- Email: [email protected]
Economics of health and nutrition in Kenya
sot. SC. .Jf
7, pp. 7~i30.
ECONOMICS
1986
OF HEALTH WILFRED
‘Department
O2-'-3536 56 53.00+O.OO Pergamon Press Ltd
of Agricultural
AND NUTRITION
M. MWASGI
and GERMASO
IN KENYA
h4. \~W.ABV’
Economics and ‘Institute for Development Studies, University of Nairobi. P.O. Bon 30197, Nairobi. Kenya
.Ibstract-This paper starts by outlining the present state of health and nutrition in Kenya. Health status for Kenyans has shown spectacular improvement since Kenya’s attainment of independence in 1963. Both the infant mortality and crude death rates fell by about 30% between 1963 and 1982. Life expectancy at birth has risen dramatically from 40 years to 54 years over the same period. However, this picture could be misleading because it is possible for the morbidity rate to have risen. or declined only slightly. over the twenty-year period during which mortality rates in Kenya fell substantially. It is further indicated that despite Kenya having per capita availability of nutrients exceedin, 0 that recommended by FAO;WHO: about one-third of Kenya’s population is unable to meet its food or nutritional requirements.
I.ISTRODUCTION
The vvell-being (welfare) of a nation is determined by many factors. One of these factors is the quality of its human resource. Simply defined, human resource is a community of individuals. The quality of these individuals, i.e. their ability to perform social, political and economic functions effectively, depends among other things, on the status of their health and nutrttton. Ill-health or malnutrition will not only cause them suffering (disutility), but will also reduce their work effort and/or impair their efficiency. From this follows the obvious and often unheeded fact that improvements in health and nutritional status increase a community’s welfare and hence its human resource value [I]. The purpose of this paper is to highlight the magnitudes of problems of ill-health and malnutrition in Kenya, and to suggest some ways of overcoming them. Section II of the paper depicts the current situation of health and nutrition in Kenya. Section III contains a conceptual framework to an understanding of problems of ill-health and malnutrition in society. In Section IV, this model is used to evaluate some practical policies of dealing with problems of inadequate health and nutrition in Kenya. In our illustrative examples particular attention is paid to the following: cost-effectiveness of curative vs preventive health services; budgetary implications of achieving universal coverage of the population with health services, inequalities in health care, food policy and the trade-offs involved in the policy of self-sufficiency in food production. Section V contains conclusions of the paper. II(
HEALTH
The World Health Organization (WHO) defines health as a state of physical, social and mental well-being, and not merely the absence of disease (WHO, 1974). At the moment, it is not clear how this *One should treat the infant mortality rate of Kenya estimated as 86 with caution given the incomplete death registration. Thus. the comparison given above assumes that the degree of underreporting independence to be about the same.
before
and
after
775
‘state or status of well-being’ is to be measured. What is clear is that death or illness disrupts it. Thus. health or health status can roughly be measured by mortality and morbidity rates. The most recent estimates of mortality rates in Kenya are for 1952. In that year, the death rate in Kenya was I4 per 1000 of the population, as compared to 20 in 1963. The estimate for infant mortality rate for the same year was 56. or about 28% smaller than the figure of I20 in 1963.* In both of the above cases, death rates in Kenya fell by about 30% over the 20 years since independence. But the risk of death among children under the age of 5 years continues to be much higher than in other age groups. For example, from the registration of roughly 200,000 deaths in 1975, about 50% of those deaths occurred among the children under the age of 5 years [I]. Obviously, this statistic does not give an entirely correct picture of the relative risk of death for Kenya’s under 5s. The year 1975 might have been a bad year for Kenya’s child health. Moreover. even if this was not the case, since death registration in Kenya is only about 25% complete. the above statistic contains some bias. But the point remains that the distribution of risk of death or illness in Kenya is highly skewed against children under the age of 5 years. In contrast. in a developed nation such as Sweden, less than 1% of deaths would occur among children under the age of 5 years. About 50% of all deaths reported in Kenya in 1975 were due to relatively small number of diseases. The same diseases are still among the main causes of morbidity and mortality today. They are: diarrhea1 diseases, measles. malaria, pneumonia and nutritional diseases. In a typical developed country, only about II % of the population would suffer from these diseases. The picture that emerges from the comparisons in the preceding two paragraphs is that the basic cause of our particular health problems could be the low level of our socio-economic development. This view becomes even more believable when we look at survivorship probabilities for infants across Kenya’s provinces. The survivorship probabilities shown in Table 1 indicate the likelihood of an infant surviving to t,~o
TAble
I
Infant ?ur%iiorship prob&bll:tiej Ksnya‘s prorincrs Sur>l\orship L1
P:O\inCe
ior
Probabllit> L3
SVarobt CZltrA syanza wsstem Coast Rift Valley Eastern Sorth Eastern
0.9030 0.9 I OS 0.7563 0.8276 0.S 116 0.8921 0.9316 0.8366
0.8895 0.8396 0.7544 0.7950 0.8024 0.8661 0.8309 02309
Kenya
0.8507
0.s309
Source: Anker and Knowles [3].
years (L2) or to rhree years (L3) in each of Kenya’s provinces. The provinces vary greatly in their levels of socio-economic development. The calculations are based on 1969 census data. Table I provides some interesting information. For instance, the probability of an infant surviving to 2 years in Nairobi is lower than that for central province. This is so in spite of the heavy concentration of medical personnel and health facilities in Nairobi [3]. The survivorship rates match very closely the infant mortality rates obtained from the 1978 Kenya Fertility Survey. The provincial mortality rates from this survey (per 1000 live births) were as follows: Central Province 72, Rift Valley 80, Nairobi 82. Eastern 89, Western 114. There are no mortality estimates for the other provinces. Implied in Table I is life expectancy at birth by province. The higher the probability of surviving during the first year of life, i.e. the lower the infant mortality rate, the higher the life expectancy is. The most recent estimate of life expectancy in Kenya is for 1982. In that year life expectancy at birth in Kenya was estimated at 54 years compared with 40 years in 1963 [4]. Thus life expectancy has increased dramatically by I4 years over the 20 years that Kenya has been independent. This increase in average longevity has been attributed to medical interventions and to improvements in socio-economic conditions [3]. But what are yet to be studied, are the relative contributions of these two factors in raising life expectancy in Kenya. An increase in life expectancy does not necessarily imply an improvement in health status. This is because it is possible for morbidity rate to increase in the face of a falling mortality rate. e.g. as more people are covered with health services, more people escape death. but without quite recovering from their illnesses. This increases the proportion of sick people in the community and also the amount of illness-related suffering. The third Rural Child Nutrition Survey [5], for instance. reported high morbidity rates among children under the age of 5 years, in spite of a relatively low estimate for their mortality rate. II(B).
