Jump to ContentJump to Main Navigation
Social Security in Developing Countries$

Ehtisham Ahmad, Jean Drèze, John Hills, and Amartya Sen

Print publication date: 1991

Print ISBN-13: 9780198233008

Published to Oxford Scholarship Online: September 2011

DOI: 10.1093/acprof:oso/9780198233008.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 19 October 2020

Social Security in the SADCC States of Southern Africa: Social-Welfare Programmes and the Reduction of Household Vulnerability

Social Security in the SADCC States of Southern Africa: Social-Welfare Programmes and the Reduction of Household Vulnerability

Chapter:
(p.415) 10 Social Security in the SADCC States of Southern Africa: Social-Welfare Programmes and the Reduction of Household Vulnerability1
Source:
Social Security in Developing Countries
Author(s):

Morgan Richard

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780198233008.003.0010

Abstract and Keywords

This chapter premises that the greatest single factor now affecting the development and effectiveness of social-welfare policies and programmes in the Southern African Development Co-ordination Conference (SADCC) States is the external one of South African military aggression and promotion of military/economic destabilization. It reviews social-service policies and programmes in the post-independence SADCC States, discusses the discernable impact of these programmes using health indicators of welfare, and then looks at the characteristics of the main target groups for these programmes: the poor and vulnerable. The chapter presents case studies of State responses to external welfare-threatening events that occurred in the 1980s in two countries, one relatively at peace (Botswana) and one thoroughly subjected to war conditions (Mozambique). Some generalized conclusions from these studies and the review as a whole are drawn for the prioritization and design of welfare policies in the region.

Keywords:   social-welfare policies, SADCC, State responses, Botswana, Mozambique

1. Introduction

The ‘SADCC’ countries are defined by their membership of the Southern African Development Co-ordination Conference, established following the independence of Zimbabwe in 1980 as a body aimed at regional co-operation with the purpose of reduction of economic, technological, and transport dependence on South Africa. The body is not a trading block—the Preferential Trade Area (PTA) exists to promote and facilitate trade among a large number of central and southern African countries—but an alliance of countries with a disparate set of political, economic, and historical characteristics which have recognized a common interest in reducing their dependency on the dominant sub-regional power. The fact that the SADCC has been successful in the mobilization of international donor funds for projects with a regional character or for national projects with potential regional benefits, in a wide range of economic sectors, has increased its attraction for the member States, each of which have co-ordinating responsibilities for specific sectors.

For the purposes of discussion of social-welfare policies and programmes, the SADCC programme per se is of limited relevance (with the main interest being in the Food Security Programme co-ordinated by Zimbabwe). The nine SADCC countries2 do not present a coherent set of features for easy common analysis: there are two lusophone members which achieved independence in 1975 through a liberation war, one anglophone member which achieved the same in 1980, three former British Protectorate territories which became nation states peacefully in 1966 and retain close economic relations with South Africa including membership of a common Customs Union, and three anglophone (p.416) central and east African states which became independent in the 1960s, whose relations with South Africa range from close to reluctant to virtually non-existent (Malawi, Zambia, and Tanzania respectively).

Economic indicators also suggest a massive disparity within this geographically almost-contiguous block (Fig. 1). Table 1 shows a range of per capita GNP from US$840 in the case of mineral-rich Botswana to US$ 160–70 for landlocked, crowded Malawi and war-torn Mozambique. Variations in health and welfare outcomes, and their relation with economic indicators such as these, are discussed later in this chapter. Natural conditions including fertility of land, water availability, proximity to sea routes, as well as internal political configurations and economic policies are also highly disparate between these countries.

Social Security in the SADCC States of Southern Africa: Social-Welfare Programmes and the Reduction of Household Vulnerability

Fig. 10.1. SADCC member States.

Table 10. 1. SADCC members: Basic data

Population (millions)

Area (000 sq. km.)

GDP ($ million)

GNP per capita ($)

Angola

8.8

1 247

5 700

690

Botswana

1.1

600

830

840

Lesotho

1.5

30

260

470

Malawi

7.0

118

970

170

Mozambique

13.8

802

3 230

160

Swaziland

0.8

17

590

670

Tanzania

22.2

945

5 600

290

Zambia

6.7

753

2 330

390

Zimbabwe

8.4

391

4 530

680

TOTAL

70.3

4 903

24 040

Note: Data refer to 1986, except GDP for Angola and Mozambique which are 1984.

Sources: World Bank, World Development Report 1987 (Washington, DC: World Bank) SADCC, Macro-Economic Survey 1986 (Gaborone: SADCC) EIU, SADCC: Progress, Projects and Prospects (London: Economist Intelligence Unit, 1984) (GDP Angola and Mozambique).

However, the premiss of this chapter is that the greatest single factor now affecting the development and effectiveness of social-welfare policies and programmes in the SADCC States is the external one of South African military aggression and promotion of military/economic destabilization. The degree of impact of this factor, present in some form for all SADCC countries, has become a major constraint on the extent to which social-service programmes could be developed and implemented, at any given level of policy commitment and enthusiasm. Combined in many cases with the effects of severe droughts and unfavourable external economic circumstances, the destabilization phenomenon (p.417) (p.418) affects social services particularly through the medium of reduced resource availability to finance such programmes—whether due to absolute resource loss or the pressure to divert to types of expenditure designed to bolster security and protect trade routes. In the cases of Angola and Mozambique (and to a lesser extent Zimbabwe) social assets have been directly destroyed and programmes massively disrupted by externally backed insurgents, whilst the displacement of millions of people within and across national borders has increased the demand for urgent welfare provision in these countries and their neighbours. Over half a million additional infant and child deaths are estimated to have been caused by insurgency between 1981 and 1986 in Angola and Mozambique alone (Green et al. 1987). By 1988, Malawi was accommodating a number of displaced Mozambicans reaching almost 10 per cent of its own population.

The undermining effect of South Africa’s regional destabilization policy on the SADCC States, documented in particular detail in Hanlon (1986), has been interlinked with a series of natural events which combined to increase vulnerability of (particularly) rural households in the region in the first half of the 1980s. The period 1981–5 was characterized by rainfall insufficiencies of historically severe proportions, lasting in most cases for more than one year, in all the SADCC States. The resultant reductions in food and livestock production and access, when coupled with disruption of rural society and transport and trading links due to war, turned ‘normal’ hungry seasons into droughts, and in the countries worst affected by both phenomena, to famines. Whilst drought receded as a widespread occurrence after the 1985/6 agricultural season, and food production and flows in most countries began to recover, the destabilization wars have tended to intensify. In the last few years, most of the larger SADCC countries have also adopted economic ‘adjustment’ programmes, which, although the initial record is mixed, have tended further to reduce the share of national resources allocated to social-service sectors.

For many countries, the experience of the 1980s represents a sharp reversal of earlier progress. Significant economic growth in the late 1960s and through most of the 1970s, despite for some countries declining terms of international trade, coupled with post-independence inflows of donor funds specifically for social-sector programmes, allowed even the weakest economies in the region to support a rapid build-up of social-service coverage. This was sometimes associated with community-level mobilization policies achieved through political channels and the use of literacy programmes and health education activities. Voluntary labour mobilization was in some countries a contributory factor in extending access to social infrastructure, and settlement policies encouraging rural concentration took this process further (Tanzania is an example). In some countries (Botswana, Lesotho, and later Zimbabwe), the social infrastructure facilities put in place during this period were able to support the implementation of nutrition-based food and income supplementation programmes in rural (p.419) areas, which were introduced or expanded in times of stress to support vulnerable groups and households.

Following the achievement of Zimbabwean independence under a regime unfavourable to South African interests, and the realization of the SADCC ‘project’, South Africa’s regional policy became increasingly militarized and a ‘contextual shift’ took place for development work in the entire region. With hindsight, the question for the subsequent period increasingly became one of how the gains in promotion of greater social security could now best be protected (rather than one of rapid expansion): a question posed in domestic economies affected by a hostile external military environment and a partly unfavourable international trading context. The question of protection of welfare gains is posed, for the most thoroughly destabilized countries, by the need to respond to deliberate and direct targeting of welfare programmes themselves, through attacks on health posts and schools, kidnap, murder and mutilation of the professionals who staff them, and attacks on the vehicles that conduct their mobile programmes, deliver relief food, and facilitate water-point maintenance.

This chapter proceeds with a review of social-service policies and programmes in the post-independence SADCC States, goes on to discuss the discernable impact of these programmes using health indicators of welfare, and then looks at the characteristics of the main target groups for these programmes, the poor and vulnerable. Case-studies are presented of State responses to external welfare-threatening events which occurred in the 1980s in two countries, one relatively at peace (Botswana) and one thoroughly subjected to war conditions (Mozambique), and some generalized conclusions from these studies and the review as a whole are drawn for the prioritization and design of welfare policies in the region.

2. Social-Welfare Programmes in Sadcc States

The definition understood here of programmes adopted by the State to promote or increase family welfare (or the welfare of specific groups within families) is broader than that of not directly productive ‘social sectors’, principally health, education, and water. Also included are interventions and policies established by the State to promote or regulate production by families, and to influence access to ‘basic’ household items (those necessary to healthy life and social dignity) and to incomes. Such definition is largely based in a rural area perspective, where more than 70 per cent of the SADCC population are still found, mostly engaged in some form of agriculture, even if this may not comprise the largest source of their income.

(p.420) 2.1. Broadening Access to Primary Health Care

Having inherited health care systems biased towards urban facilities and high-cost, curative treatments, the SADCC countries have since independence shown a strong recognition of the importance of extending health services to the mass of their populations. Some convergence of health sector strategies, at least in terms of officially stated intentions, can be seen. In most countries, the rate of construction of lower-level health facilities was greatly increased in the years after independence, training of nursing staff and paramedical personnel expanded, and greater shares of generally increasing health sector expenditures allocated to preventive programmes.

This shift of emphasis accorded with the desire to bring greater proportions of rural and poorer populations within the coverage of the health network. Generally, the adoption of a primary health care (PHC) approach to health service provision allowed much greater numbers to be reached without impossibly large budgetary implications, especially to the extent that different types of programmes could be integrated (for example, public and personal health programmes). Typically, PHC has as its main elements: training and posting of frontline village health workers, community mobilization (for example, though village health committees), establishment of referral systems, mother-child health care, communicable disease control, young child growth monitoring, treatment of common diseases, child immunization, and water supply and environmental sanitation programmes.

Progress in these areas has however been highly variable between countries and regions. Some countries (Malawi, Swaziland) continue to show marked bias towards higher-level, urban-based curative facilities in their health sector expenditure patterns. In most PHC programmes, community involvement and village-based health activities have been weak in practice, and the roles of community health workers often unclear. Shortages of transport, medical equipment, fuel, and drugs, have played havoc in some areas with efficient delivery of services. None the less, the picture presented in the region is one of highly expanded coverage of physical facilities (subsequently reduced by war in Mozambique and Angola), with a less than consistent pattern of improvement in their efficiency of operation and rates of usage.

The health sector in SADCC States generally received some 5–8 per cent of total central government expenditure over the 1972–83 period (World Bank 1986), and the percentage in Mozambique was until recently considerably higher. However, despite the ability of health expenditures in most countries to retain their shares, the real values of expenditures have tended to fall in several cases as financial stringency affected the entire government budget (Mozambique since 1986, Zambia and Tanzania since the late 1970s, and Zimbabwe to some extent since 1982). This has resulted in substantial cuts in construction programmes in particular. On the other hand, the implications (p.421) for poorer groups are counterbalanced to some extent where increasing emphasis is being given within health sector budgets to PHC programmes (as, for instance, in Tanzania and Zimbabwe). Finally, without consideration of relative cost-effectiveness in health programme delivery, the absolute financial base of governments is also a major determinant of the real value of services provided: despite devoting a higher proportion of the total budget to health, Lesotho in the mid 1980s spent about US$3 per person per year, and Mozambique US$4, compared to Botswana’s roughly US$18.

Some degree of cross-country comparison is possible for one of the central PHC programmes under implementation in all SADCC countries, child immunization (although the data are not complete or fully consistent). As shown in Table 10.2, progress has been registered in immunization coverage over recent years in almost all countries, with particularly good results achieved in Lesotho and Botswana. The Zimbabwean coverage figures are also higher on aggregate than indicated here, and have increased rapidly since (p.422) independence. Mozambique too made rapid progress up to 1983; since then, war-related insecurity has dictated concentration of vaccination programmes in towns and easily accessible rural locations only, where they have been highly successful.

