Public Policy

Public Policy

As a concept in political science, “policy” has at least two quite distinct meanings. On the one hand, “policies” are often considered to be ways of doing things, decision rules—for example: “It is the policy of this office to handle requests in such-and-such a way.” In this sense policy would answer the questions “How do you do things around here?” and “What are your rules and procedures?” and may be called administrative policy. Alternatively, policies are often regarded as substantive programs, referring specifically to the content of what is being done, and not necessarily to how it is being done. Policy in this sense answers the questions “What do you do around here?” and “What kinds of problems do you handle?”

Studies of public policy—the policies of governments—often employ both meanings. Such studies explore not only what agencies of government are doing but how they are doing it, not only the content of the program and its history but its administration as well. The way a program is administered will certainly affect its content and results. Still, the two meanings should be kept distinct, since they refer to analytically separate things. This article will concentrate on the study of the content of governmental activities and will be concerned with administrative policy only insofar as it is studied in conjunction with substantive policy.

The field of public policy has as yet been left relatively untouched by the rigorous scientific investigation that has begun to characterize much of contemporary political science. Although political scientists have spent an enormous amount of time, energy, and resources studying political processes and the contexts in which those processes occur, little attention has been paid to policy studies by those in the discipline whose aim it is to apply scientific criteria to the study of political problems. A major reason for this is the obviously overt value content of public policy questions and the temptation, therefore, of any citizen, professional political scientist or not, to comment on such questions. Indeed, some in the profession would regard the solution of value questions as the major function of the political scientist. But even among those who have such a preference there may be disagreement as to the most fruitful method of proceeding. One could opt for a direct strategy of immediately engaging in value discussion and proposing reforms of one sort or another. Or, alternatively, one could take the position that careful empirical and theoretical analysis is a prerequisite to a proper understanding of the problems.

Characteristics of policy studies

Although there is some indication of a developing interest in policy analysis by political scientists who are concerned with constructing theory, the most prevalent mode of conducting research on questions of public policy may, by and large, be characterized in the following manner.


Most studies of public policy take as their format a chronological ordering of events, often describing the beginnings and changes in a particular program up to the present time. Breaking up the program into convenient historical periods marking major changes in public policy is also a common method of presentation.

The historical approach to public policy is often usefully employed to organize a large body of material, and it provides a logical structure for the discussion of the policies under study. Whether chronological ordering of data is useful for theoretical purposes poses quite a different question. If the historical analysis suggests a trend or pattern of behavior which will then be analyzed and explained, the contribution to theory may be quite large. If, as is more often the case, a historical framework is chosen simply to order the data in some convenient way, then the justification for such an ordering lies elsewhere. The point is that historical discussions of public policy, although undoubtedly informative, may not increase our understanding of the relationship between the policy and other significant factors.


By and large, studies of public policy stress richness of descriptive detail. Generalizations about political processes or about comparable public policies are quite rare. Also, in relating our observations of any event we select some features and leave out others, sometimes quite inadvertently. One may legitimately ask, then, of any study, why certain things were described and others not—in other words, what are the criteria employed for the inclusion or exclusion of data? For the most part, in studies of public policies the answer to the criteria question would be “Because it helps to tell the story.” But why certain events rather than others? Most researchers, in fact, do have implicit theories which help them to distinguish between the relevant and the irrelevant, the important and the unimportant. To make such criteria explicit is often the beginning of theory and will certainly help the reader decide whether he shares the same generalizations and premises which the author used in choosing events to describe and situations to evaluate. There is no reason why one should have to search for or induce such generalizations from the study. If they are clearly presented by the author, the argument may center on their validity rather than their existence.


As often as not, in policy studies the legal character of the policy is at issue, including its legislative, executive, and judicial history. Other aspects of the political process are sometimes mentioned (such as public opinion and the activities of political parties and interest groups), but the legal descriptive history, including changes in the policy over time, is the more usual framework for policy analyses. Although the legal framework undoubtedly acts as a constraint on those involved in the making of public policy, its major effects will be constant for most public policies. Researchers will have to turn to other variables, such as the social, economic, and political setting, the personalities of the actors, and organizational factors, to explain variations within and among public policies.


