A Critical Account of Policy Implementation Theories: Status and Reconsideration

Narendra Raj Paudel's picture
This paper critically examines the theory of public policy implementation, discusses the issues of policy implementation studies and examines the applicability of such theory. The first generation implementation researchers find out the problems of policy implementation, i.e. uncertain relationship between policies, decisions and implemented programs. Similarly, the second-generation implementation studies focus on the 'development of an analytical framework of implementation', which includes the top-down, bottom-up perspectives and their synthesis. Therefore, the third generation implementation research should concentrate on explicit implementation theory building, which has not yet been realized.
Main Article: 

1. Introduction

Research on policy implementation has been a hot discourse among social scientist since 1970s when Pressman and Wildavsky (1973) brought the issue of policy implementation to the forefront. The main aim of research at that time was to find out a concrete theory of policy implementation. It argues that policy implementation studies are not value-free due to socio-cultural, political and economic variations in the country’s context. It may lead to new forms of policy implementation not yet well understood.

This paper tries to understand the concept of policy implementation, critically analyze the theories of policy implementation, discuss the issues of policy implementation studies and examine the applicability of such theories.

2. Concept of Implementation

Implementation inevitably takes different shapes and forms in different cultures and institutional settings. This point is particularly important in an era in which processes of ‘government’ have been seen as transformed into those of ‘governance’ (Hill and Hupe, 2002, p1).

Implementation literally means carrying out, accomplishing, fulfilling, producing or completing a given task. The founding fathers of implementation, Pressman and Wildavsky (1973) define it in terms of a relationship to policy as laid down in official documents. According to them, policy implementation may be viewed as a process of interaction between the setting of goals and actions geared to achieve them (Pressman and Wildavsky, 1984: xxi-xxiii). Policy implementation encompasses those actions by public and private individuals or groups that are directed at the achievement of objectives set forth in policy decisions. This includes both one-time efforts to transform decisions into operational terms and continuing efforts to achieve the large and small changes mandated by policy decisions (Van Meter and Van Horn, 1975, p447).

According to Mazmanian and Sabatier (1983, p20-21), policy implementation is the carrying out of a basic policy decision, usually incorporated in a statute, but which can also take the form of important executive orders or court decisions. The starting point is the authoritative decision. It implies centrally located actors, such as politicians, top-level bureaucrats and others, who are seen as most relevant to producing the desired effects. In their definition, the authors categorize three types of variables affecting the achievement of legal objectives throughout this entire process. These variables can be broadly categorized as: tractability of the problem(s) being addressed; the ability of the statute to favorably structure the implementation process; and the net effect of a variety of political variables on the balance of support for statutory objectives.

O’Toole (2003, p266) defines policy implementation as what develops between the establishment of an apparent intention on the part of government to do something or stop doing something and the ultimate impact of world of actions. More concisely, the writer remarks that policy implementation refers to the connection between the expression of governmental intention and actual result (O’Toole et al., 1995, p43). As part of policy cycle, policy implementation concerns how governments put policies into effect (Howlett and Ramesh, 2003, p13).

Elmore identified four main ingredients for effective implementation: (1) clearly specified tasks and objectives that accurately reflect the intent of policy; (2) a management plan that allocates tasks and performance standards to subunits; (3) an objective means of measuring subunit performance; and (4) a system of management controls and social sanctions sufficient to hold subordinates accountable for their performance. Failures of implementation are, by definition, lapses of planning, specification and control (Elmore, 1978, p195).

Successful implementation, according to Matland, requires compliance with statutes’ directives and goals; achievement of specific success indicators; and improvement in the political climate around a program (Quoted in Hill and Hupe, 2002, p75). In this line, Giacchino and Kakabadse (2003) assess the successful implementation of public policies on decisive factors. According to them, these are the decisions taken to locate political responsibility for initiative; presence of strong project management or team dynamics and level of commitment shown to policy initiatives.

Besides this, the success of a policy depends critically on two broad factors: local capacity and will. Questions of motivation and commitment (or will) reflect the implementer’s assessment of the value of a policy or the appropriateness of a strategy. Motivation or will is influenced by factors largely beyond the reach of policy environmental stability; competing centers of authority, contending priorities or pressures and other aspects of socio-political milieu can also profoundly influence an implementer’s willingness. This emphasis on individual motivation and internal institutional conditions implies that external policy features have limited influence on outcomes, particularly at lower level in the institution (Matland, 1995).

From the above discussion, implementation can be conceptualized as a process, output and outcome. It is a process of a series of decisions and actions directed
towards putting a prior authoritative decision into effect. The essential characteristic of implementation process is the timely and satisfactory performance of certain necessary tasks related to carrying out of the intent of the law. Implementation can also be defined in terms of output or extent to which programmatic goals have been satisfied. Finally, at the highest level of abstraction, implementation outcome implies that there has been some measurable change in the larger problem, that was addressed by the program, public law or judicial decisions (Lester et al., 1995, p87).

3. Evolution and Critical Understanding of Policy Implementation Theories

In general, implementation research is supposed to have evolved through three generations. The first generation of research ranged from the early 1970s to ’80s; the second generation from the 1980s to 90s; and the third generation research from 1990 and onwards (Matland, 1995).

