The Very Basis of Reasons: Groups, Social Identities, and Political Psychology

What, then, is a “group”? How does it acquire the capacity for exercising such a decisive influence over the mental life of the individual?
— Sigmund Freud, Group Psychology and the Analysis of the Ego (1921, 6)

 

In the preceding parts of this book, we took up two broad classes of democratic theory. The first class we called “populist,” conveying the notion that the political preferences of ordinary people should be the foundation of good government. In the simplest folk theory version, those preferences are taken at face value, and the majority should rule. In more sophisticated doctrines, preferences might need to be enlightened by education or deliberation. But in all versions of this style of democratic thinking, preferences are the starting point and the foundation. Hence this category of democratic theory is variously called “liberal” (in the 19th- century sense deriving from utilitarianism), or sometimes “individualist” or “intellectualist,” for it is founded on and relies on careful thought by individuals in their role as democratic citizens (Mannheim 1936, 219– 225; de Grazia 1951, 172– 175). As we have seen, this is a role that people struggle to fulfill in everyday political life. The attentive, judicious, unprejudiced individuals attuned to the common good that populate the folk theory appear only too rarely in real life, and the barriers to erecting a satisfactory democratic theory on that foundation are formidable indeed.

 

The second class of theories we have considered focus less on policy choices and more on good management. In a Schumpeterian vein, they look for tools by which citizens with modest information can control the actions of their leaders. This retrospective theory of voting once seemed to offer great promise as a democratic theory responsive to the findings of modern social science. But as we have seen, this theory, too, asks too much of ordinary people. They cannot meet its demands, and the result is capricious judgments at election time.

 

Then what remains? In this chapter we outline an alternative view of political psychology. We argue that it provides a more scientifically accurate and politically realistic foundation for democratic theory. In one sense, not much of this chapter will be new. Indeed, we will argue that most of it was better understood by the Founders, and certainly by the social scientists of a century ago, than it is today. How was it lost? Alas, folk- democratic thinking and myopic retrospection afflict political science, too. As we will show, those forces have proven stronger than our strongest theories. Particularly when the popular research tools of the day have fit bad ideas better than good ones, we have readily exchanged our powerful intellectual inheritance for a mess of folk thinking.

 

The Realist Political Tradition in the West

 

Realist thinking about politics can be found in the Hebrew Bible, in Aristotle and Thucydides, and in many other sources from the earliest years of Western thought. Its application to democracy appears prominently in the Federalist Papers. With respect to political psychology, here is Madison in Federalist Number 10:

 

A zeal for different opinions concerning religion, concerning government, and many other points, as well of speculation as of practice; an attachment to different leaders ambitiously contending for pre- eminence and power; or to persons of other descriptions whose for tunes have been interesting to the human passions, have, in turn, di vided mankind into parties, inflamed them with mutual animosity, and rendered them much more disposed to vex and oppress each other than to co- operate for the common good. So strong is this propensity of mankind to fall into animosities that where no substantial occasion presents itself the most frivolous and fanciful distinctions have been sufficient to kindle their unfriendly passions and excite their most violent conflicts.

 

As we will show, that passage prefigures most of the key psychological ideas of the group theory of politics— the powerful tendency of people to form groups, the ensuing construction of “us” and “them,” and the powerful role of emotion rather than reason in directing group activity. Madison also anticipated the experimental finding of 20th- century psychologists that group attachments are easily generated and profoundly felt. People are naturally group- oriented.

 

The subsequent theoretical development of group theory derives from a great achievement of 19th- and early 20th- century social science— the explicit recognition as a foundational principle that human beings everywhere live in groups and that human thought is deeply conditioned by culture, including group subcultures. Of course, national cultures had long been recognized: the Israelites were not like the Canaanites, the Athenians were not like the Spartans, and none of them resembled the Romans. Then, too, within civil society, 18th- century writers had already pointed out the profusion of groups, each with their demands on government. For example, decades before Madison, the Scottish Enlightenment figure Adam Ferguson (1767) emphasized the importance of “civil society” for thinking about government.

 

The group- theoretic strand of 19th- century thought went further, however, breaking from the rationalistic liberalism that underpins the folk theory of democracy. Karl Marx saw that while entire societies possessed a culture, economic classes within a society generated subcultures, too, and that those subcultures influenced what members of each class believed, what they valued, and what actions they were willing to take in defense of their class. Other scholars soon extended the argument to competing, conflictual civil groups of all kinds— national, racial, ethnic, religious, and professional. In Europe, Gabriel de Tarde (1890) emphasized individuals’ widespread imitation of others, not just in manners and dress, but in their ideas as well. Gustave Le Bon (1895) pointed to the ubiquity of groups and their frequently irrational psychology. Georg Simmel (1908) focused on the inevitability of group conflict, and he identified overlapping and crosscutting group memberships as possible exacerbating or moderating influences. In the United States, Edward Ross (1905) and especially Albion Small (1905) brought this European tradition to the attention of their fellow citizens, emphasizing the central role of groups in sociological theory, the limitations of individual rationality, and the omnipresence of group conflict. And before any of these, the great Polish sociologist Ludwig Gumplowicz (1885) pioneered most of the central ideas.

