Somalia - Population, health and nutrition sector review

Report No. 5468-SO Somalia Population Health and Nutrition Sector Review September 5, 1985 Population, Health and Nutrition Department FOR OFFICIAL USE ONLY X74::' .Z L f ;. d-: tM X Ad -k.3-.-.* . . . , .- . - . . , U f z --'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ '~ .r ...' - 0;;1 _ _ _ = . B . . _ . .C,..j.. o'' ' "- '- ' ' ' ~~~~~~~~~~~~. - '''-.' ' .'' f ' ~~~~~~~~~~~~~- - . ' ,. ' ' . ' * . '' '.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~, ~ - It' ' ' '3 4 Document of the-World Bank- ..- -.:= ' : This: document has: a' restricted distinbution and may be- used by recipients. ., -,.,' ',.~onlyirithe pefrnnceof their official duties. Its cntents may not otherwise .............................:.--... .... bediscoe withotWorldBank authoization..-. , 'C' ' '. @_ ,,' s 'a " -t' < ': .' ". *' ; ' 7 Douet-f-& ol Mn-  FOL OFFICA USE ONLY ABSTRACT The population of Somalia is estimated to be increasing at an annual rate of 3.OZ and to have reached 5.4 million as of 1985. If the present growth rate continues, the population will reach nearly 15 million in 30 years. The rapid rate of population growth is impeding efforts to move beyond low agricultural productivity and extend cultivation by diverting investment funds and foreign exchange to consumption and social infrastructure. The most critical population issue is the development of sufficient government and public support to mount an effective population program. An intensified information and education campaign should be mounted and accompanied by extension of 1CR1!P services. The health status in Somalia is among the worst in Africa with an infant mortality rate of 145 per thousand births. Commonly prevalent disease problems could be reduced through preventive care and simple curative measures, but delivery of services is impeded by extremely low coverage and urban bias in the distribution of personnel and facilities. Improvement of services is, in turn, hampered by low absorptive capacity. The highest priority in the health sector should be placed on removing the planning, institutional management and financial management constraints to efficient use of capital expenditures in the health sector. Protein calorie malnutrition in children less than five years of age remains a problem but the evidence of the magnitude is, incouclusive. Given the large volume of food aid received by Somalia, a nutritional survey is needed to guide the management of the food supplementation program. Micronutrient problems, especially anemia, have been neglected, and low cost food additive interventions should be considered. Intersectoral nutrition issues are particularly important. Food marketing, agricultural development policies and nomad resettlement all have nutritional implications that need to be considered in the design of policy. This document has a retricted distribution and may be used by recipients only in the performance of their official duties. Its contents mnay not otherwise be disclosed without World Bank authorization.  SONALIA BASIC DAI Population: 5.4 million Annual Rate of Population Growth: 3.0Z Dependency Ratio: 92.0 GNP per capita (1982): US$2902 Population Density (per Ki2): 8.4 Crude Death Rate: 19 Crude Birth Rate: 49 Infant Mortality Rate: 145 Life Expectancy at Birth: 47 years Total Fertility Rate: 6.7 Key Service Indicators:3 Contraceptive Prevalence Rate: <1% (modern and traditional methods) Immunization Rate: Children Under Five (DPT, measles): 11% Protected Birth (Tetanus): no estimate available Percentage of population with access to and knowledge of oral rehydration therapy: no survey information is available. ORT is part of the PEC and MCHIFP programs currently reaching less than 202 of the population. Proportion of children less than five vith acute protein-calorie malnutrition: 26Z 1 All data are estimates for 1985 unless otherwise indicated. See Chapter II and Annex I for details of the population estirates. 2 In 1982 US$ 3 Arbitrary estimates based on available data and circumstantial evidence.  ABBREVIATIONS ASPDNA Parastatal Pharmaceutical Agency ASR Agricultural Sector Review (World Bank) CBR Crude Birth Rate CDR Crude Death Rate CIR Community Health Worker CRS Contraceptive Retail Sales ECA Economic Cnmmision of Africa EEC European Economic Community EPI Expanded Program on mmmnnization FP Family Planning FYDP Five Year Development Plan GDP Gross Domestic Product IEC Iuformation, Education and Communication DMi Infant Mortality Rate IPPF International Planned Parenthood Federation iDn Intrauterine Device KAP Knowledge, Attitudes and Practice (Family Planning Survey) MNC Maternal and Child Health MR Ministry of National Planning MOE Ministry of Education MOF Ministry of Finance MOR Ministry of Health NHP National Health Plan ORT Oral Rehydration Therapy OXFAM Oxford Committee for Famine Relief PCM Protein-calorie Malnutrition PEC Primary Health Care PHCU Primary Health Care Unit PHI Population, Health and Nutrition PIP Public Investment Programme RMO Regional Medical Officer SERCA Somalia Family Health Care Association STD Sexually Transmitted Diseases SWDO Somali Women's Democratic Organization TBA Traditional Birth Attendant TFR Total Fertility Rate UNICEF United Nations Childrens Fund UNESCO United Nations Educational, Scientific and Cultural Organization UNFPA United Nations Fund for Population Activities UPE Universal Primary Education USAID United States Agency for International Development VDR World Development Report WES Women's Education Service VFP World Food Program WHO World Health Organization  GLOSSARY Crude Birth Rate: The number of live births per year per 1,000 population. Crude Death Rate: The number of deaths per 1,000 population in 8 given year. Ciassifications of Malnutrition: Two major classification systems are currently used to describe the severity and the type of malnutrition among children. The Gomez classification system identifies 3 increasingly severe degrees of malnutrition by comparing weight-for-age to a reference standard. Those children vhoses weight-for-age falls between 75Z-90X of the reference median are mildly malnourished (10); those who fall between 60-75Z of the median are moderately malnourished (20); and those who fall below 60% of the median are severely malnourished (30). The Waterlow classification differentiates between chronic and acute malnutrition, based on measures of weight-for-height and height-for-age relative to reference standards. Children with relatively low weight-for-height but adequate height-for-age are wasted, the result of acute, or short-term malnutrition; adequate weight-for-height, but low height-for-age indicates stunting or chronic, long-term malnutrition; children with low weight-for-height and low height-for-age are concurrently wasted and stunted. Dependency Ratio: The population 14 years of age or under and 65 years or over as a percentage of the population aged 15 to 64 years. Infant Mortality Rate: The number of deaths of infants under one year of age in a given year per 1,000 live births in that year.  Life Expectancy at Birtb: The number of years a newborn child would live if subject to the age-specific mortality rates prevailing at time of birth. Morbidity: The incidence of disease in a population. Nortality: The incidence of death in a population. Protein-calorie Malnutrition: A clinical spectrum of suboptiuum nutritional states ranging from mild undernutrition and growth retardation to the most severe states of kwashiorkor and marasmus. Rate of Natural Increase: The rate at which a population is increasing (or decreasing) in a given year due to a surplus (or deficit) of births over deaths, expressed as a percentage of the total population. Rate of Population Growth: The rate of natural increase adjusted for (net) migration and expressed as a percentage of the total population of a given year. Total Fertility Rate: The average number of children that would be born alive to a woman during her lifetime if Mhe passed through her childbearing years conforming to the age-specific fertility rates of a given year.  SONALIA POPULATION. HEALTH AND NUTRITION SECTOR REPORT Table of Contents Pale No. MAP .................... ......................................... 15413R SUMMARY ...................... ..... ...... . I. Economic and Social Settin g.......*ee.es*.e.ee.******.*..e.** 1 II. Population .....e.e...c.ee......eeeceeeesecseseeseeecseeeee 2 A. Present Demographic Situation and Trends ..se....e...e.ee*ee 2 B. Population Projections .6e.ue... ...............se... 6 C. Consequences of Population Growth ...ese....eesem.eseeeessse 8 D. Population and Family Planning Policies ... cc... seine..... 14 E. Family Planning Services ............ e me eec.. 55m sees 16 III. BHealth ---e--e.e-----scec.......cece........ ...e..e 17 A. Health Status .... ........c................ 17 B. Health Policies and Strategies cscecsecsceece.esccc.secs. 20 C. Organization and Management smeme..... 22 D. Facilities and Services s. c sc ................... 24 E. Health Support Systems see.........cc.. see.c.e.som... cc s.... 28 F. Realth Personnel and Training ....c.c..cessecsesecee.esssc.. 30 G. Health Expenditures ....... ccc ............se. esesessec 33 IV. Nutrition .......ec.....ee.c...e.essc.......... 38 A. Current Nutrition Status ....c..............c.....e.ss.m.c...... 38 B. Causes of Malnutrition ...see....es.see.se...... . 39 C. Nutrition Activities ....................e42 V. Issues and Recommendations ........ ................. 45 A. Population .....c....c.. .... c..........s 45 B. Realth L..c.sce .... ..e... s.e...ece 47 C. Nutrition ..e...e sc... ......e.. ..... c.. 53 This report is based on the findings of a PHN mission that visited Somalia in May 1984. The missior was composed of Mr. Howard Barnum (economust, mission leader), Ms. Althea Hill (demographer), Dr. Atiqur Rabman Than (public health specialist), and Mr. Christopher Stevens (nutritionist).  ANNEXES Page No. ANNEX IA Table 1.1: Size of Population by Sector, 1975 Census ............... 55 Table 1.2: Mortality and Fertility Assumptions used in Population Projections ................ * . ....... ....... 56 Table 1.3: Projected Size of the Population of Somalia by Economic Sector: 1975-2015 ........................... 57 Table 1.4: Projected vital rates of the Population of Somalia: 1980-2015 .... * ................................ 58 Table 1.5: Projected Distribution by Age of the Population Somalia: 1980-2015 ...............*.........*...** 59 Table 1.6: Projected Growth of the Working Age Population of Somalia: 1980-2015 ....................... 60 Table 1.7: Projected Growth of the School Age Population of Somalia: 1980-2015 ...................... . 61 Table 1.8: Projected Growth of MCI Target Populations of Somalia: 1980-2015 ..... ................ 62 ANNEX 1B Technical Note on the Population Projections ........................ 63 ANNEX 2 Table 2.1: Mortality Indices and Birth Rate Quoted by Different Sources ............................. 65 Table 2.2: Common Causes of Child Death in Different Age Groups by Urban/Rural Residence Status .......o....... 66 Table 2.3: Selected Evidence of Childhood Diseases ................. 67 Table 2.4: Prevalence of Selected Diseases per 1000 Population in Somalia by Sex and Overall, 1980-82 ....... 68 Table 2.5: Health Facilities in Somalia ........ ....................... 69 Table 2.6: Population/Health Personnel for Different Regions of Somalia for 1983 ... .......................... 70 Table 2.7: Health Personnel Production Targets of the National Health Plan for 1980-85 and Actual Manpower Trained for 1980-1983 .......................... 71 ANNEX 3 Table 3.1: Trends in Nominal NOR Expenditures ...................... 72 Table 3.2: Trends in Real NOR Expenditures ........ ................. 72 Table 3.3: Trends in Real Government Health Expenditures Per Capita .... ... 73 Table 3.4: Percentage Relationships Among Gross Domestic Product (GDP) Total Government Expenditures, and MOR Expenditures ...........*............ 74  ANNEXES (continued) Paste No. Table 3.5: Trends in Government Recurrent Expenditure by Functional Category ........ ........................ 75 Table 3.6: Foreign Assistance in Health and Population ............. 76 Table 3.7: Summary of Planned Public Investment Programme .......... 77 Table 3.8: Projected Net Availability of Recurrent Funds with Supplemental Expenditures for Maintenance and PIP Recurrent Expenditures **......... .................... 78 Table 3.9: Projected Potential Funds from Cost Recovery ............ 79 ANNEX 4 Table 4.1: Incidence of Child Malnutrition from Selected Surveys ... 80 Table 4.2: National Morbidity survey Findings on Endocrine and Nutritional Diseases ............ ...... a cc.. 80 FIGURES Figure I: Organizational Chart of the Ministry of Health .......... 81 Figure II: PlC Organizational Structure .......... .................. 82  SOMALIA EECUTIVE SUMOARY A. PODU1atiOn Current Situation 1. The population of Somalia is estimated to be increasing at an annual rate of 3.0X and to have reached 5.4 uillion as of 1985, ezcluding approzimately 500,000 zefugees. The high rate of population growth results from high fertility (the total fertility rate is about 6.7 compared to 8.0 in Kenya and 6.8 in Ethiopia) coupled vith high mortality (the crude death rate is approximately 19 per thousand compared to 12 in Kenya and 15 in Ethiopia). Improvements in health services, especially preventive care, can be expected to result in an even higher population grovth rate in the future if fertility continues at the current high level. 2. Approximately 502 of the population is nomadic vith the remaining 50 almost evenly split between rural settled and urban. Most of the rural settled are located in the southern river regions of Juba and Shebelle, vhile in the 'entral and Northern regions 622 to 802 of the population are nomads. The large percentage of the population that is nomadic presents a special challenge for the delivery of health and population services, and programs for Somalia, especially in the North, must be uniquely tailored to meet this challenge. Although Somalia is among the most sparsely populated countries in Africa, with an estimated national density of 8 per km2, the rapid rate of population growth is impeding efforts to move beyond low agri- cultural productivity and extend cultivation by diverting investment funds and foreign exchange needed for agriculture, nomad resettlement and economic growth to consumption and social infrastructure needed for a larger population. The PoDulation Problem 3. The tremendous potential for population growth in Somalia is clear from Table 1 on the next page. Without major government intervention, the population will increase by around 602 to 8.5-8.7 million within the next 15 years and will multiply to at least two and a half times its present size, to reach 13-15 million, in 30 years time. Moreover, massive capacity for future growth will remain at the end of the projection period, with growth rates permitting a further doubling of population within another 19 to 27 years. A fertility decline on the scale likely to occur spontaneously would have only a limited impact, reducing the 2015 population size by 132. Strong government action will be required to reduce population increase to any really significant degree. An accelerated fertility decline would reduce the 2015 population size by 32Z.  -ii- Table 1: PROJECTED SIZE AND GROUTH OF nATIONAL POPULATION Ratio of Population Year size to 1985 Base in: ].85 2000 2015 20 1 Total Ponulation (millions) No fertility decline 5.40 8.67 14.81 1.6 2.7 Gradual fertility decline 5.40 8.51 12.89 1.6 2.4 Accelerated fertility decline 5.40 7.91 10.12 1.5 1.9 Doubling Time at 1980-85 1995-2000 2010-15 2010-15 Growth Rates AUrIrLe APuAl Rate of Population Growth (Z) No fertility decline 2.9 3.3 3.7 19 years Gradual fertility decline 2.9 3.0 2.6 27 years Accelerated fertility decline 2.9 2.1 1.4 50 years Source: World Bank Projections. 