SciELO - Scientific Electronic Library Online

vol.1 issue3Primary health care in the 21st centuryIntegrated management of childhood illness (IMCI): an innovative vision for child health care author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Revista Brasileira de Saúde Materno Infantil

Print version ISSN 1519-3829

Rev. Bras. Saude Mater. Infant. vol.1 no.3 Recife Sept./Dec. 2001 



Developing an operational framework for health policy analysis


Desenvolvendo uma estrutura operacional para análise de políticas de saúde



José Luiz C. de Araújo Jr.I; Romulo Maciel FilhoII

INúcleo de Estudos em Saúde Coletiva. Centro de Pesquisas Aggeu Magalhães, Fundação Oswaldo Cruz, Ministério da Saúde.Av. Moraes Rego s/n, Campus da Universidade Federal de Pernambuco. Recife, Pernambuco, Brasil. CEP: 50.670-420. E-mail: Departamento de Medicina Social. Faculdade de Ciências Médicas, da Universidade de Pernambuco
IINúcleo de Estudos em Saúde Coletiva. Centro de Pesquisas Aggeu Magalhães, Fundação Oswaldo Cruz, Ministério da Saúde



ABSTRACT: This article develops an original framework supporting health policy analysis that may be carried out by distinct research on various situations. The Walt and Gilson model for health policy analysis and its categories: Context, Content, Actors and Process, was taken as the basic framework of analysis. However, to be applied in practice that model needs to be unravelled. Its four major categories need to be operationalised and transformed into a matrix, with each one revealling the elements or sub-categories, and the necessary procedures to conduct a systematic analysis on the Context, Content, Actors and Process of a given health policy. Such an initiative was treated in this work.

Key words: Health policy, Models, theoretical

RESUMO: Este artigo desenvolve uma estrutura original para análise de políticas de saúde que pode ser aplicada por distintas pesquisas em situações diversas. O modelo de Walt e Gilson, e suas categorias: Contexto, Conteúdo, Atores e Processo, foi assumido como a estrutura básica. Contudo, para ser aplicável na prática, esse modelo necessita ser processado. Suas quatro grandes categorias precisam ser operacionais e transformadas em matriz, com cada uma revelando os elementos, sub-categorias e procedimentos necessários para a condução de uma análise sistemática do Contexto, Conteúdo, Atores e Processo de uma dada política de saúde. Tal iniciativa foi elaborada nesse trabalho.

Palavras-chave: Política de saúde, Modelos teóricos




Arguments pointing out the need for studies in the field of health policy analysis are increasingly present in the current specialised literature. Different authors from various points of view have often demanded attention to this subject. Frenk,1 for instance, concludes that "... under the current wave of international interest in health system reform, it is necessary to establish a mechanism for shared learning at the global level. Only in this way will it be possible to reproduce the analytical skills and accumulate the body of evidence that health systems require for their sustained improvement" (Frenk; 1995: 257).1 Collins et al.2 are also concerned with the international development of Health Policy Analysis and implementation. In focusing on the management of the health sector and its policy they state: "Priority should be given to the need for appropriate policy learning between countries, the innovative generation of alternative systems, and the international exchange of basic information and research. Only in this way can informed, knowledge-based policies and decisions occur which, in turn, will result in health sector reforms based on actual evidence of what works, and not on fashion or bold assertion" (Collins et al.; 1994: 54-5).2

Health Policy Analysis (HPA) is naturally concerned with policy-making processes. According to Walt (1994: 35)3 "They are modes of analysis which focus on decision-making processes but some take a macro-view and others a micro-view." Macro theories focus broadly on the whole policy process, concentrating on major issues like the flows of power, interaction among actors and their effects on the context. Micro theories are more concentrated in some stages of the policy process and in techinical and administrative aspects of policy formulation and implementation.3

This review has identified two main types of policy analysis. The first may be called Partial Analysis, due to its concentration on a stage or particular issue, and the second, Comprehensive Analysis, due to its wide angle focus.

In Partial Analysis are classified all types of HPA which aim to concentrate on a particular stage or issue of a chosen policy. The model presented by Parsons4 might be taken to illustrate this case. He divides the policy analysis process into four chapters.

Meta analysis. According to Parsons (1995:1),4 this is like analysing analysis "When we are engaged in meta analysis we are considering the methods and approaches used in the study of public policy and the discourse and language which it employs."

Meso analysis. This type of analysis is "... concerned with how problems are formed and framed, and how they become - or not become - items or issues on the policy agenda." It is: "The analysis of problem definition, agenda-setting and the formation of policy" (Parsons; 1995: 85).4

Decision analysis. This type focuses attention on the decision-making process, which is understood as something in between policy-formation and implementation.4

Delivery analysis. This is the analysis of policy implementation, evaluation and its outputs and outcomes.4

By Comprehensive analysis is meant those kinds which aim at analysing the whole process of a chosen policy. Three examples have been chosen to illustrate this case.

The Reich5 health policy reform scheme.

In this example every health policy reform is above all political ones. It proposes that to succeed they have to be analysed by an effective method to assess their relevant political conditions. Reich5 develops the political mapping technique, which he argues is a very effective one.

The Frenk framework for policy analysis of health system reform.1

After describing the comprehensive character of this approach, he suggests the application of a combination of newly developed analytical tools (such as political mapping, measurement of the burden of disease, national health accounts, consumers surveys, cost-effectiveness analysis of essential interventions, assessment of system performance) to explore a health policy. He proposes five headings that shape the analysis. The five P's are: the problems; the principles; the purposes; the proposals; the protagonists.

The Walt and Gilson6 analytical model.

This model also calls for comprehensive HPA, especially in developing countries where it seems to have been neglected, despite the poor health sector performance and the failure of many attempts at reform. The model recommends analytical attention to four inter-related aspects, which comprise the entire policy. These are: the actors; the content; the context; the process.

The key to this model is not to accept these categories as independent, but also look at their inter-relations and mutual influences.

This review has considered that the later framework seems to be wider and incorporates the others. It ensures a political analysis of policy, as suggested by Reich,5 it also takes into account all the points raised by Frenk,1 in a clear and parsimonious way. Thus, this will be the approach taken to be developed as an operational framework to be promptly applied in various cases of health policy analysis.

In order to make it operational, the issue of context within the health policy analysis is explored in terms of its importance and relation to the policy. The method of carrying out a detailed analysis on the context is also developed and explained by defining what to look in at the sub-categories created by this work: Macro-context and Micro-context: in its political, economic and social spheres of the Macro-context; and sectorial politics, finances health services'structure and outcomes in the Micro-context.

The content of a policy is delimited in terms of the problems it aims to change, its programmes, projects, actions, targets and resources required. The relevance of analysing the content is expressed in terms of identifying the policy's perspective, possibilities of impact and assessing the adequacy of its measures and resources.

This article also defines how to analyse the actors involved in the policy process by applying a model which allows for: the identification of major actors; their position in relation to the policy; how they are mobilised; the correlation of power among them and their possible alliances and coalitions.