Table 10.2. Estimated child immunization coverage rates for various SADCC countries in recent years (% coverage)a

Vaccination

BCG (Scar)

DPT (3rd)

Polio

Measles

Full immunization coverage

Zimbabweb 1982/84

53/76

32/66

31/66

51/53

25/42

Malawi 1977/81

97/80

46/60

23/53

11/60

n.a./55

Lesotho 1982/86

76/86

56/82

54/80

49/73

40/65

Swaziland 1982/84

n.a./72

31/43

22/43

31/37

n.a./24

Zambia 1984

68

44

44

55

35

Botswana 1985

68

68

67

68

68

Mozambique 1983/85

59/47

38/29

38/25

51/39

n.a.

(a) Of children up to one year, except Zimbabwe 12–32 months, Lesotho and Zambia 12–23 months.

(b) Zimbabwe, rural children only.

Sources: UNICEF, Situation Analyses; Mozambique Ministry of Health (unpublished data).

In countries with relatively low rates of full vaccination coverage, such as Zambia and Tanzania, shortages of fuel to maintain the cold chain and of transport for outreach purposes are major constraints on these programmes. Basic drugs are often generally in short supply in these countries, particularly at the periphery. With depreciating currencies, costs of items not produced locally or supplied directly by donors have been rising rapidly, further constraining supplies to the health sector.

2.2. Widening Access to Education

Most SADCC nations have since independence rapidly expanded the primary-education system in an effort to provide universal access. Although qualified by the inclusion of over-age pupils, Table 10.3 suggests that this coverage is virtually achieved in Lesotho, Zimbabwe, and Swaziland, and nearly so in Botswana and Zambia. Zimbabwe is particularly notable for an annual increase of over 20 per cent between 1979 and 1985 in numbers attending primary school, and Tanzania and Mozambique have also made substantial advances from very poor colonial bases.

Table 10.3. Numbers enrolled in primary school as percentage of relevant age groupa

% Enrolment overall

% Enrolment of girls

1965

1983b

1965

1983b

Angola

39

n.a.

26

n.a.

Botswana

65

96

71

102

Lesotho

94

110

114

126

Malawi

44

63

32

52

Mozambique

37

79

26

68

Swaziland

n.a.

106

n.a.

105

Tanzania

32

87

25

84

Zambia

53

94

46

89

Zimbabwe

110

131

92

127

(a) Most countries include significant numbers of over-age children in enrolment.

(b) Some years differ (±2).

Sources: World Bank 1986; UNICEF Situation Analyses.

It is noteworthy that, in similar manner to the health sector, these indicators show Malawi lagging behind. In the case of Mozambique, rapid expansion (p.423) since independence has been followed by war-related decline: the numbers of children enrolled and primary schools operating rose by 122 per cent and 37 per cent respectively between 1975 and 1979, and then fell by 2 per cent and 37 per cent respectively from 1982 to 1987 (Mozambique Ministry of Education). A similar trend has probably occurred in Angola. ‘Universal access’ as a target cannot however be taken as referring to enrolment alone, and indications are strong that the quality of education tends to fall far short of functional levels in many regions. Retakes and drop-outs are common. In Mozambique less than 20 per cent of entrants complete their primary education (Grades 1–4) and less than 5 per cent graduate in the minimum number of years (Johnston et al 1987). In Swaziland only 50 per cent of children entering Grade 1 complete Grade 7 and only 21 per cent of entrants do not retake a Grade during primary education. Conditions of education contribute to poor results, with overcrowded classrooms and shortages of textbooks and classroom materials being common, particularly in countries with severe budgetary constraints. Shortages of teachers in Malawi and Mozambique, for example, led to recent pupil-teacher ratios of over 60:1, and even in those countries better endowed, qualified teachers are often in short supply.

In contrast to primary-education coverage, pre-school child care and educational institutions are few in number and low in coverage; only in Swaziland do they cater for more than 5 per cent of 3–5 year olds, and here, as elsewhere, the bias is strongly towards urban areas and parents who can afford fees. This area has not been a conspicuous priority for SADCC States, despite the potential benefits to working mothers.

Expansion of secondary-education facilities has also tended to take a lesser priority than the achievement of universal primary access, but some countries have expanded rapidly in this area (Table 10.4). Zimbabwe has been able to build to some degree on an inherited base, whilst Botswana is now adding quickly to its stock of junior secondary schools. Even so, a relatively low percentage of secondary school age children in these countries, and even lower proportions elsewhere, are able to enter secondary education. Although this is slightly compensated by the existence of small-scale vocational or ‘education with production’ training organizations in some States, the SADCC countries face an overall ‘primary school leaver’ problem of substantial proportions, and a rapidly growing incidence of secondary school leaver unemployment as well.

Table 10.4. Numbers enrolled in secondary schools as percentage of the relevant age group

Enrolment as % of age group

1965

1983a

Angola

5

12

Botswana

3

21

Lesotho

4

19

Malawi

2

5

Mozambique

3

6

Swaziland

n.a.

n.a.

Tanzania

2

3

Zambia

7

17

Zimbabwe

6

39

(a) Or nearby year.

Sources: World Bank, 1986; UNICEF: Situation Analyses.

Improvement in adult literacy rates has proceeded extremely rapidly where State policy has given priority to this objective (most notably in Tanzania). Recent estimates of adult literacy are given for some countries in the UNICEF Situations Analyses (Table 10.5).

Table 10.5. Estimates of adult literacy rates, recent years

% of all adults literate

% of women literate

Botswana

40

n.a.

Malawi

27

13

Mozambique

27

12

Zambia

59

53

Tanzania

85

80

Source: UNICEF, Situation Analyses.

(p.424)

2.3. Water Supply Provision

Lack of access to ‘safe’ or improved water supplies is a major health hazard in SADCC countries, as suggested by the high incidence of diarrhoeal diseases and skin, eye, and parasite infections. A further dimension is that of the many hours of work and rest forgone due to long walking distances from home to natural water sources in rural areas. This affects women and children particularly: the UNICEF Zimbabwe Situation Analysis estimates that, in 1984, 73 per cent of rural water collectors were women and 25 per cent children.

Despite substantial public-sector investments in most countries, it has proven difficult rapidly to increase the access to improved water sources of rural populations in SADCC. Botswana, with its concentrated rural settlements (p.425) and large expenditures on relatively expensive borehole-drilling for village water supplies, is the exception to this rule (Table 10.6).

Table 10.6. Estimates of percentage of rural population without access to improved/ protected water source or latrine

% without access to improved water

% without access to latrine

Angola

n.a.

n.a.

Botswana (1981)

23

77a

Lesotho (1984)

76

89b

Malawi (1977)

75

47c

Mozambique (1987)

87

n.a.

Swaziland (1983)

70

73

Tanzania (1981)

61

n.a.

Zambia (n.s.)

67

70

Zimbabwe (1983)

61–7

75

(a) Figure refers to 1983/4.

(b) Recent figure.

(c) National figure for 1982.

Source: UNICEF, Situation Analyses and country offices.

A variety of technologies, from high-cost borehole-drilling to shallow wells fitted with handpumps, gravity systems, and protected springs, have been used in the region, with a tendency towards lower-cost solutions in recent years. Maintenance capacity has been a further important consideration, both in choice and design of system, and in the creation of back-up facilities and technical capability, whether in local authorities or among communities themselves. Experience has shown that switching water utilization to high-technology systems without ensuring local maintenance capability can leave communities highly dependent and in a worse position than before. Unfortunately, resources, whether national or international, to ensure the functionality of water systems have been given less priority in water supply programmes than those for the initial investment in construction.

A similar problem of technical design appears to retard the adoption of improved sanitation methods, as indicated by the low latrine availability in rural areas. However, some latrine construction programmes have been usefully based on the provision of materials by the state, with labour and local material inputs contributed by the household (or community) concerned. (p.426)

2.4. Agricultural Programmes for Enhancing the Productivity of the Poor

A wide range of policy approaches towards the agricultural sector, in which the majority of the region’s population is still engaged, exist in the SADCC countries, with objectives of support to family-level producers and improvement of rural incomes often less than fully reconciled with concerns for improved levels of national self-sufficiency in food. As explored in Section 4 below, food security at the household level is, for much of the region, a question of stabilizing and improving levels not only of domestic production but also of off-farm incomes. Some development policies in the region have recognized this feature, with attention being paid to income diversification through small-scale irrigation and horticulture, cattle and small livestock raising and marketing projects, and craft or small industry promotion, as well as off-season household income supports introduced through public works and other means (see discussion of food and income support below).

In the areas of relatively high and reliable rainfall, agricultural development programmes have tended to adopt an approach designed to increase output of cereals, notably maize, through a combination of extension advice, small-scale credit provisions, and input distribution (improved or hybrid seed, fertilizer). The more successful of these in terms of raising short-term outputs, particularly Zimbabwe and Malawi, have combined such packages with substantial price increases and the effective extension of marketing points to smallholder areas.

In Zimbabwe, success in the 1980s in raising overall levels of small-scale farm maize output and marketing has been particularly rapid, one explanation being the extent of pre-independence investment in research on breeding and use of hybrid maize varieties and fertilizers in high rainfall areas. Only after independence in 1980 were these technologies made extensively available to smallholders (Rohrbach 1987). In other countries, notably those with labour constraints to increases in smallholder crop production, efforts have been made to promote the use of hired or government-provided tractors for land preparation and sowing (Botswana, Lesotho, and Swaziland), on the grounds that this ensures timely use of soil moisture. A further variant has been the use of area-based development programmes, incorporating community development activities and the construction of roads and social service infrastructure (such as the Swaziland Rural Development Areas). The example of Zimbabwe, however, suggests strongly the importance of sustained investment over time in agricultural research, with the emphasis on technological enhancement of the productivity of existing mass small-farmer production systems, whether in high rainfall or drought-prone conditions. Raising efficiency of soil moisture use, development of drought-tolerant foodcrop varieties, low-cost irrigation, and soil fertility improvements are among the research areas which have tended to remain neglected.

In three countries, where the rural production and income base is under (p.427) threat from war (Angola and Mozambique) and severe drought (Botswana), a more ‘minimalist’ technical approach has been taken to try to avoid large-scale detachment of rural families from the land. In the first two countries, the provision of basic seeds, handtools, and consumer goods as production incentives, has been given increasing prominence in recent years. In Botswana, seeds have also been provided on a yearly basis to all farmers, as well as field clearance grants, and subsidies on the cost of basic farming implements and fencing materials.

Agricultural and rural development policies have in some nations been integrated with programmes aimed at social reorganization and the restructuring of production systems: the formation of communal villages and promotion of village-based co-operatives for the purposes of marketing and input provision in Tanzania, Mozambique, and Angola are examples. For various reasons, no great success in production terms can be claimed for these programmes, but by concentrating rural settlement patterns, they have facilitated a much wider access to social services than was possible previously. In Mozambique too, the formation of new village settlements has improved the security of some populations and facilitated the resettlement of displaced people. In Zimbabwe, resettlement in commercial farming areas, although so far relatively small in scale (involving some 32,000 families by 1986), has been a basic strategy to relieve land pressure in the ‘communal’ areas.

The drought period of 1982–5, coupled with concerns about national food security and with increasing international criticism of policy bias in favour of large-scale State farms, parastatal marketing systems, and export crop producers, led to some important shifts in policy in certain of the SADCC countries. It has become increasingly recognized that ‘drought-staple’ food crops play an important role in improving household food security. Hence price rises (in some cases through lifting of controls) have been effected for relatively drought-resistant crops such as sorghum, millet, and cassava. Research on seed-breeding and variety-testing for such crops has also been given greater priority in national programmes and by SADCC. A further realization has been that both production incentives through pricing policies effective at the farm level and the guaranteeing of the conditions to ensure productive potential through access to seasonal inputs and farming equipment are needed for sustained and broad-based increases in output. To make price incentives effective has required, in turn, simultaneous improvement in the efficiency and coverage of marketing systems, in some cases by allowing a greater role for co-operative and private sector agents alongside dominant parastatals, and also the provision of basic consumer goods in village shops or in direct exchange for agricultural products, to make marketing worthwhile. For economies with extensive shortages of foreign exchange, donor assistance in the provision of such goods has been an important condition for the realization of incentives to revive agricultural output.

(p.428) Efforts to achieve this combination of apparently necessary conditions for increases in small farm production have been frustrated by several factors. In Tanzania, Zambia, Mozambique, and Angola, poor infrastructure and long distances coupled with fuel and transport shortages have hindered the delivery of already inadequate quantities of inputs and equipment. The reintroduction in the former two countries of co-operative and private marketing agents in 1985/6, with a reduction in the roles of the respective parastatals, has also not proceeded smoothly, although it appears likely that the trends towards increased competition within marketing systems will be maintained. Moreover, whilst most governments have recognized the economic potential and social benefits of providing more favourable producer prices for small farmers’ products, some are highly constrained in their ability to do so, particularly for internationally traded crops where important parity prices (for example, from South Africa) or limited export market possibilities (such as for white maize) have to be observed.