3.1 First Generation Implementation

The first generation implementation research was focused on how a single authoritative decision was carried out, either at a single location or at multiple sites (Goggin et al., 1990, p13). Pressman and Wildavsy’s work is a prime example of this generation of research (Hill and Hupe, 2002, p66). Their analysis discovered the problem of policy implementation--the uncertain relationship between policies and implemented programs--and sketched its broad parameters.

The first generation was a more systematic effort in the 1980s to understand the factors that facilitated or constrained the implementation of public policies (Sabatier and Mazmanian, 1981). This analysis shows how local factors such as size, intra-organizational relationships, commitment, capacity and institutional complexities mould responses to policy (McLaughlin, 1987, p172). The first generation research was characterized by pioneering but largely atheoretical, case-specific, and non-cumulative studies such as that of Pressman and Wildavsky (Googin et al., 1990, p13).

3.2 Second Generation Implementation

The second generation implementation studies focused on describing and analyzing the relationships between policy and practice. These researches generated a number of important lessons for policy, practice and analysis. For example, policy cannot always mandate what matters to outcomes at local level; individual incentives and beliefs are central to local responses; effective implementation requires a strategic balance of pressure and support; policy-directed change ultimately is a problem of the smallest unit (McLaughlin, 1987, p176).

The second generation research also taught researchers the importance of time periods: at what point in history implementation occurs and over what period of time (Van Horn, 1987, quoted in Goggin, et al., 1990, p15).

The second generation studies recognized implementation’s variability over time and across policies and units of government. Thus, it concerned itself with explaining implementation success or failure and relied heavily on an explicit or implicit model of policy implementation process (Goggin, et al., 1990, p183).

The second generation research was engaged in ‘the development of analytical frameworks’ (Goggin et al., 1990, p14). The construction of models and research strategies, however, immediately led to a major confrontation between the so-called top-down and bottom-up perspectives of policy implementation (Winter, 2003, p213). Till now, no general implementation theory has emerged, although many implementation scholars have had the development of such a theory (Winter, 2003, p205).. However, as implementation research evolved, two schools of thought developed for studying and describing implementation: top-down and bottom-up (for comparison of both perspectives, see table 1).

Top-down perspective: The top-down perspective assumes that policy goals can be specified by policymakers and that implementation can be carried out successfully by setting up certain mechanisms (Palumbo and Calista, 1990, p13). This perspective is ‘policy-centered’ and represents the policymaker’s views. A vital point is the policymaker’s capability to exercise control over the environment and implementers (Younis and Davidson, 1990, p5-8.) Van Meter and Van Horn (1975) and Mazmanian and Sabatier (1978, p18) see implementation as concerned with the degree to which the actions of implementing officials and target groups coincide with the goals embodied in an authoritative decision. Jeffrey Pressman and Aaron Wildavsky (1973) discuss the extent to which successful implementation depends upon linkages between different organizations and departments at local level. Brian Hogwood and Lewis Gunn (1978) offer recommendations to policymakers about effective implementation.

Table 1. Differences between Top-down and Bottom-up Implementation Perspectives

Variables Top-down perspective Bottom perspective
Policy decision-maker Policymakers Street-level bureaucrats
Starting point Statutory language Social problems
Structure Formal Both formal and informal
Process Purely administrative Networking, including administrative
Authority Centralization Decentralization
Output/Outcome Prescriptive Description
Discretion Top-level bureaucrats Bottom-level bureaucrats

The top-down perspective exhibits a strong desire for ‘generalizing’ policy advice. This requires finding consistent and recognizable patterns in behavior across different policy areas (Matland, 1995, p146). The top-town perspective emphasizes formal steering of problems and factors, which are easy to manipulate and lead to centralization and control. Interest will be directed towards things such as funding formulas, formal organization structures and authority relationships between administrative units, regulations and administrative controls like budget, planning and evaluation requirements (Elmore, 1978, p185, 189, 191).

‘It begins at the top of the process, with as clear a statement as possible of the policy-maker’s intent, and proceeds through a sequence of increasingly more specific steps to define what is expected of implementers at each level. At the bottom of the process, one states, again with as much precision as possible, what a satisfactory outcome would be, measured in the terms of the original statement of intent’ (Elmore, 1978).

The top-down perspective largely restricts its attention to actors who are formally involved in the implementation of a specific program (Winter, 1990, p28). The top-down researchers focus on a specific political decision, normally a law. They follow the implementation down through the system, often with special interest in higher-level decision-makers. They would typically assume a control perspective of implementation, trying to give good advice on how to structure the implementation process for the above in order to achieve the purpose of legislation and to minimize the number of decision points that could be vetoed (Winter, 2003, p213).

However, the top-down perspective of implementation is not free from criticism. It faces the following criticisms. First, the top-down models take the statutory language as their starting point. This fails to consider the significance of actions taken earlier in the policy-making process. Second, top-downers have been accused of seeing implementation as a purely administrative process and either ignoring the political aspects or trying to eliminate them (Berman 1978; March and Sætren, 1986). Besides, this prescription fails to recognize the political realities that account for policies with multiple goals, vague language and complex implementations structures (May, 2003, p224). Third, top-downers put exclusive emphasis on statute framers as key actors. This criticism has two primary variants. One argues from a normative perspective that local service deliverers are experts and have the knowledge of the true problems; therefore, they are in a better position to propose purposeful policy. Another criticism is that top-downers neglect the reality of policy modification or distortion at the hands of implementers. They object to the implicit assumption that policymakers control processes that affect implementation. This model also assumes that all priorities are known and can be ranked. Another weakness is that it has no behavioral basis. As the rational model is unachievable in practice, the result will always be implementation failures (Elmore, 1979,603-4). Similarly, Berman argues that choosing the top-down strategy can lead to resistance, disregard and pro forma compliance.