NUTRITION
Nutritional status is the extent to which a population is well nourished or malnourished. Measurement of nutritional status of an individual or a community involves an assessment of a complex set of interrelated elements. These include dietary. clinical, biochemical, functional and growth factors. For
Table 2. Perccnrags of chlldrcn stunted or ux,ted Stunred (H-.4 < 90)
by pro\ince’
Waite? / \V.H < 90)
PrO~lnCC
19’7
iY7S I979
1982
19i-
coast Eastern Central Rift Valley Nyanza Western
19.0 3S.O 31.0 29.0 X.5 21.0
37.0 13 8 12.0 24.7 32.0 26.7
36 2 22.6 20.4 IY 8 x.9 25 7
7.5 50 30 4.5 2.5 5.0
IO.8 12 2.2 70 5.7 I.8
50 2.7 ‘S j.0 36 2.0
National
21.0
26.7
14.0
45
4.8
3.0
lY78 1979
1332
Source: Economic Surbey. 1981. *This table actualI\ shows that the children hais adapted \cry well whateverchro& foodshortage there is by rtiiiucme their suture. This ‘adaptation’ to food avnilability 1s qa:rs common, is not damagmg. i.e. it is physiological. and can Ly observed in nun:. countries where food supply, espcctally prors:n rich foods are not so ample.
the sake of simplicity, we will use growth and dietary factors in illustrating the extent of nutritional problem in Kenya. Nevertheless, the measures used here-anthropometric indices and availability of nutrients-give a fairly good picture of the nutritional status of a community. First we assess the nutritional status among children. and then in the entire population. To assess whether a child is well nourished or not, standard growth indices are used. The most widely used indices are height-for-age (H-.4) and weight-forage (W-H). H-A is obtained by div-iding the actual height of a child by the height he is expected to have given his age. W-H is obtained similarly. These indices can be used, among other things to indicate whether a nutritional deficiency is body wasting or stunting. As is displayed in Table 2. the most prevalent nutritional problem among Kenyan children is stunting. The ages of the children for which nutritional status is reported in Table 2 ranges from 1 to 5 years. As is clear from the last row. the nutritional status of this group of the population improved slightly between 1977 and 1982. But in some provinces, children’s nutritional status dropped. In Nyanza for instance, problems of stunting and wasting were more prevalent in 1982 than in 1977. That is in 1982 children in Nyanza were both mom stunted and wasted than in 1977. In Coast and Western Provinces. the change in nutritional status is ambiguous. As can be seen from the table, in 1982. children in Coast and Western Provinces were more stunted but less wasted than in 1977. The state of malnutrition for these children might have worsened, remained the same or become less severe. The data in Table 2 do not permit a conclusive state on this matter. There are, also large differences in nutritional status across various age groups of children l-5 years [j]. However, at the national level the degree of under-nutrition (that is acute) has declined significantly from 4.5 to 3%. a decline of 33%. Nevertheless. recent studies on food poverty in Kenya have estimated that between 31 and 36% of Kenyan households are food poor [6-S], the most recent estimate of food poverty in Kenya is that of Greer and Thorbecke [9]. By making allowances for differentials in provincial food pricej. household attributes, and by accounting for variations in con-
777
Economics of health and nutrition in Kenya Table 3. Daily per capita
nutrient
availability.
1965-1981’ Protein
CAXkS
Period 1965-1970 1971-197s 19761980 1981 196Sl981 FAO’WHO
2412 2453 2385 2428 (Average) recommended
average
7
(g) 62.9 65.6 646 72.6
2428
64.8
2362
46.0
Source: Kenya’s Development Plan 1984-1988. *It is important to note there is no consensus on the use of these criteria to denote food poverty.
sumer preferences, Greer and Thorbecke estimated that 25% of Kenya’s smallholders are food poor. A household is denoted as food poor if its food budget or nutrient intake is below the amount recommended by FAO/WHO. Surprisingly, the nutritional problem depicted above with results from household data is not visible at the national level. This is an indication of serious distributional problems in Kenya. Kenya’s daily per capita nutrient availability for the period 1965-1981 was greater than the average daily allowance recommended by FAO/WHO for that period. Table 3 displays this fact. In a detailed calculation of availability of nutrients in Kenya for the period 1971-1979, Mwangi and Adholla [IO], came up with the same conclusion as the one obtained from Table 3-that at the national level, Kenya has a surplus of nutrients. The main point that has emerged from this section is that in dealing with problems of inadequate health and nutrition in Kenya, distributional issues must be addressed seriously. At the moment, there is a sharp rural-urban differential in income levels. Incomes are about 5 times higher in urban than in rural areas. Skewness in income distribution also exists in each of these areas. The next section outlines a conceptual framework (model) that might help to explain problems of ill-health and malnutrition in Kenya. Ill.
MODEL
Nutritional and health statuses of an individual or a community are inter-dependent. For this reason, in order to determine either of the two, the other one must be known. The interdependence between health and nutrition arises from the following: (1) Infections adversely affect nutritional an individual. (2) Malnutrition increases susceptibility tions.
status of to infec-
Symbolically, the relationship between nutritional status (NUS) and health status (HES) can be expressed as NUS = NUS (HES, 0) (1) HES = HES (NUS, 0) (2) where 0 = a vector of socio-economic factors.