The policy process is analysed by focusing on the decision-making and policy implementation. The way decisions are made may be explained through pluralist or elitist views, and whether they are made by open or closed processes in terms of involving the main interested groups. The type of policy is also assessed and related to the decision-making process. There is a description of how policies may be implemented, whether in a more top-down or more bottom-up manner. Relations amongst distinct levels of governments, strategies for policy implementation and resources implications are recommended to be examined as well. Finally, the rationale in which the policy is conceived; whether it is compatible with a rational, incremental, mixed or strategic model is also taken into account.


The issue of Context within the Health Policy Analysis 

As indicated before, this analysis will be carried out by applying the model proposed by Walt and Gilson,6 in which special attention is confered to the Context where a policy takes place, to the Content of the policy proposals, to the Process of formulating and implementing the policy and to the Social Actors playing a part in this process. This approach is schematically shown in the Box 1 by taking the Brazilian Health Sector Reform (HSR) initiatives during the 90s as an example of policy.



Why it is necessary to analyse the Context surrounding a Health Policy

Policies are formulated in a multifaceted reality. A variety of different aspects of such a reality interacting in a particular context certainly influences the whole policy process, as recognised by Collins et al. (1999: 70)7 "Policy formulation and implementation take place in a context which gives explanatory and historical meaning to that policy." In other words, a particular policy formulated in a particular context is certainly meant to interact with its context and to produce some effect on it. Consequently, exploring and understanding the context of a particular policy would be useful, first to understand the possible intricate features of such a policy, but also to enable an assessment of the policy in terms of its coherence; whether it has a good chance or not to impact positively in the context; and its appropriateness, whether the policy is really necessary and sufficient in itself to produce the desired effect.

The fundamental importance of context analysis when carrying out a Health Policy Analysis is widely recognised. When the literature is examined, it appears like as consensus. Although it may vary in terminology and approach, the category Context seems to be a key element in everything cited below.

In presenting his "Comprehensive policy analysis for health system reform", after an analysis of the Mexican experience, Frenk (1995: 262)1 points that "There are five fundamental factors that must be analysed in order to decide the form that reform will take: the problem; the principles and the purposes; the proposals and the protagonists." When he further details each of these factors, it becomes evident that the factor expressed as "the problems" is all about the context analysis. There he claims that analysis should focus on: demographic, epidemiological, educational, technological, cultural, political and economic changes. These are clear feature of the so-called Context.

Barker,8 explaining health care policy analysis, emphasises the issue of power as a decisive factor to provide the means for an understanding of a policy: "I would argue that resources, ideas and technology are all important, but that the way in which all are used depends upon the distribution of power in society" (Barker; 1996: 79).8 Therefore, a clear understanding of the concept "power" is necessary in order to understand how the distribution of power in a particular reality works. This is fundamental for policy analysis and this is a part of the context analysis, as taken by this study.

Reich5 also concentrates his approach on the political component of health policy analysis. He justifies this by saying: "Politics affect the origins, the formulation, and the implementation of public policy, especially when significant changes are involved" (Reich; 1995: 48).5 The political setting is precisely part of the larger Context.

Walt and Gilson (1994: 354)6 point out that often policy analysis "...wrongly focuses attention on the content" of policy, neglecting the main issues which may explain whether a particular policy failed in producing the desired effect. They argue in favour of a comprehensive analysis in which the Context of a policy, along with the Content, the Process and the Social actors, is very much taken into account, especially to understand the extremely political character of the policies.

Collins et al.,7 also recognise the significance of Context analysis for a clear understanding of the policy process. "Health Sector Reform (HSR) policies respond in a complex fashion to issues in the social, political and economic environment. Failure to understand it can lead to misconceptions over the role, appropriateness, policy process and political feasibility of HSR" (Collins et al.;1999: 70).7

How to carry out the Context Analysis of a Health Policy

Within of this article, the category Context is to be understood as a synthetic one. It brings together in one word different aspects of a given reality. As discussed below, different types of recommended context analysis often focus on some of these aspects.

Frenk,1 for example, calls for attention to what he denominates: demographic, epidemiological, educational, technological, cultural, political and economic changes of a particular reality. Although it may seem a wide overview already, one may add to these some more aspects to be looked at Walt and Gilson,6 for instance, when focusing on context, consider other issues such as the role of the state in the policy process and the relations between the state and the market. More recently, Collins et al.,7 building upon previous formulations, included more aspects to be selected in a context analysis. They mention six "categories of factors" to be identified as the "... context of contemporary health sector reform: demographic and epidemiological change; processes of social and economical change; economic and financial policy; politics and the political regime; public policy and the public sector; external factors" (Collins et al.; 1999: 74).7

In this work, sub-categories of the main Context will also be chosen in an attempt to cover most aspects. An endeavour will be made to gather all these above-mentioned aspects of context included in the framework created in the present stuy (Figure 1).



Although the health sector is the focus of this work, it will be analysed as part of a broader context, which means the whole of society and its major spheres of activity. Walt9 acknowledges that there are macro-contextual factors that go beyond the health sector, the strength of the state, for instance, and she draws attention to its importance: "But health policy analysts ignore them at their peril. If policies are to be implemented to achieve effective outcomes then they need to be taken into account." (Walt; 1996: 231).9

In order to analyse these fundamental aspects of the macro-context and the health sector context as well, within the scope of this work, the category Context will be schematically divided into a Macro-context, meaning society as a whole, and a Micro-context, focusing on the health sector. This study drew from the idea guiding the concepts of micro and macro theory and analysis, in order to originally develop the category of Context into a Macro-context and a Micro-context', as discussed below.

According to May (1997: 39):10 "Micro theory is more concerned with understanding face-to-face interactions between people in everyday life whereas Macro theory is concerned with the behaviour of collections of people and the analysis of social systems or structure." The same idea also supports the concepts of "macroeconomic" and "microeconomic theory", in which, macroeconomics, as coined by Ragnar Frisch in 1933, is used to the analysis of the relationships amongst broad economic aggregates such as Growth National Product (GNP), balance of trade, inflation, unemployment, national budget. Whilst microeconomics is applied to the study of small parts of national economy, as the performance, for instance, certain industries or households.11

Barker8 already recalled the use of the terms macro and micro in the study of policy. In that case macro policies involves global interests within society, the long-term objectives of the government and it may affect everyone, and micro policies involves the more localised or sectorial interests.

The idea of macro and micro is also identified as a tradition within applied ethnography. As Denzin and Lincoln12 explain, the applied ethnography is used to explore contexts of decision-making and directed towards the interest of some client. And in that case, micro and macro analysis examine the relations between local and general contexts with a view to generalise to larger, more macro realities.

Each of the sub-categories created, Macro and Micro contexts, will focus on the political, economic and social spheres of activity, as dimensions which encompass all those main accepted reasons for HSR: economic, political, and ideological motivations, as well as demographic, epidemiological, educational, technological and cultural changes, and also external factors.

The Macro-Context

 The Macro-context is here arbitrarily organized into the political, economic and social spheres, in order to facilitate the identification of the most aspects present in this dimension of the context, as shown in Figure 1.

The political sphere is here considered as the space in which the flow of power occurs amongst distinct groups in society. Who is ruling, who is in support, who is in opposition; all viewed in a dynamic way so as to observe who is benefiting and who is losing throughout the policy process. Political power, to the scope of this work, will be accepted as a synthetic notion: "... its application produces results" (Hardin; 1995: 708-9).13 However, recogni-sing the academic debate around the concept of power, and the importance of such concept, as Hay (1997: 45)14 expresses: "Power is probably the most universal and fundamental concept of political analysis." This study will discuss below how it develops its understanding of power.

