Enriching Public Discourse:
Policy Analysis in Representative Democracies
All governments produce policy analysis. It has always been predictive and value-laden. Consider the story in Genesis of Joseph, perhaps the first professional policy analyst celebrated by name. He predicted seven years of plenty followed by seven years of scarcity in interpreting the Pharaoh's dreams. The Pharaoh readily appreciated the value, both for himself and his people, of the policy of stockpiling during the years of plenty to avoid famine during the years of scarcity, and he delegated all his authority (save the throne itself) to Joseph so he could implement it. Since the Enlightenment, the basis for prediction has become more rational and the consideration of social values has become more explicit. The emergence of representative government has reduced the likelihood that any policy analyst will ever enjoy Joseph's immense influence and power. It has also resulted in a proliferation of roles for policy analysis. The project of designing institutions to promote the good society ought not ignore these roles.
Yet the task of assessing the social desirability of the possible roles for policy analysis is complicated because it cannot be isolated from the context of particular political regimes. In order to assess appropriate roles for policy analysis in the good society, one must assume at least the general form of the political institutions of the good society. Would these forms be radically different from those we observe in mature representative democracies today? Several considerations make me think that they would not. On the one hand, the Western democracies have done quite well overall during the last century. Although there have been some setbacks, they have generally expanded the politically enfranchised proportions of their populations, defended themselves from aggressive totalitarian states, protected human rights, facilitated greater health, wealth, and longevity, and improved the quality of the physical environment. I thus agree with Karol Soltan that, as Madisonian constitutionalism and capitalism have been successes, there is no need for a Third Way (Soltan, 1996: 78-79). On the other hand, there are certainly limitations to the performance of representative government. At the deepest theoretical level, the problems of social choice implied by Arrow's General Possibility Theorem argue that no democratic system can be relied upon to reveal consistently the "will of the people" (Riker, 1982). With respect to representative government, the problems of collective action (Olson, 1973, 1982) and the costs of monitoring representatives suggest it is unlikely that institutions can be discovered that guarantee equal voice to the various interests in society. The successes of representative democracies, coupled with the inherent theoretical limits to their perfection, lead me to assume that the good society would have a political regime not too dissimilar from those found in Western Europe and the United States. To facilitate a more concrete discussion, I take the contemporary political institutions of the United States as the general context for considering the role of policy analysis in the good society. 2
Within this context, I make the following arguments. First, policy analysis can improve public discourse by contributing policy alternatives to the political process, by providing better predictions of the consequences of proposed policies, and by making explicit arguments for the consideration of the full range of social values, especially those that tend to be underrepresented in representative democracy. Second, institutions to promote these contributions can be imagined. Indeed, some already exist. Third, although greater participation in policy analysis is desirable, the most socially beneficial institutional changes will be those that expand participation beyond the most prominent stakeholders. Finally, a professionalization of policy analysis that promotes the virtues of analytical integrity, humility, patience, and fortitude would enable policy analysts to play more effective roles in promoting the good society.
What Policy Analysis Offers
Policy analysis takes many forms. Its essence as professional craft, however, lies in the systematic comparison of alternative policies in terms of social values. Neither policy alternatives nor relevant social values always arise fully formed in public discourse. At least from a consequentialist perspective, assessing policy alternatives requires prediction of what effects the alternatives [End Page 61] are likely to produce and how these effects impinge upon social values.
Ideas for policy alternatives come from many sources. Stakeholder organizations and political representatives frequently propose policy changes, usually to further the interests of their constituencies. Academics and policy intellectuals sometimes apply theories, analogies, or simple cleverness to inject ideas about different policy approaches into public discourse. These contributions typically lack the detailed specification needed to make them either immediate candidates for political choice or amenable to systematic assessment. For example, the relatively simple idea that electromagnetic spectrum for personal communication devices could be allocated efficiently by auction required theoretical work by many economists and 130 pages of regulatory rules before it was used with success by the Federal Communications Commission (McAfee and McMillian, 1996). Policy analysts, who themselves sometimes originate novel policy ideas, play a valuable role in translating public discourse about possible policy changes into fully specified policy proposals.
