Thesis statement (for Hifsa)

382 Paul Sabatier and Daniel Mazmanian /The Conditions of Effective Implementation

The Conditions of Effective Implementation

PAUL SABATIER AND DANIEL MAZMANIAN

... It is our contention that a statute or other major policy decision seeking a substantial departure from the status quo will achieve its objectives under the following set of conditions:

1. The program is based on a sound theory re - lating changes in target group behavior to the achievement of the desired end-state (ob- jectives).

 1979 by the Regents of the University of California. Reprinted from Policy Analysis, Vol. 5, No. 4, Fall 1979, pp. 481 -504, by permission of the Regents.

2. The statute (or other basic policy decision) contains unambiguous policy directives and structures the implementation process so as to maximize the likelihood that target groups will perform as desired.

3. The leaders of the implementing agencies possess substantial managerial and political skill and are committed to statutory goals.

4. The program is actively supported by organized constituency groups and by a few key legislators (or the chief executive) through

Paul Sabatier and Daniel Mazmanian /The Conditions of Effective Implementation 383

out the implementation process, with the Most basic policy decisions are based upon an courts being neutral Or Supportive. underlying causal theory that can be divided into

5. The relative priority of statutory objectives two components -the first relating achievement is not significantly undermined over time by of the desired end-state(s) back to changes in tar- the emergence of conflicting public policies get group behavior, the second specifying the or by changes in relevant socioeconomic means by which target group compliance can be conditions that undermine the statute's obtained. Both the "technical" and the "compli- "technical" theory or political support. ance" components must be valid for the policy ob-

jectives to be attained.

The conceptual framework underlying this set of At this point, we are concerned only with the conditions ...is based upon a (proto) theory of former ("technical") component, as the remaining

public agencies that views them as bureaucracies four conditions in our framework relate primarily with multiple goals that are in constant interaction to the latter. In particular, we wish to emphasize with interest (constituency) groups, other agen- that target group compliance and the costs in cies, and legislative (and executive) sovereigns in volved in obtaining it -may be wasted if not cor their policy subsystem. rectly linked to the desired end-state. For example,

Before elaborating on each of these conditions, the "technical” component of the theory underly- we should note that obtaining target group com- ing the 1970 Clean Air Amendments relates air pliance is obviously much more difficult in some quality levels back to emissions from various sta situations than in others. The greater the difficulty, tionary and mobile sources (the target groups). It the greater the legal and political resources that assumes that human activities are the major source must be marshalled if compliance is to be achieved. of air pollutants and that pollutant emissions from In the terms of our framework, the required various sources within an air basin can be related, "strength" (or degree of bias) of the last four con- via diffusion models, to air quality levels at specific ditions is a function of several factors, including locations. To the extent that nonhuman sources, the amount of change required in target group be- such as volcanoes, constitute a major emission havior, the orientation of target groups toward the source or that little is known about pollutant inter mandated change, and the diversity in proscribed action and transport in the atmosphere, target group activities of target groups. In other words, the compliance with legally prescribed emission lev- greater the mandated change, the more opposed the els will not achieve air quality objectives (or will target groups, and the more diverse their proscribed do so only very inefficiently). M oreover, the ad activities, the greater must be the degree of statu - ministrative and other costs involved in obtaining tory structuring, The skill of implementing offi- compliance are likely to be resented-with a cor- cials, the support from constituency groups and responding decline in political support for the pro sovereigns, and the stability in socioeconomic con- grain-to the extent that promised improvements ditions if statutory objectives are to be attained. in air quality are not at least approximated. In short, Within this context, the set of five conditions an invalid technical component has both direct and should always be sufficient to achieve policy ob- indirect effects on the (non)achievement of policy jectives. Moreover, each condition is probably nec- objectives. essary if the change sought is substantial and We should note, however, that there are some requires five to ten years of effort; in easier situa- programs for which target group compliance can tions, however, it may be possible to omit one of be interpreted as the policy objective. In such in the last three conditions. stances, the absence of any explicit attempt to link

Condition 1: The program is based on a sound target group behavior to some subsequent end-state theory relating changes in target group behavior to means that the first of our five conditions would not the achievement of the desired end-state (objectives). apply (as the underlying "technical" component