4. In the absence of a major government population program, Somalia must plan for a minimum population of at least 13 million vithin 30 years time, well over double its present size. It is vital to assess the consequences of such rapid growth. The key question is whether Somalia can develop its natural resources fast enough to outstrip population growth and to support two or three times its current numbers, at improved standards of living, within the next 20 to 30 years. 5. Over the decade of 1971 to 1981 the average annual rate of economic growth of approximately 1.2Z was sufficient to give only a modest increase in per capita GDP. Nost of this increase was in the service sectors. The increase in the livestock herd just kept pace with population growth while  -iii- per capita food production decreased. Bank projections of an average rate of growth in GDP of 4Z in the is.ediate future are more sanguine, but these projections assume improved agricultural productivity, increased industriali- zation and no serious drought. If these assumptions do not prove valid the rate of growth of GDP would be less than the current 2.8Z rate of population growth. Even beyond the next decade, the rate of population growth will de- termine whether annual growth in GDP is translated into true increases in economic well being or merely into continued subsistence levels for an in- creased number of people. The economic implications of population growth are particularly significant for nomad agriculture and growth in social services. 6. Imlications for Nomad Atriculture. Without strong government population policies and programs, the nomad population vill increase by almost 50% over the next 15 years and more than double over the next 30 years; another 1 to 1.1 million Deople would be added by 2000 and a total of 2.5 to 3 million by 2015. Similar increases in herd size would therefore be required merely to maintain current subsistence and income levels among nomads; the rangeland would have to support another 15 million animals within the next 15 years and a total of 37 to 45 million extra animals in 30 years time. There may be some scope for improving herd productivity. The possibilities for berd improvement are being examined in on going Bank livestock sector analysis. But even with plausible productivity improvement the rangeland may not be able to accomodate herd increases on the scale required by projected nomad population growth. 7. If the number of nomads is to be stabilized, up to a million nomads would have to be absorbed into the agricultural or urban sectors within the next 15 years, and up to 2 or 3 million within 30 years from nov. The agricultural sector could at most absorb about 1.2 million additional populAtion by 2000, at a total capital cost (whicb must be largely borne by the government) of US$1.5 billion. However, natural increase in the agricultural sector must also be taken into account. If tbere is no strong government intervention to control population growth, the agricultural population would itself increase more than 602 over the nezt 15 years and would multiply to two and a half to three times its present size within 30 years from now. Thus, out of the expected 1 to 1.1 million nomad natural increase, only about one quarter could be absorbed by settlement into the agricultural sector, leaving about three quarters of a million still to be accommodated elsewhere. B. Movement into the urban sector is another possible option for the surplus nomad population. Again, however, the absorptive capacity of the urban sector is limited, and must in addition accommodate urban natural increase which will be high because of relatively low mortality, still elevated fertility, and a large proportion of women of childbearing age. The urban population vould grow through natural increase alone by around 70S over the next 15 years, and would reach 2.7 to 3.3 times its current size 30 years from now. If the surplus 0.75 to 1.25 million from the nomad sector were to move into the urban areas as well, the urban population by the year 2000 vould more than double to between 3.3 and 3.9 million, or at least 40S of the total population. Given the lack of industrial and urban development in Souslia. it is difficult to imagine how such a rapid increase in the urban population could be accommodated at reasonable levels of income and  -iv- living standards for urban dwellers. Urban growth on this scale would demand heevy public expenditure on housing and basic services. 9. The fisheries sub-sector remains as one other possible option for the surplus nomad population. Hovever, there have been difficulties in resettling nomads ii fishing communities in the past (see the 1981 ASR, Annex 4, pp. 17 and 18), and this is not considered likely to be a popular option. The necessary changes in attitudes and culture may be slover than the timeframe of 20 to 30 years imposed by the rapidity of nomad population growth. Moreover, heavy capital investment in infrastructure and services would again be required, amounting to US$1.5 billion if there is no change in fertility. 10. The difficulty of accoimodating the growth that will occur in the nomad population in the absence of a fertility decline can be illustrated by a comparison of the cost of resettlement with the potential increment in annual GDP. If the overall economic growth r_ee of 3.51 to 4Z per annum projected by the Bank for the period 1982-19861 could be maintained up to the year 2000, the average annual gain in GDP of US$34-40 million (1984 Prices) would be more than absorbed by the annual costs of the nomad resettlement, which would amount to approximately US$90 million annually over the next 15 yearsO2 11. Implications for Social Infrastructure. The government of Somalia places a high priority on improvement of educational levels, including the attainment of universal primary education (UPE), and has made considerable efforts in education over the last decade. But rapid population growth has diluted the efforts and limited the educational coverage achieved. From 1975 to 1980, for example, primary school enrollments increased by 241; but because the population of eligible children grew by 15X during the same period, only a modest improvement in the enrollment ratio from 26Z to 28Z was possible. Rapid population growth will continue to hamper progress in the future. If there are no government population policies or programs, the number of children eligible for primary schooling would increase by 641 over the next 15 years and vould swell to between two and three times its present level within 30 years from now. Primary educational expenditures would have to be correspondingly increased to reach So. sh. 210 million (US$12 million) in 2000 and So. sh. 300 to 370 million (US$17 to 21 million) in 2015, just to maintain the current enrollment ratio and quality of schooling. To achieve UPE within the next 30 years, funding would have to increase by 6.5Z to 7.21 per annum to a level of So. sh. 900 to 1100 million (US$50 to 65 million) or 8 to 10 times the present levei in real terms. A substantial expansion of primary schooling in both coverage and quality would be considerably eased by an early and rapid fertility- decline. With an accelerated fertility decline, educational expenditures over the period 1985 to 2015 would need to increase by only 1.21 per annum to maintain the current enrollment ratio. UPE could be achieved by 2010 1 World Bank, Policy Measures for Rehabilitation and Growth, May 1983. 2 Assuming one quarter of the resettlement is in agriculture and three quarters in the fishing sector.  with an increase of real primary schooling expenditures to 4 times the present level compared to the 8 to 10 times needed without fertility reduction. 12. Rapid population growth would also have a substantial irpact on the demand for health services, which, like education services, are disproportionately utilized by the youthful and the dependent sections of the population. If fertility does not decline, the MCI target population and annual numbers of deliveries would almost triple within 30 years from now. The burden on NCH services would increase proportionately, forcing government to devote a considerably higher percentage of total expenditure to health care if coverage and quitity of services are to be improved over the current inadequate leve.Ls. In contrast, with a more rapid fertility decline, the MCR population would grow by only 70X and the demand for obstetric services would fall by 14Z, thus permicting faster extension and improvement of NCI services with an unchanged sbare of total government spending. Policy Issues 13. A concerted government effort to develop an information and education campaign to stimulate public demand for family planning and to develop a delivery network to supply family planning services will be required to significantly reduce populaLion in growth 20 years. If a contraceptive prevalence rate of over 20X could be achieved within the next ten years, the 2015 population could be reduced by over 30x. 14. The Government has not adopted a national population policy and family planning (FP) is not promoted as a measure to reduce the rate of population growth. Somalia has only recently emerged from a pronatalist policy. The prevailing socio-cultural situation also presents a difficult environment for the implementation of a FP policy. Nevertbeless, there is some government support for PP linked to maternal child health programs and over the last three to four years child spacing has been promoted as part of the Ministry of Health (MRE) maternal and child health program. 15. In 1982 a Family Health Divisioi was created in the MNE to plan and coordinate family health and PP-activities. Acting on a pLan of action set out by the Family Health Division, PP 5ervices have been introduced in 31 of the existing 95 NCR (maternal and child health) centers. Use of services has remained lowv with a contraceptive prevalence rate of less t'oat 1 in regions where services are offered. The use of contraception in other regions is negligible. In spite of the low level of current contraception, innovative family life programs being implemented by the ministry of education (MOE) and the information, education and communication (IEC) activities being started by the Somali Women's Democratic Organization (SWDO) and the Somali Family Health Care Association (SFHCA) hold potential for increased demand in the future. 16. The most critical population issue is whether sufficient government and public support can be cultivated to implement an effective population program. The weak delivery infrastructure is also a major constraint. Population and health strategies are closely linked and MOE capacity to offer services is restricted by the low coverage of the rural and nomadic  -vi- population vith health services, poor logistical support and lack of trained manpover. Recommendat ions 17. Immediate Initiatives. The %.Ljor constraints in proceeding with population programs are the pronatalist traditions of the Somali culture the slow change in the government position from pronatalist to active promotion of family planning. A country-wide information, education and communication (IEC) campaign needs to be developed to promote child spacing and to lay the ground work for an expanded service program to reduce population growth. The implications of population growth should be clarified among government policy makers and middle level staff involved in program implementation through seminars and policy papers developed by the Ministry of National Planning (MNP) and MOR. It is suggested that the SWDO and the SFHCA should be used at the forefront of IEC activities and should be given expanded staff and resources for this purpose. The population education activities of the MOE, currently supported by UNESCO, should be continued and expanded. 18. Subsequent Steps. In tandem with the IEC effort, the national CHC/FP program and rural health service programs should be increasir4ly strengthened to provide a higher volume of FP services. However, the population and health strategies are closely interlinked. The urban bias of the health system and extremely low coverage of the population with basic health services make it difficult to use the system of rural health centers as the delivery vehicle for family planning services. Important areas for improvement are physical facilities, logistics, management, and training. If specifically programmed and provided with the necessary resources, the delivery of FP services will be enhanced as progress is made in strengthening MOE services. A higher rate of expansion of the NCR/PP service facilities, strenghthened logistics and training should accompany the efforts to improve health management recommended below. Additionally, the urgency of the population problem requires that community based programs, contraceptive retail sales and peristatal agencies be used to deliver services outside of government clinical programs. 19. Culturally compatible non-government channels of promoting and providing family planning services for child spacing should be developed. Program design should incorporate anthropological and sociological advice. Consideration should be given to the establishment of a social marketing program. The incorporation of traditional birth attendants (TBAs) into the strategy for IEC and service delivery, possibly by allowing TBAs to market contraceptives, should be considered. B. Health Situation and Issues 20. The health status in Somalia is among the lowest in Africa. Estimates made by the Bank mission are that the crude death race is 21 per thousand population, the infant mortality rate is 150 per thousand births and life expectancy is 46 years. Commonly prevalent disease problems are diarrhea, communicable childhood diseases, *malaria, tuberculosis, obste-  -vii- trical complications and nutritional aeficiencies. Many of the diseases accounting for the high mortality are avoidable with known preventive health meaaures, low cost curative care such as oral rehydratiou, increased birth intervals, and environmental control, but all of these interventions require effective organization and infrastructure to reach the people. Female circumcision is also an important factor contributing to birth complications and other medical problems. 21. The National Health Plan (NHP) prepared in 1980 recognizes the past neglect of preventive services and -the inaccessibility of health services to 85% to 90Z of the population. To correct these problems the plan emphasizes the extension of basic health services to the rural population by linking health with other development activities at the comunity level and strength- ening services at the regional and district levels as a back up to the basic community services. Overall, the NHP represents an important step to reorient the health services to meet the needs of rural areas. The Development Strategy and Public Investment Programe (PIP) includes a set of projects for primary health care, nutrition, TB control, immunization, health manpower development and maternal child health that are all consistent with the NHP. However, the PIP also includes a teaching hospital and large rural hospital that are not consistent with the emphasis of the NEP on primary health care. 22. The abrupt increase in capital expenditures since 1981 has exceeded the absorptive capacity of the health sector. The highest priority in the health sector should be placed on removing the current obstacles to effective use of donor funds in support of the major plan goals. Donor funding has been forthcoming to support the development of the health sector with primary health care projects in seven regions (there are 12 regions in Somalia), maternal and child health projects, construction of facilities, and massive food aid. Unfortunately, coordination of donor aid has been weak and not all of these projects have been mutually consistent. Nor have the available resources been absorbed effectively. 