Incognizance and Perceptual Deviation: Individual and Institutional Sources of Variation in Citizens’ Perceptions of Party Placements on the Left-Right Scale

John Aldrich, Gregory S. Schober, Sandra Ley and Marco Fernandez

Electoral behavior scholars and analysts often view and interpret elections through the lens of the left-right scale. But under what conditions is the left-right scale a meaningful concept for voters?

Building on previous work by Campbell et al. (1960) and Abramson et al. (2012; 2014), we identify two ways that the meaning and utility of the scale may be degraded perhaps sufficiently to become non-meaningful for voters. The first way is if voters lack awareness or understanding of the scale, which we refer to as “incognizance.” The second way involves the deviated perceptions of party positions. As the distance between a voter’s perception of a party and the true party position grows, her/his ability to effectively use the scale diminishes. We refer to this distance as “perceptual deviation.”

Using Comparative Study of Electoral Systems (CSES) data and multilevel models, we estimate the effects of individual-, party-, and institutional-level variables on incognizance and perceptual deviation. We find that all three levels influence whether the scale is meaningful to respondents. We also find that incognizance and perceptual deviation have important consequences for citizens’ thermometer evaluations of political parties. By including incognizance and perceptual deviation in the models, we are able to improve our estimates of the effects of key variables on party thermometer evaluations. Thus, the paper illustrates the importance of studying the variability in information for correct specification of models of judgement and choice that rely on reports of political perceptions, opinions, and beliefs.

Interestingly, our results suggest that the standard deviation of party placement is negatively and significantly associated with party thermometer evaluations. In other words, respondents penalize parties that present vague or unclear positions. This finding runs counter to prior research (Tomz and Van Houweling 2009; Somer-Topcu 2015) and contributes to an important discussion on the limits of using perceptual deviation or vagueness for political gain.

Critical Realignment is Indeed Rare: Why the Common Wisdom about the “Earthquake” Effect of Prop 187 is Counter-Intuitive

Iris Hui and David O. Sears

V.O. Key (1955) proposes the concept of “critical election” where abrupt disruption in partisan alignments occurs. In a critical election, voters are “unusually deeply concerned”. Their electoral involvement is relatively high and the “decisive results of voting reveal a sharp alteration of the pre –existing cleavage within the electorate” (1955, p.4). In any introductory courses to American politics, we learn that critical elections seem to be rare in American history.

In 1994, Proposition 187 in California banned undocumented immigrants from accessing health care, education and a variety of other public services.  In 1996, Proposition 209 rejected state governmental affirmative action programs that used race or sex in selecting candidates for employment, contracts, or public education. In 1998, Proposition 227 proscribed a limited time period in which Limited English Proficiency students could be taught in a language other than English. Proposition 187 in particular is often described as a “tipping point” in both California and national Latino politics, leading to their partisan realignment, and marking a new era in which the Democratic Party dominates the state.

Bowler et al. (2006) described Proposition 187 as an “earthquake,” followed by the “aftershocks” of the other two propositions. They contend that “partisan change among Latinos accumulated across a series of contentious ballot propositions that targeted Latinos” (p. 146). The propositions led to a drastic increase of Latinos’ Democratic identification. And the Democratic gain came primarily through chipping away Republican support. Although the authors did not use the term “realignment,” such large changes in four years would merit that description. Second, Proposition 187 was said to politically mobilize the famous “sleeping giant,” the large but often politically inert potential Latino electorate, by getting Latinos more politically engaged.

If so, that would be a major exception to what is known about electoral change in America. Immigration and race were hardly new issues in California or nationally. In addition, party identification is usually seen as quite stable throughout voters’ lifetimes. It usually requires a massive external shock or radically new information to make people convert their partisanship to another party. These well-established findings might suggest some skepticism about the possibility that these racial propositions had such power that they produced a major partisan realignment of Latinos in California.

We reassess the conclusion that Proposition 187, and the other racial propositions, represented a rare “tipping point” that marked an abrupt realignment of partisanship among Latinos in California. We argue that the existing empirical literature on the effect of the racial propositions on the support of the Republican Party might have been somewhat misled, since the studies were based mainly on modest number of surveys with quite small samples of Latinos. Our findings contradict the conventional wisdom about the role of Proposition 187 described earlier. In the early 1970s, over seventy percent of Latino registered voters identified themselves as Democrats. That strong Democratic Party loyalty would seem largely to be a remnant of the New Deal era. The support for the Democratic Party gradually waned through the 1970s and early 1980s, as earlier research has suggested. Yet previous research somewhat underestimates the Democratic advantage going into the 1990s. Our considerably larger data set shows that the Democrats had a 2 to 1 advantage over Republicans prior to Proposition 187.  Support for the Republicans surged during the Reagan and George H.W. Bush administrations.  Then it had already begun to reverse after the Gulf War in 1991, well before the Proposition 187 campaign kicked off in 1994.

