Conditional Motivated Reasoning

Abstract

Using motivated reasoning, voters rely on partisanship as a heuristic for evaluating the economy in belief-preserving ways. Yet recent findings show that these motivations may be restricted by a range of contextual factors. We argue that partisan motivations in economic perceptions are moderated by the local economic context. As conditions worsen, a negative information environment leads in-partisans to political ambivalence that reduces confidence in party cues when evaluating the economy. As conditions improve, the motivation for in-partisans to rely on party cues is restored. As positive information has been shown to be less influential for opinion formation than negative information, and out-group members tend to be most prone to motivated reasoning, the economic context should moderate the political motivations of out-partisans to a lesser extent than in-partisans. A multilevel analysis of the 1980 to 2012 American National Election Studies supplemented with state-level data on unemployment and per capita disposable income supports this argument. The effects of in-party attachments on economic perceptions are diminished as economic conditions deteriorate and grow stronger as conditions improve. Moreover, the conditional effects of economic performance on subjective perceptions are stronger for in-partisans than for out-partisans.

Keywords

economic voting motivated reasoning asymmetry state economic conditions

The use of partisanship as a cognitive shortcut for making political judgments has been well-documented by scholars of public opinion and political psychology (Bartels 2000; Campbell et al. 1960; Goren 2005; Rahn 1993; Sniderman 2000; Zaller 1992). Through motivated reasoning, citizens tend to process information in ways that confirm their prior beliefs while rejecting or counter-arguing information that contradicts those beliefs (Kunda 1990; Lodge and Taber 2000; Lord, Ross, and Lepper 1979; Nir 2011; Taber and Lodge 2006). Identification with a partisan group provides a heuristic for processing information in belief-preserving ways with minimum cognitive effort. One area in which motivated reasoning processes have been routinely demonstrated is public opinion toward the economy; citizens tend to view the economy more favorably when they identify with the incumbent political party (Bartels 2002; Enns, Kellstedt, and McAvoy 2012; Evans and Andersen 2006; Evans and Pickup 2010; Gerber and Huber 2010; Wlezien, Franklin, and Twiggs 1997). But under what conditions might voters become more or less sensitive to partisan motivations when asked to evaluate the economy? The extent to which political predispositions shape economic judgment raises important implications for the quality of representative democracy. A staple in the economic voting literature is that democratic accountability hinges on the ability of voters to evaluate the economy independently of their prior political beliefs. If partisan bias in economic perceptions varies across economic contexts, then scholars need to take a closer look at contextual effects on public opinion and exercise caution when making generalizations about the quality of democracy.

A range of mechanisms have been established that restrict the motivation for citizens to process information in belief-preserving ways. First, political ambivalence—the possession of competing political considerations—has been shown to decrease the confidence provided by partisanship when evaluating policy-relevant objects (Basinger and Lavine 2005; Lavine 2001; Lavine, Johnston, and Steenbergen 2012). Second, when exposed consistently to belief-incongruent facts, individuals eventually reach a tipping point at which the motivation to preserve prior beliefs is weakened (Redlawsk, Civettini, and Emmerson 2010). Finally, the strength of motivated reasoning processes varies with the direction of group attachments and the general sentiment of the information environment. Out-group members tend to be more sensitive to the motivation to preserve prior beliefs than in-group members (Ditto et al. 1998; Goren, Federico, and Kittilson 2009), and negative information tends to exert a stronger impact on political thought than positive information (Baumeister et al. 2001; Soroka 2006; Taylor 1991). We extend these findings to argue that the local economic environment can strengthen or weaken the motivation for partisans to view the economy in ways congruent with their prior political beliefs.

A recent line of research has provided evidence that the economic context matters for the relationship between political and economic beliefs. The effect of political attitudes on perceptions of the economy tends to be weakest during times characterized by national economic crisis (Chzhen, Evans, and Pickup 2014; Dickerson 2016a), and the economic beliefs of partisans tend to diverge as conditions worsen (Parker-Stephen 2013; Stanig 2013). Yet most of these findings either rely strictly on national economic data or do not model objective economic conditions at all. To better understand how the economic context matters for partisan motivations in economic perceptions, the health of the economy needs to be more directly modeled. Moreover, national economic performance represents the aggregation of widely varying local economic environments. As national economic conditions do not vary across individuals, they contribute little to explaining variation in perceptions of the national economy at a single point in time. If objective performance truly matters for how citizens view the economy, then the influence of partisanship on economic perceptions should vary to some degree across more localized economic contexts.

We argue that as state-level economic conditions worsen, identification with the incumbent presidential party will have less of an impact on economic perceptions. As in-partisans are consistently exposed to a negative economic environment, they lose confidence in their partisanship as a cue for forming economic opinions. As conditions improve, in-partisans should become more strongly motivated to evaluate the economy in terms of their partisan identities. However, out-group members tend to be more strongly influenced by the motivation to preserve prior beliefs than in-group members (e.g., Ditto et al. 1998; Goren, Federico, and Kittilson 2009). Thus, changes in the economic environment should matter less for how out-partisans perceive the national economy than for in-partisans. Relying on survey data from the 1980 to 2012 American National Election Studies (ANES) supplemented with state-level unemployment and income data, we demonstrate that the influence of in-party attachments on economic perceptions is weakest in states with rising unemployment. These findings are asymmetrical for in-partisans and out-partisans, suggesting that out-partisans respond less strongly to objective changes in the economic environment than in-partisans.

Partisan Motivations in Economic Perceptions

The influence of partisanship for the formation of political judgments can be traced back to the early work of Campbell et al. (1960).¹ There remains a strong consensus that partisanship is a highly influential force for how individuals perceive the world and form political opinions (Bartels 2000, 2002; Gerber, Huber, and Washington 2010; Groenendyk 2013; Hetherington 2001; Richard Johnston 2006; Lewis-Beck et al. 2008; Rahn 1993; Tilley and Hobolt 2011). Whether grounded in psychological attachments formed early in life (Campbell et al. 1960; Lewis-Beck et al. 2008) or identities with specific social groups (Berelson, Lazarsfeld, and McPhee 1954; Miller and Wlezien 1993), the influence of partisanship on the processing of political information is one of the best-documented findings in political science.

