Side by Side,Worlds Apart

Abstract

The degree to which people desire policy change is a function of two factors: preferences for future policies and perceptions of current policies. Political scientists, pollsters, and pundits know a good deal about people’s policy preferences but surprisingly little about the distance of those preferences from policy perceptions . In this article, we assess the distance between policy perceptions and policy preferences to calculate the amount of policy change desired. The data come from an original survey tapping respondents’ preferred and perceived policies and from those few National Election Surveys where parallel items on policy preferences and perceptions were posed. By incorporating policy perceptions alongside of preferences, our findings provide a better indication of the gulf between the policy change desired by liberals and the policy change desired by conservatives. The findings help explain polarization in the United States where differences in policy preferences alone often do not indicate extreme diversity.

Keywords

policy perceptions policy preferences desired policy change ideology political behavior

Democratic accountability rests on the assumption that citizen input will guide the direction of current and future policies and a key feature of citizen input is the extent to which people desire change in current policies. To accurately assess the public’s desire for policy change, it is not enough to know what individuals want; it is also necessary to know what they believe current policies to be. To illustrate, imagine two people who believe governmental policies should strictly punish rule-breakers but differ in the extent to which they believe current policies do in fact “strictly punish.” One individual is convinced current policies are firm and punitive; the other is equally convinced lax sentences and cushy penitentiaries do nothing but mollycoddle criminals. Though the policy preferences are identical, the substantial difference in perception of existing conditions produces distinct perceptions of the degree of policy change needed. Individuals whose perceptions and preferences align are likely to be satisfied with the status quo given that the policy world they see resembles the one they desire. Individuals whose preferences significantly diverge from their perceptions see a very different policy world from the one they desire and, accordingly, are more likely to want the status quo to change, perhaps radically.

In this paper, we marshal survey evidence that poses parallel questions on people’s policy preferences (e.g., to what extent do they want policies that strictly punish criminals?) and perceptions (e.g., to what extent do they believe current policies strictly punish criminals?) to more fully understand the degree to which individuals desire policy change. Most surveys ask only about policy preferences and a few ask about perceptions, but it is rare for surveys to ask about both with parallel items. As a result of the paucity of parallel items, we draw here on data from an original survey of ours and supplement this analysis with the American National Election Surveys (ANES), which in selected years included the necessary parallel items. The results are consistent across surveys in showing that variations in policy perceptions contribute in important ways to the overall degree to which people desire policy change. Though our initial concern is with variation in preferences and perceptions across ideological groups, we also present the results of multivariate analyses (with demographics and partisan affiliation included among the explanatory variables) designed to account for variation in people’s policy preferences, their policy perceptions, and, most notably, the degree of policy change they desire.

Literature Review

There are several reasons to expect perceptions of political reality to differ from person to person and from ideology to ideology. One obvious source of perceptual variation is that people tend to gravitate toward particular media outlets (e.g., Druckman, 2005; Iyengar & Hahn, 2009). Relative to a steady dose of MSNBC, regular viewing of Fox News is likely to lead to quite different perceptions of political reality. Variations in the political coloration of work and family environments likely have similar effects. A drumbeat of water cooler anecdotes about poorly equipped armed forces or abuse of food stamps may lead to a perception of policy reality that would be very different if those anecdotes were replaced with tales of corporate greed and police misconduct. Another possible reason for variations in perceptions is that people’s differing physiological constitution may lead them to sense and to experience the world differently (Amodio, Jost, Master, & Yee, 2007; Kanai, Feilden, Firth, & Rees, 2011; Oxley et al., 2008; Smith, Oxley, Hibbing, Alford, & Hibbing, 2011). As a result, people may have different policy perceptions not just because of the lenses provided by family, coworkers, and media personalities but also because of their internal cognitive and physiological lenses.

Regardless of the source of perceptual differences, prior research on people’s desire for policy change has tended to emphasize one part of the larger story. Public opinion scholars have long been interested in understanding people’s policy preferences (Miller & Stokes, 1963; Monroe, 1998; Page & Shapiro, 1983; Stimson, MacKuen, & Erikson, 1995; Wright, Erikson, & McIver, 1987) but policy perceptions are not commonly measured on standard surveys and when such items are present, they are rarely included in a way that makes possible direct comparison with policy preferences. For example, seminal efforts to assess policy responsiveness typically calculated congruence measures where the public’s reported policy preferences were compared with the policy stances or actions of their corresponding representative. Similarly, efforts to assess policy responsiveness at the macro level have commonly used Stimson’s measure of the public’s mood that captures changes in the ideological direction of the public’s preferences across a wide range of policies. What each of these earlier approaches has in common is the decision to operationalize the public’s desire for policy change as mapping onto policy preferences. We contend that a singular focus on preferences, valuable as it is, underestimates the extent of policy change desired by the electorate. By not incorporating potential differences in perceptions of current policy realities, the implication is that everyone’s preferences begin from the same perception of the policy status quo and that differences in preferences are tantamount to differences in the desire for policy change.

To provide a more nuanced measure of the degree of policy change desired, perceptions must be measured alongside preferences. This is not to say that prior research has completely neglected the importance of perceptions broadly or of policy perceptions more specifically. In fact, the relevant role of the “pictures in our heads,” these perceptions of reality, has been recognized for at least 90 years (e.g., Lippmann, 1922). In terms of perceptions of existing policies, extensive research has explored the connection of perceptions and preferences. Perhaps most notably, the role of perceptions regarding economic performance in shaping preference for incumbent candidates has been widely assessed (e.g., Evans & Andersen, 2006; Rudolph, 2003; Sanders, 2000). Examining the effects of perceptions on policy preferences, rather than on candidate preferences, Hurwitz and Peffley (1990) uncover an intriguing pattern in which conservatives are more likely than liberals to see the international community as being populated by belligerent and expansionist states, with the policy desires of both ideological groups best seen as flowing from these perceptual differences.

