1975-vanmeter-policy implementation process- conceptual framework

8/9/2019 1975-Vanmeter-Policy Implementation Process- Conceptual Framework

1/45

http://aas.sagepub.com

Administration & Society

DOI: 10.1177/0095399775006004041975; 6; 445Administration & Society

Donald S. Van Meter and Carl E. Van HornThe Policy Implementation Process: A Conceptual Framework

http://aas.sagepub.com The online version of this article can be found at:

Published by:

http://www.sagepublications.com

can be found at:Administration & SocietyAdditional services and information for

http://aas.sagepub.com/cgi/alertsEmail Alerts:

http://aas.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://aas.sagepub.com/cgi/content/refs/6/4/445Citations

at Nanjing University on July 4, 2010http://aas.sagepub.comDownloaded from

http://aas.sagepub.com/cgi/alertshttp://aas.sagepub.com/cgi/alertshttp://aas.sagepub.com/subscriptionshttp://aas.sagepub.com/subscriptionshttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsPermissions.navhttp://www.sagepub.com/journalsPermissions.navhttp://aas.sagepub.com/cgi/content/refs/6/4/445http://aas.sagepub.com/http://aas.sagepub.com/http://aas.sagepub.com/http://aas.sagepub.com/http://aas.sagepub.com/cgi/content/refs/6/4/445http://www.sagepub.com/journalsPermissions.navhttp://www.sagepub.com/journalsReprints.navhttp://aas.sagepub.com/subscriptionshttp://aas.sagepub.com/cgi/alerts


2/45

[445]

THE POLICY IMPLEMENTA TIONPROCESS

A ConceptualFramework

DONALD S. VAN METERCARL E. VAN HORN

Departmentof Political ScienceOhio State University

Political scientists have turned their attention to the study of

public policy with increasing frequency. The topics withwhich they deal and the methodologies they employ reflectthe diversity of their interests. The disparate character of thepublic policy literature complicates the task of defining thescope of policy analysis and bringing order to the concerns ofpolicy analysts. Some order can be achieved by utilizing amodel of the policy delivery system, which facilitates an

organization of the policy literature (see Figure 1 ).1 More-over, it identifies relationships among the diverse concerns ofpolicy analysts, directs attention to the determinants of-andthe consequences of-public policy, and gives emphasis to the

AUTHORS’ NOTE: This is a revised version of a paper delivered at the 1974 Annual Meeting of the American Political Science Association, Chicago,Illinois, August 29-September 2, 1974. Research support provided by the graduate

school of the Ohio State University greatly facilitated the completion of thispaper. We are indebted to Gillian Dean, William Gormley, Michael Reagan,Randall Ripley, and Stephen Wasby for their helpful comments on an earlierversion of this paper. We also wish to express our appreciation to Rebecca Howewho typed the final version of this paper.


http://aas.sagepub.com/http://aas.sagepub.com/http://aas.sagepub.com/http://aas.sagepub.com/


3/45

[446] ADMINISTRATION & SOCIETY / FEBRUARY 1975

often imperfect correspondence between policies adoptedand services actually delivered.

The components of this model are: (1) an environmentthat both stimulates government officials and receives theproducts of their work; (2) demands and resources that carrystimuli from the environment to policy makers; (3) aconversions process, including the formal structures andprocedures of government, that transforms (converts) de-mands and resources into public policies; (4) the policies thatrepresent the formal goals, intentions, or statements ofgovernment officials; (5) the performance of the policy as itis actually delivered to clients; and (6) the feedback ofpolicies and performances to the environment, which istransmitted back to the conversions process as demands andresources of a later point in time. In most respects thisframework differs little from other adaptations of political

systems models first introduced by Easton.2 Its distinguish-ing feature is that it considers &dquo;policy&dquo;and &dquo;performance&dquoastwo distinct categories.

Some features of the policy delivery system clearly havebeen explored more fully than others. During the pastquarter century, a disproportionate effort has been made toanalyze the character of economic and social problems, the

Figure 1. THE POLICY DELIVERY SYSTEM




4/45

[447]Van Meter, Van Horn / POLICY IMPLEMENTATION PROCESS

claims made for government action, and the processes bywhich policy decisions are made. More recently, policyanalysts have turned their attention to the impacts or effectswhich policies may have on the people and problems atwhich they had been directed.

These foci on the determinants and consequences of publicpolicy have added much to our understanding of the policyprocess. Yet neither tells us a great deal about how policydecisions are transformed into public services: they tell us

little about the application or implementation of publicpolicy. In short, they have not given sufficient attention tothe linkage between policy and performance. It is one thingto examine the determinants of policy decisions and toidentify their impacts or consequences-it is another toprovide explanations for these observed consequences.

Our purpose in this paper is to explore the process of

policyimplementation. To this end we define the concept of

implementation and survey the current literature in an effortto ascertain whether what is presently known will inform aconceptualization of this process. We present a model thatcan be used to analyze policy implementation, both where itinvolves actors within a single organization and acrossorganizational boundaries.

DEFINING POLICY IMPLEMENTATION

Several often-conflicting uses of the concept of implemen-tation are found in the existing literature.3 Our definition isquite explicit: policy implementation encompasses thoseactions by public and private individuals (or groups) that aredirected at the achievement of objectives set forth in prior

policy decisions. This includes both one-time efforts totransform decisions into operational terms, as well ascontinuing efforts to achieve the large and small changesmandated by policy decisions. Williams (1971: 144) states itmost succinctly:




5/45


In its most general form, an inquiry about implementa-tion ... seeks to determine whether an organization can bringtogether men and material in a cohesive organizational unit andmotivate them in such a way as to carry out the organization’sstated objectives.

We should emphasize that the implementation phase doesnot commence until goals and objectives have been estab-lished (or identified) by prior policy decisions; it takes placeonly after legislation has been passed and funds committed(or after a judicial ruling and accompanying decree). AsPressman and Wildavsky (1973: xiv) argue: &dquo;Afterall, theworld is full of policy proposals that are aborted. You can’tfinish what you haven’t started. Lack of implementationshould not refer to failure to get going but to inability tofollow through.&dquo;Therefore, the study of implementationexamines those factors that contribute to the realization or

nonrealization of policy objectives. A clear distinction is made between policy implementa-tion, performance, and what generally has been referred to aspolicy impact. These are distinct though not unrelatedconcepts. The study of impact searches for the consequencesof a policy decision. Do school prayer practices change as aresult of a Supreme Court decision? Do disadvantagedchildren improve their reading or math skills as a conse-quence of an innovative education program? By focusing onthose activities that affect the rendering of public services(i.e., performance), the study of policy implementationhighlights one of the forces that determines policy impact.Hence, as Dolbeare (1974) observes, impact studies typicallyask &dquo;Whathappened?&dquo;whereas implementation studies ask&dquo;Whydid it happen this way?&dquo;

We should point out that we are not maintaining that thestudy of policy implementation will resolve the issuesrevolving around the question of &dquo;real&dquo;impact (e.g., doescompensatory education have salutory effects). Our modeland the research that flows from it are not designed to




6/45


measure and explain the ultimate outcomes of governmentalpolicy, but rather to measure and explain what we prefer tocall program performance (i.e., the degree to which antici-pated services are actually delivered). We recognize that someservices could actually be delivered without having anysubstantial impact on the problem to which the policy issupposed to be related. A policy may be implementedeffectively, but fail to have a substantial impact because itwas ill-conceived-or because of other circumstances. Hence,successful program performance may be a necessary-but notsufficient-condition for the attainment of positive ultimateoutcomes.

