Score.su.se

Regulating NeedDeciding on public financial intervention withinthe fields of healthcare and development aid Anna Krohwinkel-Karlsson and Ebba Sjögren Regulating NeedDeciding on public financial intervention withinthe fields of healthcare and development aid Anna Krohwinkel-Karlsson and Ebba Sjögren Score working paper 2006:1ISBN 91-89658-38-8ISSN 1404-5052 Deciding on public financial intervention within the fields of healthcare and development aid Stockholm centre for organizational research AbstractNew Public Management involves a movement from hierarchical planningtowards increased reliance on market-like forms of coordination and con-trol. Yet for public organizations to adopt a market approach is broadly ac-knowledged to not automatically create a linear matching of individual andpublic interests. Growing individualization may favor selection of ‘clients’and unequal provision on part of service providers. At the same time, cli-ents may claim services without or above their actual need. An answer tothe dual problems of overspending and supply-led provision has been theorganizational separation of public service provision and assessment ofservice needs. In principle, the practice of ‘neutral’ needs assessment hasbeen thought to stimulate fair and just demand responsiveness, with anequal and efficient allocation of resources without (over) representation ofany single interest group(s). The call for fair and just needs assessment hastherefore been articulated in similar ways across a variety of sectors. How,then, do public organizations operating in different functional areas prac-tice the notion of needs assessment? This topic is investigated by comparing two government agencies in Sweden: Sida and LFN. The former organization administers Sweden’s in-ternational development aid. The latter decides which prescription drugs toinclude in the national public pharmaceutical benefit. A comparison of theorganizations’ work in assessing needs informs a discussion about the im-pact of organizational setting and process logic on the regulatory practicesof the studied agencies.
Abstract. 11. Public management and the assessment of needs . 62. Needs assessment by two Swedish agencies . 83. Needs assessment in principle: development aid vs. pharmaceutical sub-sidization. 9 3.1 Development aid. 93.2 Pharmaceuticals in healthcare services. 12 4.1 What triggers needs assessment? . 154.2 What are relevant needs? . 174.3 How are needs for intervention determined? . 19 5.1 Organizational setting . 245.2 Process logic. 255.3 Sida and LFN: management by rules or by organization . 26 6.1 Public material. 296.2 Interviews . 30 1. Public management and the assessment of needsThe past few decades have witnessed similar changes in the organizationand role of the public sector in most Western countries. 1 One way to un-derstand New Public Management (NPM) is as a movement from hierar-chical state planning towards increased reliance on market-like forms ofcoordination and control (Pollitt 1993; Hood 1995; Christensen & Lægreid2002). An underlying assumption behind the NPM reform model is that de-sirable states of aggregated public utility can best be attained through thesatisfaction of individual preferences. That is to say, allocation of resourcesshould not be made through planning and top-down decision-making, butshould be driven by demand. In this context, the concept of needs has be-come more widely used and widely understood within disparate fieldsadopting market-inspired styles of resource allocation.
Yet for public organizations to adopt a market approach and seek to be demand-driven is not without problems. A market approach is broadly ac-knowledged to not automatically create a linear matching of individual andpublic interests. Notably, growing individualization may favour selectionof ‘clients’2 and unequal provision on part of service providers. At the sametime, on a ‘free market’ clients may claim services without or above theiractual need.3 An answer to the dual problems of overspending and supply-led provision which has increasingly become a central component of the‘regulatory state’ (Majone 1994; MacGowan & Wallace 1996) is the orga-nizational separation of public service provision and assessment of serviceneeds. In principle, the practice of ‘neutral’ needs assessment has beenthought to stimulate fair and just demand responsiveness, with an equal andefficient allocation of resources without the (over) representation of anysingle interest group(s).4 As Mohr (2005) argues, discourses about the nature of needs are funda- mental components of social welfare systems because the interpretation of 1 The authors gratefully acknowledge the constructive and supportive comments received fromCarmen Huckel, Jim March, C-F Helgesson and other participants at the Score conference Or-ganizing the World – Rules and rule-stetting among organizations, 13-15 October 2005 inStockholm, Sweden.
2 A key tenet of marketization reforms in the public sector has been the creation of customers,clients or users – as opposed to citizens or recipients.
3 This is particularly problematic when there is a distributed customer role, i.e. when the cost ofservices is not directly incurred by the actor partaking of services.
4 As Stone (1997) points out, independency can be considered a paradoxical concept: it servesgoals or interests by not serving (opposite) interests in particular.
needs and of how people of various status identities are linked to needs dis-courses has a profound impact on the relief practices that are considered le-gitimate.5 While the understanding of the term needs has varied over timeand with the application of different national and political models so hasthe notion of what constitutes appropriate needs assessment (see discus-sions in Bradshaw 1971; Campell 1976).6 Within the discourse of NPM,there are a number of different organizational practices and forms whichare justified as appropriate means for assessing needs for public interven-tion. In some cases, specialized assessment bodies have been instituted toperform assessments. These bodies have been comprised of both politi-cians, civil servants and/or various ‘experts’. Elsewhere, policy formulationand implementation have been divided through various contracting-out ar-rangements. In contrast to the traditional welfare state, which integratedregulatory, operating and policy-making functions, the new models havetended to create autonomous agencies responsible for different tasks (cf.
Christensen & Lægreid 2005).
While the discourse of NPM may seem consistent, it is likely to underpin different actions in practice (cf. Brunsson 1989, 1995; Fernler 1996; Clark2004). One reason for such differences in how ideas are enacted is that re-form models often include only vague and general guidelines for applica-tion (although they tend to be precise in the way they are labelled). Thisvagueness makes concepts applicable in many settings, but it also opens upfor different organizations to adopt different practical usage of the models(cf. Czarniawska & Joerges 1996). Differences in organizational and insti-tutional structure, for example between countries (Clark 2004), has beenseen as one source of interpretational variation.The idea of neutral needsassessment, as practiced within two organizational bodies with the sameformal status as autonomous governmental agencies, will be the topic ofthe present paper. For, as will be shown below, the organizations’ inter-pretation of appropriate needs assessment varies considerably.
5 We emphasize the importance of needs assessment, although there are different conceptions ofsocial justice which view need as but one of many relevant concepts (see overview Boyne et al.
2001).
6 Bradshaw’s (1971) typology, for example, differentiates between who should determine needsand on what basis unfulfilled needs are delineated. Normative need is determined by experts’evaluation, whereas felt need is based on self-perception. This, in turn, is different from ex-pressed need, which is linked to the notion of demand for services, and so-called comparativeneed which bases need assessment on a comparison of parties with similar characteristics butdifferent service access or use.
2. Needs assessment by two Swedish agenciesWe have compared the work of two Swedish governmental agencies -- theSwedish International Development Cooperation Agency (Sida) and thePharmaceutical Benefits Board (LFN) -- in assessing and determining theform and content of their specific policy choices. The former organizationmanages Sweden’s international development aid. The latter decides whichprescription drugs to include in the national public pharmaceutical benefit.
