Elizabeth Anne VanderPutten
Because social service policies typically specify both intended outputs and intended outcomes, slippage can be stated in terms of either or both. Outcomes are the goals to be achieved by the policy; outputs are the services which are the means for achieving the desired goals. It is suggested in this chapter that twenty years of evaluations of federal social policies have documented the existence of unacceptably high levels of slippage between the policy-makers' outputs, as well as outcomes. Considered are a wide range of evaluation studies and research reports that have documented the page at all levels and sought to explain the phenomenon. Among the foci of these studies are the policy-making process itself; the Branch where policies are translated into programmatic terms; the relationships between the two; the content of the policies; and implementation of programs at the local level.
Vague and sometimes conflicting goals resulting from the policy making process, the not-infrequent distortions of policy goals by program development personnel, and the type and extent of administrative enforcement procedures are shown to affect the final outcomes and outputs. Nevertheless, they do not seem to account for the mechanisms by which slippage occurs. What appears most probable is that persons, at all levels, are guided by factors other than or in addition to overall policy goals. It is suggested that, among these factors, the most important may be the subgoals of the individual.
Implementation research, that is, those lines of inquiry involving what happens at the local level in the federal social policy implementation chain, is then considered. Implementation is found to be a function of the policy, programmatic goals, and the local organizational context within which the policy is implemented. Three theories of organizational behavior are presented, and the argument is made that the social service agencies are most appropriately described as loosely coupled.
Individual decision-making within a particular type of organization becomes the central focus. Based on work by Herbert Simon, a model is developed postulating that, faced with ambiguities of goals, technologies and authority, staff in social service agencies define problems in terms of subgoals that can be achieved through individual effort. These subgoals, and the strategies used to achieve them, influence the extent to which slippage or congruence exists between policy goals and policy outputs and outcomes. Individual subgoals are classified in three categories: personal, organizational, and professional. "Decisions" are defined as the more or less routine allocation of resources, including teacher and student time and effort, to the achievement of these goals.
An analysis was made of ethnographic case-study data from a federal social policy which were gathered at ten sites in various parts of the country by researchers from Cornell University. The chapter suggests that the analysis will support the propositions that: (1) subgoals are central to the decision-making of social service staff, and (2) the greater the similarities between the subgoals of staff and the goals of policy, the greater the congruence between policy and outputs.
B. POLICY ASSUMPTIONS ABOUT OUTPUTS AND OUTCOMES
Berman defines slippage as the lack of "consonance between ... local interests and federal program goals."32-1 While the research reported in this study is limited to slippage between policy intent and policy out puts, other related research has involved intended outcomes as well. Although policies often stipulate both outputs and outcomes, as well as inputs and rationale, given policies tend often to emphasize either outputs or outcomes. Much of the social legislation of the 1960s, for example, tended to place greater emphasis on general outcomes, while many of the education bills of the 1970s tended to emphasize service outputs. Consequently, the literature reviewed in this chapter includes slippage in both senses. "Results" is used in this chapter to include both outputs and outcomes.
2. Consequence Will Equal Policy Intent
Each policy-maker, whether serving on a local school board, as a corporate director, member of Congress, or elsewhere, must face the essential policy-making question: "Given that I desire the following results, what policy can I develop that will yield the desired consequences?" As Wise put it, "the two questions to be asked of every education policy are: will it have the intended effect? and what other effects will it have."33-1 The term "policy" here refers to "nonstandard decisions," and suggests a departure from routine behavior.33-2
3. Illustrative Federal Social Policies
During the quarter of a century ending in 1980, the federal government initiated a massive and sweeping set of social reform laws. Among these were numerous policies aimed at planned educational change.33-3 The term "planned educational change" is one of a variety of phrases used primarily, although not exclusively, to refer to federal programs aimed specifically and directly at altering either the process or the output of education, primarily in the public sector. In part, the term reflects the activist stance of federal education policy-makers during the past quarter of a century marking a significant departure from their previous supportive, non-interventionist role.33-4
Analyses of the legislation implementing these policies show that all contained, with varying degrees of vagueness and ambiguity, statements of desired results and at least an assumed relationship between the intended results and the policy.
The National Defense Education Act (NDEA) of 195834-1, for example, begins by stating its rationale and broad desired outcomes:
The succeeding parts of the Act elaborate various aspects of what the policy would mean in practice, including the level of authorization (federal inputs), the types of activities to be supported (outputs), and proscriptions against other activities.
The Elementary and Secondary Education Act (ESEA) of 196534-3 begins with an ambitious statement of policy intent:
As with NDEA, the rest of the Act elaborates the policy by stipulating various and specific purposes or policy subgoals for which the funds were to be used, the methods of allocation and the levels of federal funds authorized.35-1
These examples could be followed by hundreds of others. The purpose, however, is merely to make the point that policies contain the implicit assumptions that results will equal policy intent and will follow from the policy.
The reason for this is obvious. Not to contain such assumptions would make the policy irrational. If a given consequence is desired, it would be unreasonable to develop a policy aimed at not producing the desired results.
The examples also give evidence of the lack of clear, non-ambiguous, goals. The multiple actions supported by the policies, often in no particular order of priority, indicate an unwillingness, if not inability, of policy-makers to select the most efficient means of allocating resources.
In reality, not even the most optimistic social policy-makers assume an equality between the policies they adopt and the results desired. The point is that policy-makers must and do: (1) posit a technical identity of policy and desired consequences and a causal connection between the two; (2) hope for the closest fit; and (3) expect to make adjustments in either or both at some later time, if the degree of slippage is too great. Evidence of this can be found in the legislative history and subsequent record of amendments of almost any federal social policy.
