Authoritarianism and American Political Behavior from 1952 to 2008

Abstract

Using American National Election Studies (NES) data from 1952 to 2008—a longer timespan than any analysis to date—we evaluate the leading claims about growing polarization along authoritarian/nonauthoritarian lines and the reasons for that growth. We find authoritarianism’s impact has grown for partisanship and voting but has been consistent for policy attitudes—usually present for “social” and defense issues, but less so for social welfare and foreign policy. This suggests that authoritarianism’s importance is related to strategic politicians advancing issues that touch on the authoritarian/nonauthoritarian divide, and varies across election years.

Keywords

authoritarianism public opinion partisan polarization

Influential in the 1950s (Adorno et al. 1950; Lane 1955), then largely moribund for nearly a half-century, the concept of authoritarianism—a set of personality traits associated with aversion to difference and conformity to authority—has reemerged as a prominent topic in recent research on American political behavior (Altemeyer 1996; Feldman and Stenner 1997; Huddy et al. 2005; Mockabee 2007; Stenner 2005). Two of the most important recent offerings contend that authoritarianism or closely related “parenting values” are at the heart of the growing ideological and partisan divide in American mass politics. Barker and Tinnick (2006) argue that parenting values—measured with the same indicators typically used to measure authoritarianism—are related to preferences on a wide range of policy issues and thus underlie the growth of mass ideological constraint and party polarization. In contrast, Hetherington and Weiler (2009) contend that authoritarianism shapes views on a limited number of issues, but those issues have moved to a more central place on the political agenda, thus tying authoritarianism more closely to partisanship and vote choice.

The difference between the Barker–Tinnick and Hetherington–Weiler accounts points to a broader question about how authoritarianism and other core values and psychological dispositions have contributed to the growth of political polarization in the United States. Has growing polarization along authoritarian/nonauthoritarian lines resulted from developments outside of politics—for example, growing societal divides about how best to raise children and structure family life—or from politics itself: party politicians devoting more attention to the issues most clearly structured by authoritarianism, thus making authoritarianism and nonauthoritarianism more politically relevant?

In this article, we argue for the political explanation. Authoritarianism’s impact on policy attitudes should be consistently related to a limited number of issues and should grow only as those issues grow more politically important. While this supports Hetherington and Weiler, it also suggests that they may have overstated the degree to which authoritarianism is consistently important for electoral behavior. If authoritarianism’s impact is contingent on the salience of particular issues, it will vary with the context of campaigns and be inconsistent across election years. Evaluating this account requires data from both recent years and over a broader time frame. Using data from several presidential-election-year surveys conducted by the American National Election Studies (NES) from 1952 to 2008, we assess the political impact of authoritarian perspectives over a much wider time horizon than any analysis to date. We find that authoritarianism’s impact on partisanship and voting behavior has grown, but its influence on policy attitudes has been highly consistent over time—nearly always structuring attitudes toward “social” issues, broadly defined, as well as national defense, but only indirectly shaping social welfare attitudes and inconsistently affecting views on foreign conflicts. A more detailed examination of the effect of authoritarianism on recent political orientations finds that the effect was noticeably larger in 2004 than in 2008.

Authoritarianism, Policy Attitudes, and the Growing Partisan Divide

Barker and Tinnick and Hetherington and Weiler suggest that the influence of authoritarianism (or closely related values) on American partisanship and voting behavior has increased considerably in recent years. Where their accounts differ is on authoritarianism’s impact on Americans’ policy attitudes and changes over time in the depth and breadth of that impact. Barker and Tinnick’s work implies a growing and broadening influence of authoritarianism on policy attitudes. Hetherington and Weiler argue that authoritarianism’s political import has increased because the limited set of issues for which it consistently matters has grown more politically salient.

Barker and Tinnick (2006) assess the argument of cognitive linguist George Lakoff (2002) that the principal divide in contemporary American politics is between individuals with “strict” and “nurturant” parenting values and moral visions. Although neither Lakoff nor Barker and Tinnick term their core concept “authoritarianism,” the concepts are very closely related—for example, the disciplinarian preference for strong parental authority is closely related to a general desire for and conformity to authority. As Feldman (2003, 55) argues, “People . . . who desire social conformity should want children to be taught to be good, obedient citizens. Conversely, those who value autonomy should want to encourage it in their children.” Indeed, Barker and Tinnick measure parenting values with the same indicators used to measure authoritarianism (Hetherington and Weiler 2009; Stenner 2005).¹ They argue that parenting values structure attitudes on a wide range of policy issues, and their analysis of 2000 NES data shows that proponents of strict parenting fall to the right of nurturers on nearly all major policy dimensions, including immigration, gay rights, welfare, defense, and tax policy. Thus, Barker and Tinnick contend that the strict–nurturant divide is central not only to increasing partisan polarization but also to the growth of ideological constraint in mass policy attitudes.² They conclude, “The stronger one’s views regarding childrearing—either in terms of nurturance or discipline—the more consistently liberal or conservative one’s political attitudes tend to be” (Barker and Tinnick 2006, 259; italics in original). While Barker and Tinnick (2006, 251) note that they are “officially agnostic as to the temporal aspects of Lakoff’s story,” their contention that parenting values are now connected to attitudes on the whole range of policy issues suggests either that these values always have been relevant for citizens’ views on a wide variety of policy issues or that their influence has grown and expanded to new issue agendas. Barker and Tinnick (2006, 251–52) seem to lean toward the latter interpretation, suggesting that the rise of new cultural issues, ideologically oriented media, and elite leadership of mass opinion may have made parenting values more relevant to a wide range of policy concerns.

