The Effectiveness of Renewable Portfolio Standards in Reducing Carbon Emissions in the US Electricity Sector

Open Access
Gautam, Suman
Graduate Program:
Energy and Mineral Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
May 31, 2012
Committee Members:
  • R J Briggs, Thesis Advisor
  • Renewable Portfolio Standards
  • Carbon Emissions
  • Selection bias
Do renewable portfolio standards (RPS) – a state level policy that requires utility companies to include a minimum percentage of renewable or “alternative” electricity – lower CO2 emissions? A major goal of RPS is to reduce carbon emissions, but to our knowledge no prior study quantifies this impact. The purpose of this paper is to analyze how RPS policy affects carbon emissions and how this impact varies with RPS characteristics. We develop a panel dataset integrating state-level annual data on RPS levels, CO2 emissions, electricity generation, electricity market restructuring, electricity price, fuel prices, and demographic characteristics. The final dataset consists of annual state level data for 48 states comprising from 1997 to 2010. Among 48 states considered in the study, 22 states have mandatory RPS yearly goals in effect by 2010. We use a reduced form econometric analysis with state fixed variables and time fixed effects to study the impacts of RPS yearly targets on carbon emissions. In addition, we address the possible selection problem of state’s decision to adopt and design its RPS policy with the help of three-part Heckit model. The first stage of the Heckit model is the probit selection equation and the second stage is the linear regression with the RPS yearly targets as the dependent variable. From this second-stage, we calculate the linear prediction of RPS yearly targets by restricting observations to non-zero RPS yearly target variable. We use this linear prediction in the main structural equation instead of the actual RPS yearly targets data to find the effect of RPS on CO2 per MWh. The ordinary least squares results show that RPS yearly targets are statistically significant in reducing CO2 per MWh – a ten percentage point increase in RPS yearly mandates improves carbon efficiency by at least 10 percent. While the OLS results show that RPS yearly targets are statistically significant in reducing CO2 per MWh, the regression results of the base model after considering selection problem fail to find the significance of RPS yearly targets in affecting carbon emission efficiency. The closer analysis suggests that the state’s decision to adopt and design its RPS policy is influenced by factors such as neighboring states’ RPS status, shares of fossil-fired generation, and electricity price. These results suggest that state-level RPS policies do not show any effect in reducing carbon emissions after taking account of the selection problem. This study’s findings do not claim that RPS is not effective in reducing CO2 per MWh, rather we conclude that that the underlying characteristics of states enacting RPS policies have more to do with the apparent success of RPS in reducing CO2 per MWh.