Paul Sabatier and Daniel Mazmanian /The Conditions of Effective Implementation 384

deals directly with that linkage). For example, the for implementation to an existing agency clearly goal of desegregation policy in the South could be indicate the relative priority that the new direc construed as the elimination of dual schools -in tives are to play in the totality of the agency's pro which case the compliance of local target groups grams. If this is not done, the new directives are (school boards) would be tantamount to success- likely to undergo considerable delay and be ac ful implementation. Insofar, however, as the goal corded low priority as they struggle for incorpora- of desegregation was not simply the elimination of tion into the agency's operating procedures. dual schools but also the improvement of black (b) The financial resources provided to the im- children's reading scores, the "technical" assump - plementing agencies are sufficient to hire the staff tion that unified schools improve reading scores and conduct the technical analyses involved in the would have to be valid for the policy objective to development of regulations, the administration of be attained. Permit/service delivery programs, and the moni-

Condition 2: The statute (or other basic policy toring of target group compliance. Although this decision) contains unambiguous policy directives condition is fairly obvious, ascertaining what con and structures the implementation process so as to stitutes "sufficient" resources presents enormous maximize the likelihood that target groups will difficulties in practice. As a general rule, however, perform as desired a threshold level of funding is necessary for there

This is the condition most under the control of to be any possibility of achieving statutory objec- policy formulators (such as legislators). Unfortu- tives, and the level of funding above this threshold nately, its importance has often been overlooked by is (up to some saturation point) proportional to the behaviorally oriented social scientists. For these probability of achieving those objectives. Financial reasons, we will briefly examine its constituent resources are perhaps particularly problematic in parts. labor-intensive service delivery programs and in

(a) The policy obje ctives ar eprecise and clearly regulatory programs with a high scientific or tech ranked, both internally (within the specific statute) nological component, where implementing agen- and in the overall program of implementing agen- cies often lack the funds to engage in the research cies. Statutory objectives that are precise and and development necessary to examine critically clearly ranked in importance serve as an indis- the information presented by target groups and, in pensable aid in program evaluation, as unambigu- some cases, to develop alternative technologies. ous directives to implementing officials, and as a (c) Implementation is assigned to agencies sup- resource available to supporters of those objec- portive of statutory objectives that will give the tives both inside and outside the implementing new program high priority. Any new program re agencies. For example, implementing officials quires implementing officials who are not merely confronted with objections to their programs can neutral but also sufficiently committed and persis- sympathize with the aggrieved party but neverthe- tent to develop new regulations and standard op- less respond that they are only following the leg- erating procedures and to enforce them in the face islature's instructions. Clear objectives can also of resistance from target groups and from public serve as a resource to actors outside the imple- officials reluctant to make the mandated changes. menting institutions who perceive discrepancies Thus it is extremely important that implemen- between agency outputs and those objectives (par- tation be assigned to agencies whose policy orien- ticularly if the statute also provides them formal ac- tation is consistent with the statute and which will cess to the implementation process, such as via accord the new program high priority. This is most citizen suit provisions). likely when a new agency is created with a clear

While the desirability of unambiguous policy mandate after an extensive political struggle, as directives within a given statute is normally un- the program will necessarily be its highest priority derstood, it is also important that a statute assigned and the creation of new positions opens the door

Paul Sabatier and Daniel Mazmanian /The Conditions of Effective Implementation 385

to a vast infusion of statutory supporters. Altenatively, implementation can be assigned to a prestigious existing agency that considers the new mandate compatible with its traditional orientation and is looking for new programs. In addition to selecting generally supportive agencies, a statute can sometimes stipulate that top implementing officials be selected from social sectors that generally support the legislation's objectives. Even if this cannot be done through legislation, legislative supporters can often play a critical role in the appointment of non-civil-service personnel within the implementing agencies.

In practice, however, the choice of implementing agencies and officials is often severely constrained. In many policy areas (such as education) there is little option but to assign implementation to existing agencies that may well be hostile or whose personnel may be so preoccupied with existing programs that any new mandate tends to get lost in the shuffle. In addition, most positions within any governmental agency are occupied by career civil servants who are often resistant to changes in existing procedures and programs and only moderately susceptible to the sanctions and inducements available to political appointees. In fact, the generally limited ability of policy formulators to assign implementation to agency officials committed to its objectives probably lies behind many cases of suboptimal correspondence of policy outputs with statutory objectives.