Hetherington and Weiler (2009) document a growing relationship between authoritarianism and party identification—with Republican identifiers growing more authoritarian and nonauthoritarian values becoming more common among Democrats—and view this development as critical to the growth of polarization. They attribute this growing partisan divide to the issues on which authoritarians and nonauthoritarians always have differed growing more central to American politics. Specifically, they argue that authoritarianism has long been closely connected to views on issues that define Americans’ “worldviews,” such as civil rights for women, minorities, and homosexuals; immigration; and the appropriate balance between security and liberty in national defense efforts. These connections have not changed or come to include other policy domains such as economic and social welfare issues. Instead, changing societal circumstances and strategic manipulation of the policy agenda by political elites has created a “worldview evolution” in which “politics is increasingly contested over issues for which preferences are structured by authoritarianism” (Hetherington and Weiler 2009, 9).

In contrast to the view that authoritarianism’s impact has extended to the entire range of salient political issues (Barker and Tinnick 2006), we suspect that authoritarianism has a stable and limited impact on policy attitudes. Authoritarians’ aversion to difference and staunch desire to preserve the prevailing social order make them quite likely to be conservative on cultural issues such as gay rights and abortion, racial issues concerning government help for African Americans, and issues related to immigration (Lane 1955; Mockabee 2007; Stenner 2005). Because of authoritarians’ heightened sense of threat and desire to deal with threats in an unambiguous way, they also should support strong defense capabilities, the death penalty, and less restrictive gun laws (Altemeyer 1996; Hetherington and Weiler 2009; Lipset 1959).

However, connections between authoritarianism and attitudes toward social welfare issues and the proper course of U.S. foreign policy may be more tenuous. The authoritarian desire for strength and forceful action by political leaders does not suggest a preference for a limited government role in economic and social welfare matters. In fact, Lipset (1959) finds a clear connection between authoritarianism and support for parties favoring a larger social welfare state in the 1950s. While this relationship may have changed over time as social welfare attitudes became more closely connected to attitudes toward African Americans (Edsall and Edsall 1992; Gilens 1996), such a change at most suggests an indirect and weak relationship between authoritarianism and social welfare opinions. Moreover, while authoritarians should support maintaining strong military capabilities, they may not share the commitment of modern conservatives to intervention in foreign affairs. Instead, given the authoritarian distaste for out-groups, they may tend to be isolationists (Janowitz and Marvick 1953; Lane 1955).

If authoritarianism’s influence on policy attitudes is limited in this way, then the chief explanation for the growing polarization of American politics along authoritarian and nonauthoritarian lines is probably not factors outside of party politics—such as the growing popularity of new parenting orientations or the decreasing length of political sound bites—making authoritarianism more relevant for all policy attitudes. Rather, it is party politics itself. Strategic party politicians have seen a potential advantage in emphasizing issues that are closely related to authoritarianism, particularly Republicans, who have stressed “wedge issues” such as race, culture, and crime to counter the Democrats’ traditional advantage on economic issues (Carmines and Stimson 1989; Hillygus and Shields 2008; Leege et al. 2002). As those issues have moved to the forefront of the agenda, authoritarianism is more important for partisan political behavior.

While this argument supports Hetherington and Weiler’s account of strategic politicians and worldview evolution, it also suggests that authoritarianism’s political impact may be inconsistent across elections. As Leege and his colleagues (2002) note, political campaigns are increasingly characterized by “the politics of cultural differences,” emphasizing conflicts over cultural and moral values. Such campaigns highlight “the conditions most threatening to oneness and sameness . . . lack of conformity to or consensus in group values, norms, and beliefs” (Stenner 2005, 17–18). They also activate nonauthoritarians’ commitment to diversity (Hetherington and Weiler 2009), thereby contributing to a larger electoral impact for authoritarianism. At the same time, the importance and effectiveness of appeals to cultural difference vary across campaigns (Leege et al. 2002). Cultural differences take center stage in some campaigns, as they seemed to in 2004, with its emphasis on terrorism, homeland security, and patriotism, and with the issue of gay marriage being quite important in some states (Abramson, Aldrich, and Rohde 2007; Campbell and Monson 2008). In other campaigns, economic performance, social welfare issues, or other matters not tied closely to authoritarianism take precedence, much as they did in 2008 when a severe economic downturn pushed issues of cultural values, terrorism, and immigration to the background (Conley 2009).³ This suggests that even while the partisan and electoral impact of authoritarianism is on the rise, that impact may well be variable across electoral contexts—for example, dipping in 2008 relative to 2004.

Data and Measures

To assess the influence of authoritarianism and parenting values on policy attitudes over time, we use data on the white respondents to all of the NES surveys with questions on parenting orientations—those in 1952, 1956, 1972, 1992, 2000, 2004, and 2008.⁴ For our measure of authoritarianism from 1992 to 2008, we use the same set of items used by Hetherington and Weiler (2009) and by others in the recent literature (e.g., Mockabee 2007; Stenner 2005). The specific wording of the question set is as follows:

Although there are a number of qualities that people feel that children should have, every person thinks that some are more important than others. I am going to read you pairs of desirable qualities. For each pair please tell me which one you think is more important for a child to have: Independence or respect for elders? Curiosity or good manners? Obedience or self-reliance? Being considerate or well behaved?⁵

We code responses as 0 for the nonauthoritarian (or nurturant) choices (independence, curiosity, self-reliance, and considerate), 1 for authoritarian (or strict) choices (respect for elders, good manners, obedience, and well behaved), and 0.5 for volunteered responses that “both” qualities are desirable.⁶ To measure authoritarianism,⁷ we take the mean of all of the valid responses on the items for each respondent.⁸