As Philp15 has pointed out, there is no unified precise definition of power, and there are three main sources of disagreement explaining that. Some disciplines emphasise different bases of power; wealth, knowledge, force, for example; others may focus on the different types of power; influence, command or control, for instance; another may concentrate on the different uses of power; economic and political ends, individual or collective means. Moreover, Philp15 explains that definitions of power are rather theory-dependent and that certainly implies in certain amount of interpretation and interest.

According to Lukes,16 there are two major views about who or what possesses power. There are those who see power as a property of agents, either as individuals or groups. And on the other side, those who see power as an impersonal asset, as the capacity of the social system to achieve generally required objectives, or as social mechanisms set to regulate persons.

This study analyses power within the policy process which is strongly dominated by decision-making events, involving social relationship among various actors. Therefore, this study focuses on the power of social agents, including institutions, as used within the decision-making processes.

In terms of power related to decision-making, Lukes16 discusses and points out limitations on two earlier leading understandings of power, and then he expresses a complementary view which argues for a three-dimensional "radical view" of power.

The first view of power that Lukes16 discusses is the one which achieved great influence among American political scientists in the sixties through the work of Robert Dahl, and which has historical roots in the thought of Max Weber. As believe by Dahl, power was clearly expressed, and empirically observable in the decision-making process. He understood power simply as: "A" has power over "B" to the extent that he/she can get "B" to do something that "B" would not do otherwise, the power is especially observed where there is a conflict of interests between "A" and "B".16 Despite the great appeal that such an approach had in terms of showing power as something visible and catalogued, it was strongly criticised in the early sixties by Peter Bachrach and Morton Baratz.16

According to Lukes,16 Bachrach and Baratz did not deny the power relation identified by Dahl. Their criticism is on the restrictive character of Dahl's view. They argued that power has two faces, the first face is that already considered. However, Dahl's narrow approach concentrates exclusively in the decision-making and misses the other face of power, the one which is exercised via "non decision making." In fact, power may be, and often is exercised by limiting the range of decision-making to secondary and relatively innocuous issues. In that case, "A" make efforts to mobilise predominant values, social rules or institutional procedures intending to limit to public consideration only those issues secondary and safe to "A". Lukes16 considered this two-dimensional view of power, "decision-making and non decision-making" a major advance over the first one-dimensional view. However he considered that a "qualified critique", since this type of agenda-setting, through non decision-making, is still "... a form of decision-making" (Lukes; 1974: 20).16

Finally, Lukes16 calls for a comprehensive, or radical, three-dimensional view of power. Acknowledging those two views considered before, he says that a view of power to be complete must go beyond the formal decision-making process and cover also the situations where there is consensus. Situations where instead the conduct is set as in the one-dimensional view, or the agenda is set, as in the two-dimensional view, but where the preferences are shaped, must also consider the exercise of power. This preference-shaping view of power is made clear in Lukes'question (1974: 24)16 "Second, and more important, is not the most insidious exercise of power to prevent people, to whatever degree, from having grievances by shaping their perceptions, cognitions, and preferences in such a way that they accept their role in the existing order of things, either because they can see or imagine no alternative to it, or because they see it as natural or unchangeable, or because they value it as divinely ordained and beneficial?" As warned by Philp,15 Lukes'radical view of power also raised disagreements.

Hay14 considers Lukes'three-dimensional approach problematic. He notes that within Lukes'approach, one must make the difference between the own genuine interest and the fake induced or shaped ones. 'Who defines that? Hay (1997: 47)14 notes that: "The analyst thus becomes the supreme arbiter of the genuine interest of the victim." Hay also accuses Lukes of "resurrecting the spectre of the false consciousness" into contemporary theory. Moreover, Hay challenges the possibility of such a view be empirically demonstrated.14

In the scope of this work, all the three-dimensional view of power, the "conduct-setting", the "agenda-setting" and the "preference-shaping" are used in the analysis of power within the "decision-making" processes. These views of power are suitable and updated to support the type of context analysis recommended in this study. Furthermore, Hay's criticisms towards Lukes'approach do not challenge the two other dimensions of power, neither does he disprove the plausibility of Lukes's assumption.14 Even agreeing with the difficulties to empirically demonstrate Lukes's view, this is not sufficient to contradict the existence of such a dimension of power.

In the political sphere of society, issues like: political changes, political regime, ideology, public policies and international influence, shall be focused on with a view to analyse their relations with the given Health Policy process (Box 2).

Throughout this work, the Economic sphere of society will be considered as the production, distribution and/or concentration of income, goods and services within society. In this sphere the economic reasons determining a health policy are identified. In examining this, it will analyse the government's economic policy, focusing on general indicators of its performance, such inflation rate, GNP and GNP per capita, level of growth and stability, position in the international ranking of economies, amount of external and internal debt, balance of international trade, evolution of direct foreign investments and the evolution of national reserves in US$ dollars. These will allow a critical examination linking economic constraints to the health policy process.

By Social sphere is meant the space in which social facts, whether achievements or problems, take place. Many of these facts, like population growth and distribution, level of income and goods distribution, prevalence of poverty and general conditions of life such as housing, employment, education, are closely related to the health conditions in all societies. It is in this sphere where these phenomena will be identified and examined in their possible relations with the health policy process.

Particular attention shall be paid to demographic changes, to people's quality of life and level of human development, to the prevalence of poverty, and to the distribution of income and consumption among the population (Box 2).

The Micro-context

The Micro-context is the health sector itself. It must also be analysed by paying attention to its sectorial politics and policy. This shall focus on the political aspects of the health sector's recent history, its major structure and its guiding principles. It will look at the public private relations within the sector. Special attention must also be given to the political conflicts at the top of the system and their effects on the policy process.

The sectorial finances analysis shall be carried out by looking at the resources available to the sector, its funding mechanisms and its expenditures. The sources of funds must be identified and discussed, the yearly budgets and the main expenditures shall be examined in relation to the policy's priorities officially expressed. The financial performance of the private sector and its relations with the public sector must be looked at through the available data, attempting to find out its role and the implications for health policy process.

The health problems and the health services are described and analysed through the main health problems shown by the health indicators and the main components of the health system and their outcomes, viewed through the health service indicators. These should be discussed in order to understand the appropriateness and the need for a given Health Policy and the kind of organization, performance and further arrangements of the health services it may suggest.

Possibilities and limits of Context Analysis

As Context is such a synthetic category, it may give the impression that by performing a Context Analysis of a policy the whole policy is already analysed, which is not the case. Therefore, it is important to bear in mind what is and what may be not covered by the Context analysis itself of a health policy.

The content of a policy, for instance, deserves its own analysis. Although aspects of the content may have been discussed in the Context Analysis, it certainly preserves its own identity. It is indeed formulated under the intense influence of its context but it is something else which comes from inside the context, takes on a specific format, then returns added to the context, and may possibly change that as well. Because of this perceived separation, it is worth doing the individual Content Analysis of a policy in order to capture most the greatest part of such a policy.