and environmental
Expressions (1) and (2) are simultaneous equations. For a given value of 0. once HES is known, so also is NUS; and conversely. In order to formulate practical health and nutrition policies, the factors denoted as 0 above, should not only be identified, but their effects on HES and NUS should also be estimated. Estimation of these effects is outside the scope of this paper. In a remarkable paper. Jonsson [l I] has identified some key factors in 0. In Jonsson’s framework, the factors in 0 affect only the nutritional status. But from our model, it is obvious that they also affect health status. The main idea of the model just presented is that the observed problems of health and nutrition are due to certain causes or factors. Jonsson classifies these factors as follows: (i) Immediate factors (a) inadequate intake (b) disease
of nutrients
(ii) Underlying factors (a) lack of food (b) wrong feeding practice (c) lack of water (d) inadequate health services (e) low incomes or purchasing power (f) lack of land (g) inadequate or poor housing (h) poor hygiene (i) inequalities in the distribution of commodities and services (j) inappropriate education or illiteracy. (iii) Basic factors (a) formal and non-formal institutions (b) socio-economic structure (c) resource availability. It is clear that our model throws considerable light on possible causes of ill-health and malnutrition in Kenya. Furthermore it shows that although problems of health and nutrition in Kenya are caused by a complex bundle of factors, these factors fit into a simple classification. This classification should be helpful in determining intervention levels for health and nutrition policies, e.g. in deciding whether a policy focus should be at the underlying causes of the problem at hand, or at some other level. IV. ASSESSMEST
OF POLICIES
Health policy Curative, preventive and promotive health services. Kenya’s health policy is predominantly curative in nature. That is, its focus is mainly on immediate and underlying causes of our health problems. As can be seen from Table 4, the bulk of our health budget in the next 4 years will be spent on treatment of the existing diseases. The ultimate objective of government’s health policy as depicted in Table 4, is to increase health status mainly through chemotherapy. There are two basic problems with this policy. First of all, it focuses only on one of the underlying causes of poor health. That is, on inadequate health services. If this is not the dominant determinant of health status in Kenya, then the policy will not achieve its objective. More-
775
WILFRED
41. MMWA?~GI
Table 1. Percentaee breakdowns oi zovernment health budget for the p&&J 1983 198-r-19&5 1988 (ewmatesi Activity (excludes research) Administration and planning Preventive and promotlve serbIces Rural health senwzs Curative services (hospital) Training Equipment and supplies Percentage total Total health expenditure (estnnates) Total eovernment budcet Source: Own calculations, Derslopment
Recurrent expenditure
100.0
0.2 j.9
4l.S
39 6 8.2 1.3 100.0 KS71.7m 4.7
Plan, lY8&19SS
over, in the long run, the current health policy is less cost-effective than the alternative policy of achieving better health largely through preventive and promotive health services. In the immediate future. the alternative policy will be more costly than the current policy because it will have to be concerned with such additional issues as housing. health education, water supply, food production and income distribution among others. Thus, in the shorter term the health sector will require additional resources in order to coordinate effectively with the other sectors. But in the longer term, the preventive health policy will require smaller budget for the disease problems that exist at the moment because their basic and underlying causes will have been removed or weakened. Health planners should resist instituting curative health policies as palliative measures in response to public demands for them. In fact, historical data show a very weak relationship between improvements in health status and provision of curative health services [IL!]. Costeffectiveness and whenever possible, the related analytical concept of cost-benefit analysis should guide planner’s choice of policy. Social, political and other considerations in health care should be taken as constraints in the application of analytical economic concepts in evaluating soundness of policy, rather than as hindrances in the use of these techniques. Inequalities in health care. Contrary to numerous pronouncements, the rural sector which certainly includes the vast majority of the population does not appear to be receiving a reasonable share of the health budget. Table 5 illustrates this point. As indicated by Table 5 there exists a wide variation in per capita health expenditures across Kenya’s provinces. Now suppose that under the current curative health care policy, in the pursuit of egalitarian goal, the government decides to provide every Kenyan with the ‘Nairobi-type health care’. What would be the health budget implication of this policy? Assuming for illustrative purposes the cost for curative services per person in Nairobi is still the same as in 1974; the however. that even under the concept of equal distribution of health services, Nairobi would still have a hisher budget being the national referral centre. But the point here is that Nairobi serves relatively small proportion of the national population. Furthermore, the bulk of this population comes from the City of Nairobi and its periphery.
*It is to be noted,
MWABL
Table 5. Selected per ca?rrs sxpendlruxs on curau\e heslrh senses b) prwince (Kf). IQ’:-1971
Derelopmenr expenditure
5.6 6.0 10.0 67 0 9.5 I .9
KS331.Sm 5.7
and GERM~XO M.
Nairobi Cent4 Coast Eastern North Eastern N)anza Rift ‘idlIe> WWWl National
Sourcs: Blgstrn
Development sxyndirure
Rscurrenr xpenditure
6.jY 0.50
-0.76 9.69
0.97
13.07
11.64
I) 04 0.59 0 34 0.18
6.12 3 54 3.28 8.84 4.09
1.23
11.96
[131
health budget for this policy w-ould be Kf469.7 million*. This budget would be greater than the entire government budget for 198-t and would absorb about 40% of GDP for that year. This extreme example shows that hospital based health care is not affordable for the whole population and is therefore inherently inequitable. Complete unirersal corerage cs selectice rinicersal coverage. In 1983, approx. 12% of Kenyan households lived less than 2 km from a health centre. That is, only about 360,000 households were covered with accessible ‘health centre-type’ of health services [l4]. Suppose now the government decides to extend this health centre service to the remaining 2.64 million households, how large will be the health budget? Based on 1984 estimates of health budget. it costs about Kf25.5 to provide the above type of health care to a household of 6. Thus, universal coverage of the population with health centre services would almost double the budget for l9S-t from Kf76.5 million to about Kfl44 million. The government can lower this cost by pursuing a policy of selective universal coverage. Under this policy, the government would cover all households with ‘health centre-type’ services. but would not cover all members of a household. For example, the government might decide to cover only children under the age of 5 years. This would mean for instance ensuring that drugs for the illness that commonly strike children are ala-ays available in health centres. The preceding discussion of the economic reality in health care, points to the need to increase the budget for the health sector. Additional funds to the health sector would increase health care coverage in the population, However, it is well known that no government budget anywhere in the world is allocated solely on the basis of economic criteria. Other considerations for allocation include such non-economic intangibles as national security and political expediency. Thus health planners should not expect a sure increase in the allocations for health and nutrition programmes. Nutrition Ioolicv , The aim of Kenya’s nutrition policy is to ensure that every Kenyan has an adequate intake of nutrients. For this policy to succeed, it must deal with the immediate as well as the underlying. and basic causes of malnutrition in Kenya. Nutntton policy is in-
Economics timately
related
policy depends The Xational tives:
of health
and nutrition
to food policy. Success of nutrition on that of food policy. Food Policy has the following objec-
(1) to maintain a position of broad self-sufficiency in the main foodstuffs in order to enable the nation to be fed without using scarce foreign exchange on food imports; (2) to achieve a calculated degree of security of food supply for each area of the country; and (3) to ensure that these foodstuffs are distributed in a manner that every member of the population has a nutritionally adequate diet. The basic aim of this policy is to ensure that nutritious food is available in adequate amounts to every Kenyan. It is not intended to achieve selfsufficiency in all food products within the same time period. Acailability of food and nutrients. Recently calculations of food availability in Kenya shows that per capita food availability in Kenya was fairly steady from 1974 to 1979. From the food data was derived per capita availability of nutrients for the same time period. The results indicate that Kenya does not have a shortage of nutrients. On the contrary the cause of Kenya’s nutritional problems between 1971 and 1979 and in subsequent years is mainly maldistribution of food rather than its shortage. To alleviate the problem of malnutrition, more efficient ways of distributing food must accompany increases in food production. Purchasing power. There is empirical evidence to indicate that an increase in the purchasing power of low income households would increase their nutritional status proportionately more than that for higher income households. This point is illustrated in Table 6. The main point to notice from Table 6 is the behaviour of the expenditure elasticity for calories. This elasticity declines very rapidly as we move from very low income groups to very high income groups. The policy implication of this is that for any additional shilling received by these groups, the low income groups spend a greater proportion of it on consumption of calories than is the case with the rich groups. Food self-sujiciency. A strategy of self-sufficiency in food is faced with a set of climatic, socio-economic and political constraints among others. A realistic strategy of self-sufficiency in food production should take cognizance of certain trade-offs. Table 6. Food consumption by income group
Group Rural low income Rural middle income Urban poor Rural rich Urban middle V’rban rich
Share of income spent on food
Share of expenditure spent on food
Expenditure elasticitv for calohes
0.80
0.77
0.74
0.45 0.62 0.32 0.39 0.19
0.75 0.45 0.73 0.37 0.21
0.67 0.35 0.48 0.34 0.25
Source: Income and expenditure shares are computed from the Integrated Rural Suney I (1974-1975) and the Urban Food Purchasing Survey (1977). Elasticities are computed from calories expenditure data derired by Frahkrg and Shah [Ij].