THE STUDY OF IMPLEMENTATION

At presentwe

know relatively little about the process ofpolicy implementation. This is an unfortunate deficiency inour understanding of the policy process and may lead toill-advised conclusions on the part of policy makers. Whenfaced with an unsuccessful program, many observers willattribute its failure to insufficient planning or the inadequacyof the program itself. This attribution of blame is oftenunjustified. Viewing the Great Society’s social policiesgenerally, Levine (1968: 86) has concluded that most of thetrouble with the War on Poverty resulted &dquo;notso much fromthe nature of the programs as from difficulties of administra-tion.&dquo;Pointing to the possible gap between the intentionsand statements of public officials (policy) on the one handand the delivery of public services (performance) on theother, Dolbeare and Hammond (1971: 149) have argued that

very little may really be decided by the words of a decision or astatute: the enunciation of such national policy may be just thebeginningof the decisive process of determiningwhat will happento whom, and understandingthis further stage is essential to a fullunderstandingof politics.




7/45


Hence, the study of implementation adds a new dimension topolicy analysis. It gives the student of politics and the policymaker a new understanding of how the system succeeds orfails in translating general policy objectives into concrete andmeaningful public services.

Why do we know so little about the problems of policyimplementation? If this process is so important, why has nota greater effort been made to identify its basic features? Thisneglect is due in part to the naive assumption, implicit inmany studies, that &dquo;oncea policy has been ’made’ by agovernment, the policy will be implemented and the desiredresults of the policy will be near those expected by thepolicy-makers&dquo;(Smith, 1973: 197-198).The implementationprocess is assumed to be a series of mundane decisions andinteractions unworthy of the attention of scholars seekingthe heady stuff of politics. Implementation is deceptively

simple; it does not appear to involve any great issues. Most ofthe crucial policy issues are often seen to have been resolvedin the prior decisions of executives, legislators, and judges.

Second, the growth of Planning Programming BudgetingSystems (PPB)-the leading analytic technique of the1960s-may have encouraged policy analysts to ignore theproblems of policy implementation. PPB concentrated policymakers’ attention on the choices between competing meth-ods of achieving selected goals and objectives, focusingattention primarily on decisions made by policy makers inWashington to the exclusion of the lower echelons ofagencies responsible for implementation. PPB’s concern withimproving the basis for policy-making did not address theproblems of delivering public services. As Williams (1971:139) points out, PPB emphasized program objectives and

alternative means of reaching those objectives. Such a focus,however, did not require that serious attention be given tothe implementation or operation of public programs.

Third, the difficulty of the task has discourageddetailedstudy of the process of policy implementation. The problems




8/45


of implementation are overwhelmingly complex and scholarshave frequently been deterred by methodological considera-tions. Relative to the study of policy formulation, theanalysis of the implementation process raises serious bound-ary problems. It is often difficult to define the relevantactors. Furthermore, many of the variables needed tocomplete an implementation study are difficult-if notimpossible-to measure. Unlike legislative and judicial arenaswhere votes are often recorded, decisions in an administrative

setting are frequently difficult to isolate. Finally, a com-prehensive analysis of implementation requires that attentionbe given to multiple actions over an extended period of time,thus involving an enormous outlay of time and resources.

This is not meant to suggest that the implementationprocess has been ignored by policy makers and policyanalysts, nor that these obstacles are insurmountable. Policymakers have turned their attention to the problems ofimplementation, perhaps as a consequence of the disappoint-ing results of the Great Society’s social programs. Among themost important published studies of policy implementationare Kaufman’s (1960) study of the U.S. Forest Service,Bailey and Mosher’s (1968) examination of the administra-tion of the Elementary and Secondary Education Act of1965, Derthick’s (1970) analysis of federal grant-in-aidprograms, Gross and associates’ (1971) examination ofplanned organizational innovation, Berke and Kirst’s (1972)study of federal aid to education programs, Derthick’s (1972)analysis of the Johnson Administration’s effort to create newcommunities on federally owned land in metropolitan areas,and Pressman and Wildavsky’s(1973) study of Oakland’scommunity development program These and other studies

have helped identify factors that contribute toan

under-standing of the process of policy implementation.While these studies have been highly informative, their

contributions have been limited by the absence of atheoretical perspective. To date, no one has advanced atheoretical framework within which policy implementation




9/45


can be examined. Without such a framework, it is difficult tofurther our understanding of this process in disparate policyareas and jurisdictional settings. The few efforts (Smith,1973; Bunker, 1972) to provide such a perspective are lessthan adequate. For example, Smith asserts that implementa-tion is a problem in Third World nations, where privateinterests seek to prevent a beleaguered bureaucracy fromimplementing public policies. Smith (1973: 199) argues thatthis situation is found only infrequently in Western societies.

As we have already observed, this position is difficult todefend. The problems of implementation are profound inWestern and non-Western nations alike: they are generic tocomplex organizations.

THE FOUNDATIONS OF A THEORETICAL PERSPECTIVE

In their study of the Economic Development Administra-tion’s Oakland Project, Pressman and Wildavsky (1973: 166)write :

There is (or there must be) a large literature about implementa-tion in the social sciences-or so we have been told by numerouspeople.... It must be there; it should be there; but in fact it isnot. There is a kind of semantic illusion at work here becausevirtually everything ever done in public administration must, inthe nature of things, have some bearing on implementa-tion.... Nevertheless, except for the few pieces mentioned in thebody of this book, we have been unable to find any significantanalytic work dealing with implementation.

While we share Pressman and Wildavsky’s concern that fartoo little attention has been

paidto the

questionof policy

implementation, their criticism of the literature is unneces-sarily harsh and short-sighted. Our argument is put simply:there is a rich heritage from the social sciences that is oftenoverlooked by those purporting to discuss the policy imple-




10/45


mentation process. This literature includes theoretical and

empirical work in several disciplines, including sociology,public administration, social psychology, and politicalscience. While most of these studies do not examine

specifically the policy implementation process, close inspec-tion reveals that it takes little imagination to comprehendtheir relevance.

In developing our theoretical framework we have beenguided by three bodies of literature: (1) organization

theory-and more specifically, the work in the general area oforganizational change (innovation) and control; (2) theimpact of public policy, particularly judicial decisions; and(3) selected studies of intergovernmental relations. Our taskin this section is to delineate the contributions of these fieldsof study. Primary attention will be given to the literature onorganizational change and control since we believe that it hasthe

greatesttheoretical contribution to

make,and since it has

been ignored generally by others studying the policy imple-mentation process.

