PL16CH14-Malhotra ARI 17 April 2013 14:49
Retrospective Voting
Reconsidered
Andrew Healy1
and Neil Malhotra2
1
Department of Economics, Loyola Marymount University, Los Angeles, California;
email: andrew.healy@lmu.edu
2
Graduate School of Business, Stanford University, Stanford, California;
email: malhotra_neil@gsb.stanford.edu
Annu. Rev. Polit. Sci. 2013. 16:285–306
First published online as a Review in Advance on
February 28, 2013
The Annual Review of Political Science is online at
http://polisci.annualreviews.org
This article’s doi:
10.1146/annurev-polisci-032211-212920
Copyright c 2013 by Annual Reviews.
All rights reserved
Keywords
citizen competence, voting behavior, economic voting, democratic
accountability
Abstract
We review advances in the study of retrospective voting, or how citizens evaluate
and act on their perceptions of government performance. As a whole,
the recent literature provides a more complete and nuanced picture of the
retrospective voter as sometimes, but not always, effectively incentivizing
elected officials to enhance public welfare. Leveraging examples of retrospective
voting in areas other than the economy, the field is heading toward
a middle ground in which voters resemble decision makers in many other
domains. In many cases, a coherent logic governs voters’ choices. In other
instances, voters make mistakes, often in predictable ways subject to wellknown
psychological biases. Understanding the circumstances under which
retrospective voting achieves effective democratic accountability and when
it fails to do so is an important task for subsequent research. We discuss two
additional issues for future exploration: a better understanding of normative
benchmarks, and increased attention to the relationship between retrospective
voting behavior and policy outcomes.
285
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
Click here for quick links to
Annual Reviews content online,
including:
• Other articles in this volume
• Top cited articles
• Top downloaded articles
• Our comprehensive search
FurtherANNUAL
REVIEWS
PL16CH14-Malhotra ARI 17 April 2013 14:49
INTRODUCTION
The evaluation of incumbent performance at the ballot box plays a key role in democratic accountability
(Besley 2006, Ashworth 2012). Through this retrospective behavior, citizens can incentivize
politicians by sanctioning poor performance (Ferejohn 1986) and selecting leaders who will govern
competently and honestly (Fearon 1999). Democratic governance is thus normatively appealing
because elections should be an effective means of enhancing public welfare. Underlying these theoretical
conceptions is a model of retrospective voting wherein citizens consider information on
past government performance in order to make forward-looking decisions. Many models of political
economy assume that voters behave in such a fashion (e.g., Persson et al. 1997, Canes-Wrone
et al. 2001).
Accordingly, scholars of political behavior have long analyzed how citizens consider events and
outcomes in deciding whether to retain the incumbent (party). Challenging the Michigan School’s
conception of the American voter as lacking both knowledge about political issues and a coherent
ideological structure (Campbell et al. 1960, Converse 1964), V.O. Key stated in The Responsible
Electorate (1966) that “voters are not fools.” He argued that voters were capable of executing the
vital tasks of citizenship required in a republican democracy. In Key’s view, voters updated their
beliefs based on government performance.
Nearly every article and book on retrospective voting—the study of how citizens evaluate and
act on their perceptions of government performance—cites The Responsible Electorate. As a result,
the retrospective voter has generally been conceived of as a foil to the Michigan voter. Scholars in
this research tradition have argued that if we observe that voters respond to changes in objective
performance indicators—such as an increasing unemployment rate or contracting gross domestic
product (GDP) growth, as shown by Kramer (1971) and Fair (1978)—then the electorate as a
whole holds policy makers accountable in an effective manner even though most individual voters
possess little political knowledge and are generally disinterested in politics (Page & Shapiro 1992).
As Fiorina argued in Retrospective Voting in American National Elections (1981), citizens can
implement effective accountability even if they lack basic political knowledge by focusing on simple
performance metrics. Voters just have to answer the kinds of questions that Ronald Reagan posed
in his closing statement in his 1980 debate against Jimmy Carter: “Are you better off than you were
four years ago? Is it easier for you to go and buy things in stores than it was four years ago?” The
answers to these questions and related ones represent easily accessible and digestible information.
Encyclopedic knowledge is not needed; these shortcuts are sufficient heuristics for voters to judge
overall government competence (Lupia 1994). Voters may judge incumbents based on personal
pocketbook considerations or sociotropic attitudes (Kinder & Kiewiet 1981, Kramer 1983, Markus
1988). Regardless, all the key pieces of information are easily accessible. The pocketbook voter
need only consult his bank account; the sociotropic voter can consult the news to assess the health
of the economy.1
However, recent research casts doubt on this sanguine view of retrospective
voting. Voters do not merely lack knowledge. They appear to make substantial, consistent, and
correlated errors, often holding politicians accountable for actions beyond their control. The effect
sizes are often small, suggesting that these errors may make a difference only in close elections.
No one would claim that voters are either perfectly sophisticated optimizers or completely
clueless rubes. Nonetheless, both the Michigan model and the rational choice approach have
1
At the same time, the media can distort economic perceptions. For example, Hetherington (1996) shows that the way the
media frames the health of the economy can significantly affect how voters construct their sociotropic assessments. Similarly,
Vavreck (2009) argues that campaign strategy not only is shaped by economic conditions but also influences voters’ perceptions
of the economy.
286 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
been mischaracterized and caricatured over the decades (Cox 1999, Converse 2006). Perhaps the
problem is that much of the literature frames retrospective voting as a salvation for democratic
competence and accountability. This divide between the Michigan voter and the retrospective
voter as portrayed in the literature is unfortunate and unnecessary. Instead, recent research suggests
a middle ground where voters sometimes, but not always, make mistakes. Moreover, these
errors are consistent with psychological biases that influence people’s decisions in domains outside
of politics. A rich and accurate model of voter behavior needs to account for sensible voter responsiveness
but also these important biases, identifying the conditions under which retrospective
voting achieves effective democratic accountability and when it fails to do so.
To provide the background for this hybrid model, we review the literature on retrospective
voting over the past decade or so. The field has experienced a renaissance in recent years largely
driven by an emphasis on research design (e.g., controlled and natural experiments) to address issues
of causality and to combat persistent concerns such as omitted variables bias, reverse causality,
and simultaneity (Angrist & Pischke 2009, 2010). Coincident with the increased focus on design
has been the flourishing of creative approaches and applications to study retrospective voting.
Scholars have moved beyond studying how the health of the economy (perceived or actual) influences
political behavior. We have learned a great deal over the past decade about how voters
actually make decisions and what their behavior implies for democratic accountability and public
welfare. Yet these advances in knowledge have also generated more open questions and areas for
future research.
Before we begin our summary of the recent literature, we mention topics related to retrospective
voting that this article does not discuss. We do not review research on “policy performance,”
that is, whether policy actions taken by the incumbent government reflect promises made during
the campaign or represent citizens’ preferences (e.g., Downs 1957, Miller & Stokes 1963,
Canes-Wrone et al. 2002). Rather, we restrict our attention to studies of how voters assess whether
incumbent governments have competently managed the economy, war, public safety, and the like.
By and large, the domains we discuss are ones in which all citizens have common interests, such
as a prosperous and peaceful society. Although research on this topic has been increasingly and
interestingly studied from a comparative perspective (and we do review some studies from non-US
contexts), we generally position our discussion in relation to the American politics subfield.
This article builds on Anderson’s (2007) review of economic voting in this journal. He argues
that retrospection is moderated by the cognitive limitations of voters as well as political institutions.
Anderson contends that economic voting sometimes does not take place, thereby hindering
democratic accountability and creating “contingency dilemmas.” We depart from Anderson by
arguing that voter responsiveness to the economy (and other performance metrics) does not enhance
public welfare per se. Recent research that has explored retrospective voting both within and
outside of the economic domain has sometimes found that voter responsiveness creates bad incentives
for leaders. Therefore, the relevant question may not be “Do voters react to government
performance or not?” but rather “Are voters reacting in the right way?”
The research that we describe here answers the following questions: How do voters evaluate
government performance? What does this imply about the health of American democracy and the
policies it produces? These two questions suggest that this research agenda is both positive and
normative. From a positive perspective, we aim to understand and analyze voters’ behavior and the
cognitive bases of their decisions. Yet, a primary reason we care about this behavior is to compare it
to some normative benchmark of how voters should be incentivizing politicians to pursue policies
that maximize public welfare. Beyond ensuring that citizens have rights, a main argument in favor
of liberal democracy is that elections produce good outcomes via the processes of sanctioning
and selection. However, if voters’ decisions sometimes fail to provide appropriate incentives to
www.annualreviews.org • Retrospective Voting Reconsidered 287
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
elected officials or distort those incentives altogether, democracy may produce inefficient policies
in those cases. Also, similar to classic collective action problems, individually rational behavior
by voters can sometimes produce bad outcomes for society. The literature thus suggests that
retrospective voting does not automatically ensure good incentives, particularly given voter errors.
Understanding exactly how and under what conditions voter biases impact public welfare remains
an important research agenda.
