Conducting Interpretive Policy Analysis

Dvora Yanow

• Publisher: SAGE Publications, Inc
• Year: 1999
• Print ISBN: 9780761908272
• eBook ISBN: 9781506319735
• Edition: 1
• Topic: Social Sciences
• Subtopic: Social Science Research & Methodology
• Index: Social Sciences

This book in the QRM series is designed for a wide variety of research methods courses taught in various departments. It will be of most interest to those in Public Policy, Political Science, and Public Administration departments, but will also be of interest to researchers in Sociology, Anthropology, Communication and Education departments, among others. The book fills a gap in the traditional policy analysis coverage, which is usually heavily quantitative. It will also fill a gap in the QRM series in covering the discipline of political science, which is warming to qualitative methodology…slowly.

There has been much in the journal literature in the past 15 years calling for more interpretive approaches to the study of public policy; Yanow has been in the middle of it.

1. UNDERLYING ASSUMPTIONS OF AN INTERPRETIVE APPROACH: THE IMPORTANCE OF LOCAL KNOWLEDGE

What is policy analysis? What does a policy analyst do? Where is she employed?

Common answers to these questions would be that a policy analyst researches a policy issue to advise a policymaker on some decision relative to that issue. The policy analyst provides to the policymaker information that the latter lacks and is unable, for various reasons, to obtain personally. This information has traditionally been some form of technical or other expert knowledge required to craft a policy or to assess the likelihood of and evaluate its projected outcomes, but it may also entail procedural knowledge useful for implementation activities.¹

Policy analysis has traditionally been undertaken in advance of legislative or other policy-making decisions or acts, but the sphere of activity has also extended to evaluating policies after they have been enacted, and to the evaluation of implementation activities themselves. So Fischer (1995, p. 2), for example, treats policy evaluation as the applied social scientific activity “typically referred to as policy analysis’ or ‘policy science.’” In this view, policy analysis is “designed to supply information about complex social and economic problems and to assess the processes by which a policy or program is formulated and implemented.” It can focus on a policy’s anticipated outcomes or, in retrospect, on its actual results. Ideally, as Fischer notes, we expect policy analysis to provide both policymakers and citizens “with an intelligent basis for discussing and judging conflicting ideas, proposals, and outcomes.”²

Professional training in these applied social scientific analytic modes typically takes place in graduate programs in public policy, public administration, planning (urban, city, and regional), social work, education, and other issue-focused departments or schools (such as environmental studies). Even more broadly, as Heineman, Bluhm, Peterson, and Kearny (1990, p. 5) note, physicians, attorneys, and chemists could be said to engage in policy analysis when their work shapes policy decisions. Business and management schools have long taught “business policy” as a central aspect of their curriculum. Even in its current incarnation as business “strategy,” this decision-making process follows the same steps as the policy process, even using the same terms (formulation and implementation; see Hatch, 1997, pp. 105-110). Traditional strategic analysis in the business world draws on many of the same tools as traditional policy analysis in assessing and shaping policies with respect to individual and organizational stakeholders, and these traditional approaches face many of the same limitations.

Although policy research can be university-based, it is also supported or conducted by and within independent agencies and consulting firms across the political and topical spectrum (e.g., the Rand Corporation, which also has its own training program; the American Enterprise Institute; the Brookings Institution; and Abt Associates), as well as by legislative aides for state and federal representatives and committees, and by interest groups (e.g., Sierra Club; Peace Officers Research Association of California). Private sector businesses and corporations have increasingly been employing policy analysts to monitor governmental regulations that may affect their activities.