The social actors are certainly touched by the Context Analysis, particularly the more important ones. However, as they are the subjects of the whole matters, they are also entitled to an specific analysis. The content and most of the context of a policy depends on the social actors'actions and interaction. Being such an important feature of a policy, it needs an individual analysis, in support of a more complete understanding of the entire situation.

Taking the process of a policy as the entire span and movements of a policy, from its inputs to its formulation, implementation and outcomes, the process becomes then the dynamic narrative of the policy's life story. The Process Analysis may allow an understanding of the entire trajectory of the policy, from its birth, paying attention to the resources - material, know-how and power resources - used to actually make it, as well as how it was formulated and implemented, and to what impact it made on its surroun-dings.

Walt and Gilson6 warns that a schematic model of analysis, which naturally is much simpler than the real life, may give the impression that these categories should be considered as essentially separate from each other. This would be a big mistake. The effort just made above, in order to consider the specific points of each category, is in fact an intellectual exercise. It was built as a tool to facilitate a deep and comprehensive analysis of a health policy in which the policy process is understood as a whole and it is multi-dimensionally analysed through the categories above, as if context, content, actors and the process were merely different angles from which to view a previously shaded surface of the same object.

In fact, the actors are highly influenced by the context in which they act nevertheless they also affect the context by their actions. Yet the content and the process of a health policy are drastically influenced by the context and by the social actors as well. In practice, these categories are linked components of a complete phenomenon. And the Context analysis should be carried out with an awareness of its limitations and of its partial character. The health policy as a whole, to be fully appreciated, has to be viewed from others angles, which are the other three categories of analysis.

The next topic will discuss the Content analysis of a given health policy.


What is the Content and why analyse the content of a health policy?

The content is the body of a policy and it is expressed through all of its components, such as: programmes, project, specific activities, aims, objectives and targets. According to this understanding, to analyse the content of a policy is a fundamental element in its comprehension. Thus, in this topic, the Content Analysis will be described by explaining the policy's main proposals and practical measures, its goals, aims, objectives, priorities and targets. The analysis will consider the resources required by the policy proposal and their availability.

Drawing upon Matus,17 in his Situational Strategic Planning (SSP), for the aims of this review, the content of a policy will be, in practice, seen as a set of measures combined, in a particular sequence, in order to transform a real situation (RS) into a desired one (DS), as schematically shown in Figure 2.

The analysis of health policy Content, shall lead to: inferences on the policy's competence to change reality and in which sense; conclusions on whether the proposals meet the needs shown by the Context Analysis; deductions about the group of measures as to whether they are sufficient and necessary ones, and if they are or not set up in the most coherent form.

Conclusions about the perspective of the policy shall also be made, by classifying it among the most common international approaches to social policy. As described in "the Student's Companion to Social Policy",18 there are several perspectives about social policy. Some of them are: the neo-liberal, the conservative, the social democratic, the socialist, the feminist, the racial and the green perspective. For the purposes of this review, the key perspectives on social policy which should be taken into account at least, are described below in their essential elements.

The neo-liberal perspective: "Neo-liberals favour a competitive market economy and oppose an economic system planned and directed by the state. They contend that a market economy is an essential bulwark of democracy because, by dispersing property ownership, it prevents the concentration of power in the hands of few" (Green; 1998: 60).19 "Neo-liberals historically did not make a sharp distinction between economic questions and other issues" (Green; 1998: 62).19 They also advocate that government have important but limited duties and that the State should increasingly reduce its role within society.

The conservative perspective: "The Conservative approach to social welfare starts not from the formulation of abstract principles, but from the reality of established social institutions such as the family, community, class, religion, private property and government. It expresses a preference for order and continuity and a spirit of deference to the past rather than an assumption that change is necessary because existing institutions are deficient or that progress has more to offer than does the status quo" (Pinker; 1998: 64).20

The social democratic perspective: this approach is classically based on the Marshal's conception of citizen rights. According to his idea, full membership of a society is defined by the ownership of three sets of rights, these being: civil rights, political rights and social rights. State assurance of social services and welfare is understood by the social democratic approach as decisive components of the package of social rights, which is itself one of the citizenship rights.21

The socialist perspective: in this case, the provision of social services is seen as a permanent duty of the state and it should be provided aiming to eliminate social inequality.22

The feminist perspective: as described by Lewis (1998: 86)23 "... it would be a mistake to think that it is possible to identify a single feminist approach to social policy " because this approach may vary according to someone's political point of view or ethnic background, for instance. However, it is possible to observe that a common feature is identifiable as the basis of the distinct feminist approach, which is the belief that the whole foundation of society is gendered. "Thus access to income and resources of all kinds - for example, education - is gendered, as are concepts that are crucial to the study of social policy: need, inequality, dependence, citizenship" (Lewis; 1998: 86).23

These five perspectives on social policy shall be of good value in order to compare and classify distinct cases of health policy.

How to analyse the Content of a health policy

According to Matus,17 a social policy and its proposals shall be formulated towards tackling a problem or a group of problems, identified by the social actor commanding the process. Walt3 also expresses a view on the policy oriented towards the problem. "Policy involves the decision to act on some particular problem, but it includes subsequent decisions relating to its implementation and enforcement" (Walt; 1994: 41).3 In this case, the review suggests that the identification must be made of what are the main problems motivating a given social actor to embark on a specific health policy. These problems should be revealed in each specific topic analysing the different health policies that may take place in a same period of analysis.

After identifying the problems, the Content Analysis will concentrate on the proposals themselves, their aim, goals, objectives and targets. Then it will draw attention to the resources required by the policy. Again based on Matus17 formulations, the main resources that should make up a policy are considered as: material resources, like equipment and money; technical and organisational resources, these sorts of resources meaning the know-how and organisational and managerial capabilities to carry out the proposals; and political resources, meaning basically the power to make the policy viable. Internal organisation of the proposals, in terms of sequence of actions and their external coherence, is also examined.

Finally, the proposal is appraised in terms of outputs and outcomes, the results or possible results, and also whether it matches or not the problems it was meant to tackle, namely external coherence, as well as how it relates to the problems identified in the Context Analysis performed before. As Matus17 advocates, an analysis of the policy's impact shall be also be carried out whenever it is the case. This whole sequence is schematically shown in Figure 3.

The problems generating the policy may be identified by this research, either from primary or from secondary sources in the Context Analysis. The policy is then described and analysed, in terms of resources, its impacts, internal and external coherence and its social policy perspective.

Problems in analysing the content

In contrast with the three other categories of this health policy process analysis, the context, the process and the social actors, there are few references to frameworks and procedures on how to carry out an analysis specifically on the content of a policy.

Matus,17 in his framework for "Situational Strategic Planning" (SSP), emphasizes the formulation of proposals. In fact, this is a "Moment" of his planning method, "The Normative Moment" (Momento Normativo), in which the "objective situation" as well as all proposals seeking to build it are formulated. The content of such proposals are methodologically and meticulously dealt with in that chapter of his influential work, particularly on Latin America.17 That was taken by this study as a reference for analysing the content of health policies.


The Actors: why analyse them in relation to the Health Policy process?