In the 1952, 1956, and 1972 NES surveys, there is only one question about child qualities or parental authority. However, the question taps directly into the strict-versus-nurturant parenting values that characterize theories of authoritarianism. The question asked respondents to react to the statement “What young people need most of all is strict discipline by their parents.” In 1952 and 1956, respondents were asked to place themselves in one of four categories ranging from “disagree quite a bit” to “agree quite a bit.” The 1972 NES only provided two choices: “agree” and “disagree.” Our measure of disciplinarian orientations in these three years is respondents’ levels of agreement with that statement.⁹ While it is less than ideal to use a different measure of authoritarianism in these earlier years, and is additionally unfortunate that the 1972 response choices differ from those in 1952 and 1956, we argue that the underlying construct in the earlier and later years is sufficiently similar to draw meaningful conclusions about the relationship between authoritarianism and political behavior over time.¹⁰ Still, we want to highlight this change in the measurement of authoritarianism over time, as subsequent conclusions should be interpreted with all due caution.¹¹

In assessing the impact of parenting values on political behavior and policy attitudes over the period from 1952 to 2008, we include a fairly limited range of control variables because the number of variables was much smaller in the early NES surveys than in recent surveys and we needed comparable models over time. Controls include education, income, age, gender, region (dummy variables for residents of the South, Midwest, and West), frequency of worship attendance, and dummies for Catholics and Jews.¹² Models of recent years include a much broader range of control variables.

Authoritarianism and Political Orientations from the 1950s through the 2000s

Partisanship and Presidential Vote Choice

We begin with models of the impact of authoritarianism on whites’ party identifications and presidential voting decisions from 1952 to 2008. Table 1 presents the effects of authoritarianism (coded to range from zero to one) in these models—in the form of either unstandardized coefficients from ordinary least squares (OLS) regression models of partisanship or changes in the predicted probability of voting Republican from logit models of the vote—and shows a clear growth in the political impact of authoritarianism over time.¹³ In the 1950s, authoritarianism has no discernible impact on either party identification or the vote, and the same is true for party identification in 1972. Authoritarianism does have a statistically significant and positive impact on the likelihood of voting for Richard Nixon in 1972. Given the cultural liberalism of Democratic candidate George McGovern and his active supporters (Kirkpatrick 1976; Layman 2001), this effect is in keeping with Hetherington and Weiler’s argument that authoritarianism’s political importance grows as the distance between parties and candidates on social and cultural issues increases.

Table 1.

The Effect of Authoritarianism/Strict Parenting Values on Party Identification and the Presidential Vote among Whites, 1952–2008.

Dependent variable	1952	1956	1972	1992	2000	2004	2008
Party identification^a	.06 (0.05)	−.24 (0.28)	.03 (0.02)	.09** (0.03)	.09 (0.05)	.24*** (0.05)	.14*** (0.05)
Presidential vote^b	.08 (0.35/0.35)	.02 (0.09/0.34)	.11** (0.49/0.15)	.29*** (1.21/0.27)	.31** (1.29/0.44)	.44*** (1.95/0.37)	.34*** (1.43/0.38)

Source. 1952–2008 American National Election Studies (whites only).

All party identification and vote models contained controls for worship attendance, education, income, age, gender, region (dummy variables for residents of the West, Midwest, and South), and religious affiliation (dummy variables for Catholics and Jews). All variables in the models range from 0 to 1.

Ranges from strong Democrat to strong Republican. Entries are unstandardized regression coefficients, with standard errors in parentheses.

Coded 1 for Republican voters and 0 for Democratic voters. Entries are the change in the probability of voting Republican for an increase from 0 to 1 (minimum to maximum value) in authoritarianism/strict parenting values while all of the control variables are held constant at their mean values. In parentheses is the ratio of the logit coefficient to its standard error.

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

In the 1990s and 2000s, the effect of authoritarianism on partisanship and vote choice is statistically significant in every year. This effect is consistently positive, producing more Republican Party attachments and a greater likelihood of voting Republican for president. The political influence of authoritarianism is also noticeably larger in more recent years. Even the statistically significant effect on the vote in 1972 is noticeably smaller than the effects in the 1990s and 2000s. Increasing authoritarianism from its minimum to maximum values (while holding all of the other independent variables at their mean values) is associated with an increase in the predicted probability of voting Republican of .10 in 1972 and from .29 to .45 over the 1992–2008 period.

In short, Barker and Tinnick and Hetherington and Weiler are correct that authoritarian or parenting orientations are consequential for contemporary American political behavior. They are also right that the polarization of American politics along authoritarian and nonauthoritarian lines has increased markedly in recent years. As expected, the peak of authoritarianism’s influence is 2004 and weakens in 2008.

Policy Attitudes

We turn next to the impact over time of authoritarianism on policy attitudes, estimated through models in which party identification and the sociodemographic variables included in the partisanship and presidential vote models are the controls. The dependent variables include attitudes toward the three major issue dimensions in post–World War II domestic politics: social welfare issues such as government’s role in ensuring citizen welfare and in providing social services, racial issues such as government protection of African American civil rights and affirmative action, and cultural issues such as abortion, school prayer, and lesbian, gay, bisexual, and transgender (LGBT) rights (Layman and Carsey 2002). Social welfare issues have been a central part of the agenda since the New Deal era and thus have appeared in NES surveys since the 1950s. Issues of race and civil rights were just starting to emerge on the national stage in the late-1950s and appear in the 1956 and all subsequent NES surveys. Cultural issues did not appear on the national agenda until the late-1960s and thus appear only in the surveys from 1972 to the present.¹⁴

In addition to these broad issue dimensions, we examine attitudes toward two domestic issues—immigration and the death penalty—that, like racial and cultural issues, are important components of the broader realm of “social issues” that have traditionally produced divisions between authoritarians and nonauthoritarians. Because these issues have become important in national politics only in recent decades, they appear only in the 1992–2008 surveys.¹⁵ We also assess the impact of authoritarianism on a number of indicators of defense and foreign policy attitudes. These include views on spending on national defense (included in all of the NES surveys since 1972), isolationism (measured in all of the NES surveys by how strongly respondents agree that the United States would be better off staying out of problems in other parts of the world), and opinions about five U.S. military actions: the Korean War in 1952, the Vietnam War in 1972, the Persian Gulf War in 1992, the war in Afghanistan in 2004, and the Iraq War in 2004 and 2008.¹⁶

