Research Statement

My research broadly addresses democratic accountability. Much of this research focuses on elections as instruments of accountability by examining one measure of accountability—the economic vote—through an alternative lens focused on strategic politicians.

Below, I describe three strands of this research. The first, a book project based on my dissertation, focuses on means by which politicians can skirt accountability by manipulating voters’ tendencies to engage in economic voting. The second is a series of papers exploring additional implications of the democratic accountability literature. The final strand explores how monetary policy makers and ordinary citizens form inflation expectations.

Dissertation Book Project
My dissertation, entitled “It’s About Time: Election Timing’s Effects on the Economic Vote,” connects the literatures on economic voting and election timing. Early work in the economic voting literature found the real economy’s effect on election outcomes to be highly variable. Scholars have explained this fragility by considering the effects of political institutions on how much responsibility for the economy voters attributed to incumbent politicians. While measures of institutional constraints (e.g., clarity of responsibility (c.f., Powell and Whitten, 1993; Whitten and Palmer, 1999)) are still commonly used in studies of economic voting, the first chapter of my dissertation shows that these fundamental findings in the literature suffer from model misspecifications and misinterpretations that, when corrected, produce shockingly little empirical support for these theories under numerous alternative specifications.

To revive the literature on the economic vote at the level of election outcomes, the dissertation explores an alternative causal mechanism through which institutions may affect economic voting: the strategic timing of elections. The endogenous election timing literature assumes economic voting—politicians surf the economy when calling elections because they expect voters to reward them during upturns—but largely fails to test this assumption.

I begin by assuming that voters hold incumbents more responsible for the economy when it is performing poorly than when it is performing well (e.g., Singer 2011, Weyland 2003), and that they neither know nor care how constrained politicians are by political institutions (e.g., Bartels, 1996). Politicians, however, are highly aware of these constraints and of voter responses to the economy. Politicians—both in and out of government—have an interest in the timing of elections during a business cycle (e.g., Ito, 1991; Kayser, 2006; Smith 2004). If variations in the magnitude of the economic vote vary based on the type of election, then these variations ought to correlate with the incentives and capacities of different political actors to call early elections.

I go on to categorize 255 elections in 22 OECD countries between 1967 and 2010 into one of three types: end of term elections (held in the final 3 months of the maximal parliamentary term (Samuels & Hellwig, 2008)); early elections called by the prime minister (premier timed elections); and early elections called by the opposition (opposition timed elections). I find that the underlying assumption of the opportunistic election timing literature that early elections are always the consequences of choices made by the prime minister is not valid: nearly 20% of all early elections are brought about by opposition parties.

With this new dataset in hand, I investigate what institutional and economic factors produce different types of elections. Using multi-state hazard models, I find evidence that the constitutional right of prime ministers to unilaterally dissolve parliament and its combination with low unemployment strongly predict premier timed elections. Semi-presidentialism and high unemployment both increase the likelihood of an opposition timed election.

The final substantive chapter of the dissertation then returns to the initial question of whether politicians, through the timing of elections, affect the magnitude of the economic vote and, ultimately, improve their grasp on power. I argue that the effects of the economy on election outcomes are mediated by the type of election. I find evidence that the economy has a statistically significant effect on the retention of the prime minister following elections only through its effect on the type of election held. This evidence provides additional evidence that focusing on the actions of strategic politicians under constraints can help scholars better understand the puzzling variation in the importance of the economy in election outcomes. I conclude that, at least where election timing is fungible, economic voting may undermine mechanisms of accountability.

The dissertation posits election timing as an avenue by which politicians might pursue their goals of holding power when manipulating voters’ proclivity to use recent economic performance in their vote choice. However, another plausible mechanism exists: the manipulation of macroeconomic policy levers. By reconsidering the literatures on political business and political budget cycles in light of the dissertation’s findings about election timing and outcomes, a number of additional implications for the use of both macroeconomic policy and election timing arise. By surfing the economy, prime ministers who have significant capacity to call early elections have fewer instances in which they face the end of the term and the associated incentivve to boost output, reducing the frequency of electorally motivated use of fiscal and monetary policy instruments. However, because there are fewer instances of macroeconomic policy manipulation, those instances in which they do manipulate ought to be more of a ‘surprise’ to markets, increasing the magnitude of their effect relative to other similarly sized manipulations in countries where such policies are more commonplace.

Additionally, changes in the ability to engage in short-term economic management due to changes in exchange rate regime and/or central bank independence (Fleming, 1962; Mundell, 1960) ought to produce changes in the frequency of premier timed elections and the economic circumstances under which incumbents are willing to risk early elections. As incumbents become more constrained in their ability to manipulate the economy, they are willing to dissolve parliament earlier in the term, less willing to go to the end of the term, and dissolve under less beneficial economic conditions than when they have these levers at their disposal. Where dissolution is more difficult, these changes ought to reduce average support for, and retention of, incumbents ought to fail. The book will empirically examine these implications for politician behavior and economic performance by accounting for exogenous changes in capacity for policy manipulation (e.g., introduction of the euro or the end of Bretton Woods).