The actors are the agents of any policy. As formulators, implementers, supporters or opposition, they add their ideological values, political views and practical proposals to the process and, by doing this they determine the course of a policy process. As Walt and Gilson (1994: 362)6 pointed out: "For many writers concerned with policy analysis, the key determinant of a policy change is the group of actors involved, and the focus is often on government."

Drawing on Matus,17 in the context of this work, any institution, personality or social movement capable of producing events altering in some way a particular scenario are referred to as "actors". As explained by Matus,17 one common characteristic of all social actors is that they all possess a certain amount of power. Power being understood, according to the conceptualisation described earlier in this article and generally, as the capacity to bring about events or to affect the events brought about by others. In practical terms, the social actors identifiable by this method may be a person, such as a country's President or the Minister of Health; or an institution, like the national parliament or a particular non-governmental organisation (NGO); or even a social movement.

How to analyse the Actors

This work adopts as reference two influential formulations on how to conduct a dynamic analysis focusing on the actors, their inter-relations and their connections with the social environment being studied. These two formulations are very much alike, as will be shown below, but surprisingly, there is no connection of any type, assumed between them.

The first is the Actors Analysis model, proposed by Matus,24 in his SSP method, first published by the Pan American Health Organization (PAHO) in 1987, entitled: "Política, Planificación y Gobierno"."24 As expressed in the "iterative calculation at the political viability analysis", a topic in the 'Strategic moment' (Momento estrategico)17 chapter. This author and his work are very widely published and well known in Latin America, particularly in Brazil, as he acknowledges himself in the preface to the Brazilian edition of his quoted book.17 The issues that this model proposes to cover are shown in Box 3, and are compared with issues from the other statement referred.

The second formulation is the "players table" model, as part of the: Policy maker: Computer-Assisted Political Analysis Software and Manual package, by Reich and Cooper,25 issued in 1996. As Reich (1996: 2)26 explains: "... the software uses political mapping techniques to analyse the political actors in a policy environment. These techniques assess the power and position of key political actors, and then display the supporters, opponents, and non-mobilized players in a political 'map' of the policy. This mapping provides the basis for designing strategies of political management. Political mapping techniques have been used by Lindenberg and Crosby for assessing development policy reforms, from the perspective of government policymakers, and by Austin for analysing business-government relations in developing countries, from the perspective of private company managers."

Although the reference cited above by Reich,26 Lindenberg and Crosby, is from 1981 it was not quoted by Matus24 in his book, and the other, Austin, is from 1990.

Box 3 shows the sequence of steps for carrying out the actors'analysis according to both analytical models.

This work adopts these steps and suggest them to be adapted in order to suit the specificity of distinct cases, nevertheless they should be used in full accordance with the recommendations of both models.


The Health Policy Process Analysis

Within the limits of this work, the analysis of policy processes will mostly focus on the issues related to policy-making or decision-making. Although some analysts may differentiate policy from decision, as suggested by Walt,3 mainly by considering policy as something much bigger than a decision, it may in fact involve a series of decisions. However, when looking at decisions made in the policy-making process, it amounts to much the same feature, in the sense that, policy-making is indeed a sequence of decisions such as: deciding which problems are important to tackle; deciding which measures and resources should be put into action; deciding on how and when actions shall take place; deciding who to co-operate with or who to oppose or benefit. Actually, "... in everyday life the distinction between policy and decision is often blurred …" (Walt; 1994: 40).3

The sequences of decisions composing the policy-making process seem to be well identified by most frameworks set to analyse them. According to Foltz27 the policy process is often divided into phases or stages, such as: policy formulation, policy implementation, and policy evaluation and feedback. Otherwise, some analysts may subdivide the process into six stages: initiation, estimation, selection, implementation, evaluation and termination. Foltz27 observes that these divisions may suggest that the policy process has clear stages with a precise sequence and boundary in between them. She maintains that this is not exactly the case: "Some stages may be skipped altogether; others may occurs simultaneously as when a policy is adopted and implemented and modified at the same time" (Foltz; 1996: 208-9).27

Bearing these considerations in mind, models framing the policy process are largely used and, possibly because of that, they seem to be convenient for the systematic analysis of the whole process.

As Walt3 points out, the most consensual frameworks suggested to analyse the policy process are those which basically describe policy-making in four stages: "Problem identification and issue recognition. How do issues get on the policy agenda? Why do some issues not even get discussed? Policy formulation. Who formulates policy? How is it formulated? Where do initiatives come from? Policy implementation. Arguably the most important aspect of policy - yet often gets short shift. What resources are available? Who should be involved? How can implementation be enforced? Policy evaluation. What happens once a policy is put into effect? Is it monitored? Does it achieve its objectives? Does it have unintended consequences?" (Walt; 1994: 45).3

Taking that framework as reference, most aspects of the first and fourth stages are dealt with in the Content Analysis of the health policy. For that reason, the topic chosen to analyse the "process" of the health policy will consider the aspects within the second and third stages: policy formulation and policy implementation respectively, and the issues related to them. Also, according to Walt,3 the main aspects related to the policy process are: the issue of power, in terms of who decides and who influences it; the concept and types of policy, in terms of what is policy and how is policy made; and the rationale of the policy, in terms of its formulation as a rational process or a more incremental or mixed one.

Drawing upon that idea, the topic on process shall analyse the health policy precisely according to the following criteria.

Who decides and who influences

There are basically two major groups of opinion on who possess the power to make decisions and influence the health policy process. These are:

The pluralist view: this view a shared understanding is that power is spread and distributed amongst groups, institutions and individuals within society. Although it may be unevenly distributed, it is never concentrated or monopolised by just a few. Matus,17 offers a classical example of a pluralist view of power when he treats it as a resource that can exchange between distinct actors in society. "Power is the product of a social accumulation." Distinct social actor in conflict will loose power whilst others may gain it (Matus; 1996: 440-1).17

The elitist view: as described by Barker (1996: 83),8 "Elite theories place a central emphasis on the existence of a political elite which in any society will hold power, and which will consist of a network of those most powerful not only in political but also in business, military, aristocratic and bureaucratic circles." "There is here a concept of a ruling class, and it is assumed that the state will follow the preference of the elite. Agencies of government will fall within the elite network." Within this broad group, Marxism offers a classical example of such a concept when it considers that ruling society, and consequently policy-making, is something performed by a dominant class over the dominated one, plus the fact that the principal duty of the state is to preserve that class dominance.3

In this review, these two views are considered in order to explain the actors' influence over different health policy cases. However, as discussed below, it may be broadened and combined to provide a more adequate analytical approach to these processes.

How decisions are made: whether by an open pluralistic approach or a closed elitist one

It is understandable that those adopting the pluralist view should accept that the policy's aims and outcomes tend to favour the public interest, because directly or indirectly, via representatives, the majority have somehow influenced and supported the policy processes. The government in this case would, at least, be sufficiently open to deal and negotiate with representatives - sometimes remotely leading a big passive group - of various social groups, such as employers or employees. As recalled by Walt,3 the emergence of such powerful corporate groups may facilitate the appearance of another political phenomenon, which is corporativism. "It was argued that governments welcomed corporativism because it was easier to deal with one powerful group than with many - and the state could also then expect leaders to control any conflicting interest between members" (Walt; 1994: 37).3 O'Donnel28 also described the occurrence of corporatism in Latin America, particularly under authoritarian governments, as the direct relation between the government and distinct corporate groups within society, overlapping the regular representative body, like parliament, usually to negotiate their particular interests.

