Behavioralism, which was one of the dominant approaches in the 1950s and ’60s, is the view that the subject matter of political science should be limited to phenomena that are independently observable and quantifiable. It assumes that political institutions largely reflect underlying social forces and that the study of politics should begin with society, culture, and public opinion. To this end, behavioralists utilize the methodology of the social sciences—primarily psychology—to establish statistical relationships between independent variables (presumed causes) and dependent variables (presumed effects). For example, a behavioralist might use detailed election data to argue that voters in rural areas tend to vote for candidates who are more conservative, while voters in cities generally favour candidates who are more liberal. The prominence of behavioralists in the post-World War II period helped to lead political science in a much more scientific direction. For many behavioralists, only such quantified studies can be considered political science in the strict sense; they often contrasted their studies with those of the so-called traditionalists, who attempted to explain politics by using unquantified descriptions, anecdotes, historical analogies, ideologies, and philosophy. Like behaviourism in psychology, behavioralism in political science attempted to discard intuition, or at least to support it with empirical observation. A traditionalist, in contrast, might attempt to support intuition with reason alone.

Perhaps the most important behavioral contributions to political science were election studies. In 1955 American political scientist V.O. Key, Jr. (1908–63), identified as “critical,” or “realigning,” several elections in which American voters shifted their long-term party affiliation massively from one political party to another, giving rise to the dominance of the Republican Party from 1860 to 1932 and of the Democratic Party after 1932. Pioneering statistical electoral analyses were conducted by the University of Michigan’s Survey Research Center (SRC), which was developed in the 1940s. In The American Voter (1960), Angus Campbell, Philip Converse, William Miller, and Donald Stokes used the results of studies by the SRC to develop the concept of party identification—the long-term psychological attachment of a voter to a political party. The long-recognized influences of religion, social class, region, and ethnicity, they argued, contribute to voting behaviour only insofar as the voter has been socialized, primarily by his parents, to adopt a particular party identification.

Behavioral approaches were soon adopted outside the United States, often by scholars with connections to American universities. The University of Oxford initiated election studies in the 1960s, and David Butler and Donald Stokes—one of the authors of The American Voter—adapted much of the American study in Political Change in Britain: Forces Shaping Electoral Choice (1969). They found that political generation (the era in which one was born) and “duration of partisanship” also predict party identification—that is, the length of time one has been a partisan heavily predicts one’s vote. They also found that party identification, initially transmitted by one’s parents, may change under the impact of historic events. The influential Norwegian scholar Stein Rokkan pioneered the use of cross-national quantitative data to examine the interaction of party systems and social divisions based on class, religion, and region, which in combination explain much voting behaviour. Rokkan identified the importance of “centre-periphery” tensions, finding that outlying regions of a country tend to vote differently from the area where political and economic activities are centred. The extensive Eurobarometer series—public-opinion surveys carried out in European Union countries since 1973 on behalf of the European Commission—have given European behavioralists a solid statistical base on a range of political, social, economic, and cultural issues; the surveys have provided valuable data for examining trends over time, and they have shown, among other things, that modern European ideological opinion clusters around the political centre, suggesting that stable democratic systems have taken root. More recently, Transparency International, founded in 1993 in Berlin, has conducted worldwide surveys that attempt to quantify corruption. In Latin America, Guillermo O’Donnell and Arturo Valenzuela used public-opinion surveys and voting, economic, and demographic data to examine the forces that have destabilized democracy there.

The behavioral approach was also central to the work of the American sociologist and political scientist Seymour Martin Lipset, whose influential Political Man: The Social Bases of Politics (1960) used statistical and historical data to demonstrate that social class is one of the chief determinants of political behaviour. Lipset infuriated Marxists by portraying elections as “the democratic class struggle” in which the working class finds its true voice in moderate leftist parties. Lipset also contributed to modernization theory by identifying factors that explain why countries adopt either authoritarian or democratic political systems. Specifically, Lipset found a strong relationship between level of affluence and type of political system, demonstrating that less-affluent countries seldom establish democratic structures.

Behavioralism also influenced international relations, though it did not achieve the same dominance in this area that it enjoyed in domestic and comparative politics. The Correlates of War Project, founded at the University of Michigan in 1963, gathered much quantitative data and became one of the leading sources for scholars studying the causes and effects of war and international tension. Behavioralism also established itself in studies of judicial and bureaucratic systems.

By the 1960s behavioralism was in full bloom, forcing the traditionalists into retreat in much of the discipline. By the late 1960s, however, criticism of behavioralism had begun to grow. One charge leveled against it was that the statistical correlations uncovered by behavioral studies did not always establish which variable, if any, was the cause and which the effect. The fact that two variables change together does not in itself show which causes which; indeed, the changes exhibited by both variables may be the effects of an underlying third variable. In order to make sense of the actual relationship between the variables, the researcher must often use intuition—a tool that behavioralists expressly sought to avoid. A study of white blue-collar Roman Catholics in Detroit, Michigan, for example, might find that during a certain period they were more likely to vote Republican as they became more affluent and suburbanized. However, whether the change in their voting patterns was due to their race, their religion, their increased affluence, or their suburban lifestyle—or whether they simply responded to the message or personality of particular Republican Party candidates—may be unclear.