Policy analysts often inform public discourse by making arguments about what social values are relevant to particular policy contexts. Stakeholders and other organized interests can usually inject values consistent with their goals and beliefs into public discourse; unorganized interests and the general public are often silent. An important role for the policy analyst is to make sure relevant values not raised by organized interests receive consideration. Specifically, policy analysts can speak for "silent losers" (Weimer and Vining, 1999: 156-7). Some silent losers do not speak because they do not follow policy debates. Others remain silent because they do not anticipate being affected by policies, or because they do not connect the harm they suffer to specific policies. Some, such as those in future generations, cannot speak for themselves. 3
One of the most persistent normative attacks against rationalist policy analysis challenges its foundations in welfare economics and its emphasis on the value of efficiency. These attacks, however, suffer from the nirvana fallacy in the sense that they implicitly contrast a world in which the institutions of representative government work perfectly to a world in which policy is decided solely on the basis of cost-benefit analysis. A more appropriate comparison recognizes the imperfection of representative government and the limited influence of rationalist policy analysis within it. Concentrated and organized interests enjoy much more political influence in representative government than do diffuse and unorganized interests. Economic efficiency takes account of the costs and benefits accruing to these diffuse interests. As economic efficiency rarely has a vocal constituency, policy analysts play an important role in giving it a voice, which also often contributes to distributional values by speaking for the less well-represented interests (Vining and Weimer, 1992). At the same time, the limited influence of rationalist policy analysis suggests that speaking for efficiency is unlikely to overcompensate for the tendency of representative government to give disproportionate voice to concentrated and organized interests.
Specifying policy alternatives and arguing for the consideration of a full range of values in their assessment are inherently eclectic and discursive processes—ideas and values come from a great variety of sources and evolve through discovery, exchange, contemplation, and argument (Weimer, 1998). These processes constitute much of the art and craft of policy analysis. Yet policy analysis borrows heavily from the sciences to predict.
Prediction is essential, difficult, and imperfect. Sometimes policies can be assessed solely in terms of their means, especially if the means conflict with some important value such as constitutionality or human rights, but in most circumstances it is necessary to predict their consequences to assess whether they are desirable. Both the methods and substantive findings of the social sciences provide resources for prediction. Taking full advantage of social science knowledge does not guarantee a good prediction, but it increases the chances of a better prediction than would otherwise be made.
Policy analysts can make a contribution to debate by encouraging a more dynamic view of the world. Government programs often create incentives for people to change their behaviors. Sometimes the behavioral changes are anticipated and, indeed, intended. Yet often, without consideration of knowledge drawn from the social sciences, the behavioral changes will not be anticipated and therefore not given consideration in the assessment of alternative policies. A more dynamic perspective can also have relevance for assessing the desirability of alternative states of the world. For example, recognizing that between 25 and 40 percent of the U.S. population changes income quintiles each year (McMurrer and Sawhill, 1998: 33) may alter our normative assessments of changes in the distribution of income. More generally, policy analysts make an important contribution when they encourage a more nuanced and longer-term perspective on social problems than would otherwise result from the exigencies of electoral politics. [End Page 62]
Institutions for Promoting the Contributions of Policy Analysis
Arguments in the guise of policy analysis fill the public arena. Most stakeholder and issue-oriented organizations produce assessments of social conditions and proposals for improving them; government agencies produce analyses and research reports relevant to their missions. Sometimes these products are predictively neutral in the sense that they make plausible assumptions in applying appropriate methods to predict the consequences of the status quo and alternative policies. They may also be balanced in the sense of considering all the major consequences. Organizational interests, however, often push these products away from predictive neutrality and balance.
Lack of neutrality and balance in individual analyses may sometimes be compensated for in their aggregation in open policy forum (Jenkins-Smith, 1990: 99-101). In dealing with complex issues, however, the public and its representatives may not have sufficient expertise, information, time, or inclination to delve sufficiently deeply into the analyses to reconcile their differences. The analyses then become conflicting voices in public discourse. No analyses may come through with clear messages, and those that do are likely to be the ones with more skillful sponsors or some special media appeal, say because they provide interesting, but not necessarily balanced, stories with heroes, villains, and innocent victims (Stone, 1988: 108-126).
