CHAPTER 6
The Policy Analysis Movement
MICHAEL MINTROM
Since the mid-1960s, an increasingly large number of people have come to devote their professional lives to producing policy analysis. This is a global phenomenon, although the intensification of activities associated with policy analysis has been most pronounced in the United States. Here, I term this amassing of personnel and resources the policy analysis movement. Use of the term is intended to imply a deliberate effort on the part of many people to reconceive the role of government in society and renegotiate aspects of the relationships that exist between individuals, collectivities, and governments. However, the term should not be taken to imply either consistency of purpose or a deliberate striving for coordination among producers of policy analysis. While not directly comparable in a political sense with other social movements, the policy analysis movement has been highly influential. It has served to transform the advice-giving systems of governments and, as a consequence, challenged informal yet long-established advising practices through which power and influence flow. The profundity of this transformation has often eluded the attention of social and political commentators. That is because the relevant changes have caused few immediate or obvious ruptures in the processes and administrative structures typically associated with government or, more broadly speaking, public governance.
Early representations tended to cast policy analysis as a subset of policy advising. As such, policy analysis was seen primarily as an activity conducted inside government agencies with the purpose of informing the choices of a few key people, principally elected decision-makers (Lindblom 1968; Wildavsky 1979). Today, the potential purposes of policy analysis are understood to be much broader. Many more audiences are seen as holding interests in policy and as being open to – indeed demanding of – appropriately presented analytical work (Radin 2000). Beyond people in government, people in business, members of non-profit organizations, and informed citizens all constitute audiences for policy analysis. While policy analysts were once thought to be mainly located within government agencies, today policy analysts also can be found in most organizations that have direct dealings with governments, and in many organizations where government actions significantly influence the operating environment. In addition, many university-based researchers, who tend to treat their peers and their students as their primary audience, conduct studies that ask questions about government policies and that answer them using forms of policy analysis. Given this, an appropriately encompassing definition of contemporary policy analysis needs to recognize the range of topics and issue areas policy analysts work on, the range of analytical and research strategies they employ, and the range of audiences they seek to address. In recognizing the contemporary breadth of applications and styles of policy analysis, it becomes clear that effective policy analysis calls for not only the application of sound technical skills, but also deep substantive knowledge, political perceptiveness, and well-developed interpersonal skills (Mintrom 2003). Although producing high-quality, reliable advice remains a core expectation for many policy analysts, advising now appears as a subset of the broader policy analysis category. The transition from policy analysis as a subset of advising to advising as a subset of analysis represents a significant shift in orientation and priorities from earlier times. In what follows, I first review the sources of increasing demand for policy analysis. I then review the growth and adaptation in the supply of policy analysis that have occurred in response to this demand. I conclude by discussing the current state of the policy analysis movement and its likely future trajectory.
The Evolving Demand for Policy Analysis
Demand for policy analysis has been driven mostly by the emergence of problems and by political conditions that have made those problems salient. Early in the development of policy analysis techniques, the people who identified the problems that needed to be addressed tended to be government officials. They turned to academics for help. Frequently, those academics deemed to be most useful, given the problems at hand, were economists with strong technical skills, who had the ability to estimate the magnitude of problems, undertake statistical analyses, and determine the costs of various government actions. During the twentieth century, as transportation, electrification, and telecommunications opened up new opportunities for market exchange, problems associated with decentralized decision-making became more apparent (McCraw 1984). Meanwhile, as awareness grew of the causes of many natural and social phenomena, calls emerged for governments to establish mechanisms that might effectively manage various natural and social processes. Many matters once treated as social conditions, or facts of life to be suffered, were transformed into policy problems (Cobb and Elder 1983). Together, the increasing scope of the marketplace, the increasing complexity of social interactions, and expanding knowledge of social conditions created pressures from a variety of quarter for governments to take the lead in structuring and regulating individual and collective action. Tools of policy analysis, such as the analysis of market failures and the identification of feasible government responses, were developed to guide this expanding scope of government. Yet as the reach of policy analysis grew, questions were raised about the biases inherent in some of the analytical tools being applied. In response, new efforts were made to account for the effects of policy changes, and new voices began to contribute in significant ways to policy development. To explore the factors prompting demand for policy analysis, it is useful to work with a model of the policymaking process. A number of conceptions of policymaking have been developed in recent decades. Here, I apply the ‘stages model,’ where five stages are typically posited: Problem definition, agenda setting, policy adoption, implementation, and evaluation (Eyestone 1978).