(d) The statute (or other basic policy decision) provides substantial hierarchical integration within and among implementing agencies by minimizing the number of veto clearance points and by providing supporters of statutory objectives with inducements and sanctions sufficient to assure acquiescence among those with a potential veto. Surely one of the dominant themes in the implementation literature is the difficulty of obtaining coordinated action within any given agency and among the numerous semiautonomous agencies involved in most implementation efforts. The problem is particularly acute in federal statutes that rely on state and local agencies for carrying out the details of program delivery and for which some field-

.

level implementors and/or target groups display considerable resistance toward statutory directives. Thus one of the most important attributes of any statute (or other basic policy decision) is the extent to which it hierarchically integrates the implementing agencies. To the extent the system is only loosely integrated, there will be considerable variation in the degree of behavioral compliance among implementing officials and target groups as each responds to the incentives for modification within its local setting-and thus a distinctly suboptimal attainment of statutory objectives.

The degree of hierarchical integration among implementing agencies is determined by the number of veto/clearance points involved in the attainment of statutory objectives and the extent to which supporters of statutory objectives are provided with inducements and sanctions sufficient to assure acquiescence among those with a potential veto. Veto/clearance points involve those occasions in which an actor has the capacity (quite apart from the question of legal authority) to impede the achievement of statutory objectives. Resistance from specific veto points can be overcome, however, if the statute provides sufficient sanctions and/or inducements to convince role occupants (whether implementing officials or target groups) to alter their behavior. In short, if these sanctions and inducements are great enough, the number of veto points can delay -but probably never ultimately impede-behavioral compliance by target groups. In practice, however, the compliance incentives are usually sufficiently modest that the number of veto/clearance points becomes extremely important. As a result, the most direct route to a statutory objective-such as a negative income tax to provide a minimum income-is often preferable to complex programs administered by numerous semiautonomous bureaucracies.

(e) The decision rules of implementing agencies are supportive of statutory objectives. In addition to providing unambiguous objectives, generally supportive implementing officials, few veto points, and adequate incentive for compliance, a statute (or other basic policy decision) can further bias the implementation process by stipulating the formal

Chapter 13 / Implementation 386

decision rules of the implementing agencies. The decisions of implementing agencies are likely to be consistent with statutory objectives to the extent, for example, that the burden of proof in permit/ licensing cases is placed on the applicant and that agency officials are required to make findings fully consistent with statutory objectives. In addition, a statute can assign authority to make final decisions within implementing institutions to those subunits most likely to support statutory objectives. Finally, when multimembered commissions are involved, the statute can stipulate the majority required for specific actions. In the case of regulatory agencies that operate primarily through the granting of permits or licenses, decision rules that make the granting of a permit contingent upon substantial consensus, such as a two-thirds majority, are obviously conducive to stringent regulation.

(f) The statute (or other basic policy decision) provides ample opportunity for constituency (interest) groups and sovereigns supportive of statutory objectives to intervene in the implementation process through, for example, liberal rules of standing to agency and judicial proceedings and requirements for periodic evaluation of the performance of implementing agencies and target groups. While a statute can take steps to assure that implementing officials are generally supportive of statutory objectives and that the decision process involving implementing agencies and target groups contains few veto points, adequate incentives for compliance, and supportive formal rules, we nevertheless contend that implementing officials cannot necessarily be trusted to act in a manner consistent with statutory objectives. What is also required is constant oversight and intervention from supportive constituency groups and legislative (and executive) sovereigns.

A statute (or other basic policy decision) can take a number of steps to maximize the probability of such intervention. First, it can require opportunities for public input at numerous stages in the decision process of implementing agencies and even require that the agencies take positive steps to assure the participation of unorganized potential beneficiaries. Second, it can provide for liberal

rules of standing to appeal agency decisions to the courts. For example, the citizen suit provisions of the 1970 Clean Air Amendments have been used on several occasions to compel the U.S. Environmental Protection Agency to carry out statutorially mandated provisions that it had failed, for one reason or another, to do. Third, requirements for periodic reporting of agency performance to legislative and executive sovereigns and for evaluation studies by prestigious independent organizations (such as the National Academy of Sciences) are conducive to external oversight of the implementing agencies and probably to the achievement of statutory objectives.

In sum, a carefully formulated statute (or other basic policy decision) should be seen as a means by which legislators and other policy formulators can structure the' entire implementation process and maximize the probability that the policy outputs of the implementing agencies and the behavior of target groups (whether outside or inside those agencies) will be consistent with statutory objectives. This requires, first, that they develop unambiguous policy objectives and incorporate a valid technical theory linking target group compliance with the desired impacts. In order to maximize the probability of such compliance, they should then assign implementation to supportive agencies, provide implementing officials with adequate financial resources, hierarchically integrate the implementation process through minimizing veto points and providing sufficient incentives to overcome resistance, bias the formal decision rules of implementing agencies, and provide opportunities for outsiders to participate in the implementation process and to evaluate accurately agency (and target group) performance.