THREE MODELS OF RETROSPECTIVE VOTING
As mentioned above, scholars have extensively documented correlations between performance
indicators, particularly economic ones, and election outcomes. From a theoretical perspective, at
least three models of voting behavior are consistent with these empirical patterns. Two of these
are the familiar sanctioning and selection models, while the third is rooted in psychological biases.
These models are not mutually exclusive. Voters whose decisions are subject to some psychological
biases may be attempting to implement decisions that accord with a rational choice model.
The most traditional interpretation of retrospective voting is a reward-punishment, or sanctioning,
model. In the language of rational choice, voters (principals) are attempting to reduce
moral hazard on the part of elected representatives (agents). This was Key’s (1966) informal
description of retrospective voting, later formalized by Barro (1973) and Ferejohn (1986). By reelecting
high-performing politicians and throwing out poor performers, voters incentivize good
behavior on the part of politicians. As Key noted, it is the threat of marginal voters abandoning
their party (and nonaligned voters swinging back and forth) that induces accountability.
A second model, also in the rational choice framework, describes retrospective voters as selecting
the leaders who will perform most competently after being elected (Fearon 1999). After
learning about an incumbent’s quality through his or her performance in office, voters can choose
to reelect a competent leader or take their chances on an unknown challenger (Persson & Tabellini
2002, Duch & Stevenson 2008). In the long run, elections serve the process of selecting good performers
(Ashworth 2005, Padro-i-Miguel & Snyder 2006). Using the identification strategy of
comparing term-limited and non-term-limited governors, Alt et al. (2011) found empirical evidence
consistent with both mechanisms. Importantly, both the sanctioning and selection models
suggest that retrospective voting serves the normatively beneficial function of reducing moral
hazard and adverse selection.
A third interpretation of the relationship between performance indicators and election outcomes
incorporates cognitive and emotional biases. Voters may find it difficult to process all
relevant information and instead rely on cognitive shortcuts when faced with difficult problems.
Conceiving of voters using heuristics follows in Kahneman and Tversky’s tradition (see e.g.,
Kahneman et al. 1982, Kahneman 2003) and corresponds to how people appear to make decisions
in fields ranging from finance to consumer behavior (e.g., Shefrin 2002, Tufano 2009). Heuristics
often lead people to make optimal economic decisions but sometimes lead them astray. The same
pattern appears to be true in politics: party cues or endorsements can be reasonable guides in most
cases (Lupia 1994, Lupia & McCubbins 1998, Lau & Redlawsk 2001), but such heuristics can also
produce significant mistakes. As many authors have discussed, the election-year economy has a
large impact on voting behavior whereas conditions in other years have little effect on election
outcomes (e.g., Kramer 1971, Fair 1978, Tufte 1978, Achen & Bartels 2004, Bartels 2008). This
behavior is consistent with either of the two rational choice models above (e.g., MacKuen et al.
1992). For instance, the explanation could be that voters gain the most information about incumbent
quality from recent performance, so it may be sensible for people to ask, “What have you
done for me lately?” However, Healy & Lenz (2012) and Hill et al. (2012) argue that at least part of
288 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
the explanation for voters’ election-year emphasis comes from a pure cognitive bias. Voters find it
hard to think about total growth over a president’s entire term. Even though people say that they
would like to base their decisions on the full set of information, they substitute a more easily available
attribute—recent economic performance—for the less accessible attribute they are searching
for (Healy & Lenz 2012). Therefore, voters’ decisions may be subject to the “availability heuristic”
(Tversky & Kahneman 1973). Such reasoning can also help explain the power of the status quo
because it may be a powerful reference point for voters to consider (Kahneman & Tversky 1979).
In addition to predictable cognitive biases, emotions may influence voters so that they sometimes
reward or punish politicians simply because they feel happy or sad for reasons that have
nothing to do with incumbent performance (Achen & Bartels 2002). In other words, what may
look like sanctioning and selection may reflect voters’ emotional reactions to the current state of
affairs. There are two potential mechanisms for this. First, voters in a bad mood may misattribute
their emotional state to the incumbent (Schwarz & Clore 1983). Second, voters may be more
likely to remember positive (negative) actions of the incumbent when they are in a good (bad)
mood (Bower 1981).
This psychologically based third model paints a less sanguine portrait of retrospective voting
than traditional interpretations based in rational choice. Cognitive and emotional biases may
compromise democratic accountability. The use of conditions at the end of a president’s term to
stand in for overall economic performance at a minimum may lead elections to be highly dependent
on the vagaries of the business cycle (Achen & Bartels 2004) and can also distort incentives toward
inflationary policies in election years (Tufte 1978, Rogoff 1990, Abrams 2006). Similarly, mood
may be a useful heuristic in many decision-making environments, but it is generally not closely
tied to actual government performance. For instance, if the president is managing the economy
effectively but it is still not performing well owing to external factors or not achieving some
arbitrary threshold of performance, voters in a bad mood may punish the incumbent even if
sanctioning or selection would call for the incumbent to be retained. Furthermore, events that
are loosely related or unrelated to government performance can significantly affect mood but may
not be grounds for reward and punishment.
Although a great deal of research adopted the two rational choice mechanisms as theoretical
starting points, the third model has been gaining some empirical traction. However, it is important
to make a conceptual distinction between what voters are trying to do and what voters actually do.
Research in the psychological tradition does not argue that voters lack a desire to minimize moral
hazard and adverse selection in the electoral process. Instead, the psychology literature suggests
that even if voters have this motivation, they may not be doing what they intend or even what they
think they are doing.
Retrospective voting can be considered a four-step process that forms a feedback loop. Voter
psychology is particularly important in the first two steps and thereby can affect the entire process.
Voters observe events in the world (e.g., disasters, wars), outcomes (e.g., macroeconomic statistics,
test scores), and policy actions taken by elected officials (Step 1). They then attribute responsibility
for these events, outcomes, and actions to particular elected officials (Step 2). These attributions
then lead people to evaluate the performance of officeholders, thereby influencing their voting
decisions (Step 3). The manner in which events, outcomes, and actions are translated into election
results creates incentives for elected officials, thereby influencing the formation of policy both
before and after elections take place (Step 4). Finally, these policy choices feed back into the
events and outcomes that voters observe.
The research we review can be thought of as examining various pieces of this overall process.
Existing studies, which most often focus on Steps 2 and 3, generally do not consider the policy
implications of voter behavior. There is growing evidence that voters make mistakes in Steps 1–3
www.annualreviews.org • Retrospective Voting Reconsidered 289
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
according to a strict definition of rationality. It is unsurprising that voters fail to adhere to such
a strict standard; decision makers in other areas with much more at stake often display bounded
rationality. However, the question of normative importance is whether voters make the kinds of
mistakes that will impact Step 4. Here, the evidence indicates that voting behavior often introduces
a considerable amount of noise into democratic accountability, potentially distorting policy-maker
incentives in some cases.
RECENT ADVANCES IN ECONOMIC VOTING RESEARCH
Before describing new directions that the retrospective voting literature has taken beyond the
classic economic voting research agenda, we describe the work on that much-explored topic since
the publication of Anderson’s (2007) comprehensive review. In particular, we consider (a) recent
evidence on the influence of local economic conditions on voting behavior and (b) the comparative
politics literature exploring how institutions condition the magnitude of the economic vote.
The Local Context
Conditions prevailing in the local economy appear to bias people’s perceptions of the national
economy. In part due to the measurement error associated with measures of local economic conditions,
this topic has received relatively little attention. However, creative research designs and new
geographic analysis techniques have added to what we can say about the local economy’s influence.
Gas prices, the housing market, and local unemployment all appear to influence perceptions of
national economic conditions (Mondak et al. 1996, Books & Prysby 1999, Ansolabehere et al.
2012, Reeves & Gimpel 2012).
Nonetheless, the effects are fairly small. For instance, Reeves & Gimpel (2012) find that a
one-standard-deviation increase in the county-level home foreclosure rate in the four months
before the 2008 presidential election increased negative evaluations of the national economy by
0.026 units on a five-point response scale, with somewhat larger effects in states most directly
affected by the housing crisis. In Nevada, a one-standard-deviation increase in foreclosure rates
was associated with a 0.15-unit increase in negative economic perceptions, greater than the effect
of unemployment status. Local peak gas prices in the summer before the election did not affect
economic attitudes overall, but among independents a $1 increase in the price of gas was associated
with a 0.062-unit decrease in the economic perceptions scale. Similarly, analyzing American
National Election Studies (ANES) data from 1980–2008, Ansolabehere et al. (2012) find that a
one-percentage-point increase in the state unemployment rate decreases national retrospective
economic evaluations by 0.05 points on a five-point scale.
Not only do local conditions affect economic perceptions, but those conditions may affect
election outcomes differently according to the type of shock that occurs. Margalit (2011) finds
that American voters respond to layoffs in their county and somewhat more strongly to layoffs that
are related to trade. He reports that a one-percentage-point increase in the share of the county
workforce losing a job because of foreign competition reduced President Bush’s vote share by
0.15 percentage points in the 2004 election, double the effect of job losses caused by other factors.
Voters, he finds, may particularly hold presidents accountable for losses due to outsourcing and
similar phenomena that voters may (perhaps incorrectly) trace to trade-related policies.