Initial demands for policy analysis were prompted by growing awareness of problems that government could potentially address. Questions inevitably arose concerning the appropriateness of alternative policy solutions. Thus, in the United States in the 1930s, as the federal government took on major new roles in the areas of regulation, redistribution, and the financing of infrastructural development, the need arose for high quality policy analysis. With respect to regulatory policy, concerns about the threat to the railroads of the emerging trucking industry prompted the expansion of the Interstate Commerce Commission (Eisner 1993). This body employed lawyers and economists who helped to devise an expanding set of regulations that eventually covered many industries. Although the American welfare state has always been limited compared with the welfare state elsewhere, especially European welfare states, its development still required concerted policy analysis and development work on the part of a cadre of bureaucrats (Derthick 1979). Initially, much of the talent needed to fill these positions was drawn from states like Wisconsin, where welfare policies had been pioneered by Robert LaFollette. As the role of the United States government in redistribution expanded, policy analysts swelled the ranks of career bureaucrats in the Treasury, the Office of Management and Budget, and the Department of Health and Human Services. Meanwhile, benefit-cost analysis, a cornerstone of modern policy analysis and a core component of public economics, was developed to help in the planning of dam construction in the Tennessee Valley in the 1930s. Politicians at the time worried that some dams were being built mainly to perpetuate the flow of cash to construction companies rather than to meet growing demand for electricity and flood control (Eckstein 1958). The broad applicability of the technique soon became clear and its use has continually expanded. At the same time, efforts to improve the sophistication of the technique, and to develop variations on it that are best suited to different sets of circumstances have also continued to occur (see Vining and Broadman in this volume).
Several features of policy development, the politics of agenda setting, and the policymaking process have served over time to increase demand for policy analysts. The dynamics at work have been similar to those through which an arms race generates on-going and often expanding demand for military procurements. In Washington, DC, the growing population of policy analysts employed in the federal government bureaucracy led to demand elsewhere for policy analysts who could verify or contest the analysis and advice emanating from government agencies. A classic example is given by the creation of the Congressional Budget Office (CBO). This office was established as an independent resource for Congress that would generate analysis and advice as a check on the veracity of the analyses prepared and disseminated by the Executive-controlled Office of Management and Budget (Wildavsky 1992). The General Accounting Office was also developed to provide independent advice for Congress. In the latter case, the scope of the analysis has always been broader than that of the Congressional Budget Office.
As the analytical capabilities available to elected politicians grew, groups of people outside of government but with significant interests in the direction of government policy began devoting resources to the production of high quality, independent advice. Think tanks, like the Brookings Institution and the American Enterprise Institute, established in 1920s and 1940s respectively and both still going strong, represent archetypes of many independent policy shops now located in Washington, DC (Smith 1991). These organizations engage in original research and policy analysis, the synthesis of findings from prior studies, and the marketing of policy ideas, all in the service of more informed policy debate. Among government agencies, the growing quality of the technical advice generated by some has typically promoted renewed effort on the part of others to build their analytical and advice-giving capabilities. When agencies must compete among themselves for funding of new programs – or for on-going funding of core functions –a degree of learning occurs, so that incremental improvements are observed in the overall quality of analysis and advice being produced.