The mechanisms through which partisanship shapes political judgment can be found in the psychological literature on motivated reasoning. Citizens tend to process information in ways that uphold their prior beliefs, accepting belief-congruent facts while rejecting or counter-arguing belief-incongruent facts (Kunda 1990; Lodge and Taber 2000; Lord, Ross, and Lepper 1979; Nir 2011; Rousseau and Snehal 1999). This motivation is also reflected in the tendency for individuals to process more quickly and respond more strongly to information that confirms prior beliefs, while taking longer to process contradictory information (Lodge and Taber 2005; Morris et al. 2003; Redlawsk 2002). In their review of the psychological literature on motivated reasoning, Lavine, Johnston, and Steenbergen (2012, 27–29) describe three goals that guide human reasoning and cognition: efficiency, accuracy, and belief perseverance. First, people want to be efficient when processing information and making decisions—that is, they wish to preserve as many cognitive resources as possible. Second, while expending as little cognitive effort as possible, individuals still strive for accuracy when processing information and making decisions. Third, citizens have a subconscious desire to preserve their own prior beliefs that often overshadows accuracy goals. To do so, citizens often tend to form appealing conclusions before accurately processing factual evidence. The trade-off between efficiency, accuracy, and belief perseverance lies at the core of the social and psychological models of partisanship. Partisanship is one of the most stable political identities and an easily accessible cognitive shortcut for evaluating political information. As a result, partisan motivations to preserve prior beliefs often overshadow accuracy-driven goals to reach correct conclusions (Groenendyk 2013; Jacobson 2010; Jerit and Barabas 2012; Kim, Taber, and Lodge 2010; Lebo and Cassino 2007; Nir 2011; Rahn 1993; Taber, Cann, and Kucsova 2009; Taber and Lodge 2006; Westen et al. 2006).²

One of the clearest demonstrations of partisanship’s influence on political judgment is public opinion toward the economy. Despite normative arguments for an economic voter who evaluates economic performance and subsequently updates political preferences (Fiorina 1981; Key 1966; Lewis-Beck 2006; Weinschenk 2010), a growing consensus shows that members of the incumbent presidential party routinely view the economy more favorably than those opposed to the incumbent party (Bartels 2002; Evans and Andersen 2006; Evans and Pickup 2010; Gerber and Huber 2010; Wlezien, Franklin, and Twiggs 1997). Citizens tend to be polarized along partisan lines in how they interpret the same economic facts, as well as how they attribute responsibility for economic performance (Bisgaard 2015; Rudolph 2003, 2006; Tilley and Hobolt 2011). Toward the end of the Reagan administration, despite improvements in both unemployment and inflation, Democratic identifiers overwhelmingly believed that both conditions had worsened over the past two presidential terms while Republican identifiers indicated that conditions had improved (Bartels 2002). Other evidence suggests that partisanship, presidential approval, and vote choices exert stronger effects on economic perceptions than economic perceptions exert on political attitudes (Evans and Pickup 2010). Immediately following the success of the Democratic Party in the 2006 midterm elections, Republican and Democratic identifiers were shown to abruptly change their perceptions of the economy despite very little objective change in economic conditions (Gerber and Huber 2010). On an aggregate level, findings have demonstrated that consumer confidence is strongly shaped by factors such as presidential approval, party identification, and media coverage (De Boef and Kellstedt 2004; Lebo and Cassino 2007). This abundance of evidence for partisan biases in economic perceptions casts doubt on the ability of voters to effectively hold elected officials accountable for economic performance (Anderson 2007), an inability which often translates to a poor assessment of the quality of democracy in the United States. Even still, several recent findings suggest that all hope may not be lost for the economic voter.

A growing body of research argues that partisan motivations in how voters view the economy may vary with the economic context. As macroeconomic conditions worsen, partisans have been shown to converge in how they perceive economic performance (Parker-Stephen 2013; Stanig 2013). In-party members become more skeptical of party cues when evaluating the economy during times of economic crisis, while out-party members are already motivated to view the economy unfavorably. The implication is a smaller distance between the economic opinions of in-partisans and out-partisans during times of crisis. Yet even as partisans converge in how they perceive economic realities, other evidence suggests that they simultaneously diverge in how they attribute blame for economic performance (Bisgaard 2015). Other findings on the dynamic relationship between political and economic attitudes show contextual variation in how partisans view the economy. During time periods characterized by national economic recession, effects tend to move more strongly from economic perceptions to government approval than from government approval to economic perceptions (Chzhen, Evans, and Pickup 2014; Dickerson 2016a). Local economic conditions have also been shown to contribute to heterogeneity in national economic perceptions (Books and Prysby 1999; Weatherford 1983). Research on the influence of local contexts in the United Kingdom shows that citizens approving of the incumbent government become less likely to vote accordingly as local area unemployment rises (Ron Johnston et al. 2000). This recent line of research suggests a dynamic middle ground between the traditional model of economic voting and contemporary models of partisan motivated reasoning. More importantly, contextual effects on partisan motivations in economic perceptions suggest that the complexity of assessing democracy has been taken for granted. If the mechanism through which partisans form national economic perceptions varies with local contexts, then so too should the tendency for voters to hold incumbent governments accountable for economic performance.

The State Economy as a Constraint on Motivated Reasoning

A number of recent findings suggest that subjective biases in national economic evaluations may be conditional on objective performance below the national level. First, political ambivalence has been shown to reduce the motivation to make judgments in ways that preserve prior political beliefs (Basinger and Lavine 2005; Lavine 2001; Lavine, Johnston, and Steenbergen 2012). Generally defined as “an individual’s endorsement of competing considerations relevant to evaluating an attitude object” (Lavine 2001, 915), political ambivalence reduces the level of confidence in a decision made based on party cues. Psychologists describe a sufficiency threshold, or a desired level of confidence in a decision or judgment (Lavine, Johnston, and Steenbergen 2012; Payne, Bettman, and Johnson 1993). When partisan motivations lead to a judgment that falls below the sufficiency threshold, the resulting confidence gap leads individuals to seek out other sources of information to increase subjective confidence in the decision.³ Such circumstances might arise when an individual holds conflicting opinions on two of a candidate’s policy stances or when an individual identifies with the governing party but disapproves of its handling of the economy. Second, the motivation to preserve prior beliefs has been shown to decrease as the information environment increasingly exposes an individual to belief-incongruent facts (Redlawsk, Civettini, and Emmerson 2010). At some point, a tipping point is reached at which prior beliefs no longer shape the processing of subsequent information. Third, out-group members have been shown to be more prone to motivated reasoning than in-group members (Ditto et al. 1998; Goren, Federico, and Kittilson 2009). This tendency can be thought of as a sort of cognitive defense mechanism through which out-group members are more strongly motivated to defend their position as “outsiders,” while in-group members are already the “winners” and thus face weaker motivations to defend their current position. Finally, negative information tends to be more influential for opinion formation than positive information (Baumeister et al. 2001; Soroka 2006; Taylor 1991). Thus, the general sentiment of the local economic environment should bear some impact on how partisans form evaluations of the national economy.