Recent research casts a somewhat different light on the complex relationship between perceptions and preferences, indicating that individuals can mold perceptions to their preexisting preferences—preferences that are rooted in ideology or motivated social cognitions (e.g., Gaines, Kuklinski, Quirk, Peyton, & Verkuilen, 2007; Galdi, Arcuri, & Gawronski, 2008; Jost, 2006; Jost, Glaser, Kruglanski, & Sulloway, 2003; Taber, Cann, & Kucsova, 2009; Taber & Lodge, 2006; Parker-Stephen, 2011). Much of the burgeoning work on motivated reasoning, hot cognition, implicit attitudes, and related concepts looks at how misperceptions and corrected misperceptions subsequently affect policy preferences (e.g., Gilens, 2001; Howell & West, 2009; Kuklinski, Quirk, Jerit, Schweider, & Rich, 2000; Nyhan & Reifler, 2010; Sides & Citrin, 2007). These misperceptions extend all the way to factual knowledge as Shapiro and Bloch-Elkin (2008) demonstrate that liberals and conservatives differ even with regard to the basic facts that they accept, such as whether Saddam Hussein actually possessed weapons of mass destruction at the time of the U.S. invasion. Thus, the important political role of variations in people’s perceptions has been demonstrated in previous research.

In some cases, scholars have taken the next step and measured the distance between people’s perceptions and other important political variables. Typically, however, these distance measures involve perceptions of the gap between particular parties or candidates. For instance, measures of the perceived distance between political parties or major political actors on given policy issues have been widely used as predictors of political behaviors, most notably vote choice (e.g., Alvarez & Nagler, 1998; Downs, 1957; Granberg & Brown, 1992; Stokes, 1963). Somewhat relatedly, literature on authoritarianism has used an individual’s perceived ideological distance from major political actors as a measure of the “normative threat” they feel (e.g., Feldman & Stenner, 1997; Stenner, 2005).

As informative as this research has been, note that our approach is quite different. None of the distance measures described above quantifies the gap between people’s perceptions and preferences on a given policy area. In fact, the existing distance measures tend not to deal with policies. Rather, they usually are used to show that the ideological distance people perceive between the respective parties (or candidates) correlates with variables such as vote choice and party identification. Our approach is distinctive in that we are analyzing not the difference in the distance an individual perceives between the policy stances of parties (or candidates) but rather the distance between that individual’s own preference for a policy area and his or her perception of the current state of that policy area. Very few examples of this approach are to be found in previous research no doubt largely because the necessary survey items are rarely asked. Perhaps the closest to an exception is Grofman’s (1985) integration of expected changes from the status quo into a formal model of spatial issue voting.

Having parallel measures of policy perceptions and preferences allows us to pursue several interesting possibilities. For example, in addition to the well-documented ideological differences in policy preferences, ideological differences in policy perceptions deserve greater attention. In this respect, our work is building on a handful of important studies (most notably Gaines et al., 2007; Shapiro & Bloch-Elkin, 2008) that show how partisan and ideological differences can lead to different interpretations of the same objective policy reality. The novel contribution of the current article is to pair policy perceptions with policy preferences to calculate the degree of policy change desired. Differences between liberals and conservatives in the amount of policy change desired may be larger (and more accurately measured) when perceptions are taken into account. Finally, the improved measure of desire for policy change makes it possible to construct more meaningful multivariate models. We investigate these possibilities below, paying special attention to the role of ideology, party identification, and standard demographics.

Assessing the Role of Perceptions Using an Original Survey

One of the challenges of conducting research on policy perceptions in a federal system such as the United States is that policies often are different from one part of the country to another. For example, the degree to which the criminal justice system stresses compassion rather than stern punishment is quite different in Texas than it is in Massachusetts. Perceptions of the policy status quo, in other words, could vary simply because the policy status quo differs by geography. Twenty-four states currently have a “three strikes” or “habitual offender” statute and 26 states do not, and it is possible that variations in key policies such as these could filter through to policy perceptions. Perhaps this is why few extant national surveys query respondents about their perceptions of existing policies. Consequently, we were motivated to design our own survey and administer it to a random sample of adults from a single state.

In the summer of 2010, we hired a professional survey organization to recruit a sample (N = 343) of residents in the vicinity of Lincoln, NE. Initial contact was via phone subsequent to random selection from lists of telephone numbers (an appropriate mix of landline and cell phone lists). However, the recruited subjects had to be willing, in exchange for US$50, to travel to a location in the city to complete the computerized survey as well as other tasks. These procedures were not designed to result in a random national sample, rather by randomly selecting from the adult population from a single metropolitan area, our goal was to obtain a demographically diverse sample that was drawn neither from a convenience sample of undergraduates nor from a prerecruited bank of on-line responders. Most importantly, the restriction to a reasonably small geographical area means that, even with federal variations, at the time they completed the survey, all respondents in the sample were immersed in roughly the same objective policy realities.

The overall response rate was 26% (American Association for Public Opinion Research, Response Rate 1 [AAPOR RR1]) and the demographics of the sample are consistent with other adult samples (54% female, mean age 45, modal family income category [27% of sample] US$40,001 to US$60,000, 55% having at least some college experience). According to the Statistical Abstract of the United States comparable figures for the entire U.S. population are 51% female, median age 39 (this includes those younger than 18), 17% with a family income of US$40,000 to US$59,999, 56% having at least some college experience. Demographically, our sample is notably nonrepresentative of other samples of U.S. adults only in the sense of race—It is overwhelmingly (greater than 90%) White, a pattern that reflects the population from which the sample was drawn. With that one potential exception, we can think of no reason why the sample used here should produce results that are fundamentally misleading with regard to the relationship between policy perceptions and preferences. Furthermore, as a preview of the discussion to come, we are able to show that our central findings are generalizable across time and for nationally representative samples.

Upon arrival, participants completed an informed consent form before being ushered to a computer terminal so that they could begin answering the survey items. The first substantive battery posed five parallel items in the following format: “Setting aside what you prefer, the most accurate description of the political system in our society is that . . . ” Subjects were then to place their perception of the policies of the political system on a scale from 1 to 10. The first item ranged from “its policies are guided by traditional values” (1) to “its policies openly tolerate new lifestyles” (10); the second from “its policies do everything possible to protect against external threats” (1) to “its policies do not stress protection and security” (10); the third from “its policies strictly punish rule-breakers” (1) to “its policies display compassion for rule-breakers” (10); the fourth from “its policies benefit the rich even if they are undeserving” (1) to “its policies benefit the poor even if they are not making an effort” (10); and the fifth from “government is only minimally involved in society” (1) to “government is involved in most every facet of society” (10).