in Kenya
779
The following trade-offs have been identified regard to food policy in Kenya:
with
(i) Food’Food trade-offs; mainly in Western Kenya where the continued pursuit of self-sufficiency in sugar is in conflict with that of maize. (ii) Food, Export crop trade-offs; these are mainly in Central Kenya where self-sufficiencv in maize is in conflict with the foreign exchange ea&ing crops such as coffee and tea. (iii) Food/Animal trade-offs; these are mainly in areas where livestock and crops must compete for the same piece of land. v. CONCLUSION In this paper, we have shown that there exists serious problems of ill-health and malnutrition in Kenya. These problems are caused by many and interrelated factors. These factors can be classified as direct, intermediate, and ultimate or basic. The manner in which these factors affect health and nutritional statuses must first be known before health and nutritional policies are designed and implemented. The existing data on health and nutrition in Kenya are sparse and unreliable for use in the design of practical policies. We suggest a systematic collection of data on health and nutrition at the household level be started. In particular, information should be collected regarding people’s perceptions of their health statuses, their expenditures on health care, the health facilities they use in the event of an illness. etc. With regard to data on use of facilities. special attention should be given to data on use of traditional healers with a view of finding ways of officially integrating traditional healers with modern health care sector. As for nutrition, data about growth profiles of children and the socio-economic conditions of their households is now available at the Central Bureau of Statistics, Ministry of Finance and Planning. Efforts to collect this kind of data should be continued. We should also like to call attention to the fact that the capability to analyse the collected data should be developed. We see no need to collect data that will not be analysed. An implicit and important point in our paper is that public spending on health and nutrition is a productive investment. It is an investment in health capital. Scarcity of health capital can impede the process of socio-economic development exactly the same way that shortages of physical capital. foreign exchange or skilled labour can hinder it. Another implicit and critical point in our paper is the issue of the relationship between improvements in health/nutritional status and the rate of population increase. We assume that a healthy and well nourished population will use the available resources more efficiently, thus increasing the supply of the available goods and services. Further, since infant mortality rate would be low in such a population. the well documented motive of couples having many children in order to increase the chances of some of them surviving would not be in effect. We will now conclude this paper with irs three major findings: (1) Health and nutritional statuses of a community are interdependent. None can be attained indepen-
780
WILFRED Xl. MU’,~SGI and
dentlv of the other. This idea of simultaneity m the pursuit of optimal states of nutrition and health, implies that in order to be effective our efforts to remove problems of disease and malnutrition in society must be coordinated with our other efforts elsewhere in the economy. (2) Attainment of self-sufficiency in food production is a necessary, but not a sufficient condition for security against malnutrition. Equitable distribution of food and of nutrients must also accompany food self-sufficiency. (3) At the farm level, the national policy of food self-sufficiency might be in conflict with the farmer’s rational management of his farm. Faced with relative costs of growing various foodstuffs and cash crops. a farmer will choose the crop combination from which he expects the greatest net return; or the combination that guarantees him some minimum amount of food. Needless to say his choice of crops might be different from that which would be consistent with a national policy of food self-sufficiency. A good understanding of farmers’ decision making at the farm level, and of how the decisions farmers make might be influenced by policy, is one of the first steps in a successful implementation of a programme of self-sutKciency in food production.
GERMANO M. MWABC
4.
5.
6.
7.
8.
9.
IO.
REFERENCES I. A
recent measure of social welfare is the Physical Quality of Life Index (PQLI). This is a composite index of a community’s life expectancy at birth. infant mortality and literacy rates. PQLI as a measure. attaches a significant weight on the provision and distribution of basic human needs. The traditional measure of social welfare, the Gross Domestic Product (GDP), ignores distributional problems in society. For details about PQLI see, Morris M. D. ,Measuring the Condition of Ihe World’s Poor: The Physical Quality of Life Index. Pergamon Press, New York, 1979. 2. Koinange W. A Report by the Director of Medical Services on Status of Health in Kenya, 1979. Ministry of Health, Nairobi, 1982. 3. Anker R. and Knowles J. An empirical analysis of mortality differentials in Kenya at the macro and micro levels. Econ. Dec. Cult. Change. 29, 165-155. 1980. The official survivorship in Nairobi is probably not lower because of adverse socio-economic conditions but because of more complete reporting, i.e. deaths are re-
II.
12.
13. 14.
15.
ported more completely. hence the survivorship goes down in comparison with a region where (especially early child) deaths are reported mcompletely. Ministry of Finance and Economic Planning. Kenya. Daelopmenr Plan 19511985. Central Bureau of Statistics. Third Rural Child :Vurrifion Srtmey. Ministry of Finance and Planning, Bairobi. 1982. While the mortality rate for this age group was 56!1000, their morbidity rate was 465,lOOO. The percentage of sick children varied greatly by province. It ranged from 37.1% in central province to 58.8% in Nyanza province. In some districts. e.g. Kisumu, Kilifi, Tana and Lamu. those percentages were as high as 64%. McCarthy F. D. and Xlwangi W. M. Kenya Agriculture: Toward 2000. International Institute for Applied System Analysis, Laxenburg. Austria, l9S2. Crawford E. and Thorbecke E. Employment Income Distribution, Poverty and Basic Seeds in Kenya. Report of an IL0 Consulting Mission. Cornell University. Ithaca, N.Y.. 1978 (mimeo). Collier P. and Deepak L. Porerr) and Growrh in Kenya. Staff Workinn Paper No. 389. World Bank. Washington, D.C.. 1980. . Greer J. and Thorbecke E. Patterns of Food Consumption and Poverty and Effects of Food Prices. Cornell University, Ithaca, N.Y.. 1983 (mimeo). Mwangi W. M. and %ligot-Adholla S. E. Food selfsufficiency in sub-Saharan Africa: approaches. problems and constraints: a case of Kenya. Paper presented at a Workshop on Food Self-Sujiciency in Sub-Saharan Africa, University of Dar es Salaam. Tanzania. 1984. Jonsson U. A conceptual approach to the understanding and explanation of malnutrition in society. Paper presented at the Workshop on “Hunger and Sociefy‘in Soliwayo”. Iringa. Tanzania, 1983. Garcia D. The effect of ownership and control of health systems and the concept of health. Paper presented at a Workshop on Health Policy on Eastern and Southern Africa, Arusha, 1981. Bigsten A. Regional Inequalilies in Kenya. Institute for Development Studies. DP., No. 330. Nairobi, 1977. We arbitrarily define any health centre that is more than 2 km from a household as inaccessible. Furthermore, access to a health centre includes access to a referral facility. This view assumes a terrain with homogeneous physical features. But in a non-mountainous area such as northern Kenya a health centre that is 4 km auay can be considered accessible. Frohberg H. C. and Shah M. M. Nutrition status of rural urban Kenya. \Iarketing Development Project. FAO, Nairobi, 1978.