Table 2 shows the models’ estimates of the effects of authoritarianism—in the form of unstandardized regression coefficients or changes in the predicted probability of taking the most conservative issue position from binary or ordered logit models—on the ten different dependent variables.¹⁷ As noted, we expect authoritarian tendencies to be consistently related to conservative attitudes on social issues broadly defined—including racial and civil rights issues, moral and cultural issues such as abortion and gay rights, immigration, and capital punishment—and to support for military strength and increases in defense spending. However, when it comes to social welfare issues, general views about American involvement in foreign affairs, and assessments of specific military conflicts, the effects of authoritarianism should be weaker, less consistent over time, and perhaps even in the opposite direction of positions typically associated with American conservatism and the Republican Party. The results support these expectations.

Table 2.

The Effect of Authoritarianism/Strict Parenting Values on Policy Attitudes among Whites, 1952–2008.

	Dependent variables
Year	Racial issues	Cultural issues	Defense spending	Immigration	Death penalty	Social welfare
1952	—	—	—	—	—	.002^a (0.17/0.29)
1956	.08* (0.05)	—	—	—	—	−.06 (0.04)
1972	.13*** (0.01)	.13*** (0.01)	.10**^b (0.42/0.18)	—	—	.07*** (0.02)
1992	.05** (0.02)	.21*** (0.02)	.08*** (0.02)	.13***^a (0.74/0.18)	.28***^a (1.22/0.20)	−.01 (0.02)
2000	.13** (0.04)	.21*** (0.03)	.05 (0.04)	.28***^a (1.52/0.32)	.38***^a (1.61/0.34)	−.01 (0.03)
2004	.14*** (0.04)	.18*** (0.03)	.09** (0.03)	.17***^a (1.29/0.29)	.38***^a (1.59/0.31)	.03 (0.03)
2008	.13*** (0.03)	.11** (0.04)	.17** (0.05)	.25***^a (1.73/0.30)	.37***^a (1.59/0.34)	−.09** (0.04)

Year	Foreign policy interventionism	Korean War	Vietnam War	Persian Gulf War	Afghan War	Iraq War

1952	−.01^a (−0.27/0.29)	−.15*^a (−0.61/0.32)	—	—	—	—
1956	−.06^a (−0.24/0.29)	—	—	—	—	—
1972	−.07**^b (−0.53/0.18)	—	.08*** (0.02)	—	—	—
1992	−.15***^b (−0.96/0.25)	—	—	.06*^a (0.48/0.27)	—	—
2000	−.22**^b (−1.33/0.42)	—	—	—	—	—
2004	−.17***^b (−1.47/0.44)	—	—	—	.02^b (0.14/0.41)	.05^b (0.22/0.39)
2008	−.23***^b (−1.22/0.37)	—	—	—	—	.04^b (0.19/0.41)

Source. 1952–2008 American National Election Studies (whites only).

All of the dependent variables are coded so that higher values indicate more conservative (or hawkish or interventionist) positions. Unless otherwise noted, the top entries in each cell are unstandardized ordinary least squares coefficients with standard errors in parentheses. The models for all issues included controls for party identification, worship attendance, education, income, age, gender, region, and religious affiliation. All variables in the models range from 0 to 1.

Because the dependent variable was ordinal, an ordered logit model was estimated. The top entry is the change in the probability of taking the most conservative/hawkish/interventionist position on the issue (i.e., the highest value of the ordinal variable) for an increase from 0 to 1 (minimum to maximum value) in authoritarianism/strict parenting values while all of the control variables are held constant at their mean values. In parentheses is the ratio of the ordered logit coefficient to its standard error.

Because the dependent variable was dichotomous, a binary logit model was estimated. The top entry is the change in the probability of taking the conservative/hawkish/interventionist position on the issue for an increase from 0 to 1 (minimum to maximum value) in authoritarianism/strict parenting values while all of the control variables are held constant at their mean values. In parentheses is the ratio of the logit coefficient to its standard error.

p < .10. **p < .05. ***p < .001.

In the first five sets of estimates in the top half of the table, we see disciplinarian orientations consistently and significantly producing greater social conservatism and support for increases in defense spending. Even as the content of the racial issue agenda has shifted over time—from a focus only on desegregation and civil rights for African Americans in the 1950s, to a combination of civil rights and government efforts to improve the welfare of blacks in the 1970s, and to a mix of those questions and affirmative action for African Americans in the 1990s and 2000s—authoritarianism has been associated unfailingly with opposition to government efforts on behalf of blacks and other racial minorities. Authoritarianism has positive and highly significant effects on conservative attitudes on abortion, women’s rights, and other cultural issues in 1972, 1992, and in the 2000s. It is also clearly related to opposition to increases in immigration and to support for the death penalty in every year of the 1992–2008 period. Finally, authoritarianism is significantly related to preferences for increased defense spending in 1972, 1992, 2004, and 2008, and remains positive and approaches significance in 2000.

The pattern of consistent and statistically significant relationships between authoritarianism and conservatism comes to an abrupt halt in the estimates for social welfare issues, shown in the last column of the top half of the table. The impact of disciplinarian orientations on social welfare attitudes is generally indistinguishable from zero. The effect is statistically significant in 1972 and 2008. However, in 1972, it is smaller than it typically is for the various types of social issues, and in 2008, it is negative, increasing support for liberal social welfare policies.¹⁸ Lakoff’s (1996, 179–80) claim that “liberals . . . see the federal government as a strong nurturant parent, responsible for making sure that the basic needs of its citizens are met,” while “conservatives . . . apply the Strict Father model [in which] . . . social programs amount to coddling people—spoiling them” does not withstand empirical evidence.