But a statute, no matter how well it structures implementation, is not a sufficient condition for assuring target group compliance with its objectives. Assuring sufficient compliance to actually achieve those objectives normally takes at least three to five, and often ten to twenty, years. During this period, there are constant pressures for even supportive agency officials to lose their commitment, for supportive constituency groups and sovereigns

Paul Sabatier and Daniel Mazmanian /The Conditions of Effective implementation 387

to fail to maintain active political support, and for the entire process to be gradually undermined by changing socioeconomic forces. In short, while a statute can go a long way toward assuring successful implementation, there are additional conditions that must be fulfilled if its objectives are to be attained.

Condition 3: The leaders of the implementing agencies possess substantial managerial and political skill and are committed to statutory objectives.

As already indicated, legislators and other policy formulators can take a number of important steps-both in the drafting of a statute and in the subsequent appointment of non-civil-service personnel-to increase substantially the probability that the leaders of implementing agencies will be supportive of statutory objectives. In practice, however, statutory levers are often somewhat limited (except where creation of a new agency is feasible), and the process of appointing political executives is heavily dependent upon the wishes of the chief executive and important legislators -several of whom may well not be committed to implementation of the basic policy decision. In short, the support of top implementing officials is sufficiently important and problematic to warrant being highlighted as a separate condition for successful implementation.

Moreover, policy support is essentially useless if not accompanied by political and managerial skill in utilizing available resources. Political skill involves the ability to develop good working relationships with sovereigns in the agency's subsystem, to convince opponents and target groups that they are being treated fairly, to mobilize support among latent supportive constituencies, to present the agency's case adroitly through the mass media, and so forth. Managerial skill involves developing adequate controls so that the program is not subject to charges of fiscal mismanagement, maintaining high morale among agency personnel, and managing internal dissent in such a way that dissidents are convinced they have received a fair hearing.

Finally, there is some evidence that maintain -

ing high morale, commitment, and perhaps even skill becomes increasingly difficult over time. Innovative policy initiatives often attract committed and skillful executives to implementing institutions, particularly in the case of new agencies. But such people generally become burned out and disillusioned with bureaucratic routine after a few years, to be replaced by officials much more interested in personal security and organizational maintenance than in taking risks to attain policy goals.

Condition 4: The program is actively supported by organized constituency groups and by a few key legislators (or the chief executive) throughout the implementation process, with the courts being neu- tral or supportive.

It is absolutely crucial to maintain active political support for the achievement of statutory objectives over the long course of implementation. If the first three conditions have been met, this essentially requires that sufficient support be maintained among legislative and executive sovereigns to provide the implementing agencies with the requisite financial resources annually, as well as assuring that the basic statute is not seriously undermined but instead modified to overcome implementation difficulties.

This seemingly rather simple requirement is, however, exceedingly difficult to accomplish, for a variety of reasons. First, the rather episodic issue-attention span of the general public and the mass media tends to undermine diffuse political support for any particular program among both the public and legislators. Second, there is a general tendency for organized constituency support for a wide variety of programs -including environmental and consumer protection, as well as efforts to aid the poor-to decline over time, while opposition from target groups to the costs imposed on them remains constant or actually increases. This shift in the balance of constituency support for such programs gradually becomes reflected in a shift in support among members of the legislature as a whole and the committees in the relevant subsystem(s). Third, most legislators lack the staff resources and/or the incentives to monitor program implementation

Paul Sabatier and Daniel Mazmanian /The Conditions of Effective Implementation 388

actively. The exception is constituent casework, which tends to be heavily skewed towards complaints. Without active political support from a few key legislators, implementing officials supportive of the program find it difficult to overcome the constant drumbeat of constituent complaints, as well as the delay and resistance inherent in implementing any program requiring substantial behavioral change (except in those instances where target groups support such change).