Institutional Context and the Economic Vote
Recent research has supplemented earlier work on the importance of institutional context in
moderating retrospective voting. Duch & Stevenson (2008) show how institutional context affects
290 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
the magnitude of the economic vote in theoretically plausible ways. Examining 163 election surveys
from 18 countries, they provide some of the clearest evidence that economic voting appears to
be more prevalent in places where responsibility for economic policy is most easily assigned to
one party and where economies are less subject to external forces, among other factors. Although
the authors argue forcefully for the selection model, they more generally contend that voters
respond sensibly to the conditions that they observe.
Voters, the authors find, manage to incorporate subtle information about potential future
governments in their decision making. Specifically, voters think about the alternative governments
that might be formed when deciding how economic conditions should affect their vote for a given
party. Voters respond less strongly to economic performance when the incumbent party can
potentially belong to many possible coalitions, not only to the existing incumbent coalition.2
Such
behavior is predicted by a selection model, but not by a sanctioning one, where the responsibility
for past policy is all that matters. Duch & Stevenson’s (2008) model places significant demands
on voter awareness of coalition outcomes. Given limited political knowledge among voters, such
findings are perhaps surprising but suggest that citizens may figure out what they need to know
to select effective governments. European voters, in comparison to American voters, may need to
be more politically sophisticated in order to deal with the more complex forms of parliamentary
government.
Duch & Stevenson (2008) find evidence that voters respond more strongly to performance
when it is easier to assign responsibility for economic outcomes to a given party. For example, their
model predicts the economic vote to be stronger in economies that are less influenced by foreign
trade because open economies are more subject to global forces beyond the government’s control.
Their empirical analyses support this prediction. At the same time, the theoretical expectation
may be unclear because the executive or government may control trade policies and therefore
may be held responsible for trade-related outcomes (per Margalit 2011). However, other results
are more unambiguously consistent with the theory that greater clarity of responsibility enhances
economic voting. For example, Duch & Stevenson find that the concentration of responsibility
for economic policy within a single party is positively associated with the strength of the economic
vote. Moving across the range of their concentration variable increases economic voting by 15–20
percentage points.
Altogether, Duch & Stevenson make a compelling case that voters’ decisions have a coherent
logic that governs their responses to the economy. We view these results as entirely consistent
with other findings that voters’ decisions sometimes make less sense in different contexts.
UNPACKING THE BLACK BOX OF ATTRIBUTION
The first wave of retrospective voting research, using both aggregate and individual-level data,
sought to understand how economic performance indicators influence voting behavior. Attribution,
the process by which voters allocate blame and credit to political actors, while not the focus
of that initial research, was assumed to be an intermediate step. Social psychology provides important
insights into how people make these kinds of attributions. In particular, psychologists have
highlighted the “group-serving bias” (Taylor & Doria 1981). People tend to make internal attributions
for positive outcomes enjoyed by their own group and external attributions for negative
outcomes (e.g., Taylor & Jaggi 1974, Fletcher & Ward 1988). With respect to out-groups, the
reverse pattern holds. Hence, much attributional research has focused on how partisanship colors
2
Other evidence also suggests that voters respond to the potential outcomes of coalition formation (Kedar 2005).
www.annualreviews.org • Retrospective Voting Reconsidered 291
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
voters’ allocations of credit and blame to political actors (e.g., Rudolph 2006, Malhotra & Kuo
2008, Brown 2010, Tilley & Hobolt 2011).
Peffley (1984) and Peffley & Williams (1985), who first modeled the attributional process
explicitly, found that the relationship between economic conditions and vote choice was strongest
for people who attributed those conditions to the incumbent’s performance (see also Kinder et al.
1983, Abramowitz et al. 1988, Stein 1990). Several recent research agendas have also leveraged the
central role of attributions in retrospective voting. For example, Gomez & Wilson (2001, 2003)
argue that political sophisticates are more likely to understand complex economic processes and
apportion blame across multiple political institutions.
A series of papers by Rudolph (2003a,b, 2006) argues that both individual-level factors (e.g.,
partisanship) and contextual ones (e.g., divided government) influence attributions of economic
performance. For instance, Rudolph (2003a) showed that in 1998, Democrats who thought the
economy had improved were 13% more likely to attribute credit to President Clinton than to
the Republican Congress; conversely, Republicans were 10% more likely to credit Congress for
economic improvement. Marsh & Tilley (2010) and Tilley & Hobolt (2011) find that these partisan
rationalizations not only affect attributions but also spill over into vote intentions.
Leveraging the different party affiliations of federal and local officials responsible for responding
to Hurricane Katrina, Malhotra & Kuo (2008) confirm that partisanship can affect attributions
but find that citizens’ partisan rationalizations are mitigated by contextual information about officials’
roles. In their experiment, receiving a party cue made Democratic respondents 1.75 times
more likely to blame Republican officials for the deaths and damage stemming from the hurricane;
however, the effect of the party cue disappeared when participants observed simple information
about the Republican officials’ titles. Arceneaux (2006) similarly examines the complex task of
assigning responsibility in a federalist system. He finds that citizens do make meaningful distinctions
among different levels of government in assessing performance but that these attitudes only
translate into vote choice for easily accessible issues. Additionally, citizens’ views of the responsibilities
of local, state, and national governments affect attitudes on which levels of government
should have policy-making influence (Arceneaux 2005).
Similarly, partisanship can bias people’s perceptions of economic conditions themselves and
thereby affect attributions (Conover et al. 1986, Wlezien et al. 1997, Bartels 2002, Evans &
Andersen 2006). In an influential paper, Bartels (2002) showed that Democrats and Republicans
have drastically different views of objective economic realities such as changes in unemployment
rates, inflation rates, and the budget deficit.3
For instance, over half of strong Democrats in
1988 claimed that inflation had gotten worse since 1980 even though it had actually plummeted
nearly 10 percentage points during that time. This bias also affects subjective perceptions of the
incumbent’s economic performance (Evans & Andersen 2006, Evans & Pickup 2010). For instance,
controlling for lagged partisanship reduces the estimated effect of current sociotropic perceptions
on vote choice by a little more than two-thirds in the 1997 British elections (Evans & Andersen
2006). Subsequent research has shown similar patterns of partisan bias in performance evaluations
for other issues such as war and foreign policy (Gaines et al. 2007).
An open question is whether partisan bias in economic perceptions has been increasing over
time as voters have become increasingly well sorted ideologically into parties (Levendusky 2009).
Using ANES data from 1980–2012, we examine Democrats’ and Republicans’ answers to the
3
Bullock et al. (2012) explored whether these partisan differences are simply an artifact of “cheap talk” partisan cheerleading
in survey responses. They asked people similar questions about objective economic conditions but paid some respondents
for answering the question correctly. Offering respondents monetary incentives for providing correct responses substantially
reduced partisan gaps but did not eliminate them entirely.
292 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
Year
1980 1984 1988 1992 1996 2000 2004 2008 2012
0.0
0.5
1.0
1.5
Absolutevalueofpartisangapfor
previousyear'seconomicperformance
Figure 1
Partisan bias in economic perceptions (1980–2012).
following item: “Now thinking about the economy in the country as a whole, would you say that
over the past year the nation’s economy has gotten better, stayed about the same, or gotten worse?”
Respondents who did not select “stayed about the same” were then asked: “Has the economy
gotten much better (worse) or somewhat better (worse)?” We analyzed the resulting five-point
scale of economic perceptions. In each election year, partisans of the incumbent presidential
party evaluated the economy significantly more favorably than did people who identified with the
challenger’s party. In Figure 1, we plot the absolute value of the difference between Democrats’
and Republicans’ responses to the item across the last nine presidential elections along with 95%
confidence intervals.4
In every single election the partisan gap is highly statistically significant and
substantively meaningful, ranging from 7% of the length of the scale to 27%. There does not
seem to be any systematic pattern suggesting that partisan bias has been increasing; however, the
larger partisan divides in 2004 and 2012 may reflect an upward trend that was masked in 2008 by
an economy that was performing so poorly that there was little room for interpretation.
More generally, partisan bias seems strongest when economic performance is middling (as
in 2012) as opposed to being clearly good (as in 1996) or bad (as in 1980, 1992, and 2008). In
Figure 2, we plot the partisan divide against the absolute value of the mean deviation in all
respondents’ overall assessment of the economy from the midpoint position on the response scale
(“stayed the same”). There appears to be an inverse relationship between the size of the partisan
divide and the extremity of the perceived economic state. This analysis is simple but suggestive of
potential influences on partisan bias. It seems clear that partisanship colors economic perceptions,
but future scholarship can examine the conditions under which these partisan divides are large or
small.
4
Following Keith et al. (1992), we pool leaners and partisans together, excluding pure independents from the analysis. Because
the 2012 time-series study was not completed at the time of writing, we used the fourth wave of the ANES Evaluations of
Government and Society Study (EGSS) conducted between February 18 and 23, 2012.
www.annualreviews.org • Retrospective Voting Reconsidered 293
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
0.0 0.5 1.0 1.5
0.0
0.5
1.0
1.5
1980
1984
1988 1992
1996
2000
2004
2008
2012
Absolute deviation in overall assessment
of economy from “stayed the same”
Absolutevalueofpartisangapfor
previousyear'seconomicperformance
Figure 2
Partisan bias and the extremity of the state of the economy.