In addition, though behavioral research yielded important insights into the political behaviour of individuals, it often explained little about actual governance. Voting studies, for example, rarely provided an understanding of public policy. Because behavioral research tended to be limited to topics that were amenable to quantitative study, it was often dismissed as narrow and irrelevant to major political issues. Indeed, intense methodological debates among behavioralists (and within the discipline more broadly) often seemed arcane, filled with esoteric jargon and addressed to issues of little concern to most citizens. Because behavioralists needed quantitative survey and electoral data, which were often unavailable in dictatorships or less-affluent countries, their approach was useless in many parts of the world. In addition, the reliability of behavioral research was called into question by its dependence in large part on verbal responses to questionnaires. Analyses of survey results have shown that respondents often give socially desirable answers and are likely to conceal their true feelings on controversial topics; moreover, the wording of questions, as well as the ordering of possible answers, can affect the results, making concrete conclusions difficult. Finally, many behavioral findings revealed nothing new but simply restated well-established or obvious conclusions, such as the observation that wealthy people tend to vote conservative and poor and working-class people tend to vote liberal or left-of-centre. For all of these reasons, behavioralism did not become the sole methodology in political science, and many behavioralists eventually acknowledged the need for the unquantified insights of traditionalists; by the late 1960s political scientists called this the “postbehavioral synthesis.”

Political culture

Political culture may be defined as the political psychology of a country or nation (or subgroup thereof). Political culture studies attempt to uncover deep-seated, long-held values characteristic of a society or group rather than ephemeral attitudes toward specific issues that might be gathered through public-opinion surveys. Several major studies using a political culture approach appeared simultaneously with the behavioral studies of the late 1950s, adding psychological and anthropological insights to statistical covariance. The study of political culture was hardly new; since at least the time of Plato, virtually all political thinkers have acknowledged the importance of what Tocqueville called “habits of the heart” in making the political system work as it does. Modern political culture approaches were motivated in part by a desire to understand the rise of totalitarian regimes in the 20th century in Russia, Germany, and Italy, and many early studies (e.g., The Authoritarian Personality) focused on Nazi Germany; one early political culture study, Edward Banfield’s The Moral Basis of a Backward Society (1958), argued that poverty in southern Italy grew out of a psychological inability to trust or to form associations beyond the immediate family, a finding that was long controversial but is now accepted by many.

Perhaps the most important work of political culture was Gabriel Almond and Sidney Verba’s The Civic Culture: Political Attitudes and Democracy in Five Nations (1963), which surveyed 1,000-person samples in the United States, the United Kingdom, Germany, Italy, and Mexico. Almond and Verba identified three types of political culture: (1) participant, in which citizens understand and take part in politics and voluntary associations, (2) subject, in which citizens largely obey but participate little, and (3) parochial, in which citizens have neither knowledge of nor interest in politics. The authors found that democratic stability arises from a balance or mixture of these cultures, a conclusion similar to that drawn by Aristotle. In Almond and Verba’s edited volume The Civic Culture Revisited (1980), several authors demonstrated that political culture in each of their subject countries was undergoing major change, little of which was predictable from the original study, suggesting that political culture, while more durable than mere public opinion, is never static. Critics of The Civic Culture also pointed out that political structures can affect culture. The effective governance and economic policies of West Germany’s government made that country’s citizens embrace democracy, whereas Britain’s economic decline made Britons more cynical about politics. The problem, again, is determining causality.

Over the decades Lipset, who served as president of both the American Sociological Association and the American Political Science Association, turned from explanations of political values based on social class to those based on history and culture, which, he argued, displayed consistency throughout history. American political scientist Robert Putnam followed in this Tocquevillian tradition in Making Democracy Work: Civic Traditions in Modern Italy (1993), which demonstrated that the historical cultures of Italy’s regions explain their current political situations. In Bowling Alone: The Collapse and Revival of American Community (2000), Putnam claimed that the American tendency to form citizen groups, a characteristic that Tocqueville praised, was weakening. Americans were less often joining groups and participating in politics, Putnam argued, leading to a loss of “social capital” (the collective value of social networks) and potentially undermining democracy, a worry shared by other political observers in the United States.

Adopting what became known as the “path-dependent development” approach, advocates of the historical-cultural school maintained that contemporary society is a reflection of society in ages past. The political culture approach declined in the 1970s but was later revived as political scientists incorporated it into explanations of why some countries experienced economic growth and established democratic political systems while others did not. Some suggested that the rapid economic growth and democratization that took place in some East Asian countries in the second half of the 20th century was facilitated by a political culture based on Confucianism. In Africa and Latin America, they argued, the absence of a culture that valued hard work and capital accumulation led to the stagnation of much of those regions. This viewpoint was captured by the title of Lawrence E. Harrison and Samuel P. Huntington’s edited volume Culture Matters: How Values Shape Human Progress (2000).