The Emergence of Renewable Portfolio Standards: An Empirical Investigation Tom Lyon (Joint with Haitao Yin)
The Adoption of RPS As of 2005, 22 states and DC have adopted RPS. State Goal ☼ PA: 18%¹ by 2020 ☼ NJ: 22.5% by 2021 CT: 10% by 2010 MA: 4% by % annual increase WI: requirement varies by utility; 10% by 2015 Goal IA: 105 MW MN: 10% by 2015 Goal + Xcel mandate of 1,125 MW wind by 2010 TX: 5,880 MW by 2015 *NM: 10% by 2011☼ AZ: 15% by 2025 ☼ NV: 20% by 2015 ME: 30% by 2000; 10% by 2017 goal - new RE State RPS *MD: 7.5% by 2019 ☼ Minimum solar or customer-sited requirement * Increased credit for solar or customer-sited ¹PA: 8% Tier I, 10% Tier II (includes non-renewable sources) HI: 20% by 2020 RI: 15% by 2020 ☼ CO: 10% by 2015 ☼ DC: 11% by 2022 ☼ NY: 24% by 2013 MT: 15% by 2015 *DE: 10% by 2019 IL: 8% by 2013 VT: RE meets load growth by 2012 *WA: 15% by 2020 Note: Renewable portfolio goal is voluntary, as opposed to a renewable portfolio standard, which is generally enforced by an appropriate state regulatory agency.
The Adoption of RPS: RI PAVT WINYMT NVTXMDIL MNMENJNMHIDC IAMACTAZCACODE
RPS: The Right Path to Sustainable Electricity? Dallas Burtraw and Karen Palmer, “Cost- Effectiveness of Renewable Electricity Policies,” RFF Working Paper, Main findings –RPS raises electricity prices and primarily displaces natural gas –RPS is not as cost-effective as cap-and-trade for reducing carbon emissions.
Research Questions Why have state governments moved to RPS’s without federal mandates, especially since they appear to be solving a global externality problem? Why do states choose RPS’s instead of carbon caps? Goals –Improve our understanding of RPS as an environmentally friendly energy policy tool –Shed light on the role of state governments in American federalism Assuming leadership role and launching bold initiatives (race to the top)? or Setting lower regulatory standards in hope of inducing business to move from other states (race to the bottom) ? –Extend work of Rabe (2006) using a more systematic econometric analysis
Statehouse and Greenhouse Rabe (2006) Case studies on five states: Texas, Massachusetts, Nevada, Pennsylvania, and Colorado Driving Forces: –Transcending Partisan Boundary: “Regardless of partisan composition of state government, these policies (RPSs) have consistently drawn a rather broad coalition of support.” –Renewable Energy Developers: “Renewable energy developers are far more visible and influential in RPS deliberations than conventional environmental advocacy groups” –Economic Benefits: “perhaps one of the biggest factors…has been a commonly held perception that promotion of renewable energy through an RPS is in the economic interest of an individual state.” Especially, “…this labor benefit has fostered discussion in many state capitals about an anticipated ‘job multiplier’ effect of renewable as opposed to conventional sources.”
Factors Driving RPS Adoption Public Interest Private (Special) Interests Political Ideology
Factors Driving RPS Adoption: The Public Interest Environmental Benefits: –Research hypothesis 1: States with poor air quality conditions are more likely to adopt RPS. Economic Benefits: –Research hypothesis 2: States with higher unemployment rates are more likely to adopt RPS.
Factors Driving RPS Adoption: Special Interests Conventional Environmental Groups: –Research hypothesis 5: States with more environmental group members are more likely to adopt RPS. Renewable Energy Developers and Advocates: –Research hypothesis 6: States with more renewable energy developers and advocates are more likely to adopt RPS.
Factors Driving RPS Adoption: Political Ideology Ideological Preference of Constituents: –Research hypothesis 7: States with higher LCV scores are more likely to adopt RPS. Ideological Preference of State Legislatures: –Research hypothesis 8: States with Republican-controlled legislatures are less likely to adopt RPS.
Model Specification We model the state decision of adopting RPS from 1991 to 2005 as a hazard model that defines the probability that state i adopt RPS in year t, given it did not adopt RPS in year t-1, as a function of relevant variables. Proportional odds model – the discrete time hazard rate for year t – is a vector of explanatory variables – the baseline hazard arising when = 0 This generates a logistic hazard model:
Explanatory Variables Air Conditions: –Nonattainment Index –Greenhouse Gas Emissions Unemployment Rate Renewable Energy Developers –Presence of an ASES Chapter in the state –We are collecting data for AWEA membership Environmental Groups –the number of Sierra subscriptions per 1,000 state residents in 2000 Ideological Preference of the State’s Constituents –Average LCV scores of U.S Senators and U.S. House Representatives in each state Ideological Preference of the State Legislature –Republican-controlled or not the population (in 2000) in state i living in areas where air pollution exceeds national ambient air quality standard for pollutant j; is the population in state i in 2000.
Results (1)(2)(3)(4)(5) RPS Average Electricity Price in Previous Year (2.53)*(2.41)*(2.21)*(2.59)**1.88 Nonattainment Index (2.72)**(2.39)*(2.35)*(2.22)* Unemployment Rate (2.62)**(3.01)**(2.93)**(3.08)** Presence of ASES Chapter (2.29)*(2.00)*(2.09)* Republican-controlled (3.02)** LCV Score0.034 (2.70)** Other Variables Included Wind Potential Solar Potential Median Income Regulatory Restructure Wind Potential Solar Potential Median Income Greenhouse Gas Emission Regulatory Restructure Electricity from Coal and Nuclear Coal Reserves Wind Potential Solar Potential Median Income Regulatory Restructure Electricity from Coal and Nuclear Coal Reserves Sierra Subscription Wind Potential Solar Potential Median Income Regulatory Restructure Electricity from Coal and Nuclear Coal Reserves Wind Potential Solar Potential Median Income Regulatory Restructure Electricity from Coal and Nuclear Coal Reserves Pseudo R220.57%29.10%31.49%38.24%35.50% Observations668 Absolute value of z statistics in parentheses * significant at 5%; ** significant at 1%
Key Findings (1) States with higher average electricity price are more likely to adopt an RPS. –This is consistent with the argument that the states with higher electricity price take renewable energy development as a way of complementing conventional energy sources and ensuring long-term energy supply. RPS is in part a responses to poor air conditions within the state; States with higher unemployment rates are less likely to adopt an RPS, contrary to Rabe’s (2006) notion of RPS as an economic development tool Renewable energy developers are among the driving forces of RPS;
Key Findings (2) Ideological preference matters a lot; –In fact, a closer look at the data finds that only two of the 23 RPSs were enacted by a Republican-controlled legislature – New Jersey and Pennsylvania. Many things do NOT seem to matter: –Solar and wind potential –State income –GHG emissions –Presence of coal or nuke supply in state –Sierra Club membership Public interest theory, private interest theory and theory of ideology all have some valuable insights on the adoption of RPS. But none of them can explain the whole story of the emergence of RPS.
Next Steps Add AWEA Members; explore possible endogeneity between ASES/AWEA membership and RPS Explore state policy regarding Renewable Energy Credits (RECs) and their fungibility Interplay of Sierra Club membership, LCV ratings, and state party control