Retrospective Voting Research Articles

Static and Dynamic Models of Retrospective Voting: A Clarification and Application to the Individual Level

This study draws attention to an important distinction between different types of models in the field of retrospective voting, namely static and dynamic models. Both types of models are often used interchangeably, but their interpretation and implications are fundamentally different. Furthermore, while both types have been used in aggregate‐level studies, so far, most studies on the individual level rely on static models. However, there is a specific research interest in individual‐level dynamic models. This article introduces the dynamic model at the individual level, and discusses its methodological advantages and disadvantages. It concludes with an empirical application, indicating how different model specifications lead to a different interpretation of model coefficients, and the applicability of previous research findings on the individual level.Related ArticlesDyck, Joshua, and Annika Hagley. 2012. “Political Geography, Direct Democracy, and the Reasoning Voter: Spatial Proximity, Symbolic Politics, and Voting on California's Proposition 83.” Politics & Policy 40 (2): 195‐220. https://doi.org/10.1111/j.1747‐1346.2012.00346.xNielson, Lindsay. 2017. “Ranked Choice Voting and Attitudes toward Democracy in the United States: Results from a Survey Experiment.” Politics & Policy 45 (4): 535‐570. https://doi.org/10.1111/polp.12212Stegmaier, Mary, and Michael S. Lewis‐Beck. 2009. “Learning the Economic Vote: Hungarian Forecasts, 1998‐2010.” Politics & Policy 37 (4): 769‐780. https://doi.org/10.1111/j.1747‐1346.2009.00197.x

Populism, the Backlash Against Ruling Politicians and the Possible Malfunctioning of Representative Democracy

The aim of this paper is to investigate the links between lack of trust in ruling politicians and the functioning of a representative democracy. Within a standard principal-agent model of democracy, we show how lack of trust by citizens as reflected by passive beliefs updating may lead to the malfunctioning of representative democracy. We highlight how de facto accountability crucially depends on out-of-equilibrium beliefs, and that this is indeed descriptive of a substantive feature of public opinion that affects the functioning of democracy. Specifically, we show that effective accountability needs more than simple retrospective voting, as it requires voters to believe in the existence of good politicians that always choose according to voters' interests, so that a deviation from bad policies can happen only because the leader is congruent. In this case, the unique equilibrium is an efficient one that maximizes voters' welfare. However, if, on the other hand, the citizens share an overall lack of trust in ruling elites, then there is another inefficient equilibrium, where even the congruent politician behaves badly because of the adverse but rational voters' behavior. This inefficient equilibrium does not depend on fake news or on distorted beliefs or, again, on voters' heterogenous preferences, since the voters' perfectly observe the quality of the policy implemented by the government, are fully rational and share the same interests. This result might contribute to explain the increasing negative perceptions on the working of democracy as due to a self-fulfilling equilibrium.

Voting retrospectively: Critical junctures and party identification

This paper provides evidence for retrospective voting in the very long-term by exploiting a unique quasi-natural experiment of history. We trace the origins of party identification to a critical juncture in the local history of Sasun, a mountainous region of the Ottoman Empire located in Eastern Turkey. Sasun received vital support from the Armenian Revolutionary Federation (ARF) both during the Great Massacres against Armenians at the end of the 19th century and during the Armenian Genocide (1915–1917). With the help of the ARF, some of the survivors from Sasun were resettled in various villages in modern-day Armenia. Although the party was not active in Armenia during seven decades of Soviet rule, we find that villages with Sasun ancestry display substantially higher electoral support for the ARF than other villages. Evidence from first names of current residents and our field work suggest that this differential support can, at least in part, be explained by historical gratitude and sympathy for the party. We offer suggestive evidence to explain why this sympathy might have endured over generations.

Electoral systems and immigration

Unexplored stylized facts on OECD countries suggest that plurality electoral systems are associated with higher openness to immigration. We propose an explanation based on a retrospective voting model where immigration hurts voters but benefits a rent-seeking policymaker who appropriates part of the income generated by immigrants. To be reappointed, the policymaker must distribute the compensation. With respect to proportional systems, plurality systems make it possible to compensate only a few decisive districts and leave after-compensation rents higher, therefore producing higher immigration. In our model, non-decisive districts receive no compensation at all under both electoral systems, providing a rationale for widespread anti-immigration attitudes. Notably, our results also help to explain why governments often seem more pro-immigration than do voters. Finally, our model predicts that opposition to immigration is more geographically dispersed in plurality systems. Basic evidence supports this prediction.