Are there no organizational settings that consistently promote predictively neutral and balanced policy analysis? Legislative settings appear to provide the possibility for institutionalizing neutral and balanced policy analysis. Legislatures generally operate at an informational disadvantage relative to executive agencies, providing a motivation for creating analytical offices. Anticipating that different parties will command majorities in the legislature at different times, those who lead the analytical offices have strong incentives to be neutral and balanced, and to be perceived as neutral and balanced. As Elizabeth Hill, the California legislative analyst, notes: "The viability of the office is dependent on its ability to provide independent analyses based on reliable information on an ongoing basis" (Hill, 2001). The office she heads has existed since 1941, surviving many changes in legislative control. It also served as the model for the Congressional Budget Office, which has played a similar role at the federal level since 1975, and many state-level analytical offices. At the federal level, most observers would agree that the Congressional Research Service also provides neutral and balanced analyses.
Although analytical offices in the executive branch may also at times develop reputations for neutrality and balance, they face strong pressures to advance the missions of their agencies and the policies of the current administration. Rather than abolish analytical offices that served the previous administration well, new administrations are likely to try to use them to promote their interests.
Requirements for the routine production of certain types of analysis may help push analytical offices toward neutrality and balance. In particular, requirements for cost-benefit analysis may play such a role because, as argued above, cost-benefit analysis forces consideration of diffuse interests that often remain silent. Early in the last century Congress required that the Army Corp of Engineers conduct cost-benefit analysis of its projects. Executive Order 12291 issued by President Reagan in 1981, and reaffirmed in Executive Order 12866 by President Clinton in 1994, requires the application of cost-benefit and distributional analyses to major regulatory initiatives. In 1993, a blue ribbon panel of social scientists assembled by the National Oceanic and Atmospheric Administration set out guidelines for the use of contingent valuation surveys to estimate the value of environmental changes in cost-benefit analyses (Arrow et al., 1993). Over the long run, the imposition of such analytical requirements and the development of norms to implement them encourages the use of acceptable methodological approaches, which in turn increases the likelihood that policy alternatives will receive neutral and balanced assessments.
Broadening Participation in Policy Analysis
Broader participation in policy analysis offers several advantages. First, conversations about public policy increase the chances that analysts will discover the full range of relevant values. Second, depending on who is involved in the conversations, they may also help analysts structure arguments about what weights the relevant values should receive in decision-making. Third, the conversations also have potential for contributing to a more informed and analytically sophisticated public discourse on policy issues.
A number of scholars have argued for the desirability of expanding participation in policy analysis. John Dryzek (1990, 2000) wishes to embed policy analysis within a discursive democracy that promotes communicative over instrumental rationality. Efforts to bring this idea closer to practice have been [End Page 63] through proposals for participatory policy analysis (Durning, 1993; deLeon, 1994). I suspect, however, that these efforts will generally not be effective in broadening participation much beyond stakeholders. If this is the case, then we should not expect participatory policy analysis to produce dramatic improvement over the political processes in which these stakeholders already have voice. For example, in the regulatory arena, there has been considerable experimentation with negotiated rule-making that seeks to engage stakeholders and representatives of the public in discourse on the content of regulations. It appears, however, that negotiated rule-making performs no better than traditional notice-and-comment procedures (Coglianese, 1997). It is simply too optimistic to believe that reasoned argument can generally produce a consensus over policy alternatives among stakeholders. Assuming it does may undercut good policy analysis. As Duncan MacRae Jr. (1993) notes, the strategies for effective advocacy discourse are the refutations of the normative guidelines for good policy analysis in consensual discourse.
In designing and redesigning institutions, it is desirable to find ways of expanding participation in policy analysis beyond well-organized stakeholders. The Internet may offer this possibility. Its potential lies in its dramatic reduction in the costs of finding, accessing, and transmitting information. By lowering these costs, many more people can be drawn into participating in policy analysis.