To illustrate this point, consider a single subsection of a specific law: ESEA, Part A, Subpart 1, Grants to Local Educational Agencies. This section was enacted originally on April 11, 1965, to provide funds to local schools for the education of disadvantaged children. In July of the same year, it was amended for the first time, and again, in November, 1965. Subsequent amendments occurred in November, 1966; January 1968; October, 1968; April, 1970; June, 1972; August, 1974; and April, 1976. Chapter II of the Educational Consolidation Act of 1981 brought further changes. Whether these changes reduced slippage or not depends, at least in part, on how much confusion the changes brought to local implementers.
C. TWENTY YEARS OF POLICY EVALUATION RESEARCH
Twenty years of evaluation research of federal social policies has repeatedly confirmed that slippage is considerable and significant between the policy intent and the policy outcomes.36-1 A major study of NDEA Title XI in 1968 demonstrated that many of the Act's subgoals, such as supporting institutes for the advanced training of teachers and attracting to those institutes thousands of teachers, were accomplished. Nevertheless, the study also showed that little progress was made toward accomplishing one of the primary outcomes of NDEA, the upgrading of science and foreign language instruction in the nation's elementary and secondary schools.37-l Moreover, the report indicated that many of the teachers who attended the Institutes were already the better prepared teachers and that many of these left public school teaching soon after training. That the more highly qualified teachers were recruited does not in itself necessarily indicate that slippage occurred. NDEA was specifically aimed at attracting such individuals as participants. As the Act states: "We must increase our efforts to identify and educate more of the talent of our nation."37-2 At the same time, the finding that many participants did not return to schools, or, if they did, returned as supervisors, was inconsistent with the legislation, at least as it was interpreted by the regulations. For example, participants were required to sign letters of intent to return to their schools for a period of time.
The National Science Foundation conducted a four year study between 1976 and 198037-3 to assess changes in science, mathematics and social science education at the pre-college level during the preceding decades of federal involvement in the area. Among the more discouraging findings was the fact that very little change had occurred. Executive Summaries of the reports contain statements such as: "Stated objectives for elementary school science have not changed significantly since 1955;"38-1 "Although the scope and sequence of the social studies curriculum has remained basically stable in general outline over the 20 year period from 1955 through 1975, there have been a few noticeable shifts ... the infusion of concepts and methodologies from the scientifically oriented social science disciplines;" and "Recent studies do not indicate clearly whether patterns of instructional methods have changed as result of the 'new social studies' and other educational innovations."38-2 Moreover, the studies found that what changes did occur were largely in teachers' knowledge, but not necessarily in student learning.
The mathematics education report states: "We do know that the NSF institute effort changed teachers' mathematical competencies and was a significant factor leading to the rapid implementation of the new curricula such as SMSG and UICSM. This does not say that teachers became more skillful and/or more effective in teaching mathematics." The report does not say that there were no effects; it says only that the studies do not show effects on student achievement, that is, outcomes.38-3
Early evaluations38-4 of the educational programs that aimed generally at increasing equal educational opportunity reported consistently discouraging findings. The Westinghouse study of Head Start, a primary and pre-primary program of health and education39-1 authorized by the Economic Opportunity Act of 1964,39-2 reported finding little, and often no, evidence that outcomes matched Congressional intent. A comprehensive review and synthesis of the finding of various studies on equal educational opportunity in 197739-3 suggested that, except in rare instances, the children's post-test scores did not improve.
More recent studies of Title I of ESEA do indicate that, after seventeen years, greater congruence exists between intent and outcome.39-4 For example, scores on standardized tests of children who received compensatory educational services are closer to the national norm than eligible children who do not receive the services. Analyses of scores from the National Assessment of Educational Progress, for example, indicate that students in Southern states, the primary recipients of Title I funds, have progressed over the past 13 years.39-5 Nevertheless, even the most optimistic observers point out a continuing achievement gap between poor minority children and advantaged majority children.39-6
2. Questioning Policy Assumptions
In general, the evaluation studies used a discrepancy model in which goals were specified in measurable form, outcomes evaluated, and the two compared.40-1 The studies showed, but did little to explain, why such discrepancies existed. What the studies did do, however, was to create coteries of researchers who called into question various assumptions contained in the educational policies and sought by their research to explain the causes of slippage. One question became: Were the goals embedded in the policies clear, or were they too ambiguous and contradictory to permit reasonable congruence between policy and outcome? Two usual federal education policy assumptions involve funding level and time. An evaluation of the "Trainer of Teacher Trainer Projects" (TTT) indicated that the federal government expected too much, too soon, and for too little.40-2 These findings have been confirmed by other studies. For example, a review of the Youth Employment Demonstration Projects Act suggested that the entitlement program, because of its large funding level in a given area and its high visibility, had greater impact than other parts of the Act.40-3
Other assumptions concerned the degree to which the programs and activities enacted to implement the policies were related to the goals of the policies, and the extent to which the resources committed to those programs were adequate. Researchers interested in these questions focused on the federal level, on the law-making process, and on legislative histories.
Equally fundamental assumptions questioned were the implied power relationships and the organizational models embedded in the policies, the programs established, and the evaluation procedures.41-1 If outcomes did not equal policy goals, perhaps the reasons could be found in what was happening at the local level. This line of inquiry has led to a loosely coupled body of research called "implementation studies," and is generally characterized by its local context and concern for outputs, rather than outcomes.