Most discussions of public policy also have a value position, or an evaluative posture. This view may be either in favor of or in opposition to the current operation of the policy, but the latter is most often the case. In the best of these studies the author will make quite explicit what his value premises are, but occasionally these values will be somewhat hidden and implicit. From this normative position the author will then criticize the program. Often these criticisms will be interlaced in the historical-descriptive-legal narrative, but occasionally they will not appear until the last chapter or two, with possibly some mention of them in the first chapter. The study will then usually conclude with some suggestions for improvement of the program, in some cases with a list of reform measures for “solving” the problem. These suggestions for reform will normally take one of two forms. For some authors they are the sine qua non of the study, the major reason why so much time and effort were expended in the collection of historical-legal-descriptive data. For others this reform section is simply pro forma; it is expected that traditional public policy studies will have such a section, and leaving it out would slight the expectations of certain readers. Actually, given the relatively straightforward nature of this kind of study, assuming it is competently done, the section on reform may be the only section which is controversial and argumentative. If the facts are correct, the only argument left is in the interpretation of those facts, and since scientific generalization is not one of the stated goals of these studies, value questions fill the gap.

Other features

In addition to the above quartet of characteristics, four related observations may be made about policy studies in general.

First, in terms of policy focus, the concern of most writers is with the merits of the policy. Is it a good policy or not? Does it satisfy appropriate values? Is it administered fairly? Is it comprehensive? Are the funds spent too large or too small? By and large, political scientists who are interested in explaining how and why policies develop the way they do have attempted to shy away from these kinds of questions. It is not that these are not questions which all citizens should be asking themselves but rather that they are inappropriate questions for a scientific discipline to be asking of its own data.

Second, in terms of approach, the research is problem-oriented, in the sense that the authors speak of problems to be solved with appropriate policies. Again, problem solving of this type is an appropriate form of behavior, but not necessarily for professional political scientists as opposed to political actors. To think in terms of solutions to problems is to direct one’s energies away from discovery (the hallmark of the scientific enterprise) and toward a way of doing things which may not be realistically related to the data being studied. How often political scientists propose reforms which are not feasible because they do not square with political realities!

Third, the research is action-oriented in that political scientists doing policy studies often would like to have an impact on the policy to which they address themselves. It is not only their fellow colleagues for whom they write—it is governmental decision makers and opinion leaders as well. It may be more useful, however, for those analyzing public policies to provide as much information as possible on how various aspects of the policy are related to one another, what interests support and oppose the policy, and how the policy is related to the institutions of government and the informal governmental processes, and to leave the activist political role to the decision makers and those whose interests are clearly affected by the policy.

Fourth, in terms of research design, the research is often on a single policy, a case study. The pitfalls of case studies for theory building are well known, but since traditional studies of public policy are problem-oriented and action-oriented, and focus the debate on the merits of specific policies, the case-study format is perfectly appropriate to the aims of the researchers. But for those interested in developing a scientific inquiry, alternative research designs may be more desirable.

Although these observations are meant to be descriptive of most policy studies, there have been departures from this format. For example, there have been several recent efforts to think about the future in certain policy areas. Rather than being concerned with past policy failures, these studies attempt to delineate expected problems sometime in the future, often based on hypothesized trends of certain data. Studies of the population explosion and what to do about it, nuclear strategy, natural and human resources, automation, and other problems have captured the imagination of some researchers. Unlike traditional studies, they are not as concerned with past or even present problems. Their major concern, rather, is what the world will be like x number of years hence and possible steps which may be desirable given certain developing problems. However, these studies, too, tend to be problem-oriented single cases differing from most traditional studies primarily in their extensive speculations on the future.

New directions in policy research

The objectives of science, in any discipline, are relatively straightforward. If one is to have some measure of control over his environment, whether that environment is physical, social, economic, or political, he needs theory that is based on empirical generalizations which, in turn, subsume particular facts. The greater the array of facts subsumed, the more general the theory and the better the understanding of the causal nature of whatever phenomena are being discussed. Science clearly aims at theory construction; and where theory has been developed, manipulation and control—engineering, if you will—have not been far behind. Clearly, the ability of economists, for example, to have as much impact as they do in advising decision makers about the economy is in no small measure due to the development of theories in that field which specify, more or less accurately, the causal mechanisms which influence the economy.