Students of organizational theory and practice have dealtextensively with the topic of change (Bennis, 1966; Downs,1967; Gross et al., 1971; Katz and Kahn, 1966). In one ofthe most insightful analyses of organizational change, Kauf-man (1971) explores a variety of impediments to innovationin organizational structure and action. He examines a numberof factors (e.g., resource limitations, sunk costs, the collectivebenefits of stability, psychic costs, and the accumulation ofofficial and unofficial constraints on behavior) that &dquo;tendtokeep organizations doing the things they have been doing inthe recent past, and doing them in just the way they havebeen doing them&dquo;(Kaufman, 1971: 39). Kaufman recognizesthe many advantages of stability and makes a serious effortto identify those forces conducive to organizational change-those that occur both involuntarily and by design. Yet heconcludes that most organizations &dquo;areimprisoned in thepresent and often cannot change, even when the future




11/45


threatens them unless they do&dquo;(Kaufman, 1971: 40).Organizational control has also been a frequent research

topic. The relevance of this subject for our focus onimplementation may be seen in various definitions found inthe literature. For example, Wilensky (1967: 3) definescontrol as &dquo;theproblem of getting work done and securingcompliance with organizational rules.&dquo; Anthony (1965: 17)defines &dquo;managerialcontrol&dquo;as &dquo;theprocess by whichmanagers assure that resources are obtained and used

effectively and efficiently in the accomplishment of theorganization’s objectives.&dquo;Finally, Etzioni (1964: 68) de-fines organizational control as a process intended to &dquo;ensurethat rules are obeyed and orders followed.&dquo;

Numerous headings have been used to explore this aspectof organizational life. Control has been discussed in terms ofleadership authority, coordination, hierarchy, human rela-

tions, democracy, incentives,and

compliance.For our

purposes the final concept is most useful. Compliance may beseen as a special case in the study of implementation-usuallyrelated to the specific obedience or lack thereof to a law ordirective. Yet studies of the process by which compliance isobtained or avoided give us insight into the problem of theimplementation of complex policies in a fragmented politicalsystem.

In A Comparative Analysis of Complex Organizations( 1961 ), Etzioni utilizes the concept of compliance as a basisfor comparing organizations. It permits one to compare manyfeatures of complex organizations: the goals they pursue,their structures, their motivational characteristics, the powerand interaction of their elites, the level and kinds ofconsensus attained, and their systems of communication and

socialization. Central to Etzioni’s thesis is the notion thatdifferent types of organizations will require different kinds ofcompliance systems. For example, where participants in anorganization are alienated and have an intense negativeorientation toward the organization, coercive power-the




12/45


application, or the threat of application, of punitive sanc-tions-may be required to achieve adherence to the organiza-tion’s rules and objectives. Where most participants haveintense positive orientations and are highly committed to theorganization’s goals and objectives, compliance can beachieved usually through the use of normative power-theallocation and manipulation of symbolic rewards and depriva-tions. Finally, where participants do not have intenseorientations toward the organization and their involvement isa function of perceived costs and benefits, remunerativepower-the allocation of material resources such as salaries,commissions, fringe benefits, and services-is likely to be themost effective means of achievingcompliance.

Integral to Etzioni’s thinking and to all discussions ofcontrol is the relationship between superiors and subordi-nates in complex organizations. The classic Weberian inter-pretation of this relationship holds that the ideal role ofsubordinates is one of implementing faithfully the decisionsmade by their superiors. Policies are made at the highestlevels; they are then carried out by lower participants whosediscretion is acutely limited. (The classic &dquo;idealtype&dquo;isdescribed by Max Weber, 1946: chap. 8.) While thisinterpretation is widely accepted, most organizations deviateconsiderably from it. Numerous studies have shown that such&dquo;lowerparticipant&dquo;groups as attendants in mental hospitals(Scheff, 1961), maintenance workers in factories (Crozier,1964), and prison inmates (Sykes, 1961) can exercise powerand, therefore, affect the performance of complex organ-izations.

Not infrequently these instances are dismissed as excep-tions to the rule. Mechanic (1962) suggests, however, that

they are manifestations of a general pattern. By acquiringcontrol over persons, information, and instrumentalities,lower participants can wield considerable power that is notnormally associated with their formally defined positionswithin the organization. Hence, he (1962: 351) argues that




13/45


&dquo;organizations,in a sense, are continuously at the mercy oftheir lower participants.&dquo;Recognizing that hierarchical con-trol is never perfect and that policy implementors maypossess great power, Baum (1974) has constructed a theoryof judicial impact that reverses the traditional interpretationof hierarchy in organizations. Baum proposes that &dquo;insteadofviewing any power possessed by lower participants asaberrant, we may begin with the assumption that they alonewill determine the content of the policies they execute.&dquo;Herecognizes that this assumption cannot be defended empiri-cally, but he asserts that its use &dquo;requiresthat we discover theforces which counteract autonomy rather than taking themfor granted&dquo;(Baum, 1974: 6).

The power of lower participants in organizations isenhanced by superiors who frequently have little idea whattheir subordinates are doing. The monitoring of subordinate

behavior, hence, becomes an important question in the studyof complex organizations. There are several means by whichsuperiors keep themselves informed about activities in thefield, including reports required of subordinates, personalinspections and contacts, and formal investigations. However,as Kaufman’s (1973) study reveals, administrative feedbacksystems are often inadequate, and superiors often do notwant knowledge about activities in the field. Increasedcompliance on the part of subordinates will come only aftersuperiors have been given new incentives to use whatinformation may become available to them. (For a morecomprehensive discussion of this topic, see Kaufman, 1960;Downs, 1967: chap. 12; Wilensky, 1967.)

The literature on the impact of judicial decisions also hasmuch to contribute to the design of a theory of policyimplementation. To date, most students of judicial impacthave emphasized empirical investigation rather than con-scious theory development. A few, however, have advancedinventories of variables to explain their observations; andsome writers have constructed partial theories of judicial




14/45


impact.4 Relying on &dquo;utilitytheory,&dquo;Krislov (1965) empha-sizes the relationship between superiors (the Supreme Courtand appellate courts) and subordinates (lower trial courts andadministrative agencies). He posits that subordinates acceptand comply with the directives of superiors if the incentivesto do so are greater than are those not to comply. Krislovidentifies three areas of possible motivations for compli-ance-personal utilities, organizational utilities, and psych-ological utilities-and argues that

complianceis at its greatest when personal advantagesare highest,organizational sanctions against opposition are certain and severe,and the legitimacy of the issuing authority is acknowledged.Conversely, it will be at its minimum when the individual utilitiesall point in the direction of opposition, organizational sanctionsare lenient and-most important-erratic in application, and thelegitimacy of the higher authority is doubtful [Krislov, 1965:136] .5

Another study relevant for our purposes is Dolbeare andHammond’s (1971) analysis of the impact of the U.S.Supreme Court’s school prayer decisions. They identify fourcategories of factors which shape the response to courtrulings: the substance of the decision; institutional mecha-nism and procedures; the politico-cultural context; and theinterests, priorities, preferences, and behavior of politicalactors. In doing so, they (1971: 134) suggest that &dquo;similaroranalogous categories apply to other types of policies gener-ated by other institutions of the national government.&dquo;

Finally, there are several studies of intergovernmentalrelations that deal directly with the problem of implemen-tation and that serve to highlight the problems of intra- and

interorganizational relations. Among the most importantones are Bailey and Mosher (1968), Sundquist (1969),Derthick (1970, 1972), and Pressman and Wildavsky (1973).The literature provides numerous insights into the conflictsbetween national, state, and local officials; it points to the




15/45


interdependence of public officials at all levels of govern-ment. More important, it identifies factors that confound theimplementation process in organizations that are neither wellintegrated nor self-contained. Hence, it gives emphasis to theautonomy of subordinates-both in intra- and interorganiza-tional affairs.

A THEORETICAL PERSPECTIVE

The natural starting point in the elaboration of ourtheoretical framework is with the policy itself, where goalsand objectives are established. It is here that the implementa-tion process begins. Following Lowi (1964) and Froman(1968), we submit that the implementation process will varydepending on the nature of the policy to be carried out.

Different types of decisions will display characteristic proc-esses, structures, and relationships among factors that influ-ence the execution of public policy.