NEW DIRECTIONS
In recent years, the literature has expanded in three directions distinct from the classic economic
voting literature. First, a series of papers has examined how voters respond to conditions other
than the economy. These noneconomic domains sometimes allow clearer conclusions about the
retrospective voter. Second, researchers have tested for the presence of apparent errors in voter
decision making, most notably looking at voters’ responses to events seemingly beyond any politician’s
control. A third strand of work has explicitly considered how retrospective voting relates to
policy outcomes.
Moving Beyond the Economy
Despite the economy’s obvious importance as a source of retrospective voting, there are several
reasons why it is often difficult to interpret voters’ responses to economic conditions, and these
limitations make it difficult to establish normative benchmarks. First, political economists debate
the extent to which fiscal policy affects the economy, suggesting that federal officials may have
limited ability to influence it (Alesina et al. 1993). Moreover, the condition of the national economy
affects state and local elections (Carsey & Wright 1998), even though it is implausible that elected
officials at those levels have substantial influence on national economic outcomes. Second, it is
unclear whether voters should care more about levels or changes in economic indicators (or maybe
even the second derivative), and which baselines they should be using. Third, voters should perhaps
not only evaluate the current state of the economy but also compare it to what would have happened
if the losing candidate had been elected. Fourth, it is unclear which economic indicators should be
indicative of government performance, particularly since some are negatively correlated and have
different distributional implications (e.g., unemployment versus inflation). Fifth, it is extremely
294 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
difficult to tie the actions of politicians to particular macroeconomic outcomes. Indeed, there is a
debate on when the previous administration should stop receiving credit and blame for the current
administration’s record (see Bartels 2008 and Campbell 2011).
A great deal of contemporary research has attempted to study retrospective voting outside the
economic domain, in part to avoid many (but not all) of the problems listed above. For instance,
Berry & Howell (2007) find that voters evaluate school board members on the basis of student
test scores, but only when the media highlights testing and accountability as an issue. Under
these circumstances, moving across the interquartile range of test-score change is associated with
a three-percentage-point increase in incumbent vote share. Gasper & Reeves (2011) find that
governors are rewarded for asking the federal government for assistance after a natural disaster
(by about four percentage points) and that presidents are punished if they reject the request (by
about one percentage point). Similarly, Healy & Malhotra (2009) demonstrate that voters reward
presidential administrations for delivering relief payments after natural disasters. Examining a
comparative context, Bechtel & Hainmueller (2011) find that German voters who suffered from
the Elbe flooding rewarded the robust response of the Socialist government not only in the election
immediately following the floods but also in a future election. In the immediate aftermath of the
flood in 2002, government response increased incumbent vote share in the affected areas by seven
percentage points; about a quarter of this reward was still present in the next election in 2005.
Building on Mueller’s (1973) pioneering work in the aftermath of the Vietnam War, recent
scholars have leveraged the similarly unpopular Iraq War to demonstrate the relationship between
casualties and vote shares for presidents (Karol & Miguel 2007), senators (Kriner & Shen 2007),
and representatives (Grose & Oppenheimer 2007).5
Karol & Miguel’s results suggest the counterfactual
that if it were not for the 10,000 US casualties in the Iraq War prior to the 2004 election,
President Bush would have won two more percentage points in the national popular vote, producing
a decisive Electoral College victory. All of these examples of noneconomic retrospective
voting—test scores, disaster response, and casualties—have the advantages that the events and outcomes
in question are temporally isolated and can be clearly attributed to a particular incumbent.
Do Voters Make Mistakes?
In addition to the findings on voters’ responses to noneconomic outcomes, a second recent advance
in the retrospective voting literature has come from work challenging the presumption that
retrospective voting reflects individual and collective rationality. This line of research suggests that
biases in citizens’ evaluations of government performance can hamper electoral accountability, in
contrast to the sanctioning and selection models of retrospective voting. We highlight three areas
of voter mistakes that have been documented by this recent literature: (a) the influence of irrelevant
information on voter decision making, (b) citizens’ inability to benchmark, and (c) excessive
weighting of recent events.
An editorial cartoon at the beginning of Fiorina’s Retrospective Voting (1981) depicts two
curmudgeonly businessmen looking out a window into a blizzard. One remarks, “We never had
winters like these past two before Carter took office!” Although Fiorina included this cartoon as
a parody of retrospective voting, recent scholars have argued that voters indeed often respond
to events irrelevant to an incumbent’s performance, damaging the incentivizing function of retrospective
voting. Achen & Bartels (2002) famously offered evidence that Woodrow Wilson lost
vote share in coastal New Jersey towns in the 1916 election due to a rash of shark attacks preceding
5
See also Gartner et al. (2004) and Gartner (2008).
www.annualreviews.org • Retrospective Voting Reconsidered 295
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
the election. The extensive analysis in Achen & Bartels’ paper uses data from 1896–2000 to
suggest that voters systematically punish politicians for events such as floods and droughts that are
beyond the incumbent’s control. They find, in particular, that extreme droughts or wet spells in
the election year cost the incumbent presidential party roughly 1.5 percentage points of the vote.
The most important critique of Achen & Bartels’ examples is that although government could
not reasonably be held responsible for aberrant weather or shark attacks, voters could reasonably
expect government to be responsible for preparation, mitigation, and response to natural disasters.
In other words, it may be not the drought itself that matters, but rather the failure to deliver
relief in the aftermath.6
Consistent with this idea, Healy & Malhotra (2010) show that the way
voters respond to tornadoes is more consistent with a standard economic voting story than with
the “shark attacks” logic. Although voters punish politicians for every 1% increase in tornadocaused
monetary damage by about 0.014 percentage points of the vote (which government may be
expected to respond to), voters do not punish politicians for deaths caused by tornadoes, which are
more stochastic and beyond government response. Thus, voters do not appear to simply respond
to tornadoes emotionally; they seem to separate out economic and noneconomic concerns. A more
subtle critique of Achen & Bartels’ (2002) examples is that one could write down a model of a
rational voter who simply responds to any negative event by punishing the incumbent because he
knows that some portion of any negative event may be a signal about the incumbent’s competence.
Because it is costly to seek information that separates signal from noise, it could make sense to
simply punish all politicians for negative events.
Healy et al. (2010) demonstrate that at least some voters do reward and punish incumbent
presidents, governors, and senators for events that are completely beyond their control. They
consider voters’ responses to wins and losses by the local college football team shortly before
elections. Unlike aberrant weather, local sports outcomes are not something that citizens could
expect government to prepare for nor to respond to, so this empirical strategy isolates the effect of
mood on voting. Healy et al. (2010) find that a win by the local college football team in the week
before Election Day increases incumbent vote share by about 1.6 percentage points in Senate,
gubernatorial, and presidential elections. The effect is bigger where teams have more fan support,
where we would expect a greater shock to voters’ moods.
A second demonstration of apparent voter error in retrospective voting is improper benchmarking.
For instance, Leigh (2009) studies whether world leaders benefit from global macroeconomic
swings relative to their own country’s economic performance. He finds that voters do a poor job of
distinguishing signal from noise; they are more sensitive to the global economy than to their own
country’s economic performance net of global swings. Wolfers (2009) finds similar evidence from
US gubernatorial elections. Although voters are able to distinguish between state and national
economic performance, voters in oil-producing states base voting decisions on the international
price of oil, which should be beyond any governor’s control.
Nonetheless, analyses using different datasets have suggested that voters sometimes benchmark
more appropriately. Ebeid & Rodden (2006) use data from US states to show that in
states highly dependent on natural resources and farming, there is little relationship between
macroeconomic indicators and gubernatorial reelection. Such behavior is sensible because the
6
Mo Fiorina reminded us of the scene in The Godfather where Don Corleone tells the other Mafia bosses, “If some unlucky
accident should befall my youngest son [Michael], if some police officer should accidentally shoot him, or if he should hang
himself in his cell, or if my son is struck by a bolt of lightning, then I will blame some of the people here” (Puzo & Coppola
1972). By holding the bosses responsible for any event, even those outside their control, Corleone incentivized them to
especially look out for Michael’s safety. Similarly, voter responses to any negative event may reflect an attempt to ensure
maximum effort.
296 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
economic performance of these states is more subject to factors outside the governor’s control.
In contrast, economic voting is stronger in states less reliant on natural resources and agriculture.
Kayser & Peress (2012) examine multiple countries and, in contrast to Leigh, find that voters do
an excellent job benchmarking domestic economic performance to global swings, with the effect
of benchmarked growth exceeding that of aggregate national growth by a factor of two.7
Finally, recent research has argued that voters do not take into account the full extent of an
incumbent’s economic record. Building on earlier work by Kramer (1971), Tufte (1978), and Fair
(1978), Achen & Bartels (2004) find that voters overweight the last year of economic growth relative
to the previous three in a presidential term. A one-percentage-point increase in real disposable
income growth in the election year increases incumbent party vote share by three percentage
points, whereas income growth early in the term has no effect. The authors claim that such voter
behavior makes presidential elections a game of “musical chairs” in which politicians are selected
largely based on luck, since elected officials cannot precisely time the economy to perform well
in a given year. Further, Bartels (2008) argues that voter myopia has substantial implications for
election outcomes. Because Republicans have tended to achieve slightly higher income growth in
election years than Democrats after substantially weaker income growth earlier in their terms (see
also Alesina et al. 1993), voter myopia has helped lead to the reelection of Republican presidents
who achieve lower overall growth. Bartels (2008) thus argues that voter myopia has led to the
election of presidents who manage the economy less well.