The opening up of policy analysis through the Internet is already beginning to happen. Many federal and state agencies post research, policy reports, and regulatory dockets on their web sites. Though few sites now seek or permit responses from those who view the materials, encouraging two-way communication seems like a natural next step. In the good society, I envision sunshine laws that, absent specific reasons for confidentiality, would require government agencies routinely to post policy analyses and regulatory dockets on the Internet and to collect comments from those who access them.
I do not expect participating in policy analysis to be a popular leisure activity—most people would rather spend time with friends and family than engage in public discourse. Nevertheless, analyses might draw attention from professors, students, policy wonks, irrepressible report card makers, and random browsers, as well as from those mobilized by stakeholders. Not all of this participation will be informed or usefully informing. Nonetheless, by opening up their analyses to greater scrutiny, it is quite possible that analysts would feel pressure to conform more closely to norms of predictive neutrality and balance.
Encouraging Professional Virtues
A few years ago, I heard a professor from a prestigious university assert that the main contribution of its curriculum in policy analysis was to give its students a six-month head start in the bureaucracy. Certainly, policy analysts play many roles in the public sphere, including serving as advocates for their organizations—the general recognition of these diverse roles is what Beryl Radin (2000) refers to as policy analysis coming of age. And further, those wishing to affect policy are well advised to understand the political arenas in which decisions are made. Yet in the good society, analysts would not only advance the interests of their organizations, but they would be virtuous in ways that contributed to productive, genuine, and civil public discourse.
One virtue, analytical integrity, seems to me to be central to the professional policy analyst. Although I readily admit that the sciences are influenced by social constructions, I reject the extreme post-positivist view that these constructions render standards for empirical research meaningless. We may disagree, say, about which laws of probability apply to a particular problem, but it would be misguided to reject the laws themselves. Although not constant, at any time there are widely accepted conventions in the empirical social sciences about the proper uses of evidence and the appropriate bases for prediction. Widely held norms of analytical integrity reduce conflict over the prediction of consequences, leaving more room for discussion of values and the search for creative solutions to social problems. To foster development of the norm of analytical integrity, training in policy analysis should include substantial exposure to research methods, which will likely give new analysts something more valuable to society than a six-month head start in their careers.
Commitment to analytical integrity should be tempered by humility. As even the best methods are inherently imperfect, analysts should be explicit about uncertainties while seeking to use their skills to reduce them. They should have fortitude to protect analytical integrity from its most egregious assaults in the name of expediency—in the words of Aaron Wildavsky, they should be willing to speak truth to power. Finally, analysts should have patience. Analysis need not lead to the immediate adoption of a desirable policy alternative to play a positive social role. It can help facilitate the enlightenment function of social research by connecting social science research to public policy issues (Weiss, 1977); it can help in the interpretation of issues and the justification of actions (Shulock, 1999); it can introduce [End Page 64] policy ideas that may resurface sometime in the future (Thomas, 1987) when circumstances open policy windows favorable to their consideration (Kingdon, 1984); and it can contribute to the gradual accumulation of evidence that eventually leads to a change in policy (Derthick and Quirk, 1985).
How can these virtues be encouraged? Those of us who train policy
analysts can help by making sure that those entering the profession have
reflected upon how their work can contribute not only to the adoption and
implementation of better public policies, but also to more informed public
discourse about public policy. Perhaps we should also seek to involve
policy analysts more in the intellectual project of understanding the
role of institutions in achieving the good society in order to encourage
them to take broader and longer-term perspectives. Ultimately, however,
these virtues are unlikely to thrive unless they are reinforced by
institutions that promote predictive neutrality and balance.
David L. Weimer is professor of political science and public affairs, University of Wisconsin, Madison.
Arrow, Kenneth, et al. "Report of the NOAA Panel on Contingent Valuation," Federal Register 58 (1993): 4601-4614.
Coglianese, Gary. "Assessing Consensus: The Promise and Performance of Negotiated Rulemaking," Duke Law Journal 46 (1997): 1255-1349.
deLeon, Peter. "Democracy and the Policy Sciences: Aspirations and Operations," Policy Studies Journal 22 (1994): 200-212.
Derthick, Martha and Paul J. Quirk. The Politics of Deregulation. Washington: Brookings Institution, 1985.