Let us define “theory” as a logically related set of empirical propositions; let us define “empirical proposition” as a statement which relates one variable to another; and let us define “variable” as a concept which may assume different values. If we employ these definitions in the field of public policy, certain considerations are raised in terms of research strategies.

At the minimum we need to transform policy from a constant, an unvarying concept, to a variable. At the present time, for example, it makes little sense to assert that “the more the x policy, the more the y,” or “the more the y, the more the x policy.” “Policy” is simply not a word which has, as yet, achieved the status of a variable.

Traditionally, one possible first step in developing variables is developing typologies, classifications, or taxonomies. A typology (or classification or taxonomy) is a set of concepts or categories which differentiate among various kinds of a higher-order concept. For example, at a very simple level, a typology of a concept like “government” might be “dictatorship” and “democracy.” One may then ask a number of questions about such a typology—for example, whether the categories are exhaustive and/or mutually exclusive.

At a slightly different level, it is also possible to define a typology as a measure of a concept at the nominal level of measurement. When it is viewed as a measurement problem, one can then ask additional questions, such as whether the typology has validity and reliability, and whether operational definitions are easy or difficult to construct.

Once a typology has been constructed, it may then be possible to go further, to inspect whether an underlying factor is present which could move the typology from discrete categories to a continuous variable (increase the measurement power from nominal to ordinal or interval). If one can move in this general direction, it becomes a good deal easier to talk of theory building. However, it would be quite possible to solve all of the problems of measurement, reliability and validity, and in-clusiveness and mutual exclusiveness, and still not have contributed much to theory. The next problem to be solved is whether the categories successfully differentiate other phenomena. That is, can the categories be related to some other set of categories in such a way that relationships can be observed which might then achieve the status of theory?

Fortunately, some efforts have been made in this direction, although for the most part in a very primitive way. At a level quite close to the data, researchers have distinguished among policies in terms of their substance (agriculture, labor, education, foreign policy), the institutions making them (legislative, judicial, administrative), the target of the policy (businessmen, Negroes, lower class), time periods in which policy has taken form (ante-bellum, post-depression, prewar, Eisenhower years), ideology (secular, capitalist, first amendment, liberal, conservative), values (good, bad, dangerous), extent of support (consensus, divisive), and governmental level (national, regional, local, metropolitan).

Each of these categories is at a very low level of generalization, and unfortunately very little comparative analysis has taken place within each category. That is, labor policies are seldom contrasted with education policies, national policies are infrequently compared with city policies, etc. Instead, the categories generally are used merely to delineate an area of discourse.

At a somewhat higher level of abstraction, public policies have been differentiated in terms of style versus position issues, symbolic versus material issues; and policies have been categorized as distributive-regulatory-redistributive, areal-seg-mental, and collective-private. These distinctions have been useful in developing rather interesting empirical generalizations about how political processes differ, depending upon differences in kinds of policies. For example, it is suggested that political party preference varies more by position issues than by style issues (Berelson et al. 1954, chapter 9); that the interests of organized groups in tangible resources or in substantive power are less easily satiable than are interests in symbolic reassurance, or that the most intensive dissemination of symbols commonly attends the enactment of legislation which is most meaningless in its effects upon resource allocation (Edelman 1964, chapter 2); that political processes themselves will vary, depending upon whether policies attempt to distribute, regulate, or redistribute advantages and disadvantages (Lowi 1964); that policies which affect a whole community are more likely to be successful in homogeneous than in heterogeneous communities, and policies which affect only a segment of a community are more likely to be prevalent in heterogeneous than in homogeneous communities (Froman 1967); and that groups are much more likely to be organized successfully around private than around collective advantages (Olson 1965).

A next step, after developing categories by which different kinds of policies may be compared, is to construct hypotheses in an independent-dependent variable schema. By and large, in the policy literature to date, policies have been treated as dependent variables or outcomes of the political process. For example:

(1) The system of electing the president by the Electoral College tends to favor “liberal” interests.