3. Implementation Research
In broad terms, implementation research seeks to explain what happens between policy intent and policy output and, in its more applied form, to do something about the slippage between the two. Implementation studies seek to understand what happens inside the often "black" box that lies between policy inputs and policy outputs, and to trace the "many decisions made over time by many local actors. The sequence of decisions describes the path of micro-implementation."41-2 Pressman and Wildavsky, pioneers in the field, defined implementation as the carrying out of an authoritative decision; that is, a policy choice.41-3 Berman posits that implementation analysis is "the study of why authoritative decisions (policies, plans, and the like) do not lead to expected results."42-1 In this sense, implementation is the study of slippage. On a more positive note, it is the study of how to increase congruence, or, as Berman states, "the study of the conditions under which authoritative decisions do lead to desired outcomes."42-2
4. Slippage at the Polite-Making Level
While some implementation researchers specifically omit the policymaking process from the field of implementation, it is included here because implementation is defined above as the line of research that seeks to explain the phenomenon of slippage between federal social policy intent and local output. While policy-making process researchers were rarely primarily concerned with explaining slippage, their findings contribute much to understanding its nature. As was once said of poets, a reader is entitled to all that he or she can get from the work.
Congruence between policy and outputs requires that there be a relatively clear understanding of the policy's intended goals. Such clarity is more the exception than the rule, as many studies have shown. In 1965, Bailey analyzed the myriad steps, claims, counter-claims, and accommodations between the White House and Congress and among the various committees and actors involved in the development and passage of the Elementary and Secondary Education Act.42-3 One result of the political accommodations and trade-offs involved in getting a final document approved was a high degree of vagueness in the law's goals. The bill was broad enough to accommodate those whose professional subgoals emphasized general aid to education; those who wanted to alleviate poverty; and those who wanted to improve the educational chances of low-achieving children, regardless of economic status.43-1 Moreover, because Congress wanted to avoid entangling itself in state-local relations, the formula developed for allocating funds, besides confusing educational and economic needs,43-2 called for funneling funds through the states. As Murphy put it, "The objective was a law, not reform."43-3 Former U.S. Commissioner of Education, Harold Howe, noted later:
Early evaluations of ESEA showed it had had little effect.43-5 In other words, slippage could be understood and accounted for, in part at least, in terms of the policy-makers' refusal or inability to express clear intent or purpose.
5. Slippage at the Administration Level
One source of ambiguity of goals in policies is often the role bureaucratic offices play in shaping--some might say confusing--the intent of the policy. McLaughlin analyzed how evaluation procedures were included in ESEA. Robert Kennedy, the chief Congressional proponent of including evaluation in the law, and Samuel Halperin, the person in USOE most responsible for drafting the law, differed in their under standing of the goals and subgoals of the law and the evaluation procedures. According to McLaughlin, USOE deliberately may have misled Kennedy about what the final wording of the provisions actually said.44-1
Another cause of confusion in the goals of a law is often the speed at which a bill is pushed through Congress by the Administration. The National Defense Education Act, which effectively ended (for the time at least) the half-century debate about federal aid to education, was passed in a matter of months. The legislative history of the Teacher Corps is another case in point.44-2 President Johnson, having become convinced that a cadre of teachers trained in the special needs of urban, poor schools could work wonders, educationally and politically, rammed the legislation through Congress in eighteen weeks. During that time, no legislative analysis was performed by either the Congress or the Administration.44-3
Bailey and Nosher45-1 continued the original work on ESEA by examining its evolution from Congress to the Administration, and from P.L. 89-10 into its programmatic form within the USOE. Alterations and modifications (some subtle and some not so subtle) appeared as USOE developed the administrative rules and regulations, including evaluation procedures, that would apply and that, at the same time, gave shape and substance to the policy. Murphy notes:
So close is the connection between the legislation and regulations developed by the bureaucratic offices that the "regs" have the force of law.45-3 Nevertheless, there is sufficient evidence to suggest that the subgoals of the administrative personnel in writing regulations may be somewhat different from the policy subgoals of Congress. In fact, some analysts suggest that it is in the interests of Congress and the administrative offices to have a disjunction between the goals of the two branches. Fiorina states:
The purpose here is neither to analyze the legislative process nor deride its possible shortcomings. Rather, it is to demonstrate that among the slips between policy Intent and policy results is the policy making process, as embodied in the programs enacted at the federal level.
6. Research on Organizational Theory
Still another fundamental assumption called into question involved the power and organizational relationships between the policy-makers in Washington and the implementers at the state and local levels. The idea that Congressional policy should lead directly and unchanged to federal programs and to intended results assumes that behavior at the local level is shaped, not just in compliance terms, but also in substance, by federal policies.
When it became clear that federal policies, as translated into pro grams, still had relatively limited effect on outcomes, researchers and policy-makers began to ask the rather obvious question, "What was happening at the local level?" The answer came back: a resounding "A whole range of things."
The already noted "Change Agent Study" and NIE's study of compensatory Education,47-1 as well as a host of other studies,47-2 have suggested that states and schools varied enormously in what they did with federal funds. These studies led to the development of a number of models that showed what happened to federal policies and programs at the local level. These studies constitute what Berman calls "implementation proper."47-3 As Williams explains,
The overwhelming conclusion of these studies is that federal, state, and district policies, as translated into programs, become reality only in the school or classroom. Berman and McLaughlin conclude that "Implementation strategies are the local decisions and choices, explicit or implicit, on how to put the innovation into practice. We found that these strategies could spell the difference between success or failure almost independent of the type of method involved."47-5 In becoming reality programs are shaped by local values, priorities, goals and subgoals. This may be the heart of implementation literature. John Pincus writes, it seems safe to say that most external efforts to promote innovation in schools have floundered in part through their ignorance of the tunes to which school districts must dance."48-1 Farrar, DeSanctis and Cohen write, "
The difference between federal hopes and local action is not simply or necessarily the result of federal mismanagement or local obstinacy; it is due to differing and often contradictory local perceptions of a program and its purposes."48-2 Rist writes, "The commonly shared understanding of the project and its goals by the various parties is yet a third factor influencing CETA/school communication and collaborations. When the different parties have different expectations and assumptions about what is to be accomplished, one or another group is bound to be disappointed.48-3 Where a program is successfully adopted, "mutual adaptation" takes place. This term was probably first used by education implementation researchers on the "Change Agent Study" and refers to the changes that take place in both the program and the local institution implementing the program. Sometimes, the programs are so 'bent out of shape" by the local values--so much slippage occurs--that there Is little resemblance between policy Intent and policy results.