The idea that the government may be a neutral negotiator between the various groups within society is largely questioned. As Walt3 points out, public choice theorists fully rejects the government as acting to balance the interests of distinct groups. For them, the government has considerable power which is strongly linked to some other national and international powerful groups and therefore, it may act to suit its own interests.

Those supporting the elitist view, on the other hand, would expect that particularly in developing countries the circles able to participate and influence the policy-making process are exclusive and only few are able to realize their interests in the policy aims and objectives. However, as noted by Walt,3 there is a strong criticism of elitist views, in the sense that they overestimate the capacity of elites to manage the power and control to the majority. "Even in developing countries, where interest groups are not always sufficiently well organised to put effective pressure on government officials, certain groups will have access to government through, for instance, professional bodies or church" (Walt; 1994: 39).3 Fleury,29 identifies this phenomenon, particularly in Latin America: "Grassroots social movements and NGOs flourished in the region, becoming a channel for organizing and representing interests that did not fit into the old political order” (Fleury; 2000: 11).29

According to that observation, different levels of influence and openness are more likely to exist in different processes of policy-making. As a matter of fact, Walt3 observed in the formulations of Hall et al.,30 a variation in the level of influence over the policy-making process according to the magnitude of the policy. "Their notion of bounded pluralism suggests that issues of high politics - largely economic questions - are decided within an elitist framework, but that most domestic, routine policies on health, education, transport and housing are likely to be developed along pluralist lines, with some participation of different groups at different stages of the policy process" (Walt; 1994: 39).3 In that case, some health policy processes may be understood as being associated with high politics and magnitude, if they are perceived as related to macro-economic policies - as programmes of economic adjustment - or with low politics and magnitude, if they does not have repercussions on major politics or on economic policy.

According to that understanding, Walt3 developed a classification that should be used to analyse the policy process as treated in this study. Such a classification is shown in the adapted Box 4, in wich high politics correspond to Macro or Systemic policies and low politics to Micro or sectorial policies. Examples of these policies are also provided by the Box.

Another way to classify policies, also suggested by Walt,3 is by their main outcomes. According to this idea, the policies might be:

Distributive. When the policy promotes the delivery of goods or services to some groups in society, without disadvantaging or reducing the benefits of any other particular group. This particular type, usually tends not to be contested and therefore its process may develop openly for the groups involved on it. A programme delivering food supplements to combat children's malnutrition may be an example of this type of policy.

Redistributive. These policies are based on a transfer of resources from some groups to others, usually somehow affecting the pattern of income distribution within society. This may be done progressively, when there is a transfer from the better to those worse off, or regressively, when the transfer is in the opposite direction. In both cases, these policies tend to be strongly contested, at least by those paying for them. The creation of a new tax on particular goods consumed by the better off to support a health programme for the poor is an example of this category.

Self-regulatory. In case of health policy, for instance this would involve the Ministry of Health (MoH) introducing regulatory rules over its own organisation. This may occur as an attempt to re-arrange its organisational structure or its operating rules. As long as it does not have repercussions for other organisations, it is probably not controversial.

Regulatory. These policies frequently involve the implementation of regulatory restrictions by the health authority on other organisations, like private health insurance companies, or professional groups, such as doctors and nurses. These policies tend to be conflictive insofar as they restrict the space of the targeted organisation or group, though this might not arise the general interest, if the issue is a very particular one or if the group or organisation cannot spread out its concerns and build up alliances with major groups.

Walt3 also expresses concern that health policies often occurs as low politics, distributive or self-regulatory and regulatory policies, they are therefore less controversial and more likely to develop as pluralist processes rather than elitist.

How it is implemented: its advantages and obstacles; which strategies and activities are applied

As Foltz (1996: 209)27 points out: "It is most common that the implementation phase shapes profoundly the policy, as the practicalities of getting things working on the ground are negotiated by administrators." Walt3 also explains that implementation is arguably the most important moment in the policy process. Drawing on her formulations, some important features in policy implementation will be identified here as criteria for analysing the health policy process. These features are:

The way of implementation. There are two contrasting ways: the top-down approach or the bottom-up. A traditionalist way of implementation is the top-down manner. In this form, a clear separation between policy formulation is perceived; which would be highly political, hence a task for the national government's top officials; and a policy implementation, which would be more a managerial and administrative task, hence to be performed by managers and technicians under its command and supervision. These managers and technicians may be at the intermediate level (state governments) and at the local level (municipal governments). This is a typical military approach, applied in situations where prevail an authoritarian relationship between the national government and sub-national levels. The bottom-up form on the other hand, considers that policy implementers do play a very important role in the way policy evolves. Also, this conception does not make a clear separation between policy formulation and implementation. Although the national government may start the process, the sub-national levels are expected to add to it at each moment of the policy process. It is acknowledged that in practice, an intermediate manner of policy implementation between these two extremes, is mostly the case. Anyway, it may be evident from which of these two poles a real policy, implementation is closer.

Relations between the national government and sub-national levels in the policy process. This is very much a look at how the distinct levels of government - national, states and municipalities - share the burden and the bonus of the policy process. Are the relations bilateral or is there any sort of control and command by one level of government over the others. Mills et al.31 points out that most governments decentralise authority or transfer some power in public planning, management and decision-making, from national to sub-national levels. Although all decentralisation processes vary greatly, a certain degree of control may remain with national government over the sub-national levels. Araújo,32 in another article, pointed out important constraints on the Brazilian health sector decentralization process. "There is a gap between the type of decentralization (devolution) called for in the national constitution and the decentralization of activities actually implemented" (Araújo; 1997: 109).32 Walt3 observes that the most common forms of national government control over the sub-nationals ones are: budgetary control and legislative or regulatory control. Araújo32 has found these types of control and yet one more, the fragility of municipal governments. "The regulatory role over states and counties (municipalities) performed by the MoH, was established according to Basic Operational Rule number one…" "Leaving the counties (municípios) on their own to cope with their lack of skills, and subordinating them to a financial dependence in accordance with their accomplishment of a national level health policy, is a genuine example of a deconcentrating policy" (Araújo; 1997: 114-5).32 These three types of control are useful as references to check upon the health policy process as suggested in this review.

Advantages and difficulties for policy implementation. There are factors that may facilitate or impede policy implementation. Walt3 has identified the following: the type of policy may be one of these factors. If a policy is a systemic one, producing major political or economic changes, it is likely to be hard to implement. On the other hand, if a policy seeks only marginal changes on the whole status quo, it is an advantage for its implementation, because certainly the amount of all types of resources required for it are much less than with a policy with extensive changes. Another factor that may be an advantage is when the policy requires just one actor for its implementation. As more actors involved in its implementation are required, as greater agreement needs to be built, obstacles may upset the process. Another point is related to the policy goals and objectives. When these are clearly stated and perceived by the targeted population, as directly benefiting them, it is most likely to work out as an advantage for policy implementation. On the other hand, policies looking for bureaucratic or organisational re-arrangements, that may only indirectly benefit some groups, who cannot see these as the policy-makers do, may face more obstacles. Another factor seems to be related to the timing of policy implementation. If a policy can be quickly introduced it mostly works out as an advantage, because it does not allow time for possible hesitations or disputes about its alternatives and possible outcomes.

