Rawls and the Theory of Justice

Rawls' work offered a new theoretical foundation for a liberal polity, and for a generation political theorists have organized their research both to challenge its core claims and to work out the implications of the theory for contemporary policy. In the grip of the economistic criterion of Vilfredo Pareto, welfare could only be seen as increased in a society if and only if a change in policy would make no one less well-off, and at least one person better off. Rawls insisted that this principle of welfare cannot itself constitute a conception of justice; it must be supplemented by a theory that explains how, over time, an initial distribution of rights and assets (natural and social) is to be determined. This theory, Rawls argued, is best approached by the heuristic of the “original position,” in which a rational individual would select principles of justice from behind a veil of ignorance, unaware of his or her future standing in society. Two principles were theorized to emerge: first an equally distributed set of basic liberties and second, genuine equality of opportunity. As a corollary to the second, and most provocatively, social and economic policies (without contradicting the previous points) would target their greatest benefits to the least advantaged members of society. These principles spawned a generation of debate and refinement, a process that remains vibrant. For example, A. K. Sen and G. A. Cohen better specified criteria for equality. Along similar lines, Brian Barry and Philippe van Parijs have normatively justified radical egalitarian claims not on the communitarian assumptions as have many Marxists, but rather on Rawls' individualist assumptions.

Perhaps the most visible aspect of this research program in current research is on the issue of multiculturalism. In Rawls' “original position” there are only individuals, without history and without culture. They are asked to choose rules for a society that will inevitably be culturally diverse. Many normative theorists, led by Will Kymlicka, now ask if there is a way to insert culture into a theoretical tradition that denies its relevance to questions of justice. These theorists argue that cultural belonging is necessary for individual freedom, and thus ought to be an integral part of the ethical individualism at the heart of Rawlsian liberal theory. However, these culturalists reject the conservative Burkian model of culture, one that accepted traditional but illiberal practices; rather these culturalists seek to imagine a liberal society based on Rawlsian principles, yet in which individuals can be recognized as members of a “societal culture.” Rawls sought to modify his theory in regard to this critique, much to the chagrin of the universalists and egalitarians who saw this move as a threat to individual liberty (especially to those incarcerated in their own cultures), a diversion from a focus on redistribution, and an undesirable concession to cultural relativism. This debate illustrates the contemporary vibrancy of the normative element in political science, with clear implications for public policies having to do with inter alia welfare, language, and redistribution.

Black and the “Median Voter Theorem”

The median voter theorem has a simple and beautiful elegance. Given single-peaked preferences along some policy dimension and choice along that dimension made by majority rule, the winning policy will be the ideal point of the voter who has an equal number of fellow-citizens on his “left” and on his “right.” This powerful idea drove Anthony Downs to his Economic Theory of Democracy in which he demonstrated (with a set of plausible assumptions) that political parties in a “first past the post” electoral system would rationally move their platforms towards the preferred point of the median voter. The implications of this work have blossomed such that ever new questions, both theoretical and empirical, drive new research.

Research today, to choose but one strand of this research program, addresses the question of which electoral systems are best able to satisfy the ideal point of the median voter. In the Downsean framework, the answer was obvious. We already knew that a first-past-the-post plurality system, as Duverger's law told us (and Gary Cox's recent work has provided theoretical foundations for), would yield two parties. Downs' model would lead us to expect both parties to converge from opposite directions toward the ideal point of the median voter. However, recent work associated with G. Bingham Powell, Jr., John Huber, and Josep Colomer has shown that the voice of the median voter gets lost in the transition from elections to legislative bargaining. In real legislatures, this research shows, proportional representative systems yield policies that are consistent with the median voter, at least more so than in plurality systems. Deriving the implications of varying electoral systems in their ability to legislate in a democratic manner is thus a core research activity in contemporary political science.

Many other questions that are at the core of democratic theory and take off from the median voter theorem remain on the current agenda of political scientists. One of them has to do with the role of political parties, and the extent to which if they get a majority position in the legislature they can push public policy toward their own median member and away from the legislative median. A second has to do with the role of congressional committees, and the extent to which they can monitor agencies under their purview in a manner that serves the interest of their median committee member at the expense of the interests of the legislative median. A third has to do with the question of the endogenous selection of electoral institutions, and how far distinct selection processes deviate from the interests of the median voter. A fourth has to do with the question as to whether economic growth (because of fewer rents) is enhanced the closer public policy comes to the wishes of the median voter. A fifth has to do with the question of complex coalitions and minority governments in multiparty parliamentary systems, and the extent to which they represent the median voter. Here, a plethora of scholars collaborating across the Atlantic (Michael Laver, Kaare Strom, Norman Schofield, and others) has shown for a variety of issue areas a surprising connection between the ideal points of the median voter and minority and coalition government policies. All of these questions are now being asked in papers appearing in our leading journals. While these papers employ arcane models and complex statistical tests, they are all part of a common research program that addresses fundamental questions in democratic theory.

Rokkan's Comparative Project

In the classic corpus of Stein Rokkan and his collaborators, the pathways to the modern democratic state were explored systematically through complex matrices and branching trees. Karl Deutsch and his collaborators envisioned a program in which quantitative data would be collected such that these matrices and trees could be reduced to statistical description. In his early papers advertising the potential of this imagined method, Deutsch's datasets were filled with missing values; worse, even if he had the datasets he dreamed of collecting, he did not have the statistical programs (or theories to address the inevitable problems of endogeneity) to analyze them.

Political scientists today have reinvigorated the program envisioned by Rokkan, Lipset, and Deutsch to address with new methods the historical and comparative questions that they posed. First is the question of what differentiates countries that become democratic from those that do not. A project headed by Adam Przeworski partly confirms but substantially modifies Lipset's classic study linking modernization and democracy in Political Man. The Przeworski team relied on extensive data covering a half century to demonstrate that while democracies arise almost randomly in a cross-section of countries, they are more likely to survive once a country passes a GDP per capita threshold. Carles Boix, in advancing this program, has provided new evidence suggesting that over the past two centuries, democracies aren't randomly generated, but emerge under conditions of relative economic equality when assets become movable.

Other political scientists address not the foundations of democracy, but rather the foundations of political order: what countries maintain political order as opposed to facing civil war and/or state collapse? Several research teams, using newly constructed country/year datasets, have come to the conclusion that economic wealth and state capacity to cauterize insurgency are the factors that best assure order. What is interesting about these studies (along with those on the emergence of democracy) is that variables having to do with “civilizational divides” or “culture” have had little or no explanatory power. Thus, although in the Rokkan tradition (in seeking long-term historical patterns for the nation-state) religious and other cultural variables play a key role, recent studies of democracy and civil war have found little significance in these variables. While it is doubtful that cultural and historical variables, once they reappear in our statistical models in more subtle forms, will remain without significance, this work represents a scientific advance in a way that Rokkan would have appreciated.