We will classify policies according to two distinguishingcharacteristics: the amount of change involved, and theextent to which there is goal consensus among the partici-pants in the implementation process. The element of changeis important in at least two respects. First, implementationwill be affected by the extent to which the policy deviatesfrom previous policies. As a number of scholars (Braybrookeand Lindblom, 1963; Lindblom, 1965; Wildavsky, 1974)suggest, incremental changes are more likely to engender apositive response than will drastic ones. In similar fashion,Derthick (1970: 63) has written that the response of stategovernments to federal initiatives &dquo;dependsin part on

whether the grant is for an activity in which it is alreadyengaged.&dquo;

Second, the implementation process will be influenced bythe amount of organizational change that is required.Kaufman’s (1971) analysis suggests that effective implemen-




16/45


tation is most likely when the implementing agency is notrequired to undergo drastic reorganization. This view isreenforced by Levine’s (1968) argument that many of thefailings of the Great Society’s social programs resulted fromthe increased demands being made of existing administrativestructures and procedures. Policies which mandate changes inthe relationships among participants involved in the imple-mentation process will be more difficult to carry out thanwill policies which require only marginal change in estab-

lished relationships.The other critical feature of the policy is the degree of

conflict or consensus over its goals and objectives. To whatextent do implementing officials agree on the goals of theprogram? Dolbeare and Hammond (1971: 137-138) point tothe significance that goal conflict may have for a programwhen they suggest that the &dquo;value-basedactions of officialsand leaders...

maybe most determinative of ultimate

policy.&dquo;In reviewing the literature on planned organizational

change, Gross and associates (1971: 24-29) identify severalfactors that affect goal consensus-and thus implementation.One of these factors is the extent to which subordinates (orimplementors) have participated in the making of the policydecision. Their literature review finds support for thefollowing arguments: &dquo;( 1 )participation leads to higher staffmorale, and high staff morale is necessary for successfulimplementation; (2) participation leads to greater commit-ment, and a high degree of commitment is required foreffecting change; (3) participation leads to greater clarityabout an innovation, and clarity is necessary for implementa-tion ; (4) beginning with the postulate of basic resistance to

change, the argument is that participation will reduce initialresistance and thereby facilitate successful implementation;and (5) subordinates will tend to resist any innovation thatthey are expected to implement if it is initiated solely bytheir superordinates.&dquo;It cannot be argued, however, that




17/45


subordinate participation in decision-makingwill necessarilyresult in

goal consensus;nor can it be

concluded that theproblems of implementation will be removed once goalconsensus has been achieved.

The combination of these two features produces a typol-ogy of public policies, as depicted in Figure 2.6 Notsurprisingly, the preponderance of policies are found in the&dquo;majorchange/low consensus&dquo;and &dquo;minorchange/highconsensus&dquo;categories. Programs that require major changefrequently lead to goal conflict on the part of relevant actors,while goal consensus is usually highest where little change isinvolved. Major change/low consensus policies evolve fromprotracted controversies, as in the case of the Elementary andSecondary Education Act (1965), the Economic Opportunity

Act (1964), the Medicare legislation of 1965, major taxreforms, a radical reorganization of bureaucratic structures,and busing to achieve school desegregation. In contrast, thestruggles which minor change/high consensus policies causeare less severe. Such policies are reflected in the nature of




18/45

[461 ]Van Meter, Van Horn / POLICY IMPLEMENTATION PROCESS

incrementalism, a political routine that stipulates that current

policydecisions are

largelya function of

previousdecisions.

By definition, incremental policies discourage controversy bydeviating only marginally from previous accommodations andpractices (Braybrooke and Lindblom, 1963; Wildavsky,1974).

These first two policy types are the most common onesfound in the American political system. However, exceptionsto this pattern are not unknown. While policies adoptedfollowing the attack at Pearl Harbor in 1941 mandated majorchange-factories were converted for the manufacture ofarmaments, and the rationing of scarce goods and basic rawmaterials was introduced-there was also a wide consensusthat such actions were both desirable and necessary. The

space program and Peace Corps are additional examples ofmajor change/high consensus policies. Finally, while con-sensus accompanies most policies which involve minorchange, it is not uncommon to find minor change/lowconsensus policies. This occurs frequently when controver-sial programs, such as federal aid to education, Medicare, andpublic welfare, are given reauthorization with only marginalmodifications made in appropriations levels or the formulasused to distribute funds. Hence, while the policy adoptedmay

depart onlyslightly from the past, it may nevertheless

spark enduring controversy.In developing this typology of public policies, we are

suggesting that the probability of effective implementationwill depend-in part-on the type of policy being considered,and that the specific factors contributing to the realization ornonrealization of program objectives will vary from onepolicy type to another. More specifically, we are hypothesiz-ing that implementation will be most successful where onlymarginal change is required and goal consensus is high.Conversely, where major change is mandated and goalconsensus is low, the prospects for effective implementationwill be most doubtful. Furthermore, we anticipate that major




19/45


change/high consensus policies will be implemented moreeffectively than policies involving minor change and lowconsensus. Hence, it is our expectation that goal consensuswill have a greater effect on the policy implementationprocess than will the element of change. With these sugges-tions (or hypotheses) in mind, it is essential that we turn ourattention to an examination of those factors (or independentvariables) that are involved in the implementation process.

A MODEL OF THE

POLICY IMPLEMENTATION PROCESS

Our basic model-as depicted in Figure 3-posits sixvariables which shape the linkage between policy andperformance. This model not only specifies the relationshipsbetween the

independentvariables and the ultimate

depend-ent variable of interest, but also makes explicit the relation-ships among the independent variables. The linkages includedimplicitly represent hypotheses which could be tested empiri-cally, assuming that satisfactory indicators could be con-structed and appropriate data collected. By approaching theproblem in this manner, there is greater promise forelucidating the processes whereby policy decisions are carriedout than simply by correlating independent and dependentvariables in a relatively unthinking fashion (Van Meter and

Asher, 1973). The model has been constructed on the basisof the three bodies of literature cited above, as well as theauthors’ own research and intuitions about the implemen-tation process.

POLICY STANDARDS AND OBJECTIVES

Given our primary interest in the factors that determinethe performance of policy, the identification of performanceindicators is a crucial stage in the analysis. Essentially, the




20/45

[463]

enenLUu0a:Q.

z0

H


21/45


performance indicators assess the extent to which thepolicy’s standards and objectives are realized. Standards andobjectives elaborate on the overall goals of the policydecision. They move beyond the generalities of the legislativedocument to provide concrete and more specific standardsfor assessing program performance. These standards andobjectives are self-evident and easily measurable in somecases. For instance, the Economic Development Administra-tion’s Oakland Project sought to create jobs for the unem-

ployed through several public works projects (e.g., theconstruction of an airport hangar, marine terminal, portindustrial park, and an access road to the city’s newly builtcoliseum) [Pressman and Wildavsky, 1973 ] . To ascertainwhether implementation has been successful, one mustdetermine the number of jobs that have been created, theidentity of those who have been hired, and the progress on

the related public works projects.In most cases it is much more difficult to identify andmeasure performance. This may be due to the program’sbreadth or the complex and far-reaching nature of its goals. Itmay also be a consequence of ambiguities and contradictionsin the statement of standards and objectives. It should berecognized that ambiguity in standards and objectives may befostered deliberately by policy makers in order to ensure apositive response on the part of those responsible forimplementation at other levels of the organization or thepolicy delivery system. Yet the study of implementationrequires that goals and objectives be identified and measuredsince &dquo;implementationcannot succeed or fail without a goalagainst which to judge it&dquo;(Pressman and Wildavsky, 1973:xiv). In determining standards and objectives one could usethe statements of policy makers, as reflected in numerousdocuments such as program regulations and guidelines whichspell out the criteria for an evaluation of policy performance.In some cases, however, the policy’s standards and objectiveswill have to be deduced by the individual researcher. And one




22/45


may even wish to use the criteria for evaluation of policyperformance provided by clientele groups. Ultimately, thechoice of performance measures depends on the purposes forwhich the research is conducted (Rivlin, 1971; Rossi andWilliams, 1972).

