How effective is carbon pricing?—A machine learning approach to policy evaluation

Abstract

While carbon taxes are generally seen as a rational policy response to climate change, knowledge about their performance from an ex-post perspective is still limited. This paper analyzes the emissions and cost impacts of the UK CPS, a carbon tax levied on all fossil-fired power plants. To overcome the problem of a missing control group, we propose a policy evaluation approach which leverages economic theory and machine learning for counterfactual prediction. Our results indicate that in the period 2013–2016 the CPS lowered emissions by 6.2 percent at an average cost of €18 per ton. We find substantial temporal heterogeneity in tax-induced impacts which stems from variation in relative fuel prices. An important implication for climate policy is that whether a higher carbon tax leads to higher emissions reductions and higher costs depends on relative fuel prices.