7, pp. 7~i30.
ECONOMICS
1986
OF HEALTH WILFRED
‘Department
O2-'-3536 56 53.00+O.OO Pergamon Press Ltd
of Agricultural
AND NUTRITION
M. MWASGI
and GERMASO
IN KENYA
h4. \~W.ABV’
Economics and ‘Institute for Development Studies, University of Nairobi. P.O. Bon 30197, Nairobi. Kenya
.Ibstract-This paper starts by outlining the present state of health and nutrition in Kenya. Health status for Kenyans has shown spectacular improvement since Kenya’s attainment of independence in 1963. Both the infant mortality and crude death rates fell by about 30% between 1963 and 1982. Life expectancy at birth has risen dramatically from 40 years to 54 years over the same period. However, this picture could be misleading because it is possible for the morbidity rate to have risen. or declined only slightly. over the twenty-year period during which mortality rates in Kenya fell substantially. It is further indicated that despite Kenya having per capita availability of nutrients exceedin, 0 that recommended by FAO;WHO: about one-third of Kenya’s population is unable to meet its food or nutritional requirements.
I.ISTRODUCTION
The vvell-being (welfare) of a nation is determined by many factors. One of these factors is the quality of its human resource. Simply defined, human resource is a community of individuals. The quality of these individuals, i.e. their ability to perform social, political and economic functions effectively, depends among other things, on the status of their health and nutrttton. Ill-health or malnutrition will not only cause them suffering (disutility), but will also reduce their work effort and/or impair their efficiency. From this follows the obvious and often unheeded fact that improvements in health and nutritional status increase a community’s welfare and hence its human resource value [I]. The purpose of this paper is to highlight the magnitudes of problems of ill-health and malnutrition in Kenya, and to suggest some ways of overcoming them. Section II of the paper depicts the current situation of health and nutrition in Kenya. Section III contains a conceptual framework to an understanding of problems of ill-health and malnutrition in society. In Section IV, this model is used to evaluate some practical policies of dealing with problems of inadequate health and nutrition in Kenya. In our illustrative examples particular attention is paid to the following: cost-effectiveness of curative vs preventive health services; budgetary implications of achieving universal coverage of the population with health services, inequalities in health care, food policy and the trade-offs involved in the policy of self-sufficiency in food production. Section V contains conclusions of the paper. II(
HEALTH
The World Health Organization (WHO) defines health as a state of physical, social and mental well-being, and not merely the absence of disease (WHO, 1974). At the moment, it is not clear how this *One should treat the infant mortality rate of Kenya estimated as 86 with caution given the incomplete death registration. Thus. the comparison given above assumes that the degree of underreporting independence to be about the same.
before
and
after
775
‘state or status of well-being’ is to be measured. What is clear is that death or illness disrupts it. Thus. health or health status can roughly be measured by mortality and morbidity rates. The most recent estimates of mortality rates in Kenya are for 1952. In that year, the death rate in Kenya was I4 per 1000 of the population, as compared to 20 in 1963. The estimate for infant mortality rate for the same year was 56. or about 28% smaller than the figure of I20 in 1963.* In both of the above cases, death rates in Kenya fell by about 30% over the 20 years since independence. But the risk of death among children under the age of 5 years continues to be much higher than in other age groups. For example, from the registration of roughly 200,000 deaths in 1975, about 50% of those deaths occurred among the children under the age of 5 years [I]. Obviously, this statistic does not give an entirely correct picture of the relative risk of death for Kenya’s under 5s. The year 1975 might have been a bad year for Kenya’s child health. Moreover. even if this was not the case, since death registration in Kenya is only about 25% complete. the above statistic contains some bias. But the point remains that the distribution of risk of death or illness in Kenya is highly skewed against children under the age of 5 years. In contrast. in a developed nation such as Sweden, less than 1% of deaths would occur among children under the age of 5 years. About 50% of all deaths reported in Kenya in 1975 were due to relatively small number of diseases. The same diseases are still among the main causes of morbidity and mortality today. They are: diarrhea1 diseases, measles. malaria, pneumonia and nutritional diseases. In a typical developed country, only about II % of the population would suffer from these diseases. The picture that emerges from the comparisons in the preceding two paragraphs is that the basic cause of our particular health problems could be the low level of our socio-economic development. This view becomes even more believable when we look at survivorship probabilities for infants across Kenya’s provinces. The survivorship probabilities shown in Table 1 indicate the likelihood of an infant surviving to t,~o
TAble
I
Infant ?ur%iiorship prob&bll:tiej Ksnya‘s prorincrs Sur>l\orship L1
P:O\inCe
ior
Probabllit> L3
SVarobt CZltrA syanza wsstem Coast Rift Valley Eastern Sorth Eastern
0.9030 0.9 I OS 0.7563 0.8276 0.S 116 0.8921 0.9316 0.8366
0.8895 0.8396 0.7544 0.7950 0.8024 0.8661 0.8309 02309
Kenya
0.8507
0.s309
Source: Anker and Knowles [3].
years (L2) or to rhree years (L3) in each of Kenya’s provinces. The provinces vary greatly in their levels of socio-economic development. The calculations are based on 1969 census data. Table I provides some interesting information. For instance, the probability of an infant surviving to 2 years in Nairobi is lower than that for central province. This is so in spite of the heavy concentration of medical personnel and health facilities in Nairobi [3]. The survivorship rates match very closely the infant mortality rates obtained from the 1978 Kenya Fertility Survey. The provincial mortality rates from this survey (per 1000 live births) were as follows: Central Province 72, Rift Valley 80, Nairobi 82. Eastern 89, Western 114. There are no mortality estimates for the other provinces. Implied in Table I is life expectancy at birth by province. The higher the probability of surviving during the first year of life, i.e. the lower the infant mortality rate, the higher the life expectancy is. The most recent estimate of life expectancy in Kenya is for 1982. In that year life expectancy at birth in Kenya was estimated at 54 years compared with 40 years in 1963 [4]. Thus life expectancy has increased dramatically by I4 years over the 20 years that Kenya has been independent. This increase in average longevity has been attributed to medical interventions and to improvements in socio-economic conditions [3]. But what are yet to be studied, are the relative contributions of these two factors in raising life expectancy in Kenya. An increase in life expectancy does not necessarily imply an improvement in health status. This is because it is possible for morbidity rate to increase in the face of a falling mortality rate. e.g. as more people are covered with health services, more people escape death. but without quite recovering from their illnesses. This increases the proportion of sick people in the community and also the amount of illness-related suffering. The third Rural Child Nutrition Survey [5], for instance. reported high morbidity rates among children under the age of 5 years, in spite of a relatively low estimate for their mortality rate. II(B).