POLICY RESOURCES

Policies furnish more than the standards and objectives

against which to judge implementation: they also makeavailable resources which facilitate their administration.These resources may include funds or other incentives in theprogram that might encourage or facilitate effective imple-mentation. (For a more detailed discussion of the utility ofincentives, see Levine, 1972, and Schultze, 1969.) It isobvious that funds are usually not adequate. In fact,Derthick’s (1972: 87) &dquo;new-towns&dquo;study suggests that thelimited supply of federal incentives was a major contributorto the failure of that program: &dquo;Toinduce local governmentsto accept the burden of developing new towns in-town, thefederal government had to give them something of value. ThePresident assumed that low-cost surplus land would beavailable for this purpose, but this assumption turned out tobe wrong.&dquo;

Four additional factors are included in our model:interorganizational communication and enforcement activi-ties ; the characteristics of the implementing agencies; theeconomic, social, and political environment affecting thejurisdiction or organization within which implementationtakes place; and the disposition of implementors. Each ofthese factors consists of several variables, some of which will

be identified here.

Interorganizational Communication and Enforcement Activities

Effective implementation requires that a program’s stand-ards and objectives be understood by those individuals




23/45


responsible for their achievement. Hence, it is vital that weconcern ourselves with the clarity of standards and objec-tives, the accuracy of their communication to implementors,and the consistency (or uniformity) with which they arecommunicated by various sources of information. Standardsand objectives cannot be carried out unless they are statedwith sufficient clarity so that implementors can know what isexpected of them. Communication within and betweenorganizations is a complex and difficult process. In transmit-

ting messages downward in an organization, or from oneorganization to another, communicators inevitably distortthem-both intentionally and unintentionally (Downs, 1967:133-136). Furthermore, if different sources of communica-tion provide inconsistent interpretations of standards andobjectives or if the same source provides conflicting interpre-tations over time, implementors will find it even more

difficult to carry out the intentions of policy. Therefore, theprospects of effective implementation will be enhanced bythe clarity with which standards and objectives are stated andby the accuracy and consistency with which they arecommunicated.

Successful implementation often requires institutionalmechanisms and procedures whereby higher authorities(superiors) may increase the likelihood that implementors(subordinates) will act in a manner consistent with a policy’sstandards and objectives. As Neustadt (1960: 18) hasobserved with respect to presidential directives, orders arenot self-executing: they require the presence of an action-forcing mechanism.

Within the context of a single organization superiors haveaccess to a wide range of such mechanisms. They have the

standard personnel powers: recruitment and selection, assign-ment and relocation, advancement and promotion, andultimately, dismissal. Moreover, they have control over thebudgetary allocations of bureaus and field offices which theymay inflate or reduce in response to satisfactory or unsatis-




24/45


factory performance. While they cannot command obedi-ence, superiors have substantial capacity to influence theirsubordinates’ behavior.

In contrast, when we examine the relationships amongmembers of different organization-or among federal, state,and local officials-many of these mechanisms are absent.Schultze (1969: 202) has summarized this condition:

actions cannot be commanded. There is no hierarchy of officialsin a single line of command who can be directed toward a set ofpredeterminedobjectives. In such cases the careful specificationof plans and objectives by a public agency will not suffice toguarantee effective programs.

In the context of interorganizational (or intergovern-mental) relations, two types of enforcement or follow-up

activitiesare

most important. First, technical advice andassistance can be provided. Higher level officials can often domuch to facilitate implementation by aiding subordinates ininterpreting federal regulations and guidelines, structuringresponses to policy initiatives, and obtaining the physical andtechnical resources required to carry out a policy.

Second, superiors (or federal officials) can rely on a widevariety of sanctions-both positive and negative. We canexplore this aspect of enforcement by referring back toEtzioni’s (1961: 5-8) distinction between normative, remu-nerative, and coercive power. Even though the federalgovernment is not a &dquo;superior&dquo;in its relations with states andlocalities, by extending the analogy we can use this notion asan ordering device for thinking about interorganizationalrelations and the role of enforcement.

The use of normative and remunerative powers is mostcommon. For example, the federal government seeks toinfluence state and local activity through the allocation andmanipulation of symbolic and material rewards. One of the




25/45


most important techniques of federal influence is the

socialization, persuasion,and

cooptationof state

and localactors. By attempting to build a professional alliance aroundthe organization and its mission, federal officials will try tocultivate allies at the state and local level who will implementtheir policies willfully. The fragmentation of the federalsystem adds significance to this technique, since it makesmonitoring effective and oversight virtually impossible (see,for example, Kaufman, 1960; Derthick, 1970; Bailey andMosher, 1968; and Etzioni, 1965).

Another way to achieve influence is to get states andlocalities to participate in a program. The prospect ofreceiving federal dollars is often sufficient to secure theirparticipation and at least implicit acceptance of the objec-tives of federal policy. This is a significant beginningpoint.Research has shown that the amount of federal influence

over aspects of a program increases as the percentage of thefederal contribution rises (Porter, 1973: 85; Derthick, 1970:69-70).

Finally, federal officials may induce state and localparticipation and cooperation by rendering valuable services.For example, many grants do this by offering a healthypercentage of program funds for administration at the stateand local

level;and such crucial

supportservices-technical

advice, staff loans, and research-may be offered to partici-pating organizations.

Federal officials also have more compelling devices at theirdisposal, which range from gentle to explicit forms ofcoercive power. A common practice is to require states andlocalities to draw up elaborate plans for the administration ofa federal program. Once these assurances are made, thefederal government will allocate funds on the assumptionthat they can be withdrawn if the conditions specified in theplan are not fulfilled. Through this device, federal officialsseek &dquo;compliancein advance&dquo;(Derthick, 1970: 209).




26/45


A similar strategy is to specify conditions and procedural

requirements,such as

thorough reportingand

accountingsystems, in the regulations that accompany the acceptance offederal funds. In this manner the federal government hopesto achieve the substantive ends of federal objectives. AsDerthick (1970: 200) points out, however, there are dangersin this procedure:

Specificityentails risks.... The more

specificthe

languageof the

federal requirements, the lower the federal capacity to adapt tostate peculiarities and the greater the danger that the limitationsof federal capacity to compelconformance may be revealed.

Moreover, stringent regulations and guidelines may induce asort of goal displacement, wherein state and local officialsstrive to meet federal requirements in order to obtain fundsand avoid sanctions, while ignoring the basic mission of theprogram.

Recognizing these problems, federal officials tend toemploy more reliable forms of surveillance. Such activitiesinclude on-site visitations, program evaluations, administra-tive and management reviews, audits, and other feedbackmechanisms-including reports by nongovernmental advisory

committees set up tooversee state

and local governmentalunits (see, for example, Downs, 1967: 145-153; Blau andScott, 1962: 170-172; and Kaufman, 1973). Wilensky (1967:60-61) adds an important caveat, however, on the limitationsof oversight:

Where field or branch products and local operating conditionsvary, surveillance machinery proliferates. Such machinery is oftenineffectual-especially where the local people must submit toinspection either by those outside their professionor specialty orby those of different ideological persuasions.... Where the

... doctrinal distance between the inspectors is great, theresulting information blockagemay imperil top leaders’ awareness




27/45


of and accommodation to local problemsas well as their abilityto communicate new goals to local units.