In conjunction with the SADCC Food Security Programme, and based on agricultural, meteorological, and nutritional surveillance information systems already in place or being developed, attempts are being made to improve monitoring systems for national and household food security in the SADCC countries. These early warning systems have the potential to assist more effective crisis management and response by governments in the region, but thus far they suffer from a number of limitations. They are primarily concerned with aggregated food supply, crop, and rainfall data, which are not directly useful for monitoring stress among the food-deficit households dependent substantially on non-agricultural sources of income which, as described below (Section 4) are among the most vulnerable. Data systems to provide information on such indicators as local market prices, household food stocks, and the nutritional status of vulnerable group populations, whilst increasing in scope and coverage, are as yet incorporated into national monitoring systems in relatively few countries. Further, the systems in place tend to operate mainly at national level, reducing the scope for decision-making and timely response by provincial or district-level authorities to specific, localized problems. Even at national levels, existing early warning systems seldom feed directly into established structures for policy analysis and decision-making, meaning that much of the effort put into data collection and preliminary analysis may be wasted. None the less, the importance of sustained investment in the development of ‘wide’ information systems, including local and informal indicators of household economic status and activity, is increasingly clear in view of the complex and structural nature of household vulnerability in the region.

(p.429) 2.5. Food and Income Support Programmes for Vulnerable Groups

Despite the weakness of the data base for programme design and monitoring, a range of interventions directed to high-risk areas, groups, or households are carried out by SADCC governments in addition to programmes of agricultural output promotion, primary health care, and other social services. These are essentially direct income and welfare transfers, in recognition of special problems of poverty, food shortages, and low income, or in response to normally short-run emergency situations. The main types of programmes in operation are:

  1. (1) Institutional feeding and food distribution for nutritionally at-risk groups. These operate typically through social-service facilities and are active on some scale in almost all the SADCC countries. They break down into four basic types: (i) feeding programmes in education and training establishments, including primary schools (notably the BLS countries, Botswana, Lesotho, and Swaziland); (ii) provision of take-home rations at health facilities for selected vulnerable groups (women, undernourished children, TB patients, and so on), also widespread in Botswana and Lesotho, and to a lesser extent in Swaziland and Malawi; (iii) rehabilitative, often ‘on-site’ feeding of actually malnourished children at health facilities, which operates in Zimbabwe (Child Supplementary Feeding Programme), Botswana, and in Malawi and Mozambique to some extent; and (iv) distribution of supplementary rations to families in situations of drought or other emergency. By the mid-1980s this was confined to Botswana (with some 270,000 additional beneficiaries), Mozambique (where the total number of ‘affected’ and displaced people eligible for free food distribution stood at some 3.2 million in 1988), and Angola. Drought-hit areas of Zimbabwe also retained a limited food distribution programme of this type during 1986/7. In other countries, such operations have largely been phased out since the severe drought periods of the early 1980s.

  2. (2) Village-level food-for-work projects, which are most extensive in Lesotho, with about 17,000 participants per year, but also in limited operation in Tanzania and Malawi. In Mozambique, both food-for-work and com-modities-for-work projects have been undertaken on a small scale, as well as food sales for workers in agricultural state enterprises.

  3. (3) Seasonal rural employment creation on public-works projects at basic wage levels, undertaken in order directly to boast household income levels. These have been extensive since 1982 in Botswana, where they complement the food relief element of the government’s response to drought (see Section 5). Such public works were also used extensively in Zimbabwe in 1983/4, and on a lesser scale in 1986/7.

  4. (p.430) (4) Market intervention sales of food in order to stabilize or reduce prices in urban areas, undertaken by Grain Marketing Boards. These were used until recently in Tanzania to protect real incomes of the urban poor.

An emerging issue with regard to such programmes, explored further in Section 5, is their role not merely as short-term buffering mechanisms for emergency-affected households, but also in relation to strategies for more sustained attacks on poverty and vulnerability, through such mechanisms as community mobilization, creation of permanent income-generating opportunities, and income-protection during medium-term economic adjustment policy implementation. These potentials have not yet been fully realized in either policy formulation or practice in the region (although the trend to adoption of intersectoral food and nutrition strategy perspectives in some countries is encouraging). A particularly weak linkage has been with the area of pricing and subsidy policy, one with major implications for the welfare status of the urban poor.

2.6. Consumer Pricing and Subsidies

Some governments in the region have used a policy of subsidizing extensively the costs to consumers of basic food products, particularly staple cereals. This has been the case in countries with large urban populations: in Mozambique and Angola through city rationing systems, and in Zimbabwe, Zambia, and Tanzania through covering losses of Grain Marketing Boards and/or payments to milling companies to limit the mark-ups passed on to the consumer. Since the sales of these Boards and companies have been concentrated in urban areas, such policies have tended to benefit the urban poor, who devote a high proportion of their household budgets to purchase of staple commodities.

Levels of subsidization have not proven sustainable in the 1980s, with state finances under pressure from the effects of worsening international terms of trade and military destabilization. Mozambique’s urban rationing system for basic commodities has come to rely almost entirely on external donations, and the low-price consumer policy was abandoned in steps during 1987/8 as part of the economic adjustment programme, which envisaged rapid elimination of consumer subsidies. In Tanzania, due to the poor performance of the official internal marketing system, urban sales of maize meal and other cereals (rice and wheat) have relied increasingly on commercial imports, difficult for the country to afford, as well as food aid donations. Urban demand at relatively low prices has tended to outstrip supply, causing shortages of basic products and rationing by parallel markets, with negative implications for the poor. It is probable that such policies have also influenced consumer preference in favour of less nutritious, highly refined blends of maize, as well as cereals in which countries have no immediate prospect of self-sufficiency (wheat, rice), leading (p.431) to further import dependence. Most seriously, however, for all three countries which have used parastatal marketing boards to implement these subsidies, the costs of the programmes have taken up increasing proportions of government budgets otherwise under stress. This factor above all has led to their curtailment (Zimbabwe in 1982–3, Tanzania in 1984, Zambia over the period 1984–6), in an effort to reduce losses sustained and the often inflationary borrowing associated with such deficits.

3. The Impact of Welfare Policies on Health and Nutrition

Health indicators of the success or otherwise of the range of welfare policies and programmes adopted in SADCC countries can only give rough measures of overall changes in well-being: this is also true for sectorally based information of other kinds. However, allowing for variations in average income levels between countries, health sector indicators, for all the recognized difficulties of cross-country comparison and the poverty of data series, do provide some convenient, broadly comparable measures of impact. They are also useful in focusing the argument on those groups within the population facing the greatest vulnerabilities from a welfare point of view.

UNICEF estimates of infant and under-5 mortality rates for the years 1960 and 1986 and for life expectancy at birth for 1986 are shown in Table 10.7. The most significant achievements in life expectancy, on these estimates, have been made in Tanzania, Zambia, Zimbabwe, and Botswana, where high commitment to the health and other welfare sectors in public policy, and in the case of the latter two countries, rapidly rising income levels, have been apparent. In the two countries most affected by destabilization wars (Angola and Mozambique), and in low-income Malawi, life expectancy (for both sexes) remains under 50 years.

Table 10.7. Infant and under-5 mortality rates, and life expectancy in SADCC countries, 1960–1986

Infant mortality rate (deaths per 1000 live births under one year)

% change in infant mortality rate 1960–86

Under-5 mortality rate (deaths per 1000 live births under five years)

Life expectancy at birth (years)

1960

1986

1960

1986

1986

Angola

208

200a

-4

346

_b

44

Botswana

119

69

-42

174

96

56

Lesotho

149

102

-32

208

140

51

Malawi

206

153

-26

364

270

47

Mozambique

174

200a

+ 15

302

_b

47

Swaziland

152

120

-21

227

178

50

Tanzania

146

107

-27

248

179

53

Zambia

135

82

-39

228

132

53

Zimbabwe

110

74

-33

182

118

57

South Africa

135

75

-44

192

101

55

(a) These figures only from Green et al. (1987).

(b) Estimated by Green et al. to be in range of 325–75.

Source: UNICEF (1988) except as indicated in note above.

A similar pattern is clear, broadly, for changes in infant mortality rates (IMRs) over the same period. All countries achieved falls of 20 per cent or more, except for Mozambique and Angola where rapid improvements in welfare services after independence were reversed by the effects of war since the early 1980s. As argued in Green et al. (1987) rates of 200 per thousand for the IMR and of 325–75 for child mortality are realistic, and accord with recent if limited survey evidence. Such rates are not unknown even in parts of the region not directly affected by war: certain districts of Malawi are also estimated to have IMRs above 200 per thousand.

Fig. 10.2 shows a close relationship between GNP per capita (estimated for 1985 except Angola, 1980) and national IMRs (estimated for 1986). The impact of war on mortality in low-income Mozambique and middle-income Angola is clearly seen. Also of interest is the relatively poor performance of (p.432) non-war-affected Malawi and middle-income Swaziland, in relation to other countries with similar characteristics.2

Social Security in the SADCC States of Southern Africa: Social-Welfare Programmes and the Reduction of Household Vulnerability

Fig. 10.2. GNP per capita against infant mortality rate. Note: Infant mortality rate figures represent annual deaths of children under 1 year per 1,000 live births. Sources: World Bank and UN Population Division estimates quoted in UNICEF (1988), except Mozambique and Angola IMRs, re-estimated in Green et al. (1987) to take account of effects of war.

Figure 10.3 relating life expectancy to GNP per capita gives similar indications. The directly war-affected countries perform least well; Malawi and particularly Swaziland appear somewhat below trend; whilst considerable achievements in increasing life expectancy at respective levels of resource availability are suggested in Tanzania and Zimbabwe. As stated above, however, care needs to be exercised in drawing even limited conclusions from these data.

Social Security in the SADCC States of Southern Africa: Social-Welfare Programmes and the Reduction of Household Vulnerability

Fig. 10.3. GNP per capita against life expectancy. Source: UNICEF (1988).

An ‘intermediate’ set of welfare indicators from the health sector are shown in Table 10.8, involving measures of access to and utilization of health services.

Table 10.8. Indicators of access to and use of health services

Population (000) per

Estimated % of births in health institutions

Estimated coverage of pregnant women by ante-natal care services (%)

Physician

Nurse

1965

1981

1965

1981

Angola

12

n.a.

4

n.a.

n.a.

n.a.

Botswana

22

9

16

1

66

89

Lesotho

23

13a

5a

3a

51

61

Malawi

47

53

49

3

35

n.a.

Mozambique

22

33

5

6

28

46

Swaziland

n.a.

7

n.a.

n.a.

30

86

Tanzania

22

19b

n.a.

5

58

96

Zambia

11

7

6

2

43

80

Zimbabwe

5

7

1

1

67

89

(a) These figures from UNICEF Situation Analyses, 1983/4.

(b) This figure is for 1984.

Sources: For columns 1–4, World Bank (1986), except as indicated in note above. For columns 5–6 UNICEF Situation Analyses.

Whilst here again the higher-income countries such as Botswana, Zimbabwe, and Zambia tend to have the most favourable results, rural-urban and inter-provincial disparities in access within countries are likely to be almost as great as those between them. Such figures also do not reflect the existence of medical and healing services outside the State sector (namely, the significant private medical sectors in Botswana, Zimbabwe, and Swaziland, and the ‘traditional (p.433) healers’ who are widespread in the region). The impact of policy, however, is particularly apparent in the case of Tanzania, where extraordinarily high levels of ante-natal coverage and institutional births have been achieved for a country so populous and poor.

Improvements in nutritional status and reduction in the incidence of common diseases are both more difficult to measure and more complex to achieve. Rising real levels of personal incomes and per capita health expenditures by the State are not necessarily enough to effect these improvements, since disease and nutrition patterns tend to reflect underlying conditions of household and communal poverty, which require to be addressed by specific intervention in a range of sectors, even as incomes rise.

Diarrhoeal diseases, and intestinal infections generally, remain a common cause of death in young children throughout the region, and acute respiratory infections are also still prevalent. Measles has not yet been eradicated as a cause of death in all countries, and malaria is endemic in most of the central SADCC area. Perinatal complications and infections are prime contributors to high IMRs. Bilharzia and tuberculosis remain problematic in countries such as Botswana and Swaziland, despite rising income levels; and sexually transmitted diseases, including AIDS, exist throughout the region and appear to be increasing rapidly.

