Which liberalism?

It is less what political science has taught us in the recent past that is most impressive, than how its multifaceted focus on the Western liberal tradition has reopened inquiry to the most vexing, difficult, and important question in political life: how is a decent politics possible in the face of violence, fear, discrimination, and recurring instances of barbarism? After the Second World War, some of our most esteemed colleagues—including Harold Lasswell, Hannah Arendt, and Robert Dahl—created an all too brief moment when political thought, comparative politics, American political studies, international relations, and policy analysis were cross-hatched by this large inquiry, all the while animated by anxiety for the fate of constitutional democracy. Theirs was a search to understand how the Western liberal tradition could become more durable, less credulous, and more realistic in the face of unprecedented violence, radical evil, and entrenched exclusions.

By the mid-1960s, it was clear that this moment had passed. The discipline had become complacent. Its scientific advances, as David Truman warned in his APSA presidential address of 1965, had substituted an increasingly arid rigor for this and other large normative and institutional questions. Truman's critique was echoed even more loudly by that generation of graduate student critics, of which I was one, and a small number of key role models, who rightly critiqued the discipline for its neglect of racism, class, and power in all of its dimensions—and who refused any model of American or Western triumphalism. The period that immediately followed produced many specialized advances and a deepened ability to mine data and build models, but despite the heroic work of such scholars as David Greenstone, Judith Shklar, Aristide Zolberg, and Juan Linz, who kept an anxious quest for decent politics at the front of their sophisticated and careful writing, all the while refusing to confine their scholarship to well-trod paths, overall there was a collective degradation of focus and purpose.

The greatest gain of the past quarter century has been a return to the big issues of how to organize and sustain decent regimes in the face of many varieties of human depredation and challenges to the best legacies of Enlightenment. These interests animate work on democratic transitions, forms of participation, nationalism, the politics of global and local inequalities, and many other rich subjects. But most important, more and more work in the discipline has been probing the characteristics and prospects of the liberal tradition with a renewed joining of positive and normative purpose. Like the best work of the postwar period, this scholarship has reconnected the big issues of political theory to systematic inquiry. There has been an appreciation of the historical and institutional features of the earliest years of political science, but without a return to the complacency that took Western and white superiority for granted while limiting inquiry mainly to issues of public law and the history of ideas.

Especially important has been a focus by students of such pressing concerns as violence, ethnicity, race, and civil war not only on what happens and why, but also on the points of intersection between liberalism's best features of consent, toleration, and rights and its inherent difficulty in grappling coherently with the scope of its membership, with intense rather than superficial pluralism, and with the definition of proper conduct in the pursuit of security at home and abroad. Recent disciplinary trends have put us in a position to revisit more effectively than ever before—on the basis of far more knowledge, a wider repertoire of tools, and a more realistic sense of human capacities—how the combination of states and deep human diversity both threatens and, under some institutional conditions (not all of one kind), can enhance the prospects for a decent politics.

But if this impetus is to be sustained, much more needs to be accomplished. The powerful insights of critical theorists who have reminded us that for all that consent, toleration, and rights have liberated human potential, Western liberalism has also abided racially grounded slavery and colonial rule, religious bigotry, the subjugation of women, military conquest, and alliances with very ugly regimes must be brought to bear in a far more subtle way on our empirical inquiries than various dichotomies that too contentedly distinguish “us” from “them.” Studies of such key liberal subjects as the character and practice of political representation must attend more closely to the underside of the liberal experience. The relative absence of studies that highlight the often decisive role played by the slave and Jim Crow South in shaping American political development, as compared, say, to work on bureaucracies and state building has produced a far too sanitized version of the liberal tradition.

But in the United States, as elsewhere, no decent future is on offer without a more capable, more grounded, more open version of this set of ideas and institutions. So we might take heart from the manifest gains our discipline has been making in holding this tradition up to multiple types of systematic inquiry, and in starting to lower the barriers that divide political theory from empirical scholarship. As a result, political science, finally, is approaching the chance to realize its best long-standing feature as an organized discipline under the umbrella of the American Political Science Association: the wish to critically probe, assess, and advance political liberalism.

Failed and failing states

The recent elections in Iraq and Afghanistan and the on again–off again resolution of several major conflicts in Africa seem to confirm to the Bush administration that its approach to state-building is feasible. I wish I could share its optimism, but I cannot. All we know from social science and from history suggests that Iraq, Afghanistan, the Sudan, and many other countries in the world have a long and difficult process in front of them with no certainty that they can succeed.