Partisanship as Tradition: Critical Junctures, Collective Memory and Persistent Party Identification

This paper provides evidence for retrospective voting in the very long-term by exploiting a unique quasi-natural experiment of history. We trace the origins of party identification to a critical juncture in the local history of Sasun, a mountainous region of the Ottoman Empire located in Eastern Turkey. Sasun received vital support from the Armenian Revolutionary Federation (ARF) both during the Great Massacres against Armenians at the end of the 19th century and during the Armenian Genocide (1915-1917). With the help of the ARF, some of the survivors from Sasun were resettled in various villages in modern-day Armenia. Although the party was not active in Armenia during seven decades of Soviet rule, we find that villages with Sasun ancestry display substantially higher electoral support for the ARF than other villages. Evidence from first names of current residents and our field work suggest that this differential support can, at least in part, be explained by historical gratitude and sympathy for the party. We offer suggestive evidence to explain why this sympathy might have endured over generations.

Do Perceptions Match Provision? Investigating the Gap between Levels of Service Provision and Citizen Perceptions in Sub-Saharan Africa

Do citizens' evaluation of government services accord with actual service provision by the government? Some argue that citizens are able to reward or sanction politicians on the basis of service provision and governance quality. Known as retrospective voting, this theory allows even uninformed voters to evaluate politicians while still not following politics or party platforms. Many authors, however, have questioned the core assumption of retrospective voting theories: that citizens are even capable of accurately evaluating service provision. When faced with the same objective level of service provision, two voters may evaluate that service differently, leading to a in evaluating services. This project investigates the perception gap, using a finite mixture model to test competing explanations for why citizens might mis-evaluate government service provision. We find that while the presence of the perception gap is difficult to predict, the direction of mis-evaluation is explained by sociotropic effects more than respondent identity or respondent engagement, though with major inter- and intra-country heterogeneity.

한국 지방선거의 다층적 회고적 투표 : 2018년 광역단체장선거를 중심으로

한국 지방선거의 다층적 회고적 투표 : 2018년 광역단체장선거를 중심으로

Election or Disaster Support?

In this study, I explore whether the public spending provided in response to a natural disaster is affected by upcoming elections. The theory on retrospective voting argues that politicians are kept responsible by the electorate for the damage caused by a natural catastrophe. At the same time, voters reward government officials when they react promptly by taking actions that limit the negative consequences. Thus, disaster support in the aftermath might not always only be provided to improve the economic situation after a disaster, but also for political purposes. My estimates seem to confirm this idea since roughly 10 per cent of the disaster related public spending provided in an election year is attributed to rent seeking rather than need. It turns out that the existence of these election cycles can partly be explained by cross-country differences. For instance, elections have a stronger effect in countries that are characterised by fewer checks and balances, presidential system and majority elections. One interesting result is that the role of media in electoral manipulation changes when a disaster has happened in the year before.

Retrospective Voting on the Party Level: An Icelandic Test

This research note adds to the emerging body of literature arguing that retrospective voting works on the level of political parties – for government and opposition parties alike – by investigating the generalisability of previous research findings. Furthermore, it tests whether there is a knowledge gap in retrospective voting on the party level. Using the data of the Icelandic National Election Study (ICENES), support is found for the argument that mechanisms of electoral accountability work both for incumbent and opposition parties. Second, while previous research raised doubts on the electorate’s ability to hold governments accountable, there is no evidence of a knowledge gap in retrospective voting on the party level.

Voting for secular parties in the Middle East: evidence from the 2014 general elections in post-revolutionary Tunisia

ABSTRACTArab uprisings paved the way for democratic elections in the Middle East and North Africa region. Yet countries in this region, except for Tunisia, were not able to maintain further democratisation. Tunisia, regardless of economic turbulence and security problems, managed to hold its second parliamentary elections in October 2014, and Ennahda, the party of the popular Islamist movement, could not keep mass support. A large number of studies have examined the rise of the Islamist parties as their electoral success in the post-Arab Uprisings elections by focusing on their organisational strength as well as their social services. However, the social basis of secular parties in the region has been overlooked in the democratisation literature. In this study, four competing arguments, religious–secularism cleavage, nostalgia for the old regime, negative campaign targeting Islamists, and retrospective voting, are considered as the key determinants of citizens’ party choices. By using original election survey data, this study asserts that secular Nidaa Tounes derived its support from secular people, who, at the same time, sympathised with the old regime and disfavoured Islamists.