NUTRITION
Nutritional status is the extent to which a population is well nourished or malnourished. Measurement of nutritional status of an individual or a community involves an assessment of a complex set of interrelated elements. These include dietary. clinical, biochemical, functional and growth factors. For
Table 2. Perccnrags of chlldrcn stunted or ux,ted Stunred (H-.4 < 90)
by pro\ince’
Waite? / \V.H < 90)
PrO~lnCC
19’7
iY7S I979
1982
19i-
coast Eastern Central Rift Valley Nyanza Western
19.0 3S.O 31.0 29.0 X.5 21.0
37.0 13 8 12.0 24.7 32.0 26.7
36 2 22.6 20.4 IY 8 x.9 25 7
7.5 50 30 4.5 2.5 5.0
IO.8 12 2.2 70 5.7 I.8
50 2.7 ‘S j.0 36 2.0
National
21.0
26.7
14.0
45
4.8
3.0
lY78 1979
1332
Source: Economic Surbey. 1981. *This table actualI\ shows that the children hais adapted \cry well whateverchro& foodshortage there is by rtiiiucme their suture. This ‘adaptation’ to food avnilability 1s qa:rs common, is not damagmg. i.e. it is physiological. and can Ly observed in nun:. countries where food supply, espcctally prors:n rich foods are not so ample.
the sake of simplicity, we will use growth and dietary factors in illustrating the extent of nutritional problem in Kenya. Nevertheless, the measures used here-anthropometric indices and availability of nutrients-give a fairly good picture of the nutritional status of a community. First we assess the nutritional status among children. and then in the entire population. To assess whether a child is well nourished or not, standard growth indices are used. The most widely used indices are height-for-age (H-.4) and weight-forage (W-H). H-A is obtained by div-iding the actual height of a child by the height he is expected to have given his age. W-H is obtained similarly. These indices can be used, among other things to indicate whether a nutritional deficiency is body wasting or stunting. As is displayed in Table 2. the most prevalent nutritional problem among Kenyan children is stunting. The ages of the children for which nutritional status is reported in Table 2 ranges from 1 to 5 years. As is clear from the last row. the nutritional status of this group of the population improved slightly between 1977 and 1982. But in some provinces, children’s nutritional status dropped. In Nyanza for instance, problems of stunting and wasting were more prevalent in 1982 than in 1977. That is in 1982 children in Nyanza were both mom stunted and wasted than in 1977. In Coast and Western Provinces. the change in nutritional status is ambiguous. As can be seen from the table, in 1982. children in Coast and Western Provinces were more stunted but less wasted than in 1977. The state of malnutrition for these children might have worsened, remained the same or become less severe. The data in Table 2 do not permit a conclusive state on this matter. There are, also large differences in nutritional status across various age groups of children l-5 years [j]. However, at the national level the degree of under-nutrition (that is acute) has declined significantly from 4.5 to 3%. a decline of 33%. Nevertheless. recent studies on food poverty in Kenya have estimated that between 31 and 36% of Kenyan households are food poor [6-S], the most recent estimate of food poverty in Kenya is that of Greer and Thorbecke [9]. By making allowances for differentials in provincial food pricej. household attributes, and by accounting for variations in con-
777
Economics of health and nutrition in Kenya Table 3. Daily per capita
nutrient
availability.
1965-1981’ Protein
CAXkS
Period 1965-1970 1971-197s 19761980 1981 196Sl981 FAO’WHO
2412 2453 2385 2428 (Average) recommended
average
7
(g) 62.9 65.6 646 72.6
2428
64.8
2362
46.0
Source: Kenya’s Development Plan 1984-1988. *It is important to note there is no consensus on the use of these criteria to denote food poverty.
sumer preferences, Greer and Thorbecke estimated that 25% of Kenya’s smallholders are food poor. A household is denoted as food poor if its food budget or nutrient intake is below the amount recommended by FAO/WHO. Surprisingly, the nutritional problem depicted above with results from household data is not visible at the national level. This is an indication of serious distributional problems in Kenya. Kenya’s daily per capita nutrient availability for the period 1965-1981 was greater than the average daily allowance recommended by FAO/WHO for that period. Table 3 displays this fact. In a detailed calculation of availability of nutrients in Kenya for the period 1971-1979, Mwangi and Adholla [IO], came up with the same conclusion as the one obtained from Table 3-that at the national level, Kenya has a surplus of nutrients. The main point that has emerged from this section is that in dealing with problems of inadequate health and nutrition in Kenya, distributional issues must be addressed seriously. At the moment, there is a sharp rural-urban differential in income levels. Incomes are about 5 times higher in urban than in rural areas. Skewness in income distribution also exists in each of these areas. The next section outlines a conceptual framework (model) that might help to explain problems of ill-health and malnutrition in Kenya. Ill.
MODEL
Nutritional and health statuses of an individual or a community are inter-dependent. For this reason, in order to determine either of the two, the other one must be known. The interdependence between health and nutrition arises from the following: (1) Infections adversely affect nutritional an individual. (2) Malnutrition increases susceptibility tions.
status of to infec-
Symbolically, the relationship between nutritional status (NUS) and health status (HES) can be expressed as NUS = NUS (HES, 0) (1) HES = HES (NUS, 0) (2) where 0 = a vector of socio-economic factors.