Nor can policy makers turn to social scientists for recipes for success, despite the numerous theories of the state.¹

We now know that modernization theory and other grand evolutionary typologies may help us categorize countries and eras (and I even wonder about that),²

From historians and area experts, we have the detailed knowledge of cases essential for in-depth understanding of key processes and comparisons across time and place. From economists, political economists, and formal theorists—most importantly Mancur Olson, Michael Taylor, Robert Axelrod, and Elinor Ostrom⁷

We now have some confidence about necessary preconditions for state formation and democracy.¹¹

We have understood since Adam Smith that the “wealth of nations” depends on well-functioning markets. We have much more recently acknowledged that well-functioning markets require stable government. But government does not just happen, even when there are well-designed constitutions and especially when there are not. We think of France as a highly centralized state, but it was not easy to make it so. Forging a national identity required the conquest of provinces and their lords. Gaining compliance with laws required the loss of lives, including those of tax collectors, census takers, and draft registrars. Britain and the United States might now have nearly zero levels of official corruption, but that took centuries to achieve. How France centralized, how Britain and the United States eradicated most government corruption, and how all three gained relatively high compliance with their laws offers lessons for countries tying to build states today. It is not a question of following a particular grand path, but of learning how to solve specific problems.

When we turn to the contemporary world, the same question that plagued historical states infects today's: how to ensure that government has sufficient power to tax and provide security while inhibiting government from predation on the population it is supposed to be serving. Mancur Olson framed the issue as transforming “roving bandits” into “stationary bandits,”¹³

Another issue must be resolved if states are to be strengthened or rebuilt. We are ever more conscious that well-ordered institutions and the bases for cooperation require small steps that allow people to learn what institutions and which people are reliable and in what settings. Individuals can then take increasing risks and broaden the range of those with whom they can productively interact. However, there are also real dangers that tight local networks, effective bilateral exchanges, or even some of the more sophisticated and long-distance multilateral arrangements can become more constraint on than facilitator of widespread cooperation.¹⁶

We know quite a lot about how to govern communities, and we are learning why states fail. The next step is to figure out how to prevent them from failing and how to rebuild them if they do. Yet, even this grand task is not enough. We should no longer be satisfied with the Hobbesian solution, a state that provides only security. Our goal is not social order alone but an equitable, just, and democratic government that elicits from its citizens willing compliance and, better yet, contingent consent.

Answering “mother-in-law” questions can move the discipline forward

Major advances in political science have often come from puzzling about the causes and consequences of major events in the real world and seeking answers to questions that our fellow citizens could readily understand.¹⁹

From international to world politics

My marker for what we knew about international relations in 1980 is provided by Kenneth Waltz's masterful synthesis of structural realist theory, Theory of International Politics, published in 1979. The field is now more aware than it was 25 years ago of aspects of world politics that Waltz downplayed or ignored, including the roles played by nonstate actors and the ways that international regimes establish rules and practices that shape state behavior. I will emphasize, however, two more fundamental changes in our understanding: (1) the role of information and beliefs in world politics; and (2) the phenomenon of the “democratic peace.” The broad implication of these advances in knowledge is to transform the field from “international politics” to “world politics.”

Variations in information and beliefs did not play a major role in structural realist theory, although in the 1970s Robert Jervis had already done pioneering work on the role of perceptions and uncertainty in world politics.³³

Game theory taught us—as only became clear in the international relations literature in the 1980s and early 1990s—that deterministic theories of interaction could not succeed. Outcomes depend not only on power and interests but on the information available to the players: what they know about others' types, what they know about what others know about their own types, and so on, ad infinitum. Just as there is no unique solution to children's games such as “scissors, rock, paper,” there is no unique solution even to relatively simple interactions among states. Furthermore, credibility and commitment are crucial to understanding political strategy and the role of institutions, which means that beliefs (and interpretation) are central to politics.

The second big thing that students of world politics know now that they did not know in 1980 is that whether a pair of states is democratic, rather than both autocratic or of different types, has an enormous impact on their propensity to fight. Stable democracies fight other states and unstable democracies may fight each other. However, stable democracies do not fight one another except when virtually forced (for example, Finland during World War II) to do so. This “democratic peace” proposition is stylized and is not fully explained, but it is highly robust. We are also learning more about how democracies, or states governed by the rule of law, behave differently on issues other than war and peace, and within multilateral institutions. Our growing knowledge of this “democratic difference” renders untenable the realist view—almost unchallenged in 1980—that only interests and power matter, not regime type.