If the constraints to motivated reasoning described above are taken together, they paint an interesting picture of how partisans form perceptions of the economy. Consider four partisans in distinct economic environments: an in-partisan in poor conditions, an in-partisan in prosperous conditions, an out-partisan in poor conditions, and an out-partisan in prosperous conditions. An underlying assumption of this study is that the economic environment can foster political ambivalence, at least for in-partisans.⁴ First, an in-partisan surrounded by poor economic conditions experiences conflicting signals. The resulting ambivalence, combined with the asymmetrically stronger influence of negative information over positive information, reduces confidence in partisanship when it comes to evaluating the national economy. For an in-partisan surrounded by more prosperous conditions, these signals fall back into sync with one another and the influence of partisanship on economic perceptions returns. For an out-partisan surrounded by poor conditions, partisan identities and economic signals are congruent with one another and so the effect of partisanship on economic perceptions remains unchanged. Finally, for an out-partisan surrounded by prosperous conditions, which are conflicting signals, the influence of partisanship still remains. This latter tendency should result from the fact that out-group members are more prone to motivated reasoning and the positive information coming from the prosperous local economy is less influential for opinion formation than negative information. The observable implications of these expectations should be an asymmetrical influence of the local economy on how in-partisans and out-partisans come to view the national economy. Local conditions should moderate the influence of partisanship on national economic evaluations for in-partisans, but less so for out-partisans.

These assumptions align well with recent findings on the moderating role of the macroeconomic context for the formation of economic perceptions (e.g., Chzhen, Evans, and Pickup 2014; Dickerson 2016a; Parker-Stephen 2013; Stanig 2013). However, most of these studies either rely strictly on national-level economic data or do not directly model economic conditions at all. Although the health of the national economy undoubtedly receives substantial attention from politicians and pundits alike, national-level indicators of economic performance fail to capture the objective environment in which partisans actually exist. Most typical measures of national economic performance, such as unemployment and income, are little more than abstract aggregations of many local environments (Gimpel and Reeves 2012). For example, an 8 percent national unemployment rate reveals very little about the distinction between states with 4 percent unemployment and states with 15 percent unemployment. This is not to suggest that citizens possess specific knowledge about the health of their state’s economy, but simply that national figures fail to capture true variation in how the economy is actually performing. Given that citizens tend to process the same national economic figures in vastly different ways, it is plausible that voters possess no more objective knowledge of national economic performance than they do for local economic performance. More importantly, voters do not need to possess specific knowledge about their local economic environment to be influenced by those surroundings.

Previous findings suggest that local economic circumstances shape both evaluations of the national economy and vote choices (Ansolabehere, Meredith, and Snowberg 2014; Books and Prysby 1999; Ron Johnston et al. 2000). More generally, political behavior is influenced by both individual characteristics as well as characteristics of the local environment (Przeworski 1974). But defining what constitutes “local” is difficult. Individuals could exist in multiple local economic contexts depending on how local is defined. Too small of a geographic unit will underestimate the influence of subjective variation across individuals, while too large of an area will fail to capture the “element of similar experience” that leads to similarities of opinion (Weatherford 1983, 871). Although the most appropriate unit is open to debate, we use state-level economic conditions to model the local economy. The state-level economy offers a middle ground that allows for contextual effects of economic performance without washing out the effects of individual characteristics. Moreover, the state represents a distinct economic environment that is to some extent driven by state-level economic policy, the effects of which connect the economic conditions of communities within states. At the same time, state-level economic performance represents the aggregation of more localized community experiences. Furthermore, state economic performance is salient enough for us to reasonably expect individuals to be aware of these conditions, while economic information on smaller geographical units is likely much harder to come by. Empirically, measuring performance at the state level provides variation in economic conditions both spatially and temporally. Finally, state-level economic performance has been shown to shape national evaluations more strongly than county-level performance (Ansolabehere, Meredith, and Snowberg 2014; Books and Prysby 1999).⁵ None of this is to suggest that a state-level measure will perfectly identify the “local” economic context, as any measure is bound to miss some variation. In the case at hand, contextual differences within states (i.e., counties or communities) could also influence perceptions of national economic performance. However, we rely on state-level performance for the theoretical and empirical reasons described here and the methodological advantages demonstrated by the following analyses.⁶

Based on the assumptions and findings described above, we propose the following hypotheses. First, we expect the effect of identifying with the incumbent party on economic perceptions to be conditional on state-level economic performance. Attachments to the incumbent party should have a weaker effect on economic perceptions in states with rising unemployment and below-average changes in per capita disposable income (PCDI) and a stronger effect in states with declining unemployment and above-average changes in per capita income. Second, we expect the effect of state-level economic performance on economic perceptions to be conditional on the strength of an individual’s identity with the incumbent party. Changes in economic conditions should have a stronger effect on economic perceptions for individuals who identify with the incumbent party, while this effect should diminish as partisan identities shift toward the out-party.

Data and Methodological Approaches

Data for the following analyses come from the 1980 to 2012 ANES, supplemented with state-level economic data on unemployment and PCDI. The 1980 study is the earliest of the ANES surveys to include items that measure evaluations of the economy. Surveys were then conducted at two-year intervals from 1980 until 2004 and at four-year intervals from 2004 until 2012. Net changes in state-level unemployment rates and PCDI are used to measure economic performance. Unemployment data were obtained from the U.S. Bureau of Labor Statistics (http://www.bls.gov), and PCDI data were obtained from the U.S. Bureau of Economic Analysis (http://www.bea.gov). As each of the ANES surveys was conducted during November of each survey year, net changes in seasonally adjusted state unemployment rates are measured as the change from November of the year prior to each survey year. Changes in PCDI are measured as the percent change from the period prior to each survey.⁷

The dependent variable is retrospective evaluations of the national economy.⁸ In each survey year, respondents were asked whether the national economy had gotten better or worse during the last twelve months, as well as a follow-up question asking how much worse or how much better. Responses were coded as a five-point ordinal scale ranging from “much worse” to “much better.” The primary individual-level explanatory variable is a measure of partisanship based on the traditional seven-point scale used in the ANES surveys.⁹ The traditional measure is a seven-point scale ranging from strong Democrat to strong Republican. We reverse coded this measure for all survey years in which a Democratic president held the White House so that higher values indicate identification with the incumbent presidential party and lower values indicate identification with the out-party. This new measure, in-party strength, was then centered to have pure independents at zero so that positive values indicate in-partisans and negative values indicate out-partisans, with the strength of partisanship increasing with each absolute value. For example, a strong Republican in the 1998 ANES survey is coded as a −3 as the Democratic Party controlled the presidency. For the 2004 survey, strong Republicans are coded as a 3 since the Republican Party controlled the presidency. Centering pure independents at zero facilitates easier interpretation of intercepts and the main effects of interaction components. Other control variables include individual employment status, educational attainment, race, sex, and age. All variable codings and question wordings can be found in the appendix.

