Beyond liberal and conservative: Two-dimensional conceptions of ideology and the structure of political attitudes and values

Transcription

1 Journal of Political Ideologies (June 2008), 13(2), Beyond liberal and conservative: Two-dimensional conceptions of ideology and the structure of political attitudes and values BRENDON SWEDLOW Department of Political Science, Northern Illinois University, DeKalb, IL, USA ABSTRACT This article analyzes conceptual similarities and differences in selected prior work on ideological multi-dimensionality and finds substantial conceptual convergence accompanied by some provocative divergence. The article also finds that evidence from a recent survey of the American public largely validates areas of conceptual convergence. Respondents political attitudes vary in two dimensions that are associated with different value structures specifically, with different rankings of liberty, order, and caring for those who need help. As a result, liberals, conservatives, and libertarians are identified more fully than previously possible. But the evidence does not allow the validation of one conception over another in the area where they diverge most markedly. Does the fourth ideological type value order and equality, making them communitarians, or are these respondents better understood as a humanitarian, paternalistic, hierarchical subtype, the inclusive social hierarch, since they value order as well as caring for those who need help? Although politicians, philosophers, and social scientists often discuss politics as if it were organized on a single left-right dimension, 50 years of research on public opinion shows that a unidimensional model of ideology is a poor description of political attitudes for the overwhelming proportion of people virtually everywhere. 1 Introduction 2 Students of US electoral behavior noticed that the liberal-conservative debate had expanded beyond the economic issues of New Deal politics to encompass or be displaced by a new kind of social/moral issue at least as early as the late 1960s and early 1970s. 3 These scholars argued that conflicts over government intervention in ISSN print; ISSN online/08/ q 2008 Taylor & Francis DOI: /

2 brendon swedlow the economy had expanded to include or were superseded by conflicts over government intervention in social issues such as civil rights, abortion, and marijuana smoking. The best known of the general findings on issue dimensionality is that of separate, not necessarily correlated, social and economic dimensions, observes Kathleen Knight in reviewing US public opinion research on ideology. Indeed, the American public at large may deserve to be described in this fashion. 4 More recently, scholars have argued that liberal-conservative debate has appeared unidimensional because it tends to occur in either the economic or social dimension but not both at any given time. These students of public opinion and political ideology trace this oscillating conflict back to Still others have argued that issue and ideological differences are additive, expanding to three dimensions social welfare, racial, and cultural in recent times in the mass public, while remaining one dimensional for the mostly partisan, reflecting similar unidimensional structuring of ideological conflict among political elites. 6 Further complicating analysis and confounding understanding, other students of public opinion and ideology have provided alternative specifications of the relevant ideological dimensions. For some, the most important dimensions are defined by attitudes toward government intervention in the economy and personal freedom. 7 Others propose that different views of capitalism and democracy 8 or differing support for the political values of liberty and equality 9 or equality and order 10 define dimensions generating four ideological types. Still others think that the most important dimensions are related to the extent of individual autonomy and collective action in social relations and are accompanied by functionally related ideologies. 11 Finally, there are students of public opinion and ideology who, rather than focusing on ideological multi-dimensionality, study the relationship between value structures, including different rankings of liberty, equality, and order, and left-right, liberal-conservative ideology. 12 Many other scholars have studied the relationship between political values taken one at a time and political attitudes and/or unidimensional ideology. 13 None of these proponents of different two-dimensional conceptions of ideological organization and/or value structure have compared their conceptions to those of the other scholars cited, much less tried to arbitrate among contending conceptions through empirical research. This article seeks to identify and where possible reduce apparent conceptual variation by analyzing conceptual similarities and differences in selected prior work on ideological multi-dimensionality, beginning with the most well-known US attempt to move beyond liberal and conservative, William Maddox and Stuart Lilie s study of the same name. 14 It then discusses Kenneth Janda, Jeffrey Berry, and Jerry Goldman s efforts to develop Maddox and Lilie s concepts. 15 The article next introduces a much less well-known to Americans (but more familiar to Europeans) account of why political ideologies, values, and attitudes should exist in two dimensions, the cultural theory of Mary Douglas and Aaron Wildavsky. 16 The article finds substantial convergence among these contending conceptions accompanied by some provocative divergence. Most interestingly, Janda and colleagues suggest the existence of a communitarian ideological type that simultaneously holds 158