Recent experimental work has attempted to unpack the mechanisms underlying voters’ overemphasis
on recent events. Hill et al. (2012) use an experiment to strip political institutions out of the
equation, examining whether voters decide to retain an “allocator” based on a stream of random
payments they receive. They find that voters overweight recent payments from the allocator and
thus make voting decisions that fail to maximize their welfare. The conclusion is that recent events,
even in an experiment that takes place over a matter of minutes, exert undue influence on voting
decisions.8
Controlling for performance overall, an incumbent who does better than average in
the last few quarters is 6.5% more likely to win reelection. Healy & Lenz (2012) conduct a series
of survey experiments to provide evidence on the reasons why voters put so much weight on the
election-year economy. They find that people intend to evaluate incumbents based on their overall
performance over an entire term but rely on election-year income growth because it is a more
easily available metric than the information on overall performance that they are searching for.
Retrospective Voting and Policy
Although the majority of the literature on retrospective voting studies how voters respond to
outcomes, some recent work considers the relationship between retrospective voting and policy.
Some of these papers consider voters’ responses to policy decisions, others look at how policy
responds to the implied incentives provided by voters, and some research considers both of these
7
There are several possible reasons why Leigh (2009) and Kayser & Peress (2012) obtain different results. Leigh’s sample
includes 58 countries; Kayser & Peress examine 22 OECD states, mainly European. Perhaps European voters in wealthy
nations are especially politically sophisticated owing to the complexity of coalitional government. Leigh’s time series covers
only 1978–1999, whereas Kayser & Peress examine 1948–2012. The studies also differ with respect to data analysis; Leigh
excludes both the United States and Japan because of their large economies, whereas Kayser & Peress exclude only the United
States. Leigh’s dependent variable is whether the leader is reelected; Kayser & Peress predict the vote share of the leader
party. Therefore, it may be that Kayser & Peress’ results are driven by uncompetitive elections.
8
They also replicate Healy et al.’s (2010) finding by showing that exogenous shocks to experimental payments in the form
of lotteries—even when respondents are explicitly told that the allocator had no impact on the lottery—predict retention
choices, suggesting that emotional responses unrelated to performance influence evaluations.
www.annualreviews.org • Retrospective Voting Reconsidered 297
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
aspects together. The empirical results can often only be suggestive, but some important insights
emerge from these analyses.
A series of papers has considered the effect that federal spending in the aggregate has had on
election outcomes, particularly for the US House of Representatives. Most of these papers have
found that spending has little effect in House elections, although sometimes spending influences
certain subgroups of incumbents and voters (e.g., Alvarez & Schousen 1993, Bickers & Stein 1996,
Alvarez & Saving 1997, Levitt & Snyder 1997, Lazarus & Reilly 2010). Kriner & Reeves (2012)
look instead at the effect of spending on the president’s vote share and find clear positive effects.
Specifically, they find that an 80% increase in federal grant spending increases the incumbent
presidential party vote share by more than 0.5%. They also find that this effect appears to depend on
the ideology of individual voters, with more liberal voters being more likely to reward presidents for
spending. The variation in voter responses indicates the importance of accounting for context when
describing the retrospective voter, in keeping with the conclusions of Duch & Stevenson (2008).
Other research has looked at particular kinds of spending to provide additional insights into the
relationship between retrospective voting and policy. In a particularly thorough example, Chen
(2012) combines detailed data on FEMA disaster spending and wind intensity during the 2004
Florida hurricane season with individual-level voter files. He finds that disaster relief spending
mobilizes incumbent party turnout but demobilizes the opposing party. Specifically, Governor
Jeb Bush’s targeting of hurricane relief aid increased Republican turnout by 5.1% and decreased
Democratic turnout by 3.1%. The logic is that members of the opposing party are less motivated
to show up to vote against the incumbent because relief payments mitigate their ideological
disagreement. The results have important implications for policy-maker incentives as incumbent
administrations may choose to direct resources strategically in advance of elections. In fact,
policy makers often appear to respond to those incentives. For example, in election years, governments
provide more agricultural credit (Cole 2009), pay back wages owed to government workers
(Akhmedov & Zhuravskaya 2004), hire more police officers (Levitt 1997), and provide more relief
in response to natural disasters (Cole et al. 2012). Much like there are political business cycles
(Nordhaus 1975), there may be political relief and policy cycles.
Other scholars have also utilized government response to natural disasters to identify the relationship
between policy action and voter response (e.g., Bechtel & Hainmueller 2011). Our paper
(Healy & Malhotra 2009) studies retrospective voting and public policy in a unified framework;
we combine data on voting behavior (election results), elite actions (disaster preparedness and
relief spending), and public welfare outcomes (disaster damage). We find that voters reward the
incumbent presidential party for disaster relief spending (a 10% increase in relief spending increases
incumbent vote share by 0.045 percentage points) but not for disaster preparedness efforts.
As our paper describes, this may be due to the fact that preparedness spending is not very salient
compared to relief payments (see also Ashworth 2012). We also find that politicians appear to
devote insufficient resources to disaster preparedness. On average, we estimate that a dollar of
disaster preparedness spending was worth between $3 and $27 in terms of reduced future damage.
Even though our results are consistent with voters and elites acting rationally at the individual
level, the results suggest that the collective failure to adequately provide disaster preparedness
leads to welfare losses of about $4.1 billion per year.
The intersection between natural disasters and elections recently received a great deal of attention
due to Hurricane Sandy, which hit the mid-Atlantic in October 2012. President Obama’s wellreceived
response, and New Jersey Governor Chris Christie’s endorsement of that response, may
have helped him win a closely decided election. The storm and its aftermath dominated the news for
days and gave Obama the opportunity to appear as a strong leader in a time of crisis. The political
effects of Sandy can be contrasted with the perceived poor federal response to Hurricane Andrew
298 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
in 1992 and Hurricane Katrina in 2005, which damaged the standing of incumbent presidents.
Elected officials appear not to face such clear consequences for failing to build needed infrastructure
that might make power delivery more resilient in a hurricane or that might mitigate the climate
changethatmakesstormsmoresevere.Whateverthereasonwhyelectionsdonotturnonissueslike
disaster preparedness, the important point is that voters’ actions help to explain policy outcomes.
OPEN QUESTIONS
Can Responsiveness Decrease Welfare?
Besides the few studies listed above, we uncovered almost no literature that explores how retrospective
voting actually influences policy outcomes. Identifying this link in the retrospective voting
chain is empirically difficult but also crucially important if we are to interpret retrospective voting
as a normatively appealing phenomenon. For example, a natural extension of Berry & Howell’s
(2007) study of school boards (described above) would be to see if voter (un)responsiveness to
test scores improves future educational performance, not only with respect to test scores but also
dropout rates, college attendance, personal income growth, juvenile crime rates, and the like.
Further, to close the feedback loop, future research can explore whether policy changes that may
arise from retrospective voting then affect voting behavior itself.
A theme in many of the papers we have discussed is that voters can be too responsive and place
too much accountability pressure on government. Healy & Malhotra (2009) provide evidence
that responsiveness per se may not always be associated with good normative outcomes and public
welfare gains, as much research has often argued. For example, Gasper & Reeves (2011) conclude
that voters reward politicians for delivering disaster relief and interpret this as an example of
democracy functioning well. Yet, disaster relief spending may not be as efficient as preparedness
spending. Similarly, Berry & Howell (2007) suggest that voters should be rewarding and punishing
school board incumbents based on test score performance. However, school districts can easily
manipulate test scores by shuffling students in and out of districts (Bryk et al. 2010). Additionally,
it is uncertain whether test scores reliably and validly measure student achievement or whether
schools that “teach to the test” are actually providing a better education. Hence, there may be
cases where we would prefer that voters did not respond at all and did not place accountability
pressure on government.
Consistent with these ideas, some recent evidence suggests that making politicians less accountable
to voters can sometimes increase public welfare. Achen & Bartels (2004) show that a
policy they consider unambiguously beneficial—fluoridation of tap water—was undertaken earlier
in cities where mayors had longer terms and were thus more insulated from voters. Likewise, Dal
Bo & Rossi (2011) examined a natural experiment where Argentine legislators were randomly assigned
to different term lengths to show that longer terms enhanced legislator effort. Moving from
two-year terms to four-year terms increased the number of bills passed, a measure of legislative
productivity, by 45%. In some cases, less electoral pressure may lead to improved government
performance.