Some analysts refer to this as "co-optation,"48-4 a situation that barely qualifies as implementation.49-1 In other cases, the federal program is adopted intact, but has little effect on the adopting school or district-- a case of installation, but, again, barely implementation. Berman elaborated these ideas into four possible processes: non-implementation in which no change occurs; co-optation, in which there is an adaptation in routine behavior that may serve as a defense mechanism against change; technological learning, in which there is no adaptation in the project, but adaptation of routinized behavior to accommodate the plan; and mutual adaptation on the parts of both the project and the deliverer's behavior. It is the last case that Berman considers to be successful implementation.49-2
The question for researchers then becomes, of what is local implementation a function? For many, the answer is the organizational structure in which the program is administered.
D. THEORIES OF ORGANIZATIONAL TYPES
Various models, ideal types, constructs and typologies of organizational behavior have been suggested. Elmore, for example, proposed four "ideal types," which emphasize different features of organizations and give different views of the implementation process. Those included the "systems management model," which treats organizations as value- maximizing units; the "bureaucratic process model," which emphasizes discretion and routine; the "organizational development model" in which members' commitments and needs for participation are central; and a "conflict and bargaining model," which views organizations as battlefields, where life is a matter of constant negotiations and bargaining.50-1
Elmore suggests that these "ideal types" may be useful for analyzing operational decisions, and further suggests that implementation theory would be better served by using existing theories of organization than developing new ones. The particular types he offers, however, are some what situation bound and contain a mix of prescriptive, descriptive, and normative elements. Elmore writes of his models: "Some models are basic ally normative--they're based on strongly held opinions about how organizations ought to operate. Some models are descriptive--they attempt to capture the essential objective attributes of organizations. In some instances, it is difficult to distinguish normative from descriptive elements. But in all instances, the model is a simplification of reality.50-2
A different typology was developed by James G. March and Herbert Simon in 1958 in their work, Organizations.50-3 In their three-part prepositional inventory of organizational behavior, each set of propositions posits a relationship between the organizational type and the behavior of its members. According to economist Robert Wolfson in 1982, these propositions "form a central core of notions to be found in all organizational theory; and furthermore, they are as applicable today as they were originally. "51-1
2. Three Theory Types
a." Passive-unit Type A theories
According to the March-Simon typology, the first category of theories is composed of "propositions assuming the organizational members ... are primarily passive instruments... "51-2
Theories of this type are characteristic of the structural approach to organizational theory, such as the scientific management theories of Taylor51-3 and the bureaucracy theories of Max Weber.51-4 Weber speaks of the "monocratic type of bureaucratic administration",51-5 in which the individual is little more than a machine part. The more the organization "develops, the more completely the bureaucracy is 'dehumanized,' the more completely it succeeds in eliminating from official business, love, hatred, and all purely personal, irrational and emotional elements which escape calculation."52-1 In all such organizations, although various inter mediate levels exist between policy-maker and policy-implementers, it Is the function of various levels to understand policy goals and faithfully transmit them downward. If the intermediaries and implementers are provided with appropriate and sufficient resources, the output can be expected to correspond closely with the intent. Weber claims that an organization of this type is capable of "attaining the highest degree of efficiency and is in this sense formally the most rational means of carrying out imperative control over human beings."52-2
b. "Reluctant participant" Type B theories
The second type of organizational propositions are those "assuming that members bring to their organizations attitudes, values, and goals; that they have to be motivated or induced to participate in the system of organizational behavior; that there is an incomplete parallelism between their personal goals and organizational goals, and that goal conflicts ... make power phenomena, attitudes and morale internally important."52-3
The reluctant participant category is characteristic of the motivational approach to management.52-4 Management, for example, allocates re sources to helping workers feel a part of the firm, to getting along with each other, and to internalizing the goals of the company.53-1
c. "Decision-making" Type C theories
In Type C theories, the propositions assume that "organizational members are decision-makers ... and that perception ... is central to the explanation of behavior in organizations."53-2 What is involved are complex organizational systems, characterized by multiple and semi-autonomous layers, and which operate with considerable freedom from those above, below, or adjacent to them.
Each layer or unit or subunit has multiple goals and subgoals, which it transmits to other layers where they are more imperfectly understood, and reshaped by new decision-makers.53-3 Within such systems, the front-line worker (teacher, nurse, welfare worker, unemployment counselor) has considerable discretion and is, of necessity, a decision-maker. Terms like "loosely coupled systems" and "street-level bureaucrats" are associated with this type of system. Each is discussed in detail below.
The question now is: if implementation is a function of organizational types, into which category do schools and other social service agencies fall?
3. Policy Assumptions Regarding Organizational Types
Most policies in the area of planned social change have implicitly assumed either Type A or Type B theories. The "institute model used by the National Defense Education Act Title XI programs and by various National Science Foundation programs during the early- and middle-196Os, for example, assumed Type B models; that is, the "reluctant participant" motivational model. Financial incentives were offered to the teachers (stipends, tuition), to university faculty (salaries, travel, research support), and to universities (tuition reimbursement, overhead) to motivate them to participate. Implicit incentives included community and professional recognition and openings on career ladders. If congruence still eluded the policy-maker, it was concluded, then either the type of motivation was incorrect (teachers would be motivated more by salary increases than summer stipends, for example), or the level of the reward was too low.