3 beyond liberal and conservative equality and order in high regard. By contrast, most other scholars, including Douglas and Wildavsky, think that liberty, equality, and order must be traded off against each other, which would preclude the existence of communitarians. 17 Thus, this article will pay particular attention to the role values and value-structure play in two-dimensional conceptions of ideology. This article also seeks to evaluate these contending conceptions of ideological multi-dimensionality to the extent possible with available evidence. The article finds that a recent survey of the American public largely validates areas of conceptual convergence. Respondents political attitudes vary in two dimensions that are associated with different value structures specifically, with different rankings of liberty, order, and caring for those who need help. 18 As a result, liberals, conservatives, and libertarians are identified more fully than previously possible. But the evidence does not allow the validation of one conception over another in the area where they diverge most markedly. Does the fourth ideological type value order and equality, making them communitarians, or are these respondents better understood as a humanitarian, paternalistic, hierarchical subtype, the inclusive social hierarch, since they value order as well as caring for those who need help? Implications of the existence of this fourth ideological subtype for partisan conflict and for coalition and institution-building are discussed before concluding with suggestions for future research to reconcile unidimensional conceptions of ideology with evidence of ideological multi-dimensionality. Throughout, the relationship among values, attitudes, and ideology will be investigated and analyzed, both as conceptualized and measured, beginning with a discussion of some important distinctions and hypothesized relationships. Ideology, values, value structure, and political attitudes in public opinion This article looks to Milton Rokeach s ground-breaking work on values and value structure 19 as developed more recently by Shalom Schwartz for definitions of key concepts and as a jumping off point for analyzing two-dimensional, value-based accounts of ideology. 20 Schwartz thinks that values (1) are concepts or beliefs, (2) pertain to desirable end states or behaviors, (3) transcend specific situations, (4) guide selection or evaluation of behavior and events, and (5) are ordered by relative importance. Values... differ from attitudes primarily in their generality or abstractness (feature 3) and in their hierarchical ordering of importance (feature 5). 21 In short, values are the criteria people use to select and justify actions and to evaluate people (including the self) and events. 22 Or, as William Jacoby puts it, [values] effectively define what is good and bad in the world. 23 The potential value-added to studies of ideology by focusing on the value-basis and value-structure rather than unidimensional ideological structure of political attitudes is well-described by Stanley Feldman and the scholars he cites. 24 As he observes, public opinion research shows that people have fewer values than political attitudes and more values than can be explained by left right ideological commitments. Thus [values] could provide a basis for reducing the complexity 159

4 brendon swedlow of political judgments and for creating consistency among attitudes while at the same time potentially addressing the need for a two or more-dimensional conceptualization of ideology. Moreover, [if] there is indeed an organization to the values people hold, this may provide an even simpler structure for political attitudes and an underlying basis for political ideology. 25 As we will see, values and value structure figure prominently in the conceptions and/or measures of ideological two-dimensionality considered here. Going beyond liberal and conservative with Maddox, Lilie, and Janda In Beyond Liberal and Conservative, Maddox and Lilie argue that the single liberal-conservative dichotomy and the resulting two-way analysis of American politics is inadequate for understanding belief systems or ideologies in the United States. While [most] observers of public opinion recognize that economic and social (or personal freedom) issues are two distinct dimensions, Maddox and Lilie believe that previous researchers have made a fundamental, consistent error: They have ascribed the terms liberal and conservative to the individual s place on the issue dimension, which leads them simultaneously to remain with a single-continuum definition of belief systems and to find that people are inconsistent. The two terms, however, refer to belief systems; that is, what a person thinks about several types of issues... One does not have a liberal position on any one issue, such as government health insurance, for example; one simply supports that policy to some degree. The terms liberal and conservative refer only to a total belief system, not a position on any one issue or even a cluster of issues. 26 Maddox and Lilie distinguish between ideology at the philosophical level and at the level of mass belief systems. At the philosophical level, an ideology involves the elaboration of a worldview and of desired processes of political change to reach desired goals or values. The purpose of an ideology may be to provide a guide to action, to persuade others, to give legitimacy to a set of social structures, to engender passive acceptance of a set of socio-political arrangements, or some mix of these purposes. 27 Maddox and Lilie acknowledge that The masses of ordinary citizens seldom articulate an ideology, if ideology is defined solely in these terms, but they may have shared sets of consistent attitudes that, at least in a latent sense, relate to established traditions of political thought. 28 Maddox and Lilie go on to discuss various European political theorists influences on the development of American liberal, conservative, and libertarian political ideologies, acknowledging that their fourth political ideology, populism, has been associated more with protest and political action than with theoretical writings. 29 In Maddox and Lilie s two-dimensional conception of ideology, attitudes toward government intervention in the economy and attitudes toward the maintenance or expansion of personal freedoms are held in different configurations by different people. Liberals and conservatives have opposite attitudes on both dimensions. Liberals support government economic intervention and expansion of personal freedoms; conservatives oppose both. But we have to go beyond liberal and conservative to discover those who support one position and oppose the other. 160