Of course, parallel item pairs could be constructed for a variety of additional policy areas and we very much hope this will be done in future work. We began with these particular items because issues of lifestyle, intergroup security, miscreant treatment, resource redistribution, and state-individual relations arise in any society and, thus, are the most likely to be transportable to other political systems (Petersen, 2010; Schwartz, Caprara, & Vecchione, 2010; Smith et al., 2011). The wording was kept broad and somewhat vague (“policies benefit the rich/poor”) because this is the level at which most people tend to think about politics. Many individuals are unfamiliar with narrow policy proposals (e.g., the fine-print of workfare programs). The vagueness of the wording opens the possibility that liberals and conservatives could have different interpretations of the perception scales, which could have measurement implications (see, for example, Hopkins & King, 2010; King, Murray, Salomon, & Tandon, 2003). However, similar results to those obtained with these scales are reported below when using NES perception items that involve more concrete wording.

We are initially interested in variations in policy perceptions across the ideological spectrum and the strategy we use to identify people’s ideological location is the most straightforward: We asked participants to self-identify their political ideology. The specific item was worded as follows: “Labels are often misleading, but in general, do you consider yourself liberal, conservative, or something in between?” The five available options were liberal, moderate leaning liberal, moderate, moderate leaning conservative, and conservative. Lumping together liberals and moderates leaning liberal, and lumping together conservatives and moderates leaning conservative creates a reasonably balanced distribution with 97 liberals, 115 moderates, and 128 conservatives. This ideological composition is reflective of the slightly conservative proclivities of the region of the country from which the sample was drawn as well as the larger reticence of many people across the country to label themselves as liberal.

Differing Perceptions of Political Reality

Partisanship, ideology, and even ethnicity have long been recognized as serving as perceptual screens through which individuals interpret the political world (Brady & Sniderman, 1985; Campbell, Converse, Miller, & Stokes, 1960; Conover & Feldman, 1981; Farwell & Weiner, 2000). Partisan and ideological differences can even lead individuals to reach different interpretations of the same objective set of facts (e.g., Gaines et al., 2007; A. Gerber & Green, 1999; Jerit & Barabas, 2012). Thus, our initial hypothesis is that liberals and conservatives differ markedly in the way they perceive the political world.

Furthermore, it seems likely that partisans and ideologues perceive that reality is stacked against them. More specifically, we believe liberals perceive the policies and practices of the country to be more in line with “conservative” positions while conservatives perceive these policies and practices to be more in line with “liberal” positions. The reason for this expectation stems from the same human tendencies evident when the fans of sports teams believe, as they so often do, that the conditions and decisions of referees are tilted in favor of the opposing team (e.g., Hastorf & Cantril, 1954). Similarly, most individuals perceive media content to be “hostile,” or ideologically at odds with, their personal preferences (e.g., Arpan & Raney, 2003; Baum & Gussin, 2007; Christen, Kannaovakun, & Gunther, 2002; Dalton, Beck, & Huckfeldt, 1998; Gunther & Schmitt, 2004; Turner, 2007; Vallone, Ross, & Lepper, 1985). This perceived hostility arises because neutral information tends to be evaluated as being further away from an individual’s preferred position and thus closer to the “other” end of the ideological spectrum. Consistent with this pattern, it seems likely that liberals perceive policy reality to tilt toward conservative positions just as conservatives perceive it tilts toward liberal positions.

Panel A of Figure 1 shows that these expectations are accurate when respondents are separated into three ideological groups. For each of the five policy items, compared with liberals, conservatives place the status quo closer to the preferences of liberals. For their part, liberals place policy reality closer to that desired by conservatives. These differences in liberals’ and conservatives’ perceptions of political reality are statistically significant for all five items. In sum, when it comes to the nature of governmental policies in the United States, liberals and conservatives appear to have quite different visions and as these respondents all came from the same state, the results cannot be attributed to differing policies from state to state.

Figure 1.

Differences in policy perceptions and policy preferences across ideological groups on five issue areas, 2010 sample.

Introducing Differing Political Preferences Into the Mix

After the participants answered several other items dealing with personal choice matters, especially their decisions in a set of hypothetical economic games, they were then presented with a battery of items parallel to the five analyzed above only this time the lead-in read, “Setting aside the way the political system actually is, which of these captures the way you would most like it to be (mark any number from 1 to 10)?” Thus, in this battery, participants reported their preferences rather than their perceptions but otherwise the five items are perfectly parallel. Panel B of Figure 1 provides the comparisons of the preferences of liberals, moderates, and conservatives.

It will come as no great surprise that liberals have more liberal political preferences than do conservatives on these five issues. Our purpose in presenting Panel B of Figure 1 is not to demonstrate the painfully obvious but rather to provide some perspective on the importance of the differences in perceptions that were reported in Panel A of Figure 1. Interestingly, there is one policy on which liberals and conservatives have much larger differences in preferences than they do on the others: whether or not the country should be guided by traditional values or, alternatively, should be tolerant of new lifestyles. On a scale of 1 to 10 with 10 being tolerant of new lifestyles and 1 being guided by traditional values, the mean liberal preference would have the country at a 7.52 whereas the mean conservative preference would have the country at 3.91, a difference in preferences of 3.61. Other areas give evidence of much smaller differences in liberal and conservative preferences, ranging from 1.64 on preferences for minimal government to 1.03 on policies that tilt toward the rich rather than the poor. As was the case for perceptions, the preferences of moderates again tend to be tucked between liberals and conservatives. In fact, across all five items, moderates are roughly equidistant from liberals and conservatives.

Comparing the liberal–conservative differences in preferences to the differences in perceptions indicates that, as would be expected, differences in preferences tend to be greater than differences in perceptions. In one area—traditional values as opposed to new lifestyles—the gap is enormous (3.61 for preferences to 0.64 for perceptions). What is more surprising, however, is that for one policy—protection from external threats—the perceptual difference is basically the same as the preference gap (1.16-1.15) and in the final area—policies benefiting the rich as opposed to the poor—liberals and conservatives are substantially more different in their perceptions than in their preferences (2.54-1.03). This suggests that when it comes to policies affecting redistribution, such as taxes and welfare, the main reason liberals and conservatives are so different is more a function of different perceptions of reality than different preferences for an ideal policy. This finding would not have been possible with traditional approaches focusing only on preferences and, if generalizable, raises the possibility that the supporters and detractors of the “Occupy Wall Street” movement are driven more by differing perceptions of wealth distribution and deservedness in the United States than by beliefs about the most appropriate distribution patterns.

Desired Policy Change

Change means departing from existing conditions. Measuring the degree of change desired thus requires information on preference for final destination and perception of existing location. Only by combining information on perceptions and preferences is it possible to gain an understanding of the differences in the degree of policy change desired. The survey items introduced here make it possible to merge perceptions and preferences, thereby permitting conclusions on the mean policy change desired.