These contributions require substantial refinement. Constructivists, psychologists, and cognitive scientists are enriching our concepts of information and information processing. The best students of democracy analyze it as a practice that varies contextually, not as a stylized fact. As refinement of these advances continues, we need to address a related question that is of crucial policy import. Under what conditions can democracy in a given society be successfully promoted by advocacy networks, nongovernmental organizations, and states outside the society itself?

During the last twenty-five years, enormous effort has been devoted to promoting democracy around the world. The means by which such external democratizing has proceeded can be divided into seven categories:

• persuasion and example—as in John Quincy Adams's famous defense of America's “voice,” proclaiming “to mankind the inextinguishable rights of human nature,” updated as “soft power”;³⁵

  Adams 1821; Nye 2004.

• publicity and criticism—as in the activities of human rights activists;

• pressure by a great power on client states, as in U.S. relationships with South Korea, the Philippines, and Taiwan;

• offers of accession to international institutions, typically accompanied by positive economic incentives, such as those used by the European Union;

• military support for nominally democratic incumbent regimes, which failed in Vietnam in the 1960s and had mixed results in the Caribbean;

• multilateral peace building, as was attempted (with mixed results) in Angola, Bosnia, Cambodia, East Timor, El Salvador, Kosovo, Liberia, Mozambique, Namibia, and Somalia;

• invasion and conquest designed to overthrow tyrannical regimes and create the conditions for democracy. Occupation policies succeeded in Germany and Japan in the 1940s but failed for the United States in the Caribbean and the British in Egypt and Iraq during the first half of the twentieth century. It is an open question whether they will succeed in Iraq.

Controversy now rages about how ardently the United States should promote democracy, even when this objective comes into conflict with other interests. Behind these debates lie deep disagreements about both theoretical and empirical issues. Political science is already beginning to provide answers to some of these questions, but we need to do much more. In the process, the field of international politics will be transformed. It has already become more than interstate politics, since nonstate actors are so important today. Over the next decade, it will increasingly merge with the field of comparative politics, to become the field of world politics.

In this new world politics, zones of interstate conflict are likely to remain. Sovereignty will not disappear, and in Latin America and Asia it may even, for a time, be strengthened. Analysis of interdependence and interaction will require sensitivity to regional differences and complexity, as well as to nuances of information and institutional incentives. Yet the subject of our study will be less individual states and their policies than governance at various levels, from town meeting to institutions with a global scope. Sovereignty will be increasingly disaggregated by function; that is, the ability to make binding rules will be dispersed—devolved downward, spun off to transgovernmental networks or quasi-governmental entities, delegated to authoritative multilateral organizations that supersede traditional governments in failed states, or pooled and institutionalized to cover larger geographical areas. These arrangements will vary in their effectiveness in performing various tasks, in their legitimacy with different audiences, and in the degree to which they are accountable, democratically or in other acceptable ways. Understanding the impact of external agency on democracy within traditional national societies will be only a first step toward understanding democratic, quasi-democratic and nondemocratic governance in world politics.

To get beyond the cult of the robe

“Where are there high-priority intellectual opportunities?” Norton Long's principle that power is the life blood of administration is my guide. Of power, he says, “Its attainment, cultivation, maintenance, increase, dissipation and loss are subjects the practitioner and the student can ill afford to neglect.”³⁶

The controversy (in the first quarter of 2005) about the nomination and confirmation of Alberto Gonzales as attorney general of the United States is a case in point. The topical part of the controversy will pass. What will not pass is the need for the president, attorney general, and Department of Justice to attain, cultivate, and maintain power through the system of law. Nor will the dissipation and loss of power pass. Our high-priority intellectual opportunity lies in the fact that knowledge of the executive role in law is so ill-developed.

We now know much about courts, but there is a tremendous knowledge gap concerning the initiation of action under the color of law and also the withholding from initiation. Of course, only a small proportion of what is initiated ever gets into the courts.³⁷

A little of this, a lot of that

Overheard through thin office walls. An enterprising political science student (Student) interviews a political science professor (Professor).

Student: As I said on the phone, I'm doing a term paper on the state of political science, using interviews with a sample of professors as my main source. I'd like you to tell me some of the accomplishments and failures of the discipline over the last 15 to 20 years.

Professor: Do you mean in my field or in general?

Student: Just however you want to define it.

(Listener thinks: “This student must have taken our methods course.”)

Professor: Well, I'd better stick to a couple of areas, say, public opinion and political participation, mainly on the American side.

Student: Alright.