To test the hypotheses described above, we estimate a mixed effects model of economic perceptions as a function of partisanship and state-level economic performance, with state-level random effects and intercepts and cross-level interactions between partisanship and each economic indicator. To control for unmeasured factors that vary over time, such as the national economic context, we include survey-year fixed effects with the 2008 survey excluded as the baseline group. In hierarchical form, with other controls excluded for space purposes, the full model can be expressed as follows:

Individual

E c o n_{i j} = β_{o j} + β_{1 j} P a r t y_{i j} + \sum_{t = 1}^{k - 1} T_{t} + e_{i j} .

State

β_{o j} = γ_{00} + γ_{01} U n e m p_{o j} + γ_{02} P C D I_{o j} + v_{o j},

β_{1 j} = γ_{10} + γ_{11} U n e m p_{1 j} + γ_{12} P C D I_{1 j} + v_{1 j} .

In the individual-level equation, Econ_ij indicates the economic perception of an individual i in state j, Party_ij indicates the strength of identification with the incumbent president for individual i in state j, and T_t represents a vector of dummy variables for survey years with 2008 excluded as the baseline survey. The two state-level equations model the intercept and slope coefficient from the individual-level equation as a function of net changes in state unemployment rates and percent changes in PCDI. Thus, the first state-level equation models the intercept β_oj as a function of each economic indicator, which allows the expected economic perception for pure independents (in the baseline year) to vary with state-level economic performance. More substantively interesting, the second state-level equation allows the relationship between partisanship and economic perceptions, β_1j, to vary with state-level economic performance. The hierarchical setup shown here can be rearranged to its mixed effects form by substituting the two state-level equations for β_oj and β_1j in the individual-level equation, so that

\begin{array}{l} E c o n_{i j} = γ_{00} + v_{o j} + v_{i j} + e_{i j} + γ_{01} U n e m p_{j} + \\ γ_{02} P C D I_{j} + γ_{10} P a r t y_{i j} + γ_{11} (U n e m p_{j}) (P a r t y_{i j}) + \\ γ_{12} (P C D I_{j}) (P a r t y_{i j}) + \sum_{t = 1}^{k - 1} T_{t} . \end{array}

Expressed in mixed effects form, the model now includes cross-level interactions between partisanship and each state-level economic indicator. For testing the theoretical expectations of this study, the primary parameters of interest are γ₁₁ and γ₁₂, which represent marginal changes in the effect of partisanship on economic perceptions based on one-point net changes in unemployment and 1 percent changes in PCDI, respectively. In the error component of the mixed effects equation, γ₀₀ represents the grand mean economic perception, v_0j represents the error variance of the intercept across states (in standard deviations), v_1j represents the error variance of the slope coefficient across states (in standard deviations), and e_ij represents the remaining idiosyncratic error term. In the analysis that follows, the parameters from the mixed effects equation are estimated via full maximum likelihood. To facilitate interpretation of the survey fixed effects, education and age are standardized to have a mean of zero and a standard deviation of 1. By doing so, all variables in the model can be interpreted as having a meaningful zero.

Analysis

We begin by examining the individual-level relationship between economic perceptions and partisanship. Not surprisingly, retrospective evaluations of the national economy are strongly correlated with the strength of an individual’s identification with the incumbent President’s party. For the entire pooled survey sample, retrospective evaluations and partisanship have a correlation coefficient of 0.30. The yearly survey averages for retrospective economic perceptions and partisanship have a correlation coefficient of 0.60. To better visualize this relationship, Figure 1 plots the average retrospective economic perception for strong Democratic and Republican identifiers in each survey year from 1980 to 2012. The dashed line in the figure represents respondents identifying as strong Democrats; the solid line represents respondents identifying as strong Republicans. At first glance, Democratic and Republican perceptions of the economy seem to follow one another fairly closely. However, a closer look reveals an interesting trend. Shortly following each election year in which the party of the President switched from one party to the other (indicated by vertical dashed lines), the average economic perceptions for Democratic and Republican identifiers cross paths. Under Republican presidents, Republican identifiers view the economy more favorably and under Democratic presidents, Democratic identifiers view the economy more favorably. This trend offers support for the notion that predisposed political differences shape how citizens view economic performance (Bartels 2002; Evans and Andersen 2006; Evans and Pickup 2010; Gerber and Huber 2010). More substantively interesting, and the purpose of this study, is an investigation into how the trend shown in Figure 1 varies across citizens in different economic environments.

Figure 1.

Economic perceptions by partisanship, 1980 to 2012.

Three requirements need to be met for a certain level of economic performance to serve as an appropriate proxy for the economic environment in which partisans’ judgments are made. First, enough variation in economic performance needs to exist across states to conduct a multilevel analysis. Second, state-level performance needs to follow national performance closely enough to justify its use as a proxy for how citizens perceive the national economy. And third, if these first two requirements are met, then aggregate subjective evaluations of the national economy should follow changes in state-level performance relatively closely. The following three figures illustrate the extent to which state-level performance meets all three of these requirements.

Figure 2 plots each state’s observed net changes in unemployment and PCDI during each survey year, along with the yearly means of each economic indicator. In the top panel, observed twelve-month net changes in state unemployment rates are plotted with the yearly average trend across all states per year. For example in 1986, the average net change in unemployment across all fifty states and the District of Columbia was just below zero, indicating that, nationally, unemployment did not change much between November 1985 and November 1986. During that same year, net changes in unemployment varied wildly across states, with some states seeing almost 2.5 percent increases in unemployment and other states seeing almost 2 percent decreases in unemployment. The same can be said for percent changes in PCDI from one period to the next, shown in the bottom panel Figure 2. In 2008, for example, the average change in PCDI across all states was just under 5 percent, yet some states saw no change at all, while other states saw increases of up to 13 percent. Figure 2 demonstrates significant variation in unemployment and PCDI across states at a given point in time and across survey years. This is methodologically appealing because it provides the variation needed for a multilevel analysis of economic perceptions, partisanship, and objective performance, while also providing a stronger grounding for the notion that local economic performance might serve as a proxy for national-level performance.

