“Conditionality and the politics of climate change” (with Mark Buntaine and Michaël Aklin)
Conditional commitments are thought to be a stepping stone toward deeper cooperation between states. However, while states frequently make conditional policy pledges during international negotiations on climate change, their empirical effects remain unclear. We conducted three experiments in ten of the largest carbon-emitting countries to test whether conditional pledges made by national governments to mitigate climate change increase public preferences for ambitious climate action in other countries. The results reveal that only unconditional pledges made by foreign countries increase public preferences for policy ambition, and that countries seeking financial and technical transfers only gain support from the public in sending countries when they couple conditional pledges with ambitious unconditional pledges. We also find that the public in most countries only prefers to make part of their country's climate pledge conditional on other countries' action when at high levels of unconditional ambition. Overall, conditional bargaining between countries does not appear to significantly shape public preferences for cooperation on climate change.
“Global geographic variation in climate concern at national and sub-national scales” (with Parrish Bergquist, Peter Howe, Jennifer Marlon, and Clara Vandeweerdt)
In December 2023, the world's leaders will meet in Dubai to assess the world's progress towards the goal of limiting global warming to 2 degrees Celsius. Scholars have extensively discussed the complex and thorny geopolitics of this convening, but the micro-foundations of those geopolitics are poorly understood. This is due to the patchy availability of survey data concerning climate change around the world. In this paper, we present a comprehensive dataset estimating concern about climate change in 164 countries around the world, on a common scale. We describe our dataset of responses to 134 questions from 101 surveys measuring concern about climate change between 1998 and 2020. We then develop a group-level item-response-theory model to aggregate these data into a single latent-variable estimate of concern about climate change and support for policy to address it in each country represented in our dataset. We present preliminary results from this model, which we validate by comparing the model outputs to cross-sections of our underlying data. The estimates we present have the potential to open exciting new avenues for deepening scholarly understanding of the drivers of climate concern around the world.
“Process-tracing, counterfactual comparisons and causal inference”
Process tracing is now the dominant method used by qualitative political scientists. However, the nature of process-tracing remains the subject of substantial debate, including the mechanism through which the method supports causal inference-making. Here, I offer a methodological elaboration of process-tracing within the potential outcomes framework. Emphasizing the shared importance of counterfactual comparisons for both qualitative and quantitative inference, I argue qualitative scholars can also manage the Fundamental Problem of Causal Inference. While quantitative approaches to causal inference typically use statistical techniques to estimate group-level causal estimands, particularly causal effects, qualitative approaches can use process-tracing to make informed judgements about unit- level causal estimands, particularly causal mechanisms. Both approaches involve structured frameworks to estimate the value of unobserved counterfactual states of the world, and can be understood as achieving their causal inferential leverage by comparing realized outcomes with assumption-dependent representations of unrealized counterfactual outcomes. Accordingly, I conceptualize process-tracing as a tool to estimate bounds on unit-level counterfactual outcomes and describe how the method can mitigate common biases associated with counterfactual reasoning. In this account, qualitative causal inferences are generated by within-case analysis; correspondingly, qualitative research that draws from multiple cases should be understood as a form of empirical replication to generate conjectures about the scope conditions under which particular causal mechanisms realize. Acknowledging the implicit role of counterfactual reasoning in process-tracing can also increase the transparency and falsifiability of qualitative political science analysis.
“Policy bundling and the political economy of climate policymaking” (with Parrish Bergquist)
Climate action has been stymied by a challenging political economy: climate policies can impose salient, short-term costs in exchange for uncertain, long-term benefits. Public support for many climate policies has, as a result, remained uneven. We investigate the effects of policy bundling - linking one contentious policy to other programs - on public support for climate action. Using conjoint experiments in 10 of the world’s largest carbon-emitting countries, we show that policy bundling increases support for climate policy across diverse political and economic contexts. Linking climate policy to other economic and social programs expands coalitions of support, particularly on the left, without reducing support on the right. Achieving the Paris Climate Agreement's goals will require every country navigate contentious domestic climate policy conflicts; we find that policy bundling offers a broadly applicable strategy for strengthening public coalitions in favor of a contentious policy.
“Perceived costs dominate objective costs in predicting carbon pricing opposition” (with Alice Lépissier, Chloe Boutron, Erick Lachapelle and Kathryn Harrison)
Economists view carbon pricing as a cost-effective way to reduce carbon pollution, but extant carbon pricing policies have often generated significant political opposition from both interest groups and national publics. Often, these political debates revolve around the potential policy costs and benefits - both objective and subjective. Here, we investigate whether Canadians accurately perceive the costs and benefits of the country's existing carbon tax and dividend policy. Using surveys fielded before, during and after policy implementation, we find stronger opposition to carbon pricing among those who drive to work and spend more on transportation fuels. However, partisanship and ideology dominate assessments of cost. Conservative party supporters overestimate the impact of carbon pricing on gas prices by more than other parties' voters. Moreover, perceived price changes better predict carbon tax opposition than actual changes in the price of gasoline and other fuels. Our results suggest that, partisan contestation can undermine public support for carbon pricing, even with a progressive policy design.
“How constituent contact reinforces the status quo: Evidence from administrative records on contact to Congress” (with Olivia Quinn, David Broockman, Alexander Hertel-Fernandez, and Leah Stokes)
Citizens communicating to their representatives is a key mechanism for representation. However, who contacts politicians, what they express, and how this may affect representation remains unclear. We offer an unprecedented window into these questions based on a collaboration with a constituent communications software vendor to the United States Congress. Analyzing data from over 3.4 million constituent contacts from 14 Congressional offices over four years, matched with an original survey of voters, we document three biases that lead contact with Congress to reinforce the status quo. First, we find \textit{demographic bias}, with socioeconomically advantaged voters more likely to contact Congress. Second, we uncover \textit{organizational bias}, with organizations mobilizing the vast majority of constituent contact. Third and most originally, we document \textit{reactive bias}, wherein opponents of changes to the status quo are especially likely to contact Congress. We show how these biases distort which issues and viewpoints Congress hears about.