The negative effect that authoritarianism had on adherence to the tenets of modern conservatism in the case of social welfare in 2008 is seen more consistently in the estimates for support for U.S. involvement in world affairs. At each of the time points, authoritarianism is positively associated with isolationism. Moreover, in all of the years from 1972 to 2008, the effect of parenting outlooks is highly statistically significant. Thus, disciplinarian values are associated with a conservative perspective on U.S. involvement in world politics, but it is the isolationist perspective of the Republican Party in the 1950s, not the interventionism favored by modern conservatives. That authoritarianism is associated with support for military strength and spending, and with opposition to U.S. intervention into global problems points to an expectation of weak and inconsistent effects when it comes to opinions about the particular armed conflicts in which the United States has been involved over the past sixty years. That expectation is confirmed in the last five columns of Table 2, where we assess the effect of parenting values on support for the Korean War, Vietnam War, first Persian Gulf War, and the recent wars in Afghanistan and Iraq. In 1952, there is a statistically significant negative relationship between authoritarianism and support for an escalation of the Korean War. Authoritarians in 1952 actually were less likely to favor a bombing campaign in China and more likely to support U.S. withdrawal from the Korean peninsula.

In stark contrast, parental strictness in 1972 was significantly related to hawkishness on Vietnam, with authoritarians being more likely than nonauthoritarians to support policies that would lead the United States to pursue “everything necessary to win a complete military victory.” A possible explanation for this reversal in the effect between 1952 and 1972 is the long-term development of a linkage between authoritarianism and support for American military ventures. However, in keeping with our argument about the role of strategic political elites, we suspect that a better explanation is the way in which the issue of Vietnam was connected to domestic issues in 1972. By labeling Democratic presidential nominee George McGovern as the candidate of “acid, amnesty, and abortion,” the Republican Party and incumbent President Richard Nixon may have succeeded in connecting support for an expedited withdraw from Vietnam with liberal positions on salient cultural issues and with the countercultural tendencies of McGovern’s most active supporters (Kirkpatrick 1976). The hawkishness of authoritarians in 1972 may be more due to their negative reactions to antiwar protestors than from a steadfast belief in continued involvement in Vietnam.

The results for the three major military conflicts after 1972 point to further reservations about the idea of an emerging connection between authoritarianism and foreign policy hawkishness because the impact of parental strictness on attitudes toward all of them fails to meet standard levels of statistical significance. Authoritarianism’s positive connection to support for the first Persian Gulf War in 1992 is significant, but parenting values bear no relationship to support for either the Afghan War in 2004 or the Iraq War in 2004 and 2008. In sum, the consistently isolationist impulses of authoritarians together with parental strictness being clearly related to support for only one of the five major war efforts in which the United States has been involved over the past sixty years cast considerable doubt on Lakoff’s contention that a “Strict Father” mentality inevitably leads to preferences for an aggressive and interventionist American foreign policy.

A Closer Look at Authoritarianism’s Impact in 2004 and 2008

The results thus far support Hetherington and Weiler’s (2009) argument that political polarization along authoritarian and nonauthoritarian lines has grown because the policy issues to which this divide is most closely connected have become more politically important. The influence of strict parenting values has not broadened to include attitudes on issues to which those values were not related in the 1950s or in the early-1970s. Rather, authoritarianism’s effect on policy attitudes has been remarkably consistent over the past sixty years, always pushing citizens toward greater support for the modern Republican Party’s conservative positions on the variety of “social” issues and on military and defense spending, but very rarely boosting support for the contemporary Republican Party’s conservative positions on social welfare or for its promotion of aggressive military intervention in foreign affairs. Thus, the growing impact of authoritarianism on Americans’ electoral decisions likely has resulted from the rising political salience of the policy attitudes for which it is most relevant.

Of course, if the mounting electoral influence of authoritarianism is contingent on changes over time in the political importance of particular policy agendas, then that influence should vary from one election to the next as the salience of those policy agendas waxes and wanes with changing societal circumstances and presidential campaign strategies. In particular, the impact of authoritarian orientations should have been greater in 2004 than in 2008. The 2004 election fell in the aftermath of the 2001 terrorist attacks and included the presence of antigay marriage initiatives on many statewide ballots. As a result, the Bush campaign emphasized the President’s execution of the war on terror and mobilizing evangelical voters through culturally conservative appeals. This made issues such as terrorism, national defense, and gay rights more salient than in 2008, when the economic downturn and the focus of the Obama campaign on the economic failures of the Bush administration highlighted social welfare concerns.

To assess the degree to which authoritarianism is connected to electoral decisions by attitudes on social issues, defense spending, social welfare, and foreign policy, and whether the electoral impact of authoritarianism was stronger in 2004 than in 2008, we estimate a path model of presidential candidate choice with data from the 2004 and 2008 NES surveys. As we illustrate in Figure 1, our model follows Miller and Shanks’s (1996, 192) identification of various “causal stages” in explaining voting behavior by separating the influences on citizens’ electoral decisions into five distinct causal stages. We allow all of the variables in earlier causal stages to exert direct effects on variables in later stages.

Figure 1.

Causal stages in the path model of political orientations.