Strategy and actions for policy implementation. Matus17 argues that the context surrounding policies is usually characterized by a scarcity of political, financial, managerial and technical resources. Policy-makers have then three possible attitudes to adopt before such a reality: to ignore these constraints and go ahead in an attempt to implement the policy, which may turn into an adventure; or to recognise the constraints and then reduce policy' objectives, adjusting it to the scarcity, which may lead to immobility and to be led by the circumstances; and finally, to analyse the constraints and understand the reasons for the shortage of resources, in order to build bridges leading to them, which would in the end be a strategic attitude towards policy implementation.17 In this case, strategies are actions taken in order to obtain the scarce resources impeding the policy implementation, whether they are political resources, financial, managerial or technical ones.

According to Matus,17 there are essentially three types of strategy. In his own words: "... having chosen to recognise the constraints and to look for a way of increasing the scarce resources needed, then is question of designing a strategy" (Matus; 1996: 409-10).17 Generally, the possibilities are: a) strategies of co-operation with other actors; b) strategies of co-optation of others actor; and c) strategies of conflicting with other actors.

Co-operation means negotiating with others in order to find ways of acting together. The actors give away part of their interest to achieve a shared result as a consequence of their concerted action. Co-optation, on the other hand, signifies that an actor will bring other actors to his position, by attracting them because of his political power, by convincing them or by misleading them. Conflict seeks to reduce the relative power of actors with opposite interests by confronting them in various ways. Co-operation, co-optation and conflict are not mutually exclusive among the different actors involved in a policy process, nor among the same actors in relation to various activities of the policy process.17

Resource implications. Similar to Matus,17 Walt3 also identifies as the main resources of the policy process: political, financial, managerial and technical resources. In order to assess the availability of political resources, suggests a list of questions. "How legitimate is the government? How stable is it? Is there a possibility of an opposition arising? How popular or unpopular is the policy? Is there sufficient time and general conditions for the government to master a political base of support? How independent is the government? Does the policy affect important interest groups? What is the position of elites groups in relation to the policy?" (Walt; 1994: 174).3

For the most processes, the political resources are the main constraint on policy implementation. Thus, policy-makers actually have to be able to conduct a political analysis of the policy consequences on the political forces, as well as to manoeuvre in order to mobilise those potentially in favour of the proposal and to neutralise those likely to oppose it.17

Certainly policy also needs financial, managerial and technical resources to be feasible. Often the availability of financial resources is also a matter of political viability, in the sense that, once it is assured political support to a policy, the financial resources necessary for its implementation just comes as consequence of such a support. Walt (1994: 176)3 gives a good example of such a case: "... anything to do with liberal-based policy - say reform in the health sector - is likely to attract some external, donor funding in the ideological climate of the 1990s."

Managerial skills such as budget control, cost systems and human resource development are fundamental tools to facilitate policy implementation and the achievement of aims and objectives. This is also very much a matter of technical support. For instance, in areas like data-processing and health information, administrative techniques and policy analysis, whenever these resources are lacking, despite other issues, policy is at serious risk of distortion and failure. Accordingly, policy makers have to acquire the ability to set up strategies that facilitate the availability of enough political, financial, managerial and technical resources for adequate policy implementation.3

Rationale for policy implementation: rational,incremental, mixed or strategic

 According to Walt3 there are basically three models explaining how policy develops from the first stage, problem identification, to the last one, policy evaluation. These are:

Rational or synoptic model. Essentially, those sharing this understanding believe that the policy is developed in a rational way, by well-prepared policy-makers in control of the whole set of variables conditioning the process. This is a very normative approach, in which, from problem identification to policy evaluation, the best technical decisions can be made. Unexpected facts may be foreseen, the resources can be well calculated, and in the end the aim and objectives should mostly be effectively achieved. This is also a prescriptive model. However, there are a lot of criticisms of this view. They state that: policy-makers are not in control of reality; the problems are not always clearly identified; their skills cannot overcome all sorts of obstacles;33 policy-makers are not neutral in making the best technical choice, in fact they are also guided by their own values and beliefs;8 and also that reality is not as malleable as the policy-makers might wish. In conclusion, it is claimed to be an ideal model, which contradicts the dynamism of real life. This also corresponds to most of the criticism made by Matus17 in his recommendations on how not to plan, criticising what he calls the traditional planning.

The incrementalist model. Fundamentally this is a pragmatic understanding of how the policy process develops. It considers that, in practice, policy-makers do not separate clearly the four stages of the policy process. Thus, the problems selected are often the ones which the policy-makers already domain the resources to act on them and in which their values guide them to select, and frequently some action has already been implemented on them. Policy-makers also look only to the main alternatives of implementation and the best option is often the one that all the policy-makers involved agree with, which is not necessarily the best technical choice. Also, the incrementalist idea considers that policy is mainly corrective and looking for gradual changes on the already established policies, rather than operating great instant transformations. The problems are transformed, not exactly solved. Therefore, a continuous intervention on the successive unfolding situations is more likely to be the real attitude.3

Most distinguishable in the incrementalist pragmatism is the recognition of a inevitably pluralist policy process. As pointed out by Walt,3 Lindblom described as " ... the best known of the incrementalists", noted the policy process was "... affected by partisan mutual adjustment", as he called "... the process of negotiation, bargaining and adjustment between different groups to influence policy" (Walt; 1994: 49).3 Ultimately incrementalism is concerned with a factual policy process rather than its theoretical idea. It focuses on what is rather than in what ought to be.

The criticism against the incrementalist idea accuses it of lacking in commitment towards major change and of being conservative, as well as reinforcing inertia and the status quo.34

Mixed scanning and normative optimal models. This understanding seeks a position between the rational and incrementalist schemes. Basically, mixed scanning splits the decisions made in the policy process into two categories: the macro or fundamental decisions and the micro or small decisions. Etzioni35 believes that this model is both a good interpretation on how the policy process develops and also a good prescription to guide the policy-makers in action. According to this idea, to deal with the macro decisions there are an enormous number of variables and the degree of uncertainty and "circumstantial blindness" is very high.17 Therefore, it is impossible to analyse in detail all the alternative decisions and to foresee the consequences of them all. The attempt to do so could consume a great amount of time and other resources and would in the end be ineffectual. In that case, policy-makers should explore only the major and most probable alternatives and make decisions on that basis. On the other hand, micro decisions are usually made with few possible alternatives, which allows a quick detailed analysis of each one, enabling a more precise decision.3

Strategic approach. In this work, this other rationale for the policy process will be added to the previous three. This is an attempt to specify those approaches that concentrate on the issue of making the policies viable. In one sense this approach may be close to the rational one, in terms of usually aiming at big changes. However, it is not rigid nor over-normative not to consider all the limitations conditioning the policy process, and not to seek incremental changes, like shortcuts for major ones. In fact, the emphasis of this approach is to find actions that work out as mediations allowing for a way in between of giving up the great change policies, due to their initial lack of viability or feasibility, or working purely in the field of incremental policies which are already feasible, thence little altering the reality. Strategy and strategic action should be this mediation. Matus17 and his SSP model is very representative of this view. Dror34 resembles this approach when he favours the idea of extra-rationality, as innovative ideas, creative inventions, brainstorming, finding acute judgements; as a tool to strengthen and improve the policy process.