In Figure 2, we combine the information from the two panels of Figure 1 to calculate the direction and magnitude of policy change liberals and conservatives desire using each group’s perception of the policy status quo as the starting point (we omit moderates so as not to clutter the figure). By aligning liberals’ preferences and conservatives’ preferences to each group’s perception of the status quo, it becomes clear that the biggest difference in the policy change desired by the liberals and conservatives in our sample is in the area of traditional values versus tolerance for new lifestyles. Liberals want policies to move 2.82 closer to the “tolerance for new lifestyles” pole from where they believe the country is at now. For their part, conservatives want policies to move 1.43 in the opposite direction, toward the “guided by traditional values” pole. Thus, the total difference in direction and magnitude of change desired by liberals and conservatives is 4.25.

Figure 2.

“Perception adjusted” desire for policy change for liberals and conservatives, 2010 sample.

The next largest difference in policy change desired by liberals and conservatives is in the area of redistribution. Liberals want policies to move 2.75 closer to the “benefiting the poor” pole from where they believe the country is now. Conservatives, however, want policies to move .82 in the other direction, toward the policies that “benefit the rich” pole. In this case, recognition of the sum total to which liberals and conservatives want change is grossly affected by including perceptions. If only policy preferences were tapped, the difference between liberals and conservatives would be taken to be a relatively modest 1.03, but when the substantial differences in perceptions of political reality are included, the real difference of 3.57 becomes evident. Compared with conservatives, liberals want policies to be only slightly more beneficial to the poor, so it would seem disputes on this issue should not be so bitter. However, when account is taken of the marked differences in liberals’ and conservatives’ perceptions of the current situation, it becomes apparent that, because of vastly differing beliefs about the current nature of U.S. policy, the changes conservatives feel compelled to advocate are dramatically different from those of liberals. In fact, 77% of the difference in policy change desired on this issue comes from differences in perceptions not preferences. Differences are smaller for the other three policy areas. Another observation made possible by our approach is that, as of the summer of 2010 and for this particular sample, conservatives tended to want significantly more change than liberals perhaps, as we will investigate soon, a function of the fact that Democrats controlled the houses of Congress as well as the White House at that time.

The larger point is that a single-minded focus on preferences leads to a significant underestimation of the extent to which liberals and conservatives want change in the status quo. Quantifying the extent of this underestimation is possible by adding the preference differences across the five areas (3.61 + 1.15 + 1.49 + 1.03 + 1.64 = 8.92), the perception differences (0.64 + 1.16 + 1.09 + 2.54 + 0.54 = 5.97), and computing the percent of the total desired policy change (8.92 + 5.97 = 14.89) that is due to differences in perceptions (5.97 / 14.89 = 40%). In other words, perceptual differences between liberals and conservatives account for 40% of the total difference in the policy changes desired by the two competing ideological groups in these five policy areas.

Accounting for Individual-Level Variations in Policy Change Desired

Ideological groupings are anything but homogeneous and the reasons for individual variations in the degree of policy change desired remain underspecified. To assess these reasons, we summed each respondent’s scores on the five policy areas to create measures of policy perceptions, policy preferences, and perception–preference differences.¹ We then accounted for variations in these concepts with a reasonably standard range of variables, including self-professed ideology, partisanship, overall level of political interest, and standard demographics (age, gender, and education). Given the potential nonlinearity of ideological identification (e.g., Inglehart & Flanagan, 1987; Knight, 1999, 2006; Marcus, Tabb, & Sullivan, 1974; Sidanius, 1985), we included dummies for each self-identification option (liberal and conservative), with moderates serving as the omitted category for estimation purposes.²

We first account for variance in overall policy perceptions (see Column 1). This variable is simply the additive total of each individual’s five perceptions (see Panel A of Figure 1) with more liberal perceptions scored higher. Even controlling for ideology, partisan identification has the expected effect, with strong Republicans more likely to perceive policies as tilting toward liberal positions (10 individuals listing “other” as their party identification were excluded from this portion of the analysis). Political interest and gender are unrelated to variations in policy perceptions but older people tend to perceive existing public policies as more liberal (but p > .05) and educated people tend to perceive existing policies as more conservative. With regard to ideology and with moderates as the comparison group, as in the bivariate analysis, liberals perceive existing public policies as more conservative, and conservatives perceive existing public policies as more liberal. The coefficient for conservatives is substantially larger than the coefficient for liberals, indicating that in mid-2010, relative to liberals, conservative perceptions were more different from the perceptions of moderates.

The second column in Table 1 repeats the analysis for policy preferences. In this multivariate model, the effects of age and gender are negligible but education predicts liberal preferences. Political interest, party ID, and ideology all have significant effects. Greater political interest correlates with more conservative policy preferences and, more predictably, self-professed liberals and Democrats have more liberal policy positions than moderates and self-professed conservative. The differential effects of aging for perceptions and preferences are interesting in that they suggest that the widely acknowledged increasing conservatism that accompanies the passing of the years may be attributable more to a growing perception that policies are liberal than to a building desire for policies to be conservative.

Table 1.

Accounting for Variations in Perceptions and Preferences, 2010 Sample.

Variable	Perceptions	Preferences	Desire for conservative policy change	Absolute degree of policy change desired
Liberal	−1.40	3.97***	−5.34***	1.31
	(0.91)	(0.79)	(1.20)	(1.05)
Conservative	3.01***	−3.18***	6.20***	4.72***
	(0.86)	(0.74)	(1.13)	(0.99)
Party ID	0.48*	−0.64**	1.11***	0.86**
	(0.22)	(0.19)	(0.29)	(0.26)
Political interest	−0.04	−0.14*	0.09	0.27*
	(0.08)	(0.07)	(0.10)	(0.09)
Age	0.05*	−0.03	0.08*	0.03
	(0.03)	(0.02)	(0.03)	(0.03)
Education	−0.52**	0.37*	−0.88**	−0.53*
	(0.19)	(0.17)	(0.25)	(0.22)
Male	−0.06	0.48	−0.53	−0.50
	(0.19)	(0.56)	(0.86)	(0.75)
Constant	25.99	27.28	−1.28	−1.44
N	329	329	329	329
F	10.90***	27.68***	34.67***	12.97***
R ²	.19	.38	.43	.22

Note. Ordinary least squares regression. Standard errors are in parentheses.

p < .10. **p < .01. ***p < .001. Two-tailed test.