and environmental
Expressions (1) and (2) are simultaneous equations. For a given value of 0. once HES is known, so also is NUS; and conversely. In order to formulate practical health and nutrition policies, the factors denoted as 0 above, should not only be identified, but their effects on HES and NUS should also be estimated. Estimation of these effects is outside the scope of this paper. In a remarkable paper. Jonsson [l I] has identified some key factors in 0. In Jonsson’s framework, the factors in 0 affect only the nutritional status. But from our model, it is obvious that they also affect health status. The main idea of the model just presented is that the observed problems of health and nutrition are due to certain causes or factors. Jonsson classifies these factors as follows: (i) Immediate factors (a) inadequate intake (b) disease
of nutrients
(ii) Underlying factors (a) lack of food (b) wrong feeding practice (c) lack of water (d) inadequate health services (e) low incomes or purchasing power (f) lack of land (g) inadequate or poor housing (h) poor hygiene (i) inequalities in the distribution of commodities and services (j) inappropriate education or illiteracy. (iii) Basic factors (a) formal and non-formal institutions (b) socio-economic structure (c) resource availability. It is clear that our model throws considerable light on possible causes of ill-health and malnutrition in Kenya. Furthermore it shows that although problems of health and nutrition in Kenya are caused by a complex bundle of factors, these factors fit into a simple classification. This classification should be helpful in determining intervention levels for health and nutrition policies, e.g. in deciding whether a policy focus should be at the underlying causes of the problem at hand, or at some other level. IV. ASSESSMEST
OF POLICIES
Health policy Curative, preventive and promotive health services. Kenya’s health policy is predominantly curative in nature. That is, its focus is mainly on immediate and underlying causes of our health problems. As can be seen from Table 4, the bulk of our health budget in the next 4 years will be spent on treatment of the existing diseases. The ultimate objective of government’s health policy as depicted in Table 4, is to increase health status mainly through chemotherapy. There are two basic problems with this policy. First of all, it focuses only on one of the underlying causes of poor health. That is, on inadequate health services. If this is not the dominant determinant of health status in Kenya, then the policy will not achieve its objective. More-
775
WILFRED
41. MMWA?~GI
Table 1. Percentaee breakdowns oi zovernment health budget for the p&&J 1983 198-r-19&5 1988 (ewmatesi Activity (excludes research) Administration and planning Preventive and promotlve serbIces Rural health senwzs Curative services (hospital) Training Equipment and supplies Percentage total Total health expenditure (estnnates) Total eovernment budcet Source: Own calculations, Derslopment
Recurrent expenditure
100.0
0.2 j.9
4l.S
39 6 8.2 1.3 100.0 KS71.7m 4.7
Plan, lY8&19SS
over, in the long run, the current health policy is less cost-effective than the alternative policy of achieving better health largely through preventive and promotive health services. In the immediate future. the alternative policy will be more costly than the current policy because it will have to be concerned with such additional issues as housing. health education, water supply, food production and income distribution among others. Thus, in the shorter term the health sector will require additional resources in order to coordinate effectively with the other sectors. But in the longer term, the preventive health policy will require smaller budget for the disease problems that exist at the moment because their basic and underlying causes will have been removed or weakened. Health planners should resist instituting curative health policies as palliative measures in response to public demands for them. In fact, historical data show a very weak relationship between improvements in health status and provision of curative health services [IL!]. Costeffectiveness and whenever possible, the related analytical concept of cost-benefit analysis should guide planner’s choice of policy. Social, political and other considerations in health care should be taken as constraints in the application of analytical economic concepts in evaluating soundness of policy, rather than as hindrances in the use of these techniques. Inequalities in health care. Contrary to numerous pronouncements, the rural sector which certainly includes the vast majority of the population does not appear to be receiving a reasonable share of the health budget. Table 5 illustrates this point. As indicated by Table 5 there exists a wide variation in per capita health expenditures across Kenya’s provinces. Now suppose that under the current curative health care policy, in the pursuit of egalitarian goal, the government decides to provide every Kenyan with the ‘Nairobi-type health care’. What would be the health budget implication of this policy? Assuming for illustrative purposes the cost for curative services per person in Nairobi is still the same as in 1974; the however. that even under the concept of equal distribution of health services, Nairobi would still have a hisher budget being the national referral centre. But the point here is that Nairobi serves relatively small proportion of the national population. Furthermore, the bulk of this population comes from the City of Nairobi and its periphery.
*It is to be noted,
MWABL
Table 5. Selected per ca?rrs sxpendlruxs on curau\e heslrh senses b) prwince (Kf). IQ’:-1971
Derelopmenr expenditure
5.6 6.0 10.0 67 0 9.5 I .9
KS331.Sm 5.7
and GERM~XO M.
Nairobi Cent4 Coast Eastern North Eastern N)anza Rift ‘idlIe> WWWl National
Sourcs: Blgstrn
Development sxyndirure
Rscurrenr xpenditure
6.jY 0.50
-0.76 9.69
0.97
13.07
11.64
I) 04 0.59 0 34 0.18
6.12 3 54 3.28 8.84 4.09
1.23
11.96
[131
health budget for this policy w-ould be Kf469.7 million*. This budget would be greater than the entire government budget for 198-t and would absorb about 40% of GDP for that year. This extreme example shows that hospital based health care is not affordable for the whole population and is therefore inherently inequitable. Complete unirersal corerage cs selectice rinicersal coverage. In 1983, approx. 12% of Kenyan households lived less than 2 km from a health centre. That is, only about 360,000 households were covered with accessible ‘health centre-type’ of health services [l4]. Suppose now the government decides to extend this health centre service to the remaining 2.64 million households, how large will be the health budget? Based on 1984 estimates of health budget. it costs about Kf25.5 to provide the above type of health care to a household of 6. Thus, universal coverage of the population with health centre services would almost double the budget for l9S-t from Kf76.5 million to about Kfl44 million. The government can lower this cost by pursuing a policy of selective universal coverage. Under this policy, the government would cover all households with ‘health centre-type’ services. but would not cover all members of a household. For example, the government might decide to cover only children under the age of 5 years. This would mean for instance ensuring that drugs for the illness that commonly strike children are ala-ays available in health centres. The preceding discussion of the economic reality in health care, points to the need to increase the budget for the health sector. Additional funds to the health sector would increase health care coverage in the population, However, it is well known that no government budget anywhere in the world is allocated solely on the basis of economic criteria. Other considerations for allocation include such non-economic intangibles as national security and political expediency. Thus health planners should not expect a sure increase in the allocations for health and nutrition programmes. Nutrition Ioolicv , The aim of Kenya’s nutrition policy is to ensure that every Kenyan has an adequate intake of nutrients. For this policy to succeed, it must deal with the immediate as well as the underlying. and basic causes of malnutrition in Kenya. Nutntton policy is in-
Economics timately
related
policy depends The Xational tives:
of health
and nutrition
to food policy. Success of nutrition on that of food policy. Food Policy has the following objec-
(1) to maintain a position of broad self-sufficiency in the main foodstuffs in order to enable the nation to be fed without using scarce foreign exchange on food imports; (2) to achieve a calculated degree of security of food supply for each area of the country; and (3) to ensure that these foodstuffs are distributed in a manner that every member of the population has a nutritionally adequate diet. The basic aim of this policy is to ensure that nutritious food is available in adequate amounts to every Kenyan. It is not intended to achieve selfsufficiency in all food products within the same time period. Acailability of food and nutrients. Recently calculations of food availability in Kenya shows that per capita food availability in Kenya was fairly steady from 1974 to 1979. From the food data was derived per capita availability of nutrients for the same time period. The results indicate that Kenya does not have a shortage of nutrients. On the contrary the cause of Kenya’s nutritional problems between 1971 and 1979 and in subsequent years is mainly maldistribution of food rather than its shortage. To alleviate the problem of malnutrition, more efficient ways of distributing food must accompany increases in food production. Purchasing power. There is empirical evidence to indicate that an increase in the purchasing power of low income households would increase their nutritional status proportionately more than that for higher income households. This point is illustrated in Table 6. The main point to notice from Table 6 is the behaviour of the expenditure elasticity for calories. This elasticity declines very rapidly as we move from very low income groups to very high income groups. The policy implication of this is that for any additional shilling received by these groups, the low income groups spend a greater proportion of it on consumption of calories than is the case with the rich groups. Food self-sujiciency. A strategy of self-sufficiency in food is faced with a set of climatic, socio-economic and political constraints among others. A realistic strategy of self-sufficiency in food production should take cognizance of certain trade-offs. Table 6. Food consumption by income group
Group Rural low income Rural middle income Urban poor Rural rich Urban middle V’rban rich
Share of income spent on food
Share of expenditure spent on food
Expenditure elasticitv for calohes
0.80
0.77
0.74
0.45 0.62 0.32 0.39 0.19
0.75 0.45 0.73 0.37 0.21
0.67 0.35 0.48 0.34 0.25
Source: Income and expenditure shares are computed from the Integrated Rural Suney I (1974-1975) and the Urban Food Purchasing Survey (1977). Elasticities are computed from calories expenditure data derired by Frahkrg and Shah [Ij].
