Renewable portfolio standards (RPSs) for electricity generation are politically popular in many U.S. states although economic analysis suggests they are not first-best policies. We present an empirical analysis of the political and economic factors that drive state governments to adopt an RPS, and the factors that lead to the inclusion of in-state requirements given the adoption of an RPS. Although advocates claim an RPS will stimulate job growth, we find that states with high unemployment

more »

... are slower to adopt an RPS. Local environmental conditions and preferences have no significant effect on the timing of adoption. Overall, RPS adoption seems to be driven more by political ideology and private interests than by local environmental and employment benefits, raising questions as to when environmental federalism serves the public interest. Why Do States Adopt Renewable Portfolio Standards? / 133 bottom." Yet RPS adoption appears to be a case of a "race to the top," suggesting that-contrary to received theory-decentralization may be desirable. Rabe (2004) offers fascinating case studies of the growing role of the U.S. states in addressing climate change. He identifies a host of factors that appear to motivate states to adopt RPSs, including policy entrepreneurs supportive of the policy and a desire to wring economic development benefits from an RPS. Our paper is complementary to Rabe's work, offering a quantitative analysis that allows us to test the generality of some of the factors identified by Rabe. Huang, Alavalapati, Carter, and Langholtz (2007) provide a cross-sectional analysis of state adoption of RPSs, and find that Democratic party legislative dominance is the most significant explanatory variable. However, cross-sectional analysis ignores information on adoption dynamics, and cannot employ time-varying explanatory variables. Utilizing a duration model and panel data, this paper avoids these caveats. We also include a wider range of critical explanatory variables, including renewable potential, organized renewable interests, and state electricity market characteristics. In addition to the question of why states adopt RPSs in the first place, there are interesting questions regarding the structure of different states' policies. One important difference across states lies in the tradability of Renewable Energy Credits (RECs). A REC is a certificate of proof that one kWh of electricity has been generated by a renewable-energy source. An RPS typically requires all electricity generators (or electricity retailers, depending on policy design) to demonstrate, by surrendering RECs, that they have provided an amount of renewable energy generation equivalent to the mandated percentage of their total annual kWh sales. Generators are typically allowed to decide for themselves whether to invest in renewable energy projects and generate their own RECs, or simply to purchase RECs from others. The tradable REC system is meant to create incentives for generators to find the cheapest way possible to meet regulatory requirements. However, some states have imposed restrictions, called in-state requirements, on the system. They either disallow REC trading, require that a certain percentage of RECs must be purchased from within the state, or give extra credit to in-state renewable generation. These restrictions attempt to localize the economic and environmental benefits from RPS programs but risk a higher electricity price by foregoing the opportunity to use cheaper renewable resources from outside of the state; they may also risk running afoul of the U.S. Constitution's Commerce Clause, which reserves to Congress the power to regulate commerce between the states. In this research, we study not only what drives states to adopt RPSs, but also what factors lead them to impose in-state requirements. This is particularly relevant considering the growing interest in creating a national market for RECs. The rest of the paper proceeds as follows. In section two, we provide a theoretical discussion of the factors that may have an impact on the adoption and design of RPSs, and discuss the data we use to measure these factors. Section three presents our empirical model, and our findings regarding the relative importance