Climate policy has gained significant momentum in many countries in recent years. In theory, putting a globally aligned price on carbon emissions, equal to the social cost of carbon, is the optimal approach to internalize climate externalities and steer the green transition. Following this theoretical idea, a list of countries have implemented carbon pricing schemes in practice in recent years, either through cap-and-trade systems or through carbon taxation. As a result, we are now entering a phase where we can evaluate the effectiveness and the efficiency of carbon pricing policies empirically ex post. As time passes, we are gathering more and more data at reasonable granularity. The rich cross-sectional variation, the large number of policy changes, and the ever-expanding time series open up lots of opportunities for causal inference. Recently, a few papers have made significant progress on two important research questions: 1) Is carbon pricing effective in reducing emissions? 2) What are potential channels and mechanisms, and are there any negative side effects for regulated firms?
Carbon pricing and investments in abatement technologies
Colmer et al. (2024) evaluate the effectiveness of carbon pricing using very rich and granular French administrative data. They study Phases I and II of the EU ETS (2005-2012). Their granular data allows them to obtain causal evidence through a quasi-experimental, matched difference-in-differences research design at the firm level. First and foremost, the authors find that market-based regulations can be an effective and economically reasonable tool to mitigate climate change. More precisely, the EU ETS has reduced both firms’ emissions and emission intensities. Regulated firms reduced their emissions by 14\% during Phase I and by 16\% during Phase II, as compared to their unregulated peers. In total, carbon pricing has caused (at a minimum) one-third of the entire emission reduction that we have observed in the EU since 2005. The EU ETS also did not induce a negative effect on regulated firms’ output or employment during Phase I and II. Instead, the authors find a positive effect on firms’ installed capital. This evidence supports the narrative of fixed up-front costs of investing in abatement technologies. Carbon pricing significantly increases the incentives of firms to undertake such investments and thus leads to more efficiency after all.
Incentives and firm heterogeneity
Martinsson et al. (2024) studied the case of Sweden, which introduced a carbon tax (on heating) in 1991. The setting allows them to work with a very long time series of 26 years and with a rich cross-section of manufacturing firms, giving them lots of statistical power. Moreover, the large number of policy changes and firm-level exemptions as well as the introduction of the EU ETS towards the end of their sample open up many opportunities for causal inference. Equipped with this rich data, the authors compute the elasticity of carbon emissions with respect to the carbon price at the firm level. On average, a 1% increase in marginal emission cost share reduces carbon emissions per unit of (real) sales by roughly 2% over 3 years. The authors also document pronounced heterogeneity in carbon price elasticities. Financially constrained firms are more inelastic. Carbon pricing thus particularly hurts those financially constrained firms for which abatement requires significant investment. Turning towards aggregate effects, the authors estimate that, at a minimum, carbon pricing is responsible for one-third of the entire emission reduction in Sweden over the sample period from 1991 until 2015. Relying on counterfactual scenarios, they report that carbon emissions would have been 30% higher in Sweden in 2015 without carbon pricing.
Time-varying carbon pricing schemes – EU ETS
A big caveat of these studies is the fact that they analyze rather early carbon pricing schemes which may not have been stringent enough to really “bite” the regulated firms. For instance, the EU ETS has become much stricter in Phases III and IV, with more sectors being covered and fewer free allowances being issued. Against this background, Bolton et al. (2023) provide a first analysis of Phase III of the EU ETS. They study the response of firms’ stock prices to changes in carbon prices on a daily basis. The use of daily data here is key to allow for causal interpretations. The authors compute a firm-level measure of exposure to the EU ETS as the difference between free allowances and regulated emissions of firms. This emission-allowances-shortfall turns out to be a significant driver of the stock price response to carbon price changes. The authors document a negative effect of the EU ETS on stock prices of firms with a pronounced emission allowances shortfall, but a positive effect for firms that have a surplus in emission permits. The key channel determining the stock price responses seems to be firms’ profitability. The results also indicate that carbon pricing induces firms to reduce their carbon emissions covered by the EU ETS, but the authors do not find evidence that the firms also reduce their emissions globally, hinting at possible carbon leakage. Overall, the findings suggest that the EU ETS was stringent enough during Phase III.
Expansion to Carbon Border Adjustment Mechanisms
Finally, Shi et al. (2024) study stock price responses to a new and very recent EU carbon pricing tool, the EU Carbon Border Adjustment Mechanism. The CBAM is the world’s first carbon border tax. In a nutshell, it requires the importers of certain goods to pay an import tax proportional to the carbon footprint of these goods. The CBAM is supposed to disincentivize carbon leakage and to level the playing field for production inside and outside the EU with respect to carbon pricing. It is part of the EU Fit-for-55 package, which comes with a bunch of other reforms associated with Phase IV of the EU ETS. The official legislative procedure was completed in 2023, and the first carbon border tax payments will be due in 2026. Because the policy is so recent, there is currently no way to evaluate the effectiveness of the policy in terms of emission reductions. But Shi et al. (2024) studied the responses of stock prices of regulated firms to the key legislative decisions. To do so, they combine financial market data with granular international data on firms’ supply chains. Most importantly, the authors document a negative abnormal stock return of treated firms of about 1.3 percentage points over a short time window around the key event date. Although small at first glance, the finding is consistent with the aforementioned studies, in the sense that the EU carbon pricing policy seems to “bite” more and more during Phases III and IV of the EU ETS. The paper also suggests that the CBAM – although targeting producers outside the EU – has negative spillover effects on firms within the EU.
Conclusion
To summarize, there is rich causal evidence in the literature that carbon pricing is an effective tool to reduce carbon emissions and thus mitigate climate change. At the same time, carbon pricing schemes that have been implemented vary greatly in their structure, coverage, and magnitude. This heterogeneity is key to understanding the channels, mechanisms, and potential negative side effects of carbon pricing. In particular and unsurprisingly, the EU ETS seems to have a much larger effect on regulated firms’ financial performance during the most recent Phases III and IV.