Type A theories are implicit in, and indeed characteristic of, much of "equity" and "equal opportunity" social legislation. Building in part on a growing recognition that some slippage could be explained in terms of confusing, vague, and sometimes contradictory goals, as well as by lack of enforcement procedures, and motivated by a variety of political and economic factors, including the high price tag attached to motivational models, federal policy-makers turned to a "systems analysis" model. Districts and schools might be motivated to participate because they received substantial sums of funds, but would have to meet fairly rigid rules for spending the federal funds, as well as complying with civil rights in allocating their own local funds. Key concepts in these policies were: (1) clarity, meaning understandable, detailed, and specific rules; (2) accountability, meaning a tight system of evaluation and feed back.; and (3) compliance, in order to increase congruence between policy and output.55-1
As has been suggested in Chapter I, these regulations did lead to increased formal compliance, but had limited affect in reducing slippage and increasing congruence. In fact, the very number of rules and regulations made compliance difficult and, as some analysts suggest, made it almost impossible to deliver educational services at all. This recognition of continued slippage--and a growing understanding of how schools work within themselves, within their communities, and within the federal system--has led to a developing consensus that the most appropriate organizational model is that which posits the members as decision-makers.
4. Loosely Coupled Systems
In an influential article on school organizations, Karl Weick suggested in 1977 that schools displayed few of the patterns of bureaucratic systems, but were instead "loosely coupled." Loosely coupled systems have multiple goals, multiple layers of decision-makers with different goals and subgoals and often with few common values, fluid decision- making in which negotiation and consensus become more important than hierarchical control, and much independence of persons at each level. 56-1
Persons with different positions in a school district often have different, sometimes contradictory, goals from others in the district. There are often, at best, tenuous connections between policy-makers (school board, superintendent, principal) and teachers who implement policies. As a result, there is a split between intention and action.56-2 "If there are several diverse intending components, then the relationship between any one intention and any one action will be imperfect," according to Weick,56-3 This split means that setting clear goals, when that Is possible, may have little effect on actions.
Managers of loosely coupled systems cannot, therefore, expect that once they have set goals, all they need do is communicate those goals clearly to lower levels. As Clark wrote, this would require:
Nor can managers rely on authority. In Elmore's apt phrase, staff In such systems are "almost generically immune to hierarchical control."56-5
Lipsky explains this immunity in terms of "street level bureaucracy."57-1 The staff persons in a social service agency are virtually the only ones with direct client contact, and have considerable discretion in the de tailed decisions of the client's treatment.57-2 It is at this point, at the lowest level, if one were to use a hierarchical metaphor, that implementation begins and the key variables between policy, program, and outputs are to be found.
To say that the street-level bureaucrat is the missing link, however, is not to explain the behavior of the bureaucrat. It is in this area that the field of implementation, the field of explaining slippage, is least developed. Murphy, for example, states that, in its brief history, implementation research has "generated some useful insights into the characteristics of intergovernmental relations, into how organizations respond to external intervention, and into the nature of the implementation process," but is still short on "theory, advice, and breadth of coverage."57-3
Some analysts, such as Elmore,57-4 suggest that slippage can be explained by three factors: (1) discretion; (2) a large number of vague and often conflicting formal (e.g., reporting) and substantive (client service) demands; and (3) the need to develop coping mechanisms.57-5 Unfortunately, for analysis, the first two are largely descriptive and probably not very amenable to policy intervention.
The third is of more interest. Elmore suggests that the "major concern for the street-level implementer is how to control the stress and complexity of day to day work. Out of this concern grows a broad set of informal routines that students of street-level bureaucracies call "coping mechanisms." The result, Elmore claims, is that, faced with a broad legislative mandate and a large, complex set of new responsibilities, local implementers Invent operating routines that help simplify their work, but which often run counter to the intent of the legislation.58-1
Indeed, as Lipsky writes, the street-level bureaucrat "may be inherently incapable of responding favorably to contemporary demands for improved and more sympathetic services to some clients."58-2 The staff persons, confronted with limited resources and perceived threats to personal safety, resort to stereotyping and simplifying procedures.58-3
It may be, however, that both Elmore and Lipsky are overly pessimistic. As decision-makers, street-level bureaucrats appear to define problems in terms they can understand and solve. Higher level policy- makers may have some choice In selecting organizations that are most likely to have staff who define the problem In ways compatible with policy objectives, and then can attempt to phrase their goal statements and define their programs in terms that are consistent with those of the Implementers. Moreover, they may be able to create the conditions that encourage local-level agency directors to recruit staff whose goals are consistent with those of the policy, and that encourage street-level bureaucrats to focus on professional rather than personal subgoals. In any event, if Type-C organizations are characterized in part by their requirement that staff must be decision-makers, then individual decision- making within Type-C organizations becomes the central question.
E. THE SIMON HYPOTHESIS: STREET-LEVEL DECISION-MAKING IN LOOSELY COUPLED SYSTEMS
1. Herbert Simon and Subgoals
In his Nobel Laureate address in 1978,59-1 Simon related an incident involving urban public recreational facilities which were jointly managed by the local school district and the city public works department. The heads of the two agencies were apparently in complete agreement about the goal of the program--providing youth with the best recreational facilities. The heads did not appear to be engaged in territorial battles; yet, there was constant tension between them at the operational level as to the allocation of funds between park maintenance and playground supervision.
In the language of rational economic decision-making, the answer would be quite simple. The leaders would simply "balance off the marginal returns of one activity against that of the other." The reason for the tension, Simon suggested, was that they did not equate at the margins because "they simply could not do so."59-2 Theories of rational economic decision-making contain a number of powerful assumptions. Among these are the presence of clear and measurable goals, technologies that are consistently related to those goals, and means of measuring marginal costs and marginal returns. These, in turn, imply that the decision-maker is certain, not only about the consequences of each action, but also about present and future evaluation of these consequences. Rationality also assumes perfect knowledge of alter natives and complete freedom of choice among them.
