed influence as justification for restricting competition along such lines. But in the case of environmental electioneering, according to List and Sturm it does not seem that campaign contributions are the central issue. Gubernatorial campaign contributions from the top environmental groups amount to but 2–3 percent of what electric utilities or oil and gas companies give in such elections. Despite this, the evidence in this paper shows that environmental issues are clearly responsible for substantial policy swings by governors, presumably because these issues resonate strongly with single-issue environmental voters.

In addition to illustrating the importance of environmental issues in elections, this study sheds light on the impact of term limits. There are many reasons offered for these caps on service in elected office, one of them being that by bringing fresh blood to office, they make politicians more responsive to the wishes of the people. Based on the evidence in this study, term limits in fact insulate politicians in their final terms from their responsibilities to the electorate. All the more reason, I suppose, to be careful about for whom you vote.


List, John A., and Daniel Sturm. 2006. How Elections Matter: Theory and Evidence from Environmental Policy. Quarterly Journal of Economics 121(4): 1249–81.

Related Content