 

For all these scholars, mental life was group life. Human beings are cultural animals, and they spend their years absorbing in great detail the various subcultures in which they participate. From this viewpoint, all the great issues of life— religion, nationality, gender roles, popular scientific ideas, partisan loyalties, the value of different occupations, the appeal of different foods— are taught to the vast majority of people by their family, their culture, and their subcultures. At a slower pace, socialization to group norms continues through out adulthood. Indeed, it is a commonplace that what passes for “going your own way” or “thinking for yourself ” turns out to be, in practice, just switching from one set of culturally conditioned ideas to another, equally preformulated set. “Between the cradle and the grave . . . , a chain is extended whose every link is wrought into the preceding. Man may choose to break it by violence, but not to re- weld it” (Gumplowicz 1885, 160).

 

By the early 20th century, then, a powerful realist tradition in social science had arisen in opposition to the rationalistic Enlightenment assumptions of liberalism. It is that realist tradition that we adopt and extend in the remainder of this book. We argue that liberalism is simply too shallow to bear the weight put on it in conventional democratic theorizing. Put another way, for thinking about democracy, rational choice liberalism is a scientific error.

 

This is not to say, of course, that 19th- century liberal ideas have no uses. When people choose between receiving more or less money for similar tasks, or between cheap or expensive food of the same quality, the liberal assumption of rational choice will work well. Much of economics is a highly elaborated structure for dealing with situations of that kind. But, as Joseph Schumpeter (1942, 262– 263) observed, once one leaves economics for politics, the choices take on a completely different character. Most people have little or no direct experience with the complexities of politics, and their thinking is far from the folk- democratic ideal, as Graham Wallas (1908) pointed out long ago. In consequence, when economics is carted into political science without attention to the evidence in the new field, as with the spatial model of elections, those foreign ideas will perform very poorly empirically, as we saw in chapter 2.3 Much the same logic applies to the theory of deliberative democracy, which is an idealistic version of liberalism proposed by political theorists (Habermas 1994; Gutmann and Thompson 1996; Macedo 1999; Cohen 2003).

 

The limitations of rationalistic liberalism were clear to early 20th- century political scientists, and yet the implications for democracy and for its study were only slowly realized. For the early sociological theorists had spoken to politics only in part. They were intent on showing that states arose from violent conquest, not from some imaginary social agreement. And they argued that domestic social conflict arose as soon as states were formed. But the work of government— the policy- making process— got little attention from sociologists.

 

American political scientists of the 19th century were no different. They also ignored the role of group forces in government in spite of the omnipresence of “the lobby,” the powerful collection of interest groups that so appalled observers of 19th- century American national and state politics. To take just three classic examples of this oversight, in volumes that are still in print, John Burgess (1933, 145) complained only briefly at the end of his important book that “everywhere, a designing class or group” deflects government to “the profit and privilege of the few.” Woodrow Wilson’s 333- page classic, Congressional Government (1900, 189– 190), dedicated just a page and a half to interest groups, and then only to condemn their influence in general terms. And James Bryce’s thousand- page second volume of The American Commonwealth gave just one chapter to an interest group— the railroads— a chapter that, he said, “No one will expect to find in a book like this” (Bryce 1894, 643).

 

It was left to Arthur Bentley, a journalist who had seen the rough and tumble of Chicago politics close up and who had taught at the University of Chicago for part of one year, to lay out the detailed implications for politics of the new ideas in social theory. Bentley had studied in Europe, and he explicitly acknowledged his deep debt to the Continental European sociologists who had preceded him. For him as for them, groups were fundamental to politics. Yet Bentley pushed well past those who came before him.

 

In The Process of Government, Bentley (1908) emphasized the importance of groups in politics— all sorts of groups, including organized interest groups, self- conscious but politically unorganized social groups, potential groups not yet self- conscious, and governmental actors. He saw most governmental decisions as decisively influenced by the power and intensity of rival groups. He also described the ability of intense minorities to get their way when the majority is apathetic, uninterested, or unorganized. The strategic choice of techniques by pressure groups, the legal and normative rules structuring group conflict, the role of parties as group coalitions, and the impact of government as both referee and independent actor— all these are discussed. Indeed, there is not much of theoretical import in subsequent group theories of the policy- making process that is not anticipated in Bentley’s book. Though not an easy read, it is a monument of creative thought, in our estimation the most important book of 20th- century political science.