The analytic reports of working policy analysts typically circulate in-house (within whatever agency, community group, or governmental entity commissioned the study, and perhaps among policy-relevant stakeholders). In a revised form, reports may find wider audiences through research papers presented at conferences of such professional associations as APPAM (the Association for Public Policy Analysis and Management), IPSA or APSA (International Political Science Association, American Political Science Association) and regional affiliates (especially in their public policy and public administration sections), and ASPA (American Society for Public Administration), and through more conceptual, theoretical treatments published in such journals as JPAM (Journal of Policy Analysis and Management), the Policy Studies Review, Policy Studies Journal, and Policy Sciences (not to mention conferences and journals devoted to specific policy areas). Policy analyses are also written initially (and sometimes exclusively) in an academic mode, for conference presentation and journal publication. Whereas the intended reader of “in-house” reports is the policymaker (including implementors), the intended reader of the academic paper is the research community.³

In short, policy analysis is conducted at federal, state, regional, and local levels of government; in private, public, and nonprofit realms; by liberals and conservatives; and by single interest groups and those that cover a broader spectrum. This text is written with the understanding that policy analysts work in many capacities:

• as advisers to policymakers;
• as advocates for community groups or as community organizers;
• as staff in nonprofit agencies or lobbying groups.

The text is also written with the understanding that policy analysis is a form of research, and so it speaks to those training or involved in the academic pursuit of policy research as well as to those training or involved in the practices of policy analysis.

Both analytic acts and the spheres of activity, however, have been conceptually curtailed and directed by the research and analytic methods available in training programs and acceptable to policy analysis scholars. If, for example, cost-benefit analysis is accepted as the tool, that is what policy analysts are trained to do. By “accepted” I mean that a method is the basis of papers delivered at associational meetings and published in mainstream journals, thereby developing expectations for professional training and hence becoming the subject of textbooks and the focus of graduate school and other training curricula.⁴

Authors of two recently published textbooks, for example, list a policy analytic process of eight steps and six steps, respectively, one that is fairly commonly identified. The first steps of both are: establish the context, formulate/define the problem, specify objectives/determine evaluation criteria, and explore/evaluate alternatives (Bonser, McGregor, & Oster, 1996; Patton & Sawicki, 1993). This appears to suggest that the policy analyst generates issue knowledge in a void, from his head. Nowhere does it ask which or whose policy knowledge should be included in these steps or how to access such knowledge.

If, however, we could get out of this cycle of expectations and back to the kinds of acts (including decisions) that the policy process actually entails, we would see that policy actions are not restricted to the sorts of questions amenable to cost-benefit analysis, decision analysis, and so on. There is a realm of activity that policymakers need to have evaluated, systematically, rigorously, and methodically, which centers not only on values but also on other forms of human meaning, including beliefs and feelings. Rebuilding bridges to make them earthquake-ready, or adopting welfare programs, entails judgments about what people in the situation find meaningful. For example, a recent health policy survey sought to explore attitudes across American race-ethnic groups about sending infirm elderly parents to nursing homes. The question was meaningless, however, to those who take it as a given that parents will be cared for at home until death: they had no conceptual framework within which to understand the question, which had been generated from the context of a different meaning system of values, beliefs, and feelings. In another example, an American-initiated comparative study of U.S. and Japanese policies included the seemingly innocuous, factual question, “What is your age?” Because of American sensitivities about age, the question was placed at the end of the questionnaire, allowing the researcher and the informant time to establish some rapport that would “cushion” the impact of the question’s “personal” nature. When administered in Japan, however, this placement cost researchers key information relative to the elderly informants’ status (age being more venerated in Japan than in the United States, broadly speaking), which led to mis-cues in the appropriate phrasing of preceding questions. Moreover, the comparative reliability of these data could be challenged: because of the link between age and status, Japanese answers are more likely to be chronologically accurate (the elderly having no reason to make themselves appear older), whereas Americans are more likely not to respond or to bias their answers toward lower numbers, given the societal value placed on youth.⁵ In both cases, survey research missed what was meaningful to policy-relevant publics.

Policy analyses that seek to avoid these sorts of errors and pitfalls require another set of analytic tools, ones based on philosophical presuppositions that put human meaning and social realities at their heart. To understand the consequences of a policy for the broad range of people it will affect requires “local knowledge”—the very mundane, expert understanding of and practical reasoning about local conditions derived from lived experience.⁶ Had researchers in either of these examples first sought to understand the values, beliefs, and feelings of informants (about parental aging, about age), and then used that local knowledge in designing the surveys, these problems would likely have been avoided.