The top-down models, however, see local actors as impediments to successful implementation--agents whose shirking behavior needs to be controlled. The second variant argues from a positive perspective that discretion for street-level bureaucrats is inevitably so great that it is simply unrealistic to expect policy designers to be able to control the actions of these agents.

The bottom-up perspective: The bottom-up perspective directs attention at the formal and informal relationships constituting the policy subsystems involved in making and implementing policies (Howllet and Ramesh, 2003, p190). This perspective has as its starting point a problem in society. The focus is on individuals and their behavior, and in this respect street-level bureaucrats are made central in the political process.

The street-level bureaucrats are considered to have a better understanding of what clients need as it is they who have direct contact with the public. Michael Lipsky (1980) propounds a theory of ‘street-level bureaucracy’. Lipsky’s theory focuses on the discretionary decisions that each field worker or ‘street-level bureaucrat’--as he prefers to call them--makes in relation to individual citizens when they are delivering policies to them. This discretionary role in delivering services or enforcing regulations makes street-level bureaucrats essential actors in implementing public policies. Indeed, Lipsky claims that street-level bureaucrats are the real policymakers (Winter, 2003, p214). However, implementation failure is connected with discretion and routine, together with personal malfunctions, and one has to identify where the discretion is congregated and which organization’s repertoire of routines needs changing (Elmore, 1978, p200).

Similarly, in Hull and Hjern (1981), the bottom-up perspective is to identify the many actors that affect the problem and to map relations between them. In these network analyses, both public and private actors become essential, and the analyses often include several policies that affect the same problem, whether or not it is intended in those policies.

Hull and Hjern (1987) focus on the role of local networks in affecting a given problem in the implementation process, and also propound a way of identifying the networks. It is a combination of a snowball and socio-metric methods (Quoted in Winter, 2003, p214). This method enables them to map a network that identifies the relevant implementation structure for a specific policy at local, regional and national level, and allows them to evaluate the significance of government programs vis-à-vis other influences such as market. It also enables them to see strategic coalitions as well as unintended effects of policy and the dynamic nature of policy implementation (Matland, 1995, p149). According to them, central initiatives are poorly adapted to local conditions. Program success depends in large part on the skills of individuals in the local implementation structure, who can adapt the policy to local conditions. It depends only to a limited degree on central activities. Therefore, their analysis is important in drawing attention to implementation activities and structures at local operational level.

According to Berman (1978), policy implementation takes place at two levels: macro and micro. At macro implementation level, centrally located actors devise a government program; at micro implementation level, local organizations react to macro-level plans; develop and implement their own programs. However, he argues that, most implementation problems stem from the interaction of a policy with micro-level institutional settings. Central-level actors can indirectly influence micro-level factors. It is because the rules created by central actors are dominated by local implementing contextual factors.

However, the bottom-up perspective does not provide satisfactory solutions to the problems of public policy, as its rejection of the authority of policymakers is questionable in the light of standard democratic theory. Policy control should be exercised by actors whose power derives from their accountability to sovereign voters through their elected representatives. The authority of local service deliverers does not derive from this power base (Matland, 1995, p150).

Another criticism is that this perspective cannot successfully explain why coping strategies occur and why they vary. It is difficult to think of ways to change the street-level behavior in the context of this model, and no thought is given to how to use discretion as a device for improving the effectiveness of policies at street level (Elmore, 1978, p208). It has also been demonstrated that people with very little education and poor social background are less likely to benefit from social services compared to more educated and wealthier people. This is the case even when these social services are targeted primarily at the former category. Hence, creaming is done not only by street-level bureaucrats but also by the self-selection of the target groups themselves (Winter, 1990, p32).

The methodological perspectives overemphasize the level of local autonomy. Michael Lipsky (1978) gives great importance to street-level bureaucrats. Similarly, Hull and Hjern (1987) emphasize on local networking in service delivery in one or more local areas. In such situations, variations in action can be explained largely by local-level differences; yet, all actions may fall within a limited range where borders are set by centrally determined policies. While central actors do not act in detail or intervene in specific cases, they can structure the goals and strategies of those participants who are active. The institutional structure, the available resources and the access to an implementing arena may be determined centrally, and substantially affect policy outcomes (Matland, 1995, p150).

Synthesis of both perspectives: Both top-down and bottom-up perspectives draw attention to the implementation process. However, there is a conflict between the two perspectives. Each tends to ignore the portion of the implementation reality explained by the other. Here, some of the synthesizers of both the perspectives are explained.