Professor: Let's start with the good news about political participation in the general public. Most importantly, we've gone way beyond a standard political resource model. We know a lot more about the avenues by which people come to participate, when and how participation skills and interests develop, the connections between political and nonpolitical activities, the rewards of participation, and the issue biases associated with different types and levels of participation. We also rediscovered political mobilization and how important it is in determining participation, what kinds of organizations are involved, the strategies they use, and …

Student: Excuse me for interrupting, Professor, but hasn't there been a decline in participation during this period?

Professor: Good question. There's healthy controversy about that and whether some new forms of participation—especially via the Internet—are replacing older modes. We've also looked at how changing life styles, demographics, and the seeming decline of social capital have affected participation.

Student: Well, that sounds pretty impressive. Does that mean the discipline covered all the important participation bases?

Professor: Don't I wish. People usually participate because of how they feel about some issue, but issues as such don't figure very large in our explanatory models. If issues are motivators, we should be paying more attention to what kinds of issues stimulate what kinds of people.

Student: Hmmm. Does this also mean that political scientists haven't paid much attention to activists in particular issues?

Professor: Yep. These people are the cogwheels of issue activism. Yet there are very few studies of domain-specific issue activists, their composition, how they're recruited, their motives, what they actually do, and their movement from issue to issue.

Student: Sounds like you're getting pretty close to what my sociology professor called social movements? Weren't political scientists looking at these movements?

Professor: You would think so, wouldn't you? Strangely enough, we have largely left the study of social movements, big and small, to other disciplines. Most of these movements have evolved into, or originated from, political action. They have had enormous impact on the politics of this country as well as others. But most of the theories and findings in this area come from other disciplines.

Student: Well, I guess each discipline has its own agenda. I think you mentioned the public opinion area also.

Professor: Yes, if anything I think our innovations in public opinion were greater than in participation. I'll mention one in particular. As you probably know, much of our knowledge about public opinion comes from survey research. Theoretical and methodological breakthroughs regarding the psychology of survey responses helped resolve some longstanding issues about the meaning and use of survey responses, including the controversy about non-attitudes. A key aspect of these developments was the unparalleled use by political scientists of experiments out in the “real world” and in the laboratory.

Student: So were you able to communicate those findings to a general public that's always hearing about the latest polls?

Professor: That's a tough sell, unfortunately.

Student: What would you say was a big oversight during this period?

Professor: I have to go a little beyond public opinion as such on this one. Since the end of the Vietnam War millions of immigrants from all over the world have come to this country. This influx represented the best opportunity to study the reaction to and impact of newcomers since the great immigration waves around the turn of the twentieth century.

Student: I can see why some other disciplines might be interested in that, but why political science?

Professor: I think it would have revealed a lot about the basic political values and norms of the American public and the nature of our institutional responses. Resettling in a new country has inherent political components, including such diverse topics as regulations about who may enter, citizenship criteria, tolerance on the part of the hosts, and the acquisition of political norms on the part of the newcomers.

Student: As you said, though, there are all kinds of immigrants, so …

Professor: Exactly. We would have been interested in such matters as country of origin, reasons for immigrating, perceptions of treatment by others and by government agencies, knowledge attainment, the desire for citizenship, and how they organized themselves politically.

Student: And what about how the country responded?

Professor: Sure, we could have looked at such topics as the content of media coverage, opinions and behaviors by the general public, leadership cues, actions by the relevant government agencies, and what impact the newcomers had on the workings of the political system.

Student: I have to go to my next class now. Thanks for your time, Professor. Overall, you seemed to talk more about the downside than the upside. Does that mean you don't feel that the discipline did very well during this time?

Professor: No, not at all! It just means our really good work inspires more ideas about what could have been, and what might yet be … By the way, could I see your paper when you're finished?

(Listener thinks: “I bet you want to see how you compare with the other professors.”)

What we have learned from multiple challenges to the study of political science

The last half century of theoretical and empirical research in political science has witnessed immense intellectual turmoil over the appropriate foundations for our discipline and for policy advice based on our discipline. In the 1950s political scientists studied how executive, legislative, and judicial structures were organized at national, state, and local levels and the role that various positions in these structures held. Every student of political science in the United States would have studied “how a bill becomes a law” at least five times during their schooling. Graduate students studying public administration were taught core “principles of administration,” which stressed the importance of strong bureaucratic control in the governance of modern democracies. Those studying comparative government would have learned the names and political orientations of core parties and how to compare the structure of upper and lower houses in major European countries, some Latin American countries, and perhaps a few in Asia. Africa was still the dark continent for most American political scientists—at least until the Peace Corps had returned many future political scientists, and African Studies programs were initiated in several major universities. Political theorists would have studied the lives, loves, and miscellaneous thoughts of diverse political philosophers.