In Simon's example, none of these conditions were present. The goal of a "best recreational facility" is open to a multitude of operational interpretations. There were, according to Simon, "no measurable production functions from which quantitative inferences about marginal productivities could be drawn; and such qualitative notions of a production function that the two managers possessed were mutually incompatible."60-1
According to Simon, the situation is "ubiquitous in human decision- making."60-2 It is, however, more intense in social service agencies where there is little pretense that the conditions of rationality apply and where the range of behavior is broad. Later in the speech, he said:
There can no longer be any doubt that the micro assumptions ... of perfect rationality are contrary to fact. It is not a question of approximation; they do not even remotely describe the processes that human beings use for making decisions in complex situations.60-3
Several alternatives to the rational decision-making model have been developed in various disciplines falling under the general rubric of bounded rationality. Two concepts are central to these theories: "search" and "satisficing." "Search"61-1 involves the extent to which a person can and does invest in identifying solutions to problems. "Satisficing"61-2 refers to the criteria used in making a selection. In making "bounded decisions" persons have two choices. The first is to choose satisfactory means to achieve optimal goals. There is no assumption that the goal will be reached or reached in an efficient manner. In the second case, abstract, global goals are replaced with tangible subgoals whose achievement can be observed and measured.61-3
2. Subgoals and Decision-Making
Simon suggests that the behavior of the park and school heads can best be explained by decisions based on subgoal identification.61-4 It is the thesis of this study that subgoals form the basis of decisions made by street level bureaucrats about the services to be delivered to clients. It is this procedure that immediately and directly accounts for slippage or congruence.
Faced with complex situations in which there are no clear guidelines and in which those guidelines that do exist are often contradictory, staff define problems in such ways that they can be solved with the resources available. Further, these subgoals, rather than being derived from broad, abstract goals of public policy, are derived from the needs of the individuals and the community and organizational setting in which members operate.
To say that the decisions are made by identifying subgoals does not mean necessarily that the individual decision-maker was aware that this was the basis of a decision. In rational economic theory, it is often said that whether or not individuals "think" they are making decisions so as to maximize profit, they "act" as if they are.
3. Types of Decisions
Street level bureaucrats are called on to make decisions continually. Indeed, it is the characteristic of Type C, loosely coupled organizations that members are required to make decisions. The decisions of interest in this study are those involving the allocation of resources for the provision of services to clients (such as the time to be spent on a given subject, or the relative emphasis on academic skills vs. on the job training). Other decisions are important to the extent they influence resource allocation and the services offered.
Most of the decisions street-level bureaucrats make fall under the general rubric of normal operating procedures, that is, the repetitive, routine decisions of daily professional life.62-1 While perhaps not as spectacular as innovative policy decisions, these routine decisions are, as Inbar states, "the central datum in understanding the output of bureaucratic decision making,"62-2 or, in the terms of this study, in explaining slippage.
Subgoals are operational objectives, the means of achieving which are perceived by the individual to be within the individual's control. While subgoals are perceived to be connected to broader goals, they are not necessarily derived from them. Rather, they are the product of decision- makers' efforts to make sense of complex situations and to convert that information into a problem that can be solved. Formulations of subgoals, according to Simon, "depend on the knowledge, experience and organizational environment of the decision maker. In the face of ambiguity, the formulation can also be influenced in subtle and not so subtle ways by his self-interest and power drives."63-1 It should be expected that persons with different backgrounds, different community and organizational set tings and different levels of personal needs will have different subgoals and will, therefore, make different decisions about the types of services to deliver to clients.
5. Types of Subgoals
Based on Simon's hypothesis, this study classified subgoals into three categories. These are "personal," "professional," and "organizational." "Personal" subgoals are those derived from "self interest and power drives;" "professional" subgoals are those derived from "knowledge and experience; "organizational" subgoals are those derived from the organization within which the person works.
a. "Personal" subgoals
Personal subgoals are those objectives related to an individual's needs. Like other subgoals, they are connected to actions that can be controlled by the individual. Living a long and happy life may be a personal goal, but there are few actions an individual can undertake that will immediately, or even in the long term, guarantee such a condition. A person can, however, control the actions and allocate time, skills and other resources necessary to maintain friendly relations with colleagues or superiors.
Personal subgoals are derived from the inherent needs of individuals, which are characterized in the relevant literature in various ways, but which seem to be constant across humans.64-1 There are three assumptions about the relationship between basic needs and the identification of subgoals in this study. One is that, in general terms, needs are the same across individuals. Persons in all walks of life, in all institutions, and in all organizational settings have survival, social and self development needs. Second, some needs take precedence over others. When the need for economic security is severely threatened, the need for professional recognition becomes less important, at least in the short term. Third, while needs are similar, the emphasis people place on the achievement of any one subgoal will vary.
Facing complex situations requiring decisions, the street-level bureaucrat will define the problem, in part at least, in terms of personal needs. Indeed, this is the core of Lipsky's thesis. The street- level bureaucrat, motivated by perceived threats to personal security, resorts to coping mechanisms including stereotyping and simplification.