Elmore (1982 and 1985) have attempted to combine two perspectives. They argue that policy designers should choose policy instruments based on the incentive structure of target groups. Forward mapping consists of stating precise policy objectives, elaborating detailed means−ends schemes, and specifying explicit outcome criteria by which to judge policy at each stage (Elmore, 1980, p602). Backward mapping consists of stating precisely the behavior to be changed at lowest level, describing a set of operations that can ensure the change, and repeating the procedure upwards by steps until the central level is reached. By using backward mapping, policy designers may find more appropriate tools than those initially chosen. This process ensures consideration of micro implementers’ and target groups’ interpretations of policy problems and possible solutions. However, Elmore’s model has no predictions as to generalize behavior (Elmore, 1980, p604-5). No specific interrelationships are hypothesized; effectively there are no hypotheses to test. As a tool, Elmore’s discussion is useful; as a theory, however, it lacks explanatory power (Matland, 1995, p151).

Richard E. Matland (1995) presents the ‘ambiguity and conflict model’ as a combination of the top-down and bottom-up perspectives. His model suggests that their relative value depends on the degree of ambiguity in goals and means of a policy and the degree of conflict. Four policy implementation paradigms are: low conflict-low ambiguity (administrative implementation), high conflict-low ambiguity (political implementation), high conflict-high ambiguity (symbolic implementation) and low conflict-high ambiguity (experimental implementation) (1995, p145). He spells out how ambiguity and conflict affect policy implementation. His model provides a more theoretically grounded approach to implementation. However, Matland also avoids seeing the level of policy discretion as something explicitly chosen by policymakers, recognizing how it may be a function of policy conflict. The question about his argument is then: How easy is it to label policies in the way he does? (Hill and Hupe, 2006, p77)

Malcolm Goggin et al. (1990) develop a model which is based on the communications theory perspective of intergovernmental implementation, but which also includes many variables from the top-down and bottom-up approaches. The model indicates that implementation in the states is influenced by a combination of inducements and constraints from the federal, state and local level; by a state’s decisional outcomes; and by a state’s capacity to act. How implementation exactly proceeds in specific policy areas is dependent on the interaction of these elements of the model.

Thomas and Grindle (1990) propose an interactive model of implementing policy reform. The process of implementing policy reform is seen as interactive rather than linear (Thomas and Grindle, 1990, p1166). Their framework for policy study looks at 'how reform proposals get on the agenda for government action, what factors influence decision makers and the linkages between agenda setting and decision-making process.’ The central element in the model is that a policy reform initiative may be altered or reversed at any stage in its lifecycle by the pressure and reaction to it. This model views policy reform as a process, one in which interested parties can exert pressure for change at many points. Some interests may be more effective at influencing high-level officials in government, others at affecting the managers of the implementation process or those who control the resources needed for implementation. Understanding the location, strengths and stakes involved in these attempts to promote, alter or reverse policy reform initiatives is central to understanding the outcomes (Thomas and Grindle, 1990). Analysis commences with a look at the characteristics of any public policy in terms of the reaction it will generate. Then, governments must assess what their resources are and how they can mobilize available resources to promote successful implementation. Decision-makers must evaluate political resources while public managers attend to bureaucratic resources. Such analysis can lead to a more realistic approach to policy where the question of implementation feasibility assumes major importance. Failure can be better anticipated, modifications can be better judged, and resources can be more efficiently and effectively allocated (Turner and Humle, 1997, p79).

However, the second generation is not exempt from criticism. Researchers cannot agree on a common definition of the term ‘implementation’. There are vast differences in the role of implementers, especially with respect to the degree to which they are autonomous actors. Furthermore, it has not been able to explain why implementation occurs as it does or predict how implementers are likely to behave in the future. Indeed, some middle range theorizing with substantial potential utility has emerged from this line of work. However, a number of propositions available in research literature have proverbial rather than scientific characteristics (Goggin et al., 1990, p183).

The main critique of the second generation model is again based on their approach: too many case studies, not enough validation and replications (Goggin, 1986). Matland (1995) suggests that some first and second generation models failed to provide a comprehensive synthesis or a unifying approach to implementation analysis (quoted in Schofield, 2001, p250).

3.3 Third Generation Implementation

While both first and second generation implementation researches have added much to our knowledge of what implementation is how and why it varies as it does, it has been much less helpful in differentiating between the types of implementation outcomes, or in specifying the causal patterns that occur, and the relative importance and unique effects of each of the various independent variables that are part of any multivariate analysis of implementation performance (Lester, 1995, p79). However, these researches have not succeeded in sorting out the relative importance of the explanatory variables (Winter, 2003, p216). A substantial part of the studies could be criticized as merely presenting--often long--checklists of variables that might effect implementation. According to Goggin (1986), this problem had hampered the development of implementation theory. He, therefore, suggests a third generation of implementation studies that would test theories on the basis of more comparative case studies and statistical research designs which could increase the number of observations (Quoted in Winter, 2003, p216).

The unique trait of the third generation research is its research design--an explicit theoretical model; operational definitions of concepts; an exhaustive search for reliable indicators of implementation and predictor variables; and the specification of theoretically derived hypotheses, with analysis of data using appropriate qualitative and statistical procedures as well as case studies for testing them (Goggin, et al., 1990, p19).