(p.434) (p.435) In consideration of trends in the nutritional status of children, insufficient and often inconsistent time series surveillance or survey data in almost all countries permit not more than a sketching of broad characteristics. It is clear that undernutrition is widespread throughout the region, and is largely structural and poverty-related in nature, tending to be higher in remoter areas such as the Western Kalahari and Lesotho Highlands, in small villages, among families of low-waged workers on agricultural estates, and in some cases in newly settled peripheral urban areas. Widespread stunting (low height for age) is mainly the result of growth faltering during the weaning period (normally the second year of life), associated with abrupt termination of breastfeeding, poor feeding practices, and lack of protein supplements. Despite this, breastfeeding practices, promoted by most governments, remain favourable in most countries, with little decline associated with urbanization. Whilst the lowest incidence of breastfeeding is found amongst the most highly educated women, these are often the best weaners. In general, a strong inverse relationship is found in nutrition research in the region between the mother’s educational attainment and the incidence of young child malnutrition. The 1983 Swaziland Nutrition Survey also found negative associations between child stunting and household ownership of consumer durables as well as access to improved water supplies (UNICEF, Swaziland Situation Analysis, 1987).

Since few countries carry out regular monitoring of nutritional status from a reasonably consistent sample population, trends in this indicator are difficult to determine. The case of Botswana, with a national surveillance system in place since 1978, indicates a major worsening of young child nutritional status in the early 1980s, coinciding with the onset of repeated drought years, and a stabilization and subsequent dramatic improvement after 1984 coinciding with the implementation of large-scale drought relief measures (see Section 5 and Fig. 10.5 below; and Cornia et al. 1987: vol. ii). In Lesotho, however, rural area monitoring of child weights at clinics with food distribution by the Catholic Relief Service over the period 1980 to 1986 showed that initial improvements up to 1983 in the percentage of children found underweight were rapidly reversed by 1986, despite the lessening of drought conditions and continued growth in imports of food aid (Food Studies Group 1988). Admissions and deaths in the main central hospital associated with protein energy malnutrition also rose over the period. Possible determinants of these phenomena in Lesotho include declining crop yields, land access, and work opportunities for rural families in the 1980s, and rises in the price of the main staple food, maize, relative to other commodities (ibid).

Social Security in the SADCC States of Southern Africa: Social-Welfare Programmes and the Reduction of Household Vulnerability

Fig. 10.5. Child malnutrition and crop production in Botswana, 1980–1986. Note: Broken line shows estimated malnutrition incidence adjusted to take account of changes in recording system in 1985. Sources: Cornia et al. (1987); Botswana Ministry of Agriculture (1986).

Survey data elsewhere show high levels of chronic malnutrition persisting in the low-income countries (Malawi, Mozambique, and Tanzania). In Malawi, a 1981/2 national survey found over 50 per cent of pre-school children to be stunted (below 80 per cent of expected height for age) (Quinn 1986), whilst in Mozambique, limited surveys suggest stunting in the range of 35–55 per cent (p.436) in rural areas. In other SADCC countries, the available indicators show levels of stunting or low weight in the 25–30 per cent range.

Severe weight loss is found to reach levels of up to 11 per cent in children of recently displaced families in Mozambique and in areas of recent conflict (see 1987 Bulletins of the Nutrition Section, Ministry of Health, Maputo). Whilst this phenomenon can be quite extensive also in Tanzania, it is by contrast relatively rare in Botswana and Swaziland (usually 1–2 per cent). Surprising anomalies still exist, however, cautioning against aggregation to the national level for comparative purposes: despite a relatively low IMR and high per capita income and food production levels, widespread undernutrition appears to persist in Zimbabwe, and various surveys between 1980 and 1982, before the onset of drought, suggested that some 9 per cent of children were severely underweight (data summarized in Davies and Saunders 1987). There is little evidence of a consistent pattern in nutritional trends in Zimbabwe since independence, although young child rates of mortality and perhaps morbidity have been reduced in association with the rapid expansion of primary health care programmes and implementation of relief measures in response to drought reviewed above.

As posited by Davies and Saunders, and suggested by the existence of widespread undernutrition in other relatively high-income SADCC countries, selective approaches to welfare, including health care, are not sufficient to address the root causes of these and other aspects of child (and vulnerable group) ill-health. As described in the next section, such causes are found also in the systematic patterns of household vulnerability, dependence, and income insecurity that persist in wide areas of all SADCC countries.3

4. Vulnerability of Sadcc Households to Food Deficits, Drought, and Dependence

4.1. National and Household Food Security

A clear characteristic of food and especially cereal production in the region is that of large year-to-year fluctuations in both marketed and, to a lesser extent, total output, associated with apparent medium-term declines in production per capita. At the height of the recent droughts, in 1983, total cereal production in the nine countries fell to some 7.3 million tonnes as against pre-drought levels of 8–10 million, and in Botswana, Lesotho, and Mozambique (the latter partly due to war), cereal imports exceeded estimated levels of internal production. At the country level, Zimbabwean maize production fell by over 70 per cent between 1980 and 1983, turning the group’s major cereal exporter into a food aid recipient (Morgan 1986 b).

(p.437) By 1985, however, the group as a whole had recovered an overall surplus production position relative to basic estimated requirements; this surplus was largely due to Zimbabwe and Malawi cereal output, which formed about 45 per cent of the total. Through the intermediary of foreign-exchange backed donor purchases (or swap arrangements providing wheat for maize), these two countries became major suppliers to relief programmes in the neighbouring States, particularly Mozambique. This is one of four countries in SADCC persistently unable to meet internal needs for staple cereals. Like Angola, the major reason is that of war and destabilization, whilst Botswana and Lesotho are able to compensate for their increasing structural food deficits through commercial imports based on diamond and migrant labour earnings, topped up with food aid receipts targeted mainly to nutrition-oriented programmes.

War has therefore now superceded drought as the primary factor threatening the region’s food supply position: progressive collapse of cereal and probably cassava production in both Mozambique and Angola, and growth of the non-agricultural population (largely the displaced and the urban dwellers) reliant on food aid imports to meet basic supplies, have led to a current annual food aid requirement of more than one million tonnes for the two countries. This is associated with the war-related collapse of foreign exchange availability, and, in the case of Angola, with falls in the price of oil, its main export. In 1987/8, Malawi also became a large-scale recipient of food aid, mainly due to the presence of some 400,000–600,000 Mozambicans within its borders, displaced by insurgency.

Whilst the highly precarious access to food for households in the war-affected parts of the region is clear, indicators of such access in other areas are more difficult to establish, and are not necessarily provided (as seen in the review of nutrition data above) by degrees of food production self-sufficiency at national level. Although measures of calorific supply per capita against basic requirements are some guide, and show even before the droughts and intensified warfare of the mid-1980s that most SADCC countries had serious shortfalls (Table 10.9), they provide little sense of disaggregated and household level food access.

Table 10.9. Some estimates of calorie availability per capita in SADCC countries (per day)

FAO estimate of daily calorie needs per capita

Average calorific supply per capita

Calorie supply 1981–83 as % of estimated need

1979–80a

1981–3

Angola

2 350

2 093

2 204

94

Botswana

2 320

2 242

2 180

94

Lesotho

2 280

2 248

2 313

101

Malawi

2 320

2 185

2 237

96

Mozambique

2 320

1 885

1 725

74

Swaziland

2 320

2 510

2 618

113

Tanzania

2 320

2 004

2 502

108

Zambia

2 310

1 981

2 000

87

Zimbabwe

2 390

1 944

2 154

90

(a) Quoted from FAO Food Balance Sheets in ‘SADCC Regional Food Reserve, Pre-Feasibility Study, Main Report’, Technosynesis, Rome, Jan. 1984.

Note: Supply is estimated after allowance for usage for seed, animal feed, wastage and industrial purposes, as well as stock changes. No account is taken of the possible extent of maldistribution of access to food between families, population groups, etc. Given this, and other factors such as pricing, special distribution schemes, etc., these figures should be taken as broad indicators only.

Source: Calculated from FAO Food Balance Sheets (FAO, Rome, mimeo).

The need to consider overall incomes, multiple sources of food supply, access to services which help increase the efficiency of intake and use of food (such as elimination of internal parasites through health and sanitation programmes), and possibilities for fallback strategies such as migration and livestock sales in times of stress, vastly complicate analysis of household vulnerability to food shortage. That such elements are important, in a region where some 25 million or more people live in normally food-deficit or drought-prone rural areas (see Fig. 10.4; also Morgan 1986ft), can be briefly illustrated by two countries:

  1. (1) Malawi was cereal self-sufficient and produced modest export surpluses during much of the 1980s. Most cereal production came from small (p.438) farms. Yet, partly due to maize pricing policies, and the small size of farm holdings, farming families continue to rely heavily on off-farm income, including remittances from South Africa and from seasonal work on commercial estates. As seen earlier, infant mortality and malnutrition rates are among the highest in the region.

  2. (2) In Botswana, crop production plays a small role in the national economy and rural incomes rely heavily on casual employment, remittances, and, for a minority of households, sales of livestock. Although repeated droughts have reduced agricultural sector incomes, large-scale state transfers financed from diamond revenues and aid programmes have not only compensated for the effects of drought, but have also greatly widened access to health, water supplies, and other services, with notable effects on welfare outcome indicators.

Social Security in the SADCC States of Southern Africa: Social-Welfare Programmes and the Reduction of Household Vulnerability

Fig. 10.4. High-risk food security areas in SADCC member States.

4.2. Income and Migration Dependence

The contrast drawn above is between countries each with low rural household incomes, one with a strong State redistribution policy and the means to effect (p.439) (p.440) it, through both ‘development’ and ‘emergency’ programmes, and the other without these conditions. Even if through the medium of the State, therefore, a general relation can be assumed in broad terms for SADCC households between total household income access and degree of food security, or, alternatively, of vulnerability to collapse of food entitlement. The problem of vulnerability is related closely also to the instability of the various sources of income themselves, within which those deriving from the State, directly or indirectly, can play an important buffering role.

Whilst droughts and other natural disasters, as already noted, have powerful short-term, and often more lasting, effects on household incomes, a further area of vulnerability in wide areas of SADCC is the reliance of families on remittance earnings from migrant workers in South Africa. Recruitment of foreign workers by the South African mining and farming sectors has been a feature of the region since the late nineteenth century. Modern-day Lesotho, Botswana, Swaziland, southern Mozambique, and Malawi provided successive generations of workers to the dominant economy, under the supervision of the British and Portuguese colonial regimes, a process which had profound consequences for the semi-subsistence agricultural sector and assisted the penetration of the cash economy.

As dependence on incomes derived from migrant earnings has persisted, States in the region facing constraints on economic growth and employment generation also remain substantially reliant on this migration as a source of jobs and of foreign exchange. The most prominent cases are Lesotho and Mozambique, where remittances form the largest single source of export earnings. In Mozambique, the reductions by South Africa of the numbers of mine-workers recruited has already had serious consequences for household incomes and national revenues in the period since 1975. By late 1987, the number of Mozambicans employed in the South African mining sector had fallen to some 40 per cent of 1975 levels. This fall in recruitment has deprived the families of mine-workers, concentrated in the largely semi-arid three southern provinces of Mozambique, of basic, long-established means of economic survival. Income and job losses have been further increased by South African repatriation of unregistered Mozambican migrant workers.

Lesotho presents a picture of even greater dependence on, but also of sustained and increasing benefit from, migrant workers in South Africa. With 80 per cent of households relying directly or indirectly on employment and incomes generated by South African mine employment, and over 85 per cent of foreign exchange earnings derived from remittances, the number of Basotho mine-workers in South Africa has shown only slight declines in the 1980s, whilst average earnings grew some 16.5 per cent in real terms between 1979 and 1985 (Food Studies Group 1988). With cereal production showing a decline over the same period, the vulnerability of Lesotho households to the (p.441) effects of possible changes in South African labour import policies has reached extreme proportions.

Other countries are now far less vulnerable, although Botswana and Swaziland (with about 25,000 and 17,000 migrants respectively) still have significant proportions of their work-force employed in South Africa. Malawi also retains a similar number of migrants in that country, much reduced from levels of the early 1970s. All national States, except perhaps Botswana, would face difficulties in helping affected families to ease the transition to other sources of income, were the bulk of these migrants to be suddenly repatriated.

4.3. Characteristics of the Rural Poor

Little comprehensive information is available on the nature and scale of rural poverty in the SADCC region, and the data base on levels, sources, and distribution of incomes is particularly poor. The wide differences in per capita GNP levels already noted give no more than a first indication of family-level welfare, tending to mask, for example, the relatively good access of Tanzanian rural households to basic services, and understating heavily the levels of rural poverty in countries like Botswana and Angola, where much of national income arises from enclave extractive industries which generate little local employment.