(2) Apportionment of state legislatures on bases other than population tends to have a “conservative” bias.

(3) Political activities of elites will tend to have a greater pro-civil liberties bias than will political activities of nonelites.

(4) Wealthy, industrial nation-states are more likely to have stable democracies than are poorer, agricultural states.

This use of policy as a dependent variable or outcome need not be its only use, however. It is just as possible to treat policy as an explanation of other phenomena, that is, as an independent variable. One could ask whether different kinds of policies help to structure different types of political processes. At an intuitive level this seems quite plausible, but there has been little or no research into the problem.

Comparative research is also a prerequisite to theory building. Case studies are useful if information about a particular policy is desirable, or if one wishes to explore possible relationships among a wide range of variables. But case studies are manifestly not useful for testing hypotheses or for rigorous theory building. For this, comparative data is essential. Studying two, three, or many policies simultaneously would mean sacrificing factual detail, but it would greatly enlarge the scope of the inquiry and would allow the separation of the unique from the similar.

Once one has constructed variables in a comparative framework, empirical generalization with expanded validity may then take place. These generalizations may be about differences as well as similarities. For example, one may observe that direct federal aid to schools in the United States has, in the last thirty years, proceeded from graduate schools to colleges, and finally to secondary and primary schools. Interestingly enough, racial integration has proceeded in the same order. How does one explain these common patterns? Are there also differences? Are there other policies which exhibit the same pattern? Are there policies which indicate an opposite pattern? If so, how does one explain these differences?

At this point it becomes quite possible to speak of theory construction in the field of public policy. Once variables are identified and placed in an independent-dependent variable framework, comparative data collected and generalization undertaken, placing generalizations in a rigorous logical framework is certainly a possible next step. Given the time and resources which have been trained on political processes in the last twenty years, there is, therefore, no reason to believe that policy analysis cannot assume its rightful place as an important and necessary part of political theory.

Lewis A. Froman, Jr.


Bauer, Raymond A.; Pool, Ithiel de Sola; and Dexter, Lewis A. 1963 American Business and Public Policy. New York: Atherton.

Berelson, Bernard; Lazarsfeld, Paul F.; and Mcphee, William N. 1954 Voting: A Study of Opinion Formation in a Presidential Campaign. Univ. of Chicago Press.

Braybrooke, David; and Lindblom, Charles E. 1963 A Strategy of Decision: Policy Evaluation as a Social Process. New York: Free Press.

Brodie, Bernard 1959 Strategy in the Missile Age. Princeton Univ. Press.

Dahl, Robert A. (1961) 1963 Who Governs? Democracy and Power in an American City. New Haven: Yale Univ. Press.

Edelman, Jacob M. 1964 The Symbolic Uses of Politics. Urbana: Univ. of Illinois Press.

Froman, Lewis A. Jr. 1967 An Analysis of Public Policies in Cities. Journal of Politics 29:94-108.

Hitch, Charles J.; and Mckean, R. N. 1960 The Economics of Defense in the Nuclear Age. Cambridge, Mass.: Harvard Univ. Press.

Huntington, Samuel P. 1961 The Common Defense: Strategic Programs in National Politics. New York: Columbia Univ. Press.

Kahn, Herman (1960) 1961 On Thermonuclear War. 2d ed. Princeton Univ. Press.

Lipset, Seymour M. 1960 Political Man: The Social Bases of Politics. Garden City, N.Y.: Doubleday. → A paperback edition was published in 1963.

Lowi, Theodore J. 1964 American Business, Public Policy, Case-studies, and Political Theory. World Politics 16:677-715.

Munger, Frank J.; and Fenno, Richard F. Jr. 1962 National Politics and Federal Aid to Education. The Economics and Politics of Public Education, No. 3. Syracuse Univ. Press.

Olson, Mancur Jr. 1965 The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, Mass.: Harvard Univ. Press.

Schelling, Thomas C. 1960 The Strategy of Conflict. Cambridge, Mass.: Harvard Univ. Press.

Wildavsky, Aaron 1962 Dixon-Yates: A Study in Power Politics. Yale Studies in Political Science, Vol. 3. New Haven: Yale Univ. Press.

  • Recommend Us