Many challenges have led us to develop substantially since the turn of the last century. The first challenge to the traditional study of political science came from “the behavioral revolution,” which stressed the importance of studying the real behavior of legislators, executives, judges, and citizens rather than their idealized roles. Immediately after World War II, survey research blossomed and substantially enriched our understanding of observed political behavior. This era was also one of turmoil resulting from the strong advocacy of different “ideal systems” struggling to convince others that there was “one best way” to design all systems. Some political scientists tended to view the world through the dichotomized lenses of Marxism versus capitalism.

The 1960s brought still further challenge to the discipline of political science. Economists interested in political phenomena challenged not only the implicit conceptual model that political scientists used and the results of the behavioral movement, but the notion that order stemmed only from central direction. The early research testing these theories in metropolitan areas, in legislatures around the world, and in analyzing the bargaining between executives and legislatures provided considerable empirical evidence for the new political economy or public choice approach. Considerable research on local public economies and federal systems demonstrated that social order could emerge from polycentric systems and did not need to be imposed by a centralized system.

One of the healthiest aspects of this challenge was a new commitment to the use of theory in the design of research and real cumulative growth in our work. Instead of studying diverse voting, legislative, bureaucratic rules so as to memorize the interesting structural differences and ponder a bit about them, political scientists began to study institutions from the perspective of rational actors, who made calculated choices of how they could maximize their own utility.

In the 1980s a new research method—experimental studies in the laboratory—came into use in both economics and political science to complement the use of voting records, survey research, and qualitative analysis. For economists, the evidence from the laboratory was highly consistent with predictions about outcomes in competitive market settings for private goods. As scholars began to design experiments to study the provision of public goods, the use of common-pool resources, bargaining, and the extension of trust, however, evidence in many settings challenged the assumptions that individuals behave in all settings as if they were short-run, own-utility maximizers. In many experimental designs, individuals do contribute to public goods, do restrict their use of a resource, do make fair offers, and do trust one another. Findings from experiments conducted on major U.S. campuses were repeatedly replicated by scholars in many countries. Experiments conducted in Asia, Latin America, and Africa offered payoffs that averaged several months of local wages rather than the several hours of work normally offered on U.S. and European college campuses. Increasing the size of the payoffs did not change the basic findings. This evidence has generated another period of turmoil, as scholars have tried to understand why predictions derived from the theory of individual rational behavior are supported in some settings and not in others.

When the Berlin Wall fell in Germany and the former Soviet Union disintegrated, Marxism was shown to be a failed design. Instead of achieving the liberation of mankind, it contributed to the tragic subjection of major populations. Unfortunately, this has led some policy analysts to presume that imposing a market order is “the” solution. They did not recognize the extensive institutional infrastructure needed to support a vibrant market economy and political system that facilitates solving public problems at multiple levels in diverse ways dependent on local ecologies, cultures, and institutions. Thus some of the efforts at simple reforms of the 1990s have also failed to achieve their ambitious goals, and some have even been counterproductive. Additionally, the oversimplified view of how one creates the conditions needed for the development of a vibrant political economic system haunts us today as we witness the tragedy in Iraq.

Currently, we are more open to exploring diverse theories and their policy implications than at earlier times. Exciting developments in multiple fields—including evolutionary theory—are opening up our discipline to an opportunity for more rapid theoretical and empirical advances and a new synthesis. The synthesis involves several assumptions:

• There are multiple types of individuals who are initially attracted to different types of structures.

• In repeated settings, individuals may learn how to be successful in that setting by adopting the strategies of others who are successful.

• In repeated settings, reputations for trustworthy behavior and productive activities bring higher payoffs over time than short-term benefits forgone.

• Multiple equilibria are present in many settings.

• The pattern of achieved outcomes depends on the initial distribution of types of players, the structure of the setting in which they are located, and the information they learn about the benefits achieved by others in this setting, leading to changes over time in the distribution of strategies adopted.

Instead of presuming that effective political orders require a single type of governance structure, theory and empirical research is opening the possibility of thinking of complex designs that enhance innovation and learning as well as creating incentives better tailored to helping people solve public problems in different environments. While a new synthesis has not fully been achieved, this is an exciting time to be a social scientist engaged in a much broader search for a core theory that relates the structure of institutions to the assumptions made about individual preferences, information, and strategies. Thus in the first decade of this millennium we have a real opportunity to build a science that is more complex than either simple descriptions of political structures or simple assumptions of narrow rationality.