5 Figure 1. beyond liberal and conservative Maddox and Lilie s attitudinal dimensions and ideological types. Libertarians support expanded individual freedom but oppose government economic intervention; populists oppose expansion of individual freedom but support government intervention in the economy 30 (see Figure 1). To validate their two dimensional conception of ideology, Maddox and Lilie analyze national survey data compiled by the Center for Political Studies (CPS) at the University of Michigan from 1952 to 1980 and National Opinion Research Center (NORC) survey data from the 1970s. Some of the CPS data are panel data. Maddox and Lilie measure attitudes toward government intervention in the economy through questions about government involvement in health care and government aid for citizens who need jobs... [and] when necessary [through questions about] the general extent of government power, government involvement in the power and housing industries, the progressive taxation system, and whether governmental services in general should be curtailed. In the 1950s and 1960s, they measure attitudes toward the maintenance or expansion of personal freedoms through questions regarding civil rights and in the 1970s through questions regarding the legalization of marijuana, women s rights, and abortion. 31 David Boaz and David Kirby have attempted to replicate and extend Maddox and Lilie s analysis from 1990 to 2006 using national survey data collected by the American National Election Studies (ANES) for , the Gallup Poll (Governance Surveys, ), and the Pew Research Center (Political Typology Survey, 2004). 32 Boaz and Kirby do not distinguish between value and attitudinal measures, 33 but by Schwartz s definition they sought to measure the extent to which the public valued government interventions in the economy with two PEW questions that asked respondents to choose between these alternative beliefs:. Government is almost always wasteful and inefficient; OR, Government often does a better job than people give it credit for. 161

6 brendon swedlow. Government regulation of businesss is necessary to protect the public interest; OR, Government regulation of business usually does more harm than good. Similarly, to measure the extent of public support for government interventions in societal morality, Boaz and Kirby used this measure asking respondents to indicate their agreement with one of these beliefs:. The government should do more to protect morality in society; OR, I worry the government is getting too involved in the issue of morality. 34 Boaz, Kirby, and Pew sought to measure attitudes toward government interventions in the economy through questions on government health insurance, government regulation, and private retirement accounts for Social Security. To measure attitudes on social/moral issues, they used questions on gay marriage, banning books in school libraries, and government promotion of morality. Boaz and Kirby used value-based measures similar to those developed by Pew to tap Maddox and Lilie s ideological dimensions in the ANES and Gallup Poll data. For example, Boaz and Kirby turned to this ANES question forcing a choice between beliefs to measure support for government economic interventions:. ONE, We need a strong government to handle today s complex economic problems; or, TWO, The free market can handle these problems without government being involved. Similarly, to measure the extent of public support for interventions in societal morality, Boaz and Kirby used this measure asking respondents to indicate the extent of their agreement or disagreement with this statement of values:. We should be more tolerant of people who choose to live according to their own moral standards, even if they are very different from our own. To help identify libertarians, they also used this question forcing a choice between beliefs about government, which does not distinguish economic from social/moral government interventions:. ONE, The less government the better; or TWO, There are more things that government should be doing. 35 Values are also central to Kenneth Janda and colleagues two-dimensional conception of ideology. Drawing on Milton Rokeach s research on political values, they re-conceive the dimensions in Maddox and Lilie s analysis as valuebased, while largely retaining their labels for the resultant ideological types. 36 Starting from a reference point that approximates a Lockean state of nature, a hypothetical condition in which freedom is theoretically absolute, Janda and colleagues argue that people have the option of trading off increments of freedom 162

7 beyond liberal and conservative to achieve either greater social equality or greater social order (or both) through government interventions. In their two-dimensional, value-based scheme, Libertarians choose freedom over both order and equality. Communitarians are willing to sacrifice freedom for both order and equality. Liberals value freedom more than order and equality more than freedom. Conservatives value order more than freedom and freedom more than equality. 37 These rankings are graphically portrayed in Figure 2. A communitarian, they say, quoting the Oxford English Dictionary (1989), is a member of a community formed to put into practice communistic or socialistic theories. Citing studies of philosophical similarities and differences between communitarianism and socialism, they claim that [communitarians] favor government programs that promote both order and equality, in keeping with socialist theory. 38 As a contemporary American example of communitarianism (representative of but narrower than socialism), they offer the Communitarian Network, a political movement founded by sociologist Amitai Etzioni. The Communitarian Network rejects both the liberal-conservative classification and the libertarian argument that individuals should be left on their own to pursue their choices, rights, and self-interests. Like liberals, Etzioni s communitarians believe that there is a role for government in helping the disadvantaged. Like conservatives, they believe that government should be used to promote moral values preserving the family through more stringent divorce laws, protecting against AIDS through testing programs, and limiting the dissemination of pornography, for example. Indeed, some observers have labeled George W. Bush as a communitarian. 39 Like Maddox and Lilie, Janda and colleagues turn to CPS data (from 2000) to validate their two dimensional, value-based conception of ideology. 40 A question about attitudes toward abortion gave respondents four possible scenarios. The first three options called for various degrees of governmental restriction of abortion. Figure 2. Janda, Berry, and Goldman s value dimensions and ideological types. 163