Perhaps the most threatening form of federal influence isthe power to withdraw or withhold funds from states andlocalities. This is the ultimate weapon in the federalgovernment’s arsenal of influence. However, this weapon israrely used. It may cause embarrassment for all concernedand damage the only ally which the federal government has

in the area-thestate or

local implementing agency. Gener-ally, the federal government negotiates with state and localofficials in an effort to attain the greatest possible compli-ance without withholding funds. Thus, federal officialsusually refrain from overt threats which could underminecooperative relations with implementors and generate con-gressional hostility-at the expense of program goals (Der-thick, 1970: 207-214). A more common practice is the auditexception where certain amounts of money are required tobe returned to the federal treasury (see The National

Advisory Council on the Education of Disadvantaged Chil-dren, 1972, for an example of Office of Education action).Finally, the mere knowledge of the ability of the federalgovernment to withhold funds and the awareness of aregularized process of audit discovery can act as powerfuldeterrents to errant behavior (errant, that is, from thesuperior’s perspective).

The Characteristics of the Implementing Agencies

Numerous factors are included in this component of themodel. Students of bureaucratic politics have identified manycharacteristics of administrative

agenciesthat affect their

policy performance. Ripley et al. (1973: 10), for example,speak of bureaucratic structure as those &dquo;characteristics,norms, and recurring patterns of relations inside the execu-tive agencies that have either potential or actual relation towhat they do in the way of policy.&dquo;




28/45


Like Ripley, we view this component as consisting of boththe formal structural features of organizations and theinformal attributes of their personnel. We are also interestedin the implementing agency’s ties to other participants in thepolicy delivery system. Without trying to provide an exhaus-tive listing of these elements, we offer the followingsuggestions of characteristics that may impinge on anorganization’s capacity to implement policy:

(a) the competence and size of an agency’sstaff;(b) the degree of hierarchical control of subunit decisions and

processes within the implementingagencies;(c) an agency’s political resources (e.g., support among legislators

and executives);(d) the vitality of an organization;(e) the degree of &dquo;open&dquo;communications (i.e., networks of

communication with free horizontal and vertical communica-

tion, and a relativelyhigh degreeof freedom in communicationswith persons outside the organization) within an organization;

(f) the agency’s formal and informal linkages with the &dquo;policy-making&dquo;or &dquo;policy-enforcing&dquo;body.

Economic, Social, and Political Conditions

The impact of economic, social, and political conditionson public policy has been the focus of much attention duringthe past decade. Students of comparative state politics andpublic policy have been particularly interested in identifyingthe influence of these environmental variables on policyoutputs (see, for example, Sharkansky, 1967, 1971; Sharkan-sky and Hofferbert, 1969; Cnudde and McCrone, 1969; Dye,1966; Hofferbert, 1964). Although the impact of thesefactors on the implementation of policy decisions hasreceived little attention, they may have a profound effect onthe performance of implementing agencies.




29/45


For illustrative purposes, we propose that consideration begiven to the following questions regarding the economic,social, and political environment affecting the jurisdiction orthe organization within which implementation takes place:

(a) Are the economic resources available within the implementingjurisdiction (or organization) sufficient to support successfulimplementation?

(b) To what extent (and how) will prevailing economic and social

conditions be affected by the implementationof the policy inquestion?(c) What is the nature of public opinion; how salient is the related

policy issue?

(d) Do elites favor or oppose implementationof the policy?(e) What is the partisan character of the implementingjurisdiction

(or organization); is there partisan opposition or support for the

policy?(f) To what extent are private interest groups mobilized in support

or opposition to the policy?

The Disposition of Implementors

Each of the components of the model discussed abovemust be filtered through the perceptions of the implementorwithin the jurisdiction where the policy is delivered. Threeelements of the implementors’ response may affect theirability and willingness to carry out the policy: their cognition(comprehension, understanding) of the policy, the directionof their response toward it (acceptance, neutrality, rejection),and the intensity of that response.

The implementors’understanding of the general intent, aswell as the specific standards and objectives of the policy, isimportant. Moreover, successful implementation may befrustrated when officials are not aware that they are not infull compliance with the policy. We have already dealt withsome aspects of this phenomenon. Yet we want to emphasize




30/45

[4731Van Meter, Van Horn / POLICY IMPLEMENTATION PROCESS

that implementors may screen out a clear message when thedecision seems to contradict deeply cherished beliefs. Undercircumstances of cognitive dissonance (Festinger, 1957), theindividual may attempt to bring the displeasing message intobalance with his perception of what the decision ought tohave been (see, for example, Wasby, 1970: 98).

The direction of implementors’dispositions toward thestandards and objectives is crucial also. Implementors mayfail to execute policies faithfully because they reject the goals

contained in them (see, for example, Peltason, 1961; Dol-beare and Hammond, 1971 ; Etzioni, 1961; Wasby, 1970; andDerthick, 1970). Conversely, widespread acceptance of thepolicy’s standards and objectives, on the part of thoseresponsible for administering it, will enhance greatly thepotential for successful execution (Kaufman, 1960). Atminimum, it would seem that shared attitudes will makeimplementation easier. The goals of a policy may be rejectedfor a variety of reasons: they may offend implementors’personal value systems, extraorganizational loyalties, sense ofself-interest, or existing and preferred relationships. Summa-rizing this phenomenon, Petrick (1968: 7) has written that it&dquo;arisesfrom the fact that human groups find it difficult to

carry out effectively acts for which they have no underlyingbeliefs.&dquo;

Finally, the intensity of implementors’dispositions mayaffect the performance of the policy. Those holding intensenegative preferences may be led to outright and opendefiance of the program’sobjectives. When this occurs theimplementation question may become moot-subordinates(e.g., states and localities) may refuse to participate in theprogram altogether (see Bailey and Mosher, 1968). Less

intense attitudes may cause implementors to attempt surrep-titious diversion and evasion, a more common pattern (see,for example, Lazin, 1973). In these circumstances one mayhave to look to the role of oversight and enforcement toexplain variations in the effectiveness of implementation.




31/45


What all this suggests is that the researcher must gathermultiple indicators of various elements of the dispositions ofpolicy implementors.

Hypothesized Linkages Between Components of the Model

Although our discussion will be presented in static terms,it is important that the dynamic character of the implementa-tion process be

recognized.Factors that may affect the

execution of a policy in its initial stages may be of littleconsequence at a later point in time. Hence, it is vital that thestudy of implementation be conducted longitudinally; rela-tionships identified at one point in time must not beextended casually to other time periods. With this in mind,let us describe and justify briefly some hypothesized relation-ships (see Figure 3).

The standards and objectives of policies have an indirecteffect on performance; what influence this component has onthe dependent variable is mediated by other independentvariables. Obviously, the delivery of public services will beinfluenced by the manner in which standards and objectivesare communicated to implementors and the extent to whichstandards and objectives facilitate oversight and enforcement.

Standards and objectives have their indirect impact on thedisposition of implementors through interorganizational com-munication activities. Clearly, implementors’ responses to thepolicy will be based, in part, on their perceptions andinterpretations of its objectives. This is not to suggest thatgood communication necessarily contributes to a positivedisposition on the part of implementors. However, variationsin

implementors’supportfor federal

policies ultimatelymay

be explained partially in terms of their understanding andinterpretation of these standards and objectives, as well as themanner by which they are communicated.