We further provide extensive new data and pursue three different empirical strategies in this paper.  We re-affirm the long-standing conclusion in the literature that critical elections are indeed rare. Latinos in the state historically have had a low propensity to support the Republican Party. The lack of enthusiasm for the Republicans was evident especially among younger, unregistered Latinos. Proposition 187 was a political decision by GOP to secure its conservative base. It did not significantly affect Latino identification with the Republican Party and it also did not reverse the fortunes of the GOP in the state. We find that the claim that the racial propositions led to the demise of a political party in the largest state in the country has been somewhat exaggerated.



The Political Implications of American Concerns about Economic Inequality

Graham Wright

Economists and political scientists have often argued that, in a well-functioning democratic system, an increase in economic inequality should lead the public to become more supportive of government redistribution. On an intuitive level, it seems logical that as inequality grows the proportion of the population who would benefit from it being reduced should also grow. In line with this intuition, recent scholarship exploring American attitudes towards inequality (Bartels, 2008; Hayes, 2014; McCall, 2013; Page & Jacobs, 2009) argues that American concerns about economic inequality lead to increased support for government action in the economy.

However, this research suffers from two methodological problems. The first is the difficulty in actually measuring American attitudes about economic inequality. Political scientists have long known that American responses to individual survey questions about political issues are highly unstable, so questions that appear to be measuring “concern about economic inequality” may be measuring something else entirely. Developing a valid measure of “concern” requires using methods such as factor analysis to extract a latent construct from the responses to a set of related items, but this has not been done in prior research.

The second problem is that past work relies mainly on cross-sectional analyses, which cannot account for the possibility that the “causal arrow” between concerns about inequality and desire for a government response runs in both directions simultaneously. When previous research has found a correlation between concern and support for government action it has interpreted this result as implying that increased concern causes increased support for government action. But it could also be the case that it is Americans’ support for (or opposition to) government intervention that causes them to become more (or less) concerned about inequality.

In this article I address both of these limitations by extracting a latent construct “national concern about economic inequality over time” from a set of marginals of national survey questions using the dyad-ratios algorithm, which was originally developed by James Stimson to construct a national level measure of Americans’ “domestic policy mood.” This analysis shows that American concerns about inequality seem to reflect not only concerns over unequal outcomes (viz. money and wealth) but also concerns about unequal opportunities (viz. a lack of social mobility). However, it also finds that a set of items asked in the GSS, which a number of scholars have used to measure American attitudes about inequality, are poor measures of this construct.

To overcome the limitations of cross-sectional analyses noted above, I use an error correction model to test the hypothesis that an increase in national concern about economic inequality leads to a subsequent increase in national support for government action, as measured by Stimson’s “policy mood.” I find no evidence that an increase in national concern about inequality leads to an increase in support for government intervention in the economy. This null result is robust to a wide variety of alternative model specifications. In fact, once confounding factors are accounted for, there is limited evidence that an increase in concern could lead to reduced support for government intervention. This finding suggests that rhetorical efforts to “problematize” economic inequality without connecting this problem to specific policy solutions (the strategy used by the Occupy movement) may produce unintended consequences. In general the political implications of Americans’ views on economic inequality appear far more complex than previously assumed.

Motivated Responding in Studies of Factual Learning

Kabir Khanna and Gaurav Sood

Observers of contemporary public opinion often lament the seeming inability of the political left and right to agree on even basic facts. Democrats and Republicans, for example, seem to hold different beliefs about a range of facts, from the number of Americans who are unemployed to the existence of global warming to the number of people who voted illegally in the past election.

A prominent explanation for these differences is motivated learning: even when people are given the same information supporting an unambiguous conclusion, they are more likely to learn the correct conclusion when it reflects positively on their core attachments and identities. Motivated learning fuels concerns about citizens’ ability to hold governments accountable and the prospect of democratic deliberation. For instance, how can people on different sides of the aisle engage productively with one another when they see the exact same information and yet walk away with diametrically opposed beliefs about what it says?

In our article in Political Behavior, we reexamine the evidence for motivated learning. In a series of experiments, we presented people with tabular data from a putative study on a social policy, either gun control or raising the minimum wage. Following Kahan et al., we manipulated the congeniality of the result supported by the data (e.g., whether the result supports or undermines the effectiveness of gun control). We find that in some cases respondents are indeed more likely to learn the correct result when it is congenial, or in other words, when the result is consistent with their position on the issue. On the surface, this looks like textbook motivated learning.