 

In chapter 11 we return to the policy- making process, which was Bentley’s central focus. For now, our key point is that he portrayed groups not just as vehicles for pressuring government, but as the mainsprings of political attitudes and interests. “When we go down to the group statement,” he wrote (1908, 241), “we get down below mere reasoning to the very basis of reasons.” However, he had little to say about the processes by which groups shape individuals’ political opinions. He took the previous sociological treatises, which he discussed at considerable length, as having established the profound impact of groups on individuals, and he moved on.

 

For the next few decades, other scholars of public opinion followed Bentley’s lead. They oriented themselves to group influences on public opinion, but paid little attention to the underlying causal mechanisms. For example, W. Brooke Graves’s (1928) reader on public opinion took an explicitly group- theoretic approach. It contained selections from Freud, A. Lawrence Lowell, and Walter Lippmann. Other readings explained how the schools, religion, the press and radio, the arts, business, political parties, and pressure groups each affected citizens’ thinking and their political preferences. Conceptually speaking, this was a more sophisticated book than virtually all modern readers on public opinion, which pay little attention to group attempts to influence the citizenry or to the processes by which people absorb their ideas from the subcultures to which they belong.

 

The arguments of the public opinion specialists of this period, like the analyses of the pioneering sociologists who preceded them, were primarily humanistic, illustrated by colorful, even compelling examples from the politics of the time, it is true, but with none of the inferential rigor that social scientists began to expect under the growing influence of empiricism. The theoretical arguments were quite general and undifferentiated, too. Considerable scope was left for subsequent scholars to fill in the mechanisms, using close observation and laboratory experiments to expand and validate group theory in a more systematic way. And that they proceeded to do.

 

In monumental multivolume studies of “Middletown” by Robert and Helen Lynd (1929; 1937) and of “Yankee City” by W. Lloyd Warner and his associates (1963), ethnic, gender, and religious memberships were studied in great detail, with attention to each of the corresponding subcultures, and to “training the young” into their roles. Participant observation in these towns (Muncie, Indiana, and Newburyport, Massachusetts, respectively) confirmed just what the 19th- century sociologists had said: groups are fundamental to social life. People took their views from the groups to which they belonged, often because the people around them made it difficult not to do so. Ideas that individuals had picked up elsewhere were disapproved of, and most people were eventually convinced to discard them. One Middletown attorney told the Lynds, “Why, when I was in the university I believed all the professors told me. . . . Now I realize it was downright wrong for them to talk to us the way they did” (Lynd and Lynd 1929, 493).

 

These findings regarding the power of groups to shape thinking have been reinforced by the results of multiple studies in the decades since. For example, the impact of families and schools in the first years of life has been extensively documented. The work of Herbert Hyman (1959), Fred Greenstein (1965), Robert Hess and Judith Torney (1968), Kenneth Langton (1969), and M. Kent Jennings and Richard Niemi (1974; 1981), among others, established empirically the powerful effects of early socialization on adult political allegiances. The effects were not immutable, but neither were they easily changed. Most adults felt the same ties to nation, religion, race, ethnicity, and political party as their parents.

 

Researchers also took to the laboratory to demonstrate the power of group norms to override individuals’ own judgments. Muzafer Sherif (1936, chap. 6) and Solomon Asch (1951) carried out famous experiments showing that even ad hoc groups of strangers who announced a faulty factual judgment could induce people to go along. For example, Asch (1951) asked a group of previously unacquainted male undergraduates to judge which of three line segments was closest in length to a fourth reference line. Left to themselves, more than 99% of students gave the correct answer. But in groups, the outcome was quite different. Unbeknownst to the experimental subjects, the other group members were confederates of the experimenter, instructed to give wrong answers at some points in a sequence of trials. The confederates were seated so that they spoke first, leaving the remaining student to either give the correct answer in defiance of everyone else, or to go along. A large majority of the experimental subjects conformed on at least some of the trials, and some conformed all the time. In debriefings afterward, the conformists ranged from those who knew their answers were wrong but thought they should go along, to those who felt their eyes must be deceiving them and so adopted the group’s perception. A few subjects conformed so completely that they professed not even to have noticed the evidence from their own eyes.

 

Outcomes of this kind would have come as no surprise to James Madison, nor to the thinkers who pioneered group theory, nor indeed to any thoughtful observer of group life, but they confirmed central sociological argu ments within the new framework of experimental validation. Backed by a great many similar studies, the results were widely influential. Theorizing, too, became more sophisticated in this period. For example, Peter Blau (1964) set out a proto- formal theory of individuals’ search for approval, status, and power, the formation of groups, and their exchange and conflict in society and politics.