Terborg and Komocar65-1 suggest a person's perception of a situation is a function of four factors: the characteristics of the organization (structure, resources, history and administration); the characteristics of the individual (demography, tenure, functional specialty, attitudes); characteristics of social-interpersonal environment (communication, decision making, trust, support); and other environmental factors (government intervention, community response).65-2 These factors are used in this study to explain why individuals in some situations are more apt to devote resources to satisfying survival needs and not to achieving higher level goals, such as the sense of doing a good job or gaining the respect of colleagues.
b. "Organizational" subgoals
These are the discrete subgoals held by the individual that relate directly to the individual's perception of his or her mission and role as a member of the particular organization within which he or she works. For this analysis, the organizations are the school, the community based organization, or the particular firm at which an employee works.65-3
In any organization, routine decisions are rarely made on the basis of their immediate and direct relationship to the organization's primary goals; in the case of schools, the education of children, however defined. Rather, the primary subgoals of individuals as related to their organization include the survival of the organization, its welfare, and their position in it. Carol Weiss, for example, after evaluating a program that was a forerunner of VISTA, suggested that while program goals were global--including improving community services, helping people, making agencies responsive to the needs of people--program staff were concerned with survival issues--seeing that the "federal grant was continued, drumming up and maintaining support from community groups, recruiting new trainees, engaging a competent staff."66-1 Chambers suggested school administrators act to "maximize the perceived quality of the school in the eyes of the school board. They do so in such way as to avoid controversy and abrupt changes in policy."66-2
Organizational subgoals derive from the structure, mores and workways of the organization as perceived by the individual. The extent to which organizational subgoals are important in decision-making depends on a number of factors. One factor is the degree to which the organization is tightly or loosely coupled. Another involves the length of time the individual has been with the organization. A related factor concerns how strongly the individual feels committed to an organization. Sometimes this is the result of a strong leader, capable of setting goals and direction;67-1 a sense of a unified, supportive community; or a shared sense of mission.67-2
The strength of organizational subgoals also depends on the strength of other subgoals. For example, a person with strong professional subgoals that are independent of the organization may not have as strong organizational subgoals as one whose professional commitments are weak.
It should be noted that "organizational subgoals" as used here do not refer to the particular content of an organization's goal--making a profit for a business firm or getting a person placed on an election committee. Rather, what is of concern are those organization-related subgoals held by the individual, and the strength of the individual's commitment to the organization.
c. "professional" subgoals
Professional subgoals derive from what might be called the institutional, as opposed to organizational, affiliation of the person. Education and medicine are examples. Institutions have goals such as the education of children or the improvement of health. They also have generally accepted means of obtaining those goals. In addition, they expect from their members certain behaviors, attitudes and values.
Members of one institution are likely to define problems in somewhat different ways than persons in another institution. The superintendent of schools viewed the problem as relating to the education and safety of children; the park superintendent as the maintenance of facilities. The sources of institutional subgoals are mainly the educational and institutional experiences of the individual.
6. The Interaction of Subgoals and Policy Outputs
The goals of any given federal social policy are but one set of the many sets of vague, global goals that staff of social service agencies are asked to achieve. Other, for example, may include those from different federal programs, state directives, local mandates, and so on. As such, the goals contribute relatively little to how the street level bureaucrat defines the problem of which services to deliver to clients and, consequently, how to allocate resources.
This study examines a number of propositions about how subgoals influence the implementation of federal policies:
This chapter has attempted to show that policy-makers must and do posit an equality between policy intent and both outcomes and outputs, but that, in fact, results in the area of federal social policy rarely come close to matching the expectations of policy-makers. "Slippage" is, for the purposes of this research, limited to and defined as the incongruity of policy-maker Intent and policy-implementer output.
A variety of possible explanations were explored and, while each accounts for some aspect(s) of slippage, even when taken together, they seem to leave an important area unexplained. Neither the federal policy making, goal-setting, program-making, regulation-setting, enforcement processes nor the traditional theories of organizational behavior have offered sufficient explanation of the slippage phenomenon. It is there fore logical to consider alternative theories of organizational types that emphasize decision making by staff of social service agencies.
Herbert Simon has suggested that, in loosely coupled systems, individual decision-making may be the key to understanding organizational behavior. He further suggests that this decision-making can best be understood in terms of subgoals, the setting of achievable objectives and the allocation of resources to attain them. By extension, this suggests that slippage may be reduced and congruence between policy and outcomes increased If subgoals of street level bureaucrats are taken into consideration throughout the policy-making, policy-implementing process.
To test these assertions, a model derived from theories relating to loosely coupled systems, street-level bureaucrats and satisficing decision-making is used. Ethnographic data from the implementation of a federal policy in different organizational settings, with staff of different professional backgrounds, are used. Analysis of these data appears to support the following assertions:
1. Three types of subgoals exist in the decision-making process of street-level bureaucrats. 2. These subgoals are central to understanding the decision-making of street-level bureaucrats. 3. Subgoals differ depending on the institutional background, the community setting, and the organizational context of the staff. 4. Slippage and congruence can be explained, in part, by the compatibility of the subgoals of staff charged with implementing social policies with the goals and subgoals of the policy.
In Chapter 3, the methodologies used are discussed. Subsequent chapters include presentations, analyses and Interpretations of the data and tentative findings. Implications for future federal/state social policies, as well as for research relating to both decision-making within loosely coupled systems and research in other areas, are discussed in the concluding chapter.
32-1Paul Berman, "The Study of Macro- and Micro-Implementation," 7 26 (Spring, 1978): 170.
33-1Arthur Wise, Legislated Learning: The Bureaucratization of the
American Classroom (Berkeley: University of California Press, 1979),
34-1P.L. 85-864. It is interesting to note that this legislation,
which ushered in a new, more active role for the federal government in
education, was justified, not in terms of education, but in terms of
national defense. As will be discussed later, it was evaluated
primarily in terms of academic achievement.
35-1Title I, which accounted for the largest portion of allocated funds, supported programs for disadvantaged children and youth. Other Titles allocated funds to school libraries; supplementary educational centers; guidance, counseling and testing services; strengthening state departments of education; and state-run, innovative projects.
36-1The term "evaluation studies" is somewhat of a misnomer as applied to many of the reports on federal education programs. Most of the studies cited in the Annual Evaluation Report produced by the U.S. Office of Education (now the Department of Education) consist of little more than self-reports of contractors or grantees. As to be expected, most of these reports are positive and indicate that recipients are happy, but few meet even the minimal requirements of research. What are reported below are some of the studies that were designed with the principles of research in mind.
