A review of The History of the Social Sciences since 1945, edited by Roger E. Backhouse and Philippe Fontaine

Social science is an effort to explain human behavior. A useful history of it would weigh the many different efforts that economics, geography, psychology, political science, sociology, and social anthropology have made since the Second World War to offer compelling accounts of that behavior in its many dimensions. It is a bit odd that the study of history was omitted from this new collection edited by Roger E. Backhouse and Philippe Fontaine, but quarrels about whether history qualifies as a social science have long divided historians as well as their colleagues in other fields.

The History of the Social Sciences since 1945 describes the organizational changes that have occurred in the social sciences over the past six decades, and the rise and fall of many of its leading practitioners. We learn a bit about how the Office of Strategic Services (OSS) mobilized social scientists (including historians) during World War II, the development of survey research (that is, opinion polls), the creation of such ventures as the Social Science Research Council and the RAND Corporation, and the prominence of scholars such as Robert Dahl, Milton Friedman, Clifford Geertz, and Friedrich Hayek.

But what we learn is just the bits and pieces. None of these organizations or personalities receives more than a glancing reference, and of these some are deeply puzzling, such as the reference to Hayek as a person who was "hardly taken seriously" until he received the Nobel Prize in 1974. (This will come as news to the tens of thousands of people who had read his books, not to mention to the entire school of Austrian economics.)

Part of the problem is that Backhouse and Fontaine, economics professors at the University of Birmingham and the École Normale Supérieure de Cachan, respectively, recruited as authors of the various chapters scholars who were either historians or whose principal writings were devoted to organizational histories of their disciplines. Part arises from the fact that this is a trans-Atlantic volume comparing the evolution of social science fields in England, France, and the United States. But the chief difficulty is that the editors did not focus the book on questions that would attract most readers interested in the social sciences: namely, which intellectual and theoretical approaches have heightened our understanding of human behavior and which have been either passing fancies or a reflection of the academic knots into which many academics have tied themselves.

* * *

Consider my own field, political science. In its early decades it sought to explain how government worked by describing its institutions; today it has embraced political behavior and made use of rational-choice theories. Studying political behavior means you attempt to quantify and explain how people vote and acquire politically relevant attitudes and how legislatures and courts reach decisions. To understand voting and attitudes, one measures the factors, such as age, gender, race, or education, that influence voting and attitudes among people; to explain legislative action, you look at the competing claims of ideology, leadership, and constituency pressures on legislative behavior. A great deal of excellent work has been done on these subjects even though such accomplishments do not diminish the fact that more attention ought to be paid to the great and competing ideas about what constitutes a good regime.

Rational-choice theory argues that people act so as to maximize the difference between the benefits they receive and the costs they pay for a course of action, with the goals of this action being whatever people prefer. Scholars who use rational-choice theories do not claim they explain all relevant behavior, but rather that they are more useful than any rival theories in explaining as much as we can of human behavior.

There is no doubt that identifying factors that correlate with (and possibly cause) voting behavior, or explain business decisions, is often useful. But there are limits to that usefulness. Take, for example, voting. We have discovered that men and women often vote differently; but we have not discovered why they vote at all. The chance of their vote affecting the outcome of an election is effectually zero, yet voting is costly (time, effort, and money must be expended to get to the polling place, wait in line, and fill out a ballot). A rational-choice theory may explain for whom they vote, but it cannot explain satisfactorily why they vote at all.

Or take another example, that of changing legislative policies in dramatic ways. Between the 1880s and the early 1970s, Congress decided to regulate many industries, and in the 1970s and 1980s it made a complete reversal and decided to deregulate many of these same industries. One can find some evidence that there was a popular or business demand for regulation, but one will search in vain for facts that support much popular or commercial support for deregulation. Yet it happened. Why?

Rational-choice theory reflects the impact of econometrics on political science. As this book observes, by the 1980s about one fifth of all of the articles in the American Political Science Review were based on formal modeling, typically involving rational-choice theory. The American Political Science Association has been convulsed by debates over whether and to what extent formal modeling is useful, an argument that shows no signs of disappearing. Yet the book does not explore this development other than to note its existence. It mentions that political theorists worried about this change, but inexplicably (and wrongly) describes David Easton's book, The Political System (1953), as "a behaviorialist manifesto." It was no such thing; instead, it was an intellectually empty defense of the idea of a "political system" made of "inputs" and "outputs" that has left no visible trace on contemporary political science.

* * *

Consider another field, sociology. In 1950, the lead articles in the first issue of the American Sociological Review were about sociological theory, research in community leadership, social case work, and a study of Japanese population trends. In a recent issue (October 2010), the articles are about racial segregation, stratification by skin color, male sexual escorts, and the financial penalty from being a mother. Sociologists had decided that describing cleavages based on race and gender was more important than explaining how society affects individual behavior. The authors of the chapters on economics, political science, and sociology recognize that there were competing perspectives in their fields even though they did little to evaluate them. But nowhere does the book even mention, much less discuss, three major issues. First has been the rise of genetic analysis to explain human behavior. Its influence on psychology has been profound and it has made headway in economics and political science. Using studies of identical twins and adopted children, scholars have found important genetic influences on attitudes, intelligence, preferences, and religious and ideological identification. These influences are accompanied, of course, by cultural influences though these have little to do with the family and much to do with peer groups. Neither "genes" nor "heredity" even appears in the index.

Second, the authors do not address the question of what specific aspects of human behavior social science can explain. Microeconomics (that is, the study of how individuals choose among alternative courses of economic action) has made great progress, but macroeconomics (that is, the study of how the economy as a whole changes with respect to output or employment) has advanced very little. Few macroeconomists predicted the recession that now grips American and Europe. Political scientists know much more than they once did about the factors that influence our vote, but they have not gotten any better than ordinary opinion polls in predicting who will win an election. I am tempted to suggest that social science is at its best when it explains things that have already happened and its worst in predicting what will happen. There are exceptions to this sweeping generalization, but it is not a bad place to begin, especially if we are trying to understand policy changes.

Finally, the ideological quarrels that have convulsed many social scientists are mentioned, nowhere more tellingly than in a quotation from the political theorist, Quentin Skinner. He wrote that "the empiricist and positivistic citadels of English-speaking social philosophy have been threatened and undermined by successive waves of hermeneutics, structuralists, post-empiricists, deconstructionists and other invading hordes." But to this profound observation the book gives only superficial treatment. Consider sociology. It was once preoccupied with hierarchy, bureaucracy, and social structure, but now seems focused on "marginalized" groups, usually defined along lines of gender and race. Important as these topics are, they are scarcely a substitute for examining the social system, unless of course you begin your work with a prior commitment to the idea that all you need to know about that order can be found in how it treats blacks and women. That was never true in the past and is even less true today.

The editors and authors of this book have a ready response to my views: I am urging them to write a book I think they should have written instead of evaluating the book they did produce. It is a fair point. Nevertheless, the importance of social science to the human condition derives not from the names of its practitioners, the organizations that supported them, or how Americans and Europeans differ in their perspectives. What is vital is how well and at what costs social science explains human behavior. The book about that question still needs to be written.