 

Further dramatic evidence of the affinity of human beings for groups was developed by Henri Tajfel (1970; 1981). In his experiments, people arbitrarily assigned to meaningless groups quickly began to favor their fellow group members against others, even when they knew nothing about anyone else involved, they themselves did not profit from their bias, and no prior group conflict had occurred. This “minimal group” paradigm demonstrated that the human capacity for joining groups and disliking other groups was close to the surface and easily mobilized, a phenomenon particularly familiar in competitive situations. The great similarity to each other of undergraduate populations in most American college football conferences, for example, does not prevent students and alumni of each university from entertaining elaborate and emotionally felt theories about how dramatically they differ from the unfortunates at the other institutions, particularly those at their main rival school. Hard- fought, “dirty” college games have long been fodder for social psychologists studying out- group stereotyping (Hastorf and Cantril 1954). The psychological processes at work are familiar to anyone who has attended sports events. At this level, it is usually harmless.

 

Out- group stereotyping has deeper and more consequential effects when the stakes are larger. Racial and ethnic prejudices are endemic in all societies, as is nationalism in every country. Particularly when conflict arises, as in ethnic violence or in war, the stereotyping can lead to seeing the “other” as evil or even subhuman. In less dramatic conflict, such as election campaigns, group tensions can be mobilized to benefit a political party at the expense of a stereotyped minority. Gordon Allport (1954) pioneered the study of prejudice and its roots in individual and group psychology. He found that drawing boundaries between a good “us” and a bad “them” occurs widely in social life. Subsequent students of political psychology have found repeatedly that many, perhaps most, individuals organize their political thinking around social groups and their role in competing political parties. They see political and racial clashes as group conflict, and they engage in the same kind of stereotyping and moralizing that Allport had found (for example, Converse 1964; Kinder and Kam 2010).

 

In sum, an enormous body of experimental, quantitative, and qualitative evidence has accumulated that validates the realist understanding of how people think. Precise causal mechanisms remain in dispute. Are group attachments dark irrationalities or simply an often misguided attempt to pursue rational self- interest? Do people’s group- related ideas and attachments derive primarily from their own needs or from elite cues and messages? Much remains to be learned. But in our judgment, the argument that people typically think as the realists say they do, and not as theorists in the Enlightenment tradition claim— that much is beyond serious dispute. But what about politics? And what about elections in particular?

 

The High Point of Group Theory Influence and Its Decline

 

In the 1950s and 1960s, political scientists were well aware of the intellectual developments of the preceding century in psychology and sociology. The key statements from that period about American politics (Truman 1951), about empirical political theory (Easton 1953; 1965), and about comparative politics (Almond and Powell 1966) all took their starting point to be groups making demands on government, constrained by the institutional structure and the political culture. Government, in turn, had its own agenda, and it reacted back upon the society. Individuals played little role; elections received little attention except as a “feedback loop” in the informal systems- theory language popular at the time (Richardson 1991).

 

When elections were studied in greater detail, the findings matched group theory, too. The pioneering Columbia University studies (Lazarsfeld, Berelson, and Gaudet 1948; Berelson, Lazarsfeld, and McPhee 1954) took group forces as fundamental. The Columbia scholars showed that group memberships— being a Protestant rather than a Catholic, a union member rather than not, or a white person rather than an African American or other minority— powerfully shaped vote choices. People adapted their ideas to those of the presidential candidate they favored, or if they were less informed, they simply assumed (sometimes incorrectly) that his ideas matched their own.

 

The Columbia studies were subject to the criticism that group memberships might be proxying for attitudes or ideology. Thus, subsequent survey research often employed far more elaborate batteries of attitudinal questions and devoted substantial effort to disentangling the relationships among them. Critics also worried that the particular locales the Columbia researchers investigated might not be representative. Thus, most subsequent researchers have sampled the national electorate rather than specific communities, achieving greater representativeness at the cost of a much- diminished grasp of community context and social group action. National surveys of unrelated individuals were all too congenial for folk theory thinking.

 

Nonetheless, the first (and still the best) of these studies, The American Voter (Campbell et al. 1960), achieved a new depth of understanding. As we saw in chapter 2, in most respects it came to conclusions similar to those of the Columbia studies. The authors reported that the typical person lacked clear, stable attitudes toward major political issues. Most conceptualized politics in terms of groups, not ideology. But The American Voter departed from previous work in stressing that a particular political identity, party identification, proved fundamental to vote choice. Party loyalty powerfully shaped issue positions as well: again, group memberships largely drove policy views, not vice versa.