8 brendon swedlow The fourth option was By law, a woman should be able to obtain an abortion as a matter of personal choice. Janda and colleagues used responses to this question to sort those valuing freedom from those valuing order. Meanwhile, Janda and colleagues used answers to this question to sort those valuing freedom from those valuing equality: Some people feel the government in Washington should see to it that every person has a job and a good standard of living... Others think the government should just let each person get ahead on his own... Where would you put yourself...? In Janda and colleagues classification:. liberals thought women should have the right to abort and that government should guarantee jobs, while. conservatives thought government should restrict the right to abort and should let each person get ahead on his own, while. libertarians thought women should have the right to abort and that government should let each person get ahead on his own, while. communitarians thought government should restrict the right to abort and should guarantee jobs. Janda and Goldman have also created a 20-question ideological self-test in a further effort to validate their two-dimensional ideological scheme. The test that can be taken by anyone and has been taken by college students in over 750 classes since These questions seek to elicit attitudes on political issues and are used to place respondents in Janda and colleagues two-dimensional ideological space. Respondents are asked to express attitudes on gay marriage, abortion, school prayer, school vouchers, gun control, health care, social security, the death penalty, taxes, welfare, civil liberties, equal rights, affirmative action, government size and services, marijuana, and sex and violence on television. Going beyond liberal and conservative with Douglas and Wildavsky If, as Samuel Huntington and others have argued, Americans are basically antiauthority and anti-government, 42 why do liberals and conservatives both support government spending and regulatory interventions for certain purposes? The lack of adequate answers to questions like this led Aaron Wildavsky to conclude that Efforts to read the left-right distinction into American history... succeed only in making a hash of it, are equally inapplicable to... the present and deserve to be discarded. 43 Wildavsky also wondered where libertarianism fit into the liberalconservative continuum and he was unsatisfied with largely historical and atheoretical explanations of how ideologies develop, maintain themselves, and change. In the anthropology of Mary Douglas, Wildavsky found his answer: 44 Instead of considering a single American culture, I propose that analyzing American political life in terms of conflicts among at least three political cultures hierarchical, individualistic, and egalitarian will prove more satisfactory. This approach will generate fewer surprises and provide explanations that better fit the phenomena

9 beyond liberal and conservative Douglas developed the concept of cultural bias to characterize beliefs and values that were functional for particular patterns of social relations. Douglas and her students and collaborators have further developed her ideas in a number of directions, including as an explanatory theory of political cultures. 46 In this theory, the values of liberty, equality, and order are hypothesized to be functional for different patterns of social and institutional relations, as are specific beliefs about human and physical nature. 47 Douglas and Wildavsky theorize that the covariance among unique configurations of social relations and political values constituting political cultures is generated by their functional interdependence. Institutions and ideologies, social relations and values, cannot just be mixed and matched in any combination. Viable institutions and patterns of social relations depend on ideologies and values that rationalize and support those relationships. Hierarchical organizations could not long persist if their members valued liberty or equality more than order, for example. In their political cultural theory, Douglas and Wildavsky construe the extent of individual autonomy and collective action in society as independent dimensions rather than as poles on a continuum, as is customary. 48 This conceptual shift allows analysts to account for four rather than two patterns of social relations. People in individualistic and fatalistic social relations are not part of a collective undertaking, but individualists retain their autonomy, while fatalists do not. People in egalitarian and hierarchical social relations, meanwhile, are part of a collective undertaking, but egalitarians retain much more of their autonomy than hierarchs. Hierarchical social relations are highly structured, with everyone and everything having his, her, and its place. Individualistic social relations, by contrast, are highly fluid, the product of individual choice. People in egalitarian social relations retain their autonomy by giving all members an equal voice in and thus the power to veto collective decisions, popularly known as consensus decision-making. Fatalistic social relations, meanwhile, are tenuous and unreliable, driven by the whim and caprice of others. Hierarchs value order, individualists value liberty, egalitarians value equality, and fatalists value (good) luck (see Figure 3). To illustrate, consider how these political cultural types may be used to understand attitudes toward government, including the US federal government and its budgeting process. Government represents a type of collective action, which leaves fatalists indifferent and repels individualists, but attracts egalitarians and hierarchs. The US federal government, and its budgetary process in particular, has characteristics of centralization, expertise, and comprehensiveness that make it the natural home of hierarchical statists and technocrats. Egalitarians also like the collective, communitarian, public qualities of government, but they do not like centralization, large bureaucracies, or decision-making driven by experts. Thus, they have reasons to be attracted to federal budgeting and reasons to dislike it. Individualists, by contrast, generally prefer a small public and a large private sphere, where they can move about freely, and have the greatest scope to arrange their lives as they choose. 49 Douglas, Wildavsky, and others have attempted to validate their two dimensional conception of political culture, including its ideological components, through both case studies and survey research. 50 In survey research, the cultural 165

10 brendon swedlow Figure 3. Douglas and Wildavsky s dimensions of social relations and political cultural types. bias or ideology hypothesized to be associated with each political cultural type has been measured most commonly by the extent to which respondents agree or disagree with these statements: For individualistic ideology:. Competitive markets are almost always the best way to supply people with the things they need.. Society would be better off if there was much less government regulation of business.. People who are successful in business have a right to enjoy their wealth as they see fit. For hierarchical ideology:. One of the problems with people today is that they challenge authority too often.. Society works best when people strictly obey all rules and regulations.. Respect for authority is one of the most important things that children should learn. For egalitarian ideology:. The world would be a more peaceful place if its wealth were divided more equally among nations.. We need to dramatically reduce inequalities between the rich and the poor, whites and people of color, and men and women.. What our country needs is a fairness revolution to make the distribution of goods more equal