Figure 2.

Net changes in state-level unemployment and per capita disposable income, 1980 to 2012.

Figure 3 lends further support for the use of state-level economic performance as a proxy for the economic environment in which partisans exist. Although the previous figure demonstrated significant variation in economic conditions across states, Figure 3 plots the average twelve-month net changes in state unemployment rates alongside seasonally adjusted twelve-month net changes in the national unemployment rate. The two measures follow each other almost in unison. This should not be surprising, however, as most measures of national economic performance are little more than abstract aggregations of local economic performance.¹⁰ Figure 4 plots the average retrospective evaluation of the national economy of all ANES respondents from 1980 through 2012 alongside average net changes in state unemployment rates during each survey year. As expected, aggregate perceptions of the national economy tend to become more favorable as state-level unemployment rates decline and more unfavorable as unemployment rates rise. Taken together Figures 2 through 4 show that state-level economic performance provides the necessary variation for a multilevel analysis, follows national performance closely enough to serve as a proxy for how voters view the national economy, and also follows aggregate perceptions of the national economy relatively closely.

Figure 3.

Net changes in state- and national-level unemployment rates, 1980 to 2012.

Figure 4.

Retrospective evaluations of the national economy and twelve-month net changes in state unemployment, 1980 to 2012.

Table 1 shows the results of a mixed effects model of economic perceptions as a function of identification with the incumbent presidential party, state-level economic performance, survey-year fixed effects, and the other exogenous controls described in the previous section. Looking first at the intercept and year fixed effects, the results show that perceptions of the economy were worse in 2008 than any other year included in this analysis. Since the 2008 survey is excluded as the baseline group, the coefficient for the intercept indicates that an individual with no partisan attachments in a state with no net change in economic performance had an expected economic perception of about 1.68 on the five-point scale of retrospective evaluations. The random effects indicate that this expected economic perception had a standard deviation across states of about 0.05.

Table 1.

Economic Perceptions as a Function of Partisanship, Unemployment, and Per Capita Disposable Income.

Fixed effects	Coefficient	Robust SE
In-party strength	0.156^a	0.008
ΔUnemployment	−0.054^a	0.01
ΔPCDI	0.016^a	0.005
In-party strength × Δ Unemployment	−0.031^a	0.003
In-party strength × ΔPCDI	−0.003^a	0.001
Employed	−0.175***	0.027
Education	0.066***	0.008
Nonwhite	0.018	0.015
Female	−0.152***	0.012
Age	−0.009	0.007
1980	0.105*	0.049
1982	0.555***	0.043
1984	1.457***	0.051
1986	1.208***	0.049
1988	1.125***	0.048
1990	0.439***	0.052
1992	0.386***	0.044
1994	1.335***	0.051
1996	1.536***	0.044
1998	1.705***	0.05
2000	1.498***	0.048
2002	0.349***	0.039
2004	0.930***	0.041
2012	1.076***	0.047
Intercept	1.679***	0.043
Random effects	σ	Robust SE
In-party strength	0.015***	0.006
Intercept	0.054***	0.017
Residuals	0.931***	0.005
Observation	28,117
Groups	51

Entries indicate unrestricted maximum likelihood estimates with state-level random effects. PCDI = per capita disposable income.

Conditional relationship—see marginal effects.

p < .05. **p < .01. ***p < .001.

More substantively interesting are the conditional effects of partisanship and state-level economic performance on subjective economic perceptions. The coefficient for partisanship indicates that for an individual in a state with no change in economic performance and all other factors held constant, a one-unit shift toward identifying with the incumbent party was associated with about a 0.16-point more favorable economic perception. The coefficients for interaction terms, their respective components, and statistical significance should always be interpreted with caution as they reflect conditional relationships. A more useful way of evaluating interactive relationships is to estimate the marginal effects of each component explanatory variable on the outcome variable across some specified range of another component explanatory variable.¹¹

These marginal effects are plotted in Figures 5 and 6. In Figure 5, the marginal effect of the strength of identification with the incumbent party on economic perceptions is plotted on the Y-axis across the observed range of net changes in state unemployment rates (left panel) and percent changes in PCDI (right panel). In the left panel of Figure 5, it is clear that the effect of identifying with the incumbent party is substantially weaker in states with rising unemployment. For a state with either no net change or small decreases in unemployment, in-partisan strength had a marginal effect of about 0.20 on economic perceptions. Thus, in states with either improving economic conditions or no net change in economic conditions, individuals became more likely to view the economy favorably as their attachments to the president’s party grew stronger. For states with rising unemployment, this effect gradually drops off. For states with 4 percent increases or higher in unemployment, the effect of in-partisan strength on economic perceptions becomes statistically indistinguishable from zero. This trend supports the hypothesis that the motivation for in-partisans to view the economy in belief-preserving ways is reduced as the local economic environment deteriorates, at least in regard to changes in unemployment rates. The effects shown in the right panel of Figure 5 are not as strong, showing very little change in the effect of partisanship on economic perceptions across the range of percent changes in PCDI. These null findings for PCDI are intriguing, a point discussed in greater detail below.

Figure 5.

Effects of partisanship on economic perceptions conditional on state-level economic performance.

Figure 6.

Effects of state-level economic performance on economic perceptions conditional on partisanship.

We also hypothesized that economic perceptions should be shaped by objective performance more strongly for in-partisans than for out-partisans due to asymmetries in the motivated reasoning of partisans and the influences of positive and negative information. Figure 6 plots the marginal effects of net changes in state unemployment (left panel) and percent changes in PCDI (right panel) on economic perceptions across the full range of identification with (or against) the incumbent presidential party. Again, net changes in state unemployment rates play a stronger role in the relationship between partisanship and economic perceptions than percent changes in PCDI. The left panel of Figure 6 shows that for out-partisans, changes in state unemployment rates had no statistically significant effect on economic perceptions. For pure independents, rising unemployment had a marginal effect of close to −0.10, indicating that growing unemployment leads individuals with no partisan attachments to view the economy less favorably than those in states with declining unemployment. As the strength of identification with the incumbent party increases, so too did the size of the effect of unemployment on economic perceptions. For respondents identifying strongly with the party of the incumbent president, rising unemployment has a marginal effect of almost −0.20 on economic perceptions. The effect of percent changes in income levels does not significantly change across the range of party identification, as illustrated by the right panel of Figure 6.