Rural areas in SADCC are generally characterized by low yields, by world standards, in arable farming, ranging from under 150 kg to slightly over 1 tonne per hectare for cereals. Use of fertilizers and improved seed is confined to a minority in all countries (Swaziland and Zimbabwe are growing exceptions), and in the most populous countries (Mozambique, Angola, Tanzania, and Malawi) cultivation is almost exclusively by (female) hand, limiting areas which can be planted. Chronic shortages of handtools exacerbate this problem. Land access is overall very high, a major advantage for the poor in the region, but is becoming limited in some States, with Malawian farm size averaging about 1.2 hectares, and some 20 per cent of rural Lesotho households estimated to have no land access. In Botswana, poverty of soils and cattle damage to crops are major obstacles to production, whilst in Zimbabwe the historically determined distribution of land between large commercial farms and the mass of ‘communal’ farmers—concentrating the latter in areas of poorer soils and less reliable rainfall—remains fundamentally unchanged.

Access to livestock, particularly cattle, is low in many areas, and has been reduced by droughts and decimated (in Mozambique) by warfare. Where average cattle numbers per family are high, as in Botswana, distribution tends to be highly uneven, and least favourable for households managed by women. Such households in turn are common (at least 25 per cent in rural areas in most countries, and as high as 40–60 per cent in Botswana and Lesotho), due in significant measure to the labour migration phenomena already described.

(p.442) Resulting from these circumstances, a substantial proportion of rural households are unable to obtain a basic means of livelihood from agricultural activities alone, and are obliged to seek off-farm sources of income. Many of these are seasonal in nature (weeding, bird-scaring, grass collection), yield low returns (sewing, handicrafts), and are liable to sudden downturns during droughts. For poor rural families, therefore, the effects of local economic and climactic stress are often highly cumulative and interactive, leading to sudden income collapse.

Internal migration opportunities exist in most countries, but tend to be associated with the perpetuation, rather than alleviation, of poverty. Adequate data exist in Botswana and Zimbabwe to indicate that some of the most widespread child malnutrition is found among families of employees on freehold ranches in the former, and commercial farms in the latter.4

Further groups to be considered as characterized by poverty are those living in remote areas: the Lesotho Highlands, the Kalahari in Botswana, western Zambia, and Tanzania, and interior areas of Mozambique and Angola. Such areas are not only relatively lacking in social-service access and tend to have lower than average incomes and higher than average incidence of child malnutrition, but are often associated with low agricultural potential and unreliable rainfall.

Such surveys on rural incomes as have been undertaken bear out this picture of a large minority of households existing in a structurally reinforced condition of poverty. The Botswana Rural Income Distribution Survey (RIDS), possibly the most comprehensive of its kind, has already been referred to. A study of one of the largest villages in the country in 1984 (Mahalapye) illustrated further the high dependence of lower-income families on transfers compared to production, and also the seasonality of incomes (UNICEF Botswana Situation Analysis). In Swaziland, the Rural Homestead Survey of 1978/9 found that 56 per cent of total income came from cash earnings, mainly from wage employment; again, the poorer households were more dependent on remittance transfers, which were highly erratic in nature (UNICEF Swaziland Situation Analysis). The UNICEF Lesotho Situation Analysis identifies the ‘rural poor’ as the members of the 67 per cent of rural households which in 1985 had no migrant, the 20 per cent who had no land, the 60 per cent which were female-managed and the 47 per cent of the farming households which had no livestock. Given the great overlap between these categories, the picture of poverty that emerges is of widowed or abandoned mothers with young children owning few productive assets and with low-yielding, unreliable sources of income.

Estimates of the extent of rural poverty in other SADCC countries are even (p.443) harder to come by. Very few rural Angolans and Mozambicans probably now escape conditions of serious poverty, other than those able to gain employment on State farms, where at least some degree of security of income and food supply is present (even rural civil servants experience great difficulty in the purchase of food and basic goods). In the current circumstances, probably only a small proportion of farming families are relatively self-sufficient in food, and manufactured products are scarce. Large-scale displacement due to banditry has caused collapse of family production systems. In Tanzania, where conditions both in terms of family farm production and access to social services are relatively much better, and cash earnings are derived from a number of export crops (although consumer goods are also scarce), it has been estimated that some 25–30 per cent of rural people remain below an assumed poverty line. In Zambia, an ILO study of basic needs in 1981 estimated that 80 per cent of rural households were ‘poor or very poor’ based on an estimation of minimum needs and projections from the 1974–6 Household Budget Survey. Perhaps more importandy, the 1980 Census in Zambia found almost 31 per cent of households without an economically active member (see UNICEF Situation Analyses of Tanzania and Zambia).

4.4. Urban Poverty in SADCC

Consideration of urban poverty also proceeds from an inadequate data base, with little reliable information available on the generally fast-growing urban and ‘urban marginal’ (or peri-urban) settlements in the region. Again, Botswana provides one of the better sources. The national Nutritional Surveillance system has over the last few years found consistently above average levels of undernutrition in children attending health facilities in low-income and recently settled urban areas. These phenomena may be closely related to migration and economic stress caused by the continuing drought. Earlier information, from a poverty datum line (PDL) study carried out by the government of Botswana in 1976, found that some 36–47 per cent of urban dwellers in the four main towns fell below the PDL, although average monthly household income exceeded it. By contrast, a PDL survey in Maseru, capital of Lesotho, carried out in 1984, found that some 60 per cent of households fell below the PDL, including all those with ‘single person breadwinners’ (UNICEF Botswana and Lesotho Situation Analyses).

Available information on poor urban households indicates that their income derives from casual sources, short-term and unskilled employment, as well as activities such as hawking, marketeering, and the running of ‘shebeens’ (beer parlours). The picture is similar to that of the rural poor in some respects, with the emphasis more on the periodic selling of goods rather than (but not excluding) labour. Such incomes are clearly vulnerable to sudden downturns in the local economy. Added to the high density and often inadequate servicing (p.444) of newly settled urban areas, and the shortages of low-cost housing, the low and fluctuating levels of income for household without full-time wage earners are associated with widespread problems of health and sanitation.

Those families with members in full-time employment are also not isolated from economic stress. In Zimbabwe, Mozambique, Zambia, and Tanzania, real urban minimum wages have been eroded by rising official prices for basic food items and consumer goods (but see subsection on trends below). In Tanzania this process of reduction of the real minimum wage value started as long ago as 1974, and has probably been accelerated since the removal of maize meal subsidies in 1984. In Zimbabwe there were initial gains in the early post-independence period, which were somewhat eroded by wage freezes and removals of subsidies beginning in late 1982, as well as by general inflation. Real wages in Zambia have been declining since about 1977, in association with various adjustment measures taken in response to the collapse of copper export earnings in the mid-1970s. The reduction in purchasing power amongst low-income wage earners in these countries doubtless reduced in turn the demand for the services provided by the informal sector.

Net reductions in urban employment have not yet occurred on a large scale in SADCC urban areas except, to some extent, in Zambia. However, the types of economic adjustments being adopted by some countries in association with IMF/World Bank loan packages will certainly involve temporary losses of urban employment, particularly in the State and parastatal sectors; it is not yet clear how households would be expected or assisted to cope with the consequences of this.

4.5. Trends in Household Incomes and the Impact of ‘Adjustment’

Four distinct factors mentioned above combine to indicate a relatively gloomy short-term outlook for household incomes in the SADCC region, and for the welfare of vulnerable groups in turn, despite advances made in the coverage of welfare services. First, the drought conditions of the early 1980s which affected most of the region not only continue in certain areas, but probably worsened income differentiation and permanently lowered the productive capacity of some of the very poor.5 Secondly, growth of national income is in most SADCC countries not now exceeding that of population, indicating no prospect of general, widespread increases in welfare, particularly where little scope exists for redistribution of land, income, or State spending. Thirdly, external factors including low prices for major exports (oil in Angola, sugar in Swaziland, copper in Zambia, and so on), rising transport costs for imports (due partly to sabotage of the Mozambican sea routes), increased defence commitments, and high debt service obligations, have combined seriously to (p.445) undermine the budgetary position of several governments. This in turn has led to cutbacks in expenditures on areas of importance to the poor, and to currency depreciations which have further disadvantaged urban consumers. Economic decline in Zambia, Tanzania, and Mozambique has reduced the efficiency of agricultural services in rural areas, leading to virtual breakdowns by the mid- 1980s in marketing systems and the timely delivery of inputs. Although the new emphasis on price incentives to smallholder farming associated with structural adjustment policies may enable real farm incomes to increase for certain strata of net surplus producers, even these gains may prove short-lived: real prices for maize have been falling back in Zimbabwe and Malawi following increases in the early 1980s once these countries were faced with surpluses that could only be stored at high cost or exported at a loss. Finally, the massive destabilization of the two former Portuguese colonies has left economic and social turmoil and instability that will probably take decades to overcome, whilst the burden of these costs shows no sign of diminishing, and may spread to other countries as well.

This discussion, perhaps controversially, does not locate a major responsibility for recent depression of household incomes or access to welfare services in the region with the various ‘structural adjustment’ programmes now in force. It assumes that the important questions are related to the management of the adjustment process, seen to be inevitable in some form, rather than whether it is consciously undertaken at all. In a war-torn economy such as Mozambique’s, the impact of expenditure cuts in social sectors, and of the reduction of credit provision to State enterprises leading to employment losses, has almost certainly been minor since the introduction of such measures in January 1987 in comparison with the continued closure of rural schools, health posts, and agricultural units due to direct attacks, and with additional employment losses for Mozambicans in South Africa. Urban consumers with semi-indexed wages may be at least no worse off as increases in official prices divert goods from illegal channels, thereby bringing down prices facing consumers overall. Meanwhile, adjustment programmes have generally been associated, as in Mozambique, with greater inflows of international resources for rehabilitation of productive sectors, and for commodity imports of raw materials and basic goods, although not necessarily at a rate quick enough tolerably to offset the effects of reductions of demand.

The effects on sustainability of welfare services of ‘adjustment’, however, may be more difficult to justify, particularly in the absence of major distortions due to destabilization and resource diversion to defence. The ‘alternative approach’ developed by Cornia, Jolly, and Stewart in Adjustment with a Human Face is based on the argument that negative effects on nutritional levels, educational access, and low-income households are likely to occur, and to prejudice medium-term economic growth, unless careful intervention measures to protect the more vulnerable are designed and implemented during (p.446) the transition period whilst sustainable growth is being restored. Specifically, adjustment programmes concerned solely with the rectification of domestic and external monetary and trade balances are shown not necessarily to result in the restoration of economic growth. Additional resources need to be dedicated to the avoidance of steep declines in levels of household consumption and health and nutritional status, both through targeted welfare interventions and through production sector investment and rehabilitation, if such adjustments are to be sustainable. Very often, the timing of implementation of such measures has been a crucial and neglected policy question.

Most countries in SADCC have already in place some ‘targeted welfare interventions’, usually with nutrition-related objectives. These, particularly as concerned with temporary income-creating employment schemes and food distribution programmes, are discussed in more detail later. Measures to ‘stretch’ available resources in the social sectors, by increasing cost efficiency and equity in service provision, furthering auto-financing by selective user fees among groups which can afford to pay, and encouraging self-management by communities, are less common. A potential benefit of adjustment programmes, which may be overlooked, is the provision of an environment of greater incentives for social-service managers and decision-makers to consider such possibilities, which may promote greater long-run sustainability of services.6

5. Choices in Strategies for Welfare Protection to Vulnerable Groups Under Stress: two Case-Studies

The twin devastations of war and drought, their effects on vulnerable groups and families, and the responses by States whose own access to resources is differently affected by these circumstances, are illustrated by the cases of Mozambique and Botswana. Whilst the latter country, although affected by severe drought in successive seasons from 1981/2 to 1986/7, has abundant national access to financial resources by which to effect compensatory measures, the former has had to make more careful and difficult choices among types of programmes, possible beneficiaries, and limited means of implementation, given an economy in crisis. In doing so, it has been unable to rely on the economic mechanisms that have continued to function in Botswana, ensuring there the physical presence of food and consumer goods in local shops even in remote areas, and providing the possibility of using private-sector transportation, (p.447) storage, and food-processing facilities. The main common factors facing the two countries in designing responses to ‘emergency’ have been simply the persistent and structural nature of the vulnerabilities that lead to repeated crises in rural society, and the scarcity of relevantly trained and experienced nationals who can be mobilized for programme implementation.