In the third generation research, the macro world of policymakers with micro world of individual implementers is integrated (McLaughlin, 1987, p177). The macro-level research operates at system level. It stresses regularities of process and organizational structures as stable outlines of the policy process and frames individual actions in terms of position in a relational network. Micro analyses, conversely, operate at individual level. They interpret organizational action as the problematic and often unpredictable outcomes of autonomous actors, motivated by self-interest. Macro-level analyses generally provide insufficient guidance to policymakers or practitioners interested in understanding program outcomes (positive or negative), evaluating alternative, assessing internal work requirements, or developing models of how policies operate in practice. Conversely, micro-level analyses ignore systemic attainments and unanticipated consequences for the institutional setting as a whole, so cannot speak to the expected organizational consequences or system-wide effects of a policy. Micro-level analyses, thus, provide limited guidance to policymakers faced with system-wide decisions. However, some scholars argue that third generation implementation has not been realized in practice.

4. Issues of Implementation Theories

The first generation implementation researchers find out the problems of policy implementation, i.e. uncertain relationship between policies, decisions and implemented programs. Similarly, the second generation implementation studies focus on the 'development of an analytical framework of implementation', which includes the top-down, bottom-up perspectives and their synthesis. Likewise, the third generation implementation research should concentrate on explicit implementation theory-building, which has not yet been realized. Nevertheless, following conceptual, applicability, transfer of knowledge, changing context and methodological issues of implementation studies seem more relevant to be considered in theory development or useful advice for decision-makers on the basis of the literature surveyed above.

4.1 Issue of Conceptual Clarification

The meaning of implementation is loosely developed and lacks adequate specifications of causal mechanisms. One problem is that the concept 'implementation' is often used to characterize both the implementation process and the output--and sometimes also the outcome--of the implementation process (Winter, 2003, p217). Output or outcome targets are harder criteria to measure performance. Is implementation about achieving conformance or performance? Policy-centered approaches to analysis of necessity involve comparing outcomes against a priori statements of intent or targets. Performance is thus judged in terms of achieving conformance with policy targets and standards. In practice, the so-called performance criteria tend to operate more as conformance criteria--often the minimum level or standard deemed to constitute satisfactory performance (Barret, 2004, p255).

4.2 Debate on the Top-down and Bottom-up Implementation Perspectives

The features and criticisms of the top-down and bottom-up perspectives have been discussed above. However, the top-down and bottom-up perspectives raise debate on the purpose of implementation analysis: Are they prescriptive or descriptive or normative? (Barrett, 2004, p255, Saetren, 2005, p572). Do these perspectives show the causal relationship in implementation research? The top-down perspective could be regarded as prescriptive--what ought to happen, whereas the bottom-up focuses on description of the implementation process. Therefore, both perspectives seem as confusing in how normative, methodological and theoretical aspects are seamlessly and indistinguishably intertwined (Saetren, 2005). Saetren suspects that this prolonged debate has caused frustration among many scholars of policy implementation research.

4.3 Theoretical Pluralism in Policy Implementation Research

Since the beginning of policy implementation research, there is no theory of implementation that commands general agreement; researchers continue to work from diverse theoretical perspectives and to employ different variables to make sense of their findings (O'Toole and Montjoy, 1984, quoted in Lester, 1995, p84). Nevertheless, the available literature does seek to identify the common causes and cures for policy failures. As discussed above, implementation literature is heavily dominated by many variables but lack crucial variables.

4.4 Methodological Issues in Implementation Research

In addition, they are criticized on the basis of methodology. In terms of methodology, many of the researches on policy implementation have been dominated by single case studies, allowing the complex phenomena of implementation to be studied in a broad context. In each case, several data sources are often applied, such as reports and documents, qualitative survey with implementers, quantitative data on coverage of program, participation, output in terms of delivery performance and outcomes (Yin, 1982). Some even use qualitative or quantitative methods for detailed text interpretation in case studies. Other scholars have called for the replacement of single case studies with comparative and statistical research designs, which can increase the number of observations and control for third variables in order to allow more systematic theory and hypothesis (Goggin, 1986, quoted in Soren C. Winter, 2000).

4.5 Implementation Research: Time for Revival?

Scholars like Harald Saetren (2005, p574), Jill Schofield and Charlotte Sausman (2004, p236) and Susan M. Barrett (2004, p260) outline the need for a revival of interest in implementation research. First, whether concepts and ideas such as regulation, innovation management and evaluation adequately take place or ignore these newer ideas in implementation research. Second, implementation research may be explained by fashion. Is it unfashionable? Barrett argues that, perhaps implementation studies were most unfashionable in the 1990s, when one contributory factor could have been the advent of the new public management and its associate managerialism. In turn, this leads to the adoption of other disciplinary approaches to the study of policy implementation. Thus, the study and use of techniques, form, strategic management, organizational change and organizational culture served to complement and sometimes also obscure implementation studies (Quoted in Schofield and Sausman, 2004, p236). Third, there is a need for renewed emphasis on multidisciplinary working in policy studies. Multidisciplinary research provides benefits for theory development in synthesizing ideas from a plurality of disciplines addressing similar issues from different perspectives. Finally, as a separate field of study, implementation is still in its infancy (Goggin et al., 1990, p9). Some scholars like Goggin, Palumbo, Linder and Peter express optimistic views on the future of implementation research. Similarly, scholars like Bearman, Salamon and Rothstein express a pessimistic view on implementation research. It has, for example, been alleged that much of the early implementation research was essentially ‘misery research’ (Rothstein, 1998), highlighting disasters and implementation failures and providing a distorted impression of implementation difficulties (Quoted in Hill and Hupe, 2003, p119).