 

We address the group- theoretic understanding of the policy- making process in greater depth in chapter 11. Here, we simply note that one offshoot of group theory— “pluralism”— lent itself to a rosy view of American politics. Pluralists conceded the prominent role of groups in both the voters’ thinking and the process of government, and the inequalities of power among those groups, but they attempted to defend the resulting peaceful equilibrium of group pressures, managed by political leadership, as tolerably “democratic.” For example, in his classic pluralist study of politics in New Haven, Robert Dahl (1961, 228) argued that the democratic process worked reasonably well. While resources were by no means equal, almost everyone had at least some resources, those with many resources of one kind often were badly off with respect to other resources, and no one resource worked well in all or even most aspects of city politics; thus, no one group dominated all aspects of city life. Edward Banfield’s study of Chicago politics took a similar view, with a more explicit normative defense of the equilibrium of group interests. “Where there exists no concretely meaningful criterion of the public interest,” Banfield (1961, 339) wrote, “the competition of interested parties supplies a criterion— the distribution of ‘real’ influence— which may be both generally acceptable and, since it puts a premium upon effort to acquire influence, serviceable to the society.”

 

In short, by the early 1960s pluralist ideas had come to dominate how prominent American political scientists thought about domestic poli tics. But pluralism was in many respects a heretical departure from the earlier group theorists’ vision of politics as inevitably conflictual and violence- prone. Writing in the wake of the Depression and the Second World War, the pluralists put a powerful emphasis on political stability. Truman’s (1951, 535) final paragraph emphasized “stability”; Dahl’s (1961) concluding chapter was titled “Stability, Change, and the Democratic Creed.” In their view, the United States was not like Nazi Germany or Soviet Russia; it was capable of solving its problems using regular elections and peaceful policy compromises. Thus, the pluralist perspective on the American political system was largely supportive of its basic institutional framework and of its status as a constitutional democracy. Not without criticisms of American politics, nor Pollyannaish about the potential for disaster, these gifted scholars nonetheless thought that with skilled leadership and devotion to democratic norms, needed changes could be achieved in much the same way that the New Deal had reshaped American politics. They were realists in their view of people, and meliorists in their view of American democracy— then and now, an appealing vision in many respects. They had gotten well past the folk theory. But at their weakest, there was a certain 1950s complacency in their thinking.

 

Then a reaction set in. In 1960s America, African- Americans engaged in widespread peaceful demonstrations to insist on their long- denied civil rights, particularly in the South. They were met by violent resistance from police forces and from white vigilantes. Violence broke into politics at the national level, too, first with the assassination of a president, and then the fatal shootings of the country’s most prominent black leader and a presidential candidate, plus the serious wounding of another candidate. Most large American cities experienced at least one summer of violent unrest. The increasingly unpopular, unwinnable, and morally questionable Vietnam War inspired campus demonstrations, which were often met by violent police suppression. Sporadic incidents of violence and political murder by fringe groups inflamed the situation further. It seemed to many that the country was coming apart.

 

None of this would have surprised Arthur Bentley or his realist predecessors. Bentley (1908, 259) wrote, “Pressure is broad enough to include all forms of the group influence upon group, from battle and riot to abstract reasoning and sensitive morality.”8 But the standing and scholarly reputation of the pluralists was shaken by the events of the 1960s. The academic establishment was suddenly seen as overly satisfied with status quo politics and out of touch with the forces bubbling up from civil society (Walker 1966). In the view of critics, peaceful interest group conflict tending to an equilibrium did not seem to be what was going on. Instead, left- wing perspectives of various stripes were called into service, and theories of popular mobilization and moral resistance got more attention. The demands and ideas of ordinary people were taken as morally compelling (Garson 1978).

 

In the American politics of this period, the Democratic Party became divided between an old guard dedicated to Cold War anticommunism and big- government social welfare policies versus an antiwar New Left that wanted more direct popular participation in politics. These divisions played into the academic politics of the time, with the tired and ineffectual national political establishment identified with the tired and ineffectual pluralists. Truman (1971, xl) complained about careless critics treating the failures of the political system as criticisms of pluralism. But in other respects, the critics had a point. Times were bad. The academic establishment was associated with pluralism, and pluralism had little place or sympathy for the kind of politics that had arisen. The powerful group theories that would have made sense of the 1960s had been forgotten, or blurrily misremembered as early pluralism. An academic version of collective guilt resulted: the offending party and its innocent relatives were all convicted. Pluralism and group theory went into decline.

 

The ideas that replaced pluralism, after an initial period of uncertainty, debate, and eclecticism, were in most respects the old- time American political religion. That is, the folk theory of democracy returned with a new lease on life. The accumulated science against it was not refuted; indeed, it continued to accumulate, as we have seen. But it was ignored in the schizophrenic fashion we documented in chapter 2. To younger authors in the folk theo ry tradition, group theory seemed to be terra incognita: Hic dracones. Groups were implicitly moved offstage; the structure of civil society disappeared from view.