Working backward from the final stage of the model, our ultimate dependent variable is the individual’s assessment of the choice between the two major-party presidential candidates. Allowing us to estimate the path model with OLS regression and thus compute the direct and indirect effects of authoritarianism and other variables in a straightforward way, we operationalize candidate choice as the difference between the feeling thermometer ratings of the Republican and Democratic presidential candidates, a variable that is highly correlated with the presidential vote and has been used as a proxy for the vote in numerous efforts at causal modeling (Page and Jones 1979).¹⁹

Attitudes toward a wide variety of policy issues exert the most direct influence on comparative candidate ratings. These include cultural issue attitudes, social welfare attitudes, views on racial issues, opinions about immigration and illegal immigration, attitudes on gun control and the death penalty,²⁰ support for increases in defense spending, and a “foreign policy interventionism” variable that combines respondents’ answers on the NES’ traditional isolationism question with assessments of whether the “war in Iraq has been worth the cost.” Directly affecting policy attitudes are party identification and ideological identification, the two variables identified most consistently in the literature as shaping evaluations of political issues and candidates in fundamental ways (Campbell et al. 1960; Miller and Shanks 1996). Just prior to partisanship and ideology are “core values” that exert strong predisposing effects on citizens’ views of political issues, parties, and candidates (Feldman 1988; Goren 2004). These are egalitarianism (coded here as antiegalitarianism), support for a limited scope of government, and moral traditionalism (combining individuals’ levels of moral absolutism and intolerance of people with different moral perspectives).²¹

The first stage of the model consists of what we term “foundational social characteristics and values,” and includes authoritarianism and basic sociodemographic characteristics—age, gender, education, income, and region of residence (a dummy variable for residents of the South)—as well as several indicators of the religious divide that increasingly has come to structure partisanship and vote choice (Green 2007; McTague and Layman 2009). The religious indicators are level of religious commitment (the combination of frequency of worship attendance, frequency of prayer, and amount of guidance received from religion), doctrinal orthodoxy (measured through the NES’ question about the authority of Scripture), and dummy variables for evangelical Protestants, Catholics, Jews, and people with no religious affiliation.²²

Our decision to place authoritarianism at the initial causal stage along with sociodemographic and religious traits and prior to core political values and predispositions was not an obvious one. That placement is supported by work that conceptualizes authoritarianism as an innate personality trait, and thus as causally prior to value orientations such as egalitarianism and moral traditionalism (Adorno et al. 1950; Stenner 2005). However, research contending that authoritarianism is caused by childhood, religious, or class-based socialization (Altemeyer 1988; Lipset 1959) suggests that it may be inappropriate to place it at the same causal stage as religious and socioeconomic characteristics and causally prior to other socialized value orientations. Because we argue that authoritarianism’s impact on electoral behavior may be more limited in particular election years than the extant literature suggests, placing it prior to other value orientations and allowing it to exert indirect effects on political identities and evaluations through those values strikes us as a conservative decision. In a similar vein, it seems appropriate for the assessment at hand given that Barker and Tinnick (2006) allow parenting values to affect core values and predispositions in their own causal model.²³

Table 3 presents the estimates of our model. In the first three columns for each year, we show the direct, indirect, and total effects of authoritarianism and religious commitment—the most powerful of the variables representing the contemporary religious divide and thus providing an apt comparison for authoritarianism’s overall political influence—on core values, partisan and ideological identification, and policy attitudes. In the next three columns, we show the direct, indirect, and total effects of all of the variables in the first four causal stages of our model (with the exception of all of the variables other than authoritarianism and religious commitment in the first stage) on comparative candidate evaluations in 2004 and 2008.

Table 3.

Path Models of Candidate Choice in 2004 and 2008: The Effects of Authoritarianism on Core Values, Political Identifications, and Policy Attitudes and of All Variables on Candidate Evaluations.

	2004						2008
	Authoritarianism’s effects			Effects of all variables on candidate evaluations			Authoritarianism’s effects			Effects of all variables on candidate evaluations
	Direct	Indirect	Total	Direct	Indirect	Total	Direct	Indirect	Total	Direct	Indirect	Total
Authoritarianism	—	—	—	.07	.13	.20	—	—	—	0	.09	.09
Religious commitment	—	—	—	0	.09	.09	—	—	—	0	.19	.19
Limited government	0	—	0	0	.10	.10	−.10	—	−.10	0	.22	.22
Antiegalitarianism	.10	—	.10	0	.23	.23	.08	—	.08	.17	.30	.47
Moral traditionalism	.17	—	.17	.07	.13	.20	0	—	0	0	.15	.15
Party identification	.11	.07	.18	.37	.12	.49	.08	.001	.08	.39	.06	.45
Ideological identification	.08	.06	.14	0	.05	.05	.11	.001	.08	.21	.12	.33
Cultural issues	.12	.06	.18	.09	—	.09	.07	.04	.11	.15	—	.15
Social welfare issues	0	.07	.07	.08	—	.08	0	.04	.04	.14	—	.14
Racial issues	.08	.08	.16	0	—	0	.12	.02	.14	0	—	0
Immigration issues	.11	0	.11	0	—	0	.17	.02	.19	0	—	0
Guns/death penalty	.12	.07	.19	.08	—	.08	.07	.02	.09	0	—	0
Defense spending	0	.07	.07	.11	—	.11	.11	.03	.14	.14	—	.14
Interventionism	−.11	.09	−.02	.17	—	.17	−.17	.04	−.13	.14	—	.14

Source. 2004 and 2008 American National Election Studies (whites only).

Entries are unstandardized ordinary least squares path coefficients. All effects that were not statistically significant at p < .10 are listed as 0 for direct effects and were not used in computing indirect and total effects. The core values were regressed on authoritarianism, religious commitment, view of the Bible, religious affiliation, and demographic variables. Party identification and ideological identification were regressed on the core values, authoritarianism, religious commitment, and all of the control variables. All of the policy attitudes were regressed on party ID, ideological ID, the core values, authoritarianism, religious commitment, and all of the control variables. Candidate evaluations were regressed on all of the other endogenous variables, authoritarianism, religious commitment, and all of the control variables. All variables were coded to range from 0 to 1. All of the variables shown in the table range from the most liberal/Democratic/progressive position to the most conservative/Republican/traditionalist position.