The new institutionalism and constitutional design

What has political science taught us of really lasting import since about 1980? I believe that there has been great progress across the board, but I do not feel qualified to comment on fields of specialization other than my own field of comparative politics.

With this proviso, let me propose a big reason for satisfaction and pride, but also disappointment. The former is the success of the “new institutionalism”; the latter is the limited practical use that has so far been made of this new scholarship. I include in the term “new institutionalism” the two strands of deductive rational-choice and inductive-empirical research and theory. What unites them, and divides them from the old institutionalism, is their commitment to systematic research methods. The old institutionalism was also clearly based on the assumption that “institutions matter”; scholars like Herman Finer and Carl Friedrich wrote about the importance of institutions with great knowledge and often acute insight. The change from old to new institutionalism took place in two steps: first, the behavioral revolution of the late 1950s introduced systematic research and testing, but also led to an emphasis on behavioral, noninstitutional, factors; then, in the late 1970s, the assumption that “institutions matter” reemerged and prompted new lines of now explicitly systematic research on political institutions.

The new institutionalism has given us strong evidence that the assumption that institutions matter and that they matter in specific ways is correct. The differences between what we knew around 1980 and now are enormous. In particular, the progress that has been made in the study of party systems, corporatism and interest groups, cabinet coalitions, executive-legislative arrangements (parliamentary, presidential, and hybrid forms), and electoral systems is extremely impressive.⁴⁵

The book that I think can be regarded as the pioneering contribution to the new institutionalism—Douglas W. Rae's The Political Consequences of Electoral Laws—was published in 1967, well before the 1980s. It was the first study of electoral systems that systematically classified and analyzed a large number of characteristics of electoral laws and election results in a large number of countries: all national legislative elections in 20 democracies from 1945 to 1964.⁴⁶

The availability of data for research on electoral systems and on political institutions more generally has improved by leaps and bounds since then. For instance, when I was working on Patterns of Democracy in the mid-1990s,⁴⁹

The empirical findings of the new institutionalism provide powerful tools for constitutional design. In particular, we now know how important two basic institutional choices are: combining a proportional (instead of plurality) election system with parliamentary (rather than presidential) government virtually guarantees what I have called a consensus form of democracy.⁵²

What is disappointing is that so little practical use has been made of this new knowledge. Today's constitution-makers are more aware of the different alternatives than their counterparts were in the past, but they often try to combine alternatives instead of making well-informed choices between alternatives. The result has been the proliferation of mixed electoral systems and semipresidential (partly presidential and partly parliamentary) forms of government—although true mixtures are very hard to fashion and the attempts to do so almost always lead to overly complicated rules and a serious lack of transparency.

Another disappointment is that the new knowledge has failed to ignite significant discussion about institutional reform in the United States. In fact, most politicians, political commentators, and even many political scientists are not aware of how deviant the American governmental system is. A prime example is our electoral college. That the United States is unique among democracies with politically powerful presidents⁵³

Where are we now that we weren't several decades ago?

Rather than selecting this or that “finding” or understanding, I'd rather speak of the general development of the discipline as a scholarly institution with ability to understand our subject area of politics and government, and then turn to an example.

One of the major features of political science is its openness to multiple approaches, methods, and theories. In the half century since the end of World War II, the discipline has drawn on such fields as psychology, ranging from social psychology to psychoanalysis; sociology in many of its manifestations; history, with particular concern with institutions; cultural studies; statistics; experimental design, and many more. Sometimes the borrowings are fruitful, other times less so. But given the complexity of politics and government, it has made sense to draw on disciplines that focus on individuals, on social settings, on historical contexts, and on ways of observing and measuring political phenomena.

The discipline's major encounter of the last few decades has been with economics—not the study of the interaction between politics and the economy, but the application of economic, largely rational choice, analysis to politics. I phrase it as an encounter because the possible results might have been rejection or capitulation. Something in between has happened. The powerful lessons and analytical perspective of the rational choice/economic paradigm have been brought into political science and in the process adapted (perhaps co-opted) to the needs of political analysis.