32/45

[475]Van Meter, Van Horn / POLICY IMPLEMENTATION PROCESS [

Standards and objectives also have an indirect impact onthe disposition of implementors through enforcement activi-ties. They provide the foundation on which superiors can relyin their relations with policy implementors in other organiza-tions. For example, standards and objectives may establishlimits on the sanctions that can be employed legitimately bysuperiors; and they help define the amount of discretionafforded implementing agencies. Where the power to with-hold funds is authorized, various forms of coercive power arealso possible. Where withholding is not permitted, however,superiors may be forced to rely exclusively on normative andremunerative powers. Enforcement and follow-up activitiesmay alter the dispositions of implementors, causing them tosee the advantages of participation, and quite possibly thedisadvantages of resisting effective implementation. Alterna-tively, by employing coercion, enforcing officials can some-

times secure the compliance of implementing officialswithout affecting their dispositions about the program. Forexample, the use of audit exceptions can deter implementorsfrom using funds in a manner inconsistent with a program’sstandards and objectives, even though implementing officialsmay continue to question their desirability.

We posit linkages between policy resources and three othercomponents of the model. The type and extent of resourcesmade available by a policy decision will affect communica-tions and enforcement activities. Technical assistance andother services can only be offered if provided by the policydecision. Furthermore, vigorous enforcement can be achievedonly if the available resources are sufficient to support suchactivity. Similarly, the disposition of implementors can beinfluenced directly by the availability of resources. When vastsums of money or other resources are perceived to beavailable, implementors may view the program with addedfavor, and compliance may be encouraged by the prospect ofreceiving a share of these resources. Conversely, support for a




33/45


program will not be encouraged if implementors perceive thatfew benefits will be realized

byactive

participation.The linkage between resources and the economic, social,and political environment of the implementing jurisdiction(or organization) suggests that the availability of fiscal andother resources may create a demand-by private citizens andorganized interest groups-for participation in and successfulimplementation of the program. Again, the prospect ofbenefiting from the program may cause otherwise quiescentgroups to press for maximum participation. However, wherelimited resources are made available, individual citizens andorganized interests may choose to oppose the policy on thegrounds that the benefits of participation are few comparedto the potential costs (e.g., a loss of state or local autonomy,or a restructuring of established relationships).

It is also hypothesized that the economic, social, and

political environment of the implementing jurisdiction (ororganization) will affect the character of the implementingagencies, the dispositions of implementors, and performanceitself. Environmental conditions can have a significant effecton the willingness and capacity of a jurisdiction (or organiza-tion) to support well-developedbureaucratic structures, thevitality and expertise found in administrative agencies, as wellas the level of political support enjoyed by an agency.Environmental conditions will also affect the dispositions ofimplementors. Where the problems to be remedied by aprogram are severe and private citizens and interest groupsare mobilized in support of a program, it is more likely thatimplementors may accept the policy’s goals, standards, andobjectives. Conversely, where problems are not severe andorganized interests are lined up against a program, imple-mentors may be encouraged to look with disfavor on theprogram requiring implementation. Environmental conditionsmay cause implementors to execute a policy without alteringtheir personal preferences about that policy. Implementors’desire to minimize public hostility or their ideologically




34/45

[477]Van Meter, Van Horn / POLICY IMPLEMENTATION PROCESS [

based inclination to be responsive to public wishes mayinfluence their behavior, even though it may be inconsistentwith their own preferences. Finally, these environmentalvariables are seen as having a direct effect on the delivery ofpublic services. Despite the dispositions of implementors andother forces in the model, these environmental conditionsmay enhance or place limits on performance.

Several characteristics of implementing agencies can affectthe dispositions of their personnel. The nature of thecommunications network, the degree of hierarchical control,and the style of leadership can influence the individual’sidentification with the organization’s goals and objectives,either facilitating or hindering effective implementationdepending on the orientation of the implementing agency.Dispositions can also be influenced by the agency’s formaland informal ties to the &dquo;policy-making&dquo;or &dquo;policy-enforc-

ing&dquo;body (e.g., do they operate at thesame

level ofgovernment? has an effective alliance been built betweenhigher authorities and implementing officials?).

We suggest the possibility of an interactive effect betweeninterorganizational communication and enforcement activi-ties and the characteristics of implementing agencies.Enforcement and follow-up activities can provide the imple-menting agencies with added vitality and expertise-improv-ing their capacity to execute programs. They can also be asource of political support which can facilitate effectiveimplementation. The nature of enforcement and follow-upactivities, including the provision of technical assistance, willbe influenced by the characteristics of the implementingagencies. Since many of the enforcement mechanisms avail-able to superiors operating within a single organizationcannot be utilized when implementation requires interorgan-izational or intergovernmental cooperation, the type ofpower used by superiors (e.g., normative, remunerative, orcoercive) will be affected by the formal and informalrelationships between the policy-makingand implementing




35/45


organizations. Also, in choosing among alternative methodsof enforcement and follow-up actions,

superiorscan be

expected to be sensitive to characteristics of the implement-ing agencies. Agencies possessing competent staffs andleadership will require different types of assistance than thosethat are poorly staffed and led. Similarly, implementingagencies with limited political resources may be morevulnerable to coercive power than agencies that enjoyextensive support among private citizens and public officials.

SOME SUMMARY OBSERVATIONS

The model advanced here has several noteworthy features.It delineates several factors that shape the linkage betweenpolicy and performance and specifies the relationships among

these independent variables. Furthermore, it aids in thedescription of the policy implementation process and servesas a guide in research by generating suggestive hypotheses.This model is relatively complex; however, it is our conten-tion that an examination of its several linkages will lead tomore systematic explanations of policy performance. While itis not our purpose to elaborate each of the linkages includedin the model, we can illustrate its utility. By employing thethree general explanations for unsuccessful implementationfirst proposed by Kaufman (1973: 2), we can emphasize theirrelevance to a summary of our model.

THE COMMUNICATIONS PROCESS

Effective implementation requires that subordinates (orimplementors) know what they are supposed to do. Asmessages pass through any communications network distor-tions are likely to occur-producing contradictory directives,ambiguities, inconsistencies in instructions, and incompatiblerequirements. Even where directives and requirements are




36/45


clear, problems may arise as subordinates (implementors) failto comprehend fully what is expected of them.

Four components of our model address this problem:policy standards and objectives; interorganizational commu-nication and enforcement activities; the characteristics of theimplementing agency; and the dispositions of implementors.Our focus here is essentially on the degree to which policystandards and objectives are transmitted to implementorsclearly, accurately, consistently, and in a timely manner.

Several policy studies address the impact of the communica-tions process on policy performance.Judicial impact studies(Milner, 1971; Canon and Kolson, 1971; Wasby, 1973;Dolbeare and Hammond, 1971) have frequently reportedthat the failure of lower court judges to show substantialcompliance with higher court decisions can be attributedpartly to their ignorance of higher court rulings or theirfailure to comprehend fully the nature of the rulings.

The problems inherent in the communications process arealso reflected in the case of Title I of the Elementary andSecondary Education Act of 1965 (ESEA). 7 State and localeducation officials had anticipated the passage of a &dquo;generalaid&dquo;bill in 1965. Many state and local officials were led tobelieve that the enacted program provided general aid-animpression that was reinforced by a widely distributed anderroneous document listing programs which were &dquo;permis-sible under the law.&dquo;However, it left out many importantrequirements, including those that dealt with the &dquo;targeting&dquo;of aid for disadvantaged children.