Despite these difficulties, the necessary infusion of political support can be maintained if two factors are present. The first is the presence of a "fixer" (or fixers)-that is, an important legislator or executive official who controls resources important to other actors and who has the desire and the staff resources to closely monitor the implementation process, to intervene with agency officials on an almost continuous basis, and to protect the budget and the legal authority of the implementing agencies. Except in very unusual circumstances, however, any particular "fixer" is unlikely to occupy a crucial position and/or to maintain an interest throughout the long process of implementation. This brings us to the second, and ultimately the most important requirement, namely, the presence of an organized supportive constituency (interest) group that has the resources to monitor closely program implementation, to intervene actively in agency proceedings, to appeal adverse agency decisions to the courts and to the legislature, and to convince key legislators that the program merits their active support. For the paramount advantage of any organization over an individual is continuity. If the supportive constituency is present, "fixers" can generally be found and/or nurtured.

Programs involving intergovernmental relations, however, pose additional difficulties to the maintenance of political support. On the one hand, programs of intergovernmental "subordinates" (such as localities vis -a-vis states and the federal government) are often subject to revision and/or emasculation by superordinate units of government. Unless a program's representatives occupy important positions at the superordinate level, there

is little that can be done to maintain its legal (and sometimes financial) integrity. Conversely, superordinate levels are usually confronted with substantial local variation in political support for program objectives and, consequently, in the compliance of local implementing officials with program directives. While such variation can, in principle, be overcome if the superordinate statute provides very substantial incentives for comphance and sufficient financial resources to enable superordinate officials essentially to replace local implementors , in practice the system is seldom structured to that degree, and thus superordinate officials are forced to bargain with recalcitrant local implementors . The result is greater sensitivity to local demands and generally a suboptimal achievement of statutory objectives.

The discussio n thus far has focused on the need for political support among the legislative and executive sovereigns of implementing agencies. But one must not neglect the courts. In most cases, the contemporary deference of most federal and state courts to agency decision making means that they play a rather minor role in the implementation process except on procedural issues and to assure conformity with explicit statutory directives. But courts strongly opposed to a given statute have the authority to emasculate implementation through delay in enforcement proceedings, through repeatedly unfavorable statutory interpretations, and, in extreme cases, by declaring the statute unconstitutional. On the other hand, there have been some instances where courts have substantially strengthened programs through favorable rulings. Given the enormous potential role of the courts, we argue that successful implementation of statutory objectives requires that they be either neutral or supportive.

Condition 5: The relative priority of statutory objectives is not significantly undermined over time by the emergence of conflicting public policies or by changes in relevant socioeconomic conditions that undermine the statute's "technical" theory or political support.

Change is omnipresent in most contemporary societies, in part because most countries are im-

Paul Sabatier and Daniel Mazmanian /The Conditions of Effective Implementation 389

mersed in an international system over which they have only modest control, in part because policy issues tend to be highly interrelated. Pollution control, for example, is linked to energy, to inflation and national monetary policy, to transportation, to public lands, and to numerous other issues. As a result of this continuous change, any particular policy decision is susceptible to an erosion of political support as other issues become relatively more important over time. Obvious examples would be the effect of the Vietnam War and inflation on many Great Society programs and the effect of the energy crisis and inflation on pollution control programs. Change can also be so extensive as essentially to undermine the technical assumptions upon which a policy is based, as when the migration of poor people from the South and Puerto Rico to northern industrial cities brought into serious question the ability of state and local governments to provide matching funds for welfare programs.

It is in responding to such changes that support for a particular program from key legislators, organized constituency groups, and implementing officials becomes crucial. If they are sensitive to the effects that changes in seemingly tangential policies and in technical assumptions can have on "their" program, they can take steps to see that these repercussions are addressed in any new legislation.

Policy Feedback and Evaluation

Thus far our attention has been focused on the extent to which implementing agencies and target groups act in a manner consistent with statutory objectives and ultimately on the extent to which those objectives are actually attained. In this respect we have mirrored the focus on formal goals of much of the literature on implementation assessment and program evaluation.

But if one is interested in the evolution of policy and particularly with the political feedback process, a much wider range of impacts (or outcomes) needs to be considered. Of particular im-

portance are unintended impacts that affect political support for tile Program's objectives. For example, any assessment of the implementation of school desegregation policy should be concerned not only with the amount of desegregation achieved but also with the effect of desegregation on "white flight' ' and ultimately on the amount of political and financial support for the public schools. Moreover, there is some evidence that political feedback is based primarily upon perceived, rather than actual, impacts and that policy elites evaluate a program not in terms of the extent to which it achieves its legal mandate but rather in terms of its perceived conformity with their policy preferences.