11 beyond liberal and conservative Comparing Maddox, Lilie, and Janda to Douglas and Wildavsky on dimensions, types, and measures of political values and attitudes To what extent are these two-dimensional conceptions of ideology similar? To what extent do they differ? The same questions can be asked about the ideological types generated by these dimensions and the value and attitudinal measures used to identify them in the public. To what extent are these alike or different? These are the questions this section seeks to address. Ideological dimensions Maddox and Lilie s dimension tapping attitudes toward the maintenance or expansion of personal freedom may be viewed as the attitudinal counterpart to Douglas and Wildavsky s behavioral dimension measuring the extent of autonomy in social relations. Likewise, Maddox and Lilie s dimension measuring attitudes toward government intervention in the economy is to attitudes largely what Douglas and Wildavsky s dimension tapping the extent of collectivization in social relations is to behavior. However, Maddox and Lilie s dimensions cover less ground than Douglas and Wildavsky s. Maddox and Lilie s dimensions reflect attitudes toward governmental regulation of freedom and the economy, whereas Douglas and Wildavsky s dimensions can be used to assess the extent of individual autonomy and collective action in both public and private spheres. The most significant difference between these approaches becomes visible in Janda and colleagues reworking of Maddox and Lilie. In Douglas and Wildavsky s conceptualization of dimensions and their associated values, people are predicted to tradeoff order against equality and freedom, not to hold two or more of these values in high regard simultaneously. This would appear to rule out Janda and colleagues communitarians, who value both order and equality. However, as explained below, Douglas and Wildavsky do allow for an hierarchical sub-type, the inclusive hierarch, that may approximate the value and attitudinal structure of communitarians. This possibility will be discussed further after presenting evidence of ideological and value structure in American public opinion bearing on this issue. Liberals and egalitarians Maddox and Lilie s liberals correspond closely to Douglas and Wildavsky s egalitarians: combining high levels of support for collective or governmental action with high levels of support for individual freedom or autonomy. But in addition to liberal governmental intervention in the economy, egalitarians will support governmental intervention to reduce limitations on liberty provided that increased liberty furthers equality of outcome or result; if it does not, liberty itself may be attacked. 52 Egalitarians will support governmental action in any sphere economic, political, or social to liberate the oppressed and to increase equality of result and participation in government. Douglas and Wildavsky s egalitarians thus 167

12 brendon swedlow have a broader agenda than Maddox and Lilie s liberals. Meanwhile, Janda and colleagues liberals are not the same as Maddox and Lilie s. In their preference for equality of condition over liberty if these come into conflict, Janda and colleagues liberals more closely resemble Douglas and Wildavsky s egalitarians. Libertarians and individualists Maddox and Lilie s libertarians also correspond closely to Douglas and Wildavsky s individualists: supporting expanded personal freedoms and opposing government interventions in the economy. As operationalized in many interpretive applications and particularly in much survey research (as seen above), however, Douglas and Wildavsky s conception of individualism is frequently narrower than Maddox and Lilie s conception of libertarianism, focusing more on economic freedoms than civil liberties, or what Maddox and Lilie call personal freedoms. As we have seen, Maddox and Lilie used indicators in both areas to identify libertarians, with support for civil liberties measured by attitudes toward government regulation of marijuana, pornography, and abortion, among other things. But conceptually libertarians are fairly similar for these scholars, and for Janda and colleagues: Libertarians are vocal advocates of hands-off government, in both the social and economic spheres. 53 Conservatives and hierarchs Douglas and Wildavsky s hierarchs partially overlap Maddox and Lilie s conservatives, who oppose both government intervention in the economy and the expansion of personal freedoms. Hierarchs of all sorts value order and so would be expected to agree on restricting the expansion of personal freedoms. In addition to social hierarchs, who may include some law and order types and/or religious and/or social conservatives, there are nationalistic, statist types of hierarchs, who may include some bureaucrats and scientific rationalists who value expertise. However, these latter hierarchs may also support rather than oppose interventions in the economy, consequently sharing significant common ground with populists or communitarians, while coming into conflict with conservatives as defined by Maddox and Lilie and Janda and colleagues. 54 Douglas and Wildavsky s conception of hierarchy spans all of these types without distinguishing them, to the detriment of analyses relying on the concept of hierarchy. 55 In survey research, as we have seen, Douglas and Wildavsky s conception of hierarchy is operationalized to measure support for traditional institutions and obedience to rules and regulations. Populists, communitarians, and fatalists Populism, conceived by Maddox and Lilie as opposition to expansion of individual freedom coupled with support for government intervention in the economy, has no direct counterpart in the political cultural types generated by 168