The lack of attention to (at best) or outright devaluing of (at worst) local knowledge has been common in development policies. In the 1970s, for example, policies to remedy drought in a specific region had nomadic tribes-people dig more wells. Policy analysts did not understand that, because of the meaning of livestock to a tribesman’s reputation, adding wells would encourage him to increase his herd size, thereby leading to an exacerbation of the problem the policy was intended to resolve.⁷ Examples from other policy areas abound. Schmidt (1993) writes of the disastrous collapse of a bridge after site-based engineers’ “intimate knowledge” about cement requirements under local conditions was dismissed by policymakers in Washington. In connection with nuclear fallout from Chernobyl and sheep grazing in northern England, Wynne (1992) describes the local, implicit knowledge held by shepherds, which was ignored by experts advising policymakers, with detrimental economic results. Had policymakers understood what busing meant to white parents, they might have pursued differently the policy that led to “white flight,” which undermined the policy’s purpose (Paris & Reynolds, 1983, pp. 180-181).⁸ To acquire such local knowledge, policy analysts need interpretive methods.

Interpretive Presuppositions

Interpretive methods are based on the presupposition that we live in a social world characterized by the possibilities of multiple interpretations. In this world there are no “brute data” whose meaning is beyond dispute. Dispassionate, rigorous science is possible—but not the neutral, objective science stipulated by traditional analytic methods (as represented by the scientific method). As living requires sensemaking, and sensemaking entails interpretation, so too does policy analysis.⁹

Traditional approaches to policy analysis—using the tools of microeconomics, decision analysis, and others—are conducted under the assumptions of positivist-informed science: that it is not only necessary but also actually possible, to make objective, value-free assessments of a policy from a point external to it. When policy language is examined, for example, a comparison is often made between the words of legislation and the projected or implemented actions in the field, under the assumption that policy words can and should have univocal, unambiguous meanings that can and should be channeled to and directly apperceived by implementors and policy-relevant publics.

These are not the assumptions of this book. This text assumes, rather, that it is not possible for an analyst to stand outside of the policy issue being studied, free of its values and meanings and of the analyst’s own values, beliefs, and feelings. The argument assumes that knowledge is acquired through interpretation, which necessarily is “subjective”: it reflects the education, experience, and training, as well as the individual, familial, and communal background, of the “subject” making the analysis. Not only analysts, but all actors in a policy situation (as with other aspects of the social world), interpret issue data as they seek to make sense of the policy. Furthermore, human artifacts and actions, including policy documents, legislation, and implementation, are understood here to be not only instrumentally rational but also expressive—of meaning(s), including at times individual and collective identity.¹⁰

The conceptual bases for these ideas were developed by the mostly German neo-Kantian (or neo-idealist) philosophers in the late 19th century, the phenomenologists and hermeneutic scholars of the early part of the 20th century, and some of the critical theorists of the mid- and late-20th century. They were largely arguing against, first, the insistence of the late-19th-century empirio-criticists that scientific knowledge about the human, social world could be derived only through the five senses—an approach that separates values from facts, leaves values outside of the realm of scientific analysis, and takes no account of Kant’s assertion regarding the role of a priori knowledge in understanding. Second, they argued against the early-20th-century logical positivists’ and analytical philosophers’ reduction (as the critics saw it) of analytic concern to language and its forms of logic. In brief, phenomenologists understood that something—variously called “mind,” “consciousness,” a weltanschauung, “paradigm,” “frame,” or “lens”—“filters” sense perceptions and organizes perceived physical stimuli (light or sound waves, sensations of taste, touch, smell) in a process of sensemaking. Prior knowledge—whether derived from education, training, experience, or some other form of personal background—was understood to play a central role in sensemaking. Hermeneuticists argued that human meaning was projected into ...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Contents
• Series Editors’ Introduction
• Preface
• Acknowledgments
• 1. Underlying Assumptions of an Interpretive Approach: The Importance of Local Knowledge
• 2. Accessing Local Knowledge: Identifying Interpretive Communities and Policy Artifacts
• 3. Symbolic Language
• 4. Symbolic Objects
• 5. Symbolic Acts
• 6. Moving From Fieldwork and Deskwork to Textwork and Beyond
• References
• About the Author