23. The major constraints to greater absorptive capacity are a lack of trained management (para. 3.19), inadequate planning capacity (para. 3.14 and 3.15), a limited information and monitoring system (para. 3.22), maldistribution of health personnel (para. 3.43 and 3.44) and insufficient recurrent funds (para. 3.57). An additional, and highly important constraint, is the inability to provide adequate financial planning, auditability and accounting (para. 3.58). It is important that either donor expenditures be reduced or the constraints released to avoid the long run institutional- ization of cost-ineffective administrative end financial management procedures brought about by a continued input of greater funds than can be efficiently absorbed. Reconmendations 24. Immediate Initiatives. There are serious deficiencies in services that could be remedied by efficient use of capital, but there are clear indications that currently the MOR cannot adequately absorb the available donor funds. The first priority in the health sector is to eliminate the constraints to greater absorptive capacity. The constraints include inadequate planning capacity, inefficient management and monitoring, and inadequate  -viii- financial planning and accounting. Only after relaxation of these constraints, can an effort be made to bring about the substantive, durable improvements to health infrastructure required for greater coverage with health and family planning services. Long-term technical assistance is required in each of these areas to build up the absorptive capacity of the MOE. The problems in organization, management, planning and finance are all closely interrelated and remedial steps for all three areas need to be coordinated. Preparatory studies need to be carried out to: (a) outline the basic requirements in the areas of health program management, financial management, planning, and manpower; (b) develop organizational alternatives and administrative, planning and financial management procedures; and (c) identify long term technical assistance and training needs necessary to carry out the suggested improvements. 25. Recammended foci for change in the three constraining areas are set out in the text (para. 5.09-5.19) and briefly summarized below. (a) Planning. The NEP should be translated into medium-term and annual implementation plans. A planning process needs to be developed that would link planning and budgeting and provide for dynamic revision of the medium-term plan. An inter-ministerial mechanism for a project approval process coordinated by the Ministry of Planning is necessary to ensure that health planning objectives and other sectoral development objectives are complementary and consistent. The planning process should involve a detailed recon- ciliation of planned capital expenditures with recurrent cost capacity. (b) Management. Responsibilities of senior and mid-level staff need to be defined better within the existing organizational framework. The management capabilities of the MOR need to be strengthened at the center through training for senior staff who could then carry out short duration retraining of mid-level staff. The highest priority support areas for management training are logistics, maintenance, service staff supervision and accounting. Information and surveillance systems should be developed to provide management with performance indicators and planning data. The ongoing MORIIWRO project to initiate a reporting system should be further supported to include process data for all programs as well as periodic epidemiological impact data. (c) Finance. Improved financial management and accounting information are essential complements to improved planning and management. Accounting and budgeting headings need to be established to conform with MOE organization and provide knowledge of the flow of funds by function. Reconciliation of budgeted and actual flows would add greater control over programs. Appropriate financial management would also improve donor coordination by providing information necessary to reconcile donor capital expenditures and recurrent cost capacity.  -ix- (d) Recurrent Costs. The recurrent cogt problem deserves emphasis. Conservative projections of the required recurrent funds needed for the continuation of existing services and operation of planned projects over the next five years (1985-1989) are giveu. in Chapter V. When these projections of recurrent cost requirements are compared -to a projection of available funds the result is a total shortfall of So. sh. 1330 (US$76) million over the five year period. Even if the large hospital projects in the PIP are omitted, the five-year shortfall remains high at So. sh. 1080 (US$61) million. To cover the shortfall a combination of greater government financial commitment to health, revision of capital expenditures to emphasize more cost effective PHC and FP/MCH projects, donor support of recurrent cost and cost recovery from clients are needed. 26. The practice of infibulation and female circumcision poses a serious health problem. A vigorous IEC campaign as a part of MOE MCR/|P activities and MOE programs should be mounted to make the public aware of the hazards. The GOS should give material and moral support to the SWDO's efforts to curtail the practice. In formulating the IEC campaign, the experience of neighboring countries should be reviewed to derive culture-specific strategies. 27. Subsequent Steps. After relaxation of the constraints identified above, extension of primary health care (PEC) services to a broader population base and improved quality of basic referral services is needed. Additionally, special efforts should be made in affected regions to control tuberculosis and schistosomiasis. The strategy for strengthened health services should exrressly concentrate on cost effective programs of basic care to achieve a higher coverage. Construction of large scale central facilities should be avoided as it would further divert recurrent funds from this strategy. 28. The PEC program should only be extended as managerial and planning capacity is built and experience with the existing projects has been used to design improved supervision, training and coverage. The program should continue to rely on paid workers and strategies to promote broad community support, while concentrating on a minimum number of interventions-imimuni- zation, ORT, nutrition, family planning, antenatal care, and first aid. The NCHIFP program, which is not uniformly developed or accessible throughout PRC regions, should be incorporated with PRC and extended as rapidly as possible. 29. Improvement of referral services should concentrate on improvement of quality rather than increased numbers of facilities. In particular, the large scale hospitals included iu the PIP should be postponed to release investment resources for higher priority MCRI/P projects. There are selected regions where small scale facilities are needed, but the greatest current problems are bringing about a more equitable distribution of staff (especially doctors), inadequate provision of drugs and supplies, lack of running vater and sanitation facilities, and inadequate standards of care in existing facilities. Activities to attack these problems include improved housing and incertives for staff relocation, improved logistics mmangement and distribution as well as increased quantity of drugs and supplies, upgrading and repair of facilities, and staff retraining. The highest priority with  x-X regard to facilities should be given to increased funding and retraining of staff for improved maintenance. 30. To respond to the emphasis on rural health services, it is recomumended that the N0 develop a personnel planning unit headed by a trained manpower planner. A task force is needed to reviev the current personnel situation and initiate personnel planning. The amount and quality of training for MCH/FP should be increased for all categories of service staff. Recruitment of staff should emphasize cultural compatibility and commitment to delivery of rural services. C. Nutrition Situation and Issues 31. Data on the incidence of nutrition problems are scanty and come from clinical surveys or household surveys for specific geographic areas. With the exception of protein-calorie malnutrition, however, the surveys present a coherent picture. They suggest that the main nutrition related health problems are iron deficiency anemia, goitre, diarrhea, and night blindness, and that small children and pregnant or lactating women are particularly vulnerable. In the case of protein-calorie malnutrition, however, there are discrepancies among the surveys with the percent of children malnourished at degree 2 or above ranging from 18 to above 80, with the best estimate approximately 26Z. 32. An important issue is to identify the situations and groups that have the highest nutritional vulnerability. The high protein diet and access to cash income from livestock exports make the nomads as a group relatively less vulnerable to malnutrition. Also, the greater mobility of nomads renders them better able to survive short droughts than are other population sub groups. Nonetheless, the current reduction of livestock exports to Saudi Arabia coupled with prolonged drought is creating the potential for a nutritional crisis among nomads in the near future. Cultivators are more vulrnerable than nomads to drought. Small farmers are unlikely to be able to afford to hold large stocks. Their children are also more vulnerable than are the children of nomads because of the more limited availability of milk as a food supplement. Nutrition surveys indicate that the poorest groups in urban areas are the most vulnerable nutritionally. 33. The nutrition-related activities currently supported by the Government are having a moderate impact, in spite of the large volume of concessional imports (the total value of food aid in 1984 was US$45 million). The largest activity outside of refugee assistance isa WFP-assisted program of supple- mentary feeding to pregnant and lactating mothers and pre-school children through the MCR centers. Apart from this, there exist a limited nutrition education program. Government policy recognizes malnutrition as a serious problem and nutrition has been identified as a priority in the NBP. The nutrition program outlined in the NHP appropriately links with MCH- services  -xi- and other basic health programs but is too narrowly focused on a clinical program to combat protein-calorie malnutrition. Nicronutrient problems and intersectoral food policy have been neglected in Somalia's nutrition strategy. Recommendations 34. The Government has appropriately assigned a high priority to combatting malnutrition but the approach to nutrition programs has remained narrowly focused on reducing protein-calorie malnutrition (PCM) through food dictribution schemes. The nutritional efforts need to be broadened without reducing the attention given to maternal and child PCM. The narrow approach has resulted in a neglect of the intersectoral issues linked to the causes of malnutrition. Additionally, it has contributed to a relative neglect of anemia which is the most widespread nutritional disorder. 35. Imediate Initiatives. Food marketing, agricultural development policies, livestock exports, sedentarization of nomads and migrant policies have nutritional implications that need to be brought out. Development of strategies to deal with these aspects of nutrition and contingency plans to cope with drought or a drop in livestock income involve broad intersectoral issues. The MOB retains a responsibility to provide the clinical information necessary to enunciate and underline the nutritional implications of policies. The analysis of broader aspects of policies is, however, a shared intersectoral responsibility and must involve ministries and donor programs outside of health. Technical information on incidence and clinical aspects of nutrition should be made regularly available by building on the nutritional division that has recently been started in the MOB. 36. Subsequent Stens. Nutritional planning would be aided by improved nutritional information. There is a lack of clarity about the nutritional status of pre-school children, on whom most attention has been focused. Given the large size of the international assistance to nutritional programs, an additional effort is required to obtain both process and output indicators to monitor the functioning of programs and to assist in targeting and design. 37. The feasibility of measures to overcome micronutrient deficiencies in the Somali diet should be investigated. A high priority should be given to the feasibility of iron fortification of sugar, which forms a major part of the diet of the poorest urban groups and is used throughout Somalia. Iodine fortification of salt should also be investigated.  I. ECONOMIC AND SOCIAL SETTING 1.01 Somalia is located in the Northeastern corner of Africa. It faces the Gulf of Aden and the Indian Ocean towards the North and East and borders Djibouti, Ethiopia and Kenya in the West and South. With a total estimated population in 1984 of 5.2 million scattered across an area of 638,000 sq. km., the density is 8.2/sq. km. making it one of the most sparsely populated coun- tries in Africa. The varied topography includes a hot and arid coastal plain, mountains and plateaus, and lowlands of varying fertility and rainfall. Only 13Z of the land is estimated to be cultivable through rain-fed or irrigated agriculture. Rainfall is very irregular and groundwater resources are limited so that drought is a continuing risk. 1.02 Per capita income was estimated to be US$290 in 1982, placing it among the least developed countries in the world. Agriculture is the dominant sector of the economy. Approximately 46% of the population are nomads who depend on livestock for their livelihood; 29Z are farmers, and another 251 are engaged in non-agricultural occupations. Somalia's industrial base is minimal and the mineral resource potential is largely unassessed. At the present time the possibilities of diversification outside of agriculture appear limited. Government services and infrastructure, especially in trans- portation, health, education, water and sanitation, are limited. Less than 6O0 of the urban population and 20% of the rural population has access to reliable water of adequate quantity and quality. Urban sanitation covers less than 50% of the population and rural sanitation is negligible. 1.03 A number of political events have had a marked effect on the development of the economy and social sectors. Following the revolution in 1969, the Government greatly extended public sector ownership and control in the economy with a subsequent fall in productivity and stagnation in gross domestic product per capita over the 1970s. Since 1981 the Government has pursued a more liberal policy with reduced control over private activities such as agricultural marketing, but with many service sector activities re- maining largely Government monopolies. The conflict with Ethiopia in 1977 and 1978 and the continuing tensions have had an adverse impact on the Somali economy as resources have been diverted to defense expenditures. The war also caused an influx of refugees reaching a lwvel of 1,200,000 in 1980181 and subsiding to about 500,000 in 1984. 1.04 From 1977 to 1980 there was a sharp decline in commodity production accompanied by a rapid increase in domestic expenditure that led to large balance of payments deficits and high inflation. From 1980 to 1982 domestic production, especially of agricultural products, recovered dramatically, possibly aided by good weather and economic reforms. Despite this short- term improvemeut, the excessive demand expansion and the threat of poor agri- cultural output remain, and Somalia continues in a state of finncial and economic crisis. An analysis of the issues, and recomendations to achieve a higher and more stable rate of economic growth, are outlined in the forth- coming economic memorandum. However, the difficulty in achieving sustained growth in real output is coupled with a high rate of population increase (discussed in detail in chapter II, below) that is continuing to impede devel- opment of per capita income.  