An interesting aspect of these procedures is that they allow us to determine at the individual level if policy perceptions are merely a stand-in for preferences and therefore superfluous. In other words, is it simply automatic that the further to the left a person’s preferences are, the further to the right that person’s perceptions of the status quo will be? Even though Figure 1 provided evidence that group averages followed such a pattern, individual-level correlations reveal something quite different. Perceptions and preferences do indeed correlate negatively for three of the five policy areas (traditional values, decisive leaders, and protection from external threats) but are mildly positive for the other two (punishing rule-breakers and redistribution) but, regardless of direction, none of the five coefficients is statistically significant even at the more permissive .10 level. When all five policy areas are added together, the correlation between preferences and perception is negative, as predicted, and statistically significant but the size of the coefficient (−.26) indicates that variations in preferences account for less than 7% of the variance in perceptions (−.26 × −.26), thereby buttressing our point that perceptions, while related to preferences, are a largely independent concept, making independent contributions to the desire for policy change.³

The dependent variable for the model reported in the third column in Table 1, measure of the gap between perceptions and preferences, was computed by subtracting the policy preference score from the policy perception score for each of the five policy areas and then adding these differences together (larger numbers are associated with a desire for policy change in the conservative direction). The findings suggest that, ceterus paribus, age, conservatism, and identification as a strong Republican are associated with a desire for policy change in the conservative direction while education and liberalism are associated with a desire for policy change in the liberal direction.

The most telling results, however, are in the final column of the table. Here we have totaled the absolute values of the perception–preference differences on each of the five items, thus making the issue not the direction of change desired but rather the overall degree of change desired. A zero on this variable indicates that the policy world an individual perceives and the policy world that the individual desires are one and the same; higher numbers indicate greater aggregate differences between preferences and perceptions. Interestingly, education appears to reduce the distance between preferences and perceptions just as high political interest appears to increase the distance between preferences and perceptions. Moving to the ideological variables, major differences are indicated between liberals and conservatives. Liberals’ perceptions of policy reality are statistically indistinguishable from those of moderates while conservatives are substantially different, with a perception–preference gap that is 4.72 bigger than the perception–preference gap for moderates. As of the summer of 2010 at least, conservatives in our sample were clearly much more desirous of significant policy change than liberals and moderates. A key question becomes whether the type of individual desiring sizable policy changes in 2010 is similar to the type of individual wanting sizable policy changes in other election years. Fortunately, parallel items on policy preferences and policy perceptions were posed in seven ANES surveys.

Assessing Generalizability With NES Data

From 1980 through 1992, the ANES asked respondents to provide their preference and their perception “of the federal government’s activities” with respect to a range of policies. Although this battery has been used to compute policy distance measures between the respondent and particular candidates, to our knowledge, no other research has used it to measure the distance between respondent preferences and respondent perceptions of the federal government’s activity. The particular items are not identical to ours but do allow investigations in a variety of different electoral contexts.

Respondents were first provided with a seven-point scale (rather than 10) specific to each policy and asked, “Where would you place yourself on this scale or haven’t you thought much about this?” Next respondents were asked to place specific candidates and the two parties on the same scale. Finally and crucially, participants were asked, “Where would you place what the federal government is doing at the present time?” The following 10 policy areas were included, though not always in all election years (for specific item wording, see online Appendix A): (a) government services (all years), (b) minority aid (1980, 1982, 1984, and 1988), (c) U.S. involvement in Central America (1984 only), (d) defense spending (all years), (e) social/economic status of women (1984 only), (f) relations with Russia (1980, 1984, and 1988), (g) guaranteed jobs/standard of living (all years except 1986 and 1990), (h) aid to Blacks (1990 and 1992 only), (i) women’s equality (1980 and 1982 only), and (j) inflation/unemployment (1980 only). All the policy scales were coded so that 7 corresponded to the liberal response.

A disadvantage of using the ANES data is that the particular items included fluctuates from year to year. Still, the items on defense spending, government services, aid to minorities, and the role of government in guaranteeing jobs are posed in nearly every year and serve as anchors, with the other items making more occasional appearances. Our concern about respondents being affected by variations in policies from state to state is mitigated somewhat by the items’ specific focus on the “federal government’s activities.” Though the extent to which individuals can parse federal and state policies may be a matter of debate, the existence of perception and preference items in national samples is an important addition. Moreover, the ability to compare perception and preference differences in seven different electoral contexts is valuable. The timing of the inclusion of these items is a bit unfortunate in that six of the seven elections for which the ANES collected data (all but 1980) were held when a Republican was in the White House but, at the least, valuable comparisons can be made between the 1982 and 1992 NES results on one hand and the two instances we have of respondents answering in the context of Democratic Administrations on the other: the 1980 NES data and our 2010 data. An obvious additional hypothesis that becomes testable with data over several years is that conservatives will be more desirous of policy change when a Democrat is in the White House and liberals will be more desirous of policy change when a Republican is in the White House.

To assess the amount of policy change desired in various election years, we first had to create overall measures of policy perceptions and policy preferences.⁴ This we did by first summing each respondent’s scores on the policy areas available in each NES survey, subtracting the policy preference score from the policy perception score and finally taking the absolute value of that result. These procedures parallel those generating our key “absolute degree of policy change desired” variable in Table 1 (the last column) and yield a measure of the extent to which preferences deviate from perceptions, regardless of direction. To account for variations in the amount of policy change desired, we included items very similar to those included in Table 1: self-professed ideology, a standard seven-point measure of partisanship (0 = strong Democrat to 6 = strong Republican), standard demographics (age, male, and education) and overall level of interest in the campaigns. As before, indicator variables were included for liberals and conservatives with moderates serving as the omitted category. The results from the ANES data are presented in Table 2 with the relevant model from our 2010 data presented for comparison purposes.

Table 2.

Explaining Variations in the Amount of Policy Change Desired, ANES (1980-1992) and 2010 Survey Data.