Dryzek, John S. Deliberative Democracy and Beyond: Liberals, Critics, Contestations. New York: Oxford University Press, 2000.
Dryzek, John S. Discursive Democracy: Politics, Policy, and Political Science. New York: Cambridge University Press, 1990.
Durning, Dan. "Participatory Policy Making in a Social Service Agency: A Case Study," Journal of Policy Analysis and Management 12 (1993): 297-322.
Hill, Elizabeth. Personal Communication, March 1, 2001.
Jenkins-Smith, Hank C. Democratic Politics and Policy Analysis. Pacific Grove: Brooks/Cole Publishing Company, 1990.
Kingdon, John W. Agendas, Alternatives, and Public Policies. Boston: Little Brown, 1984.
Olson, Mancur. The Rise and Decline of Nations: Economic Growth, Stagflation, and Social Rigidities. New Haven, Conn.: Yale University Press, 1982.
Olson, Mancur. The Logic of Collective Action. Cambridge, Mass.: Harvard University Press, 1973.
MacRae, Duncan, Jr. "Guidelines for Policy Discourse: Consensual versus Adversarial." In Fischer, Frank and John Forester, eds. The Argumentative Turn in Policy Analysis and Planning. Durham, N.C.: Duke University Press, 1993.
McAfee, R. Preston and John McMillian. "Analyzing the Airwaves Auction," Journal of Economic Perspectives 10 (1996): 159-175.
McMurrer, Daniel P. and Isabel V. Sawhill, Getting Ahead: Economic and Social Mobility in America. Washington: The Urban Institute Press, 1998.
Radin, Beryl A. Beyond Machiavelli: Policy Analysis Comes of Age. Washington: Georgetown University Press, 2000.
Riker, William H. Liberalism Against Populism. San Francisco: W. H. Freeman and Company, 1982.
Riker, William H. and David L. Weimer. "The Political Economy of Transformation: Liberalization and Property Rights." In Banks, Jeffrey S. and Eric A. Hanushek, eds. Modern Political Economy. New York: Cambridge University Press, 1995.
Soltan, Karol Edward. "Institution Building and Human Nature." In Soltan, Karol Edward and Stephen L. Elkin, eds. The Constitution of Good Societies. University Park: Pennsylvania State University Press, 1996.
Shulock, Nancy. "The Paradox of Policy Analysis: If It Is Not Used, Why Do We Produce So Much of It?" Journal of Policy Analysis and Management 18 (1999): 226-244.
Stone, Deborah. Policy Paradox and Political Reason. Boston: Scott, Foresman, 1988.
Thomas, Patricia. "The Use of Social Research: Myths and Models." In Blumer, Martin, ed., Social Science Research and Government: Comparative Essays on Britain and the United States. New York: Cambridge University Press, 1987.
Vining, Aidan R. and David L. Weimer. "Welfare Economics as the Foundation for Policy Analysis: Incomplete and Flawed but Nevertheless Desirable," Journal of Socio-Economics 21 (1992): 25-37.
Weimer, David L. "Policy Analysis and Evidence: A Craft Perspective," Policy Studies Journal 26 (1998): 114-128.
Weimer, David L. and Aidan R. Vining. Policy Analysis: Concepts and Practice 3rd Edition, Upper Saddle River, N.J.: Prentice Hall, 1999.
Weiss, Carol H. "Research for Policy's Sake: The Enlightenment Function of Social Research," Policy Analysis 3 (1977): 531-545.
1. I thank Brian Cook, Noah Pickus, and Graham Wilson for thoughtful, if not completely heeded, comments on the first draft of this essay.
2. While it is possible to speculate generally about socially valuable roles for policy analysts in any particular type of regime, it is only possible to consider institutions for furthering these roles in the context of the specific political, social, and economic institutions of a particular regime (Riker and Weimer, 1995).
3. Future generations may also be "silent winners" to the extent that we bequeath a larger capital stock. As an important component of the capital stock, knowledge, is rapidly growing and does not depreciate, it is quite likely that future generations will be wealthier than ours as long as we avoid truly catastrophic events.