Both organizations’ activities are rooted in the tradition of the ‘Swedishmodel’, presuming a (relatively) strong role of the state in the provision ofboth healthcare services and poverty relief (cf. Premfors et al. 2003 for abroad overview).
The two studied agencies are active in different functional contexts, but have many common characteristics. They are both national governmentalagencies, and as such part of the Swedish state organization and subject tomany of the same rules and regulations. Sida and LFN are also both‘autonomous’ in the sense that they are formally separated from their re-spective ministries (cf. Pollitt 2005). They also perform a similar task: tomake decisions about resource allocation and forms of financial interven-tion. In doing this, both organizations express adherence to the principle ofneutral needs assessment.
Where Sida and LFN markedly differ is in the means that the two agen- cies have at their disposal for performing needs assessment and regulatingactivities based on such assessments. Sida’s needs assessment is part of aportfolio of activities which also includes policy-making and managerialsupervision. In contrast, LFN’s main activity is to make decisions based onassessments of pharmaceuticals’ usage (which includes assessment oftreatment needs). Differently put, LFN is an archetypical example of whatLægreid et al. (2005) define as purely ’regulatory’ agency, while Sida isakin to the notion of a ‘hybrid’ agency.7 Hence, the questions guiding this 7 The concepts of ’regulation’ and ’rules’ have be used in different ways in the literature.
Christensen and Lægreid (2005) distinguish between three meanings: In the narrowest sense re-gulation means formulating authoritative sets of rules and setting up mechanisms for monito-ring, scrutinizing, and promoting compliance with these rules. Second, regulation can be definedmore broadly as state intervention in the economy or the private sphere designed to realize pub-lic goals. This goes beyond rule-making to include areas like taxation, subsidies, and public ow-nership. Third, regulation can be seen as social control of all kinds, including non-intentionaland non-state mechanisms. Needs assessment, as performed by Sida, thus falls into the secondof the above categories. LFN, on the other hand, more clearly practices regulation according tothe first, more narrow, definition.
paper are firstly, to what extent needs assessment practices differ in thesetwo types of organizations, and secondly, how differences can be ex-plained. Assuming that organizing is extensively context-dependent, andthat one key feature of organizational context is the nature of the primaryactivities in which a particular organization is engaged (cf. Minzberg 1979;Whitley 1988), we will argue that differences in how Sida and LFN canintervene in their targeted activities influence how they practice needs as-sessment. In particular, the means of intervention play a role in how theidea of needs assessment is enacted, and how needs for intervention areelicited.
The rest of the paper is structured as follows: The next section describes the organizational and institutional settings in which Sida and LFN operate,respectively. We then move to study and reflect on Sida's and LFN’s needsassessment practices.8 In particular, we compare the means through whichthe issue of needs assessment arises in the two agencies, how relevantneeds are determined, and subsequently evaluated. Finally, we presentsome tentative empirical and theoretical conclusions based on differencesand similarities between the organizations’ needs assessment practices.
3. Needs assessment in principle: development aid vs. phar-maceutical subsidization3.1 Development aidThe Swedish government has been involved in international developmentassistance since the end of World War II. In 2004, Sweden’s total appro-priation for foreign aid amounted to approximately 2.3 billion € and wasimplemented in more than 120 countries in Africa, Asia, Latin America andCentral and Eastern Europe.9 Sida10 is the state agency responsible for allo-cating Swedish resources, both in terms of disaster relief – commonly re-ferred to as ‘humanitarian assistance’ - and long-term aid projects and pro- 8 The empirical section of this paper is based on a combination of material from two separatedissertation projects on pharmaceutical subsidization [by Ebba Sjögren] and development aidallocation [by Anna Krohwinkel-Karlsson].
9 Out of these, Sweden has in-depth bilateral programmes of co-operation with some 40 countri-es, while support to the other countries is channeled through multilateral programmes, largelyvia the UN and the EU.
10 When the organization was founded in 1965, the correct acronym was SIDA (Swedish Inter-national Development Authority). The ‘new’ Sida (Swedish International Development Co-operation Agency) was formed in 1995 through a merger of SIDA and four smaller Swedish aidagencies.
grams – ‘development co-operation’.11 Organizationally Sida is located un-der the jurisdiction of the Ministry for Foreign Affairs. Its current role isadministrative in nature: its main tasks are to formulate policies for Swed-ish support and to prepare, finance and evaluate individual contributions.
(By contrast, Sida does no longer carry out any projects of its own, butchannels implementation through various intermediaries12). Sida alsoserves an important function as a dialogue partner with other actors in thedevelopment co-operation sector, notably civil society organizations inSweden and abroad, governmental bodies in the recipient countries, andother donors.
Sida’s overall activities are managed by the director general and super- vised by a board.13 It is a matrix organization that is currently divided intofour regional departments, five sector departments, and a number of intra-agency functions. In addition, Sida has approximately 40 offices at Swed-ish embassies in the partner countries.14 Contribution management is nor-mally shared between regional and sector departments, with regional divi-sions assuming responsibility for the programming of country portfoliosand sector divisions taking care of the planning and monitoring of individ-ual projects/programs.15 Decisions about new allocations are taken on vari-ous levels in the organization; however the approval of a special ‘projectcommittee’ is required for most types of contributions over 50 mSEK (5.3million €).
The government issues instructions to its authorities, specifying their mandates. The guidelines for Sida are provided in two forms: 1. long-term ordinances that establish the basic framework. The overall goal of Sweden’s development co-operation is ‘to contribute to anenvironment supportive of poor people’s own efforts to improve theirquality of life’. Furthermore, all areas of foreign policy should ‘con- 11 All in all, Sida handles almost three-quarters of the total Swedish aid budget. Other govern-mental bodies manage the reminder.
12 See Krohwinkel-Karlsson (2005) for an historical perspective on shifts in development thin-king and in the role of aid agencies.
13 The director general is chairman of the board which has eleven members. They represent po-litical parties, trade and industry, the trade unions and organizations working with internationaldevelopment cooperation.
14 In August 2004, Sida had 769 employees of whom 165 were working abroad.
15 The picture is further complicated by the fact that Sida increasingly delegates tasks andauthority to its field offices.
tribute to equitable and sustainable development’ (Gov. bill2002/03:122, pp. 1; 58).
2. annual appropriation directives, which specify financial allocations to individual geographical regions and policy areas; the goals and pur-poses of these allocations; and directives for reporting back to thegovernment over the forthcoming financial year.
In addition, the government issues specific instructions, notably in the form of decisions on regional and country strategies. 16 Within the frame-work of its instruction, Sida has formulated normative regulations includ-ing procedures for work and decision-making, external communication,procurement and various administrative issues. In addition, a number ofinternal policy documents and handbooks that serve to guide the applica-tion and interpretation of these regulations have been produced.17 Thus,Sida both sets policy and makes decisions based on policies from outsidethe organization.
To ensure that public resources are used in accordance with overall goals and to facilitate its decision-making, Sida has set up a number of criteria tobe considered during the support preparation process. (As we will see later,there is a difference here to LFN, which does not set own decision criteriabut has significant interpretive leeway). The assessment criteria are speci-fied as: relevance (including needs assessment); effectiveness; feasibility;sustainability; quality of the development cooperation framework; and risksand risk management. One important task of the agency is to analyze andweight these aspects against each other when designing pro-jects/programmes, meaning that a mix of moral, political and economicfactors have to be taken into consideration. The specific matter of needs as-sessment is further complicated by the fact that poverty is multidimensional(i.e., it manifests itself differently depending on the specific situation and isperceived differently by those affected), and that many types of contribu- 16 Country and regional strategies are commissioned by the Swedish government and workedout in collaboration between Sida and the Ministry for Foreign Affairs. The government appro-val of a strategy normally includes a delegation to the director general of the right to decide oncontributions and other issues that fall within the framework of the strategy. Within Sida, furtherdelegation of authority takes place.
17 According to a recent inventory by Sida’s policy unit, there is a hierarchy of internal policiesthat govern operations consisting of 1) main policy documents (Sida Looks Forward, Perspecti-ves on Poverty and Sida at Work); 2) crosscutting policies (e.g. on environmental impact andgender equality); 3) sector policies (for health, education, trade, energy etc.); and 4) position pa-pers and other ’policy-like’ documents (Svensson & Holmgren 2003).
tions use indirect channels for poverty reduction.18 Thus, needs assessmentas a resource allocation criterion in its own right figures most prominentlyin humanitarian aid, which by definition should be free from any political,economic and military objectives and implemented ‘solely on the basis ofneed’ (Skr. 2004/05:52).
3.2 Pharmaceuticals in healthcare servicesSweden maintains a system of universal, publicly financed healthcare andhas provided its’ citizens with a pharmaceutical benefit since 1955. In 2003the total cost of healthcare services (excluding elderly care) was approxi-mately 22.1 billion € (Socialstyrelsen 2004, p. 11). The cost of healthcareas percentage of GDP has remained fairly constant over the past decades,but the absolute cost and cost per citizen has increased by more than fortyper cent since the early 1990s (Socialstyrelsen 2003, p. 282). In 2004, thecost of the public pharmaceuticals benefit was approximately 2.0 billion €.
The cost of the pharmaceutical benefit has increased by an annual averageof five per cent (Socialstyrelsen 2005, p. 12).19 The budgetary and operational responsibility for healthcare services, in- cluding the cost of pharmaceuticals, lies with the county councils (SFS1982:763). The county councils have independent power of taxation to fi-nance their operations. They also receive directed grants from the state. Yetwhile the county councils have operational autonomy, the central govern-ment also regulates many aspects of healthcare services such as the certifi-cation of professional personnel (ibid.). The Ministry of Health and SocialAffairs organizes a number of governmental agencies, which perform vari-ous activities to promote and ensure good health and medical care on equalterms for the entire population. One such agency is the Medical ProductsAgency (MPA), which is responsible for regulation and surveillance of thedevelopment, manufacturing and sale of drugs and other medical products.
Another agency is the Pharmaceutical Benefits Board (LFN).
LFN was established in October 2002 with the formal task to make deci- sions about which prescription drugs should be included in the publicpharmaceutical benefit (SFS 2002:160). Prior to LFN’s creation, all pre-scription drugs deemed safe for use in Sweden by the MPA were ‘auto- 18 Many recent efforts aim at influencing host country institutions, with the idea that improvedgovernance will over time lead to sustainable poverty alleviation.
19 2003 and 2004, the years following the latest reform of the public pharmaceutical benefit, sawthe first year of stagnated growth in pharmaceutical spending (Socialstyrelsen 2005, p. 12).
matically’ subsidized (Gov. bill 2001/02:63).20 Following the passing ofnew legislation, the use of a newly approved product is not subsidized untilLFN has decided that it should be. If LFN denies subsidization, a drug canstill be prescribed, however patients must assume the cost of treatment.
And if a product is granted subsidization, the budgetary and operational re-sponsibility for pharmaceutical usage lies with the county councils.
The LFN organization is made up of two parts: the bureau and the board.21 According to the organization’s formal work plan it is the job ofthe bureau to prepare a memo detailing findings about a pharmaceutical.
During its work with the evaluation memo, the bureau can be given in-structions by the board. Based on the final document, the board then makesa decision about whether the product will be subsidized or not (LFN 2002;LFN 2003a). The board can make one of three decisions: to grant the prod-uct unrestricted subsidization, to deny the product all subsidization or togrant the product-restricted subsidization. When the board is to set restric-tions and what types of restrictions to set is by law discretionary.22 In making its’ decisions, the board must adhere to the Act on Pharma- ceutical Benefits (2002:160). The law outlines the criteria that the officeand the board must consider when making their evaluations and decisions,respectively. A prescription drug should be subsidized provided: that the cost for using the pharmaceutical, with consideration given to 2 § in TheHealth and Medical Service Act (SFS 1982:763) is reasonable from medical, hu-manitarian and socio-economic perspectives’ (authors’ emphasis added).
What constitutes a reasonable usage is not specified. But that the consid-eration of the various perspectives involves needs assessment is made ex-plicit in the bill submitted by the government to the parliament. This docu- 20 Before LFN can make its evaluation of a pharmaceutical, the drug must be approved for usein Sweden by the MPA. The MPA evaluates whether a drug is safe to use and also specifies forwhich medical condition(s) the drug should be used to treat. Once the MPA has judged that aproduct is safe to use, it can be prescribed by a physician (or another professional practitionerwith the legal right to prescribe drugs).
21 The bureau employs approximately thirty individuals, many of whom hold doctorates inpharmacy and health economics. The board is comprised of eleven members who are personallyappointed (by the government), but whose personal backgrounds and formal expertise reflectdifferent interest groups within the healthcare sector (SFS 2002:719). Notably, the pharmaceuti-cal industry are not represented on the committee; pharmaceutical companies are instead theformal counterparts in LFN’s decision-making processes.
22 LFN can ‘under particular circumstances’ (11§, SFS 2002:160) chose to include a pharma-ceutical in the public benefit for certain areas of use. The board’s decision can also be ‘combi-ned with other particular conditions’ (ibid.).
ment includes repeated references to the need for imprecise instructions forthe new agency due to the difficulty of specifying practice for an organiza-tion with a new and complicated task (see Gov. bill 2001/02:63, p. 43).
However the bill does specify that LFN’s work is to be guided by threeprinciples and informed by the prioritization of illnesses approved by theparliament in 1997. The three principles are: 1. Equal human value, stating that all people have an equal right to life 2. Need solidarity, meaning that those with greatest need of treatment should have priority over those with lesser need 3. Cost-effectiveness, that the benefit of treatment must be reasonable in The principle of needs solidarity clearly makes needs assessment a matterof concern. But unlike Sida, which also formulates policy regarding its de-cision and evaluation criteria, LFN is given the job of building practice byinterpreting the vague policies and principles set by legislators (ibid. p. 47).
This practice can then be modified – or reaffirmed – if and when decisionsare appealed to the administrative courts.
In summary, Sida and LFN take part in regulating the development aid andhealth care sectors, in so much that the two organizations activities consti-tute means for the Swedish state to intervene in economic activities andsteer them to realize public goals (cf. Christensen & Lægreid 2002). How-ever LFN is arguably more specialized, since it exercises control throughthe setting of rules about pharmaceutical usage. Sida, in contrast, is more ofa hybrid organization with a wider repertory of intervention mechanisms(cf. Lægreid et al. 2005). The seeming difference in organizational scope inone aspect which creates expectations of variation in how Sida and LFNpractice needs assessment. Additional matters are discussed further below.
4. Needs assessment in practiceIn this section, we will describe aspects of Sida’s and LFN’s needs assess-ment in practice. Firstly, we look at how needs can become subject to con-sideration by each organization. Secondly, we exemplify ways in which thetwo organizations determine what needs are relevant for them to meet.
Thirdly, we consider how the organizations assess and evaluate needs whendeciding on various forms of intervention. As previously mentioned, our examples are drawn from processes of decision-making in each organiza-tion where needs assessment is one of many dimensions under considera-tion.
4.1 What triggers needs assessment?As described earlier, the two agencies both assess needs as part of their re-spective decision-making processes. However the manner in which issuesare raised for consideration vary between LFN and Sida.
In the case of LFN, the matter of needs assessment is linked to decisions regarding the subsidization status of pharmaceuticals. There are three for-malized ways in which the issue of a pharmaceutical’s subsidization can beraised for consideration. Firstly, LFN must by law evaluate the status ofproducts that receive market approval by the MPA. EU rules requires LFNto make a decision within 180 days of receiving an application for subsidi-zation from the company marketing an MPA-approved drug (Council Di-rective 89/105/EEC). Secondly, LFN is tasked with evaluating all drugsthat were subsidized prior to the agency’s inception. The evaluations of theexisting product assortment are not initiated by the pharmaceuticals’ mar-keting companies but by LFN itself. There are no formal requirements forwhen the product assortment review must be completed, although the gov-ernment has previously indicated that five to six years would be desirable.
The first two groups – products for treating migraine and stomach acid dis-orders – were completed in February 2005 and January 2006, respectively.
There is an order in which the remaining forty-seven groups of products areto be evaluated, however there is no set timeline. Finally, LFN has the dis-cretionary right to bring the matter of a pharmaceutical’s subsidization un-der consideration at such a time as it desires (10 § SFS 2002:160).
For Sida, matters involving needs assessment are raised in less structured process and through many more channels than for LFN. Requests or pro-ject/programme proposals may come from governments, organizations orindividuals. Requests could be formal or informal; there is no standardformat for a request and no formal requirements regarding its contents.23An important principle, however, is that proposals should originate fromexternal parties and not from Sida's own staff. The underlying reason forSida’s profile as a ‘responding organization’ is the current view that recipi- 23 For certain cooperation forms, formal request requirements have been established, however.
For example, only non-governmental, non-profit-making organizations can apply for Sida grantsfor humanitarian assistance.
ent country actors should, as far as possible, be responsible for their owndevelopment efforts. It is assumed that such delegation of responsibility re-quires that donors pay close attention to the ‘real’ demand for assistance.24In practice, however, there are instances where Sida more actively makessuggestions for new contributions. For example, the reassessment of largerprogrammes (which takes place every two to five years) offers possibilitiesfor Sida-initiated changes in scale and scope of interventions, which mayalso include the adding (or subtracting) of certain program components.
Sida provides and prepares many different forms of support,25 and prepa- rations vary in time and ambition depending on the size and character ofthe project/programme. In cases of large grants or complex considerations,the preparation phase normally stretches over several months (sometimeseven years) and involves consultations with several internal and externalparties. For contributions within the frame of humanitarian assistance, onthe other hand, a special regulation applies which enables decisions to betaken swiftly and without much involvement from external parties or otherSida units. The width and depth of the information provided in assessmentmemos varies accordingly. While the preparation of long-term support maypermit the contracting of consultants for extensive feasibility studies, acutehumanitarian crises often require decision-making on the basis of vagueand changeable accounts of the actual needs. In such situations, Sida isauthorized to grant untied resources to ‘reliable and experienced’ partnerswith whom they have special agreements.26 In other words, the status of animplementing organization may substitute for an ‘accurate’ needs assess-ment when the major concern is to gain time (see further discussion in nextsection).
To summarize, LFN’s needs assessment is triggered by a few defined and formally structured means. There are only certain actors and certainsequence of events, notably the market approval by the MPA or the orderof the product assortment view, which start the evaluation and decision-making process. LFN does not primarily deal directly with those who po- 24 At present, geographic closeness seems to be viewed as a proxy for insight into needs for de-velopment aid. For an overview of how development thinking in this area has shifted over time,see Krohwinkel-Karlsson (2005).
25 The main cooperation forms are: support under country/regional strategies (pro-jects/programmes); support to economic reforms and debt relief; support via Swedish NGOs;humanitarian assistance; research cooperation; and contract-financed technical cooperation.
Sida also handles concessionary credits, soft loans and guarantees.
26 So-called Framework Agreements for Minor Humanitarian Assistance.
tentially have needs (such as patient organizations) or those who directlyinteract with those with needs (such as medical practitioners). And sinceLFN’s work is triggered by the matter of a drug’s subsidization status, onecould describe that the organization’s needs assessment is driven by a sup-ply of possible needs fulfilment. In comparison, Sida’s needs assessment iscommonly triggered by occurrences that generate demand for needs fulfil-ment. It presupposes a concrete expression of this demand in the form of anexternal project proposal. Though there are many different routes throughwhich a financial intervention may become subject to evaluation, who initi-ates is an important factor for determining the legitimacy of needs at Sida.
4.2 What are relevant needs?Once a matter has been made subject of consideration by the respectiveagencies, the question arises about what are to be considered relevantneeds. Within development assistance, the question of how needs should bemet is fundamental, and a frequent topic of debate. The discussions aboutneeds assessment and fulfilment are however complicated by the fact thatthe concept of needs has multiple meanings. Darcy and Hofmann (1999)found that it is used in at least three different senses:1. To define what constitutes needs (e.g., ‘food is a basic need’). This is a matter of generic definition which indicates the ‘level of ambition’ ofaid providers, and to some extent also provides a guideline for the pre-ferred modes of intervention.
2. To describe a lack of the above (‘these people need food’). This is a situation-specific account which should preferably take the perceptionsof those ‘in need’ into account (as to create a matching between needsand demands).
3. To argue for the need for intervention by the aid community (‘these people need food aid’). This is essentially a political question whichasks in what situations aid should be considered a suitable instrument toprovide lacking resources. However, it also involves more practical con-siderations, notably the potential of those in need to deal with and im-prove their own situation.27 27 These matters may be analyzed at different levels of aggregation, from the national level (me-aning that countries that are ‘too wealthy’ in terms of GDP/capita are normally not eligible forsupport) and down to the individual level.
LFN faces similar challenges in defining what are to be consideredhealthcare needs. Should, for example, needs to be determined through tar-get groups’ deviation from states of perfect health? What considerationshould be given to general deterioration in health caused by aging? Another perceived similarity between the development aid and healthcare sectors is that the need for interventions will always outstriptheir supply. Therefore what constitutes ‘relevant needs’ for Sida and LFNis in part also determined by which needs can be met by each agency (cf.
Fernler 2004). Needs can and do exist in excess of what the two agenciescan impact and they consequently fall outside of needs assessment. Onecritical delimitation of relevant needs in the case of LFN is that the organi-zation only considers pharmaceuticals’ subsidization, even though drug-based treatments are only one of several possible forms of medical inter-vention.
For Sida, what constitutes relevant needs is less clear-cut. For example, the concept of ’basic needs’, which emerged in the 1970s in reaction togrowth-focused approaches to development, is commonly understood to in-clude certain minimum requirements of a family for private consumptionsuch as adequate food, shelter and clothing, as well as certain householdequipment and furniture. They also include essential services provided forby the community at large, such as safe drinking water, sanitation, publictransport and health, educational and cultural facilities’ (cf. Singh 1979).
While most donors would agree on this definition, they have more recentlyalso introduced more complex considerations of needs, inspired e.g. by thetheories of Amartya Sen which stress the importance of social support andgenuine democracy for sustainable poverty reduction. Needless to say, themeeting of needs of the latter kind requires interventions of a totally differ-ent type than ‘traditional’ service provision.
The definition of target groups for intervention is one way to qualify ‘relevant needs’. In the case of LFN, relevant needs are (simply put) thoseof individuals covered by the Swedish healthcare system. In the case ofSida, it is not feasible to evaluate the needs of the whole relevant popula-tion (i.e. the inhabitants of the developing world). Instead, Sida concen-trates on certain subgroups, the selection of which is influenced by politicaland competitive reasoning and constrained by financial and administrativefactors (see next section). The difference in how target groups are identi-fied influences what the two organizations consider to be ’reliable’ infor-mation about needs: For Sida, needs assessment is in part contingent on and legitimized through the direct involvement of (representatives of) those ‘inneed’. In contrast, LFN organizes representatives for generic interestgroups within the healthcare sector into the decision-making process andperforms calculations in order to mediate specific needs.28 One importantreason for this is that LFN’s mandate is to consider all citizens needs fordifferent healthcare. The needs of any one specific patient population aresubject to consideration in light of needs within the entire existing and po-tential patient population.
In summary, LFN and Sida’s delineation of needs is not primarily based on the specification of needs in general. Notably, the relevant needs in Sidaand LFN’s respective assessments are those where it is feasible for theagency to intervene. However the two organizations have different ap-proaches for gaining information about target groups’ needs. While bothagencies organize those ‘in need’, they maintain different levels of close-ness. Whereas Sida’s work is legitimized by close interaction, LFN main-tains distance in order to consider ‘society’s needs’ for drug-basedhealthcare intervention. Yet the specification of what constitutes relevantneeds and how information about needs is sought does not offer much ex-planation as to how the assessment of need for specific intervention is per-formed. This will be the topic of the proceeding section.
4.3 How are needs for intervention determined?Evaluating whether specific needs are to be met is done using a variety ofdifferent techniques and organizational routines.
In the case of Sida, the most obvious delimitation of what constitutes ‘feasible’ needs is financial and made through the annual assignment ofspecific budget posts to specific countries or regions. In addition to needs-based concerns, it is commonly held that a number of extraneous factors(notably foreign policy and domestic political interests) have an influenceon inter-country prioritizations.29 A second delimitation is administrative in nature: For capacity reasons, Sida can only prepare and monitor a limited number of contributions at thesame time. ‘There is a dividing line when quality control is made difficult 28 As previously mentioned, interest representation is evident in the composition of LFN’sboard. It includes medical specialists, clinical practitioners, health economists and individualswith extensive experience from county councils and patient organizations.
29 For an empirical review of aid allocation rationales by different donors, see e.g. Alesina andDollar (2000).
because of Sida’s internal capacity’ (interview Sida employee C 2005-05-10). Hence, the administrative constraint tends to create a ‘natural’ thresh-old of needs fulfilment also in cases where there are no pre-defined finan-cial limits (such as humanitarian crises).
Thirdly, Sida’s interventions are based on judgments about levels of un- met needs in a given situation. This means taking into account not only theavailable resources in the recipient country, but also the expected actionsby other donors in the same geographical area or sector. The latter is im-portant for reasons of ‘comparative advantage’: while Sida and other do-nors commonly agree on needs, they tend to profile their interventions byhighlighting different consequences of needs fulfilment. Sida will considerthe ability of their organization to have an impact in a particular context,and more generally how to achieve maximum impact with the expertise andcapacity available to them. For example, Sida has emphasized the envi-ronmental aspects of the recent tsunami disaster in South East Asia whereasmany other donors have focused relatively more on the rapid resumption ofeconomic activities (Interview Sida employee A 2005-02-22).
Fourthly, Sida is dependant on actors’ interest in aid implementation.
That is to say, a prerequisite for intervention is that there is a timely supplyof ‘appropriate’ project proposals that fit with the above criteria for priori-tization (especially the third). Hence, Sida’s needs assessment is partybased on what is considered a ‘proper’ distribution of funds between differ-ent applying organizations. For example, the UN system is ‘frequently re-ceiving special treatment because of its intrinsic value’ (interview Sida em-ployee A 2005-02-22).
On the other hand, Sida’s sequential evaluation of potential interventions means that a specific intervention (and attendant needs) is seldom directlyassessed with comparable forms of intervention to the same target group, orwith other aid projects. Rather Sida tends to consider needs in trying toforesee the effects of potential interventions. The agency operates with aprinciple of ‘do no harm’, making it important that there is no injustice inallocations within a target group: Even in a humanitarian context, one has to realize that one is not working in a po-litical vacuum. When distributing food, for example, you have to make sure not tooverlook some areas because of their political affiliation, which might aggravate aconflict in both the short and long run. [….] At the same time, [humanitarian as-sistance] should be independent, but it is not entirely neutral [….] Neutral doesn’tmean that you have to deliver just as much to everybody but that the factors influ-encing (allocation) should be rational, so to say. It shouldn’t be that we like the ‘Tigers’30 better and they therefore get more. But you mustn’t read into it thatthere are no effects – it’s not that kind of political neutrality (interview Sida em-ployee B 2005-04-05).
LFN’s assessment of needs is in part contingent on similar concerns. How-ever, unlike Sida, LFN is not subject to budgetary restrictions. Thereforethe organization can, at least formally, disregard the financial impact of itsdecisions. In fact, the agency can potentially make decisions that increasecosts due to an increased prescription of cost-effective drugs (see discus-sion in Lundin 2004).
And while LFN faces similar problems with administrative capacity, this limiting factor is not linked to needs assessment per se since it is alreadyformalized which matters should be made subject to LFN’s consideration(as discussed above). What seems more important when determining therelevance of meeting specific needs is, for one, the cost of not meetingneeds (LFN 2005a).31 There is an inferred link between the cost of non-treatment and the urgency of a condition. So, in line with the aforemen-tioned principle of needs solidarity, those with more costly unmet needshave the more urgent conditions that motivate financial intervention: A ‘cheap’ drug which has an acknowledged effect on [a symptomatic stomach ill-ness] may still not be included in the pharmaceutical benefit simply because thematter is not urgent enough to warrant public subsidization. We have both agreater cost tolerance and interest in subsidizing drugs which treat conditions withhigh mortality or morbidity (interview LFN bureau employee 2005-05-24).
How the matters of cost and urgency are determined is in part contingenton whether it is a new or old pharmaceutical being evaluated. In contrast toSida, where it is less common to directly compare projects (and needs) witheach other, LFN systematically evaluates specific pharmaceuticals in rela-tion to other drugs.32 In the case of new drugs, there is an incremental com-parison of each individual pharmaceutical’s effect and cost with drugs forsame therapeutic usage (if there are any). A product with a higher price willnot be granted subsidization (e.g. LFN 2003d). In the case of the productassortment review, there is a concurrent analysis of all drugs. In the latter 30 The ‘Tamil Tigers’ is a common name for the Sri Lankan separatist movement.
31 LFN’s sensitivity to the cost of unmet needs follows from the notion that the cost and benefitsof treatments are both societal in nature, i.e. that non-treatment is an alternative cost for societycompared to the financing of treatment through the public pharmaceutical benefit.
32 This comparisons is in part sought through the use of shared metrics for measuring cost-effectiveness which are removed from the medical condition for which specific drugs are used(see discussion in Sjögren and Helgesson (2004)).
case, there is no referent. So while existing subsidization of pharmaceuti-cals within the same therapy group can be seen to imply that a drug is usedfor treating relevant needs, this is no guarantee. New products have beenincluded in expectation of a coming product assortment review whichmight bring further subsidization into question (e.g. LFN 2004b).
In further contrast to Sida, LFN also considers urgency of needs in rela- tion to the target group’s own responsibility for meeting needs. For LFN,some needs are not deemed the responsibility of the public healthcare sys-tem (LFN 2003b; LFN 2004c). In short: the feasibility of an interventiondoes not make LFN responsible for this intervention and LFN can and doesdeny pharmaceuticals’ subsidization (e.g. LFN 2004a). That LFN explicitlydecides not to meet certain needs is another difference compared to Sida.
Whereas the latter agency decides not to bring certain matters under con-sideration (and therefore, by default, does not satisfy certain needs), LFNcannot avoid evaluations of pharmaceuticals’ subsidization status. How-ever, in addition to saying no, LFN has the option of delegating needs as-sessment to health care practitioners (most commonly physicians). This oc-curs when the agency grants unrestricted or restricted subsidization. Inthese cases, it becomes the task for the individual doctor to determinewhether a specific patient should be prescribed a certain drug with subsidi-zation.33 This delegation of responsibility for needs assessment is some-what different compared to the aforementioned discussion regarding Sidaand other donor agencies division of labour based on ‘comparative advan-tages’ in needs fulfilment.
In sum, we see that the principle of ’neutral’ needs assessment is not ap- plicable to LFN and Sida’s decisions on whether or not to fund specific in-terventions. Notably, in the case of LFN, certain needs are assessed in rela-tion to others (albeit in different ways). This also involves a considerationof who is responsible for meeting needs: the public (through the pharma-ceutical benefit) or the target group (out of their own pocket). In the case ofSida, specific needs assessment is contingent on other donors’ activities(and indirectly on the capacity of target countries to adequately manage 33 In contrast, a decision to deny all subsidization of a drug removes this local evaluation sincethe restriction is enforced through the National Corporation of Swedish Pharmacies’ productdatabase.
their own needs fulfilment).34 Needs assessment is therefore for both agen-cies not only a matter of ‘can do’ but also ‘how (and who) do’.
5. DiscussionIn this paper, we chose to compare two organizations that share the sameformal status as Swedish governmental agencies and that apply a commonprinciple as one basis for financial allocation, that of needs assessment. Ouraccount shows that the organizations are dissimilar in several ways with re-spect to how they practice what they perceive to be neutral needs assess-ment.
We see the differences in intervention mechanisms as one possible ex- planation for variations in practice. For, as noted earlier, the organizationshave different formal mandates and means for exercising influence. LFN’stask and organizational tools relate to the practice of regulation in a morenarrow sense, i.e. by making decisions and thereby setting rules for howothers should make subsequent decisions regarding treatment choice. Theorganization also targets a more narrowly defined group with needs. This isa consequence not only of the fact that the matters considered by LFN are’national’ in character while those of Sida are international, but also relatesto how issues are raised for consideration at the two agencies. LFN’s com-paratively narrow focus, both with respect to its target and its means of ex-ercising influence, arguably makes it possible for LFN to be more struc-tured, and seeming more rational, in its activities than Sida. The latter or-ganization is not able to disengage from its environment in the same way,both as a consequence of its fiscal responsibility and because needs as-sessment and decisions based on such assessments is only one part of itsportfolio of activities. This makes for a more eclectic process which, incombination with the organization’s financial responsibility, also offersSida a wider range of means for intervention and more powerful mecha-nisms for forcing compliance (for example through monitoring of results,conditioning of further funding and the threat of exit in the case of inade-quate use of resources).
In addition to considering intervention mechanisms, our study also sug- gests the relevance of dimensions that cut across both formal structure and 34 Following the recent tsunami disaster in South East Asia, Indonesia and Sri Lanka were sub-jects of significant financial intervention whereas Thailand was not considered a potential reci-pient of aid due to the relatively strong financial position of the country and the perceived abi-lity of the Thai government to meet the needs of its own population.
tasks.35 Below, we will broadly consider the impact of organizational set-ting and process logic on the practices of needs assessment in Sida andLFN. By organizational setting, we refer to how the respective organiza-tions delineate and relate to their environment. Process logic, in turn, refersto the timing of and impetus for needs assessment in the two agencies. Wewill end with a short comment on how Sida and LFN's regulating practicescan be seen as means of bridging these dimensions of space and time.
5.1 Organizational settingWhile the development aid and healthcare sectors are populated by differ-ent actors, the two studied agencies share a dependence on others whenmaking needs assessments and implementing subsequent decisions regard-ing financial intervention to fulfil needs.
In performing needs assessment in practice, both Sida and LFN constrain their activities to those needs that can be feasibly met through the agencies’respective activities. Needs can fall outside of the organizations’ perceivedregulatory mandate due to administrative and budgetary limitations (Sida)or formal restrictions in scope of operations (LFN). The relevant needs arealso defined by the organization’s target groups. These differ with regard tosize and stability. Sida’s potential target group is significantly larger thanLFN’s. And while Sida’s actual target group at any one time is smaller, it isalso subject to change over time (due to aforementioned aspects such ascatastrophic occurrences and political priorities). LFN has a finite numberof potentially ‘needy’, i.e. those covered by Swedish healthcare system.
However, since LFN is tasked with considering the systematic needs of thispopulation, the organization maintains an arms-length relationship with anyspecific part of the target group. LFN also more clearly views the targetgroup as potentially responsible for certain needs fulfilment. Sida, in con-trast, tends to organize sub(target)-groups closely into their needs assess-ment and decision-making processes – and not explicitly consider the re-sponsibility of the target group for meeting needs (once their needs havebeen made the subject of consideration).
With regard to needs fulfilment, Sida is dependent on a shifting number of actors which supply the agency with both targets for and means of inter-vention. While there are certain actors which have a special ‘trusted’ status, 35 Earlier studies of NPM-influenced practices have also suggested the relevance of broaderfactors than specific processes within individual organizations (cf. discussion of dangers ofcosting myopia by Lewis and Stiles (2004)).
there are many means through which projects enter Sida and are imple-mented. As with the case of target groups, LFN has a more fixed environ-ment comprised of the county councils and health care practitioners. How-ever due to the organization’s lack of budgetary responsibility and ‘fieldpresence’, the level of dependency is significant. The agency has no meansof forcing compliance to decisions (unless it denies subsidization, as de-scribed earlier).
In summary, the organizational setting for LFN is more stable and structured than for Sida. Furthermore, LFN maintains a greater distance toits environment due to its avoidance of close contact with ‘special interests’and lack of budgetary clout.
5.2 Process logicAs described earlier, there are many ways in which a matter of needs as-sessment (and fulfilment) can be raised in Sida. And when assessing needs,the agency seldom directly compares various forms of intervention for thesame target group. Evaluations (especially concerning the more explicitly‘needs-oriented’ humanitarian aid) occur sequentially, as occurrences trig-ger requests for aid. Many needs and interventions are never made the ob-ject Sida evaluation. There are other donors who might provide aid in caseswhere Sida does not, and there are forms of intervention where Sida is per-ceived to have ’comparative advantages’. So Sida seldom explicitly says noto intervention. However its needs assessment – and fulfilment – is contin-gent on there being a suitable supply of implementation projects.
LFN, in contrast to Sida, has a more clearly defined form and scope of intervention. There are fewer, and explicitly formalized, ways in whichmatters are brought under consideration by the agency and LFN does notmake a selection of issues based on demands for intervention. Needs as-sessment is instead triggered by the supply of needs fulfilment (in the formof pharmaceuticals). Furthermore, the mandate of the agency (to considerspecific needs in relation to ‘societal needs’ for specific intervention) im-plies direct comparisons of needs for intervention. This, combined with thefact that LFN also has fewer and more clear-cut choice options than Sida(there are only three possible decision outcomes), means that the organiza-tion can and does explicitly say no to needs fulfilment/intervention whichhas been brought under consideration.36 36 That LFN is able to say no does not mean it is easy to say no. As discussed earlier, LFN is ta-sked with taking a ‘societal view’ of needs and recognizes its target group as potentially respon- The proceeding section will end with a discussion of how Sida and LFN practices as regulators can be understood in light of their respective organi-zational settings and process logics.
5.3 Sida and LFN: management by rules or by organizationSida practices a broader form of regulation than LFN, using various orga-nizational and financial elements. In this way, Sida bridges a dynamic or-ganizational setting and unstructured process logic by organizing actors(e.g. through written agreements) and allocating resources, in particular fi-nancial resources but also other organizational resources such as employ-ees. It follows that the conclusions of Sida’s needs assessments, as well asthe means of subsequent needs fulfilment, might be viewed as ongoingprocesses, rather than clearly delineated outcomes (see further discussion inKrohwinkel-Karlsson 2005).
In contrast, LFN seemingly regulates more through outcomes, i.e. ex- plicit decisions regarding pharmaceutical subsidization status. The relativestability of the agency’s setting and process, where the responsibility forneeds assessment and needs fulfilment is more clearly separated (in linewith Christensen and Lægreid (2005)), supports LFN acting ‘at a distance’(Blomgren & Sahlin-Andersson 2004). Unlike Sida, LFN more clearlymanages by setting rules.
In conclusion, we note that neither organization realizes the ambition of in-dependent needs assessment as broadly articulated within an NPM frame-work. Though this is not surprising per se, it is relevant to observe the dif-ferent ways in which environmental or organizational restrictions on re-sources (notably time and money) and command of resources gives rise tovarious forms of prioritization and attendant regulation. We believe there ispotential in further exploring such processes in greater detail.
sible for their own needs fulfillment. However, LFN is arguably held more responsible for needsfulfillment than Sida since the matter of pharmaceutical subsidization is viewed as more of aright by the agency’s target group than development aid.
6. ReferencesAlesina, A. and Dollar, D., 2000, Who Gives Foreign Aid to Whom and Why? Journal of Economic Growth. 5: 33-63.
Blomgren. M. and Sahlin-Andersson, K., 2004, Ledning på distans — Att skapa kun- skap för politisk styrning av hälso- och sjukvård. Stockholm: Landstingsförbun-det.
Boyne, G., Powell, M. and Ashworth, R., 2001, Spatial Equity and Public Services: An empirical analysis of local government finance in England. Public ManagementReview, 3 (1): 19-34.
Bradshaw, J., 1971, A Taxonomy of Social Need. In McLachlan, G. (ed.) Problems and Progress in Medical Care. Oxford: Oxford University Press.
Brunsson, N., 1989, The Organization of Hypocrisy: Talk, Decisions and Action in Or- ganizations. Chichester: Wiley.
Brunsson, N., 1995, Ideas and Actions: Justification and Hypocrisy as Alternatives to Control. Research in the Sociology of Organizations. 13: 211-235.
Campell, A. K., 1976, Approaches to Defining, Measuring and Achieving Equity in the Public Sector. Public Administration Review. 36 (4): 556-62.