 

For one group of authors, direct participation in politics by individual citizens became a new ideal, as in the work of Carole Pateman (1970) and her successors.9 Another widely influential group of scholars rediscovered Anthony Downs’s (1957) attempt to inject economistic thinking into political science. For them, as we saw in chapter 2, people’s preferences were to be taken as given; there was no accounting for ideological tastes. When a basis for ideology was posited in this tradition, it was self- interest, usually economic self- interest. Social groups were absent. Citizens were rational individuals with an ideological viewpoint but no group identities, much as in a high school civics course.

 

The notion that our own thoughts and feelings are rational, or should become so, is always superficially appealing, of course. The liberal tradition celebrates that doctrine as foundational for politics, as Karl Mannheim (1936, 122– 126, 219– 229, and passim) pointed out long ago. Thus, following Downs, an entire field of collective choice and political economy was developed to play out the implications of rational individualism, often with relatively sophisticated applied mathematics. Adopting that viewpoint was sometimes just a modeling convenience for the political economists; more often it powerfully affected their conclusions, as we have seen in chapter 2. They wrote as if the “crisis of reason” that formed the modern mind a century ago had never occurred (Burrow 2000). The result was a body of work that was simultaneously advanced in its methods and antiquated in its ideas.

 

The political economists’ school of thought had devotees, but it had strictly limited impact on the rest of the discipline. In model after model, the assumptions were too uninformed about well- established social science findings, and the implications were too naive about politics and about people, to make much lasting contribution to understanding democracy. Indeed, as we discussed in chapter 2, the two big results from this tradition— Downs’s demonstration that competing political parties would converge to the ideological center, making polarization impossible, and Kenneth Arrow’s theorem establishing that no voting rule would satisfy a simple and appealing- looking list of ethical postulates— embroiled the field in empirical oversimplifications and deep conceptual tangles, respectively. Obviously, both results were substantial theoretical achievements, and subsequent work in this tradition sometimes achieved valuable insights. Any serious theory of democracy must wrestle with them. But in our view, this approach is not foundational. Rather, it underlines the need for theorists of democracy to look elsewhere for more fruitful starting points.

 

Developments in Political Psychology

 

In this same period, psychology, too, moved away from group studies. The field underwent a “cognitive turn.” Influenced by experimental findings that simple stimulus- response models of attitude formation and change were inadequate, psychologists placed more emphasis on the mental processing that their experimental subjects undertook. In social psychology, “social cognition”— focused primarily on individual thinking— became the central focus of the field (Fiske and Taylor 1991).

 

In addition, the experimental paradigm became ever more powerful in psychology, including among cognitive psychologists. Randomizing gender, religious affiliation, or social class was impossible. Thus when groups appeared in experiments, they were usually ad hoc groups created in the laboratory, not races, religions, or interest groups. In consequence, external validity came into question, and even more so, external relevance. As Tajfel (1981, 23) put it, “There are few social psychologists who have not at one time or another felt uneasy about the social vacuum in which most of their experiments were conducted.” Did the stated effects hold outside the artificial environment of the university laboratory, typically populated only by undergraduates? No one knew for certain. Many of the same concerns arose about experiments embedded in public opinion surveys, which also became popular in political science and sociology. Did isolated individuals answering survey questions respond as they would in realistic political circumstances? It was often difficult to be sure. But the preferred social science tools worked best for studying artificially isolated individuals, not the consequential groups functioning in the society outside the lab, and so the study of individuals is where psychology focused.

 

For a macro– social science like political science, therefore, borrowing from an increasingly micro– social science like psychology became more difficult. Most psychologists addressed political groups only indirectly, and making use of their results in analyzing everyday politics required complex chains of auxiliary reasoning. Some political scientists mastered the cognitive psychology literature and employed it in studies of public opinion and voting, arriving at a skeptical view of human cognitive capacities in politics (for example, Lodge and Taber 2013). Most students of politics, though, went their own way, using an informal, eclectic psychologism as the main, and not very powerful, alternative to the economists’ rational choice theories. Modern treatments of how individuals acquire their ideas from the social world continued to sell well, impress, and even shock undergraduates (notably Berger and Luckmann 1967). But it was hard to know what to do with them. A few political scientists carried out sophisticated empirical studies exploring the influence of national political culture on ordinary citizens (for example, McClosky and Zaller 1984) or the impact of party elites on the attitudes of their partisans (Zaller 1992). But for most, the group theory inheritance had little effect on their day- to- day practice of political science. Liberal individualism fit our theoretical and empirical tools all too well. Thus, when pressed about democracy, we fell back on the familiar and congenial folk theory, with occasional nods to retrospective voting.

 

Identity Theory in Politics and the Road Forward

 

In recent decades, as always in politics, new realities have intervened to disturb the conventional wisdom and reinvigorate scholarship. Beginning in the 1980s, evangelical Christians and other religious conservatives began playing an important electoral role in American politics. The terrorist attacks of September 11, 2001, themselves a consequence of religious fanaticism, also made American nationalism a potent force in elections. The increase of Latinos and Asians in the American population directed additional academic attention to the politics of racial and ethnic diversity.