13 beyond liberal and conservative Douglas and Wildavsky s scheme. Populism seeks to enlist an interventionist, nationalistic state on behalf of traditional, culturally conservative, small town, rural folks to level the concentrations of power and wealth found in big government and big business. 56 Populism, in this rendering, spans some of the hierarchical types just described and even seems to draw on egalitarian concerns. 57 Populists seem similar to what Wildavsky calls inclusive (rather than exclusive) social hierarchs because they want to inculcate traditional morality but still help the less fortunate. 58 Janda and colleagues communitarians, however, may not be assimilated so easily to Douglas and Wildavsky s framework. As discussed, communitarians are said to value highly both order and equality, while for Douglas and Wildavsky these values must be traded off against each other. Douglas and Wildavsky s fatalistic political culture also has no counterpart in Maddox and Lilie s or Janda and colleagues schemes. 59 Measures of political values and attitudes Another important area to assess similarities and differences in these accounts of ideology is how proponents have operationalized their conceptions in survey research. Maddox and Lilie and Janda and colleagues rely primarily on attitudinal measures, specifically on measures of issue or policy attitudes, to validate their two-dimensional conceptions of ideologies. The survey questions they use seek to elicit respondents attitudes toward government involvement in health care, creation of jobs, gun control, regulation of abortion, school prayer, and the like. Maddox and Lilie demonstrate that these attitudes vary in two dimensions and they infer that more general attitudes toward government intervention in the economy and in personal freedoms underlie and organize these specific issue or policy attitudes. Janda and colleagues take the same approach, but infer that the more general underlying attitudes are different valuations of liberty, equality, and order. In contrast, those who have operationalized Douglas and Wildavsky s cultural biases or ideologies have done so through value-based measures, which seek to elicit values more directly. The survey questions on which they rely ask how different distributions and control of wealth are valued, not about attitudes toward progressive taxes versus flat taxes, as Janda and Goldman do, for example. Meanwhile, Boaz and Kirby, relying on the Pew surveys, are the only proponents of a two-dimensional ideological scheme to use measures of both attitudes and values to attempt to validate their conception of ideology. However, even though they use value measures, and even though they seek to replicate and extend Maddox and Lilie s analysis, they do not re-conceptualize the dimensions underlying these attitudes as value-based in the way that Janda and colleagues do. Nor do Boaz and Kirby attempt to investigate whether and how the value measures and results are related to the attitudinal measures and results. It is also important to assess whether these studies are attempting to measure the same values, since liberty, equality, and order can have a variety of meanings. Maddox and Lilie and Boaz and Kirby do not discuss this issue directly and 169

14 brendon swedlow Douglas and Wildavksy only discuss it indirectly, but the values at stake in their analyses can be inferred from the attitudinal and value measures they use, and these values are largely the same ones that Janda and colleagues discuss and seek to measure. Janda and colleagues make the familiar distinction between freedom of and freedom from. Freedom of is the absence of constraints on behavior; it means freedom to do something. In this sense, freedom is synonymous with liberty, which is how they use the term. In contrast, Freedom from... comes close to the concept of equality. 60 Here, Janda and colleagues make the usual distinction between equality of opportunity and equality of outcome. Equality of opportunity means that each person has the same chance to succeed in life. Equality of outcome, on the other hand (which is how Janda and colleagues define equality), requires government to design policies that redistribute wealth and status so that economic and social equality are actually achieved. 61 Janda and colleagues also distinguish governmental efforts to preserve order in a familiar way: in the narrow sense of preserving life and protecting property [and]... in the broader sense of preserving the social order... [i.e. preserving] established patterns of authority in society and traditional modes of behavior... Our use of the term order... includes all three aspects. 62 Douglas, Wildavsky, and their students and collaborators define the values of liberty, equality, and order in largely the same way as Janda and colleagues do. These definitions are often implicit in how the terms are used rather than explicitly developed, as this passage illustrates: The arguments for or against environmental and land use regulation... can be categorized according to three basic beliefs: A well-ordered society is desirable and is promoted by regulatory controls; regulation violates the basic freedoms of the individual; and greater equality is desirable, and regulation should be used to redistribute resources and power in society. 63 As previously discussed, for Douglas, Wildavsky, and their students and collaborators, valuing order means valuing hierarchical order, often found in traditional institutions, but also found in bureaucracy. As we have seen, in survey research relying on their political cultural theory valuing order is measured by support for authority and obedience to rules and regulations. These scholars also share the view that valuing freedom means valuing behavioral freedom to do things. As Wildavsky puts it, Individualistic cultures seek self-regulation as substitute for authority. They prefer minimum authority, just enough to maintain rules for transaction As operationalized in survey research, however, valuing liberty is measured by opposition to regulation of business and support for markets as economic freedom alone even though their concept of freedom also extends to valuing civil liberties. 65 Finally, valuing equality is operationalized as valuing equality of outcome or results, as in the survey questions above, which ask about support for redistributing wealth and reducing inequalities between classes, races, and genders. Egalitarians value equality of outcome above other forms of equality, while individualists value equality of opportunity and hierarchs value 170