2 1.05 Social programs in general are difficult to implement in the Somali setting because of the scattered nomadic population, the low level of infra- structure and a shortage of technical manpower. However, in spite of economic difficulties and the difficult setting, hope for progress in health is held out by the progress Somalia has made in meeting educational needs. Between 1969 and 1978, the primary school enrollment ratio rose from 9Z to 30Z, and following a mass literacy campaign, the Government introduced a formal three-year primary equivalency program for adults, and the adult literacy rate is reported to have risen from 9X to about 50Z. The population is almost entirely Muslim and shares a high degree of cultural, and in spite of the wide geographical dispersion, linguistic unity although there are ethnic clan subgroups. The relative religious and ethnic homogeneity compared to much of Africa are additional positive factors that could aftect the design and ease the implementation of social programs in the future. II. POPULATION A. Present DemograDhic Situation and Trends Introduction 2.01 Penographic data are scanty and incomplete for Somalia at present. Only one census has ever been held, in 1975, and full results have still not been published, although an analytical volume vith sammry results appeared in early 1984. A national demographic survey was carried out in 1980/81, but the data have not yet been processed, and only a few results from hand tabulation are available. The only fully processed, analyzed and published set of demographic data comes from the 1980/81 Poplab (University of North Carolina) demographic survey of three regions, Banadir, Bay and lower Shebelle. Data deficiencies are particularly severe for the nomad population. Data on fertility and mortality were collected in the 1975 census but only for the sedentary population. Hence, although sufficient data do exist for a preliminary assesament of the population situation, reliable estimates of fertility, mortality and natural increase must await full results from the 1980/81 survey and the results of the forthcoming census planned for mid-1985. Population Size and Growth 2.02 The 1975 census enumerated a total population of 3.25 million, but because of problems with the census fieldwork and the difficulty of cover- ing the nomadic population, the Government estimates that an undercount of just over 20% occurred (Annez 1, Table 1.1). The total population in 1975 is estimated to have been 4.1 million, with 46Z of the population nomads, 29% rural settled and 25 urban. Because no other enumerations have been held, no direct information on population growth can be obtained from the census and the rate of population growth must be estimated indirectly from data on mortality, fertility, natural increase and migration. According to  3 Bank estinates, the population reached 5.4 millionl as of mid 1985 with a current average annual growth rate of 3.0x. Population Distribution and Urbanization 2.03 Somalia is among the most sparsely settled countries in Africa, with a national density of 5-6 persons per km2 in 1975 and an estimated 8 today. Full regional and urban figures from the 1975 census have not yet been published, but there are no heavily settled areas outside the coastal Mogadishu agglomeration and a few other large towns, primarily in the north- west. Most of the rural settled agricultural population are found in the southern riverine regions of Middle and Lower Juba, Middle and Lower Shebelle, and Bay, where they make up 48% to 64% of the population. In the central and northeastern regions, by contrast, nomads form a substantial majority (622-80Z) of the population; and in the northwest highlands make up about half. Of the estimated 1.0 million urban dvellers2 in 1975, over 40% lived in Mogadishu, with an estimated population of about 450,000, or 8% of the national population. Another 19% of urban dwellers lived in the towns of the northwestern highlands, principally Hargeisa and Burao. International and Internal Minration 2.04 International migration, both forced and spontaneous, is important in Somalia. Following drought and hostilities with Ethiopia, large streams of refugees arrived in Somalia between 1978 and 1981, with numbers peaking at 1.2 million in 1980/81. Most were displaced nomad households, who settled in semi-permanent refugee camps. Since 1981 many of the adult men have re- turned to Ethiopia, leaving women and children in the camps; total numbers remaining are in dispute but now amount to not more than 700,000, and probably closer to half a million, or about 1OZ of the national population. In early 1984 the Somali government announced a policy of acceptance of refugees as permanent residents of Somalia and resettlement of them as farmers on agri- cultural land. Economic emigration, mainly to the Gulf countries, was substantial before 1980 but has been substantially reduced since then. 2.05 The importance of nomadism in Somalia makes it difficult to define and measure internal migration. However the summary figures on lifetime emigration so far available from the 1975 census suggest that most Somalis, particularly in the heavily nomad regions, do not permanently move out of their region of birth, with typically less than 10% of regional residents born outside the region. The exceptions are some of the southernmost regions, where the development of comiercial agriculture has been important, and Mogadishu, half of whose population vas born elsewhere and which has been lNot including refugees, who are also excluded from both the enumerated and adjusted census figures. 2Urban areas are defined as all district capitals regardless of size (some with less than 100 inhabitants), plus any localities with at least 1,500 inhabitants.  4 groving very rapidly in recent years. Within regions, rural to urban migration is known to be important, with towns such as Bargeisa also growing very rap- idly, but exact figures are not available on the size of such flows. Mortality 2.06 Fully available data on mortality in Somalia are at present limited to information on child survival for the settled population (urban and rural) from the 1975 census, and information on both child and parental survival for the settled population of Banadir, Bay and Shebelle regions (rural. Mogadishu and other urban) from the 1980/81 Poplab survey. The census infor- mation proved too poor in quality to use, but the 1980/81 Poplab results were accepted as the basis for estimating national mortality levels in the urban and rural settled sectors of the population3. Life expectancy at birth in the late 1970s is estimated at 42 years (41.1 for males, 42.3 for females) in the rural settled population, and at 48 years (47.0 for males, 48.8 for females) in the urban population. The corresponding rural infant mortality rate (IMR) was 167, with 292 of children dying within their first five years of life, while the urban DIR was 140, with 23Z of children dying before their fifth birthday. Life expectancy in 1985 is estimated at 45 years in the rural settled population and 51 years in the urban population. There are no reliable data to estimate nomad mortality, but information from other countries with substantial nomad populations, such as Mauritania and Mali, suggests that mortality is unlikely to be higher among nomads than among the agricultural population and could well be lower. On the assumption that nomad mortality is similar to rural settled mortality in Somalia, national life expectancy in 1985 is estimated at 47 years (45.7 for males and 47.6 for females), with a national INR of around 145 and about 252 of children dying within their first five years of life. These levels of mortality are at the high end of the range for eastern Africa. For instance, the circa 1982 DKRs for Kenya, Uganda and Ethiopia are estimated to have been 80, 120 and 140, respectively. Fertility 2.07 As in the case of mortality, the only fully available data on fer- tility come from the 1975 census and 1980/81 Poplab survey, and consist of information from women on past and recent childbearing. Again, the 1975 census data proved to be too poor to use, and the Poplab survey results were accepted as applicable to the rural agricultural and urban sectors as a vwole. The total fertility rate (TFR) in 1980 is estimated at 7.1 for the rural settled population and 7.4 for the urban population (7.2 in Mogadishu and 7.5 in other urban areas). The level of fertility among Somali nomads is unknown at present. However, fertility is low among nomad populations elsewhere in Africa both in relation to their settled neighbors and to African fertility levels in general. On the assumption of a nomad TFR of 6.0 in Somalia, rhe national TFR around 1980 can be estimated as 6.7. This is at the low end of the range for eastern Africa, where levels of 7 or more are comon. For 3Standard World Bank Development Report methodology was applied.  5 eza-ple, the circa 1980 TFRs for Kenya, Uganda and Ethiopia are estimated to have been 8.0, 7.0, and 6.8, respectively. 2.08 The determinants of fertility in Somalia have not been studied, but the moderately high overall level is no doubt the product of universal marriage for women, a mean age at first marriage of about 20, and virtual absence of either traditional or modern child-spacing practices. Higher urban than rural fertility is uncommon in Africa and elsewhere, but here it may be the effect of better health services not yet counteracted by any widespread adoption of modern birth control among urban and educated women. Nomad fertility may be reduced by long periods of spouse separation and heavy physical work loads for women. Ase and Sex Distribution 2.09 The 1975 census and 1980 national demographic survey both reported an extremely unbalanced sex composition, with a male surplus of 114 males per 100 females in 1975 and 109 in 1980. The surplus was found in all sectors in the census (109 in the urban and rural settled sector, 123 among the nomads) but chiefly in the nomad population in the survey (104 for urban dwellers, 99 for the rural settled population and 116 for nomads). Given a normal sex ratio at birth (i.e. in the range 100-107) such a female deficit could occur only through severe excess mortality of females, substantial net ir- migration of males, or substantial undercounting of females relative to males. The latter factor is probably the most importAnt in this case, particularly among nomads. By contrast, the sex ratios reported in the Poplab 1980/81 survey are more reasonable, being 91 for the rural agricultural population, 101 for Mogadishu, and 91 for other urban areas. Further clarification of this point vill have to await another census. In the absence of better data, it is assumed in the Bank projections reported in section B below that the national sex ratio in 1975 was 100.4, the value selected for the adjusted population by the census report, with the sectoral ratios for the urban and rural agricultural populations taken from the Poplab results, and a nomad sex ratio of 106.5 derived as a residual. Similar problems are encountered with data on age distribution. The national ap distribution for the projections was chosen using a model life table consistent with estimated national levels of fertility and mortality (see Annex 1). Vital Rates and Natural Increase 2.10 Application of the estimated levels of mortality and fertility to the chosen age distribution yields a 1985 national crude birth rate (CBR) of 49, a crude death rate (CDR) of 19, and hence a current natural increase of 3Z. Natural increase is highest in the urban population (4.1%), lowest among the nomads (2.2%) and just above national level (3.0Z) in the rural agricultural population. At a growth rate of 3.0%, the population of Somalia vould double in 24 years.  6 B. Populat ion Proiections 2.11 National population projections for the period 1985-2015 were con- structed based on three alternative assumptions about trends in fertility--unchanging, gradual decline and accelerated decline. The national projections were constructed from a summation of separate projections for the urban, rural agricultural and nomad populations. The separate projections were not adjusted for migration between sectors and are, therefore, conservative because overall fertility is expected to increase as population moves from the relatively lower fertility nomad sector to the higher fertility agrLcultural and urban sectors. In all projections an annual net outflow (to the Gulf) of 10,000 was assumed to have occured between 1975 and 1980 but not thereafter. Also, in all projections mortality is assumed to decline according to a standard path used in Bank World Development Reports (WDR) projections; life expectancy rises from 48 years in 1985-90 to 59 years in 2010-15.4 2.12 In the first of the three projections, it is assumed that there is no change in fertility in any of the sectors during the projection period. (This vill in fact entail a gradual rise in fertility at national level, as the population shifts, through differential natural increase and/or migration out of the lower fertility nomad sector into the higher fertility urban and agricultural sectors.) In the second projection, fertility is assumed to decline gradually beginning from 1990 in the urban sector and from 1995 in the agricultural and nomad sectors. The national TYR is projected to decline from 6.7 in 1980-85 to 4.7 by 2010-15, with replacement level fertility not projected to be achieved until 2050. This decline follows normal Bank methodology for WDR projections and represents a possible spontaneous evolution of fertility with continuing socio-economic development but without any major government intervention either in hindering private contraceptive use or in promoting a public program. The third projection assumes an early and rapid fertility decline beginning after 1990 and resulting in a TFR of 2.4 by 2015. This third projection represents what could be achieved if a successful national family planning program were built up over the next four to five years. In order to achieve the fertility levels of the third projection, the contraceptive user rate would have to be 20Z by 1994. 2.13 The projected size and growth of the population of Somalia under each of the three fertility assumptions is sum-arized in Table 1. The tremen- dous potential for population growth in Somalia is clear. Without major 4Details of the underlying projection model are discussed in E:wan, Vu and Zachariab, "Short-term Population Projections, 1980-2000 and Long-(erm Projections, 2000 to Stationary Stage for All Countries of the World. PHNPR, World Bank, August 1981. The refugee population is not included in the projections. If it were to be included, it would add about 500,000 to the estinate of the 1985 population. Details of the mortality and fertility assumptions are given in Annex 1, Table 1.2.  7 government intervention, the population will increase by around 60Z to 8.5-8.7 million within the next 15 years and will multiply to at least two and a half tines its present size, to reach 13-15 million, in 30 years' time. More- over, massive capacity for future growth vill remain at the end of the projec- tion period, with grovth rates permitting a further doubling of population within another 19 to 27 years. A fertility decline on the scale likely to occur spontaneously vill have a limited impact, reducing the 2015 population size by 13Z. Serious government action vill be required to reduce population increase to any really significant degree. An accelerated fertility decline would reduce the 2015 population size by 32X. Table lr PROJECTED SIZE AMND GROWTH OF NATIONAL POPULATIONa/ Ratio of Population Year Size to 1985 Base in: 1985 2000 2015 2000 2015 Total Population (millions) No fertility decline 5.40 8.67 14.81 1.6 2.7 Gradual fertility decline 5.40 8.51 12.89 1.6 2.4 Accelerated fertility decline 5.40 7.91 10.12 1.5 1.9 Doubling Time at 1980-85 1995-2000 2010-15 2010-15 Grovth Rates AveraRe Annual Rate of Population Growth (x) No fertility decline 2.91 3.30 3.69 19 years Gradual fertility decline 2.91 3.01 2.59 27 years Accelerated fertility decline 2.91 2.14 1.38 50 years a/ See Annex 1, Table 1.2-1.4 for details of the projections.  