These criteria have comprised a model of analysis that involves both, a macro-view of the policy process, more concerned with the politics and issues related to power within the process, and a micro-view, more concerned with mechanisms and activities used to implement the policies.



There is an evident wave of international interest in health policy analysis, so it is necessary to establish mechanisms for shared learning at the global level. Moreover, operational frameworks for health policy analysis may workout as replicable tools facilitating national and international comparisons and providing the means for diffusion of experiences and knowledge.

The review identified several types of policy analysis. They were mainly types of partial analysis (meta-analysis, meso-analysis, decision-making analysis, impact-analysis) and comprehensive analysis (models by Frenk, Reich and Walt and Gilson.).1,5,6 However, they all seemed to lack in the right combination of sensitivity, to get the most from the analysis; generalizability, to be applicable in distinct situations; and accuracy, in order to allow the most reliable findings. The Walt and Gilson6 analytical model was taken as the basis to build upon it the framework for health policy analysis as here presented. That model calls for comprehensive HPA, especially in developing countries where it seems to have been neglected. It recommends analytical attention to four inter-related aspects, which encompass the entire policy. These are: the actors, the content, the context and the process. This study practically developed each one of these aspects into operational categories of analysis clearly pointing out the issues and evidences that should be looked for by policy analysts in each one of them.



1. Frenk J. Comprehensive policy analysis for health system reform. Health Policy 1995; 32: 257-77.         [ Links ]

2. Collins C, Hunter D, Green A. The market and health sector reform. J Managem Med 1994; 8: 42-5.         [ Links ]

3. Walt G. Health policy. An introduction to process and power. People, governments and international agencies-who drives policy and how it is made. London: Witwatersrand University Press; 1994.         [ Links ]

4. Parsons W. Public policy. An introduction to the theory and practice of policy analysis. Cheltenham, UK: Edward Elgar; 1995.         [ Links ]

5. Reich M. The politics of health sector reform in developing countries: three cases of pharmaceutical policy. In: Berman P, editor. Health sector reform in developing countries. Making health development sustainable. Boston: Harvard University Press; 1995. p. 59-100.         [ Links ]

6. Walt G, Gilson L. Reforming the health sector in developing countries: the central role of policy analysis. Health Po-licy Plan 1994; 9: 353-70.         [ Links ]

7. Collins C, Green A, Hunter D. Health sector reform and the interpretation of policy context. Health Policy 1999; 47: 69-83.         [ Links ]

8. Barker C. The health care policy process. London: Sage; 1996.         [ Links ]

9. Walt G. Policy analysis: an approach. In: Janovsky K, editor. Health policy and systems development. An agenda for research. Geneve: World Health Organization (WHO); 1996. p. 225-41.         [ Links ]

10. May T. Social research. Issues, methods and process. Philadelphia: Open University; 1997.         [ Links ]

11. Taylor M. Macroeconomic theory. In: Kuper A, Kuper J, editors. The social science encyclopedia. London: Rou-tledge; 1996. p. 489-91.         [ Links ]

12. Denzin N, Lincoln Y. Handbook of qualitative research. London: Sage; 2000.         [ Links ]

13. Hardin R. Power. In: Honderich T, editor. The Oxford companion to philosophy. Oxford: Oxford University Press; 1995. p. 708-9.         [ Links ]

14. Hay C. Divided by a common language: political theory and the concept of power. Politics 1997; 17: 45-52.         [ Links ]

15. Philp M. Power. In: Kuper A, Kuper J, editors. The social science encyclopedia. London: Routledge; 1996. p. 657-61.         [ Links ]

16. Lukes S. Power. A radical view. London: Macmillan; 1974.         [ Links ]

17. Matus C. Política, planejamento e governo. Brasília, DF: Intituto de Pesquisa Econômica Aplicada (IPEA); 1996.

18. Alcock P, Erskine A, May M. The student's companion to social policy. Oxford: Blackwell; 1998.         [ Links ]

19. Green D. The neo-liberal perspective. In: Alcock P, Erskine A, May M, editors. The student's companion to social policy. Oxford: Blackwell; 1998. p. 57-63.         [ Links ]

20. Pinker R. The conservative tradition of social welfare. In: Alcock P, Erskine A, May M, editors. The student's companion to social policy. Oxford: Blackwell; 1998. p. 64-70.         [ Links ]

21. Sullivan M. The social democratic perspective. In: Alcock P, Erskine A, May M, editors. The student's companion to social policy. Oxford: Blackwell; 1998. p. 71-7.         [ Links ]

22. Ginsburg N. The socialist perspective. In: Alcock P, Ers-kine A, May M, editors. The student's companion to social policy. Oxford: Blackwell; 1998. p. 78-84.         [ Links ]

23. Lewis J. Feminist perspectives. In: Alcock P, Erskine A, May M, editors. The student's companion to social policy. Oxford: Blackwell; 1998. p. 85-90.         [ Links ]

24. Matus C. Política, planificación y gobierno (Borrador). Washington, DC: Pan American Health Organization (PAHO); 1987.         [ Links ]

25. Reich MR, Cooper DM. PolicyMaker: coputer-assisted political analysis [;on line]; 1996. Brookline: PoliMap. Available from: <> [;2000 Mar 1]         [ Links ];

26. Reich MR. Applied political analysis for health policy reform [;on line]; 1996. Boston: Harvard School of Public Health. Available from: <> [;2000 Jun 22]         [ Links ];

27. Foltz AM. The policy process. In: Janovsky K, editor. Health policy and systems development. An agenda for research. Geneva: World Health Organization (WHO); 1996. p. 207-26.         [ Links ]

28. O'Donnell G. O corporativismo e a questão do Estado. Cad Dep Ciência Política Univ Fed Minas Gerais 1976; 3.         [ Links ]

29. Fleury S. Reforming health care in Latin America: challenges and options. In: Fleury S, Belamartino S, Baris E, editors. Reshaping health care in Latin America. A comparative analysis of health care reform in Argentina, Brazil, and Mexico. Ottawa: International Development Research Centre; 2000. p. 3-24.         [ Links ]

30. Hall P, Land H, Parker R, Webb A. Change choice and conflict in social policy. London: Heinemann; 1975.         [ Links ]

31. Mills A, Vaughan J, Smith D, Tabibzadeh I. Health system decentralization. Concepts, issues and country experiences. Geneva: World Health Organization (WHO); 1990.         [ Links ]

32. Araújo JL. Attempts to decentralize in recent Brazilian health policy. Int J Health Serv 1997; 27: 109-24.         [ Links ]

33. Motta P. Gestão contemporânea: a ciência e a arte de ser dirigente. Rio de Janeiro: Record; 1991.         [ Links ]

34. Dror Y. Public policy making re-examined. Oxford: Tran-saction; 1989.         [ Links ]

35. Etzioni A. Mixed-scanning: a third approach to decisionmaking. Public Adm Rev 1967; 27: 385-92.         [ Links ]

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License