Who is responsible? The effect of clarity of responsibility on voter turnout

Does voters’ ability to discern who is responsible for policy outcomes affect voter turnout? Although particular institutional arrangements which influence this ability – known as clarity of responsibility – appear to affect how voters form retrospective judgements, existing literature is less informed about its role on voter turnout. This article argues that voters tend to turn out less if they cannot discern who is responsible for policy outcomes. This lack of clarity hinders the process of retrospective evaluations, makes the electoral stakes less profound, and dampens the voters’ political efficacy. Using 396 elections in 34 established democracies between 1960 and 2015, it is found that lower clarity of responsibility is associated with lower voter turnout. This study highlights the importance of clarity of responsibility, as it enhances democratic accountability, not only by encouraging retrospective voting, but also by increasing political participation.

RETROSPECTIVE VOTING AND PARTY POLARIZATION

AbstractWe provide a new and favorable perspective on voter naiveté and party polarization. We contrast sophisticated (Nash) versus retrospective voting in a model where two parties commit to policies. Retrospective voters do not understand the mapping between states and outcomes induced by a policy; instead, they simply vote for the party that delivers the highest observed performance, as determined in equilibrium. We show that parties have an incentive to polarize under retrospective, compared to Nash, voting. Moreover, this polarization often results in higher welfare due to a better match between policies and fundamentals.

Weathering the Storm: Conditional Effects of Natural Disasters on Retrospective Voting in Gubernatorial Elections—A Replication and Extension

Previous research shows that uncontrollable events like natural disasters hinder incumbent leaders’ chances at re-election, but also suggests that competent responses to such crises can benefit incumbents. Replicating and extending the work of other researchers, I show that disasters are informative events in retrospective voting where leaders have the opportunity to demonstrate competence and be rewarded by voters while incapable leaders face punishment for their failure to respond to such a crisis. Governors see greater electoral rewards for demonstrating competence when disaster damage is high as well as when the political situation is more difficult. In the context of a leader’s decision-making process, this reveals a complex and conditional relationship between voters and the leaders they are evaluating.

Electoral Accountability, Responsibility Attributions, and the Democratic Deficit in Devolved Wales

Abstract This study explores electoral accountability at the devolved level in Wales. The retrospective voting model assumes that electoral accountability requires that citizens assign responsibility accurately and vote on the basis of their evaluation of government performance. Drawing on data from the 2016 Welsh Election Study, we examine whether citizens in Wales can identify devolved policy competences and office holders, and identify factors that are linked to accurate attributions. We then examine whether voters seek to use devolved elections as a sanctioning tool, even if they do not have the information required to do so accurately. The findings indicate that there is an acute accountability deficit at the devolved level in Wales: few have the knowledge or the inclination to hold those in power to account. The conclusion discusses the implications of these findings for democracy in Wales and in other multi-level settings.

Comprehensive Retrospective Voting in Mixed Electoral Systems: Evidence from the 2016 Korean Legislative Election

AbstractPerformance-based retrospective voting is a fundamental mechanism of democracy. A good deal of scholarship has examined this electoral mechanism, but the extant studies have two omissions. First, there is little research that considers several retrospective evaluations together using an incumbent voting model. Second, there is little research that examines the difference in the effects of voters’ retrospective evaluations on two different ballots in mixed electoral systems. To fill these omissions, this article tests a comprehensive retrospective performance voting model in a mixed electoral system. Specifically, this article examines the effects of voters’ retrospective economic evaluations of economic performance at the national and personal levels, human rights, corruption, welfare protection, and foreign policy on vote choice for the incumbent party in the 2016 Korean legislative election in which voters had two ballots: one for the party list vote and one for the district vote. By using multinomial logistic regression models, this article finds that among the six retrospective evaluation categories, judgments of national economic performance at the national level, human rights, and foreign policy have a statistically significant impact on the likelihood of voting for the incumbent party in party list vote choice, whereas only voters’ evaluation of foreign policy matter in the district level vote decision. The results imply that Korean voters consider various aspects of government performance, such as the conditions of human rights and relationships with other countries, rather than just focusing on the economy. The retrospective voting behavior of Korean voters differs between party list and district level ballots.