5.1. Botswana: Addressing Drought and Structural Poverty

Although Botswana is ecologically a semi-arid land where arable farming is marginal in the best of years and livestock, particularly cattle, provide the main source of rural incomes, about 70 per cent of rural households engage in rain-fed crop production. By contrast, 45 per cent or less of such households are owners of cattle (RDC 1985). Rural income surveys quoted above have shown a very high dependence among poorer families on casual employment and remittances. None the less, the loss of self-provisioning capability and opportunities for seasonal work on fields caused by drought conditions, particularly repeated over several seasons, leads to an intensification of poor health and nutritional status. Whilst between 1978 and 1984 cereal production fell from about 50,000 tonnes to some 7,000 tonnes per year, malnutrition rates in children under 5 rose from a yearly average of 25 per cent to 31 per cent in the respective years (percentage below 80 per cent of expected weight for age, based on monthly weighing at all health facilities). Income from crops per household actively engaged in agriculture fell from about 275 Pula (about US$138) in 1981 to some 39 Pula in 1984. In this period of the early 1980s, overall cattle numbers were reduced from almost 3 million to about 2.3 million, with losses concentrated disproportionately in the smaller herds, namely those of lower income households (see Morgan 1986 a).

None the less, Botswana during the period 1984–6 managed to reverse the deteriorating nutritional status of children associated with the drought (Fig. 10.5), even improving on the original pre-drought position (although changes in the recording system at health units in 1985 throw some doubt on the extent of this). It was also possible to avoid a mass exodus from rural areas to the prosperous urban centres which had sprung up around mining, meat-processing, and civil service activities, and which remain largely free of ‘slum’ areas. Equally, despite repeated harvest failures over a period of several years, the numbers of households planting crops fell only some 25 per cent at most, and recovered to near pre-drought levels by 1986. Starvation, even among extremely isolated communities, was entirely averted during the droughts.

Supported by a budget surplus and extensive foreign exchange reserves, plus a level of food aid donations per capita which stands at one of the highest in Africa (in 1983/4, for example, between 24 and 30 kg per person; see RDC 1985), the Botswana government has run a nationwide Drought Relief Programme since 1981 which simultaneously addresses income losses experienced (p.448) by rural households, targets food supplies to non-able-bodied groups presumed most vulnerable, and provides the means for agricultural recovery in the small farm sector. The basic elements of the programme have been:

  1. (1) Distribution of free food as take-home rations on a monthly basis, through the extensive national network of health facilities, to all pregnant and breastfeeding women, all children under 5 years, and registered destitutes and tuberculosis patients (6 kg per month of cereals plus cooking oil).

  2. (2) Maintenance of a school feeding programme for primary school children.

  3. (3) Therapeutic feeding on a daily attendance basis to actually malnourished children at health facilities.

  4. (4) A special trucking operation to supply maize to roughly 2 per cent of the population classified as ‘remote area dwellers’, in areas which tend to show the highest rates of malnutrition among children.

  5. (5) Public works programmes with payment on a cash basis for days worked, employing between 50,000 and 70,000 people at some time during the year. This has involved work in labour-intensive projects selected by village committees, such as dam construction, building of traditional houses for extension workers, of communal kraals, and of (p.449) local roads. A special programme has employed some 4,000 women per year to handstamp the sorghum used in the school feeding programme.

  6. (6) Provision of seeds (usually 20 kg per family per year) to farming households, as well as grants to allow for local hiring of tractors and of labour for clearance of fields before ploughing. In 1985/6, a weeding subsidy was introduced (these forms of assistance now form a major component of Botswana’s Accelerated Rainfed Arable Programme aimed at promoting post-drought recovery in the late 1980s).

A series of additional measures financed within the Drought Relief Programme included the expansion of water-drilling and maintenance capacity at national and district levels, and the purchase and distribution of stock-feed and vaccines for cattle farmers. Cattle-marketing possibilities were extended by grants to district councils for purchase of animals to be used in school-feeding.

Although the programme at its height has taken up some 12 per cent of the government’s development budget (Holm and Morgan 1985), its implications in terms of additional personnel and institutional resources have been very light. Implementation of the measures described above has been based almost entirely on existing health, agricultural extension, and local council service structures. The food distribution programmes, which use village-level health posts and schools as final distribution points, are based on an expanded version of an earlier nutrition promotion project supported by the World Food Programme. The public works schemes depend largely on community-level supervision and very limited technical support from the district councils which effect payment to workers once a month. The main additions to available cadres for these measures, which amount to a significant extension of Botswana’s Welfare State, have been the recruitment of a dozen ‘Western’ volunteers as technical officers for public works, and of a similar number of district-based drought relief co-ordinators, usually retired civil servants.

In designing this programme, however, the Botswana State decided to take full advantage—drawing on the experience of earlier, less effective .drought relief campaigns in the late 1970s—not only of its existing local institutions, but of national monitoring systems gradually being developed in the fields of nutritional surveillance, agro-meteorology, livestock and grazing conditions, and food availability (Morgan 1985). The early warning system, which was established in early 1984 based on these data sources, initially as a review committee making recommendations for resource allocation to a parent Inter-ministerial Drought Committee, was able, in compiling short-term indicators, to provide the beginnings of a national targeting mechanism for such allocations.

Furthermore, the State was not ideologically constrained from contracting out certain functions involved in the implementation of programmes to private enterprise: indeed, this was increasingly seen as attractive to political decisionmakers, often owners themselves of trucks and tractors eligible for hire. The (p.450) country’s (then) only maize mill, South African owned, received a steady flow of contracts for the milling and mixing of maize and milk powder for clinic distribution/In a perhaps healthy division of labour, rural women and, for urban areas, co-operative millers, continued to process sorghum for primary-school feeding.

Along with its political attractiveness to richer farmers and entrepreneurs, the programme—distributing food to over 60 per cent of the rural population as direct beneficiaries, seeds to almost all rural households, and employment to many others—has become a major election plank for the ruling Botswana Democratic Party (BDP), to the point where the (nominally socialist) opposition complained during the 1984 election campaign of unfair buying of votes. Given the opposition’s dominance of the small number of urban constituencies, and the BDP’s reliance on the majority rural electorate, there will be a strong political incentive, based on a coincidence of economic interest between the rural elite and the rural poor, for the government not to reduce drastically the scale of these welfare measures as climatic conditions improve. This will be the case even if donated food aid volumes decline, assuming that the budgetary position of the State remains in sufficient surplus to allow it to compensate.

As implied already, Botswana’s budget surplus and foreign exchange reserve positions, coupled with donors’ willingness to provide relatively high levels of food aid to an efficient distribution system, have enabled it to avoid certain hard choices in the design and targeting of its drought relief measures. At certain points in the budgetary process, the only apparent limitations on resource allocations in the past few years have been the implementation capacities of local government institutions and the ability of communities to identify projects not evidently absurd. The public-works programmes have proven an efficient and speedy way of redistributing mining sector State revenues directly to the rural poor in cash, thus bolstering their incomes in times of stress. The setting of a daily wage-rate high enough to provide a meaningful return but low enough not to attract members of relatively well-off families to manual work on the schemes has proven an administratively simple targeting device (from personal experience of these projects, over 80 per cent of participants have normally been women). Similarly, women from well-off rural families are less likely to join the monthly queue at health posts to have their children weighed and collect a take-home ration. The apparent effectiveness of the ‘self-selecting’ mechanisms used is partly assisted by the fairly clear wealth division prevalent in rural Botswana, between a large-herd-owning minority on the one hand, and the mass of few-or no-cattle-owning families on the other. A further contributory factor has been the significant involvement in local drought relief operations of village-based organizations—for example, the Village Development Committees in the identification of projects and selection of participants, and the Parent-Teacher Associations in school feeding and food preparation.

(p.451) Economically, however, the programme has inefficiencies that a less well-endowed national State would have been forced to address. Little productive or permanent infrastructure has resulted from the US$5 million (equivalent) or more spent yearly on public works under drought relief since the early 1980s, and labour productivity on most such projects has been low. Furthermore, the opportunity to use these resources as an initial community investment in what might later develop into much needed additions to permanent rural employment has largely been missed. It is also argued that disincentive effects on agricultural production have resulted from the high levels of free food distribution (an assertion unproven and belied by the only small reduction in the number of households planting) and that—more plausibly—understanding of health facilities among the rural population as a service for disease prevention has been undermined by their use in a food distribution role. Finally, the programme has done little to address the structural aspects of rural poverty which tend to provoke a high demand for relief intervention in drought years: the very limited access to cattle and draught animals, which form the main repository of rural wealth, and high reliance on remittances, casual employment, and low-yielding, high-risk arable farming. It is not apparent either that Botswana’s ‘Drought Recovery Programme’, coming into effect following generally good rains in 1987/8, will directly address these problems.

5.2. Mozambique: Management of Acute and Structural Crisis

Mozambique provides as different a picture as can possibly be imagined from Botswana for two countries in such geographical proximity. Emerging from a highly repressive colonialism and inheriting the remains, often sabotaged, of a notoriously inefficient State and formal sector economy, the Mozambique liberation movement FRELIMO continued to be involved with a regional war against white minority rule, first in Southern Rhodesia from 1975–80, and later, until it reduced its role under the 1984 Nkomati Accord, in support of the African National Congress in South African. With the flight of about 90 per cent of the Portuguese settler population in the mid-1970s, and as a result the virtual disappearance of the commercial sector and managerial class which had been dominated by it, the State was obliged (whether eagerly or not) to assume much of the responsibility for the running of the country’s economic enterprises, large and small. This it was very inadequately equipped to do. None the less, after the initial disruptions of the independence process, modest economic growth from very low levels of per capita income were achieved until 1981. At this point, widespread, externally supported insurgency began to affect several provinces, eventually extending its destructive effects to all districts in some form. The strategy of this insurgency focused initially on the closure of the main transport routes to the sea used by Mozambique’s SADCC neighbours, and extended to attacks on all types of productive units as well as (p.452) the social and commercial sectors. Coupled with production falls due to drought, the Mozambique economy contracted by some 33 per cent in real terms between 1981 and 1985, with modest growth resuming thereafter. Meanwhile, South Africa intensified its economic sanctions against Mozambique, reducing its use of Maputo port for international trade by over 80 per cent from 1975 to 1987, and recruitment of Mozambican mine workers by about 60 per cent over this period—thus intensifying the shortages of foreign exchange and problems of employment and income in southern Mozambique.

Without a widespread and well-endowed State structure to offer social protection, the effects of the intensified crisis on rural households in Mozambique became dramatic by 1983/4. In that period, it is commonly accepted that some 100,000 people may have died due to drought, coupled with inaccessibility and displacement caused by insurgency, and the failure of international food aid donors to provide timely support. Whilst since that time famine has not been widespread, displacement and loss of household assets and the family production base have become commonplace, with probably several million Mozambicans forced to flee their home areas. At the time of writing, roughly one million citizens are estimated to have been recently displaced within the country, and a further million are in neighbouring countries (adding to demands on Zimbabwean, Zambian, Malawian, Tanzanian, and Swazi State welfare systems in the process).

The household level crisis is indicated, as described earlier, by extremely high levels of infant and child mortality, as well as extensive chronic malnutrition (stunting) in children, high levels of acute malnutrition (wasting) in recently displaced communities, increasing incidence of low birth weight, and outbreaks of preventable infectious diseases. In contrast to Botswana, the largely externally induced crisis has not only affected welfare status at family level, but has drastically eroded the already weak financial and institutional capacity of the state to mount responses to mitigate the effects on the most vulnerable.

Alliances with what remained of the private sector after independence in operational terms have also been largely unavailable as options to assist responses. Although ideological considerations deterred this for several years, the main impediment has been the destruction of rural shops and breakdown of private transport capacity due to extreme scarcities of spare parts, servicing, and fuel.

The State has been faced with a further complicating factor: the need to intervene to secure at least minimum food supplies for urban populations, following the collapse of internally marketed food production and the inability of private channels to provide significant imports (Fig. 10.6). By 1984, the main towns (containing 2.0 to 2.5 million people) were largely dependent on food aid deliveries, supplied through rationing systems, to meet basic food needs, particularly for cereals, salt, sugar, and cooking oil. This presented the State with the constant need to make allocative decisions with inadequate (p.453) resources, between the urban population with some levels of local currency purchasing power but no food to purchase, and the rural population extensively affected both food and income shortages due to the emergency. Whilst local transport costs for supplying the urban populations are in most cases relatively low, the administrative costs of operating rationing systems for hundreds of thousands or urban families are high, not least given the State’s limited managerial capacity.

Social Security in the SADCC States of Southern Africa: Social-Welfare Programmes and the Reduction of Household Vulnerability

Fig. 10.6. Cereal market supply in Mozambique, 1956–1986. Source: Food Security Department, Ministry of Commerce, Mozambique.