4.6 Geographical Area and Sectoral Policy Issue Coverage

A review of literature reveals a number of models which relate to the implementation of public policy. The majority of these have been developed from a North American perspective and generally embedded in a discussion of the policy process, emphasizing that implementation is inextricably linked to the political process of governing (Schofield, 2004, p284). In other words, these researches suffer from a bias towards ‘westernized’ ideas and points of view. Many of the conditions and situations that can be found in a developing country cannot be analyzed through these models. The literature comes basically from a western culture with assumptions that liberal democracy fits all conditions and those materialistic notions of human needs are the most important factors to be taken into account. Market mechanisms are overrated and the critical role of the state and societal culture is underrated (Dror, 1992, p276, 279). Harald Saetren (2005, p571) makes clear the picture of the overall regional bias of implementation research. Western hemisphere accounts for close to 90 per cent of all publications on or about implementation research.

Similarly, implementation scholars consistently pay great attention to certain policy sectors such as education, health, environment, social and economic policy issues and so on (Hill and Hupe, 2002, p201-204 and Saetren, 2005, p570). Educational policy issues occupy nearly 40 per cent of total publications, whereas studies of foreign policy are conspicuously rare--1-2 per cent of all publications.

4.7 Issue of Legitimization of Policy Implementation in Developing Country

Translation of policy into practice in developing countries is a challenging and legitimate concern (Saetren, 2005, p573). The policy implementation process in developing countries shares a great deal with the process in more developed countries (Lazin, 1999, p151). However, the effects of poverty, political uncertainty, people’s participation as well as the unique character of each developing country cannot be ignored in the policy implementation process.

‘Poverty is a state of economic, social and psychological deprivation occurring among people or countries lacking sufficient ownership, control or access to resources to maintain minimal acceptable standards of living’ (UNDP, 2002, p10). Poverty has a direct influence on the policy implementation process. The intended results cannot be achieved due to poverty in a developing country. Jan Erik Lane (1999), in an article entitled ‘Policy Implementation in Poor Countries’ argues that the problems connected with policy implementation in developing or Third World countries are intertwined with basic economic and political conditions. He contends that political stability and economic development are closely interrelated. On the one hand, low level of economic development leads to political instability and, on the other hand, political instability worsens poverty. Effective policy implementation improves poverty situation in Third World countries, which need both economic development and political stability. He is optimistic about the possibility of closing the gap between the rich and poor countries, provided strong and stable regimes utilize available economic resources to foster economic growth and development.

Political uncertainty is an endemic condition to policymaking and implementation. Political uncertainty refers to military threats, domestic violence, political regime change and so on. Uncertainty is likely to be more pronounced in developing than in developed countries due to severely limited resources, extensive demands for public services and investment, weak political institutions and limited capacities for policymaking and program implementation amidst all other difficult conditions (Caiden and Wildavsky, 1974, quoted in Nagel and Lazin, 1999, p37-38). In a paper entitled ‘Policymaking and Implementation in the Context of Extreme Uncertainty: South Africa and Israel’, SX Hanekom and Ira Sharkansky (1999) confirm the relationship between political uncertainty and policy implementation. Their paper illustrates uncertainties in each country by reference to past and recent events, and links these uncertainties to the country’s political, policymaking and policy implementation traits. Extreme uncertainty is likely to affect the quality of policymaking and program implementation in both types of countries, but in different ways that show the influence of each country’s own traits. Finally, they recommend ways for other governments of developing countries to cope with uncertainties.

Participation in policy processes is not so pronounced, and the channels for participation are less well established in developing countries. At the same time, the state structures, whatever their weaknesses, are relatively powerful vis-à-vis their societies. But, the interface between state and society is constantly changing. Of all the causes of poor policy evaluation in developing countries, the most serious institutional flaws are in political systems. Furthermore, a common assumption is that implementers are involved at every stage of the policy-making process, and that they are often the most powerful groups in setting the policy agenda. In many developing countries, participation of lower level in the selection of sets of options is rare, and the choices are made by central-level policymakers. Very often, the problems the Third World bureaucracies have to deal with are more difficult to solve than those in developed countries, compounded by limited resources for implementation (Jain, 1992, p24, Moharir, 1992, p257).

Socio-administrative culture differs from country to country. Culture incorporates social values, beliefs, norms and practices. It is defined as the collective programming of the mind, which is developed in the family in early childhood and reinforced in school and organizations; these mental programs contain a component of national programs (Hofstede and Hofstede, 2005, p4). They are expressed in different values that predominate among peoples from different countries. Hofstede classifies four dimensions: power distance, uncertainty avoidance, individualism versus collectivism, and masculinity versus femininity. It can be argued that, along these dimensions, dominant value systems in different societies can be ordered, which affect policy implementation processes in predictable ways. A study of administrative culture in Bangladesh reveals that the bureaucracy in Bangladesh is characterized somewhat more by traditional than by modern norms (Jamil, 2002, p121-122). It is characterized by relatively high degree of power distance between authority and common citizens, low tolerance for ambiguities, dependence on traditional sources of information, low tolerance of bureaucrats than egalitarian, more positive towards NGOs, and
preference for employees with traditional qualities.