This has meant several things. In the first place, the subject focus of political science (which I think is the main glue that holds us together) has kept us from the trap of having research questions and answers defined by theory and method. It has also meant considering a wide range of phenomena in the framework of rational choice analysis. Political actors can not be understood as utility maximizers with any economic-benefit definition of utility. They do more than maximize material benefits. Values and culture have come to be taken seriously and have been given new understandings and meanings within the economic framework. Psychology—in such guises as attitude formation, framing, and decisional analysis—is in the mix. History and institutions—and the history of institutions—are all part of political analysis. The relationship across these various perspectives⁵⁶

Naturally, most folks have their favorite angle on analysis, and some reject one form or another out of hand. The rational choice Bolsheviks sometimes object to soft Menshevik rational choice. But the heterogeneous mix has created a discipline at once more rigorous and more open.⁵⁷

An example: Political cleavages and noneconomic issues

The request from Perspectives for comments on “what we have learned” went out to senior members of the profession who, because of their seniority, have earned the right to reminisce and be autobiographical. (They've also earned the right to—and, given Perspectives' assignment, are obliged to—make sweeping generalizations.)

When I started practicing political science at the beginning of the 1960s, the developing world and, within that world, the role of ethnicity and religion as bases of political cleavage and contestation had just been discovered. These bases of political identity and action were believed to be “primordial,” acquired at birth and not easily changed, if at all. At the same time, there was a belief in modernization, the spread of education and rationality, and the end of ideology. Fifty years later, religious and ethnic cleavages dominate politics across the world. Ideology has not died, though many adherents of one ideology have killed adherents of other ideologies.

A failure of political science? In one sense yes, but in another way, no. Our ability to understand the role of ideology, religion, and ethnicity in politics is much greater now than it was then. This is not the result of a particular seminal work, but the aggregation of much work from many perspectives. It is the result of adding an understanding of strategic choice in relation to identity, but not subsuming identity completely within a strategic framework. We have learned that identities are not primordial. They may be created. Or, more likely, preexisting identities may be invoked by the action of strategic elites seeking support in democratic elections or nondemocratic struggles for control.

The seminal addition of rational calculation to the understanding of identity—in the guise of straightforward self-interested motives for adhering to one or another identity, or the social calculations associated with cementing relations with others, or the strategic calculations of elites building support—has given sharpness to a sometimes fuzzy subject. But identity and the role it plays in political cleavage is not fully explained by such calculations. Identities that are mobilized—whether ethnic or religious or ideological—may be formed through socialization and be long-term value commitments, without the understanding of which one only has part of the story. And the content of the belief system that may be connected with the identity is consequential as well. In addition, analysis of the way in which identities are reinforced or diminished by institutional development add to the mix. Rational actor analysis seems to be more fruitful in understanding some folks' identity choices and actions than others, or some identity-relevant situations than others. Elites are more likely to calculate strategically the use of ethnic or religious identity in mobilizing voters or other supporters than are the voters or supporters themselves. The former have more information and more instrumentally at stake than the latter. But elites may have deep religious or ethnic or ideological commitments that override their strategic electoral or support-building calculations. This does not mean that political scientists have a formula for predicting or analyzing ideological, ethnic, or religious cleavages. But we have a toolbox filled with ways of coming to grips with such phenomena.

Bullish on my corner of the discipline

I am bullish on political science as I reflect on the last two decades. We have worked our way through many methodological and theoretical issues with positive results relevant to governance and politics. We are better at our craft and, by the measure of what goes on in my corner of the profession, more sensitive to applying what we find. I am aware that many departments still struggle over what is “good” political science and who knows how to do it. Yet I also know productive scholars who plug away in those settings to the benefit of the discipline.

What encourages me? My evidence is mostly anecdotal and impressionistic; very real, yet, I confess, a bit soft. Nonetheless, I start with an example that is plenty hard: the successful launching of this journal by the skillful management of its first editor, staff, and board. Many asked for an APSA-sponsored outlet that would turn political scientists loose on the issues of our time and our discipline. Perspectives on Politics has done that with articles, symposia, “catchall” perspectives, and review essays. The topics suit the declared mission of the journal. It is worth noting too that these features require supervision, not simply editing. Symposia are not usually tossed in over the transom! They require the initiative, then nurturing, of the editor and staff.

It is noteworthy that our demands were being made coincident with a quantum increase in scholarly (and other) communication, thus providing more opportunities to distribute research findings. No longer bound by the hard covers of books or the soft covers of journals, scholars have been able to connect and communicate as never before. The online journal managed by Nelson W. Polsby is but one example.⁵⁸

What are we learning? Obviously, I can only speak for those issues of greatest personal interest, central of which are the workings of the United States' separated system under conditions of split party control. The two parties sharing responsibility for governing has come to be a common feature in national and state capitals. We have even witnessed frequent changes in the configuration of the splits, five from 1993 to the present (all Democratic; all Republican; Democratic president, Republican Congress; Republican president and House, tie in the Senate; Republican president and House, Democratic Senate).