in Kenya
779
The following trade-offs have been identified regard to food policy in Kenya:
with
(i) Food’Food trade-offs; mainly in Western Kenya where the continued pursuit of self-sufficiency in sugar is in conflict with that of maize. (ii) Food, Export crop trade-offs; these are mainly in Central Kenya where self-sufficiencv in maize is in conflict with the foreign exchange ea&ing crops such as coffee and tea. (iii) Food/Animal trade-offs; these are mainly in areas where livestock and crops must compete for the same piece of land. v. CONCLUSION In this paper, we have shown that there exists serious problems of ill-health and malnutrition in Kenya. These problems are caused by many and interrelated factors. These factors can be classified as direct, intermediate, and ultimate or basic. The manner in which these factors affect health and nutritional statuses must first be known before health and nutritional policies are designed and implemented. The existing data on health and nutrition in Kenya are sparse and unreliable for use in the design of practical policies. We suggest a systematic collection of data on health and nutrition at the household level be started. In particular, information should be collected regarding people’s perceptions of their health statuses, their expenditures on health care, the health facilities they use in the event of an illness. etc. With regard to data on use of facilities. special attention should be given to data on use of traditional healers with a view of finding ways of officially integrating traditional healers with modern health care sector. As for nutrition, data about growth profiles of children and the socio-economic conditions of their households is now available at the Central Bureau of Statistics, Ministry of Finance and Planning. Efforts to collect this kind of data should be continued. We should also like to call attention to the fact that the capability to analyse the collected data should be developed. We see no need to collect data that will not be analysed. An implicit and important point in our paper is that public spending on health and nutrition is a productive investment. It is an investment in health capital. Scarcity of health capital can impede the process of socio-economic development exactly the same way that shortages of physical capital. foreign exchange or skilled labour can hinder it. Another implicit and critical point in our paper is the issue of the relationship between improvements in health/nutritional status and the rate of population increase. We assume that a healthy and well nourished population will use the available resources more efficiently, thus increasing the supply of the available goods and services. Further, since infant mortality rate would be low in such a population. the well documented motive of couples having many children in order to increase the chances of some of them surviving would not be in effect. We will now conclude this paper with irs three major findings: (1) Health and nutritional statuses of a community are interdependent. None can be attained indepen-
780
WILFRED Xl. MU’,~SGI and
dentlv of the other. This idea of simultaneity m the pursuit of optimal states of nutrition and health, implies that in order to be effective our efforts to remove problems of disease and malnutrition in society must be coordinated with our other efforts elsewhere in the economy. (2) Attainment of self-sufficiency in food production is a necessary, but not a sufficient condition for security against malnutrition. Equitable distribution of food and of nutrients must also accompany food self-sufficiency. (3) At the farm level, the national policy of food self-sufficiency might be in conflict with the farmer’s rational management of his farm. Faced with relative costs of growing various foodstuffs and cash crops. a farmer will choose the crop combination from which he expects the greatest net return; or the combination that guarantees him some minimum amount of food. Needless to say his choice of crops might be different from that which would be consistent with a national policy of food self-sufficiency. A good understanding of farmers’ decision making at the farm level, and of how the decisions farmers make might be influenced by policy, is one of the first steps in a successful implementation of a programme of self-sutKciency in food production.
GERMANO M. MWABC
4.
5.
6.
7.
8.
9.
IO.
REFERENCES I. A
recent measure of social welfare is the Physical Quality of Life Index (PQLI). This is a composite index of a community’s life expectancy at birth. infant mortality and literacy rates. PQLI as a measure. attaches a significant weight on the provision and distribution of basic human needs. The traditional measure of social welfare, the Gross Domestic Product (GDP), ignores distributional problems in society. For details about PQLI see, Morris M. D. ,Measuring the Condition of Ihe World’s Poor: The Physical Quality of Life Index. Pergamon Press, New York, 1979. 2. Koinange W. A Report by the Director of Medical Services on Status of Health in Kenya, 1979. Ministry of Health, Nairobi, 1982. 3. Anker R. and Knowles J. An empirical analysis of mortality differentials in Kenya at the macro and micro levels. Econ. Dec. Cult. Change. 29, 165-155. 1980. The official survivorship in Nairobi is probably not lower because of adverse socio-economic conditions but because of more complete reporting, i.e. deaths are re-
II.
12.
13. 14.
15.
ported more completely. hence the survivorship goes down in comparison with a region where (especially early child) deaths are reported mcompletely. Ministry of Finance and Economic Planning. Kenya. Daelopmenr Plan 19511985. Central Bureau of Statistics. Third Rural Child :Vurrifion Srtmey. Ministry of Finance and Planning, Bairobi. 1982. While the mortality rate for this age group was 56!1000, their morbidity rate was 465,lOOO. The percentage of sick children varied greatly by province. It ranged from 37.1% in central province to 58.8% in Nyanza province. In some districts. e.g. Kisumu, Kilifi, Tana and Lamu. those percentages were as high as 64%. McCarthy F. D. and Xlwangi W. M. Kenya Agriculture: Toward 2000. International Institute for Applied System Analysis, Laxenburg. Austria, l9S2. Crawford E. and Thorbecke E. Employment Income Distribution, Poverty and Basic Seeds in Kenya. Report of an IL0 Consulting Mission. Cornell University. Ithaca, N.Y.. 1978 (mimeo). Collier P. and Deepak L. Porerr) and Growrh in Kenya. Staff Workinn Paper No. 389. World Bank. Washington, D.C.. 1980. . Greer J. and Thorbecke E. Patterns of Food Consumption and Poverty and Effects of Food Prices. Cornell University, Ithaca, N.Y.. 1983 (mimeo). Mwangi W. M. and %ligot-Adholla S. E. Food selfsufficiency in sub-Saharan Africa: approaches. problems and constraints: a case of Kenya. Paper presented at a Workshop on Food Self-Sujiciency in Sub-Saharan Africa, University of Dar es Salaam. Tanzania. 1984. Jonsson U. A conceptual approach to the understanding and explanation of malnutrition in society. Paper presented at the Workshop on “Hunger and Sociefy‘in Soliwayo”. Iringa. Tanzania, 1983. Garcia D. The effect of ownership and control of health systems and the concept of health. Paper presented at a Workshop on Health Policy on Eastern and Southern Africa, Arusha, 1981. Bigsten A. Regional Inequalilies in Kenya. Institute for Development Studies. DP., No. 330. Nairobi, 1977. We arbitrarily define any health centre that is more than 2 km from a household as inaccessible. Furthermore, access to a health centre includes access to a referral facility. This view assumes a terrain with homogeneous physical features. But in a non-mountainous area such as northern Kenya a health centre that is 4 km auay can be considered accessible. Frohberg H. C. and Shah M. M. Nutrition status of rural urban Kenya. \Iarketing Development Project. FAO, Nairobi, 1978.