Overall, these findings present support for a relationship between partisanship and economic perceptions that is conditional on the local economic environment, at least when measured in terms of changes in unemployment. The motivation for in-partisans to view the economy in ways that preserve their prior beliefs significantly changes as state-level unemployment rates rise. Moreover, the impact of rising unemployment on how citizens view the economy is stronger for in-partisans and almost nonexistent for out-partisans, reflecting an asymmetry in how partisans respond to changes in the state economy. The moderating effect of the state economy was much weaker when measured as percent changes in income levels. There are both theoretical and methodological explanations for the null findings for PCDI. Theoretically, an abundance of research shows stronger evidence for sociotropic economic voting than pocketbook voting (Books and Prysby 1999; Kinder and Kiewiet 1979, 1981). Only the most sophisticated citizens tend to evaluate the economy based on their own pocketbooks (Gomez and Wilson 2001), which PCDI more closely measures. Given the general lack of sophistication among the American electorate (Delli Carpini and Keeter 1996), it is not entirely surprising that voters respond less strongly to local changes in income than changes in unemployment. Unemployment represents a more sociotropic measure of economic performance than disposable income. Methodologically, income is less useful for capturing variation across local economic contexts as it is strongly space and time trended. Income levels vary across contexts with the cost of living and over time with inflation. Thus, the null findings for state income on how partisans form economic perceptions should not be taken as an invalidation of the influence of the local economic context, but simply as evidence that unemployment provides more useful information about the local economic environment.

Conclusion

Building on previous research that focuses on constraints to political motivated reasoning, this study has posited that the motivation for partisans to view the economy in belief-preserving ways is conditional on local economic performance and the direction of an individual’s partisanship. Specifically, the preceding analysis has shown that in-partisans become less likely to view the economy favorably as state-level unemployment rises. In such contexts, members of the incumbent party are consistently exposed to negative economic information that conflicts with political identities. These conflicting signals, combined with the strong influence of negative information stemming from the poor economic environment reduce confidence in in-party cues when evaluating the national economy. For out-partisans, signals may conflict when the local economy prospers. Yet, given out-group members’ stronger tendency toward motivated reasoning and the weaker influence of positive information stemming from a prosperous local environment, out-partisans respond less strongly to the local economy than in-partisans. The null findings for a moderating role of changes in PCDI suggest that intervening contextual effects of the local economy are more sociotropic in nature than egocentric. This discrepancy also suggests that changes in unemployment represent a more suitable measure of the local economy than changes in disposable income.

For decades, scholars of economic voting have made bold claims regarding the quality of representative democracy in the United States. On one hand, voters might evaluate past economic performance or anticipate future economic events, subsequently updating their political attitudes accordingly. In this case, democracy “performs well” in the sense that voters are able and willing to hold elected officials accountable for their handling of the economy. On the other hand, evaluations of the economy might be biased by predisposed political beliefs—voters identifying with the incumbent party may be more likely to view the economy favorably than those opposed to the incumbent party, regardless of objective economic performance. In this case, democratic accountability is undermined by the lack of an ability or willingness of voters to evaluate the economy independently of their political attitudes. The findings presented here suggest that neither argument is all-encompassingly correct. Past research has encouraged scholars to take into account of the context in which the individual exists (Books and Prysby 1999; Przeworski 1974; Weatherford 1983). Only recently have scholars begun to empirically examine the contextual effects of economic performance for public opinion (Ansolabehere, Meredith, and Snowberg 2014; Chzhen, Evans, and Pickup 2014; Dickerson 2016a; Ron Johnston et al. 2000). The findings presented in this study should raise the field’s awareness of how factors outside of the individual voter can influence the formation of political and economic opinions. Before making any general claims about the quality of democracy and electoral accountability in the United States, it is essential that researchers pay closer attention to the social, political, and economic contexts in which citizens exist. Such surroundings undoubtedly have a profound impact on political thoughts and behaviors.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

Data and replication materials are available at

Notes

References

Anderson

Christopher J.

2007. “The End of Economic Voting? Contingency Dilemmas and the Limits of Democratic Accountability.” Annual Review of Political Science 10:271–96.

Ansolabehere

Stephen

Meredith

Marc

Snowberg

Erik

. 2014. “Macro-economic Voting: Local Information and Micro-perceptions of the Macro-economy.” Economics & Politics 26 (3): 380–410.

Bartels

Larry M.

1996. “Uninformed Votes: Information Effects in Presidential Elections.” American Journal of Political Science 40 (1): 194–230.

Bartels

Larry M.

2000. “Partisanship and Voting Behavior, 1952-1996.” American Journal of Political Science 44 (1): 35–50.

Bartels

Larry M.

2002. “Beyond the Running Tally: Partisan Biases in Political Perceptions.” Political Behavior 24 (2): 117–50.

Basinger

Scott J.

Lavine

Howard

. 2005. “Ambivalence, Information, and Electoral Choice.” American Political Science Review 99 (2): 169–84.

Baumeister

Roy F.

Bratslavsky

Ellen

Finkenauer

Catrin

Vohs

Kathleen D.

2001. “Bad Is Stronger than Good.” Review of General Psychology 5 (4): 323–70.

Berelson

Bernard R.

Lazarsfeld

Paul F.

McPhee

William N.

1954. Voting: A Study of Opinion Formation in a Presidential Election Campaign. Chicago: The University of Chicago Press.

Berry

William D.

Golder

Matt

Milton

Daniel

. 2012. “Improving Tests of Theories Positing Interaction.” Journal of Politics 74 (3): 653–71.

10.

Bisgaard

Martin

. 2015. “Bias Will Find a Way: Economic Perceptions, Attributions of Blame, and Partisan-motivated Reasoning during Crisis.” Journal of Politics 77 (3): 849–60.

11.

Books

John

Prysby

Charles

. 1999. “Contextual Effects on Retrospective Economic Evaluations: The Impact of the State and Local Economy.” Political Behavior 21 (1): 1–16.

12.

Brambor

Thomas

Clark

William Roberts

Golder

Matt

. 2006. “Understanding Interaction Models: Improving Empirical Analyses.” Political Analysis 14 (1): 63–82.