For 2004, our results generally reinforce those we already have presented and the broader analysis that Hetherington and Weiler present of authoritarianism’s political influence in 2004. Authoritarianism clearly mattered for political orientations in 2004. It had statistically significant direct effects on party identification, ideological identification, and presidential candidate evaluations, and its total effects on these variables were substantial. Its total effect on candidate evaluations was particularly impressive, clearly surpassed only by that of party identification and easily surpassing those of ideology, support for limited government, and any of the policy attitudes other than foreign interventionism. However, authoritarianism’s path to political influence was not all-encompassing. Just as our longitudinal analysis of policy attitudes suggests, authoritarianism mattered in 2004 largely because it shaped attitudes toward social issues, broadly defined, and the core values that are connected to those attitudes, not because it also came to structure social welfare and foreign policy perspectives. In terms of core values, parental strictness was significantly related to both antiegalitarianism and moral traditionalism. However, it bears no relationship to support for limited government, not surprisingly given the authoritarian preference for strong and centralized leadership. Authoritarianism was significantly related—directly, indirectly, or (in most cases) both—to conservatism on the whole host of social issues. It encouraged moral and cultural conservatism, opposition to liberal racial policies, immigration, and gun control, and support for the death penalty. Meanwhile, its effect on social welfare attitudes was only indirect and bore a direct negative relationship to foreign policy interventionism.

The pattern of authoritarianism’s political influence in 2008 was quite similar. Although it had no effect on moral traditionalism, it was related to antiegalitarianism, pushed citizens toward greater conservatism on all of the social issue dimensions, and had direct and indirect effects on defense spending attitudes. Meanwhile, the effect of parental strictness on support for limited government was negative, its direct and total effects on foreign policy interventionism were even more negative than in 2004, and its influence on social welfare attitudes was virtually nil. However, the size of authoritarianism’s political influence was noticeably smaller in 2008 than in 2004. Its total effects on party identification and ideology were smaller, about half of the 2004 effect in the case of partisanship. Authoritarianism’s total effect on comparative candidate evaluations in 2008 was less than half of its effect in 2004, and the effects of other variables on candidate evaluations in 2008 provide some clues as to why. With the federal government bailout of the financial industry and with Obama calling for a larger government role in health care, support for limited government and social welfare attitudes had noticeably larger effects in 2008 than in 2004. Those, of course, are orientations on which authoritarianism had either a negative influence or virtually no influence. Meanwhile, the effect of foreign policy interventionism on candidate choice in 2008 was similar to its effect in 2004. However, after four more years of American involvement in an unpopular war in Iraq and in a global war on terror, the connection between authoritarianism and interventionism had grown even more clearly negative. In short, just as the context of the 2004 election helped to make authoritarianism quite relevant, the context of the 2008 election placed a damper on that political relevance.

Conclusion

Our analysis of NES data from 1952 to 2008 strongly supports the arguments by Barker and Tinnick (2006), Hetherington and Weiler (2009), and others that authoritarianism has become an increasingly important influence on American political behavior, particularly in its impact on party identification and vote choice. However, authoritarianism’s influence on public policy attitudes has remained extraordinarily consistent over time. Strict parenting values have nearly always been connected to conservative attitudes on “social” issues such as culture, race, immigration, guns, and crime policy, and support for military strength and increases in defense spending. However, they have rarely been connected to social welfare issues or support for a limited scope of government, they have been consistently linked to isolationism in foreign policy, and their effects on opinions about the various military conflicts in which the United States has been involved since World War II have been weak and inconsistent.

This supports Hetherington and Weiler’s argument that there has been a “worldview evolution” by which the issues to which authoritarianism has always been connected have grown more politically important and thus linked the divide between authoritarians and nonauthoritarians more closely to political behavior. Conversely, our findings cast doubt on the view of Barker and Tinnick (2006) and Lakoff (2002) that parenting values have become linked to the whole range of policy issues and thus serve as an important basis for attitudinal constraint across those issues.

Thus, our analysis sheds light on the primary reason why political scientists have turned their attention back to authoritarianism: the search for explanations of growing political polarization. Our findings support the idea that the most important explanation for growing polarization is politics itself, or, more specifically, the political goals, strategies, identities, and attachments of party politicians, political activists, and mass partisan identifiers (e.g., Cox and McCubbins 2005; Layman et al. 2010; Lee 2009). Authoritarianism’s political importance appears to have grown not primarily because of factors external to politics, such as the Dr. Spock-inspired revolution in parenting styles that Lakoff (2002) identifies as central to current polarization, but due to citizens’ responses to explicitly political developments. The political parties, their candidates, and their most active supporters have devoted more space on the political agenda to issues such as race, culture, immigration, and national security and they have taken increasingly divergent positions on those issues. Citizens have responded by tying their levels of authoritarianism or nonauthoritarianism, which have always been central to their views on these issues, more closely to their party attachments and voting decisions.

At the same time, our evidence suggests that the political influence of authoritarianism is contingent on the context of elections. It is likely to be larger in elections when societal circumstances and candidates’ campaign strategies encourage a focus on issues of national defense and security or on social and cultural issues. It is likely to be smaller when the issues at the center of campaigns and public concern are economic and social welfare issues or issues surrounding military intervention by the United States into foreign conflicts. Our 2004 and 2008 results show such conditional effects. The greater focus on the economy and social welfare concerns in 2008, as well as the public’s growing weariness over American military involvement on multiple foreign fronts seemed to make authoritarianism less important in 2008 than it had been in 2004.

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.

Notes

References

Abramson

Paul R.

Aldrich

John H.

Rohde.

David W.

2007. Change and Continuity in the 2004 and 2006 Elections. Washington, DC: Congressional Quarterly Press.

Adorno

Theodor W.

Frenkel-Brunswik

Else

Levinson

Daniel J.

Sanford.