8 2.14 Trends in fertility vill also have a great impact on population structure, as is shown in Table 2. The 1984 dependency ratio (the ratio of population less than 15 years and over 65 years to the working age popu- lation 15-64 ) is 92. For comparison the 1984 dependency ratio is estimated to be 102 in Uganda and 86 in Ethiopia. As mortality falls (with gains con- centrated in childhood) the population of Somalia will become ever more youth- ful if fertility does not also decline, and the dependency ratio will rise from its present level of 92 dependents per 100 adult workers to 102 in 30 years' time. However, under the gradual fertility decline the proportion of children in the population vill fall and the dependency burden will be reduced to 80 dependents per working age adult by 2e15. With an accelerated fertility decline the dependency burden would be 53 in 2015. Table 2: PROJECTED POPULATION AGE STEUCTUREal 1985 2000 2015 Percentage of Pouulation Ated under 15 Years No fertility decline 45 46 48 Gradual fertility decline 45 45 41 Accelerated fertility decline 45 41 31 Dependency Ratio No fertility decline 92 96 102 Gradual fertility decline 92 92 80 Accelerated fertility decline 92 79 53 a/ See Annex 1, Table 1.4 for detailed projections. C. Conseauences of PoDulation Growth 2.15 Vithout any major government population program, Somalia must plan for a minimum population of at least 13 million within 30 years' time, vell over double its present size. It is therefore vital to assess the consequences of such rapid growth for the economy of Somalia and the welfare of its citi- zens. The key question is not how many people the natural resources of Somalia could ultimately support. Somalia may ultimately be able to support a popu- lation two or three times its present size. Rovever, the critical question is whether Somalia can develop its natural resources fast enough to outstrip population -growth and to support two or three times its current numbers, at improved standards of living, within the next 20 to 30 yoars. Implications for Nomad Agriculture 2.16 A central issue involves the redistribution of population that will be necessary given current population growth and the balance of resources between sectors. The livestock sector is the principal economic base of Somalia at present,-providing occupation for half the population as pastoral  9 nomads in 1975, and producing half of (DP and over 80% of export receipts. Its future is, however, critically dependent on the -reaaining potential of the rangeland. In the absence of a comprehensive rangeland survey there is much dispute over its current status and absorptive capacity. However, it is generally agreed that little, if any, rangeland reserves remain, and that the scope for, and costs of, improvements in the productivity of rangeland already in use are uncertain. Consequently, increases in national herd size will have to be severely restricted for the foreseeable future and, if the returns to nomad households in income and subsistence are to be maintained, the growth of the nomad population must be reduced. Yet despite the uncertain- ties over exact levels of nomad mortality and fertility, the rate of natural increase among the nomads is already high and will continue to be so over the next 20 to 30 years. 2.17 Without strong government population policies and programs, the nomad population will increase by almost 50X over the next 15 years and more than double over the next 30 years; another 1 to 1.1 million people will be added by 2000 and a total of 2.5 to 3 million by 2015. Similar increases in herd size would therefore be required merely to maintain current subsistence and income levels among nomads; the rangeland would have to support another 15 million animals within the next 15 years and a total of 37 to 45 million extra animals in 30 years time. A gradual fertility decline of the scale expc_ted to occur without major government intervention could make little difference to the magnitude of the population increase, cutting the 2015 nomad population size by only 102 . Even an accelerated decline would reduce the 2015 nomad population by no more than 26Z. 2.18 There may be some scope for improving herd productivity. The possi- bilities for herd improvement are being examined in ongoing Bank livestock sector analysis. But even with plausible productivity improvement the range- land may not be able to accommodate herd increases on the scale required by projected nomad population growth. Indeed, the 1981 Bank Agricultural Sector Review (ASK) recommended that the numbers of people and animals on thw-a range should be stabilized for the foreseeable future. To accomplish this, up to a million nomads would have to be absorbed into the agricultural or urban sectors within the next 15 years, and up to 2 or 3 million within 30 years from now. The 1981 ASR estimates that given a Goverment policy of voluntary settlement with public sector support (infrastructure development and provision of basic services), a maximum of 750,000 hectares could be brought 'inder rain-fed cultivation by the year 2000, supporting about 825,000 people. The capital costs to the government of this operation would amount to US$600 million in 1984 prices. In addition, controlled irrigation could be rehabilitated or expanded over 110,000 hectares to accommodate an additional 344,000 people in irrigated agriculture, at a total capital cost of about US$900 million. Thus, the agricultural sector could at most absorb about 1.2 million additional population by 2000, at a total capital cost (which must be largely borne by the government) of US$1.5 billion. 2.19 However, natural increase in the agricultural sector must also be taken into account. If there is no strong government intervention to control population growth, the agricultural population will increase more than 60% over the next 15 years and will multiply to two and a half to three times its present size within 30 years from now. Another 920,000 to 950,000 people will be added by 2000 and a total 2.3 to 2.8 million over the next  10 30 years. On the assumptions of the agricultural expansion and intensification outlined above, this natural increase of the agricultural population could easily be accommodated at least up to 2000, but surplus absorptive capacity would only amount to 250,000-280,000 persons. Out of the expected 1 to 1.1 million nomad natural increase, therefore, only about one quarter could be absorbed by settlement into the agricultural sector, .eaving about three quarters of a million still to be accommodated elsewhere. The situation becomes even worse if the possible settlement by government of half a million or more refugees is taken into account; the surplus then would amount to over 1.25 million people by the year 2000. 2.20 Movement into the urban sector is another possible option for the surplus nomad population. Again, however, the absorptive capacity of the urban sector is limited, and must in addition accommodate urban natural in- crease, which will be high because of relatively low mortality, still elevated fertility, and a favorable age structure with vomen in the childbearing ages forming a high proportion of the total. The urban population will grow through natural increase alone by around 70% over the next 15 years, and will reach 2.7 to 3.3 times its current size 30 years from now. Over one million people will be added by 2000, and between 2.5 and 3.5 million by 2015. If the surplus 0.75 to 1.25 million from the nomad sector were to move into the urban areas as well, the urban population by the year 2000 would more than double to between 3.3 and 3.9 million, or at least 40% of the total population. The inclusion of refugees in the nomad sector would raise these numbers still further. Given the lack of industrial and urban development in Somalia, it is difficult to imagine how such a rapid increase in the urban population could be accommodated at reasonable levels of income and living standards for urban dwellers. Urban growth on this scale would demand heavy public expenditure on housing and basic services. 2.21 The fisheries sub-sector remains as one other possible option for the surplus nomad population. With present fish production only about 5% of its potential and employing only about 2% of the total population, there is enough absorptive capacity in fishing for a surplus population of up to 2 million. However, there have been difficulties in resettling nomads in fishing communities in the past (see the 1981 ASR, Annex 4, pp. 17 and 18) and this is not considered likely to be a popular option. The necessary changes in attitudes and culture may be slower than the time frame of 20 to 30 years imposed by the rapidity of nomad population growth. Moreover, heavy capital investment in infrastructure and services would again be required, amounting to at least US$930 million (1984 prices) by the year 2000 if fertil- ity declines and US$1.5 billion if there is no change in fertility. 2.22 The difficulty of accommodating the growth that will occur in the nomad population in the absence of a fertility decline can be illustrated by a comparison of the cost of resettlement with the potential increment in annual GDP. If the overall ecoromic growth rate of 3.5% to 4% per annum projected by the Bank for the period 1982-19865 could be maintained up to the year 2000, the average annual gain in GDP of US$34-40 million (1984 Prices) S World Bank, Policy Measures for Rehabilitation and Growth, May 1983.  3U would be more than absorbed by the annual costs of the nomad resettlement, which would amount to approximately US$90 million annually over the next 15 years6. Implications for Social Infrastructure 2.23 The government of Somalia places a high priority on improvement of educational levels, including the attainment of universal primary education (UPE), and has made considerable efforts in education over the last decade. But rapid population growth has diluted the efforts and limited the educational coverage achieved. From 1975 to 1980, for example, primary school enrollments increased by 24Z; but because the population of eligible children grew by 15% during the same period, only a modest improvement in the enrollment ratio from 26% to 28Z was possible. Rapid population growth will continue to hamper progress in the future, as can be seen from Annex 1, Table 1.7, showing pro- jected growth in the primary school-age population, and Table 3 below illus- trating the implications for the recurrent cost of primary schooling. If there are no government population policies or programs, the number of children eligible for primary schooling will increase by 64% over the next 15 years and will swell to between two and three times its present level within 30 Table 3: REQUIRED GROWTH IN ANNUAL EDUCATIONAL EXPENDITURES (1984 So. sh. 000,000) At Constant Enrollment Universal Primary Educationc/ Ratioa/ by 2010 Average Average Annual Annual Growth Growth 1985 2000 2015 RateZb/ 1985 2000 2015 RateZbl No fertility decline 129 212 374 3.5 129 468 1126 7.2 Gradual fertility decline 129 210 301 2.8 129 464 906 6.5 Accelerated fertility decline 129 195 187 1.2 129 431 562 4.9 a/ Assuming: US$28 per primary student and a 0.58 enrollment ratio b/ 1985 to 2015 c/ Assumes current attrition rates between grades. 6 Assuming one quarter of the resettlement is in agriculture and three quarters in the urban sector.  12 years from now. Primary educational expenditures would have to be correspondingly increased at an average annual rate of 2.8Z to 3.5%, to reach So. sh. 210 million (US$12 million) in 2000 and So. sh. -300 to 370 million (US$17 to 21 million) in 2015, just to maintain the current enrollment ratio and quality of schooling. To achieve UPE within the next 30 years, funding would have to increase by 6.5% to 7.2% per annum to a level of So. sh. 900 to 1100 million (US$50 to 65 million) or 8 to 10 tismes the present level in real terms. Such rapid expansion of the education system does not appear possible if quality is even to be maintained, let alone improved to a desirable level. 2.24 A substantial expansion of primary schooling in both coverage and quality would be considerably eased by an early and rapid fertility de- cline. With an accelerated fertility decline, educational expenditures over the period 1985 to 2015 would need to increase by only 1.2% per annum to maintain the current enrollment ratio. An annual growth of 5% would allow UPE to be achieved by 2010 with an increase of real primary schooling expendi- tures four times the present level. 2.25 Rapid population growth will also have a substantial impact on the demand for health services, which, like education services, are dispropor- tionately utilized by the youthful and the dependent sections of the popula- tion. If fertility does not decline, the MCI target population and annual numbers of deliveries will grow by 60% over the next 15 years, and will almost triple within 30 years from now (Annex 1, Table 1.8). The burden on NCR services will increase proportionately, forcing government to devote a consid- erably higher percentage of total expenditure than at present to health care if coverage and quality of services are to be improved over the current inad- equate levels. Even a gradual fertility decline could reduce the needed increase in expenditures significantly, cutting the size of the NCR target population by 17% within 30 years, aud annual deliveries by 38%. However, the target population and the demand for obstetric services would still more than double by 2015. In contrast, with an earlier and more rapid fertility decline, the 1CR population would grov by only 70% and the demand for obstetric services would fall by 14%, thus permitting faster extension and improvement of MCH services with an unchanged share of total government spending. 2.26 The Bank Energy Sector Report7 notes that the most difficult long term challenge in the energy sector is to meet future woodfuel demand while preserving the existing forest reserves. Groving population and livestock numbers have combined with drought to produce increasing deforestation aud range degration. Part of the solution to this problem lies in development of alternative sources of energy and actions to change the technology of fuelwood consupmtion and production as set out in the Action Based Scenario in the Energy Sector Report. However, low rainfall, soil salination, and constraints on-alternative energy sources limit the gain from actions to 7The discussion in this section is based on reanalysis of data in World Bank, SOMALIA: Issues and Options in the Energy Sector, August, 1985 (Report No. 5796-SO).  13 change energy consumption and production. Reduction of population growth is an essential complement to these energy policies. 