But here is the crucial part of our study design: independently of the congeniality manipulation, we offered a random subset of respondents a small financial incentive to accurately report what they had learned. Importantly, we only told respondents about the incentive after they had seen the data and could no longer return to it. When we did so, respondents became significantly more likely to report the correct result when it was uncongenial. The incentive treatment significantly reduces estimates of motivated learning and in some cases, eliminates it entirely. But incentives do not alter responses universally. For instance, incentives made no difference among opponents of gun control, suggesting that they really did learn in a biased manner.

Overall, however, the data suggest that without incentives, some respondents give incorrect but congenial answers even when they have learned the correct result. This sort of behavior is what we mean by motivated responding. We also find that respondents are unbiased in recalling the precise numbers they saw in the table.

Our study builds on recent scholarship by Bullock et al. (2015) and Prior et al. (2015), who each demonstrate that motivated responding occurs in surveys of stored knowledge. We find a similar phenomenon in surveys of what people learn over the course of a study. This line of research has important implications for measuring factual beliefs – incentivizing answers to factual questions is likely to reduce bias in measurement.

We’ll end by noting an important wrinkle in our findings. When we incentivized respondents to faithfully report what they had learned, they became more biased in judging the credibility of the putative study. That is, anti-gun respondents judged a study with a pro-gun result more harshly, and vice versa. So, while our findings suggest that motivated learning is less common than what the literature suggests, there is also a whack-a-mole nature to bias: reducing bias in one place is offset by an increase in bias in another place.

The Impact of Procedural Information Costs on Voting: Evidence from a Natural Experiment in Chile

Alejandro Corvalan and Paulo Cox

How much influence may information about ‘how to vote’ have on election participation? At first glance voting rules do not attract much attention. We use to associate them to dull administrative or menial procedural tasks we care of once every four years, and which are kept hidden in some bureaucratic drawer the rest of the time. This is so however digitalized these procedures are today, and how information about them is easily spread through social networks in today’s information era. More attention is put, naturally — if any these days—, on knowledge about the candidates or their policies or scandals. So why worry?

In this article we show that however dull the matter might be, the consequences on participation are meaningful. After all, a very basic form of information is about the electoral procedure itself, which is something that one cannot bypass if one is willing to vote. In fact, frequently asked questions among citizens are ‘When?’, ‘where?’ and ‘how?’ to vote. Of course, and most importantly, the consequences that this type of information may have, may be larger in democracies with self-initiated registration systems, which have been described as more complicated  and often involving more obscure information.

We measure the effect that the cost of information acquisition about how to participate has on participation. To be clear, then, we do not study the effect that the particular rules have, but rather the effect that the costs of acquiring information about the rules have on voting. We exploit a natural experiment to identify the effect of a specific procedural information cost on the registration, and voting, of young first-time voters. We consider the link between two particular rules: the minimum age eligibility requirement and registration closing date. In many countries, first time voters’ eligibility is related to whether they have 18 years at election day or not, while registration closing date is well before the election. This temporal gap between ‘election day’ and ‘closing date’ (CD) may create uncertainty about when the minimum age eligibility requirement is requested. The key point is that information about age eligibility’s deadline has different effects across individuals. Indeed, this piece of information is only meaningful for someone turning 18 a day after CD, while completely irrelevant for another turning 18 before or at CD. The fact that this effect is changing discontinuously at CD allows identification, because nothing at the administrative level is changing across groups on either side of the discontinuity, at any rate. What does change across groups is that one of them requires more information in order to register than the other. Accordingly, the observation of a sharp fall in registration rate at CD reveals that procedural information costs have an impact on voting.

Hence, indeed, not only the procedure itself, nor information about it, can have an impact on the rate at which citizens participate. Also the costs at which this information is obtain matter. The transparency of the electoral procedure is therefore a key aspect in shaping registration and turnout. We show that beyond the direct costs of participating, information costs on procedural rules have an effect on turnout of young first-time voters, and account for about a 10% decrease of the whole cohort in their concurrency to the ballots, due to information costs that are related to this rule only.

Information about electoral procedures may be crucial in the implementation of electoral reforms as well. A change in the electoral system typically is aimed at reducing the direct costs of voting in order to increase participation. However, the very change of these rules requires voters to acquire additional information about the new process for them to be effective. Indeed, the intended effect of the reform may be partially offset if its informational costs more than compensate the direct costs that the reform intends to decrease. Our results suggest that a key component of any reform is how information about the reform itself is to be delivered to voters.