15 beyond liberal and conservative procedural equality (i.e. equality before the law) over equality of outcome. As Wildavsky notes, Between equality of opportunity, so that individuals can accentuate their differences, and equality of results, so that individuals can diminish their differences, there is a vast gulf. 66 In the next three sections, this article considers how well these often similar, sometimes divergent conceptions of ideological multi-dimensionality, value structure, and the ideal types they generate correspond to the ideological and value structure found in the American public. The analysis begins by summarizing evidence from related survey research demonstrating that political attitudes vary in two dimensions that are associated with different value structures specifically, with different rankings of liberty, order, and caring. The article then discusses the extent to which this evidence validates these varied conceptions, including, particularly, the communitarian ideological type. Evidence of ideological differences in two dimensions of political attitudes and values As noted in the introduction, to the extent two-dimensional ideological structure is acknowledged and discussed, reference is most frequently made to separate economic and social dimensions in issue attitudes. These dimensions are often couched in terms of regulation: economic regulation versus social regulation. Liberals are said to support economic regulation but oppose social regulation, while conservatives have the opposite preferences: they oppose economic regulation but support social regulation. Because these dimensions are essentially restatements of the attitudinal dimensions used by Maddox and Lilie, with social regulation varying inversely with support for the expansion of personal freedoms, and because of the greater familiarity most people have with these dimensional concepts, my co-author, Mikel Wyckoff, and I adopt the terms economic and social regulation in our analysis of ideological differences in two Figure 4. Swedlow and Wyckoff s dimensions and ideological types. 171

16 brendon swedlow dimensions of political attitudes and values. 67 Following Janda and colleagues, we also re-label Maddox and Lilie s populists as communitarians. Libertarians oppose and communitarians support both economic and social regulation (see Figure 4). The evidence summarized here is taken from our analysis of the results of a national telephone survey of registered voters (N ¼ 805) conducted March 30 April 2, 2003 by Voter Consumer Research in Washington, D.C. The survey was commissioned by the Competitive Enterprise Institute and designed by National Media, Inc. 68 Among other things, the survey sought to build on prior survey research operationalizing Douglas and Wildavsky s conception of political culture, as discussed above. New to this survey are questions pairing values and forcing choices among liberty, order, and caring for those who need help and items measuring attentiveness to and participation in politics and feelings toward and knowledge of political figures, groups, and institutions. Some agree/disagree scales measuring policy preferences included in the survey and analyzed here replicate prior work by Wildavsky and others, but most of these items are also new. My co-author and I selected eight questions about political attitudes to operationalize the dimensions of economic and social regulation. 69 The political attitudes are not all policy or issue attitudes, but also include attitudes toward reference groups. The economic dimension measures support for governmental interventions in the economy, particularly to protect the vulnerable. Liberals support (and conservatives oppose) government economic interventions:. to regulate business,. to provide health insurance,. prescription drug coverage, and. to reduce inequalities between whites and minorities. The social dimension measures support for social/moral regulation, in part by measuring affect for non-governmental organizations that help regulate society/morality. Conservatives support (and liberals oppose):. prayer in school, and. positively assess Evangelical Christians,. the Boy Scouts, and. the US Military (while liberals do not). Libertarians oppose these government economic interventions and prayer in school and negatively assess these reference groups, while communitarians support the economic interventions and prayer in school and positively assess these groups. The dimensions of economic and social regulation emerged very strongly in the structure of political attitudes on these eight survey items. 70 A large number of people in this sample are consistently liberal or conservative with regard to economic and social/moral issues. Another large group of respondents exhibit equally impressive attitudinal consistency but in a less conventional way they are consistently libertarian or communitarian in outlook. 172

17 beyond liberal and conservative Survey participants were also given six opportunities to weigh the relative merits of freedom, order, and caring. Three questions asked respondents to rank the personal importance of protecting the freedom of the individual, maintaining order and stability in society, and caring for those who need help, with these values considered one pair at a time. 71 Three additional questions asked respondents to choose the most important statement of three alternatives, where the three alternatives sought to tap values related to those just listed. These three questions and their three alternatives follow: Question 4. Which one of these three statements do you think is most important? Children must be taught to respect their parents and proper authorities. Children must be taught to work hard and make it on their own. Children must be taught to care about and be fair to others. Question 5. Which one of these three statements do you think is most important? People should have more respect for the traditional way of doing things. People should be allowed to do things their own way. People should be required to give up doing some things, if that might help the public good. Question 6. Which one of these three statements do you think is most important? Government should maintain an orderly society. Government should stay out of individuals lives. Government should provide for those who cannot provide for themselves. These six questions were used to create multiple item indices reflecting each respondent s relative preference for the three values. With these measures in hand, we were able to examine the value preferences associated with liberal, conservative, libertarian, and communitarian attitudinal structures. Results from this analysis are presented in Table 1. Table 1. Average value preferences by attitudinal structure. Value preferences Attitudinal Structure Order Freedom Caring Liberal Conservative Libertarian Communitarian Sig. level (F-test) Note: Cell entries are average value scores for each attitudinal structure category. 173