Variable	1980	1982	1984	1986	1988	1990	1992	2010
Liberal	−0.32	1.96***	3.05***	0.81***	1.03**	0.83**	0.72**	1.31
	(0.88)	(0.54)	(0.54)	(0.17)	(0.36)	(0.27)	(0.23)	(1.05)
Conservative	1.86*	0.01	0.04	−0.15	−0.11	0.10	0.13	4.72***
	(0.81)	(0.49)	(0.48)	(0.15)	(0.32)	(0.24)	(0.21)	(0.99)
Party ID	0.84***	−0.50***	−1.11***	−0.26***	−0.56***	−0.24***	−0.29***	0.86**
	(0.17)	(0.11)	(0.10)	(0.03)	(0.07)	(0.05)	(0.05)	(0.26)
Political interest	−0.05	0.62*	0.78*	−0.002	0.18	0.16	0.22*	0.27*
	(0.48)	(0.29)	(0.31)	(0.09)	(0.19)	(0.15)	(0.13)	(0.09)
Age	−0.02	−0.03*	−0.02*	−0.01**	−0.02**	−0.01*	−0.02***	0.03
	(0.02)	(0.01)	(0.31)	(0.004)	(0.008)	(0.01)	(0.005)	(0.03)
Education	−0.28*	0.18*	0.15*	−0.04	−0.12*	0.03	−0.06	−0.53*
	(0.14)	(0.09)	(0.09)	(0.03)	(0.05)	(0.04)	(0.04)	(0.22)
Male	0.95	−0.78*	−0.65*	−0.67***	−0.50*	0.15	−0.40*	−0.50
	(0.63)	(0.40)	(0.41)	(0.13)	(0.25)	(0.20)	(0.17)	(0.75)
Constant	14.96***	8.70***	13.09***	5.87***	11.16***	6.10***	7.58***	−1.44
	(2.22)	(1.43)	(1.50)	(0.42)	(0.89)	(0.68)	(0.62)
N	369	583	914	1,353	945	988	1,238	329
F	9.26***	13.87***	35.42***	27.58***	22.96***	8.16***	15.96***	12.97***
R ²	.15	.14	.21	.13	.15	.06	.08	.22

Note. Ordinary least squares regression. Standard errors are in parentheses. ANES = American National Election Surveys.

p < .10. **p < .01. ***p < .001. Two-tailed test.

The relationship between the demographic variables and degree of desired policy change in the ANES data varies markedly though the findings for these variables are difficult to interpret given that the ideological variables are in the model as well. For most of the demographic variables, the signs switch from one year to the next and are usually not significant at the .05 level. The ideological variables tell a more interesting story. Remember that in 2010, conservatives desired much more policy change than liberals and our suspicion was that this pattern could be the result of the fact that Democrats controlled the levers of power (notably, the Presidency) at the time the survey was conducted. Does the pattern reverse under Republican Administrations? Table 2 suggests the answer to that question is yes. In 2 of the 8 election years for which we have perception and preference data (1980 and 2010), a Democrat was in the White House, and these are the only 2 years in which liberals did not want more policy change than moderates. In contrast, in these 2 years, conservatives desired substantively and significantly more policy change than moderates. Moving to the years in which a Republican was in the White House (1982 through 1992), the pattern continues to be evident. In these years, liberals always desired significantly more policy change than moderates and conservatives never desired more policy change than moderates. These results suggest that the desire for policy change varies in a theoretically sensible fashion in response to changes in the political context where those ideologically at odds with the party in control of the White House are more likely to desire greater change from the perceived status quo. Similarly, Gerber and Huber (2010) show that perceptions of partisan competence with respect to the economy are sensitive to shifts in party control. However, given the limited NES data availability, we must be cautious regarding the generalizability of this particular finding and future work is needed. Our point here is to underscore the importance of attending to perceptions as a critical element of desired levels of policy change. As with our own survey, we show once again that perceptions along with preferences need to be taken into account to have a fuller picture of the extent of policy change desired by the public. We illustrate this point further below.

It might be thought that a similar result could have been obtained with more traditional data and analysis. As mentioned previously, the norm is to ask survey items only on people’s policy preferences and to ignore policy perceptions. If this practice were followed rather than the combination of perceptions and preferences that ultimately produced Table 2, the results would look like those in Table 3. As can be seen, in contrast to the data incorporating perceptions, preferences alone do not follow a clear pattern according to the party controlling the White House. Instead, the traditional, preference-based approach in Table 3 does not provide any indication that liberals want more policy change under Republican Administrations and that conservatives want more policy change under Democratic Administrations. When perceptions are incorporated, however, the pattern becomes sensible.

Table 3.

Explaining Variation in Liberal Policy Preferences, ANES (1980-1992) and 2010 Survey Data.

Variable	1980	1982	1984	1986	1988	1990	1992	2010
Liberal	3.89***	2.61***	2.52***	0.57***	1.29***	1.08***	1.01***	3.97***
	(0.71)	(0.47)	(0.48)	(0.15)	(0.31)	(0.24)	(0.19)	(0.79)
Conservative	−2.46***	−1.74***	−2.12***	−.63***	−1.36***	−.65**	−.84***	−3.18***
	(0.67)	(0.43)	(0.43)	(0.13)	(0.27)	(0.21)	(0.18)	(0.74)
Party ID	−0.85***	−0.76***	−1.23***	−0.36***	−0.62***	−0.42***	−0.54***	−0.64**
	(0.14)	(0.09)	(0.09)	(0.03)	(0.17)	(0.04)	(0.04)	(0.19)
Political interest	0.21	0.41	0.18	−0.04	−0.31*	0.05	−0.16*	−0.14*
	(0.39)	(0.25)	(0.28)	(0.08)	(0.16)	(0.13)	(0.11)	(0.07)
Age	−0.07***	−0.03*	−0.01	−0.01***	−0.02*	−0.02***	−0.02***	−0.03
	(0.02)	(0.01)	(0.01)	(0.003)	(0.01)	(0.01)	(0.004)	(0.02)
Education	0.001	0.17*	0.15*	0.008	0.06	0.12***	−0.04	0.378*
	(0.11)	(0.07)	(0.08)	(0.02)	(0.04)	(0.04)	(0.03)	(0.17)
Male	−0.32	−0.95**	−1.46***	−0.54***	−0.52*	−0.53**	−0.55***	0.48
	(0.52)	(0.35)	(0.36)	(0.11)	(0.22)	(0.04)	(0.14)	(0.56)
Constant	32.41***	21.28***	30.76***	10.34***	18.94***	12.95***	15.83***	27.28***
	(1.80)	(1.22)	(1.33)	(0.36)	(0.75)	(0.58)	(0.52)
N	464	684	1031	1489	1079	1121	1405	329
F	34.29***	44.99***	63.21***	58.03***	52.52***	40.94***	82.39***	27.68**
R ²	.34	.32	.30	.22	.26	.20	.29	.38

Note. Ordinary least squares regression. Standard errors are in parentheses. ANES = American National Election Surveys.

p < .10. **p < .01. ***p < .001. Two-tailed test.