Other factors like people’s trust of public institutions, public awareness, accessibility and availability of services, and so on should be taken into consideration for the sake of successful policy implementation in a developing country. Besides, interdependence between developed and
developing countries arising from globalization is growing up. It has direct impact on the design and implementation of policy of the country. Usually, aid conditionality as per the interest of donor country determines whether a policy is translated
into practice or not in a developing country.

4.8 Share of Implementation Research Outcomes

The adoption and application of implementation research findings from western countries for further improvement of policy implementation is a challenging concern in developing countries. There are not enough autonomous associations and institutions at work to obtain information on the impact of policies or to communicate this information to public officials. Policy evaluation tools like ‘think tanks’, university research facilities and investigative journalism are virtually unknown in developing societies (Jain, 1992, p24; Moharir, 1992, p257). The more conventional feedback mechanisms, such as political parties and interest groups, are fragmented, or not trusted by their constituents, or controlled or ignored by the government.

5. Conclusion

While reviewing the implementation literature, the concrete theory of policy implementation is still lacking. In course of theory-building efforts, the first generation implementation researches were more concerned about how a single authoritative decision was carried out, such as through case studies. This body of research was primarily directed towards describing the numerous barriers to effective policy implementation (Linder and Peters, 1987, quoted in Lester, 1995, p73). Similarly, the second generation researches were concerned with framework development, and these frameworks explained implementation success or failure (Lester, 1995, p73). However, neither of the implementation research generations succeeded in sorting out the relative importance of explanatory variables. Similarly, the third generation is going on, but it has not been realized yet.

The literature surveyed above is dominated by the top-down and bottom-up perspectives and their synthesis. The basic arguments of these perspectives concern methodologies and accountability. The ‘top-downers’ call for eliminating the ‘gap’ between formulation and output, whereas the ‘bottom-uppers’ emphasize the inevitable, and perhaps desirable, participation of other actors in later stages of policy process. Synthesizers have tried to get variables from both top-down and bottom-up perspectives. There are still questions about the methodology adopted for implementation research. Earlier researches were dominated by single case studies. There were many variables used in a single case study. More implementation researches are found normative or ideological based on case studies rather than pragmatism.

Implementation is often not value-free. It is laden with social values, norms and practices. These are location-specific or country-specific. Therefore, the applicability of the implementation theory is questionable and challenging. However, the policy implementation process in developing countries shares a great deal with the process in more developed countries.

Issues concerned with the conceptual clarity of policy implementation, theoretical debate over the top-down or bottom-up perspectives, applicability or transfer of research output from one region to another, methodology employed in implementation research and so on are seen as arguable. These need to be addressed in the forthcoming implementation studies. In addition, such studies should consider the implementation context as well. It is because these theories were developed in the Western context where more stabilized and democratic political regimes are functional. Social setting is either homogenous or fixed. But, quite a different situation often characterized by uncertain, unpredictable and ever-changing political situations diversified social settings and weak economic condition prevails in developing countries.


Amanya, Rebecca (2006), “Policy Implementation in Ghana- The Case of Free Compulsory Basic Education (FCUBE) programme”, MPA thesis submitted to University of Bergen, http://ugle.svf.uib.no/admorg/?strId=5404&kategori=867&versjon=True

Beniwal, V.S. (2005), "Challenges and Prospects of Implementing Citizens Charter: A Study of Panchkula (Haryana) Municipal Council in India", Thesis of Mphil, University of Bergen

Berman, Paul (1978), “The Study of Macro-and Micro- Implementation” Public Policy 26(2),157-184pp.

Buabeng, Thomas (2005), “Implementing Public policy at the Local Level in Ghana: The Experience of Akwapen South District Poverty Reduction”, Dissertation of Mphil, University of Bergen.

Clarke, Michael(1992). ”Implementation”, in Harrop, ed., Power and Policy in Liberal Democracies. Cambridge: Cambridge University Press.

Dror, Yehezkel (1992), Advanced policy planning for the Third Word, eds Asmerom, Hoppe, Jain, Bureaucracy and Developmental Policies, Amsterdam: VU University Press

Elmore, Richard E. (1978), Organizational Models of Social Program Implementation, Public Policy, Vol.26, No. 2: 185-228pp)

Elmore, Richard F. (1979), Backward Mapping: Implementation Research and Policy Decisions, Political Science Quarterly, Volume 94, No. 4, 601-616pp.

Giacchino S. and Kkabadse A. (2003), Successful Policy Implementation: the Route to Building Self-Confident Government, International Review of Administrative Sciences 69(2), 139-160pp.

Goggin, ML (1986) The ‘Too Few Cases/Too Many Variables’ Problem in Implementation Research. Western Political Quarterly, 328-347pp.

Goggin, M.L., Ann O’M Bowman, James P. Lester and L.J. O’Toole Jr. (1990), Implementation Theory and Practice: Toward a Third Generation, USA:Harper Collins Publishers

Googin, et al. (1990), Studying the Dynamics of Public Policy Implementation : A Third Generation Approach, in Implementation and the Policy Process: Opening up the Black Box, edited by Dennis J Palumbo and Donald J.Calista, New York: Greenwood Press, 181-197pp

Hanekom, S.X. and Sharkansky, Ira (1999), “Policymaking and Implementation in the Context of Extreme Uncertainty: South Africa and Israel” ” in The Policy Implementation Process in Developing Nations, ed. Stuart S. Nagel, Volume 6., Stamford: Jai Press Inc. 37-54 pp

Hill, Michael, Hupe, Peter (2006), Implementing public policy, Sage Publication: London

Hoftede, G. and Hoftede, G.J. (2005), Culture and organizations: Software of mind (2nd ed.), New York: MacGraw-Hill.