2.27 The implications of population growth for fuelvood consumption and deforestation are demonstrated in Table 4, using the consumption and production assumptions outlined in the Energy Sector Report and population projections summarized above. Four deforestation projections are given: (1) no fertility decline, no change in energy policy, (2) no fertility decline. Action Based Table 4: DMPLICATIONS OF POPULATION GROUT! FOR FUELWOOD CONSUMPTION AND DEFORESTUTION8/ Consumption Production Net Re-aining (000 im3) (000 m3) Deficit Forest area (000 m3) (000 hectares) I-No Fertility Decline No Energy Actions (Year) 1985 4045 3503 542 9730 2000 6719 3029 3690 8221 2015 11678 1443 10235 3299 II-No fertility Decline Action Energy Policy (Year) 1985 3947 3493 454 9730 2000 4678 4015 663 9214 2015 5792 3854 1938 8286 III-Rapid Fertility Decline No Energ Action (Year) 1985 4045 3503 542 9730 2000 6065 3029 3036 8380 2015 7980 2006 5974 5078 IV-Rapid Fertility Decline Action Energy Policy (Year) 1985 3947 3493 454 9730 2000 4223 4015 208 9332 2015 3958 4191 -234 9327 a/ Derived from projections reported in SoMalia: Trends and Outions in the Enerity Sector, World Bank, August 1985. -  14 energy policy, (3) accelerated fertility deline, no change in energy policy, (4) accelerated fertility decline, Action Based energy policy. In the first projection, without a fertility decline or change in energy policy, consumption of fuelwood would grow from 4 million cubic meters in 1985 to 6.7 million by 2000 and 11.7 million by 2015. The resu.ting deficit of fuelwood consumption over projected production would grow to 10.2 million cubic meters or 87 percent of consumption by 2015 and would result in extensive deforestation. By 2015 Somalia's forests would be reduced to only 34 percent of the 1985 forested area. In the second projection, with a change in energy policies to follow the conservation methods and forest management outlined in the energy report action sceuario, but without a fertility decline, consumption would increase to only 5.X million cubic meters by 2015 and only 15 percent of 1985 forest reserves wtould be lost. Rowever, the energy policies would only postpone an inevitable fuelvood crisis arising from population growth. From 2000 to 2015, the wood fuel deficit would increase by almost threefold to constitute more thar one third of consumption. 2.28 With an accelerated fertility decline but without a change in energy policy, consumption would double from 1985 to 2015, the fuelwood deficit would increase to 8 million cubic meters by 2015. The remaining forested area would be over 50 percent greater than it would in the absence of a change ;n fertility. However, the extensive loss of forested area, attributable to population growth in the years before fertility reduction could take effect, would be severe. The combined effects of improved population and energy policies to bring about a solution to the functional crisis are dramatic. With the accelerated fertility decline and the action energy scenario, consumption in the year 2015 would have returned to its 1985 level and less than 4 percent of the 1985 forested area would be lost. Importantly, after about 2010 the woodfuel deficit would be eliminated and forest reserves would begin to be rebuilt. D. Population and Family Planning Policies 2.29 The Government has not adopted a national population policy and family. planning (FP) is not promoted as a specific measure to reduce the rate of population growth. A few goverzment officials are aware that the present high rate of population growth is a constraint on development efforts and has adverse implications for the country's economic and food situation. The Deve:opment Strategy and Public Investment Programme, 1984-1986, states that population growth is high in relation to the nation's capability to provide social service and potential for food production. However, Somalia has only recently emerged from a pronatalist policy, and many officials below the top levels are uninformed about population issues, remain uninterested in population policy or believe that accommodation of excess population will be made possible by development of irrigated agricultural land. The prevailing socio-cultural situation does not present an easy environment for the formu- lation of practical family planning policy. Widespread acceptance of family planning, particularly in rural areas and among nomads, is hampered by a fatalistic attitude towards birth and death and a perception (incorrect accord- ing to religious authorities) that FP is contrary to the teachings of Islam.  15 In addition, the high infant and child mortality rates tend to reinforce the desire for large families. 2.30 Nevertheless, there are a number of positive factors that are leading to an improved environment for population policy in the future. First, there is support for FP linked to NCR activities from key officials, especially in the MOP, MOB and MOE, and over the last three to four years child spacing has been promoted as part of the MOB maternal and child health program. Second, in the last decade increased literacy coupled with growing participa- tion of women in socio-political activities and growing political support for women's rights has improved the climate for FP promotion. 2.31 In 1982 the Government created a new Division of Family Health/Family Planning under the Department of Comnunity Health with the responsibility of planning and coordinating all family health and FP activities. The Family Health Division immediately prepared a Plan of Action to provide family plan- uing services throughout the country, making the existing NCR centers the main service channels. The Plan of Action proposed to start family planning in five major urban areas, Banadir (Mogadishu), Bay, Northwest, Togdheer and Lover Juba. It was considered that because of higher rates of literacy in these regions, FP would be accepted more readily. To date FP services have been introduced in 31 MCH centers with financial and technical assistance primarily from the United States (USAID). The NOR has prepared a long-term plan to expand family planning services-into 95 MCI centers in the next five years. This rate of expansion appears to be slow given the changing climate for services and with sufficient donor support thirty to forty centers per year could be added to the program. Ultimately, given the present density of only one MCI center per 50,000 population, the number of MCH centers must be increased if services are to be increased substantially. 2.32 Progress with innovative family life programs is especially encour- aging. With the assistance of UNESCO, the MOE is introducing population topics into school curricula using a coordinated series of eight modules for primary and secondary school. The modules are currently reaching about 50X of enrolled children comprising 1OX of the school age population. Addition- ally, in response to the Government concern to ensure women's participation in the country's economic -development effort, the Women's Education Service (WES) under the MOE was established in 1974. The WES has since established Family Life Education Centers in all 82 districts. There is now a cadre of 600 family life personnel working as teachers, headmistresses, supervisors, trainers, and administrators. The eight- month curriculum includes nutrition, child care, sewing, handicrafts, and hygiene. Family planning is proposed to be included gradually. 2.33 In late 1983 the National Parliament established the Somalia Family Health Care Association (SFHCA), a non-governmental organization affiliated with the International Planned Parenthood Federation (IPPF). The Secretary General of the SFECA is also the Director of the Family Health Division in the MOR and the policies of the two agencies are closely coordinated. When it is fully functioning the SFRCA is intended to protect maternal and child health through creation of centers and provision of social and health services including family planning, to prepare research studies on the health and  16 socio-economic conditions of the family, and to provide information, education and comhunication (IEC) services. At present, the SFHCA has a staff of less than ten and will be able to undertake only limited activities, primarily in the Mogadishu area. IPPF initially provided financial support of US$40,000 for organizational establishment. The support bas been increased to US$80,000 for 1984. As a non-governmental organization, the association should be able to play an expanded role in the creation of demand and social legitimiza- tion of contraceptive practices and in program innovation. 2.34 The Somali Women's Democratic Organization (SWDO) was started in 1971 as the women's section of the political office of the Presidency and was later established as an autonomous organization in 1977. SUDO operates on the basis of voluntary membership and has a well organized hierarchical structure resching dona to the village level. The main organizational objec- tive is to achieve 'betterment of the social life of Somali women, their liberation from social and economic inequality, the safeguarding of their basic rights and the encouragement of their full participation in the national construction." Presently the SWDO executes a number of projects in the rel- evant areas of vomen's development supported by OXFAM, Economic Commission of Africa (ECA) and UNICEF. SWDO actively sponsored, in collaboration with the NON, the formation of the SFHCA. SWDO provides a potentially effective organizational framework for the execution of projects, designed to enhance women's role in the family and society, and can directly support family plan- ning through IEC activities. The SWDO is more firmly based and has broader support than is often the case with similar organizations elsewhere and should play a central role in any family healtb strategy. E. Family Plannina Services 2.35 Provision of family planning services only started in late 1982, and it is too early to evaluate the performance of the program. Clinical service statistics are collected by the NOR units where services are offered. As reported by the Family Health Division of the MOB, at the end of 1983 MOB services accounted for a total of about 2400 users, of whom about 40Z used pills, 15Z IUDs, and the remainder condoms or jelly. In the largest urban centers, however, contraceptives may be purchased privately, and there are a modest number of users, primarily among families vith higher income and education levels, not accounted for by the NOR data. A contraceptive knowledge, attitudes and practice survey (KAP) was carried out in the five largest urban areas in the spring of 1984 by the MOR and MNP with the assist- ance of Westinghouse and funded by USAID. Preliminary tabulation of the survey reveals that less than 1I of eligible couples in urban areas use either modern or traditional contraceptive metbods. It is presumed that the rate in the rural and nomadic population is negligible, giving an average rate for the entire population of much less than 1Z. The survey also reveals that, among the five urban areas surveyed, from 95% to 98S of married women at risk want more children and 46S to 95Z want "all that God wills" (from phrasing of survey question). In contrast to the low percent using contra- ception, the percent vho approve of family planning socially ranges from 22Z to 56Z with the approximate average at about 45%. While the program implications of the survey results are unclear because of the young age of the program, it is notable that the areas with the most advanced FP programs  17 have the highest user rates and also the greatest knowledge and social acceptance of FP. 2.36 NOR capacity to offer services is hindered by the limited coverage of the rural and nomadic population with health services, poor logistical support, and a lack of trained manpower. Attention to these constraints is being given in donor funded projects, but the difficulties of logistics and greater rural coverage are especially great given Sonalia's scattered settlement pattern and mobile population, and improvement can only be expected to come slovly. More than in many countries, the difficult cultural environment and the problems of service distribution call for an innovative strategy that capitalizes on non-governmental channels, for example the WES, SFPCA and SWDO, and, perhaps, programs such as community-based distribution of contraceptives, the retraining of traditional birth attendants and promotion of contraceptive retail sales. III. HEALTH A. Health Status 3.01 Until recently no health surveillance system existed, and reliable morbidity and health statistics are lacking. Various government agencies and international organizations have made estimates of mortality on the basis of scattered evidence, and a morbidity survey conducted jointly by WHO and the MOB during 1980-83 provides some evidence on the pattern of disease. Despite variation in estimates across agencies (see Annex 2. Table 2.1), the evidence that exists is uniform in establishing that the health status in Somalia is among the worst in the world. Mortality 3.02 The National Realth Plan for 1980 quoted a Crude Death Rate (CDR) at that date of 21 per 1000 population with a life expectancy at birth of around 43 years and an infant mortality rate (INR) of 150 to 177 per 1000 live births. The WHO Count:y Health Programme quoted a CDR, based on data from the last half of the 1970s decade, in the range of 22-24 and estimated IM=s of 190 to 280 for rural areas and 140 to 170 for urban areas. The Assignment Report on National Diarrheal Disease Control (1983 WE0 study) estimated the mortality rate per thousand children 1 to 4 years-of age at 30 for the early 19809, and the mortality of children under 5 years was estimated at 51% of all deaths, with about 25Z of all children dying before their fifth birthday. Demographic estimates and projections made by the Bank mission, and consistent with the various sources quoted above, yield an estimated current CDR of 21 per 1000 population, an INR of 150, and a mortality rate of 25% for children less than five. 3.03 Breast-feeding is universal and prolonged which provides some amount of natural contraceptive protection, but the birth interval of around two to three years is nevertheless short. Both marriage and child bearing begin early, even in urban areas, with half of women married by the age of 20 years. Short birth intervals and early child bearing together with infibulation  18 (see para. 3.08) and poorly developed MCH services adversely influence the maternal mortality risk, which has been estimated by WHO at 1100 per 100,000 live births (about 1980) compared with 82 in Egypt (1978), 210 in Thailand (1971), 570 in Bangladesh (Matlab, 1970), 103 in Ecuador (1978) and 10 in the United States (1978). 3.04 According to the National Health Plan (1980), the leading causes of mortality are diarrhea, communicable childhood diseases, tuberculosis, malnutrition, obstetrical complications, cardiovascular diseases and hyper- tension, malignancies, and accidents. A study of causes of death among infants and children less than five in Banadir and surrounding areas (Annex 2, Table 2.2 and 2.3) found that 20Z of deaths were due to neonatal tetanus, 22% to diarrhea, 14% to measles, and 72 to respiratory diseases. Among the remaining 37Z, tuberculosis and birth complications were important. The study did not report on malnutrition, but given the high mortality rates due to measles and diarrhea, malnutrition vas almost certainly an important associated cause. The pattern of mortality shows a preponderance of deaths caused by comnunicable and birth associated diseases and suggests that a substantial impact on health status can be achieved by basic NCR activities, increased birth intervals and environmental control. Morbidity 3.05 Commonly prevalent diseases among children include measles, vhooping cough, tetanus, tuberculosis, malaria and intestinal disorders (diarrbea, dysentery, and helminthiasis), and nutritional deficiency diseases. Over 75% of childhood morbidity is due to preventable causes. A 1978/79 survey found that 32n of rural and 54Z of urban children had contracted measles by the age of 3 years. The poor nutritional status of Somali children reduces their resistance to complications from measles and severely increases their mortality risk. A survey conducted in 1982 in 13 districts found a high rate of lameness (11.3 per 1000) caused by poliomyelitis among children of 5 to 14 years of age (Annez 2, Table 2.3). The most common age of onset was 1 to 4 years, with 82X of the illnesses having occurrec in the first three years of life. A survey conducted in rural areas near Mogadishu reported 25% of children below 5 years having at least one episode of diarrhea each month, averaging about three episodes per child per year. Banadir hospital data shov that 60 to 70% of children attending outpatient clinics suffer from diarrhea. 