As he has in the past, David Mayhew questioned the conventional wisdom that divided government meant gridlock, providing evidence that split-party control did not apply a tourniquet to lawmaking.⁵⁹

Presidents increasingly exercise another option in split-party politics, that of constant campaigning. The vast literature on President Clinton reveals how he took advantage of his own skills and the growth in the means of communication during the 1990s. Going outside to have influence inside was the topic of Samuel Kernell's influential book, Going Public (1986). More recent works have concentrated on the so-called “permanent campaign.” A recent treatment by George C. Edwards III, casts doubt on the effectiveness of this use of the “bully pulpit.”⁶²

Then there is the whole matter of how it is that both parties come to win majorities in separated-system elections. We have all come to rely on the work of Gary Jacobson as he demonstrated why Democrats had such a hold on the House and more recently how that lock was broken.⁶³

A still broader dimension of split-party results is that described by some as “a polarized America.” Are there deep cultural schisms that explain putting one party in charge here, another there? Again, political scientists are attending to this important separated system issue. Morris P. Fiorina and his coauthors' Culture Wars? tackles this matter in a clear and concise manner.⁶⁶

I judge these to be very positive developments. This work not only is in the best traditions of political science scholarship, but it also demonstrates that quality is not adversely affected by relevance. Granted, there remains an overlay of doubt that the system should work this way, even arguments that split-party government is a condition to be corrected. Never mind. We are learning much about how governing works in a separated system, from how voters think and choose to how laws are enacted and implemented. Perhaps there is a second David Truman among us to write The Governmental Process in a New Century.

I am not without disappointments. My hopes for an American Political Science Reviewer, a journal devoted to reviewing scholarly communication (not just books), have not been realized. In fact, demands for more relevant articles and symposia in a third APSA journal may have come at the cost of having more book reviews. A check of my last year as editor of the American Political Science Review (1981) shows more books reviewed then than in the last four issues of Perspectives on Politics (by some 50 books, 80 if notes on reference books in 1981 are included). It was in my last year's report as editor that I suggested a separate journal for reviewing scholarly communication.

I also remain anxious for political scientists to do more cross-institutional research of lawmaking and execution, based on the notion that statutes are but the beginnings of law and the features of lawmaking and execution are the bases of cross-national comparison. It strikes me that we are very close to making such moves, given the contributions of historical work on policy and institutions as well as greater appreciation for comparison among Americanists. Yet my recent sorting of book reviews shows most works on institutions, including several cited above, focus on one branch only. The founders separated and, seemingly, so still do we.

Political scientists as gatekeepers: Overcoming inequality in our own backyard

In this essay I focus on what we can and should do to overcome racial, gender, and ethnic inequality in our own backyard—our individual departments. Consistent with the norms and expectations of our roles, I suggest that we as political scientists can do much to confront our past failures in this regard through determined efforts to recruit, retain, and promote women and minorities as faculty and graduate students.

What stirred my interest in writing this essay is the recent report by a task force of distinguished political scientists entitled “American Democracy in an Age of Rising Inequality.”⁶⁷

But the report has shortcomings, particularly in that it fails to suggest what political scientists can or should do to help overcome such inequalities. Here, I'd like to suggest our next step. As I stated in my 1993 presidential address, I, like others, consider the problem primarily a moral issue—a matter to which I return later. Here I offer more practical arguments.

The report directs the collective force of our research and attention to inequality issues relating to race, ethnicity, gender, and poverty that have plagued our profession and nation from their beginnings—issues about which we knew but that we generally overlooked or marginalized in our research and teaching.⁶⁹

Overall, I find the report exciting, challenging, and encouraging. It is of special personal and professional importance to me as an African American political scientist. It brings to the fore disparaging inequalities I have experienced firsthand and have observed throughout my entire life. My experience, and that of countless others, suggests the continuing significance of race and color regardless of class, gender, and sexual orientation. It might also prove instructive to those who overlook or tend to minimize, as the report and others seem to do, the fact that “the class divide tends to aggravate the racial divide because of the strong link[age] between the two.”⁷⁰

Though I remain excited by the overall thrust and promise of the task force report, I view it from a history of reasoned skepticism. Minorities and women have long experienced how the glowing rhetoric and promising expectations of similar reports and important actions sooner or later give way to the harsh realities and tempering effects of political-social factors in public and private arenas.