 

In each of these areas of political science, the concept of identity became central (Tajfel 1982). This concept serves to distinguish groups to which an individual belongs that are not central to his or her self- concept from those that are a more integral part of the personality. Different people in the same group may differ. Thus some Catholics have a purely nominal attachment to the denomination. Others think of it as central to who they are, so that a disparaging remark about Catholics is an attack on them. The latter group have a Catholic identity. As Cantril (1941, 35) put it in an earlier period before the term “identity” had been coined, “When a person’s national flag is torn down, he is insulted; when disparaging remarks are made of his parents, he is involved; when his football team or political party loses a contest, he has been defeated. Certain standards, frames and attitudes the individual feels are a part of him.”

 

Identities are not primarily about adherence to a group ideology or creed. They are emotional attachments that transcend thinking. The Irish poet Seamus Heaney once spoke at a University of Michigan event that one of us attended. Northern Ireland was undergoing “The Troubles,” with bitter Protestant- Catholic tensions punctuated by terrorist violence on both sides. Heaney told of an Irish visitor to the North who was asked whether he was a Protestant or a Catholic. The man said he was an atheist. “Yes, yes, we understand,” his hosts replied. “But are you a Protestant atheist or a Catholic atheist?” Theology was not the issue.

 

Identity theory made sense of why people adopt the opinions of their group, as group theory had emphasized but not explained. Thus, for example, the quixotic but powerful temperance movement in American history, so difficult to understand as rational self- interest, became understandable once it was seen as a “symbolic crusade,” that is, as an attempt to reinforce Protestant identity as that identity came under threat from extensive Catholic immigration (Gusfield 1963). Self- sacrifice for the group’s mission also became understandable; in some sense, people were doing it in their own interest. Subsequent theoretical development extended identity theory in many directions and fit it more closely within the framework of cognitive psychology, as in John Turner’s (1991, chap. 6) self- categorization theory.

 

The concept of identity was widely adopted within political science. Thus, decades after group concepts from sociology had been abandoned and replaced by 19th- century liberal perspectives on human behavior, groups regained a central role in political psychology. A vast quantity of important work has been produced, far too much to cite here. Some of the most powerful work in this vein has focused on the single most powerful social cleavage in contemporary American politics— race. Michael Dawson (1994, 204– 205), for example, showed that the political behavior of African- Americans is “powerfully influenced” by their sense of “shared fate”— their “perceptions of group interests. What is perceived as good for the group . . . plays a dominant role in shaping African- American partisanship, political choice, and public opinion” despite significant disagreements regarding specific racial and economic policy issues. In the same spirit, Vincent Hutchings and Hakeem Jefferson (2014) showed that racial identity is a key predictor of Democratic Party identification among African- Americans, well beyond the impact of policy views.

 

Of course, political scientists have also studied other identity groups groun ded in ethnicity, social class, religion, nationality, or region. For example, Katherine Cramer’s (forthcoming) in- depth examination of “the politics of resentment” traced the crucial role of rural consciousness in shaping the political views of working- class whites in upstate Wisconsin. Much of what has been said in studies of this sort tracks closely with the perspectives of the Founding Fathers and 19th- century sociologists; but the new scholarship has produced far deeper evidence and more sophisticated conceptual differentiation. It represents a genuine intellectual advance.

 

Yet even now, the concept of identity is too often imperfectly integrated into the study of political behavior. Much political science research mentions the concept but fails to measure it. Group memberships are substituted for group identities, as if every group member were interchangeable. Hence degrees of identity cannot be used to explain differences in opinions and behavior within the group. Nor, without careful measurement, can its effects be separated from those of attitudes correlated with it (Abdelal et al. 2009). Do the attitudes motivate the identity, or does the identity produce the attitudes? The same inferential challenges that bedeviled the Columbia studies of voting behavior in the 1940s are still with us.

 

In our view, a realist theory of democracy must be founded on a realistic theory of political psychology. At present, nothing of that kind exists. Step one, therefore, is to begin building a framework for thinking about voters that escapes from the populist liberalism that has constrained so much recent thought. Madison pointed the way, and 19th- and 20th- century scholars advanced the subject considerably. But we have too often stalled in recent decades. It is time to set Jeremy Bentham aside and bring James Madison back in.

 

To accomplish that, identity theory will have to be brought into the macro- social world of politics. The role of political elites in structuring politically rel evant cleavages needs to be understood better. It is popular now to treat identities as malleable and socially constructed, as indeed they are. We agree that politicians make their own cleavages, but we would add: not just as they please. The limits of malleability need to be spelled out.