3.06 Gast-ointestinal diseases, hypertension, schistosomiasis, anemia, malaria and tuberculosis are leading causes of adult morbidity. The morbidity survey (summarized in Annex 2, Table 2.4) conducted during 1980-83 with WHO assistance confirmed the relative importance of these diseases but found significant regional differences in the prevalence of some diseases. Tuber- culosis, as measured by a test procedure that counted cases of advanced pulmonary disease but overlooked early stages and non-pulmonary forms, was more prevalent in Middle Shebelli (8.5 per 1000 population), Lower Juba (5.7) and Middle Juba (6). The overall prevalence rate estimated by the morbidity survey was 2.2 per 1000. Falciparum Nalaria was common in Gedo, Middle Juba, Hiran, Togdheer and West Gal, while less virulent quartan malaria was more common in Sanaag, Bay and Middle Shebelli. Prevalence of ascariasis  19 and trichuriasis among tbe nomad population were 1.5 and 2.4 times the average for Somalia. The wide regional differences in disease patterns carry important implications for the design of regional programs. 3.07 Schistosomiasis. Schistosomiasis (Eaematobium) shows striking regional variation. The regions of maximum prevalence identified in the National Morbidity Survey were Lower Juba (196.0 per 1000), Lower Shebelli (182.0 per 1000), Middle Juba (91.0 per 1000) and Middle Shebe'li (88.0 per 1000). Prevalence of schistosomiasis was either nil or very low in northern regions. The national health plan estimates that schistosomiasis infects 53% of the agricultural population along the courses of the two major soutbern rivers. Schistosomiasis was emphasized by NOR officials as an important local problem in the Juba and Shebelli areas. The disease is fostered by still bodies of water and past experience demonstrates that the incidence increases with irrigation schemes. Irrigation projects should incorporate measures to monitor and control the disease. 3.08 Infibulation. Female circumcision is almost universally practiced in Somalia. There are three types of procedures: (a) Mild sunna is the mildest form and consists of excision of the prepuce of the clitoris; (b) Modified sunna is partial or total excision of the body of the clitoris; (c) Infibulation is the most mutilating form consisting of partial or total excision of the clitoris, excision of labia minora, excision of inner walls of the labia majora and suturing, thereby virtually closing the vaginal opening, leaving only a small passage for urine and menstrual blood. In Somalia, circumcision is commonly performed on girls between the ages of 5 and 8 years, and in about 85Z of the cases the procedure is infibulat ion. Immediate complications include pain, shock, bleeding, infection, injury to the urinary tract and anus, and tetanus. Delayed complications include disfiguration of genitalia, painful scar, pain and difficulty in dexual intercourse, lack of orgasm, depressed libido, cyst, abscess, and urinary infection. Lifelong complications include sterility, prolonged and difficult labor, painful menstruation, depression, and psychosis. 3.09 The- practice of female circumcision presents a serious health and social problem and deserves more attention from the MOE, related ministries and donors. There is already growing awareness about its far-reaching social, mental and health implications. The SWDO is organizing a campaign against the practice. Under the auspices of SVDO, a National Committee has been formed with representatives of various ministries, youth organizations and doctors, with the objective of developing a coordinated strategy. The exper- ience of the neighboring countries should be revieved to identify the most suitable culture-specific policy measures.  20 B. Healtb Policies and Strateties 3.10 Until 1980, little planning of health activities had been carried out in Somalia. Earlier development plans emphasized the directly productive sectors and only minimal resources were allocated to health sector development. Health projects were designed and implemented disjointedly as funds were sporadically available and the health system was largely oriented to the urban populations of the largest towns. In 1980, a major effort was made to identify the most important health problems. set priorities and formulate health policies. Although serious weaknesses remain in the planning process, and resource flows have not been sufficient to support the plan (para. 3.54- 3.59), the initial planning effort has been valuable in providing a new focus for health development programs. Plans for the health sector are set forth in the National Health Plan (1980-85) and, briefly, in the Five- Year Development Plan (1982-86) and Public Investment Profram-e (1984-86). 3.11 National Health Plan. The NHP ,as written after a methodical identification of health priorities, using seven weighted factors to determine priority health problems and develop the focus of the health strategy. The factors assessed for each health problem were the availability of inter- vention technology, morbidity and mortality rates, the effect on the economy, political i-uplications and public concern, relationship to other diseases, and epidemic potential. The health problems identified as the focus of the strategy were maternal and childhood diseases (including diarrhea and malnutrition), tuberculosis, malaria, schistosomiasis, sexually transmitted diseases, respiratory infections and accidents. 3.12 The plan that was developed recognized the prevailing inaccessibility of health services to 85 to 90X of the rural population and the past neglect of preventive services and emphasized the extension of health coverage through primary health care (PBC). The objective is to make basic curative and preventive activities available to the rural population by integrating health with other development activities at the community level and strzngthening medical care at the regional and district levels as a backup to PRC. The plan places a particular emphasis ou improvement of health management and administration. As detailed in the plan, this entails improved planning, organization, coordination, supervision, control and evaluation. Thb strategy is to carry out management studies, reorganize at the central level, recruit middle-level staff for central and district positions, train staff i manage- went, and develop manuals specifying operating procedures. As a complement to management activities, the plan calls for development of a health information system. 3.13 Overall, the NEP represents an important step to reorient the health services to meet the needs of rural areas and provide a new emphasis on preventive pro3rams to attack the naaor health problems. owever, the NIP was carried ouF as an isolated activity and deficiencies in the regular planning process are hampering the effective implementation of the plan. The deficiencies concern coordination among sectcrs, coordination among donors, reconciliation of pl.nned projects with absorptive capacity and  21 available recurrent funding. The Social Service Sector of the Ministry of Planning which deals vith health presently does not -have any health planning capacity. Health planning expertise in the NOR is also limited. 3.14 Five-Year Development Plan and Public Investment Protram. The FYDP, covering the years 1982-86, broadly outlines twenty health projects consistent with the NHP in terms of priority health problems. The projects include PHC and small scale health facilities for selected regions, health manpower-training (including a national teaching hospital and nursing school at Kismayo), maternal and child health (NCR) and family health programs, and disease specific projects for the control of childhood diseases, tuber- culosis, diarrhea, malaria, schistosomiasis, sexually transmitted diseases (STD) and malnutrition. Priorities have not been established within the list of projects but The Development Strategy and Public Investment Programme (PIP) selects a subset of the program outlined in the FYDP for emphasis during the years 1984-86. The strategy is to continue to emphasize PHC at the village, district and regional levels, to expand the immunization program, train manpower at all levels and to improve logistics and drug supplies. 3.15 Within the PIP, the Primary Health Care (PH01)7, Family Health laitiative (PH05), Nutrition Programme (NSIO), TB Control (HS13) projects and technical assistance projects for Immunization (HS02), -Maternal and Child Health (PH08) and Health Manpower Development (HT07) are all consistent with this strategy. Of the remaining projects in the PIP, Hospital Rehabli- tation (HC05) and Hospital Training (HTO7) strengthen existing referral services and are consistent with additional needs identified below (para. 5.23). However, The Teaching Hospital Project (ECOI), which comprises slightly less than 20Z of the PIP for health (Annex 3, Table 3.7), is inconsistent with the strategy of shifting health investment from urban based curative care towards rural PHC. Additionally, Jilib Hospital (EC03), although rural based, is of lesser priority when evaluated in the light of the PHC strategy. 3.16 The health planning process is intended to be a part of the total national development planning process coordinated by the Ministry of National Planning (MMP), but in practice coordination has not been effective. The planning cycles of the health plan and national development plans do not correspond. Many health projects are developed through bilateral discussions between the MOR and the individual donor agencies withcut the MNP playing a specifically outlined role in project development. As a result, a number of vertical projects have arisen that require separate supervision at different tiers of the health care system. As the projects have been implemented, issues and problems have been identified that demonstrate the need for an smalgamation of the vertical programs. The emerging health policy in this regard has been to gradually incorporate many proiects within PHC, the latter being treated as an umbrella for the otbers. 7Numbers in parentheses in this paragraph are the project numbers in Somalia, Ministry of Planning, Development Strateav and Public Investment Programme 1984-1986, Revised, December 1983.  22 C. Ornanization and Management 3.17 Modern health care in Somalia is almost entirely a public sector service. Private practice, which Was not permitted until mid 1983, is growing rapidly but is primarily confined to the major urban areas. Reorientation of the government services from an urban centered, curative care bias toward primary health care and preventive services has only commenced recently. In late 1984, an organizational framevork to accommodate the new focus of the health services vas adopted (see Figure One at end of text), but the framework has not yet been fully realized and the actual organization of the MOR remains open to change with out undue disruption. Hence, there are opportunities to strengthen MOEH organization, as positions are filled and operational procedures worked out. 3.18 National Level. The MOR is responsible for the development of health policy and for planning and delivery of health services throughout the country. Administration of the MOR is carried out under the Primary Secretary with resposibilities divided under three Director Generals for, respectively, Preventive Services, Curative Services, *and General Administration. As outlined in the organizational chart shown in Figure 1, there are three Departments under the Director General for Preventive Services: Community Health, Public Health, and Laboratories. There are three Departments under the Director General of Curative Services: Hospital and Medical Care, Medical Personnel, and Drugs, Supplies and Equipment. There are four Departments under the Director General of Administration: Finance, Personnel, Planning and Training, and Transportation. Each Department is headed by a Director and is subdivided into sections. Functional categories of health services are located within the Directorate of Curative Services and within the Directorate of Preventive Services under the vertical programs sections within the Departments of Community Health and Public Health. Primary Realth Care (PEC), Diarrhoeal Disease Control, Family Health Care (including Family Planning and NCR), the Expanded Program of Imunization (EPI), Nutrition and School Health are located in the Department of Community Health, whereas Leprosy, TB, Schistosomiasis and Malaria control are located under the Department of Public Health. There is one Medical Officer for each of the vertical health projects working-under the overall supervision of the respective department director. The Ministry has a Central Coordinating Committee which deals with a wide range of health issues and comprises departmental directors and senior officials. 3.19 In concept the present organizational structure of the MOR is compatible with the eventual establishment of an efficient management and administrative system along functional lines. However, in practice the structure on paper is not performing as designed. There is no well developed system of administrative record keeping and filing. Lines of authority and responsibilities are not clearly defined and do not flow as intended. Major problems arise from the large number of unfilled positions and a lack of procedural guidelines. For example, vithin the MOB Planning Section (one of the three sections under the Department of Planning and Training) there is currently no professional staff member trained in health planning.  23 3.20 The lack of a procedure for project development and continuing planning bas already been noted above (para. 3.13). Similar problems affect the coordination of planning, budgeting and accounting; the monitoring and surveillance of curative services and logistical support; and the gathering of health information needed to monitor the complex systems under the Departments of Community Health and Public Health under the Directorate of Preventive Services. Lack of trained managers is a pervading problem that contributes to the failure to fill positions and establish procedures. Many of the top administrators are medical doctors well qualified in their respective disciplines, but lacking appropriate training and background in management and planning. 3.21 ReRional Level and District Level. MOE programs are administered through regional and district level offices corresponding to the eight administrative regions, 48 districts and 60 municipalities. At the regional level there is a Regional Medical Officer (RHO) in overall charge of health services. The RH0s report to the Directorate of Curative Services. In addition, with the inception of the PEC program, a Regional Primary Health Care Coordinator has been employed in the regions covered by such programs. The Regional PHC Coordinator reports through the RHO to the MOH, as well as directly to the Director of Community Health (who is also the Project Manager for PEC) on PEC matters. 3.22 The health organizational struzture below the regional level is 2ssentially that of the PHC program and exists only in regions covered by PHC (see para. 3.34). There are eight regional PHC teams, each under the Regional PEC coordinator. Each regional team includes a public health nurse, midwife, sanitarian, laboratory technician and pharmacist. At the district level, the District Medical Officer is also the head of the PEC team, which is similar to the regional team in composition, but in practice at the district level the positions of sanitarian, laboratory technician or pharmacist are often vacant. In each district, there are four PHC units, each with one public health nurse, one sanitarian and one midwife. Finally at village or community level there are health posts, at the rate of -four per PHC unit. Each health post has one Traditional Birth Attendant (TBA) and one Community Health Worker (CHW), who provide the first line of contact within the primary healLn care system. Operational aspects of PEC are discussed in paragraphs 3.31 to 3.35. 