Accountability for Realists

ABSTRACTIn Democracy for Realists, Christopher Achen and Larry Bartels argue that voters are shortsighted and punish incumbents for politically irrelevant outcomes. These failings, in the authors’ view, mean that voters are incapable of holding politicians to account. But Achen and Bartels overstate voters’ failure to engage in effective retrospective voting. The authors also understate the degree to which accountability can be compatible with voters’ being myopic, such as when early- and late-term performance are correlated. Achen and Bartels also overlook evidence that the American state acted as an insurer against social risk long before the New Deal, a fact that points to voters using relevant criteria (not irrelevant ones) when they appear to punish incumbents for natural disasters. Finally, while accountability is an important consequentialist reason for democracy, we should keep in mind non-consequentialist justifications for this system of government, in particular its instantiation of political equality.

The three stages of the anti‐incumbency vote: Retrospective economic voting in young and established democracies

AbstractEconomic prosperity is the best recipe for an incumbent government to be re‐elected. However, the financial crisis was significantly more consequential for governing parties in young rather than in established democracies. This article introduces the age of democracy as a contextual explanation which moderates the degree to which citizens vote retrospectively. It shows a curvilinear effect of the age of democracy on retrospective economic voting. In a first stage after the transition to democracy, reform governments suffer from a general anti‐incumbency effect, unrelated to economic performance. In a second step, citizens in young democracies relate the legitimacy of democratic actors to their economic performance rather than to procedural rules, and connect economic outcomes closely to incumbent support. As democracies mature, actors profit from a reservoir of legitimacy, and retrospective voting declines. Empirically, these hypotheses are corroborated by data on vote change and economic performance in 59 democracies worldwide, over 25 years.

Long-term evidence of retrospective voting: A natural experiment from the German Democratic Republic

The paper investigates long-lasting electoral punishment. Decades of communist socialization and the repressive rule of a single-party have left their left-wing fingerprint on East Germany. In this paper we show that voters act rationally: given negative life circumstances experienced under the rule of the communist party, they display retrospective voting even decades later. Our insight is based on the analysis of 19 years of revealed and stated party preferences. We argue that life at the Border Region to West Germany was particularly hard and find that East German voters who lived close to the inner-German border before the reunification of the two states are 5.9 percentage points less likely to lean toward the successor party to East Germany’s communists. Given that over the years roughly every fifth person has revealed preferences for the communists in the East, this translates to over thirty percent reduction. We confirm the preferences with administrative data: The electoral punishment estimated at the district-level amounts to a reduction of 1.3 percentage points of votes for the party.

Beyond the distinction between incumbency and opposition: Retrospective voting at the level of political parties

Retrospective voting is one of the most frequently investigated theories of voting behaviour. It postulates that voters punish or reward incumbent parties according to their performance. This approach, however, cannot explain why voters cast a stable opposition vote, or switch between incumbent parties or opposition parties, respectively. This study fills this gap by showing that this traditional view should be complemented with the approach of retrospective voting to work at the level of political parties – irrespective of their incumbency status. Using the data of the second module of the Comparative Study of Electoral Systems, it provides the first large-scale test of this new perspective on electoral accountability. The results show that satisfaction with the previously endorsed party is an important determinant of the current party choice, and that it is so for voters for all parties alike. These results provide a theoretical foundation for party switches that could not be explained by the traditional model.

The Political Impact of Monetary Shocks: Evidence from India's 2016 Demonetization

What are the political consequences of monetary shocks? While theories of retrospective voting suggest that voters should punish incumbents for negative economic shocks within incumbents' control, other works suggest voters' ability and willingness to do so are limited. We examine this issue by tracing the impact of the sudden demonetization of 86% of India's currency in 2016. We argue and show, using difference-in-difference and instrumental variables analyses, that the economic impact of demonetization was felt most acutely in relatively areas, where households and businesses lack access to the formal financial system and are most cash-dependent. Further, the BJP, the incumbent party that implemented demonetization, was penalized the least in relatively unbanked districts where people were economically hurt the most. We speculate why this might be the case. Our findings illustrate that monetary policy has important distributional consequences that shape voting.