Conclusion

Tea-party as well as Christian right conservatives undoubtedly have different preferences for public policy than ardent liberals, but this is not the extent of their differences. A striking and underanalyzed element of these groups is their distinct perceptions of existing policies. Many conservatives are convinced that today’s America is one in which external threats are blithely ignored, lawbreakers are treated far too warmly, subversive lifestyles are encouraged, the federal government is busy terminating the last of our essential, time-honored freedoms, sloth is rewarded, and hard work is punished by a crushing tax burden. Liberals are left wondering whether they are living in the same country as conservatives. Many of them see an America that is voraciously militaristic and artificially patriotic, fixated on retribution at the expense of rehabilitation, embarrassingly biased toward the well-to-do, and callously indifferent to the very real needs of the poor and the nontraditional.

Insufficient scholarly attention has been paid to the extent to which desired policy change is a function of differences not just in policy preferences but also in perceptions of existing public policies. Perhaps this lack of attention is a throwback to the assumption that perceptions are tied to reality, leaving preferences as the only element of the equation that varies meaningfully. Perhaps it is attributable to a belief that perceptual differences are just an artifact of differences in preferences, that liberal policy preferences somehow force people to perceive existing conditions as being tilted toward conservative policies and vice versa. Neither of these premises receives empirical support. Psychologists have shown that perceptions are influenced by all sorts of subthreshold and subrational factors and, as we have seen, at the individual level, preferences and perceptions on a given policy area are unrelated—and even if they were related, the causal order would be unclear.

The results marshaled here are largely consistent across the ANES samples and our own 2010 sample. They indicate the importance of varying perceptions of the policy status quo and, we believe, suggest merit in additional efforts to collect data and otherwise devote attention to the causes and consequences of the fact that some people see the political world very differently from others. Our intended contribution is to call attention to the value of studying perceptions of the policy world and to provide an illustration of the procedures by which information on perceptions can be collected and integrated. Policy perceptions vary wildly from person to person and feed directly into the extent to which people believe change in the policy status quo is necessary—a central factor in subsequent political behavior.

Footnotes

Acknowledgements

The authors gratefully acknowledge the assistance of Jessica Mohatt, Karl Giuseffi, Carly M. Jacobs, Amanda Friesen, and Michael W. Gruszczynski. Also, thanks to the APR editor and the anonymous reviewers for their helpful comments.

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was made possible by a grant from the National Science Foundation (BCS-0826828).

Notes

Author Biographies

Dona-Gene Mitchell is an assistant professor of political science at the University of Nebraska–Lincoln. Her work draws heavily on relevant psychological perspectives to better understand how and how well citizens use political information to inform their judgments.

Matthew V. Hibbing is an assistant professor of political science at the University of California, Merced. His research centers on individual-level predispositions that shape political attitudes and behaviors.

Kevin B. Smith is a professor of political science at the University of Nebraska–Lincoln. His primary research interests are the biology and psychology of attitudes and behavior.

John R. Hibbing is the Foundation Regents Professor of political science and psychology at the University of Nebraska–Lincoln where he has conducted research on Congress, public opinion, comparative legislatures, and the relationship of biology to political orientations. He is coauthor (with Kevin B. Smith and John R. Alford) of Predisposed: Liberals, Conservatives, and the Biology of Political Differences (Routledge, 2014).

References

Alvarez

R. M.

Nagler

(1998). Economics, entitlements, and social issues: Voter choice in the 1996 presidential election. American Journal of Political Science, 42, 1349-1363.

Amodio

D. M.

Jost

J. T.

Master

S. L.

Yee

C. M.

(2007). Neurocognitive correlates of liberalism and conservatism. Nature Neuroscience, 10, 1246-1247.

Arpan

L. M.

Raney

A. A.

(2003). An experimental investigation of news source and the hostile media effect. Journalism & Mass Communication Quarterly, 80, 265-281.

Baum

M. A.

Gussin

(2007). In the eye of the beholder: How information shortcuts shape individual perceptions of bias in the media. Quarterly Journal of Political Science, 3, 1-31.

Brady

H. E.

Sniderman

P. M.

(1985). Attitude attribution: A group basis for political reasoning. American Political Science Review, 79, 1061-1078.

Campbell

Converse

P. E.

Miller

W. E.

Stokes

D. E.

(1960). The American voter. New York, NY: John Wiley.

Christen

C. T.

Kannaovakun

Gunther

A. C.

(2002). Hostile media perceptions: Partisan assessments of press and public during the 1997 United Parcel Service strike. Political Communication, 19, 423-436.

Conover

Feldman

(1981). The origins and meaning of liberal-conservative self-identifications. American Journal of Political Science, 25, 617-645.

Dalton

R. J.

Beck

P. A.

Huckfeldt

(1998). Partisan cues and the media: Information flows in the 1992 presidential election. American Political Science Review, 92, 111-126.

10.

Downs

(1957). An economic theory of democracy. New York, NY: Harper.

11.

Druckman

J. N.

(2005). Media matter: How newspapers and television news cover campaigns and influence voters. Political Communication, 22, 462-481.

12.

Evans

Andersen

(2006). The political conditioning of economic perceptions. Journal of Politics, 68, 194-207.

13.

Farwell

Weiner

(2000). Bleeding hearts and the heartless: Popular perceptions of liberal and conservative ideologies. Personality and Social Psychology Bulletin, 26, 845-852.

14.

Feldman

Stenner

(1997). Perceived threat and authoritarianism. Political Psychology, 18, 741-770.

15.

Gaines

B. J.

Kuklinski

J. H.

Quirk

P. J.

Peyton

Verkuilen

(2007). Same facts, different interpretations: Partisan motivation and opinion on Iraq. Journal of Politics, 69, 957-974.

16.

Galdi

Arcuri

Gawronski

(2008). Automatic mental associations predict future choices of undecided decision makers. Science, 321, 1100-1102.

17.

Gerber

Green

(1999). Misperceptions about perceptual bias. Annual Review of Political Science, 2, 189-210.

18.

Gerber

A. L.

Huber

G. A.

(2010). Partisanship, partisan control and economic assessments. American Journal of Political Science, 54, 153-173.

19.

Gilens

(2001). Political ignorance and collective policy preferences. American Political Science Review, 95, 379-396.