R. Nevitt

1950. The Authoritarian Personality. New York: Harper & Row.

Altemeyer

Robert

. 1988. Enemies of Freedom: Understanding Right-Wing Authoritarianism. San Francisco: Jossey-Bass.

Altemeyer

Robert

. 1996. The Authoritarian Specter. Cambridge, MA: Harvard University Press.

Barker

David C.

Tinnick

James D.

III . 2006. “Competing Visions of Parental Roles and Ideological Constraint.” American Political Science Review 100 (2): 249–63.

Campbell

Angus

Converse

Philip

Miller

Warren

Stokes

Donald

. 1960. The American Voter. Chicago: University of Chicago Press.

Campbell

David E.

Monson.

J. Quin

2008. “The Religion Card: Gay Marriage and the 2004 Presidential Election.” Public Opinion Quarterly 72 (3): 399–419.

Carmines

Edward G.

Stimson

James A.

1989. Issue Evolution: Race and the Transformation of American Politics. Princeton: Princeton University Press.

Conley

Patricia

. 2009. “A Mandate for Change? Decisive Victory in a Time of Crisis.” In Winning the Presidency, 2008, edited by Crotty

William J.

, 169–84. Boulder, CO: Paradigm.

10.

Cox

Gary W.

McCubbins

Mathew Daniel

. 2005. Setting the Agenda: Responsible Party Government in the US House of Representatives. New York: Cambridge University Press.

11.

Edsall

Thomas

Edsall

Mary D.

1992. Chain Reaction: The Impact of Race, Rights, and Taxes on American Politics. New York: W.W. Norton.

12.

Feldman

Stanley

. 1988. “Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values.” American Journal of Political Science 32:416–40.

13.

Feldman

Stanley

. 2003. “Enforcing Social Conformity: A Theory of Authoritarianism.” Political Psychology 24 (1): 41–74.

14.

Feldman

Stanley

Stenner

Karen

. 1997. “Perceived Threat and Authoritarianism.” Political Psychology 18 (4): 741–70.

15.

Gilens

Martin

. 1996. “‘Race Coding’ and White Opposition to Welfare.” American Political Science Review 90:593–604.

16.

Goren

Paul

. 2004. “Political Sophistication and Policy Reasoning: A Reconsideration.” American Journal of Political Science 48 (3): 462–79.

17.

Goren

Paul

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

18.

Green

John C.

2007. The Faith Factor: How Religion Influences the Vote. Westport: Praeger Press.

19.

Hetherington

Marc J.

Weiler.

Jonathan D.

2009. Authoritarianism and Polarization in American Politics. New York: Cambridge University Press.

20.

Hillygus

Sunshine D.

Shields.

Todd G.

2008. The Persuadable Voter: Wedge Issues in Presidential Campaigns. Princeton: Princeton University Press.

21.

Huddy

Leonie

Feldman

Stanley

Charles

Taber

Lahav

Gallya

. 2005. “Threat, Anxiety, and Support of Antiterrorism Policies.” American Journal of Political Science 49 (3): 593–608.

22.

Janowitz

Morris

Marvick

Dwaine

. 1953. “Authoritarianism and Political Behavior.” Public Opinion Quarterly 17 (2): 185–201.

23.

Kirkpatrick

Jeane

. 1976. The New Presidential Elite. New York: Russell Sage.

24.

Lakoff

George

. 1996. Moral Politics: What Conservatives Know That Liberals Don’t. Chicago: University of Chicago Press.

25.

Lakoff

George

. 2002. Moral Politics: How Liberals and Conservative Think. Chicago: University of Chicago Press.

26.

Lane

Robert E.

1955. “Political Personality and Electoral Choice.” American Political Science Review 49 (1): 173–90.

27.

Lavine

Howard

Lodge

Milton

Freitas

Kate

. 2005. “Threat, Authoritarianism, and Selective Exposure to Information.” Political Psychology 26 (2): 219–44.

28.

Layman

Geoffrey C.

2001. The Great Divide. New York: Columbia University Press.

29.

Layman

Geoffrey C.

Carsey.

Thomas M.

2002. “Party Polarization and ‘Conflict Extension’ in the American Electorate.” American Journal of Political Science 46:786–802.

30.

Layman

Geoffrey C.

Carsey

Thomas M.

Green

John C.

Richard

Herrera

Rosalyn

Cooperman

. 2010. “Activists and Conflict Extension in American Party Politics.” American Political Science Review 104 (2): 324–46.

31.

Lee

Frances E.

2009. Beyond Ideology: Politics, Principles, and Partisanship in the US Senate. Chicago: University of Chicago Press.

32.

Leege

David C.

Wald

Kenneth D.

Krueger

Brian S.

Mueller.

Paul D.

2002. The Politics of Cultural Difference: Social Change and Voter Mobilization Strategies in the Post-New Deal Period. Princeton: Princeton University Press.

33.

Lipset

Seymour Martin

. 1959. “Democracy and Working-Class Authoritarianism.” American Psychological Review 24:482–501.

34.

McTague Michael

John

Layman.

Geoffrey C.

2009. “Religion, Parties, and Voting Behavior: A Political Explanation of Religious Influence.” In The Oxford Handbook of Religion and American Politics, edited by Smidt

Corwin E.

Kellstedt

Lyman A.

Guth

James L.

330–70. New York: Oxford University Press.

35.

Miller

Warren E.

Shanks.

J. Merrill

1996. The New American Voter. Cambridge: Harvard University Press.

36.

Mockabee

Stephen T.

2007. “A Question of Authority: Religion and Cultural Conflict in the 2004 Election.” Political Behavior 29 (2): 221–48.

37.

Page

Benjamin I.

Jones

Calvin C.

1979. “Reciprocal Effects of Policy Preferences, Party Loyalties and the Vote.” American Political Science Review 73:1071–90.

38.

Stenner

Karen

. 2005. The Authoritarian Dynamic. New York: 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

0.40 MB