Additional Implications for Accountability
I am engaged in several additional projects related to democratic accountability and its implications for prime ministerial replacement during parliament, campaign behavior, expectations of government responsibility in new democracies, and the differential punishment of left-partisans.

I collaborated with Rob Salmond (Polity Analytics) on “Moments of Clarity: Election Campaigns, Clarity of Responsibility, and Economic Voting,” in which we argue that political institutions affect what issues politicians focus on in campaigns. Political institutions featuring high clarity of responsibility (consolidated political power) provide politicians the clearest incentive to make the economy a central plank of their election campaigns. This elite behavior leads voters to view the economy as more salient, to learn more about the economy, and therefore to treat the economy as a more important part of their vote decision. We find generally strong support for this argument with statistical tests using data from the Comparative Manifestos Project and the Eurobarometer series. Politicians talk more about the economy when political power is highly concentrated and the economy is performing poorly than they do when it is performing poorly and power is diffuse. High clarity is associated with stronger economic voting and increased salience of the economy to voters because it increases how much politicians focus on the economy during the campaign.

I am developing a project that explores how deteriorations in the economy affect the likelihood of incumbent party leader replacement (with my adviser, William Roberts Clark, Texas A&M University). We argue that when the economy deteriorates, incumbent parties are incentivized to replace their leader (and or their finance/budget minister) in anticipation of punishment at the polls. This decision is contingent on where in the electoral cycle the downturn—and particularly the trough—occur, the leader’s length of tenure, and their degree of control over parliamentary agenda (building on the election campaign findings described above). We posit that lead party vote share and likelihood of maintaining preeminence in government will both rise for parties that oust their leaders relative to parties that retain them.

In a paper with Jennifer Miller-Gonzalez (recent Michigan Ph.D.), we explore how the legacies of old institutions (both predating and during the communist era), as well as the differing neo-liberal policies implemented in the 1990s across Central and Eastern Europe affect expectations of state responsibility and satisfaction with democracy in the post-Cold War era. Using data from the World Values Survey and the European Social Survey, we consider how culture, history, and institutions constrain and shape policy, and policy, in turn, affects culture. We find that historical institutions and cultural attributes can explain much of the cross-national variation in satisfaction with current institutions, but that cohort, income, and attachments to the old communist party all help to explain within country variation.

Finally, I am completing, “Double Jeopardy: The Luxury Vote and Asymmetries of Partisan Accountability,” that explores whether and why there exist asymmetries in the degree to which incumbents are punished based on their partisanship (with Mark Andreas Kayser, Hertie School of Governance). We argue that during downturns, voters are less supportive of parties proposing ‘luxury’ policy goods (e.g., environmental protection). Because left-wing parties are likely to be associated with ‘luxury’ policies, left-wing incumbents are punished more harshly for economic downturns than their right-wing counterparts. This runs counter to the expectations of many pundits following the 2008 financial crisis. Using survey evidence from both the Eurobarometer Trend File and the CSES, we find that left-wing parties are in fact more harshly punished during downturns. We then go further by testing our causal mechanism against several others that would produce similar observational regularities through two experiments. We find that the reputation for pursuing luxury policies harms left parties in downturns, and is the reason voters turn strongly away from the left in recession.

Formation of Inflationary Expectations
The final project in progress is a series of papers with Christopher Gandrud (post-doctoral fellow, Hertie School of Governance). Our forthcoming paper at Political Science Research and Methods entitled “Inflated Expectations: How government partisanship shapes monetary policy bureaucrats’ inflation forecasts,” is the first to our knowledge showing a partisan bias in inflation forecasts at the US Federal Reserve Bank. We find evidence that Fed staff economists use partisan heuristics that lead them to systematically predict higher than realized inflation throughout Democratic presidencies and lower than realized inflation throughout Republican administrations. We find little evidence that they are attempting to produce electoral business cycles, meaning that the recent findings of Clark & Arel-Bundock (2013) that electoral monetary cycles exist in the US cannot solely be the product of bad internal information.

We plan to expand this research program in both scope and geography. We have begun a study of inflation expectations and strategic interactions within the Federal Open Markets Committee, and are planning a book project that expands our previous US-based work to a global sample of central banks. Moreover, we will consider the special case of the European Central Bank, where the meaning of incumbency is less clear, to explore the accuracy of their inflation estimates compared to national central banks. We intend to begin a related project on partisan biases in voters’ inflation forecasts with an experimental survey research design.