18 brendon swedlow Conservatives value order above all else (and more than persons in the other three groups) and they are noticeably unconcerned with caring for those who need help. Liberals, in turn, place a premium on caring and substantially discount the value of order. Communitarians value freedom less than members of the other three groups, and they value caring almost as much as their liberal counterparts. Communitarians do not seem to value order as much as Janda and colleagues conceptual framework might suggest, but as expected they value order more than liberals do. Finally, the average preference for freedom by libertarians is probably not as high as one might expect, but libertarians are shown to place greater emphasis on freedom than on caring and order, and their average freedom score is higher than that observed for the other three groups. Discussion: findings and implications Douglas, Wildavsky, and others 72 who have researched value structure claim that liberty, equality, and order are traded off against each other. These authors argue and find that these values cannot be maximized together. More of one means less of the others. We found that liberals value caring, conservatives value order, and libertarians value liberty more than the other ideological types do (where liberals, conservatives, and libertarians were identified by their political attitudes). This would appear consistent with these authors claims as well as with Janda and colleagues claims for these three ideological types. So far, so good. Moreover, this evidence provides stronger support for these claims than has been previously developed by these and other authors, such as Boaz and Kirby, who have either not used both value and attitudinal measures, or have not demonstrated that they relate to each other in expected ways. But Janda and colleagues also claim that their fourth ideological type, communitarians, value both order and equality highly, trading liberty to get them. Do our findings also validate this claim? Respondents with communitarian political attitudes supporting both economic and social regulation also seem to have something approximating communitarian values, valuing both order and caring. But from the perspective of Douglas, Wildavsky, and others, valuing both order and equality is logically contradictory, if by order we mean hierarchical order, which Wildavsky calls the institutionalization of inequality. In William Jacoby s terms, populists or communitarians are ranking values intransitively and his findings suggest such people should only be a small part of the public instead of the substantial portion we found. 73 In Douglas and Wildavsky s conceptualization, Janda and colleagues communitarians do not have values that are institutionally viable because their values are contradictory and at war with themselves. In their view and that of Jacoby and others, it is not possible to value equality and order at the same time, if by order we mean hierarchical order. According to Douglas and Wildavsky, people with communitarian political values and attitudes will not be able to construct institutions that reflect their values. 174

19 beyond liberal and conservative Alternatively, perhaps the political and institutional consequences of holding communitarian views are found not within specific institutions but between them. It could be that communitarians are the glue of much ideological bipartisanship, helping to support liberal and conservative coalitions and the construction of institutions that combine their attitudes and values. For Wildavsky, the institutionalization of an egalitarian and hierarchical alliance resembling communitarianism is found in a political regime exemplified by European social democracies. 74 Or perhaps communitarians help fuel the war between liberals and conservatives and between egalitarians and hierarchs, by supporting the political parties and institution-building of these ideological foes. Our analysis indicates that communitarians more than the other ideological types vote for both parties equally. Challenges operationalizing our dimensions with measures of political attitudes and values indicate that another explanation may also be possible. The survey on which we relied did not ask respondents whether they valued equality over order. Rather, respondents were asked whether they considered caring for those who need help more important than maintaining order and stability in society. While order is frequently seen as a core value of conservatives and is a value that different kinds of hierarchs share, valuing order is not necessarily the same thing as valuing an hierarchical order. From one perspective, order is a rather innocuous, generic term. All of Douglas and Wildavsky s political cultural types are, after all, social orders of one kind or another. Order, consequently, is a value that could draw support from all political cultural types and so confound supporters of equality with supporters of inequality, among others. However, as we have seen, support for hierarchical values was also measured with questions about respecting tradition, parents, and proper authorities, values that are more strongly associated with the social conservative subtype of social hierarch. Moreover, order and caring were not equally valued by communitarians; order was less valued than caring, suggesting that communitarians share greater value affinities with liberals or egalitarians than with conservatives or social hierarchs. Furthermore, and perhaps more importantly, valuing caring for those who need help is not the same thing as valuing equality. While those who value equality probably chose caring over the available alternatives, they are not the only political cultural type that might highly value caring for those who need help. As discussed above, what Wildavsky calls inclusive social hierarchs seem to be like communitarians: both types value hierarchical order but also value caring. This might be called humanitarianism, paternalism, or President Bush s compassionate conservatism. Consequently, it is not clear that we have validated Janda s conceptualization of communitarians as people who highly value both hierarchical (unequal) order and equality. Perhaps we have only identified and provided evidence of an hierarchical subtype, the inclusive social hierarch. This may be the conclusion best supported by our evidence, which validates Douglas and Wildavsky s claim that there are inclusive social hierarchs (who we have been calling communitarians) and exclusive social hierarchs (conservatives). However, this interpretation of our findings becomes more difficult to sustain when we examine the political attitudes rather than values of respondents we have 175