13.

Campbell

Angus

Converse

Philip E.

Miller

Warren E.

Stokes

Donald E.

1960. The American Voter. New York: John Wiley.

14.

Chaiken

Shelly

Liberman

Akiva

Eagly

Alice H.

1989. “Heuristic and Systematic Information Processing within and beyond the Persuasion Context.” In Unintended Thought, edited by Uleman

James

Bargh

John A.

, 212–52. New York: Guilford Press.

15.

Chzhen

Kat

Evans

Geoffrey

Pickup

Mark

. 2014. “When Do Economic Perceptions Matter for Party Approval? Examining the Endogeneity of Economic Perceptions before and during the Economic Downturn.” Political Behavior 36 (2): 291–313.

16.

Clark

William Roberts

Gilligan

Michael J.

Golder

Matt

. 2006. “A Simple Multivariate Test for Asymmetric Hypotheses.” Political Analysis 14 (3): 311–31.

17.

De Boef

Suzanna L.

Kellstedt

Paul M.

2004. “The Political (and Economic) Origins of Consumer Confidence.” American Journal of Political Science 48 (4): 633–49.

18.

Delli Carpini

Michael X.

Keeter

Scott

. 1996. What Americans Know about Politics and Why It Matters. New Haven: Yale University Press.

19.

Dickerson

Bradley T.

2016a. “Economic Perceptions, Presidential Approval, and Causality: The Moderating Role of the Economic Context.” American Politics Research 44 (6): 1037–1065.

20.

Dickerson

Bradley T.

2016b. “Motivated Reasoning, Ambivalence, and the Local Economy: Asymmetric Biases in the Formation of Economic Perceptions.” PhD diss., University of Mississippi, University.

21.

Ditto

Peter H.

Scepansky

James A.

Munro

Geoffrey D.

Apanovitch

Anne Marie

Lockhart

Lisa K.

1998. “Motivated Sensitivity to Preference-inconsistent Information.” Journal of Personality and Social Psychology 75 (1): 53–69.

22.

Enns

Peter K.

Kellstedt

Paul M.

McAvoy

Gregory E.

2012. “The Consequences of Partisanship in Economic Perceptions.” Public Opinion Quarterly 76 (2): 287–310.

23.

Evans

Geoffrey

Andersen

Robert

. 2006. “The Political Conditioning of Economic Perceptions.” Journal of Politics 68 (1): 194–207.

24.

Evans

Geoffrey

Pickup

Mark

. 2010. “Reversing the Causal Arrow: The Political Conditioning of Economic Perceptions in the 2000-2004 U.S. Presidential Election Cycle.” Journal of Politics 72 (4): 1236–51.

25.

Fiorina

Morris P.

1981. Retrospective Voting in American National Elections. New Haven: Yale University Press.

26.

Gerber

Alan S.

Huber

Gregory A.

2010. “Partisanship, Political Control, and Economic Assessments.” American Journal of Political Science 54 (1): 153–73.

27.

Gerber

Alan S.

Huber

Gregory A.

Washington

Ebonya

. 2010. “Party Affiliation, Partisanship, and Political Beliefs: A Field Experiment.” American Political Science Review 104 (4): 720–44.

28.

Gimpel

James G.

Reeves

Andrew

. 2012. “Ecologies of Unease: Geographic Context and National Economic Evaluations.” Political Behavior 34 (3): 507–34.

29.

Gomez

Brad T.

Wilson

J. Matthew

. 2001. “Political Sophistication and Economic Voting in the American Electorate: A Theory of Heterogeneous Attribution.” American Journal of Political Science 45 (4): 899–914.

30.

Goren

Paul

. 2005. “Party Identification and Core Political Values.” American Journal of Political Science 49 (4): 881–96.

31.

Goren

Paul

Federico

Christopher M.

Kittilson

Miki Caul

. 2009. “Source Cues, Partisan Identities, and Political Value Expression.” American Journal of Political Science 53 (4): 805–20.

32.

Groenendyk

Eric W.

2013. Competing Motives in the Partisan Mind: How Loyalty and Responsiveness Shape Party Identification and Democracy. New York: Oxford University Press.

33.

Hetherington

Marc J.

2001. “Resurgent Mass Partisanship: The Role of Elite Polarization.” American Political Science Review 95 (3): 619–31.

34.

Jacobson

Gary C.

2010. “Perception, Memory, and Partisan Polarization on the Iraq War.” Political Science Quarterly 125 (1): 31–56.

35.

Jerit

Jennifer

Barabas

Jason

. 2012. “Partisan Perceptual Bias and the Information Environment.” Journal of Politics 74 (3): 672–84.

36.

Johnston

Richard

. 2006. “Party Identification: Unmoved Mover or Sum of Preferences?” Annual Review of Political Science 9 (1): 329–51.

37.

Johnston

Ron

Patties

Charles

Dorling

Daniel

MacAllister

Iain

Tunstall

Helena

Rossiter

David

. 2000. “Local Context, Retrospective Economic Evaluations, and Voting: The 1997 General Election in England and Wales.” Political Behavior 22 (2): 121–43.

38.

Key

V. O.

1966. The Responsible Electorate. Cambridge, MA: Harvard University Press.

39.

Kim

Sung-Youn

Taber

Charles S.

Lodge

Milton

. 2010. “A Computational Model of the Citizen as Motivated Reasoner: Modeling the Dynamics of the 2000 Presidential Election.” Political Behavior 32 (1): 1–28.

40.

Kinder

Donald R.

Kiewiet

D. Roderick

. 1979. “Economic Discontent and Political Behavior: The Role of Personal Grievances and Collective Economic Judgments in Congressional Voting.” American Journal of Political Science 23 (3): 495–527.

41.

Kinder

Donald R.

Kiewiet

D. Roderick

. 1981. “Sociotropic Politics: The American Case.” British Journal of Political Science 11 (2): 129–61.

42.

Kunda

Ziva

. 1990. “The Case for Motivated Reasoning.” Psychological Bulletin 108 (3): 480–98.

43.

Lau

Richard R.

Redlawsk

David P.

2001. “Advantages and Disadvantages of Cognitive Heuristics in Political Decision Making.” American Journal of Political Science 45 (4): 951–71.

44.

Lavine

Howard G.

2001. “The Electoral Consequences of Ambivalence toward Presidential Candidates.” American Journal of Political Science 45 (4): 915–29.

45.