Of course, if we ask “Is it always good that voters are responsive?” an important follow-up question
is “Compared to what?” In other words, what is the alternative model of voting behavior that
would produce better policy outcomes than a responsive electorate? In the case of the economy,
it is likely better that voters pay attention to performance in some form rather than completely
ignoring macroeconomic indicators. However, it seems likely that public welfare would be enhanced
if voters took a longer view with respect to the economy. If voters only consider macro-
economicindicatorsintheelectionyear,thentheyareliabletogetpoliciesthatprivilegeshort-term
www.annualreviews.org • Retrospective Voting Reconsidered 299
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
growth even at the expense of higher long-term costs. These electoral motivations have altered
pre-election monetary policy in countries such as Mexico, Russia, and the United States (Stein &
Streb 2004, Abrams 2006). Post-election crises or inflation have followed pre-election manipulations.
Evidence from the Nixon tapes indicates that his administration pushed the Federal Reserve
to pursue expansionary monetary policy in 1972 because of a desire to please voters with strong
election-year growth (Abrams 2006). The subsequent overheating of the economy in 1972 appears
to have advanced Nixon’s goal of securing reelection while decreasing overall welfare by contributing
to the inflation and recession that characterized the years immediately following the election.
Will Sorting Make Retrospective Voting Rarer?
As mentioned above, Key (1966) challenged the Michigan School’s depiction of the American
voter (Campbell et al. 1960) by showing that citizens often abandoned their party in order to
reward or punish the incumbent based on performance. However, Key was writing in a period
when parties were less strong and party identification was more loosely associated with vote
choice. Today, voters are ideologically well sorted into Republicans and Democrats and have
clear understandings of where the parties stand on policy issues (Levendusky 2009). Even though
voters may not have polarized (Fiorina et al. 2005), an electorate increasingly sorted along party
lines may be less likely to abandon party allegiances to vote retrospectively (Niemi et al. 2011).
Improving Democratic Accountability
A last suggestion for future research is to inquire how voters can make better retrospective
voting decisions. Ferraz & Finan (2008, 2011) provide models for such research (see also
Chong et al. 2011). The authors studied the effect of government audits in exposing corruption
among Brazilian mayors. Using clever natural experiments, Ferraz & Finan found that when
voters were informed about municipal corruption before the election, they threw out corrupt
mayors. Further, corruption was much lower in cases where mayors could be reelected. These
findings suggest that elections can improve accountability and that institutions that provide voters
with information can assist them in making good retrospective decisions.
Consistent with recent evidence from psychology, Healy & Lenz (2012) illustrate that even
psychological biases can be overcome if voters have the right kind of information. Specifically,
if voters are provided with the attribute—total economic growth over a president’s term—on
which they intend to base their decisions, they are able to evaluate presidents based on their
overall performance rather than relying so heavily on the election year. These results suggest that
economic voting could be improved if journalists presented more complete economic information.
Similarly, Schwarz & Clore (1983) found that people who were in a bad mood because of the
weather evaluated unrelated subjects more harshly. Simply making people think about the reasons
for their bad mood helped them think more carefully about their evaluations. Once aware of the
reason for their mood, people no longer transferred it to their other evaluations. Likewise, Healy
et al. (2010) demonstrate in a survey experiment that the effect of sporting outcomes on incumbent
evaluations can be eliminated if people are made aware of the irrelevant events that are influencing
their moods. If voters are in a bad mood for some reason unrelated to politics, these results suggest
that helping voters to think about the reasons underlying their vote choices could help them to
not transfer their mood to their voting decisions.
Future research can also explore how well the media helps voters induce good government performance
and how the media can do a better job informing the electorate. For example, Snyder &
Stromberg (2010) find that press coverage of Congress affects not only the policies that constituents
300 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
get, but also the amount of effort that legislators put into serving their constituents and the level
of knowledge those constituents have about politics. These findings suggest the media can serve
an important role in ensuring that voters receive relevant information before Election Day.
CONCLUSIONS
A plethora of important research over the decades has uncovered findings about how voters reward
and punish politicians based on their performance. We believe that future research on the topic
can benefit from three main considerations:
1. In many contexts, scholars have found evidence of voter responsiveness. What the literature
has done less well is address the normative question of whether the responsiveness that
we observe incentivizes politicians to maximize social welfare. In other words, the study of
retrospective voting may require a paradigm shift; the new question is not whether voters
are responding to government performance but whether they are responding in the right
manner.
2. In order to assess this, a normative benchmark is necessary. Some studies have implicitly
posited normative benchmarks (e.g., lower crime is better; higher test scores are better) but
have not demonstrated why these are good benchmarks or that voter responsiveness helps
reach them. The normative question of whether retrospective voting produces good social
outcomes is important in part because this was exactly V.O. Key’s response to the Michigan
School.
3. Finally, we see the potential for more research to close the circle connecting retrospective
voting behavior with policy outcomes.
Altogether, scholarship on political behavior is providing a more nuanced, realistic, and
complete picture of how retrospective voting actually takes place. Democratic accountability
often works wonders. Voters who throw out corrupt officials when made aware of their actions
clearly provide the right incentives for future policy makers in addition to showing a capacity
for learning. This behavior accords with Sen’s (1981) famous assertion that famines do not occur
in democracies. However, voters often provide the wrong incentives. In a variety of situations,
the ways in which voters assess government performance may lead to inefficient policies and
poorer-performing politicians.
DISCLOSURE STATEMENT
The authors are not aware of any affiliations, memberships, funding, or financial holdings that
might be perceived as affecting the objectivity of this review.
ACKNOWLEDGMENTS
We thank Ray Duch, Mo Fiorina, Simon Jackman, Gabe Lenz, Randy Stevenson, and participants
in a roundtable on this topic at the 2012 Midwest Political Science Association meeting for helpful
comments.
LITERATURE CITED
Abramowitz AI, Lanoue DJ, Ramesh S. 1988. Economic conditions, causal attributions, and political evaluations
in the 1984 presidential election. J. Polit. 50(4):848–63
www.annualreviews.org • Retrospective Voting Reconsidered 301
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
Abrams BA. 2006. How Richard Nixon pressured Arthur Burns: evidence from the Nixon tapes. J. Econ.
Perspect. 20(4):177–88
Achen CH, Bartels LM. 2002. Blind retrospection: electoral responses to drought, flu, and shark attacks. Presented
at Annu. Meet. Am. Polit. Sci. Assoc. Aug. 29–31, Boston, MA
Achen CH, Bartels LM. 2004. Musical chairs: pocketbook voting and the limits of democratic accountability. Presented
at Annu. Meet. Am. Polit. Sci. Assoc., Sep. 1–5, Chicago, IL
Akhmedov A, Zhuravskaya E. 2004. Opportunistic political cycles: test in a young democracy setting. Q. J.
Econ. 119(4):1301–38
Alesina A, Londregan J, Rosenthal H. 1993. A model of the political economy of the United States. Am. Polit.
Sci. Rev. 87:12–33
Alt J, Bueno de Mesquita E, Rose S. 2011. Disentangling accountability and competence in elections: evidence
from U.S. term limits. J. Polit. 73(1):171–86
Alvarez RM, Saving J. 1997. Deficits, democrats, and distributive benefits: congressional elections and the
pork barrel in the 1980s. Polit. Res. Q. 50(4):809–31
Alvarez RM, Schousen MM. 1993. Policy moderation or conflicting expectations? Testing the intentional
models of split-ticket voting. Am. Polit. Res. 21(4):410–38
Anderson CJ. 2007. The end of economic voting? Contingency dilemmas and the limits of democratic accountability.
Annu. Rev. Polit. Sci. 10:271–96
Angrist JD, Pischke J-S. 2009. Mostly Harmless Econometrics: An Empiricist’s Companion. Princeton, NJ: Princeton
Univ. Press
Angrist JD, Pischke J-S. 2010. The credibility revolution in emperical econometrics: how better research
design is taking the con out of econometrics. J. Econ. Perspect. 24(2):3–30
Ansolabehere S, Meredith M, Snowberg E. 2012. Mecro-economic voting: local information and micro-perceptions
of the macro-economy. Work. pap., Dep. Gov., Harvard Univ.,
Arceneaux K. 2005. Does federalism weaken democratic representation in the United States? Publius J. Fed.
35(2):297–311
Arceneaux K. 2006. The federal face of voting: Are elected officials held accountable for the functions relevant
to their office? Polit. Psychol. 27(5):731–54
Ashworth S. 2005. Reputational dynamics and political careers. J. Law Econ. Organ. 21(2):441–66
Ashworth S. 2012. Electoral accountability: recent theoretical and empirical work. Annu. Rev. Polit. Sci. 15:183–
201
Barro RJ. 1973. The control of politicians: an economic model. Public Choice 14(1):19–42
Bartels LM. 2002. Beyond the running tally: partisan bias in political perceptions. Polit. Behav. 24(2):117–50
Bartels LM. 2008. Unequal Democracy: The Political Economy of the New Gilded Age. Princeton, NJ: Princeton
Univ. Press
Bechtel MM, Hainmueller J. 2011. How lasting is voter gratitude? An analysis of the short- and long-term
electoral returns to beneficial policy. Am. J. Polit. Sci. 55(4):851–67
Berry CR, Howell W. 2007. Accountability and local elections: rethinking retrospective voting. J. Polit.
69(3):844–58
Besley T. 2006. Principled Agents? The Political Economy of Good Government. New York: Oxford Univ. Press
Bickers K, Stein R. 1996. The electoral dynamics of the federal pork barrel. Am. J. Polit. Sci. 40(4):1300–26
Books J, Prysby C. 1999. Contextual effects on retrospective economic evaluations: the impact of the state and
local economy. Polit. Behav. 21(1):1–16
Bower GH. 1981. Mood and memory. Am. Psychol. 36:129–48
BrownAR.2010.Are governors responsible forthe state economy?Partisanship,blame,and divided federalism.
J. Polit. 72(3):605–15
Bryk AS, Sebring PB, Allensworth E, Easton JQ, Luppescu S. 2010. Organizing Schools for Improvement: Lessons
from Chicago. Chicago: Univ. Chicago Press
Bullock J, Gerber AS, Huber G. 2012. Partisan bias in factual beliefs about politics. Work. pap., Dep. Polit. Sci.,
Yale Univ.
Campbell J. 2011. The economic records of the presidents: party differences and inherited economic conditions.
Forum J. Appl. Res. Contemp. Polit. 9(1):Article 7
302 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
Campbell A, Converse PE, Miller WE, Stokes DE. 1960. The American Voter. Chicago: Univ. Chicago Press
Canes-Wrone B, Brady DW, Cogan JF. 2002. Out of step, out of office: electoral accountability and House
members’ voting. Am. Polit. Sci. Rev. 96(1):127–40
Canes-Wrone B, Herron MC, Shotts KW. 2001. Leadership and pandering: a theory of executive policymaking.
Am. J. Polit. Sci. 45:532–50
Carsey TM, Wright GC. 1998. State and national factors in gubernatorial and senatorial elections. Am. J.
Polit. Sci. 42(3):994–1002
Chen J. 2013. Distributive spending, voter partisanship, and political participation: a theory of voter turnout
as retrospective evaluation. Am. J. Polit. Sci. In press
Chong A, De La O A, Karlan D, Wantchekon L. 2011. Looking beyond the incumbent: the effects of exposing
corruption on electoral outcomes. Work. pap., Dep. Polit. Sci., Yale Univ.
Cole S. 2009. Fixing market failures or fixing elections? Agricultural credit in India. Am. Econ. J. Appl. Econ.
1(1):219–50
Cole S, Healy A, Werker E. 2012. Do voters appreciate responsive governments? Evidence from Indian
disaster relief. J. Dev. Econ. 97(2):167–81
Conover PJ, Feldman S, Knight K. 1986. The personal and political underpinnings of economic forecasts.
Am. J. Polit. Sci. 31:559–83
Converse PE. 1964. The nature of belief systems in mass publics. In Ideology and Discontent, ed. D Apter,
pp. 206–61. New York: Free Press
Converse PE. 2006. Democratic theory and electoral reality. Crit. Rev. 18:297–329
Cox GW. 1999. The empirical content of rational choice theory: a reply to Green and Shapiro. J. Theor. Polit.
11(2):147–69
Dal Bo E, Rossi M. 2011. Term length and the effort of politicians. Rev. Econ. Stud. 78(4):1237–63
Downs A. 1957. An Economic Theory of Democracy. New York: Harper & Row
Duch RM, Stevenson RT. 2008. The Economic Vote: How Political and Economic Institutions Condition Election
Results. New York: Cambridge Univ. Press
Ebeid M, Rodden J. 2006. Economic geography and economic voting: evidence from the U.S. states. Br. J.
Polit. Sci. 36:527–47
Evans G, Andersen R. 2006. The political conditioning of economic perceptions. J. Polit. 68:194–207
Evans G, Pickup M. 2010. Reversing the causal arrow: the political conditioning of economic perceptions in
the 2000–2004 U.S. presidential election cycle. J. Polit. 72(4):1236–51
Fair RC. 1978. The effect of economic events on votes for president. Rev. Econ. Stat. 60(2):159–73
Fearon JD. 1999. Electoral accountability and the control of politicians: selecting good types versus sanctioning
poor performance. In Democracy, Accountability, and Representation, ed. A Przeworski, SC Stokes, B Manin,
pp. 55–97. New York: Cambridge Univ. Press
Ferejohn J. 1986. Incumbent performance and electoral control. Public Choice 50(1–3):5–25
Ferraz C, Finan FS. 2008. Exposing corrupt politicians: the effect of Brazil’s publicly released audits on
electoral outcomes. Q. J. Econ. 123(2):703–45
Ferraz C, Finan FS. 2011. Electoral accountabiltiy and corruption in local governments: evidence from audit
reports. Am. Econ. Rev. 101:1274–311
Fiorina MP. 1981. Retrospective Voting in American National Elections. New Haven, CT: Yale Univ. Press
Fiorina MP, Abrams SJ, Pope JC. 2005. Culture War? The Myth of a Polarized America. New York: Longman
Fletcher GJO, Ward C. 1988. Attribution theory and processes: a cross-cultural perspective. In The CrossCultural
Challenge to Social Psychology, ed. MH Bond, pp. 230–44. Beverly Hills, CA: Sage
Gaines BJ, Kulkinski JH, Quirk PJ, Peyton B, Verkuilen J. 2007. Same facts, different interpretations: partisan
motivation and opinion on Iraq. J. Polit. 69(4):957–74
Gartner SS. 2008. The multiple effects of casualties on public support for war: an experimental approach. Am.
Polit. Sci. Rev. 102(1):95–106
Gartner SS, Segura GM, Barratt BA. 2004. War casualties, policy positions, and the fate of legislators. Polit.
Res. Q. 57(3):467–77
Gasper JT, Reeves A. 2011. Make it rain? Retrospection and the attentive electorate in the context of natural
disasters. Am. J. Polit. Sci. 55(2):340–55
www.annualreviews.org • Retrospective Voting Reconsidered 303
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
Gomez BT, Wilson JM. 2001. Political sophistication and economic voting in the American electorate: a
theory of heterogeneous attribution. Am. J. Polit. Sci. 45(4):899–914
Gomez BT, Wilson JM. 2003. Causal attribution and economic voting in American congressional elections.
Polit. Res. Q. 56(3):271–82
Grose CR, Oppenheimer BI. 2007. The Iraq war, partisanship, and candidate attributes: variation in partisan
swing in the 2006 U.S. House elections. Legis. Stud. Q. 32(4):531–57
Healy A, Lenz G. 2012. Substituting the end for the whole: why voters respond primarily to the election-year economy.
Work. pap., Dep. Polit. Sci.,Univ. Calif., Berkeley
Healy A, Malhotra N. 2009. Myopic voters and natural disaster policy. Am. Polit. Sci. Rev. 103(3):387–406
Healy A, Malhotra N. 2010. Random events, economic losses, and retrospective voting: implications for
democratic competence. Q. J. Polit. Sci. 5(2):193–208
Healy AJ, Malhotra N, Mo CH. 2010. Irrelevant events affect voters’ evaluations of government performance.
Proc. Natl. Acad. Sci. USA 107(29):12804–9
Hetherington MJ. 1996. The media’s role in forming voters’ national economic evaluations in 1992. Am. J.
Polit. Sci. 40(2):372–95
Hill SJ, Huber GA, Lenz GS. 2012. Sources of bias in retrospective decision-making: experimental evidence
on voters’ limitations in controlling incumbents. Am. Polit. Sci. Rev. 106:720–41
Kahneman D. 2003. Maps of bounded rationality: psychology for behavioral economics. Am. Econ. Rev.
93(5):1449–75
Kahneman D, Slovic P, Tversky A. 1982. Judgment under Uncertainty: Heuristics and Biases. New York:
Cambridge Univ. Press
Kahneman D, Tversky A. 1979. Prospect theory: an analysis of decision under risk. Econometrica 47(2):263–92
Karol D, Miguel E. 2007. Iraq war casualties and the 2004 U.S. presidential election. J. Polit. 69(3):633–48
Kayser MA, Peress M. 2012. Benchmarking across borders: electoral accountability and the necessity of
comparison. Am. Polit. Sci. Rev. 106(3):661–84
Kedar O. 2005. When moderate voters prefer extreme parties: policy balancing in parliamentary elections.
Am. Polit. Sci. Rev. 99(2):185–99
Keith BE, Magleby DB, Nelson CJ, Orr E, Westlye MC, Wolfinger RE. 1992. The Myth of the Independent
Voter. Berkeley: Univ. Calif. Press
Key VO. 1966. The Responsible Electorate: Rationality in Presidential Voting, 1936–1960. Cambridge, MA:
Harvard Univ. Press
Kinder D, Kiewiet RD. 1981. Sociotropic politics: the American case. Br. J. Polit. Sci. 11(2):129–61
Kinder DR, Mebane WR Jr, Monroe KR. 1983. Politics and economics in everyday life. In The Political Process
and Economic Change, ed. KR Monroe, pp. 141–71. New York: Algora
Kramer GH. 1971. Short-term fluctuations in U.S. voting behavior. Am. Polit. Sci. Rev. 71:131–43
Kramer GH. 1983. The ecological fallacy revisited: aggregate- versus individual-level findings on economics
and elections, and sociotropic voting. Am. Polit. Sci. Rev. 77(1):92–111
Kriner D, Shen F. 2007. Iraq casualties and the 2006 Senate elections. Legis. Stud. Q. 32:507–30
Kriner DL, Reeves A. 2012. The influence of federal spending on presidential elections. Am. Polit. Sci. Rev.