The Influence of Religious-Political Sophistication on U.S. Public Opinion

Eric R. Schmidt

Political science will always have a place for scholars of religion and American politics. But there is perennial hand-wringing over the virtue of large-n survey instruments involving religion. And there is an urgent need for novel survey instruments that offer a substantial empirical payoff, and can be streamlined into the common content of election surveys.

I wrote my recent Political Behavior article partially in response to conversations with practitioners of some of our standard election surveys. Some feel that existing survey instruments on religion 1) waste valuable survey space, and 2) contribute very little to substantive understanding of religion and American politics. My position: religion-and-politics scholars need better (not fewer) instruments.

Traditional survey instruments assess religiosity in terms of identification, belief, and participation. But in my article, I suggest these measures are inadequate proxies for creedal understanding. Indeed, it is possible that “belonging, believing, and behaving” have few independent effects on public opinion – precisely because they imply little about whether respondents understand the political implications of their faith traditions. We can do better, and we do not need to abandon the sample survey in the process.

I introduce a new measure, religious-political sophistication (RPS), assessing whether respondents know their churches’ positions on abortion and same-sex marriage. (Data was collected for an Indiana University module of the 2014 Cooperative Congressional Election Study). The RPS measure is analogous to survey instruments that ask whether respondents know the political positions of candidates for elective office.

My findings can be condensed into some key observations and conclusions:

  • For evangelical Protestants and Roman Catholics, RPS (mainly in combination with church attendance) dramatically decreases the likelihood of taking socially liberal positions.
  • Religious participation seems relatively inconsequential for public opinion unless accompanied by substantive understanding of church teaching on political issues.
  • RPS is as (or more) consequential than knowing politically neutral religious information (measured as knowledge of the Bible’s books and figures). This is consistent with a narrative of elite-mass persuasion in the religious context.
  • The relationship between church attendance and RPS offers a religious model of elite-mass persuasion – analogous to the canonical logic by which partisan identifiers adopt political attitudes in response to elite cues.
  • The “culture wars” account of religion and public opinion – in which religion matters primarily because it engages religious identifiers with partisan politics – is overstated. Rather, religion matters because religious denominations take political positions independently from secular partisan politics.

Note that RPS is measured as a simple dummy variable assessing whether respondents provide a generally accurate understanding of their churches’ positions (e.g. Roman Catholics that correctly note that the Catholic Church opposes abortion in all instances). This is not a complicated measure. But even using a fairly small sample, RPS offers considerable empirical and theoretical yield-from-investment.

I do not claim that my results are the last word on this subject. Indeed, I hope that scholars fine-tune my RPS instruments further. But I hope I have demonstrated 1) that the sample survey can still be used to make theoretical advances in our understanding of religion and public opinion, and 2) that new survey instruments should emphasize whether respondents substantively understand what their faith traditions teach.

I invite scholars to access my publicly available replication data, and to email me at with any questions or comments.


Changes at Political Behavior


I wanted to use this opportunity to publicize a few changes that go into effect for Political Behavior starting January 1, 2017.

First, I wanted to welcome the new members to the Political Behavior Editorial Board.   The continuing members have been remarkably helpful in the operations of the journal and I look forward to working with the new members:

Rosario Aguilar, Centro de Investigación y Docencia Económicas (CIDE)

Bethany Albertson, University of Texas

Lee Ann Banaszak, Penn State University

Timothy Hellwig, Indiana University

Thomas Leeper, London School of Economics

Noam Lupu, Vanderbilt University

Seth McKee, Texas Tech University

Efren Perez, Vanderbilt University

Margaret Roberts, UC San Diego

Lily Tsai, MIT

Second, in continuing efforts to be at the forefront of data transparency, I am adopting the Open Science Framework’s Transparency and Openness Promotion (TOP) standards on data citation. Beginning January 1, 2017 the new data citation policy of Political Behavior is:
“All data sets and program code used in accepted work must be cited in the text and listed in the references. References must include a persistent identifier, such as a Digital Object Identifier. Persistent identifiers ensure future access to unique published digital objects, such as a text or data set. Persistent identifiers are assigned to data sets by digital archives, such as institutional repositories and partners in the Data Preservation Alliance for the Social Sciences (Data-PASS).”

Third, I am adopting a new policy for scholars who serially refuse to review manuscript. I reserve the right to refuse to send out manuscripts for review from authors who refuse to review for Political Behavior. In 2016, less than 44% of review requests resulted in a review.   The scholars who dedicate their time to ensuring that the peer review process functions effectively deserve our praise, but they are subsidizing the efforts of scholars who consistently refuse to review. I hope that I never have to implement this policy, but the reviewer response rate is threatening the practice of peer review and this is one of the few tools that I have available as editor.