Lavine

Howard G.

Johnston

Christopher D.

Steenbergen

Marco R.

2012. The Ambivalent Partisan: How Critical Loyalty Promotes Democracy. New York: Oxford University Press.

46.

Lebo

Matthew J.

Cassino

Daniel

. 2007. “The Aggregated Consequences of Motivated Reasoning and the Dynamics of Partisan Presidential Approval.” Political Psychology 28 (6): 719–46.

47.

Lewis-Beck

Michael S.

2006. “Does Economics Still Matter? Econometrics and the Vote.” Journal of Politics 68 (1): 208–12.

48.

Lewis-Beck

Michael S.

Jacoby

William G.

Norpoth

Helmut

Weisberg

Herbert F.

2008. The American Voter Revisited. Ann Arbor: University of Michigan Press.

49.

Lodge

Milton

Taber

Charles S.

2000. “Three Steps toward a Theory of Motivated Political Reasoning.” In Elements of Reason, Cognition, Choice, and the Bounds of Rationality, edited by Lupia

Arthur

McCubbins

Matthew D.

Popkin

Samuel L.

, 183–213. Cambridge: Cambridge University Press.

50.

Lodge

Milton

Taber

Charles S.

2005. “The Automaticity of Affect for Political Leaders, Groups, and Issues: An Experimental Test of the Hot Cognition Hypothesis.” Political Psychology 26 (3): 455–82.

51.

Lord

Charles G.

Ross

Lee

Lepper

Mark R.

1979. “Biased Assimilation and Attitude Polarization: The Effects of Prior Theories on Subsequently Considered Evidence.” Journal of Personality and Social Psychology 37 (11): 2098–109.

52.

Miller

Arthur H.

Wlezien

Christopher

. 1993. “The Social Group Dynamics of Partisan Evaluations.” Electoral Studies 12 (1): 5–22.

53.

Morris

James P.

Squires

Nancy K.

Taber

Charles S.

Lodge

Milton

. 2003. “Activation of Political Attitudes: A Psychophysiological Examination of the Hot Cognition Hypothesis.” Political Psychology 24 (4): 727–45.

54.

Nir

Lilach

. 2011. “Motivated Reasoning and Public Opinion Perception.” Public Opinion Quarterly 75 (3): 504–32.

55.

Parker-Stephen

Evan

. 2013. “Tides of Disagreement: How Reality Facilitates (or Inhibits) Partisan Public Opinion.” Journal of Politics 75 (4): 1077–1088.

56.

Payne

John W.

Bettman

James R.

Johnson

Eric J.

1993. The Adaptive Decision Maker. New York: Cambridge University Press.

57.

Przeworski

Adam

. 1974. “Contextual Models of Political Behavior.” Political Methodology 1 (1): 27–61.

58.

Rahn

Wendy M.

1993. “The Role of Partisan Stereotypes in Information Processing about Political Candidates.” American Journal of Political Science 37 (2): 472–96.

59.

Redlawsk

David P.

2002. “Hot Cognition or Cool Consideration? Testing the Effects of Motivated Reasoning on Political Decision Making.” Journal of Politics 64 (4): 1021–1044.

60.

Redlawsk

David P.

Civettini

Andrew J. W.

Emmerson

Karen M.

2010. “The Affective Tipping Point: Do Motivated Reasoners Ever “Get It”?” Political Psychology 31 (4): 563–93.

61.

Rousseau

Denise M.

Snehal

Tijoriwala A.

1999. “What’s a Good Reason to Change? Motivated Reasoning and Social Accounts in Promoting Organizational Change.” Journal of Applied Psychology 84 (4): 514–28.

62.

Rudolph

Thomas J.

2003. “Who’s Responsible for the Economy? The Formation and Consequences of Responsibility Attributions.” American Journal of Political Science 47 (4): 698–713.

63.

Rudolph

Thomas J.

2006. “Triangulating Political Responsibility: The Motivated Formation of Responsibility Judgments.” Political Psychology 27 (1): 99–122.

64.

Sniderman

Paul M.

2000. “Taking Sides: A Fixed Choice Theory of Political Reasoning.” In Elements of Reason, Cognition, Choice, and the Bounds of Rationality, edited by Lupia

Arthur

McCubbins

Matthew D.

Popkin

Samuel L.

, 67–84. Cambridge: Cambridge University Press.

65.

Soroka

Stuart N.

2006. “Good News and Bad News: Asymmetric Responses to Economic Information.” Journal of Politics 68 (2): 372–85.

66.

Stanig

Piero

. 2013. “Political Polarization in Retrospective Economic Evaluations during Recessions and Recoveries.” Electoral Studies 32 (4): 729–45.

67.

Taber

Charles S.

Cann

Damon

Kucsova

Simona

. 2009. “The Motivated Processing of Political Arguments.” Political Behavior 31 (2): 137–55.

68.

Taber

Charles S.

Lodge

Milton

. 2006. “Motivated Skepticism in the Evaluation of Political Beliefs.” American Journal of Political Science 50 (3): 755–69.

69.

Taylor

Shelley E.

1991. “Asymmetrical Effects of Positive and Negative Events: The Mobilization-Minimization Hypothesis.” Psychological Bulletin 110 (1): 67–85.

70.

Tilley

James

Hobolt

Sara B.

2011. “Is the Government to Blame? An Experimental Test of How Partisanship Shapes Perceptions of Performance and Responsibility.” Journal of Politics 73 (2): 1–15.

71.

Weatherford

M. Stephen

. 1983. “Evaluating Economic Policy: A Contextual model of the Opinion Formation Process.” Journal of Politics 45 (4): 866–88.

72.

Weinschenk

Aaron C.

2010. “Revisiting the Political Theory of Party Identification.” Political Behavior 32 (4): 473–94.

73.

Westen

Drew

Blagov

Pavel S.

Harenski

Keith

Kilts

Clint

Hamann

Stephan

. 2006. “Neural Bases of Motivated Reasoning: An fMRI Study of Emotional Constraints on Partisan Political Judgment in the 2004 U.S. Presidential Election.” Journal of Cognitive Neuroscience 18 (11): 1947–58.

74.

Wlezien

Christopher

Franklin

Mark

Twiggs

Daniel

. 1997. “Economic Perceptions and Vote Choice: Disentangling the Endogeneity.” Political Behavior 19 (1): 7–17.

75.

Zaller

John

. 1992. The Nature and Origins of Mass Opinion. Cambridge: Cambridge University Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB