ARTICLE ECONOMIC STRATEGIES, ADJUSTMENT AND HEALTH POLICY: ISSUES IN SUB-SAHARAN AFRICA FOR THE 1990s1 Rob Davies and David Sanders Introduction For sub-Saharan Africa (SSA) the 1980s were characterised by economic stagnation, crisis and adjustment. The relationship between this economic crisis and the health situation is complex and demands an understanding of the economic and social context of ill-health internationally and more specifically in Less Developed Countries (LDCs). The disease burden and pattern ex- perienced by the peoples of SSA are strikingly similar to those of nineteenth century Europe; they are primarily diseases of underdevelopment and poverty, not a feature of warm climates, ie the tropics (Sanders, 1985). Insofar as countries have industrialised and urbanised, so those sections of the society experience disease patterns more akin to those dominant in the industrialised countries. Historical and contemporary experiences have shown that there is a definite but complex relationship between economic growth on the one hand and health status on the other. In general, sustained economic growth over the long run does lead to improved health and nutritional status: in the now-industrialised countries the large and sustained decline in mortality was accompanied by reductions in morbidity (disease) and malnutrition and largely preceded any effective medical interventions (Sanders, 1985). There is not, however, a direct correlation be- tween health and nutrition indicators and GDP/capita because improved income distribution - even at low levels - can accelerate the improvements in health (eg China, Sri Lanka) (Halstead, 1986). In the snort-term, the inter-relationship is even more complex. There are examples of countries in which high growth has been associated with a decline in health status as reflected by the normal indicators (Brazil), but there are equally cases where severe economic decline has been associated with significant improvements in health status (Chile, Tanzania). An understanding of therelationshiprequiresa fairly detailed study of the particular circumstances in which the economic changes were taking place and of the context within which health status is determined. In particular issues of access and equity are important. 78 TRANSFORMATION 21 (1993) DAVIES AND SANDERS ARTICLE In this paper we intend raising some of the issues of the relationship between economic performance and health in SSA. Our observations are strongly in- fluenced by Zimbabwe's post-independence experience which is the case we have observed most closely. Although there are features of Zimbabwe which make it somewhat special, it provides the most relevant lessons for post-apart- heid South Africa. More generally, the lesson we draw from the Zimbabwean case is that health sector interventions result in limited and fragile gains, and, from a brief appraisal of performance in LDCs, that more thoroughgoing structural changes are required for lasting improvements in health. We then consider how fashions in policies - macroeconomic and health - have evolved and interact. We insist that progressive state policies, rather than adjustment towards laissez faire, are required to initiate the structural changes required. Finally, we argue that democratisation and empowerment are crucial in this process, and that there is thus an urgent need to reinstate a comprehensive Primary Health Care approach. The Economic and Health Situation in Sub-Saharan Africa Although the economic experience of SSA in the 1980s was quite varied, the decade was one of economic crisis. The selected indices given in Table I suggest that SSA performed badly both in absolute terms and relative to the two previous decades. Systematic and comparable data on trends in health and nutrition are difficult to obtain. However, it does appear that SSA's performance in the 1980s was in general poor. Table I also suggests that rates of improvement in such indicators as crude death rates, infant mortality rates and calorie intakes all slowed down in the 1980s. 2 In the 30 years between 1950 and 1980 Africa's Under 5 Mortality Rate (U5MR) was reduced by 1.83% per year, from 332 to 191 per thousand. This rate of reduction slowed down significantly in the 1980s and was consid- erably below the average rate of reduction for all developing countries in the world. In 1950 Africa had the third highest U5MR, being slightly lower than 334 for West Asia and 344 for Southern Asia (the world's highest rate). By 1980 it had the second worst rate of 191 compared with 118 for West Asia and 200 for Southern. By 1987, however, Africa's U5MR of 172 per thousand was not only the worst but lagged significantly behind that of West Asia (99) and even Southern Asia (158). It can be seen that the rate of reduction in Africa has been slower than in other regions since the 1950s or before. The rate of reduction was low even in the period of relatively high growth in the 1950s and 1960s and worsened in the decade of recession and stagnation. TRANSFORMATION 21 (1993) 79 ARTICLE DAVIES AND SANDERS TABLE 1: SELECTED ECONOMIC INDICATORS FOR SUB-SAHARAN AFRICA 1965-73 1973-80 1980-87 Average Annual Growth Rates(%) GNP per capita 2.9 0.1 -2.8 GDP 5.9 2.5 0.5 Agriculture 2.2 -0.3 1.3 Manufacturing 10.1 8.2 0.6 Private Consumption 3.9 2.6 1.1 Gross Domestic Investment 9.8 4.0 -8.2 Exports 15.1 0.2 -1.3 Imports 3.7 7.6 -5.8 1970 1980 1988 External Public Debt as % of GNP 13 21 78 Debt Service as % of Exports 5.3 7.2 16.5 1965 1980 1987 Crude Death Rate per 1000 23 18 16 Daily calorie supply per capita 2092 2152 2097 Urban Population as % of total 14 22 27 Infant mortality rate per 1000 160 120* 108 * 1983 Source: World Bank (1989); World Bank (1990) Note: These figures are given here purely for illustrative purposes. They give a rough and somewhat inaccurate picture particularly since they are largely weighted means including Nigeria. The Changing Policy Environment The economic experience of SSA in the 1980s has led to the evolution of a fairly wide consensus on economic policy. It is agreed, even amongst those who might disagree as to the solution, that many of SS A's economic problems have their origin in policy mismanagement Many of the ann'-dirigiste arguments put forward by Krueger, Lai and others in the 1970s have, perhaps in a softened form, become conventional wisdom in the 1990s. There is thus, amongst economists, a general view that the state has to be made more efficient and, probably, reduced in size. 80 TRANSFORMATION 21 (1993) DAVIES AND SANDERS ARTICLE This consensus conflicts with the widely-held view amongst non-economists working in the health and welfare sectors that inequities in society require government intervention to redress them. Transfers of income through the budgetary process are both necessary and desirable. It may be agreed that government intervention in practice may not always have been as effective as it might; there may be debates as to the most appropriate form of intervention; and it is recognised that there are many different ways in which the state may be organised, some more decentralised than others. However, there is little question that the state is the central agency through which social problems can be adequately addressed.3 This view is similar to that emanating from Pigovian welfare economics but contrasts in many ways with that of more hard-headed modern economists who dominate policy debates on SSA. The general view seems to be that the growth and stabilisation problems are so pervasive that it is necessary to address the causes of these before (and even, if necessary, at the expense of) short-term concerns over welfare provision. Given its centrality in the complex of causal factors it is paramount, therefore, that the size of the state be reduced and the efficiency of what remains be improved. The reduction in direct public provision of services or increases in user charges which this almost inevitably entails is, it is argued, an unfortunate but necessary transitional cost of adjusting to a more sustainable long run growth path.4 The experiences with structural adjustment programmes in the 1980s have led to softening of the initial concentration solely on the stabilisation aspects - which were the main reason why countries approached the Bank and Fund for assistance in the first place. After the experience of Liberia (1979), Sudan and Tunisia (1985/86), Algeria (1988), Zambia (1987) and other countries, where implemen- tation of adjustment and liberalisation policies was followed by popular unrest, it has been recognised by these protagonists that even if protection of the poor is not a good thing in itself, reform programmes themselves can be jeopardised if some protection is not offered. The critical perspectives on adjustment con- tained in UNICEF's Adjustment with a Human Face (Cornia et al, 1987) and other alternatives proposed by international agencies have increasingly been taken on board by the Bank, through the Living Standards Measurement Study and subsequent Social Dimensions of Adjustment programme. In operational terms, Bank SAPs now tend to contain components such as PAMSCAD (the Programme to Mitigate the Social Costs of Adjustment) in Ghana, and the social costs of any of its programmes have to be monitored and reported on to the Bank President annually. Despite this evolution, it is still the case that the increasingly dominant approach to health and other welfare needs emphasizes private rather than public TRANSFORMATION 21 (1993) 81 ARTICLE DAVIES AND SANDERS provision. Indeed, the changing fashion in thinking on general development policy has itself influenced policy approaches within the health sector. In the 1960s and early 1970s, in line with development thinking which advocated such strategies as the Basic Needs Approach, and influenced by an assessment of past failures in the health sector and the incontestable advances made in a few low income countries (most especially China), previous strategies within the health sector were replaced by the Primary Health Care (PHC) approach. This advo- cated accessible, acceptable and affordable health care provision which ad- dressed common problems and was based on the active involvement of communities in its planning, implementation and evaluation. PHC was to be comprehensive, in that it integrated promotive, preventive, curative and rehabilitative functions of health services. Explicit recognition was given to the fundamental importance of social, economic and political factors in health efforts, most especially in the promotion of health: here community involvement was acknowledged to be central. During the late 1970s and 1980s, partly in response to the deteriorating economic environment and accompanying changes in development thinking, comprehensive PHC came into question as a sustainable option for the health sector. Instead, a more technical approach, which targets intervention on selected diseases and stresses the primacy of cost-saving by improved management and even cost recovery by the imposition of user charges, has come to dominate. This shift does not appear to be based on evidence that the PHC approach has not worked or that the selective intervention approach works any better. Rather, it parallels and reflects the shift in approaches to macroeconomic management. The Determinants of Health and Nutritional Status In order to improve health there has to be improved access to factors which determine health. A standard economics approach to identifying these factors is to specify a 'health production function' and to test empirically which variables are significant (Thomas, 1990). For policy purposes it is useful to categorise these factors into two broad groups: those originating outside and those originat- ing inside the health sector. Evidence from many countries shows that income is probably the most important of the 'outside' factors (Martorell and Habicht, 1986). For example, in Zimbabwe, Mazur and Sanders in a study of an urban working class com- munity found that variation in the three dimensions of children's nutritional status (weight for age, height for age and weight for height) 'was explained principally through consideration of the socioeconomic status of parents (educa- tion, economic activities, income and housing status)' (Mazur and Sanders, 1988:33). Since education and housing status are themselves strongly correlated 82 TRANSFORMATION 21 (1993) DAVIES AND SANDERS ARTICLE with income, this suggests that income is a primary determinant of nutritional status. Other factors originating outside the health sector include social inputs, such as education, environmental inputs, such as access to clean water, and general economic measures such as food rationing and subsidies, etc. Factors originating inside the health sector are the usual range of health care provision: for example, hospitals, health services, health manpower, immuniza- tion, etc. TABLE 2: ZIMBABWE INCOME PER CAPITA (Constant 1980 prices) 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 423 454 484 477 448 445 465 462 447 467 475 471 Source: CSO National Accounts Early Release (Harare, February 1991) Although health sector inputs may be the most obvious determinants, the effects of non-health sector inputs are probably more important. It is, however, difficult to quantify and measure the relative impact of these different sets of factors because of the complementarity between them and because there are often long lags between cause and effect, with intervening changes which are impossible to isolate. It is also relatively easy to achieve rapid improvements in health as measured by the standard quantitative indicators (which are in reality disease indicators), but sustained improvements in the quality of life are more difficult to produce and measure. For instance, certain indicators, such as infant and young child mortality rates, may be rapidly improved by selective primary health care interventions targeted at these high risk groups. There is however, little evidence to suggest that improved nutrition levels, for example, can be maintained by the application of this technical package in the absence of more general improve- ments in access to resources. The Zimbabwe 'Best Case' Case Study5 The highly unequal society which Zimbabwe inherited was reflected not only in wide income and wealth disparities but also by extreme differences in health experience. For example, the Infant Mortality Rate (IMR) for the black popula- tion as a whole was estimated to be 100-120 per thousand live births while it was 14 for the whites. Within the black population it varied between 40 for urban TRANSFORMATION 21 (1993) 83 ARTICLE DAVIES AND SANDERS TABLE 3: ZIMBABWE: TRENDS IN AGGREGATE EXPENDITURE 1979-87 Real rates of Growth in: Share in GDP(mp) of: C G I Y C G 1 X 1979 12.4 2.7 0.2 1.7 69.1 19.0 12.1 -0.2 1980 -0.9 10.1 19.2 10.7 64.4 19.7 18.8 -3.0 1981 17.2 16.8 36.7 9.7 64.8 19.7 23.7 -8.1 1982 -1.7 12.3 9.1 1.5 64.7 22.6 20.6 -7.9 1983 16.6 5.6 -2.9 -2.1 68.5 21.6 13.8 -3.9 1984 -24.4 8.3 -19.2 2.3 57.0 25.8 16.7 0.6 1985 -7.0 3.7 -18.3 7.4 55.0 25.4 18.2 1.3 1986 3.5 8.2 2.6 2.1 53.0 26.3 15.8 5.2 1987 -13.9 24.6 -3.9 -0.5 50.8 32.5 10.4 6.6 1988 na na na 7.3 na na na na 1989 na na na 4.5 na na na na 1990 na na na 1.9 na na na na Y = GDP(factor cost); C = Private Consumption I = Gross Capital Formation; G = nei government current expenditure; X = net exports of goods and services All figures based on 1980 constant prices Source: CSO National Accounts Early Release (Harare, February 1991) dwellers to 140 for rural (Ministry of Health, 1984). An estimated 21% of the under five population suffered second or third degree malnutrition based on weight for age, 28% with stunting and 9% with wasting (World Bank, 1983). In the decade since independence changes have been effected in this situation by policies and developments both outside and inside the health sector. Between 1979 and 1990 Zimbabwe's GDP rose by 51.5% in constant price terms, an average annual compound rate of 3.9%. This meant that GDPper capita rose over the decade. The growth performance was, however, very erratic, with high growth rates (in 1980,1982,1985 and at the end of the period) interspersed with low or negative rates. Although constant price GDP per capita had risen from $423 in 1979 to $471 by 1990, Table 2 shows the uneven nature of the growth, with significant advances up to 1982 being eroded in the mid- 1980s and •not quite fully restored in the late-1980s. In the early-1980s the government adopted expansionary economic policies - induced by its social commitments and permitted by the economic growth around 84 TRANSFORMATION 21 (1993) DAVIES AND SANDERS ARTICLE independence. It increased its expenditure on education and, to a lesser extent, health and other social services. It used minimum wage legislation to raise wage levels quite significantly. It improved rural credit facilities and, through subsidies to the Grain Marketing Board, improved marketing infrastructure significantly for peasant producers. However, growing balance of payments deficits led to the adoption of more restrictive stabilisation measures in 1983. Minimum wage legislation began to be used to curtail wage costs. An active exchange rate policy was adopted, although quantitative import controls remained the central instrument for current account management. Although it was recognised that the budget deficit had to be reduced, less action was taken in this sphere. Government attempted to sustain its education, health and other programmes, at least maintaining the levels of expenditure if not expanding them as fast as previously. Expenditure cuts were made, but these tended to be on capital expenditure rather than current It was thus difficult to avoid a large and growing budget deficit Simultaneously government adopted an extremely prudent approach to foreign debt using direct foreign exchange controls to bring the current account deficit into surplus. This combination of a high public sector deficit and low foreign savings can only be accommodated through a high private sector surplus. Table 3 shows graphically that this was achieved through compression of consumption and investment in the private sector (see Davies and Rattso, 1990). The compression was achieved by a combination of import controls, wage controls, price controls and financial crowding out. Through the 1980s it was recognised that lack of foreign exchange imposed a major constraint on growth, and various ad hoc measures were undertaken to stimulate exports while retaining the quantitative controls on imports. Although these programmes did achieve some success, pressure built up from various quarters for a more systematic revision of economic policy. With hindsight, the experiments with various ad hoc macroeconomic and trade policies can be seen as constituting an uneven-paced evolution towards the more systematic adjust- ment programme adopted in 1990 (see Shaw and Davies, 1991). This stagnation in real income was accompanied by a changing pattern of income distribution. Although peasant incomes had been raised by government policies in the early-1980s, there is evidence that this was not across the board but favoured approximately 20%; rural household differentiation appears to have been increasing. The failure of formal employment to grow, even in the boom times, has placed an increasing burden on other forms of economic survival, such as the informal sector and subsistence agriculture. Although the successive removal of food subsidies, aimed at reducing the impact on the budget deficit of losses by marketing boards, may have helped that small percentage of farmers TRANSFORMATION 21 (1993) 85 ARTICLE DAVIES AND SANDERS who are net producers of food, it has contributed substantially to the inflation which has eroded real earnings through the 1980s. In absolute terms there are more people below the poverty line now than there were in 1983 and probably in 1980. The position in relative terms is not as clear, but it seems likely that real incomes have fallen for the majority of the population since 1983. Insofar as education contributes to better health behaviour, Zimbabwe's phenomenal post-independence expansion in school enrolment is highly sig- nificant. Total numbers at school grew from about 900 000 in 1979 to nearly 3-million in 1989. The bulk of this expansion has been at primary level. Although this is likely to have a positive impact on health in the long term, it has imposed a financial burden on many parents, leading to reduced expenditure on other items, including food. Primary education has been free since independence but parents are responsible for uniforms, transport, building funds and labour inputs. These economic pressures have contributed to increasing dropout rates, with 5% of pupils dropping out each year during primary school and another 5% not continuing into secondary school. Apart from education, expanded government programmes on water, sanitation and drought relief have probably also contributed to improved health. Donors have played an important role in funding water development projects: in 1984/85 they provided some 58% of the funding for water supply development in communal lands. Donor funding has increased as a proportion of total finance for water and sanitation while government's contribution has decreased in real terms since about 1984. The Drought Relief Programme was introduced by government in 1982 and expanded during the 1982-84 drought. Inside the health sector, the government began in 1980 vigorously to promote the PHC approach as its strategy. This involved integrating preventive with curative services and the development of several programmes designed to ease access to health care for those who were previously excluded. These programmes included: • introduction of free services for those earning less than $150 per month: this was more than twice the industrial minimum wage at that time and gave the majority of the population free access; • a vigorous programme of expanding and upgrading health care facilities particularly in rural areas: by January 1987,274 rural health centres had been completed and a further 49 were under construction. In addition a number of provincial and district hospitals and many rural clinics had been upgraded; • an Expanded Programme on Immunisation (EPI) against the six major childhood infectious diseases, and tetanus immunisation of pregnant women: complete vaccination coverage of young children in rural 86 TRANSFORMATION 21 (1993) DAVIES AND SANDERS ARTICLE Zimbabwe rose from 25% in 1982 to a median of 64% by 1986; • diarrhoeal disease control mainly through improved case management, especially promotion of Oral Rehydration Therapy (ORT): by late 1984 attendances for diarrhoea at health care facilities had substantially decreased and the percentage of rural mothers able to prepare a correct solution for ORT had significantly increased; • establishment of a national nutrition department respor. able for nutrition education and surveillance, and supervision of the Children's Supplementary Feeding Programme (CSFP): by 1984 83% of children aged one had been weighed at least twice (58% in 1982). At the peak of the three year drought over a quarter of a million children in over 8000 communal area feeding points received a daily meal through the CSFP; • a village health worker (VHW) and traditional midwife (TM) training programme: by early 1987 about 7000 VHWs had been trained and a large number of TMs had had their skills upgraded; • child spacing and family planning: largely as result of this programme Zimbabwe has the highest rate of contraceptive use in SS A; • an essential drugs programme; • disability programme. These programmes required expanded government expenditure: between fiscal years 1979/80 and 1982/83 the Ministry of Health's (MOH) actual expenditure rose by 74% in real terms. Within this increase the allocation to preventive services rose from 7.9% to 13.1%. The MOH makes major grants to the central hospitals in Harare, to local authorities, missions and voluntary organisations: the latter three provide primary and secondary level care to a significant propor- tion of low income urban and rural dwellers. Although there was not a significant increase in the total allocation in real terms between 1980/81 and 1982/83, the share of the total going to the central hospitals declined from approximately one-half to one third, reflecting a concern for more equitable provision of services to the poor. Although the above programmes were retained throughout the 1980s, as the performance of the economy deteriorated so it became increasingly difficult to sustain their initial momentum and some have in fact been eroded. For example, the ceiling income for obtaining free health care has been maintained at $150 per month in nominal terms, even though by 1990 the minimum industrial wage had risen to $245 per month. Although this is equivalent to less than $65 in 1980 prices, none of these workers now qualifies. In the most successful of the early health programmes, BPI, coverage has virtually plateaued since 1986: while this is due in part to poor supervision of vaccination at static facilities, the curtailment of outreach activities because of financial constraints on transport has also been TRANSFORMATION 21 (1993) 87 ARTICLE DAVIES AND SANDERS an important factor. Because of government's continued commitment to health care provision, the MOH's share of the budget remained around 5% through the eighties. The real value of its total and per capita allocation continued to grow, but at a substantially slower rate than in the early period. This slowing down reflects the growing financial and economic constraints on government's ability to devote resources to its health programmes, rather than the attainment of their objectives. In summary, therefore, after substantial initial improvements, the mid-1980s saw stagnation and erosion of both 'outside' and 'inside' inputs to health. There are not yet the data to examine the net outcome of this erosion. The patchy data available up to ihe mid-1980s indicate that significant progress was made in addressing Zimbabwe's legacy of ill-health. Most observers agree that there has been a sharp decline in both IMR and U5MR since the late- 1970s. For example, while the precise levels are not known, a review of extant studies suggests that IMR currently lies between 60 and 75. Levels of undernutrition appear to have declined significantly between 1980 and 1983/84, although there is less firm evidence of a decline thereafter. In 1982 a national nutrition survey of under three yeaf-olds showed 17.7% were underweight, 35.6% stunted (under height) and 9.1% wasted (significantly thin). In 1984, another national survey showed 14.5% of 1 to 5 year olds to be significantly under weight. However, due to differing methodologies these two surveys are not strictly comparable. By 1988 the CSO Demographic Health Survey found national levels of 11.5% underweight, 29% stunted, and 1.3% wasted. While the situation appears to have improved with respect to wasting, levels of stunting remain high. Once again the comparison with 1982 and 1984 data is made difficult because of differences in anthropometric cut-offs and the different age groups sampled. Although these data are sketchy, they suggest a divergence between mortality indicators on the one hand and nutrition (quality of life) indicators on the other. The improvement in the former has probably resulted from an energetic expan- sion and reorganisation of health care provision. The adverse effects of drought, recession and stabilisation policies have been partially offset by aid-supported relief feeding programmes and particular health care programmes. However, economic recession and stabilisation attempts have reduced real incomes for large numbers of rural and urban households since the immediate post-inde- pendence boom. This reflects itself in discrepantly high levels of childhood undernutrition which seem to have remained static or improved only marginally despite the health care drive. Zimbabwe's experience in the 1980s thus provides an illuminating case study of both the potential of effective health sector interventions and their limitations in the context of a deteriorating economic environment.7 88 TRANSFORMATION 21 (1993) DAVIES AND SANDERS ARTICLE The economic reform programme which government has embarked upon for the 1990s includes, inter alia, not only further reductions in subsidies but also cost recovery in the social sectors. In the health sector, government proposes enforcing already existing charges and the $150 qualifying ceiling more strictly, rather than introducing new charges. In addition, stress is being placed on training in the management of resources in an efficient and cost-effective way. These measures represent an attempt by government to introduce a more tech- nical, economically-oriented approach to the health sector. Furthermore, the bureaucratisation of structures at all levels through the development of a complex and centralised system of local government has subverted community involve- ment in PHC, most especially in the VHW programme. Together, these develop- ments represent a shift in government health policy away from a comprehensive PHC approach towards a more technical, management oriented approach. Discussion In the now-industrialised countries sustained declines in mortality and im- provements in health were achieved predominantly through changes in what we have called 'outside' factors: rising incomes, improved nutrition, better housing, better environmental conditions at work and home, rising mass education levels. In modern underdeveloped economies, especially in SSA, most advances have probably been achieved through health sector interventions. It is already clear that the progress thus made is fragile, since in the best case an increasing divergence is occurring between mortality and quality of life indicators, and in the worst cases earlier reductions in mortality and morbidity are even being eroded. The most important cause of this is the direct impact of the deterioration in the macroeconomy on the poor majorities, but governments and other agencies have simultaneously found it difficult to sustain the interventive health sector programmes upon which much of the progress has been based. It is now widely recognised that measures designed to promote and sustain economic growth are central to any sustained improvement in the health profile of Africa. The primacy given by the main designers of SAPs to correcting macroeconomic management to create an 'enabling environment* is at least in part motivated by this perspective. Recognition of the transitional social costs of macroeconomic policies has led to the evolution of wider programmes which increasingly incorporate policies to mitigate the negative impact of necessary reductions in government spending and transitional inflationary pressures on vulnerable groups. The question remains, however, as to whether the orthodox adjustment programmes used to deal with recession also tackle the structural problems which are ultimately responsible for poor health. The currently preferred fashion of liberating market forces, minimising govern- TRANSFORMATION 21 (1993) 89 ARTICLE DAVIES AND SANDERS ment interventions and reducing obstacles to a market-oriented integration into the international division of labour is in many respects redolent of the trickle- down approaches fashionable 30 years ago: what might be labelled the Little Bo-Peep hypothesis - leave them alone and they'll come home, wagging their tails behind them. However, while it is clear that inappropriate domestic policy regimes can - and in many instances have - obstruct economic development, economic under- development is not primarily created by such policies. Underdeveloped countries face structural constraints on economic growth: disarticulated markets, skills and capital shortages, technological dependence, underdeveloped financial markets, poor and biassed infrastructure. Furthermore, while it is no longer fashionable to speak of imperialism, it has to be recognised that the uneven development of the international economic system and the unequal political and economic power emanating from it, do restrict the development options of currently under- developed economies. The adjustment policy approach to development seems sometimes to go beyond the simple truth - that good policies are better than bad ones - to implying an untruth - that good policies are all that is needed to achieve Nirvana. Many of the interventions witnessed early on in post-colonial SSA and now being dismantled, were motivated by a perspective - supported by many develop- ment experts at the time - that free market forces at best would be too slow in tackling the structural constraints to development and at worst would reinforce them in a process of reverse cumulative causation. The dismantling of these interventions is not based on a logical or analytical demonstration of the inadequacy of that perspective but rather on a jaundiced, and often selective, view of the actual experience of interventionist policies in SSA. At an analytical level it is possible to construct models that show that development strategies with optimal intervention tackle poverty faster than non-interventionist strategies.8 The issue, however, is not primarily an analytical one but an empirical one: is it possible to find a state and bureaucracy which will intervene in an optimal way, or are all states inherently interest-ridden and venal? The emphasis placed by the Bank on issues of governance (World Bank, 1989) suggests that even in those quarters there remains some optimism that govern- ments can improve, and they recognise that this requires political development. The Long Term Perspective Study speaks of a crisis of governance: Underlying the litany of Africa's development problems is a crisis of governance. By governance is meant the exercise of political power to manage a nation's affairs. Because countervailing power has been lacking, state officials in many countries have served their own interests without fear of being called to account. In 90 TRANSFORMATION 21 (1993) DAVIES AND SANDERS ARTICLE self-defence individuals have built up personal networks of in- fluence rather than hold the all-powerful state accountable for its systemic failures. In this way politics becomes personalised, and patronage becomes essential to maintain power. The leadership assumes broad discretionary authority and loses its legitimacy. Information is controlled, and voluntary associations are co-opted or disbanded. This environment cannot readily support a dynamic economy. At worst the state becomes coercive and arbitrary (World Bank, 1989:61). While this account will strike a chord for many, we think that it mis-charac- terises the problem. It perceives it as having essentially a subjective, personal foundation - self-serving state officials. This is borne out by the continuation of the passage, in which the solution is given: These trends, however, can be resisted. As Botswana has shown, dedicated leadership can produce a quite different outcome. It requires a systematic effort to build a pluralistic institutional structure, a determination to respect the rule of law, and vigorous protection of the freedom of the press and human rights (World Bank, 1989:61). Again, no one can object to the objectives. But this interpretation suggests that the problem emanates from and can be cured by the leadership. In many ways this position suffers from the same flaw as that plaguing Utopian socialists: where are the good guys, the philosopher kings, who are to implement this noble programme? The problem is the underdeveloped nature - almost the lack - of civil society in Africa. This is hinted at in the quotation above, when the Bank attributes corruption to the lack of countervailing forces. But we need to understand that this does not simply mean the existence of multiparty systems. It requires that there is a civil society which is autonomous from the state. In SS A the pervasive- ness of the state is quite remarkable. There is a psychology that suggests that all that is done comes from the state: 'they' should create jobs for us, build better schools, extend clinics, raise wages. As the Bank argues, this view is perhaps understandable in the context of extensive patronage systems and is actively encouraged by politicians. But it is surely not something which can be attributed to the way in which power has been exercised to manage nations' affairs. It is located in the structure and history of our societies. We have 'underdeveloped' structures and institutions, not solely in the economic sphere, but also in the political. The centrality of the state is not simply something that has been imposed on the society in a voluntaristic way. It reflects the weakness of alternative institutions. The process of 'development' entails the creation of these. TRANSFORMATION 21 (1993) 91 ARTICLE DAVIES AND SANDERS So the question then becomes not whether the laissez faire approach of S APs enhances economic efficiency and competitiveness, but whether they advance or hinder the development of civil society. Do S APs really address the political dimension? Even if we accept that in principle free market capitalism creates economic and thus political democracy, it is extremely doubtful whether it will in practice have such effects, given the context of highly unequal distribution of income, wealth and economic power and the unequal access to opportunity in which SAPs are being introduced in SSA. Conclusion It is clear that health is an outcome of many complex economic, social and cultural factors. We have argued that qualitative and sustained improvements in health will require not only substantial improvements in the coverage and effectiveness of health sector interventions, but more importantly, permanent improvement in people's working and living conditions. This will require 'development' or, some would argue, the reversal of the process of underdevelop- ment. We would contend that in the context of the unequal global political economy, this necessarily requires a state which intervenes to address structural constraints on development We suggest that it is here that political democratisa- tion of society at all levels would create the conditions for the disempowered to influence state policy and ultimately their own social and economic situation. The comprehensive PHC approach has at its core the notion that people should be the subjects of their own health improvement rather than merely the objects for interventions by health professionals. Such a strategy provides a basis from within the health sector for developing the political strength, organisation and awareness of the popular classes. This struggle would not only be an important mechanism through which health and health care resources would be both more equitably allocated and more appropriate to the majority's needs, but would also constitute an important input into the wider process of bringing about progressive social change. NOTES 1. Paper originally presented to Witwatersrand Conference on Economic Initiatives: economic policy and development in sub-Saharan Africa, Somerset West, March 1991. 2. Although it must be expected that such indicators approach some asymptotic limit and therefore inevitably slow down, it is unlikely that this stage has been reached by SSA. 3. We do not here enter the important debate on the class nature of the state which is fundamental to the ideas outlined here. In brief, it could be argued that state intervention is a necessary but not sufficient condition for addressing the problems alluded to: sufficiency requires an appropriate class constitution of the state. We oelieve that a state truly representing the popular classes would also be geographically and politically decentralised. 4. This approach could be dubbed 'Adjustment with an Inhuman Phase'. 92 TRANSFORMATION 21 (1993) DAVIES AND SANDERS ARTICLE 5. The Zimbabwean experience has been discussed more extensively in Davies and Sanders (1988). 6. A study undertaken in 1988 reported that IMR averaged 52.7 and U5MR 75.1 between 1983 and 1988 (CSO, 1989). Many observers regard these findings as significantly understating the actual situation, since two other studies in the mid-1980s reported IMRs of 79 and 73 (ZNFPC, 1985; MOH, 1987). concerned with health, to reflect upon the implications of the pandemic for health and economic development. Briefly, we would observe that the spread of the epidemic in Africa is likely to be accelerated by economic recession, which is already aggravating those social factors underlying HIV transmission - such as increasing rural poverty, rapid urbanisation and the expansion of the informal sector activities including commercial sex; that the disease will impose an even greater burden on already limited health sector resources; that there is a danger of the higher profile AIDS diverting limited resources away from other diseases which are both more prevalent and, as yet, more amenable to treatment; that the implications of the epidemic for manpower are significant and complex; and that it is likely that adjustment in people's behaviour to cope with the epidemic will have economic and social repercussions (Sanders and Sambo, 1991). See Kanbur (1987) for an example of this. We are here focussing only on those aspects relevant to the concern of this paper. However, we should note inpassing that corrupt, nepotistic and personalised governments ate not confined to Africa nor, in our opinion, are the most serious examples found there. In the past few years most governments in Eastern Europe have been overthrown for essentially similar transgressions, while even in the western 'democracies' governments are regularly shaken by corruption scandals. REFERENCES Central Statistical Office (1989) - Zimbabwe: Demographic and Health Survey, 1988 (Harare: CSO/Institute for Resource Development). Cornia, GA, R Jolly and F Stewart (eds) (1987) - Adjustment with a Human Face (vol 1) (Oxford: Clarendon Press). Davies, Rob and J0m Ratts0 (1990) - 'Macroeconomic policies for medium term development: the Zimbabwe case study' (mimeo, UNU/WIDER Project on Medium Term Development Strategies, forthcoming). Davies, Rob and David Sanders (1988) - 'Adjustment policies and the welfare of children: Zimbabwe, 1980-1985', in GAComia et al (eds) (vol 2). Halstead, S (ed) (1986) - Good Health at Low Cost (New York: Rockefeller Foundation). Kanbur, SMR (1987) - 'Measurement and alleviation of poverty: with an application to the effects of macroeconomic adjustment' (IMF Staff Papers). Martorelh R and J Habicht (1986) - 'Growth in early childhood in developing countries', in F Falkner and J Tanner (eds) - Human Growth: a comprehensive treatise (vol 3, second ed) (New York: Plenum Press). Mazur, R and D Sanders (1988) - 'Socioeconomic factors associated with child health and nutrition in peri-urban Zimbabwe', in Journal of Ecology of Food and Nutrition, 22. Ministry of Health (1984) - Planning for Equity in Health (Harare: Government Printer). Ministry of Health (1987) - Primary Health Care Review 1987 (Harare: Government Printer). Sanders, David (1985) - The Struggle for Health: medicine and the politics of underdevelopment (London: Macmillan). Sanders, David and Abdulrahman Sambo (1991) - 'AIDS in Africa: the implications of economic recession and structural adjustment', in Health Policy and Planning, 2(6). Shaw, Timothy M and Rob Davies (1991) - The Political Economy of Adjustment in Zimbabwe: convergence and reform (Ottawa: North South Institute). Thomas, Duncan (1990) - 'Household Resources and child health in Zimbabwe' (mimeo, December). World Bank (1983) - Population, Health and Nutrition in Zimbabwe (Washington: World Bank). World Bank (1989) -Sub-Saharan Africa: from crisis to sustainable growth (Washington: World Bank). Zimbabwe National Family Planning Council (ZNFPCVWestinghouse Public Applied Systems (1985) - Zimbabwe: Reproductive Health Survey 1984 (Harare). TRANSFORMATION 21 (1993) 93