Despite these costs, urban rationing is administered fairly and effectively in Mozambique when food is available. In the rural areas, delivery of all basic services and needs, ranging from food to health care, clean water, and clothing, is inhibited by a wide range of physical constraints. Firstly, negligible investment in transport infrastructure during the colonial period, except in the export ‘corridors’ serving neighbouring countries, has not yet been compensated for. This implies reliance on a complex combination of transport means, including road, rail, air, and sea, depending on the locality. Secondly, poor telephone and radio connections make co-ordination of delivery systems more difficult. Most importantly, the danger of attacks on road and rail traffic, including relief deliveries, necessitate travelling in convoy with military protection in many districts, a process which is both time-consuming to arrange and usually costly in terms of the value (if not the benefits derived from) the commodities delivered.

Social-service provision has been reduced to an extraordinary degree by systematic destruction of physical units, looting of supplies in stock, and (p.454) attacks on state employees. Some 31 per cent of rural health facilities (Fig. 10.7) and 36 per cent of primary schools existing in 1981 were by 1988 not functioning for these and related reasons, depriving some 2–3 million people of access to primary health care, and roughly half a million school age children of education. Also serious has been the impact on service provision of reductions in the real value of the State budget both overall and for the functioning of social-sector programmes—although this has been partly compensated for by increased donor assistance in kind (medical supplies, school books, and so on).

Social Security in the SADCC States of Southern Africa: Social-Welfare Programmes and the Reduction of Household Vulnerability

Fig. 10.7. Numbers of primary-level health units functioning and not functioning due to closure in Mozambique. Source: Noormahomed and Cliff (1987).

The response of the Mozambique State to the increasing pressures generated by the emergency, and the willingness of international aid agencies to provide resources under an ‘emergency’ label, was initially modest. A national relief agency, the Department for Prevention and Combat of Natural Calamities (DPCCN) was established under the Ministry of Co-operation in 1981, in response initially to problems caused by floods and cyclones. However, it took several years before a reasonable logistical delivery capacity was established by DPCCN reaching to district level, and the management resources provided to its national and provincial offices were initially poor. The DPCCN also failed to establish regular operational links with other sectors of government and with donor agencies, although co-ordination meetings were held monthly with non-government organizations (NGOs). Until 1987, DPCCN had developed (p.455) no planning information or monitoring capability, nor had it recognized a need to do so. The concentration on delivery of relief goods without attention to coordination, establishment of priorities, and development of a ‘prevention’ or post-emergency strategy, was consistent with the orientations of some of the agencies supplying goods for the relief effort.

In the first half of 1987, however, the government adopted a different approach, one much more consistent with the economic adjustment programme adopted at the beginning of that year. Emphasis in strategy for emergency response began to be laid on the need to integrate this at the community level with medium-term rehabilitation and recovery measures, in order to avoid a condition of permanent crisis. Interventions were made to promote productive activities in areas of resettlement of displaced people, and low-cost rehabilitation of social-service facilities began to be undertaken, to provide a socio-economic base for rural recovery.

As a result, a multisectoral approach to the emergency became essential, and this resulted, in terms of institutional arrangements, in the creation of a National Executive Commission for Emergency (CENE) with co-ordinating responsibilities for the rehabilitation programmes which now began to be developed in the technical ministries responsible for social services and rural production. During 1987, Provincial Emergency Commissions were established along similar lines, under the political authority of the governors (with ministerial rank). At both national and provincial level, DPCCN’s role was broadened to include not only the distribution of food and relief items, but also to provide logistical services to the rehabilitation programmes initiated by the technical ministries. DPCCN was allowed to sell services to ministries and to private traders, thereby accumulating local currency which remained at the disposal of the emergency programme for support to rehabilitation projects.

Apart from the incorporation of multisectoral elements and medium-term perspectives into its emergency programme, the CENE has also begun to rationalize the involvement of by now large numbers of international agencies providing resources. NGOs and some multilateral agencies such as UNICEF and the EEC have been encouraged to focus on specific geographical areas, usually one or more of the 120 districts, to provide sustained institutional support to local structures, relief items for displaced families, agricultural inputs, and assistance in the recuperation of social services for the entire area. Although this ‘district adoption’ approach has both political and technical drawbacks, it simplifies the co-ordination tasks of national and provincial State structures, and encourages agencies to adopt multisectoral, recovery-oriented perspectives consistent with local conditions and security constraints.

A further development of the government’s emergency policy has been that of conscious limitation of free food distribution in rural areas, in order to avoid disincentive effects perceived to result. The CENE has requested food aid agencies to provide flexibility to enable food aid to be sold in any local situation (p.456) where people can apparently afford to buy it (which, if accepted by donors, gives the State more leeway in making adjustments in allocations between agricultural and non-agricultural or urban areas). Limitations on the periods for which, and areas within which, food can be distributed free have also been introduced, with the preference to be given to recently displaced people. Such people normally receive land allocations from the local authorities and agricultural inputs through the Emergency Programme, and are expected, in the absence of drought conditions, not to require food aid after the first or second harvest period. In this way, the government hopes to avoid a self-perpetuating food crisis in the countryside. The establishment of new settlements due to displacement is also taken as an opportunity to introduce irrigation on a cooperative basis, particularly in the vicinity of urban settlements (constituting so-called Green Zones). Further initiatives, aimed at the revitalization of local economies and trading relationships, have been undertaken using food aid as a ‘financial’ resource: donated commodities may be sold within enterprises to finance labour hiring, or exhanged with local producers for export crops such as fish and cashew.

A final characteristic of the Mozambique government’s emergency response has been its recognition of the limitations of its lower-level structures and unwillingness to place undue demands on them. Reporting obligations for food distribution, for example, extend only as far as the district level, and local decision-makers are left to determine priorities for lower administrative units. Health units have been required only to feed people who suffer from acute malnutrition or related illnesses, and require internment as a result. By simplifying procedures, and not overloading limited administrative capacities, the opportunity costs in diversion of personnel from other public sector programmes have been kept low. Conversely, very limited targeting of emergency programmes exists, beyond broad geographical areas and to concentrations of displaced people where these are apparent. Given the still fragmented nature of socio-economic data on emergency-affected areas and vulnerable groups, there exists little basis as yet on which systematically and explicitly to prioritize the allocation of emergency goods; however local political or State structures provide some articulation of popular ‘needs’, and can be effective in responding to them, within the limits of their human and material resources.

6. Some Generalized Conclusions

6.1. The Regional Context for Protection of Household Welfare

A predominant feature of welfare status in the Southern African region which has emerged from this review is that of insecurity of access at household level to necessities, and of household food production. As has been seen, this insecurity (p.457) is provoked by both historical and/or underlying factors and by welfare-threatening shocks such as drought and the effects of war. A further problem is the tenuous sustainability of State services in many countries of SADCC in circumstances of economic decline related to exogenous factors, adjustments of State expenditure levels and intersectoral allocation, and in the case of two (or possibly more) particular countries, warfare. Exacerbating these insecurities are the increases in demands on natural resources and State services resulting from an overall growth in SADCC population of the order of 3 per cent per year.

The common themes concerning the choice of strategies for welfare protection in times of stress or crisis, as discussed in the case studies and earlier review, suggest and emphasize (1) the need for co-ordinated, multisectoral strategies defined at national political levels; (2) the building over time of institutional capacity in the various levels of the State to implement extraordinary as well as ongoing welfare programmes, the former preferably on the basis of the latter; and (3) the range of options already available and for which experience exists within the region, to be adapted to local circumstances and the current nature of the crisis to be addressed.

The regional phenomena of structural poverty, chronic malnutrition, and shortages of trained personnel, features of the historical processes of colonial domination, labour migration, and dependency, in addition to the relative weakness both of the post-independence States and of rural communities as a result of these factors, suggest and necessitate a political element to the protection of the most vulnerable. The development of a popular politics around the process of creation of welfare-providing infrastructure, income-enhancing projects, and poverty-oriented programmes, by communities in alliance with the State, can, as shown in the early years of the FRELIMO movement and still in Mozambique today, help to reduce the need for capital-and import-intensive, and skill-demanding, programme designs, at least in the early stages of establishing or restoring mass welfare provision. In health and other areas, low-cost technologies to support this—oral rehydration treatment, child growth monitoring, community-built classrooms, and low-cost hand-pumps and latrines—can often be secured. As Botswana has shown, where the State is dependent for its political legitimacy on a regular electoral process, mass demands for welfare provision by an overwhelming poor and rural electorate are influential in determining resource allocation priorities. Furthermore, the extensive measures taken by the Botswana State to prevent economic collapse in rural areas in times of drought, indicate the potential of an effective alliance of interests between the different rurally based economic classes. By contrast, it seems no accident that those SADCC countries with the least populist, or most autocratic, political regimes are those which perform least well on a range of welfare achievements, both promotive and protective, taking into account levels of income per capita.

(p.458) Examples have been given of the relation between political and social objectives in the process of establishment of and access to welfare programmes in Southern Africa, as a region faced with the military and economic dominance of a white minority regime. These include Mozambique and Zimbabwe’s prioritization of community health and primary-education programmes; the targeting by insurgents in both countries of the very infrastructure created and personnel trained for those programmes; South Africa’s interference with the food and fuel supply flows of its neighbours (and threats to do so); and disparities in welfare-spending, child mortality, and life expectancy rates between legally designated racial groups in South Africa itself (Green et al. 1987: pt. 2). Such factors all attest to the existence of this context. As put by the governor of a Mozambican province, ‘we can only win this war if we help the population to build something worthwhile to defend.’ Such a statement contains a degree of relevance for most countries of the region, and suggests the need for further investigation of how the poor can be assisted to adapt to situations of displacement and military unrest.

6.2. Common Strategies and Approaches to Welfare Protection and the Reduction of Household Vulnerability

Against this background, the following common approaches are suggested as important for the development of more effective welfare protection measures in the region as a whole:

1. Prevention of crisis is related to the development of institutional bases in the State for the design and implementation of assistance programmes and the monitoring of vulnerability amongst families. Such bases have enabled the Botswana government to deliver effective compensatory measures for households during the 1980s’ droughts, without significant recourse to external agencies for implementation. The question relates also in part to the development of formal ‘early warning systems’, to inform decisions about such interventions; but also to the enhancement of analysis and understanding of the structural aspects of underdeveloped, dependent, and vulnerable local economies. ‘Early warning’ can equally be provided through the strengthening and harnessing of local political structures (such as FRELIMO Party’s village cells in Mozambique) or processes (such as Botswana’s liberal-democratic national and local electioneering which retains pressure on a rurally based ruling party). These in turn need to generate the political demands for effective action on behalf of the rural poor that are weak in some parts of the region.

2. The development of multisectoral welfare strategies and co-ordination bodies to implement them within the State, linked to the productive sector ministries and overall resource allocation and economic decision-making bodies, can help ensure a sustained approach to welfare requirements. Few countries in the region have even begun to achieve this (with the result, for (p.459) example, that nutritionists and their priorities remain isolated even within their health sector ministries, unable to influence choices in agricultural research, crop, or food-pricing policies). Again, Botswana provides something of an exception, incorporating ongoing nutrition sector, drought relief, and early warning activities into a National Food Strategy (RDC 1985), published as a ‘White Paper’ and incorporated, both as a strategic component and as a loose set of expenditure plans, into the Sixth National Development Plan (1986–91). The strategy is monitored by a working group, based in the Rural Development Unit of the central Ministry of Finance and Development Planning, which reports on progress and issues requiring resolution, to a group of top civil servants headed by the vice-president. Such a machinery ensures the possibility of bringing household food security and nutrition-related questions to high-level decision-makers in a coherent policy context.

3. Whilst cost-effectiveness considerations vary in their prominence from country to country, all states in the region still face some degree of shortage of trained personnel for the implementation of social-welfare programmes, and in most cases are particularly ill-equipped to provide the types of staff required for implementation of responses to crises. Strengthening of response capacity is often conceived, firstly, narrowly in terms of food provision (usually relatively easy to obtain from donors), and, secondly, in terms of financial outlays to complement delivery. Opportunities exist, however, for the process of extending welfare programmes as a development objective to incorporate routine training for the design and implementation of ‘emergency’ measures, partly by actually carrying out such measures on a smaller scale as part of a poverty alleviation programme. A range of options, depending on local economic and social circumstances, is available and tested for income supplementation in crises (food-for-work, consumer goods-for-work, commodity exchange programmes, cash-for-work, and so on), which also has relevance to households with low productivity in farming or in the informal sector in ‘non-crisis’ periods. Such activities, organized at the village level and supervised or monitored by the lower levels of local government, can be linked to community health and nutrition programmes, as well as to the improvement of ‘village assets’ such as water points and schools. Regular programmes of this sort would refine skills and organizational experience available to carry expanded income-enhancing and welfare-defending interventions, following harvest failure, an insurgent attack, or the influx of displaced people from a neighbouring district or country.