106(2):348–66
Lau RR, Redlawsk DP. 2001. Advantages and disadvantages of cognitive heuristics in political decision making.
Am. J. Polit. Sci. 45(4):951–71
Lazarus J, Reilly S. 2010. The electoral benefits of distributive spending. Polit. Res. Q. 63(1):343–55
Leigh A. 2009. Does the world economy swing national elections? Oxf. Bull. Econ. Stat. 71(2):163–81
Levendusky M. 2009. The Partisan Sort: How Liberals Became Democrats and Conservatives Became Republicans.
Chicago: Univ. Chicago Press
Levitt S. 1997. Using electoral cycles in police hiring to estimate the effect of police on crime. J. Polit. Econ.
87(3):270–90
Levitt S, Snyder JM Jr. 1997. The impact of federal spending on House election outcomes. J. Polit. Econ.
105(1):30–53
Lupia A. 1994. Shortcuts versus encyclopedias: information and voting behavior in California insurance reform
elections. Am. Polit. Sci. Rev. 88(1):63–67
304 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
Lupia A, McCubbins MD. 1998. The Democratic Dilemma: Can Citizens Learn What They Need to Know?
New York: Cambridge Univ. Press
MacKuen MB, Erikson RS, Stimson JA. 1992. Peasants or bankers? The American electorate and the U.S.
economy. Am. Polit. Sci. Rev. 86(3):597–611
Malhotra N, Kuo AG. 2008. Attributing blame: the public’s response to Hurricane Katrina. J. Polit. 70(1):120–
35
Margalit Y. 2011. Costly jobs: trade-related layoffs, government compensation, and voting in U.S. elections.
Am. Polit. Sci. Rev. 105(1):166–88
Markus GB. 1988. The impact of personal and national economic conditions on the presidential vote: a pooled
cross-sectional analysis. Am. J. Polit. Sci. 32(1):137–54
Marsh M, Tilley J. 2010. The attribution of credit and blame to governments and its impact on vote choice.
Br. J. Polit. Sci. 40:115–34
Miller WE, Stokes DE. 1963. Constituency influence in Congress. Am. Polit. Sci. Rev. 57(1):45–56
Mondak JJ, Mutz DC, Huckfeldt R. 1996. Persuasion in context: the multilevel structure of economic evaluations.
In Political Persuasion and Attitude Change, pp. 249–66. Ann Arbor: Univ. Mich. Press
Mueller JE. 1973. War, Presidents, and Public Opinion. New York: Wiley
Niemi RG, Weisberg HF, Kimball DC. 2011. Are economic factors weakening as vote determinants? In
Controversies in Voting Behavior, ed. RG Niemi, HF Weisberg, DC Kimball. Washington, DC: CQ Press
Nordhaus WD. 1975. The political business cycle. Rev. Econ. Stud. 42(2):169–90
Padro-i-Miguel G, Snyder JM. 2006. Legislative effectiveness and legislative life. Legis. Stud. Q. 31(3):347–81
Page BI, Shapiro RY. 1992. The Rational Public: Fifty Years of Trends in Americans’ Policy Preferences. Chicago:
Univ. Chicago Press
Peffley M. 1984. The voter as juror: attributing responsibility for economic conditions. Polit. Behav. 6(3):275–
94
Peffley M, Williams JT. 1985. Attributing presidential responsibility for national economic problems.
Am. Polit. Res. 13(4):393–425
Persson T, Roland G, Tabellini G. 1997. Separation of powers and political accountability. Q. J. Econ.
112:1163–202
Persson T, Tabellini G. 2002. Political Economics. Cambridge, MA: MIT Press
Puzo M, Coppola FF. 1972. The Godfather. Los Angeles: Paramount Pictures
Reeves A, Gimpel JG. 2012. Ecologies of unease: geographic context and national economics evaluations.
Polit. Behav. 34(3):507–34
Rogoff K. 1990. Equilibrium political budget cycles. Am. Econ. Rev. 80(1):21–36
Rudolph TJ. 2003a. Who’s responsible for the economy? The formation and consequences of responsibility
attributions. Am. J. Polit. Sci. 47(4):698–713
Rudolph TJ. 2003b. Institutional context and the assignment of political responsibility. J. Polit. 65(1):190–215
Rudolph TJ. 2006. Triangulating political responsibility: the motivated formation of responsibility judgments.
Polit. Psychol. 27(1):99–122
Schwarz N, Clore GL. 1983. Mood, misattribution, and judgement of well-being: informative and directive
functions of affective states. J. Pers. Soc. Psychol. 45:513–23
Sen A. 1981. Poverty and Famines: An Essay on Entitlement and Deprivation. Oxford, UK: Oxford Univ. Press
Shefrin H. 2002. Beyond Greed and Fear: Understanding Behavioral Finance and the Psychology of Investing. Boston:
Harvard Bus. School Press
Snyder J, Stromberg D. 2010. Press coverage and political accountability. J. Polit. Econ. 118(2):355–408
Stein EH, Streb JM. 2004. Elections and the timing of devaluations. J. Int. Econ. 63:119–45
Stein RM. 1990. Economic voting for governor and U.S. senator: the electoral consequences of federalism.
J. Polit. 52(1):29–53
Taylor DM, Doria JR. 1981. Self-serving and group-serving bias in attribution. J. Soc. Psychol. 113(2):201–11
Taylor DM, Jaggi V. 1974. Ethnocentrism and causal attribution. J. Cross-Cult. Psychol. 5(2):162–71
Tilley J, Hobolt SB. 2011. Is the government to blame? An experimental test of how partisanship shapes
perceptions of performance and responsibility. J. Polit. 73:316–30
Tufano P. 2009. Consumer finance. Annu. Rev. Financ. Econ. 1:227–47
www.annualreviews.org • Retrospective Voting Reconsidered 305
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16CH14-Malhotra ARI 17 April 2013 14:49
Tufte ER. 1978. Political Control of the Economy. Princeton, NJ: Princeton Univ. Press
Tversky A, Kahneman D. 1973. Availability: a heuristic for judging frequency and probability. Cogn. Psychol.
5(1):207–33
Vavreck L. 2009. The Message Matters: The Economy and Presidential Campaigns. Princeton, NJ: Princeton Univ.
Press
Wlezien C, Franklin M, Twiggs D. 1997. Economic perceptions and vote choice: disentangling the
endogeneity. Polit. Behav. 19:7–17
Wolfers J. 2009. Are voters rational? Evidence from gubernatorial elections. Work. pap., The Wharton School,
Univ. Penn.
306 Healy ·Malhotra
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16-frontmatter ARI 8 April 2013 21:25
Annual Review of
Political Science
Volume 16, 2013Contents
The Political Theory License
Michael Walzer p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1
Reconsidering Judicial Preferences
Lee Epstein and Jack Knight p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p11
Social Networks and Political Participation
David E. Campbell p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p33
Why Social Relations Matter for Politics and Successful Societies
Peter A. Hall and Mich`ele Lamont p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p49
Distributive Politics Around the World
Miriam Golden and Brian Min p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p73
Media and Political Polarization
Markus Prior p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 101
Media Bias by the Numbers: Challenges and Opportunities in the
Empirical Study of Partisan News
Tim Groeling p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 129
The Political Economy of the Euro
Paul De Grauwe p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 153
Empowerment of the European Parliament
Simon Hix and Bjørn Høyland p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 171
The Use and Misuse of the “Minorities at Risk” Project
Simon Hug p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 191
Looking Back to See Ahead: Unanticipated Changes in Immigration
from 1986 to the Present and Their Implications for American
Politics Today
Michael Jones-Correa and Els de Graauw p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 209
The Changing Landscape of US Unions in Historical and Theoretical
Perspective
Michael Goldfield and Amy Bromsen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 231
v
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.
PL16-frontmatter ARI 8 April 2013 21:25
The Analytical Foundations of Collective Action Theory: A Survey of
Some Recent Developments
Luis Fernando Medina p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 259
Retrospective Voting Reconsidered
Andrew Healy and Neil Malhotra p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 285
Cooperative Survey Research
Stephen Ansolabehere and Douglas Rivers p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 307
Regime Change Cascades: What We Have Learned from the 1848
Revolutions to the 2011 Arab Uprisings
Henry E. Hale p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 331
Terrorism and Democracy
Erica Chenoweth p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 355
Humanitarian Governance
Michael N. Barnett p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 379
Green Clubs: Collective Action and Voluntary Environmental
Programs
Matthew Potoski and Aseem Prakash p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 399
Climate Change Politics
Thomas Bernauer p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 421
The Politics of Energy
Llewelyn Hughes and Phillip Y. Lipscy p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 449
Indexes
Cumulative Index of Contributing Authors, Volumes 12–16 p p p p p p p p p p p p p p p p p p p p p p p p p p p 471
Cumulative Index of Article Titles, Volumes 12–16 p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 473
Errata
An online log of corrections to Annual Review of Political Science articles may be found
at http://polisci.annualreviews.org/
vi Contents
Annu.Rev.Polit.Sci.2013.16:285-306.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon02/12/24.Forpersonaluseonly.