Although policy analysis is often understood as work that occurs before a new policy or program is adopted, demand for policy analysis has also been driven by interest in the effectiveness of government programs. The question of what happens once a policy idea has been adopted and passed into law could be construed as too operational to deserve the attention of policy analysts. Yet, during the past few decades, elements of implementation have become recognized as vital for study by policy analysts (Bardach 1977; Lipsky 1984; Pressman and Wildavsky 1973). In part, this has been led by results of program evaluations, which have often found policies not producing their intended outcomes, or worse, creating whole sets of unintended and negative consequences. One important strain of the work devoted to assessing implementation has contributed to what is now known as the ‘government failure’ literature (Niskanen 1971; Weimer and Vining 2005; Wolf 1979). The possibility that public policies designed to address market failures might themselves create problems led to a greater respect for market processes and a degree of skepticism on the part of policy analysts towards the remedial abilities of government (Rhoads 1985). In turn, this required policy analysts to develop more nuanced understandings of the workings of particular markets. The government failure literature has led to greater interest in government efforts to simulate market processes, or to reform government and contract out aspects of government supply that could be taken up by private firms operating in the competitive marketplace (Osborne and Gaebler 1993). Where efforts to reform government have been thorough-going, an interesting dynamic has developed. As the core public sector has shrunk in size, the number of policy analysts employed in the sector has increased, both in relative and in absolute terms. This dynamic is indicative of governments developing their capacities in the management of contracts rather than in the management of services (Savas 1987). In these new environments, people exhibiting skills in mechanism design and benefit-cost analysis have been in demand. Therefore, employment opportunities for policy analysts have tended to expand, even as the scope of government has been curtailed. Fiscal conservatism, which has been a hallmark of governments in many jurisdictions over recent decades has added to this trend (Yergin and Stanislaw 1998; Williamson 1993). When budgets are squeezed, it becomes critical for all possible efficiency gains to be realized. More than any other trained professionals, policy analysts working in government are well-placed to undertake the kind of analytical work needed to identify cost-saving measures and to persuade elected decision-makers to adopt them.
Early discussions of the role of policy analysts in society often portrayed them as ‘whiz kids’ or ‘econocrats’ on a quest to imbue public decision-making with high degrees of rigor and rationality (Self 1977; Stevens 1993). Certainly, proponents of benefit-cost analysis considered themselves to have a technique for assessing the relative merits of alternative policy proposals that, on theoretical grounds, trumped any others on offer. Likewise, proponents of program evaluation employing quasi-experimental research designs considered their approach to be superior to other approaches that might be used to determine program effectiveness (Cook and Campbell 1979). That the application of both benefit-cost analysis and quantitative evaluation techniques has continued unabated for several decades speaks to their perceived value for generating useable knowledge. However, the limitations of such techniques have not been lost on critics. In the case of benefit-cost analysis, features of the technique that make it so appealing – such as the reduction of all impacts to a common metric and the calculation of net social benefits – have also attracted criticism. In response, alternative methods for assessing the impacts of new policies, such as environmental impact assessments, social impact assessments, and health impact assessments have gained currency (Barrow 2000; Lock 2000; Wood 1995). Similarly, widespread efforts have been made to promote the integration of gender and race analysis into policy development (Myers 2002; True and Mintrom 2001). In the case of evaluation studies, fundamental and drawn-out debates have occurred covering the validity of various research methods and the appropriate scope and purpose of evaluation efforts. Significantly for the present discussion, these debates have actually served to expand demand for policy analysis. Indeed, government agencies designed especially to audit the impacts of policies on the family, children, women, and racial minorities have now been established in many jurisdictions. Further, evaluation of organizational processes, which often scrutinize the nature of the interactions between organizations and their clients are now accorded equivalent status among evaluators as more traditional efforts to measure program outcomes (Patton 1997; Weiss 1998). Yet these process evaluations are motivated by very different questions and draw upon very different methodologies than traditional evaluation studies that assessed program impacts narrowly.
This review of the evolving demand for policy analysis has identified several trends that have served both to embed policy analysts at the core of government operations and to expand demand for policy analysts both inside and outside of government. These trends have much to do with the growing complexity of economic and social relations and knowledge generation. Yet there is also a sense in which policy analysis itself generates demand for more policy analysis. While these trends have been most observable in national capitals, where a large amount of policy development occurs, they have played out in related ways in other venues as well. For instance, in federal systems, expanding cadres of well-trained policy analysts have become engaged in sophisticated, evidence-based policy debates in state and provincial capitals. Cities, too, are increasingly making extensive use of policy analysts in their strategy and planning departments. At the global level, key coordinating organizations, such as the World Bank, the International Monetary Fund, the World Trade Organization, and the Organization for Economic Cooperation and Development have made extensive use of the skills of policy analysts to monitor various transnational developments and national-level activities of particular relevance and interest.
Today, the demand for policy analysis is considerable, and it comes both from inside and outside of governments. This demand is likely to keep growing as calls emerge for governments to tackle emerging, unfamiliar problems. On the one hand, we should expect to see on-going efforts to harness technical procedures drawn from the social sciences and natural sciences for the purposes of improving the quality of policy analysis. On the other hand, more people are likely to apply these techniques, reinvent them, or develop whole new approaches to counteract them, all with the purpose of gaining greater voice in policymaking at all levels of government from the local to the global.