Christensen, T. and Lægreid, P., 2002, New Public Management: The transformation of ideas and practice. Aldershot: Ashgate.
Christensen, T. and Lægreid, P., 2005, Agencification and Regulatory Reforms. Paper prepared for the SCANCOR/SOG workshop on Autonomization of the state:From integrated administrative models to single purpose organizations, StanfordUniversity, 1-2 April 2005.
Clark, D., 2004, Implementing the third way: Modernizing governance and public services in Quebec and the UK. Public Management Review. 6 (4): 493-510.
Czarniawska, B. and Joerges, B., 1996, Travels of Ideas. In Czarniawska, B. and Sevón, G. (eds) Translating Organizational Change. Berlin: de Gruyter.
Darcy, J. and Hofmann, C.-A., 1999, According to need? Needs assessment and deci- sion-making in the humanitarian sector. HPG Report 15. London: Overseas De-velopment Institute.
Fernler, K., 1996, Mångfald eller likriktning. Effekter av en avreglering. Stockholm: Fernler, K., 2004, Konsten att fastställa behov. In Helgesson, C.-F., Liljenberg, A and Kjellberg, H. (eds) Den där marknaden: om utbyten, normer och bilder. Lund:Studentlitteratur.
Hood, C., 1995, The ‘New Public Management’ in the 1980s: Variations on a Theme.
Accounting, Organizations and Society. 20: 93-190.
Krohwinkel-Karlsson, A., 2005, Doing Good or Doing Well? Organizing international development aid in pursuit of sustainability. IIB Working Paper Series 01/05.
Stockholm School of Economics/IIB.
Lægreid, P., Roness, P. G. and Rubecksen, K., 2005, Regulating Regulatory Organiza- tions: Controlling Norwegian Civil Service Organizations. Working Paper No 5.
Stein Rokkan Centre for Social Studies. Bergen: Unifob AS.
Lewis, B. R. and Stiles, D. R., 2004, How invisible are the Emperor’s new clothes? Transparent costing, cross-subsidization and costing myopia in higher education.
Public Management Review. 6 (4): 453-72.
Lundin, D., 2004, Kostnadseffektivitet som kriterium för subvention av läkemedel – en bra idé? Ekonomisk debatt. 6: 32-40.
MacGowan, F. and Wallace, H., 1996, Towards a European Regulatory State. Journal of European Public Policy. 3 (4): 560-576.
Majone, G., 1994, The Rise of the Regulatory State in Europe. West European Politics, Minzberg, H., 1979, The structuring of organizations. Englewood Cliffs; NJ: Prentice- Mohr, J. W., 2005, The Discourses of Welfare and Welfare Reform. In Jacobs, M. and Weiss Hanrahan, N. (eds) The Blackwell Companion to the Sociology of Culture.
Oxford: Blackwell Publishers.
Pollitt, C., 1993, Managerialism and the Public Services: Cuts or Cultural Change? Pollitt, C., Talbot, C., Caulfield, J. and Smullen, A., 2004, Agencies. How Governments Do Things through Semi-Autonomous Organizations. London: Palgrave.
Pollitt, C., 2005, Performance Management in Practice: A Comparative Study of Ex- ecutive Agencies. Paper prepared for the SCANCOR/SOG workshop on Autono-mization of the state: From integrated administrative models to single purposeorganizations, Stanford University, 1--2 April 2005.
Premfors, R., Ehn, P., Haldén, E. and Sundström, G., 2003, Demokrati och byråkrati.
Scott, W. R. and Meyer, J. W. (eds), 1983, Organizational Environments: Ritual and Rationality. Beverly Hills, CA: Sage.
Singh, A., 1979, The “Basic Needs” Approach to Development vs. the New Interna- tional Economic Order: The Significance of Third World Industrialization. WorldDevelopment. 7: 585-606.
Sjögren, E. and Helgesson, C-F., 2004, The Q(u)ALYfying hand: Health economics and medicine in the shaping of Swedish markets for subsidised pharmaceuticals. Paperprepared for the Society for Social Studies of Science Conference (4S), Paris, 25--28 August 2004.
Stone, D. A., 1997, Policy Paradox: the art of political decision making. New York: Svensson, A. and Holmgren, W., 2003, Sida’s Control Environment – a Feasibility Study. Internal Audit 04/01. Stockholm: Sida.
Whitley, R., 1988, The management sciences and management skills. Organization 6.1 Public materialCouncil Directive 89/105/EEC relating to the transparency of measures regulating the pricing of medicinal products for human use and their inclusion in the scope ofnational health insurance systems. Official Journal L 040, 11/02/1989 P. 0008 –0011.
Läkemedelsförmånsnämnden (LFN), 2002, Arbetsordning. Dnr 412/2002.
Läkemedelsförmånsnämnden (LFN), 2003a, Arbetsplan för den inledande fasen av ge- nomgången av läkemedelssortimentet. Dnr 1023/2003.
Läkemedelsförmånsnämnden (LFN), 2003b, Beslut om Viagra. 2003-03-26.
Läkemedelsförmånsnämnden (LFN), 2003c, Beslut om Cialis. 2003-05-14.
Läkemedelsförmånsnämnden (LFN), 2003d, Beslut om Totelle. 2003-11-06.
Läkemedelsförmånsnämnden (LFN), 2004a, Beslut om Elidel. 2004-05-28.
Läkemedelsförmånsnämnden (LFN), 2004b, Beslut om Selexid. 2004-05-03.
Läkemedelsförmånsnämnden (LFN), 2004c, Beslut om Levitra. 2004-04-06.
Läkemedelsförmånsnämnden (LFN), 2005a, Slutrapport: Genomgången av läkemedel Gov. bill 2001/02:63. De nya läkemedelsförmånerna m.m.
Gov. bill 2002/03:122. Shared Responsibility: Sweden’s Policy for Global Develop- SFS 1982:763. The Health and Medical Service Act.
SFS 2002:160. Lagen om läkemedelsförmånerna m.m.
SFS 2002:719. Förordning med instruktion för Läkemedelsförmånsnämnden.
Skr. 2004/05:52. Regeringens politik för humanitärt bistånd.
Socialstyrelsen, 2002, Hälso- och sjukvårdsstatistisk årsbok 2002. Statistik Hälso- och Socialstyrelsen, 2004, Statistik över kostnader för hälso- och sjukvården 2003. Statistik Socialstyrelsen, 2005, Läkemedelsförsäljningen i Sverige - analys och prognos. Uppföljning och utvärdering 2005-103-1.
6.2 InterviewsLFN bureau employee, 2005-05-25.
Sida employee A, 2005-02-22.
Sida employee B, 2005-04-05.
Sida employee C, 2005-05-10.

Source: http://www.score.su.se/polopoly_fs/1.26587.1320939799!/20061.pdf