37-1Walter F. Crockett, The Experienced Teacher Fellowship
(Washington, D.C.: National Science Foundation, 1968).
38-1Stanley L. Helgeson, Patricia Blosser and Robert E. Howe,
Volume I: Science Education: Executive Summary, (Athens, Ohio: The
Ohio State University, 1977), p. 4.
39-1Westinghouse Learning Corporation/Ohio University,
of Head Start: An Evaluation of the Effects of Head Start Experience
on Children's Cognitive and Affective Development (Springfield, VA:
Clearinghouse for Federal Scientific and Technical Information,
40-1For a discussion of the goals of the early evaluation studies,
as well as the limitations of the studies, see David K. Cohen,
"Politics and Research Evaluation of Social Action Programs in
Education," Review of Educational Research 40 (April, 1970): 213-38.
41-1Pressman and Wildavsky suggest that a decision to enact a
policy implies a theory or at least an assumption about the
relationship between the policy and its intended results. Jeffrey
Pressman and Aaron Wildavsky, Implementation (Berkeley: University of
California Press, 1978).
43-1Cohen, "Politics and Research," p. 220.
44-1 Milbrey Wallin McLaughlin, Evaluation and Reform: The
Elementary and Secondary Education Act of 1965 (Cambridge: Ballinger
Publishing Co., 1975).
45-1Stephen K. Bailey and Edith Mosher, ESEA: The Office of
Education Administers a Law (Syracuse: Syracuse University Press,
46-1Morris P. Fiorina, Congress: The Keystone of the Washington Establishment (New Haven: Yale University Press, 1977), p. 48.
47-1Compensatory Education Services (Washington, D.C.: National Institute of
Education, 1977), pp. v-lx.
48-1John Pincus, "Incentives for Innovation in Public
Schools," In Social Program Implementation eds. Walter Williams
and Richard Elmore, (New York: Academic Press, 1978), p. 67.
50-1Richard F. Elmore, "Organizational Models of Social
Program Implementation," Public Policy 26 (Spring, 1978):
185-228. Elmore credits Graham Allison's study, Essence of Decision:
Explaining the Cuban Missile Crisis (Boston: Little Brown, and Co.,
1971), with the original "insight." He notes, however, that
his development differs from Allison's because of "the nature of
the subject matter ... Defense strategy, not surprisingly, raises a
set of problems considerably different from the implementation of
51-1Robert J. Wolfson, "Formal Lexicon for the Social
Sciences," Syracuse Scholar 3 (Spring, 1982): 79.
52-1Max Weber, From Max Weber: Essays in Sociology translated and
edited by H. H. Gerth and C. Wright Mills (New York: Oxford University
Press 1946), 214-16.
53-1The theory was based largely on studies conducted at the
General Electric plant in Hawthorne, New York between 1924 and 1926.
See, Elton Mayo, Human Problems of an Industrial Civilization (New
York: Macmillan, 1933); Fritz J. Rotlilisberger and William J.
Dickson, Management and the Worker (Cambridge, Mass.: Harvard
University Press, 1939); For a review of the early studies, as well as
later reinterpretations, see Malcolm Warner, "Organizational
Experiments and Social Innovation," in Handbook of Organizational
Design: Volume I eds. Paul C. Nystrom and William H. Starbuck (New
York: Oxford University Press, l981) pp. 170-1.
55-1Carl Stenberg, Aid Reform: The Grants Consolidation Approach," in Grants Consolidation: A New Balance In Federal Aid to Education eds. C. Phillip Kearney and Elizabeth VanderPutten, (Washington, D.C: Institute for Educational Leadership, The George Washington University, 1979), pp. 51-62.
56-1Karl E. Weick, "Educational Organizations as Loosely
Coupled Systems" Administrative Science Quarterly 21(March,
57-1One of the first articles to discuss the phenomenon of the
street level bureaucrat was Michael Lipsky, "Toward a Theory of
Street Level Bureaucrats," in Theoretical Perspectives on Urban
Politics eds. Willis B. Hawley and Michael Lipsky, (Englewood Cliffs,:
Prentice Hall, 1976), p. 196-213.
61-1Perhaps the first to develop a theory of search was George J.
Stigler, "The Economics of Education," Journal of Political
Economy 69 (June, 1961): 213-5.
63-1Simon, "Rational Decision Making," p. 500.
64-1Some of the answers could be found in the vast psychological literature on individual behavior. While such considerations are of interest, they are outside the scope of this work. This research Assumes that individuals have similar needs and will demonstrate that the extent to which they define problems in terms of survival needs is dependent on environmental and organizational factors.
65-1James R. Terborg and John Komocar, "Individual and Group
Behavior in Schools as a Function of Environmental Stress," in
Organizational Behavior in Schools and School Districts ed. Samuel B.
Bacharach, ed., (New York: Praeger, 1981), pp. 465-493.
66-1Carol Weiss, "Evaluating Educational and Social Action
Programs: A Treeful of Owls," in Evaluating Social Action
Programs Carol Weiss, ed., (Boston: Allyn and Bacon, Inc.,
67-1Hargrove, et al., studied the extent to which schools implemented
P.L. 94-142 (Handicapped Act) effectively. They found that in all the
"high performance schools, principals were all
"authoritative democrats," capable of both persuading and
manipulating teachers. Teachers were enthusiastic about their
principals. Erwin C. Hargrove, et al., "School Systems and
Regulatory Mandates: A Case Study of the Implementation of the
Education for All Handicapped Children Act," in Organizational
Behavior in Schools and School Districts ed. Samuel B. Bacharach, pp.