Ho, Leong Khai (1999), Politics and Institutions in Economic Policy in Implementation in Malaysia, in the Policy Implementation Process in Developing Nations, ed. Stauart S. Nagel, Volume 6., Stamford: Jai Press Inc. 57-73pp

Jain, R.B. (1999), “Implementing Public Policy in India” in The Policy Implementation Process in Developing Nations, ed. Stuart S. Nagel, Volume 6., Stamford: Jai Press Inc. 23-35 pp

Jamil, Ishtiaq (2002), Administrative Culture in Bangladesh: Tensions between Tradition and Modernity, International Review of Sociology- Revve International de Sociologie Vol 12, No 1.

Lane, Jan-Erik (1999), “Policy Implementation in Poor Countries”, in The Policy Implementation Process in Developing Nations, ed. Stuart S. Nagel, Volume 6., Stamford: Jai Press Inc. 11-22pp.

Laizen, Fred (1999), “Methodological and Substantive Issues in the Policy Implementation Process” in the Policy Implementation Process in Developing Nations, ed. Stauart S. Nagel, Volume 6., Stamford: Jai Press Inc. 149-162pp

Lester P. James, Ann O’M Bowman, Malcolm L Goggin, and Laurence J. O’Toole (1995), Public Policy Implementation: Evolution of the Field and Agenda for Future Research, Research in Public Policy Analysis and Management, Volume 7, 71-94pp.

Matland, Richard E. (1995), Synthesizing the Implementation Literature: The Ambiguity-Conflict Model of Policy Implementation, Journal of Public Administration Research and Theory: J-PART, Vol.5, No.2 (Apr. 1995) 145-174 pp.

May, Peter J. (2003), “Policy Design and Implementation” In Handbook of Public Administration, ed. BG Peter and J Pierre. London, Thousand Oaks, CA and New Delhi: Sage 223-233 pp.

Mazmanian, D.A. and Sabatier, P.A. (1983), Implementation and Public Policy. Glenview, III.: Scott, Foresman

McLaughlin, Milbrey Wallin (1987), Learning From Experience: Lessons From Policy Implementation”, Educational Evaluation and Policy Analysis Vol.9, No. 2 171-178 pp.

O’Toole, L.J.R. Jr. (1995), “Rational Choice and Policy Implementation”, American Review of Public Administration, 25(1) 43-57pp.

Oswald, Rosemary Mruma (2005), Implementation of the National Environment Policy : A Case of the Local Government Authorities in Dar es Salaam City, Tanzania,”, MPA thesis submitted to University of Bergen.

Palumbo, Dennis J. & Donald, J. (1990), Implementation and the Policy Process, opening the Black Box, Greenwood Press: New York, Wetport

Parson, W. (1995), Public Policy. Aldershot:Edward Elgar.

Pressman, Jeffrey L. and Wildavsky, Aaron (1984), Implementation: 3r edn. Bekeley: University of California Press.

Priyantha, Ilandarige R. (2006), Challenges of Implementing New Public Management Initiatives in Developing Countries: A Case Study of Sri Lanka, Thesis of Mphil, University of Bergen.

Sætren, Harald (2005), Facts and Myths about research on Public Policy Implementation: Out of Fashion, Allegedly Dead, But Still Very Much Alive and relevant, Policy Studies Journal, Vol. 33, No. 4, 559-582pp.

Schofield, Jill (2001), Time for a revival? Public policy implementation : a review of the literature and an agenda for future research, International Journal of Management Review, volume 3, Issue 3, 245-263 pp.

Thomas, John W. & Grindle, Merille S. (1990), After the Decision: Implementing Policy Reforms in Developing Countries, World Development. Vol. 18, No. 8 1163-1181 pp.

Turner, Mark and Hulme, David (1997), Governance, Administration and Development: Making the State Work, Londeon: Macmillan Press Ltd.

UNDP (2002), Nepal Human Development Report 2001, Poverty Reduction and Governance, Kathmandu: United Nations Development Program

Van Meter, Donald S. & Van Horn, Cark E.(1975), The Policy Implementation Process: A conceptual Framework, Administration & Society, Vol.6 No4, February 1975, 445-487 pp.

Winter, Soren C. (1990), Integrating Implementation Research, eds. Palumbo, Dennis J, & Calista, Donald J, 1990, Implementation and the Policy Process, Opening the Black Box, Greenwood Press

Winter, Soren C. (2003), “Implementation Perspective: Status and Reconsideration”, (Chapter 16) In Handbook of Public Administration, ed. BG Peter and J Pierre. London, Thousand Oaks, CA and New Delhi: Sage 212-222 pp.

Winter, Soren C. (2003), “Implementation”, (Section 5) In Handbook of Public Administration, (ed.) BG Peter and J Pierre. London, Thousand Oaks, CA and New Delhi: Sage 206-211 pp.

Younis, Talib & Davidson, (eds) (1990), Implementation in public Policy, Dartmouth Publishing Company Ltd.