Governments can choose from a wide variety of policy instruments to control pollution, including tax and subsidy schemes, marketable permits, information provision programs, liability systems with well-defined property rights, and command and control selection of emission standards or technologies. Economists throughout the 1980s devoted significant effort to modeling the relative merits of these instruments. The question of why politicians would prefer one manner of controlling pollution over another also attracted attention, starting with Buchanan and Tullock's  positive theory of externality control in which polluters find direct regulatory controls on emissions more profitable than pollution taxes. Theories of instrument selection have been developed in which politicians choose policies based on the relative costs and benefits to shareholders, workers, and environmentalists, are affected by the constraints imposed by political institutions, and are driven to select the most efficient policy instrument to achieve a given level of redistribution.(1) To date, however, there have been few empirical tests of what drives instrument selection in environmental policy.
There is a large empirical literature devoted to explaining the level of support for environmental programs in Congress. Most articles focus on the degree to which a legislator's stand on a given environmental bill can be explained by narrow district interests (generally the material interests of constituents) or by ideology (either constituents' or the member's own policy preferences). Studies of voting on environmental legislation by Ackerman and Hassler , Crandall , Pashigian , and Yandle  offer evidence that legislators' actions are in part explained by the district- or state-level impacts of proposed legislation on their constituents. Kalt and Zupan  find that after controlling for constituents' material self-interest in proposed strip mining legislation, the ideology of constituents and the personal ideologies of their representatives influence voting on this aspect of environmental preservation. As Grier  demonstrates, debate over the general role of the legislator's specific ideology in voting continues. In studying votes in the House and Senate on the proposed size of the Superfund during debate over the program's renewal, Hird  concludes that "congressional voting on Superfund is found to represent legislator's environmental and liberal ideologies as much as (if not more than) narrowly-defined self-interest." He links the importance of ideology to voting on Superfund in part to the bill's symbolic environmental appeal.
In environmental legislation, the final impact of a bill on industry and the environment often depends on issues resolved through technical amendments. Downs's theory of rational ignorance and Arnold's theory of the degree to which constituents will link Congressional actions with real world outcomes predict that members of Congress may be less constrained by general constituent interests on amendment votes since these votes are less likely to be covered in the media than are final legislative votes. As Denzau and Munger  and Grier and Munger [ 1991] note, groups with specific interests such as affected industries may find it less costly to persuade a member of Congress to vote their way on a measure if the representative's constituents are supportive of or indifferent to the measure. In the case of technical amendments, affected parties may be more influential than on highly visible votes because the lack of media coverage translates into a lower probability that the vote will become a major campaign issue in future elections. This implies that voting on the instruments to control pollution will be driven in part by the specific district-level costs and benefits of these measures, while voting on the final passage of a bill will be more driven by …