While the Title I formula for the allocation of federalfunds to the disadvantaged was precise, the types of programswhich would qualify were not outlined in the original statuteor the

early regulations developed bythe U.S. Office of

Education: wide discretion was delegated to state and localeducation agencies to determine their own approaches foraiding disadvantaged children. According to one federalprogram officer, at least one state defined a &dquo;disadvantaged&dquo;




37/45


student as one lacking art, music, and physical educationskills. The state tested students in its school systems andfound nearly all to be &dquo;disadvantaged,&dquo;using these criteria of&dquo;needsassessment.&dquo;Subsequently, the state allocated Title Ifunds for the support of programs in these three areas andthus-whatever the merit of this action-provided general aidin contravention of the legislative purposes of the program.

This early ambiguity and confusion was in large part aconsequence of statutory provisions written primarily toensure fiscal accountability; they addressed casually thematter of program standards and objectives. In successiveauthorizations, Congress was able to provide greater specifica-tion and clarification of its intent. Eventually, the Office ofEducation rewrote its regulations specifying that programswould have to be developed that contributed to theimprovement of basic cognitive skills (e.g., reading andmath). In response, this state adopted new criteria of &dquo;needsassessment&dquo;which brought it into line with federal guide-lines.

THE CAPABILITY PROBLEM

Successful implementation is also a function of theimplementing organization’s capacity to do what it isexpected to do. The ability to implement policies may behindered by such factors as overworked and poorly trainedstaffs, insufficient information and financial resources, orimpossible time constraints. Commenting on the subordi-nate’s inability to comply with his superiors’ instructions,Kaufman (1973: 3) writes: &dquo;Confrontedby demands hecannot satisfy, he will fashion his own policies to handle thesituation. His own policies often do not coincide with thepolicies of his leaders.&dquo;

Capability problems are highlighted by four componentsof our model: policy resources (their character and quantity);




38/45

[481]Van Meter, Van Horn / POLICY IMPLEMENTATION PROCESS I

interorganizational communication and enforcement activi-ties (the provision of political support, technical advice andassistance); characteristics of the implementing agencies (staffcompetency, leadership, vitality, formal and informal ties topolicy makers); and the economic, social, and politicalenvironment (public opinion, organized interest groups,economic conditions of the jurisdiction).

Derthick’s (1972) analysis of the Johnson Administration’snew towns in-town program provides evidence of how

capability problemscan

inhibit implementation. She reportsthat federal officials faced insurmountable legal barriers whenthey tried to sell surplus land at a low price. Furthermore,the federal government was not able to use effectively whatresources and incentives it possessed. &dquo;Becauseit dependedon local officials to be the agents of its purpose, whateverflaws there were in the local officials’ ability to acteffectively-to gather public support, to overcome opposi-tion, to assemble an administrative organization-were liabili-ties for the federal government as well&dquo;(Derthick, 1972: 88).

The failure of government regulatory activities also isexplained often in terms of this capability problem. In itsreport (1971) on the federal government’s independentregulatory agencies, President Nixon’s Advisory Council onExecutive Organization (chaired by Roy Ash) concluded thatthe failings of such agencies flow from organizational andstructural deficiencies. Furthermore, the oil industry’s virtualmonopoly of information concerning existing fuel reservesand future supplies makes it exceedingly difficult, if notimpossible, for government officials to monitor industrypolicies and practices. It can be argued also that privateinterests may be unable to comply fully with the standards

and objectives of government policy. Faced withthe

de-mands of the Environmental Protection Agency and othergovernmental bodies, public utility companies have claimedthat compliance is neither economically nor technologicallyfeasible.




39/45


DISPOSITIONAL CONFLICTS

Implementation may fail because implementors refuse todo what they are supposed to do. Dispositional conflictsoccur because subordinates (implementors) reject the goals oftheir superiors. Goals and objectives may be rejected fornumerous reasons: they offend implementors’ personal valuesor extraorganizational loyalties; they violate implementors’sense of self-interest; or they alter features of the organiza-

tion and its procedures that implementors desire to maintain(Kaufman, 1971 ).While our primary concern is with the dispositions of

implementors, we are also interested in the four othercomponents of the model which directly influence thisfactor: policy resources; interorganizational communicationand enforcement activities; characteristics of the implement-ing agency; and the economic, social, and political environ-ment of the implementing organization (jurisdiction). Wide-spread attention to the impact of dispositional conflicts onpolicy performance is evident in the literature. Allison (1971 :127-132) suggests that a conflict between President Kennedyand the navy over the location of the Cuban blockade duringthe 1962 missile crisis resulted in the unsuccessful implemen-tation of the President’s policy. &dquo;TheNavy’s resistance tothe President’s order that the blockade be drawn in closer toCuba forced the President to allow one or several Soviet shipsto pass through the blockade after it was officially operative&dquo;(Allison, 1971: 130). Lazin’s (1972) study of the enforce-ment of civil rights regulations in public housing alsodemonstrates that the attitudes of local (implementing)agencies and administrators are among the principal factors

determining compliance.In contrast to the Ash council’s view that the failings ofregulation can be remedied by reforming the organizationalstructure of regulation, Noll (1971: 15) suggests that &dquo;theperformance of regulatory agencies is unsatisfactory because




40/45


regulators have chosen to pursue objectives that are contraryto the public interest.&dquo;In making this argument, Noll isimplying that structural changes designed to improve thecapacity of government agencies to regulate private activitymay be of little consequence. Stated simply, the problemmay not be one of capacity; rather, it involves the willingnessof implementors to achieve the objectives of policy.

CONCLUSION

While far too little attention has been paid to how policydecisions are transformed into public services, we havesucceeded in identifying a number of studies that lookexplicitly at the problems of policy implementation. Most ofthese studies have relied primarily on one of the three generalexplanations for unsuccessful implementation discussedabove. Few researchers have sought to integrate each of theseexplanations into their analysis.

The conceptual framework presented in this paper utilizesthese various partial and insufficient explanations in an effortto provide the basis for a more comprehensive understandingof the implementation process. Specifically, this modeldirects attention to six clusters of variables that affect thedelivery of public services: it points to the relevance of policystandards and objectives, policy resources, interorganizationalcommunication and enforcement activities, the character-istics of the implementing agencies, the economic, social, andpolitical environment affecting the implementing jurisdictionor organization, and the disposition of implementors for the

carryingout of

policydecisions.

It is our contention that this model offers a blueprint forthe description and analysis of the policy implementationprocess and that it proposes explanations for programachievements and failures. We maintain that implementation




41/45


studies, when approached in this manner, have much to offer

policy analysts and policy makers alike. For the policyanalyst, implementation studies move the focus beyond themeasurement of the impacts of public policy toward explana-tions for these observed outcomes. Implementation studiesalert policy makers to variables that can be manipulated toimprove the delivery of public services.

NOTES

1. The model of the policy delivery system was originally offered bySharkansky (1972) and is employed as an analytical framework in Sharkanskyand Van Meter (1975).

2. The theoretical work which is most frequently cited in connection with the

systems approachin

politicalscience is Easton

(1965).3. See, for example, the differing interpretations implied by Pressman andWildavsky (1973), Smith (1973), Lazin (1973), Bunker (1972), Derthick (1972),Gross et al. (1971), and Dolbeare and Hammond (1971). See also the followingstudies in which a variety of concepts are employed to explore the implementa-tion process:

1975-vanmeter-policy implementation process- conceptual framework

Documents