4. Given actual or plausible future situations of State budgetary constraints and foreign exchange shortages which impinge on sectors concerned with welfare provision, including those arising from ‘forced’ diversion of resources to defence and security, welfare programme design in SADCC should favour increasingly the identification and development of cost-effective means of welfare provision and an increased degree of community responsibility for (p.460) operation and maintenance. Programme design has tended to neglect the possibilities for achieving greater sustainability of capital investments, and coverage of recurrent costs, that may exist with ‘recipient’ communities, even when themselves under economic stress. With the opening up of opportunities for local-level expression of welfare priorities, and of channels and incentives (matching or otherwise) for community labour, cash, skill, organizational and material contributions, access to services at actual constrained levels of State budgetary provision can be improved. Use of decentralized management and investment decision-making, and of control over revenue-raising and expenditure may also (depending on the nature of the decision-making bodies) be a complement to this trend. Donor aid, including food aid in kind, should also seek to provide necessary resources for increased community (or group-specific) responsibility over welfare services.

5. Technology and building standards chosen for social investment projects, and techniques used in service provision, apparently need to be subjected to more systematic tests of cost-efficiency and relevance than is usually the case in social-sector ministries. This is certainly increasingly recognized by governments in the region, but largely under increasingly rigorous conditions of resource shortfall, and in the context of economic adjustment programmes. The tendency still exists for welfare programme design to assume that provision is an end in itself; that future maintenance and rehabilitation costs need not be built into investment analysis; that benefits (in terms of expected improved health status, additions to community or national human capital stock, and so on) need not be subject to estimation; and that the efficiency of resource use in achieving such benefits need not be examined. This tendency is encouraged by the availability, in many places, of ‘free’ international resources lobbying, and often competing, to be accepted. Whilst such characteristics have been more typical of the initial post-independence periods than of today, State expenditure patterns in some countries still seem slow to adjust to accumulated experience: for example, although the Zambian education system has recently been willing to consider cost-efficiency promotion and the transferral of some expenses to users and parents, such costs are kept unnecessarily high by the continued training of fairly large numbers of new teachers, whilst the greater need is for upgrading and retraining of existing teaching staff to improve the efficiency of educational output, and to free resources for expenditure on school books and materials (Johnston et al. 1987). Indeed, for the countries which now face the greatest limitations on public-sector resources, the national objective of provision of universal access to primary (and in some cases junior secondary) education for a rapidly growing school-age population may itself come under increasing scrutiny.

6. ‘Economic adjustment programmes’, implemented as financial and fiscal review and reallocation mechanisms, can, in a receptive political and administrative climate, be viewed and used as opportunities for increasing the (p.461) incentives to welfare policy-making and programme implementors to improve methods for cost-efficiency analysis in their sectors, thereby increasing the medium-to long-term sustainability of these programmes. Such measures, far from being necessarily detrimental to welfare access, may in fact prevent future drastic and disruptive reductions in service provision should export earnings from mining (Zambia, Botswana), oil (Angola), or agricultural (Malawi, Zimbabwe, Zambia) sectors fall; when migrant labour opportunities are suddenly reduced (Mozambique, Lesotho); when international transport costs for landlocked countries are inflated due to sabotage of low-cost routes; and when defence expenditure growth under external threat is suddenly high.

7. Regional and international experience with programmes to reduce household food and income insecurity needs rapidly to be drawn together and made available to policy-makers and programme designers. The SADCC Food Security Programme based in Zimbabwe is the obvious body to do this, on behalf of the region. SADCC has already instituted a number of important research activities in areas relevant to this concern, such as land and water management for eroded or semi-arid areas, improvement of small grain and other drought-tolerant crop varieties, and livestock and veterinary programmes. Further work is needed, however, in documenting and analysing experience obtained in SADCC and other States (such as India), in the design and implementation of income-buffering measures, particularly in response to emergency-induced stress, but also as a more sustained approach to the reduction of structural poverty related to low household access to productive assets (including land, livestock, and labour).

8. In particular, as described in the Botswana case above, further attention needs to be focused on ways in which responses to short-term stresses and household income fluctuations can be designed so as also to promote reduced vulnerability in future to such stresses, particularly through diversification of income sources. The use of food or cash injections in drought-or war-affected areas to provide start-up funds for household income-earning activities, not just consumption, is one example. Such inputs need to be complemented by the provision of incentives to improve local skills and community infrastructure in order to increase viability of household production (roads for marketing, dams for livestock and crop watering); by the distribution of scarce tools and equipment; and by a macro-economic climate which increases economic opportunities for participating households. As suggested by Hay (1986), post-relief phases need to do more than merely replace lost household assets: they need to apply resources simultaneously to prevent future collapses of income and reverse the increasing trend to vulnerability. The earlier review of the nature of rural poverty also suggests that sustained policies to pre-empt the strains caused by short-term insecurity will need to focus particularly on investment in technology development for smallholder farming (raising yields (p.462) and effective water availability), as well as broadening income sources and access to productive assets. Interestingly, despite the extreme collapse of incomes in much of rural Mozambique, the national consensus in that country for the strategy of an appropriate response is now one of the most advanced in the SADCC region, in its insistence on the integration of rehabilitation and development, both conceptually and practically, with relief responses (see Government of Mozambique 1988).

9. Related to the above point, ‘traditional’ (namely, persisting and proven) community and household coping systems in times of stress need to be better understood in Southern Africa, and should not be undermined by larger-scale response mechanisms financed by national or international resources. Village-based welfare strategies and adaptations, ranging from deployment of extended family connections, resort to hunting and veldfood gathering, use of medicinal herbs and healers, running down through local sales of livestock holdings, to migration in search of casual employment, are among the mechanisms long used in the first instance following income losses. Environmental changes and certain State policies in both colonial and more recent times tended to reduce possible recourse to such buffering mechanisms (via, for example hunting regulations, overexploitation of grazing resources, land alienation, discouragement of drought-resistant crops, restrictions on population movements). Examples of current dilemmas exist within the programmes reviewed in the previous case-studies: the feeding of people declared destitute by their communities under the Botswana Drought Relief Programme has, in the view of many local observers, tended to reduce family willingness to take responsibility for older, poorer members; whilst the Mozambique government, faced with the problem of large numbers of children either orphaned or out of contact with their parents due to war, has opted quite successfully for the promotion of fostering of such children by new families within displaced communities.

10. For improved welfare programme management and implementation, further attention is needed to human resource development in this as a specialized area. This is particularly required as methods of financing for welfare delivery become more highly decentralized, and as maintenance and rehabilitation take increasing priority over initial capital investment in infrastructure. The accountants and budget managers, supply and logistics officers, hospital, school, and training institute directors, vehicle maintenance and fleet management specialists, all of whom are needed to improve the efficiency of working conditions for the doctors, nurses, teachers, water engineers, and relief programme co-ordinators, seem to be in chronic short supply across the region. Equally, the patchiness and poor quality of data collection and analysis, in areas which would provide resource allocators in welfare-related sectors with indicators of priorities and results of their decisions, are prevalent constraints—in some countries, information systems themselves are the victims (p.463) of destabilization. Without improvements on the basis of training in information collection, communication, and use, from the household or community level upwards, more efficient design and implementation of welfare programmes, and more timely response to situations of stress and emergency, even in the presence of effective political demands, can hardly be expected.

(p.464) References

Bibliography references:

Botswana Ministry of Agriculture/Central Statistics Office (1986), Botswana Agricultural Statistics. Gaborone: Government Printer.

Cornia, G., Jolly, R., and Stewart, F. (1987), Adjustment with a Human Face: Protecting the Vulnerable and Promoting Growth, 2 vols. Oxford: Clarendon Press.

Davies, R., and Saunders, D. (1987), ‘IMF Stabilisation Policies and the Effect on Child Health in Zimbabwe’, Review of African Political Economy, 38.

FAO (Food and Agriculture Organization) Food Balance Sheets (1984), Rome: FAO (mimeo).

Food Studies Group (1988), ‘Lesotho, Food Aid, Food Security Disincentives and Dependency’, mimeo, Oxford University.

Government of Botswana (1974/5), Report on the Rural Income Distribution Survey. Gaborone: Government Printer.

Government of Mozambique (1988), Rising to the Challenge. Maputo: National Executive Commission for Emergency (CENE/DPCCN).

Green, R., Asrat, D., Mauras, M., and Morgan, R. (1987), Children on the Front Line, the Impact of Apartheid, Destabilisation and Warfare on Children in Southern and South Africa, (pt. 1). New York: UNICEF.

Hanlon, J. (1986), Beggar Your Neighbours. London: James Curry.

Hay, W. R. (1986), ‘Food Aid and Relief:.Development Strategies’, Disasters, 10/4.

Holm, J., and Morgan, R. (1985), ‘Coping with Drought in Botswana: An African Success’, Journal of Modern African Studies, 23/3.

Johnston, A., Kaluba, H., Karlsson, M., and Nystrom, K. (1987), Education and Economic Crisis: The Cases of Mozambique and Zambia. Stockholm: Swedish International Development Authority, Education Division.

Morgan, R. (1985), ‘The Development and Application of a Drought Early Warning System in Botswana’, Disasters, 9/1.

——— (1986a), ‘From Drought Relief to Post-Disaster Recovery, the Case of Botswana’,Disasters, 10/1.

——— (1986b), ‘Calculation of Size of Regional Reserve Stocks: Working Paper, Regional Reserve Study Part I’, mimeo, Vakakis and Associates/SADCC Regional Food Reserve/Food Aid Study.

Noormahomed, A. R., and Cliff, J. (1987), ‘The Impact on Health in Mozambique of South African Destabilisation’, 2nd edn., mimeo Ministry of Health, People’s Republic of Mozambique.

Quinn, V. (1986), ‘Malawi: Agricultural Development and Malnutrition’, mimeo, UNICEF, Nairobi.

Rohrbach, D. (1987), ‘The Growth of Smallholder Maize Production in Zimbabwe 1979–1985: Implications for Food Security’, paper presented at Third Annual Conference on Food Security in Southern Africa, University of Zimbabwe, Harare.

RDC (Rural Development Council, Republic of Botswana) (1985), ‘Report on the National Food Strategy’. Gaborone: Ministry of Finance and Development Planning.

UNICEF (1988), State of the World’s Children, 1988. Oxford: Oxford University Press.

(p.465) UNICEF: Situation Analyses of Children and Women (variously titled, published jointly with national governments): Zambia (1986), Zimbabwe (1985), Botswana (1986), Lesotho (draft, 1986), Mozambique (1983), Malawi (1983), Swaziland (draft, 1987), Tanzania (1985, 2 vols.), Angola (draft, 1986), available from respective UNICEF country offices.

World Bank (1986), World Development Report 1986. Washington, DC: World Bank.

Notes:

(1) Acknowledgement is made to UNICEF for permission to devote time to the paper on which this chapter is based and for use of the periodic Situation Analyses of Children and Women produced jointly with national governments, which form a major data source for the paper. The views expressed are however attributable to the author alone, writing in a personal capacity. Thanks are due to John Hills, Michael Lipton, and Joao Donato for their comments and assistance on earlier drafts, and to Reg Green and Roger Hay for general inspiration on the subject matter in recent years.

(2) Angola, Botswana, Lesotho, Malawi, Mozambique, Swaziland, Tanzania, Zambia, Zimbabwe.

(2) South Africa’s failure to improve on the IMRs achieved by Zimbabwe and Botswana with per capita GNP 2.5 to 3 times higher than those countries is also noteworthy in the regional context.

(3) This is not to suggest either that increases in household income, taken alone, are necessarily cost- (or otherwise) effective in leading to rapid improvement in child nutritional status.

(4) In Botswana’s 1974 Rural Income Distribution Survey (RIDS), for example, 72% of families based on freehold farms were found to have incomes below the calculated poverty datum line, compared with 55% in small villages and 45% among rural families overall.

(5) For example, cattle death rates in Botswana during the 1980s drought were strongly inversely related to the size of herd holdings (Botswana Ministry of Agriculture/CSO 1986: 37).

(6) Budget and expenditure maximization in themselves are clearly questionable as primary social-sector objectives when, for example, high-cost hospital care for the few, and teachers’ or nurses’ salaries make up large proportions of such expenditure, whilst basic drugs and school material remain widely unavailable. Use of low-cost technologies (e.g. for oral rehydration treatment of diarrhoea), bulk procurement policies for a reduced list of priority medical and school supplies, and community-based approaches to service delivery are among the measures proposed by Cornia et al. (1987) as appropriate adaptations to a highly resource-constrained social-sector environment.