 

Equally important, the impacts of identities need to be demonstrated, not only in the lab, but in the real political world. For example, when a presidential candidate draws surprising support in polls and primaries, we need sophisticated survey research tools that allow us to trace that support to identity groups, and not just to the preferences, attitudes, and ideology that the folk theory holds dear.

 

In addition, the political science version of identity theory needs to become more cognitive, just as social psychology has. Predictive power emerges from theoretical differentiation, and an implicit stimulus- response theory of identity acquisition will not do. Formalization and close empirical investigation of how people acquire an identity and how they use it to adopt ideas and make decisions is critical to further progress.

 

We can make no more than a start on this agenda in the remaining chapters of this book. Thus, just as births have a less finished quality than funerals, the succeeding chapters have a less finished quality than those preceding. The work of a great many scholars will be needed to complete the project we outline. What we do hope to achieve is to lay out a preliminary version of what a modern group- theoretic understanding of political attitudes and behavior might look like. Thus, in chapter 9 we argue in detail that social identities have trumped policy reasoning in shaping the politics of religion, race, and abortion in contemporary America. Hence in thinking about politics, it makes no sense to start from issue positions— they are generally derivative from something else. And that something else is identity.

 

In chapter 10 we turn from empirical investigation of identity politics to offer an account of how partisan identity shapes perceptions, not just of candidates and issues but also of simple facts. Instead of beginning from fanciful assumptions about what people know, we begin instead from realistic postulates that allow for differing levels of exposure to factual information and differing levels of partisan attachment. The result is a framework that can account with considerable accuracy for the complex differences in opinions and beliefs between Democrats and Republicans across the entire spectrum of political expertise, from the most uninformed and disengaged to the most (supposedly) knowledgeable.

 

Together, chapters 9 and 10 attempt to make the case that “groupiness” (Stenner 2005, 18) is fundamental to thinking about the beliefs, preferences, and political behavior of democratic citizens. In our concluding chapter, we sketch the implications of that fact for normative theories of democracy and for models of the policy- making process.

 

--------------------------------------

1. In the British context, Beer (1966, 39– 43) called this category of democratic thought “Radical,” and he traced its origins to the 17th- century Levellers movement. He noted that it has never been as influential in Britain as in the United States.

2. This intellectual history was insightfully interpreted by James Aho (1975).

3. As we have already noted, we have no dispute with rational choice per se, which is nearly tautological in content. What matters is what the citizen’s starting point in politics is assumed to be. The Federalist Papers, supported and enhanced by 19th- century sociology, make it clear that preferences over policies or outcomes are not where citizens start. Instead, they begin from group affiliations. From that basis, group theory can be given a rational- choice microfoundation, outlined in Truman (1951, 18– 21) and developed in different ways by Hechter (1987) and Hardin (1995), among others. Thus the issue is not rational whether, but rational how. Self- interested individualistic rational choice is the wrong “how” for understanding most of politics.

4. This description departs from many stereotypical summaries of Bentley (1908) in the literature, but it is faithful to the book itself.

5. The Lynds do not say which topics— evolution? socialism?— or which college. If the institution in question was Indiana University, as seems probable, then unfortunately for the student, Alfred Kinsey had not yet arrived there to offend small- town sensibilities in more thrilling fashion.

6. About 15 years ago, one of us attended with his wife a football game in Ann Arbor in which the town’s two principal high schools played each other. Both schools were similar in racial and social class composition. His children attended the school with green football uniforms; the bitter cross- town rivals wore purple. The fans of each school sat on opposite sides of the stadium. About halfway through the game, his soft- spoken wife said under her breath, “I know that if we had bought the first house we considered, we’d be sitting on the other side of this field. But I just can’t stand those purple people.”

7. The work of Robert Huckfeldt and his colleagues (Huckfeldt and Sprague 1995; Ahn, Huckfeldt, and Ryan forthcoming) represents a prominent exception to this trend, and it provides ample discussion and illustration of the benefits of understanding local context.

8. Bentley (1908, 216, 241, 371, 433) explicitly took note of a spectrum of political action ranging from discussion through voting to bribery and, when satisfaction of grievances was blocked, assassination, terrorism, and revolutionary violence. Contrary to claims from some opponents of pluralism, the racial demonstrations and riots of the 1960s fit his arguments perfectly, if not always those of the pluralists who claimed his mantle.

9. We do not attempt to review this voluminous literature here except to note that, in spite of its tensions with rational choice, participatory theory descends from the same liberal tradition.

10. Of course, political group affiliations can be made experimentally more or less salient to subjects, but that is not the same as altering the affiliations themselves.

11. The literature on identity is vast— far too extensive to cite here. Monroe, Hankin, and Bukovchik Van Vechten (2000) provided a helpful review emphasizing political identity and group behavior. Huddy (2001) is also an important source.