20.

Granberg

Brown

T. A.

(1992). The perception of ideological distance. The Western Political Quarterly, 45, 727-750.

21.

Grofman

(1985). The neglected role of the status quo in models of issue voting. Journal of Politics, 47, 229-237.

22.

Gunther

A. C.

Schmitt

(2004). Mapping boundaries of the hostile media effect. Journal of Communication, 54, 55-70.

23.

Hastorf

A. H.

Cantril

(1954). They saw a game: A case study. Journal of Abnormal and Social Psychology, 49, 129-134.

24.

Hopkins

D. J.

King

(2010). Improving anchoring vignettes designing surveys to correct interpersonal incomparability. Public Opinion Quarterly, 74, 201-222.

25.

Howell

W. G.

West

M. R.

(2009). Educating the public. Education Next, 9(3), 41-47.

26.

Hurwitz

Peffley

(1990). Public images of the Soviet union: The impact on foreign policy attitudes. Journal of Politics, 52, 3-28.

27.

Inglehart

Flanagan

S. C.

(1987). Value change in industrial societies. American Political Science Review, 81, 1289-1319.

28.

Iyengar

Hahn

(2009). Red media, blue media: Evidence of ideological selectivity in media use. Journal of Communication, 59, 19-39.

29.

Jerit

Barabas

(2012). Partisan perceptual bias and the information environment. Journal of Politics, 74, 672-684.

30.

Jost

J. T.

(2006). The end of the end of ideology. American Psychologist, 61, 651-670.

31.

Jost

J. T.

Glaser

Kruglanski

A. W.

Sulloway

F. J.

(2003). Political conservatism as motivated social cognition. Psychological Bulletin, 129, 339-375.

32.

Kanai

Feilden

Firth

Rees

(2011). Political correlations are correlated with brain structures in young adults. Current Biology, 21, 677-680.

33.

King

Murray

C. J.

Salomon

J. A.

Tandon

(2003). Enhancing the validity and cross-cultural comparability of measurement in survey research. American Political Science Review, 97, 567-584.

34.

Knight

(1999). Measuring liberalism and conservatism. In Robinson

J. P.

Shaver

P. R.

Wrightsman

L. S.

(Eds.), Measures of political attitudes (pp. 681-736).New York, NY: Academic Press.

35.

Knight

(2006). Transformations of the concept of ideology in the twentieth century. American Political Science Review, 100, 619-626.

36.

Kuklinski

J. H.

Quirk

P. J.

Jerit

Schweider

Rich

R. F.

(2000). Misinformation and the currency of democratic citizenship. Journal of Politics, 62, 790-816.

37.

Lippmann

(1922). Public opinion. New York, NY: Macmillan.

38.

Marcus

G. E.

Tabb

Sullivan

J. L.

(1974). The application of individual differences scaling to the measurement of political ideologies. American Journal of Political Science, 18, 405-420.

39.

Miller

Stokes

(1963). Constituency influence in congress. American Political Science Review, 57, 45-56.

40.

Monroe

(1998). Public opinion and public policy, 1980-1993. Public Opinion Quarterly, 62, 6-28.

41.

Nyhan

Reifler

(2010). When corrections fail: The persistence of political misperceptions. Political Behavior, 32, 303-330.

42.

Oxley

D. R.

Smith

K. B.

Alford

J. R.

Hibbing

M. V.

Miller

J. L.

Scalora

Hibbing

J. R.

(2008). Political attitudes vary with physiological traits. Science, 321, 1667-1670.

43.

Page

Shapiro

(1983). Effects of public opinion on policy. American Political Science Review, 77, 175-190.

44.

Parker-Stephen

(2011, January). Following the sport vs. rooting for the team: The multiple paths to political sophistication. Presented at the annual meetings of the Midwest Political Science Association, New Orleans, LA.

45.

Petersen

M. B.

(2010). Distinct emotions, distinct domains: Anger, anxiety, and perceptions of intentionality. Journal of Politics, 72, 357-369.

46.

Rudolph

T. J.

(2003). Who’s responsible for the economy? The formation and consequences of responsibility attributions. American Journal of Political Science, 47, 698-713.

47.

Sanders

(2000). The real economy and the perceived economy in popularity functions: How much do voters need to know? A study of British data, 1974-97. Electoral Studies, 19, 275-294.

48.

Schwartz

S. H.

Caprara

G. V.

Vecchione

(2010). Basic personal values, core political values, and voting: A longitudinal analysis. Political Psychology, 31, 421-452.

49.

Shapiro

R. Y.

Bloch-Elkin

(2008). Do the facts speak for themselves? Partisan disagreements as a challenge to democratic competence. Critical Review, 20, 115-139.

50.

Sidanius

(1985). Cognitive functioning and sociopolitical ideology revisited. Political Psychology, 6, 637-661.

51.

Sides

Citrin

(2007, April). How large the huddled masses? The causes and consequences of public misperceptions about immigrant populations. Paper presented at the 2007 annual meeting of the Midwest Political Science Association, Chicago, IL.

52.

Smith

K. B.

Oxley

D. R.

Hibbing

M. V.

Alford

J. R.

Hibbing

J. R.

(2011). Linking genetics and political attitudes: Reconceptualizing political ideology. Political Psychology, 32, 369-397.

53.

Stenner

(2005). The authoritarian dynamic. Cambridge, UK: Cambridge University Press.

54.

Stimson

J. A.

MacKuen

M. B.

Erikson

R. S.

(1995). Dynamic representation. American Political Science Review, 88, 543-565.

55.

Stokes

D. E.

(1963). Spatial models of party competition. American Political Science Review, 57, 368-377.

56.

Taber

C. S.

Cann

Kucsova

(2009). The motivated processing of political arguments. Political Behavior, 31, 137-155.

57.

Taber

C. S.

Lodge

(2006). Motivated skepticism in the evaluation of political beliefs. American Journal of Political Science, 50, 755-769.

58.

Turner

(2007). The messenger overwhelming the message: Ideological cues and perceptions of bias in television news. Political Behavior, 29, 441-464.

59.

Vallone

R. P.

Ross

Lepper

M. R.

(1985). The hostile media phenomenon: Biased perception and perceptions of media bias in coverage of the Beirut massacre. Journal of Personality and Social Psychology, 49, 577-585.

60.

Wright

Jr. Erikson

McIver

(1987). Public opinion and policy liberalism in the American states. American Journal of Political Science, 31, 980-1001.