The actual process of policy evaluation and feedback occurs continuously on an informal basis as the implementing agencies interact with concerned constituency groups, legislative (and executive) sovereigns, and the courts. At periodic intervals, however, the process normally becomes more formal and politically salient as attempts are made to revise substantially the basic statute. For example, major efforts to amend federal air pollution control law seem to occur every three to four years. Some of these revisions can be attributed to continued resistance from affected target groups, while others can be traced to significant changes in relevant social and economic conditions. Whatever the source of proposed changes, it is important that supporters of the original objectives provide for independent evaluation studies to accurately assess the actual impacts of the program. Such systematic evaluation serves both to correct imperfections in program design and performance and to counteract the tendency for complaints to dominate the informal feedback process.

Implementation Under Suboptimal Conditions

A frequently voiced criticism against both legislators and scholars is that they have been far more concerned with the passage of legislation than with

Chapter 13 / Implementation 390

its effective implementation. Over the past decade, however, a burgeoning interest in policy implementation and evaluation has occurred in the academic community. This has matched a corresponding shift of emphasis among legislators from the passage of major new policy initiatives to more effective implementation and oversight of existing programs. One of the principal purposes of this paper is to provide both communities an understanding of the conditions under which statutes (and other basic policy decisions) that seek to change the status quo can be effectively implemented-that is, can achieve their policy objectives.

Our discussion has shown that legislators and other policy formulators can go a long way toward assuring effective policy implementation if they see that a statute incorporates a sound technical theory, provides precise and clearly ranked objectives, and structures the implementation process in a wide number of ways so as to maximize the probability of target group compliance. In addition, they can take positive steps to appoint skillful and supportive implementing officials, to provide adequate appropriations and to monitor carefully the behavior of implementing agencies throughout the long implementation process, and to be aware of the effects of changing socioeconomic conditions and of new legislation (even in supposedly unrelated areas) on the original statute.

In practice, of course, even those legislators and other policy formulators concerned with effective implementation operate under substantial constraints that make it extremely difficult for them to perform all these tasks. Valid technical theories may not be available. Imperfect information, goal conflict, and multiple vetoes in legislative bodies make it very difficult to pass legislation that incorporates unambiguous objectives and coherently structures the implementation process. Implementation must often be assigned to agencies that are not supportive of the policy objectives. Supportive interest groups and legislators with the resources to serve as "fixers" may not be available or may go on to other things over the long course of implementation.

Nevertheless, even under such suboptimal con- ditions, several steps can be taken at least to in crease the probability of effective implementation.

1. If a valid "technical" theory linking target group behavior to policy objectives is not available or is clearly problematic, then the authors of the statute should make a conscious effort to incorporate in it a learning process through experimental projects, extensive research and development, evaluation studies, and an open decision process involving as many different inputs as possible.

2. If the legislature insists on passing legislation with only the most ambiguous policy directives, then supporters of different points of view can initiate litigation in the hopes of finding a court that will invalidate the law as an unconstitutional delegation of (legislative) authority. Authority not very promising, this strategy has been employed successfully at least once in a California case, with subsequent legislation providing much clearer guidance to the agency.

3. If implementation cannot be assigned to strongly supportive agencies, then it is absolutely crucial to provide for intervention by outsiders through citizen suit provisions, periodic reporting to sovereigns,. evaluation studies by prestigious and relatively independent outsiders, and perhaps special legislative oversight committees.

4. If there are no active supportive interest groups with the necessary resources to monitor implementation carefully, then identification and mobilization of such a group must be a major priority of supportive legislators and implementing officials -as any programs is doomed in the long run without one. While it is occasionally possible to create new organizations from scratch, a more feasible strategy is to convince an existing organization with the requisite resources to expand its program to make program monitoring a major responsibility.

Case Study 13 /Introduction 391

5. If a "fixer" is not readily available, then program supporters must make a major effort to find or develop one. This may involve convincing a competent new legislator to specialize in this area or convincing an existing legislator that constituents strongly support the program and thus require it being given higher priority. If legislators in the relevant committees having jurisdiction over the implementing agencies are apathetic (or, worse, hostile) toward the new program, then efforts should be made to reorganize

Committee jurisdiction or perhaps to create a special oversight committee with a program supporter as chairperson. Whatever the means, however, finding a "fixer" is of paramount importance for effective implementation.

In short, even if the conditions for effective implementation are not met at the time of the basic policy decision, policy formulators and other program supporters can still take a number of steps to approximate the ideal over time.