3.23 Health Information System. The health information system in Somalia is in a very early stage of development. Process data (for example, numbers of inpatients and outpatients by category) are not regularly reported and summarized. Health data for planning and management purposes have generally been drawn from a few scattered health surveys and studies of varying qual- ity. In many instances, disease prevalence data were derived from clinical impressions. Currently a WHO adviser is helping to design a reporting format and training procedures to start a continuing epidemiological surveillance system. The project is now underway in four regions of the country. Irregular distribution and collection of forms, -oor record keeping habits in the field, and a shortage of expert personnel to train field functionaries in the reporting zystem are constraints in the development of an effective  24 health information system. As the proposed system is heavily dependent on reports from institutional health facilities, its value in establishing prevalence rates for planning purposes is limited although its value for program monitoring and administration should be great. Community-level surveys should be carried out periodically to supplement the facility based information. D. Facilities and Services 3.24 Health services are provided through a two-tier system. The first tier was developed during the early 1960s and places emphasis on curative services delivered through a framework of urban regional hospitals and district- hospitals supplemented by services available through outpatient clinics and dispensaries. The second tier (see paragraph 3.30) is the primary health care program started in 1979 in four regions with USAID assistance. The intention is that the older curative based system and the PHC system will be unified as PEC is introduced over a 20-year period in each of the 16 regions. At present the two systems have separate management, supervision and service delivery structures. In concept, the first tier is intended to provide the referral back up as PHC is adopted in each region and the MOB plans to unify the two tiers eventually. In actuality access to the referral system is limited and 10X to 15% of the population is covered by modern health care services. 3.25 The Referral System. The mission assessment of facilities and services is based on MOE records, discussions with government and donor officials and mission field trips in the North to Burao and Eargeisa, in the South to Kismayc and in the center to the environs of Mogadishu and Afgoi. There is some inconsistency in the number of hospital beds reported by alternative sources. Official figures were in excess of actual counts made during mission field trips. The discrepancy in the number of beds may be at least partly attributable- to the fact that facilities are given an 8 shilling per bed day allowance based on the number of beds in service. In the absence of effective monitoring and supervision, the allowance acts as an incentive to exaggerate the number of beds. 3.26 Based on MOE records there are 16 regional hospitals, 62 district and specialized hospitals, 218 outpatient clinics, 92 maternal and child health clinics and 9 I-ray clinics (Annex 2, Table 2.5). The 16 regional hospitals have 100 or more beds each as well as outpatient services and provide a referral point for three to five districts. The 52 district hospitals are designed to have 20-50 beds as well as outpatient services, but many of the district hospitals have only a few of their beds actually in service; 50% of district hospitcls axe staffed only by nurses while another 11% are staffed by medical assistants. The specialized hospitals include five TB hospitals, three mental hospitals, one leprosy hospital and one pediatric and gynecological hospital. On the average there is one hospital bed for every 1,154 people in the country. However, this does not reflect the availability of services because of the overcount of official beds in service and the uneven distribution of hospital beds. Without adjusting for overcount, the Nogadishu area with about 10% of the total population of the country has 37Z of the hospital beds, giving one bed for only 317 people, as compared to one bed per 4,079 people in Lover Shebelli, 3,879 people in Galgadud, 3,659 in Nudug, and  25 3,520 in Middle Shebelli. The outpatient clinics and the NCR clinics are relatively more evenly distributed between the regions. 3.27 The quality of services in hospitals varies widely among regioni. Generally, the organization and quality of health facilities in the north is better than those in the south. The mission field trips revealed videly varying rates of utilization of health facilities. For example, nearly all of the 300 beds of Hargeisa general hospital and about 82X of the 180- bed Burao hospital in the north were found occupied. Compared to that, only 252 of beds in Kismayo regional hospital, only one out of 35 beds in Jilib district hospital and none of the 32 beds in Afgoi district hospital were occupied. The low bed occupancy rate is related to the limited avail- ability of drugs, lack of doctors, and poor quality of services. In many facilities, running water, toilets or latrines, cleanliness, mattresses, and sheets were seriously inadequate, particularly in the South. Maintenance of building and equipment was seriously lacking. There is an urgent need to improve the quality of services, particularly at the district level, in order to increase their effectiveness and utilization. Measures needed are additional medical staff, especially doctors, and provision of required drugs, equipment, runming water, toilet facilities and attention to cleanliness and routine maintenance. 3.28 In contrast to the poor state of affairs in the general hospitals, the tuberculosis hospitals and outpatient centers have high utilization rates and provide good quality services, including a full range of treatments delivered on an outpatient basis. Nearly all the 180 beds of Hargeisa TB hospital in the North and 60% of the 70-bed Kismayo TB hospital in the South were found occupied by patients. The better functioning of the TB hospitals and outpatient centers may be attributable to the fact that the TB program has substantial expatriate financing and technical assistance and is vertically monitored. 3.29 Maternal Child Health Program. The present NCR project began in 1977 with financial support from UNFPA and technical assistance from WHO. UNICEF has provided further support in equipment, supplies, drugs and transport facilities. There are currently 92 NCR clinics or one per 50,000 population. The centers, primarily located in urban and peri-urban areas, are specified to have a midwife, two nurses and two vaccinators, but generally have unfilled positions. The MCE clinics vary in service utilization, but typically the NCR clinics are much better utilized than referral facilities. Information on coverage and attendance is not regularly collected. The attendance of women and children at the eight MCH centers visited by the Bank mission averaged between 250 to 300 per week or 14,000 per year. Assuming three visits per year per person and extrapolating to all 93 NCR centers, the total number of people covered is roughly 430,000 (-93*15,000/3) or less than 20% of the MCR population. Services offered include immunization, antenatal clinics, nutritional counseling, growth charts, nutritional supplements, oral rehydration, family planning, and referral, but not all of these services are carried out consistently across centers, depending on staff training and availability of supplies and equipment. In particular, family planning has only recently been introduced to a limited number of centers (see para. 2.28) and oral rehydration, while  26 available at the nCH centers is not widely accepted as a home therapy. NCR centers where supplementary feeding programs and an inmunization program are underway are more effective in reaching the target population. The apparent difference in utilization and coverage between centers with and without supplementary feeding and expanded immunization programs (EPI) suggests the importance of incorporating supplementary programs vith high public perception of benefits as the NCR program is expanded. 3.30 In most NCR centers an adequate supply of drugs is available. The clinics operate only on an outpatient basis without facilities for deli- very. The total number of hospital beds and obstetric capacity of district hospitals is small and the prospect for increasing the proportion of institu- tional deliveries in the country is limited. The NCR project should therefore continue to employ a preventive aud promotive antenatal strategy through training and utilization of trained birth attendants (TEAs). The proposed incorporation of MCR activities into the PEC program will also provide the scope for expanding the effectiveness of the promotive MCR program. 3.31 Primary Health Care. The basic service delivery uuit of the PEC system is the primary health care post of which there is to be one per 3,000 population at the village level. Each post is staffed by one community health care worker (CHR) and one traditional birth attendant (TBA). Both workers are paid a salary. Services provided by the PlC post are primarily promotive and preventive and include home visits, IEC activities and referral to support disease specific vertical programs and basic NCR programs. Activi- ties include antenatal care, nutritional programs, oral rehydration, immuni- zation, sanitation, hygienic education, record keeping (birth, deaths, epidemic diseases), and first aid. The next higher level is the primary health care unit (PuCu) serving an average of four primary health posts and covering a population of about 12,000. The staff includes one public health nurse, one nurse midwife, and one sanitarian. The PHCU is the immediate source of technical and logistics support to the health posts. The third level in the system is the district health center staffed by two senior nurses and two midwives. The fourth and highest level of the PHC system is the regional health center located in the regional capital and staffed by a physician, a senior public health nurse administrator, a senior sanitarian and administrative support personnel. 3.32 Strengths of the program design are that at the lovest level it relies on paid workers, rather than volunteers, and the activities of the CEW and TBA are confined to a relatively short list of effective interventions. Care should be taken to preserve this latter strength because in practice additional duties have been assigned to the CIWs, including some vertical program activities. A further strength is that the program is only being implemented where there is substantial donor technical and financial assis- tance. Weaknesses of the program are that there has been a lack of coordination of separate donor programs and in a few instances training has been inconsis- tent, for example, with respect to family planning-included in the TSAID funded program but not in the UNICEF program. The large number of staff to be trained, approximately 350 service personnel per region of 375,000 people for the two lowest levels of service delivery, is a constraint, as is the potential recurrent cost of the PHC staff for the program when it  27 is fully functioning. Both the training burden and the recurrent cost problem will be surmountable only with continued donor technical and financial assis- tance over the foreseeable future. An additional weakness of the program is the poor quality of referral services in the backup tier of health services. An associated weakness is the lack of integration of the two tiered health system at the district and regional levels. 3.33 Current coverage of nomads with health services is negligible and reaching the nomad population with PRC services presents a particular challenge in Somalia. Health projects among nomads have particularly high risks of failure. Problems of logistics, programs supervision and monitoring are particularly great. Innovative program designs are required, borroved perhaps from other countries and sectors. Generally, mobile healtb care units serving nomad populations have substantially higher unit costs than static facilities or outreach services serving settled rural populations. Nevertheless, by concentrating on a core group of high return services, including immunization, oral rehydration, antenatal counseling, and family planning, cost effective delivery of services might be achieved. Experiments need to be carried out with innovative delivery strategies. As one possibility, the Government is creating (or identifying) satellite communities to provide a basis for implementation of agricultural, educational and health development programs in nomad regions. If, as planned, women and children remain relatively fixed to small plots of land while men move with cattle, these communities will provide the basis for intermittent contact with health programs. Another possibility is the use of mobile training units to identify and train members of the nomad community in promotion and delivery of basic services. The health workers could also be supervised by mobile units. Use of trained birth attendants, who move vith the community, would be an important element of this strategy. Finally, use of multipurpose community extension workers to carry out intersectoral programs may be more cost effective than separate programs. Again mobile supervision and a simple design, with a limited number of messages or services to be delivered in each sector, would be required. 3.34 Presently PHC projects at different stages of development are in operation in seven regions. USAID supports PHC in two regions (Togdheer and Bay), UNICEF/VHO supports three regions (Northwest, Middle Shebelli and Lower Shebelli); the Australian Government supports one region (Sannag) and the Swedish Church Relief supports one region (Middle Juba). USAID is scheduled to support PHC in another two regions over the next two years. Projects in Togdheer and Bay, initiated earlier, are generally at a relatively advanced stage of development. 3.35 The design of the PEC program is appropriate for extending health care to the unserved rural population, but there are some indications that the program is being expanded too rapidly without careful evaluation of implementation problems. CEWs and TBAs are not making wide enough contacts with the target population and supervision of primary health care posts is not done at satisfactory intervals. Lack of transportation was mentioned by officials as a constraint to supervision. Problems with inadequate facili- ties and insufficient numbers of trained tutors are also being encountered with the CHW and TBA training programs. Considering the ongoing implementation  28 problems and lack of previous experience, the short-term strategy for PHC should be that of consolidation by making the existing regional projects effective through intensified supervision, inspection and training, before further extension to other regions is undertaken. In addition, further expansion of PHC should be paced to accompany strengthening of the referral system. 3.36 Expanded ProRram on Immunization (EPI) The EPI program was started in 1978, first in the Banadir region and then folloved by another four regions in 1979. By 1983, nine regions vere covered by EPI; and in 1984, all the regions are proposed to be covered. In regions where PFC programs are under operation, EPI is included with PEC. The project aims at immunizing children against diptheria, whooping cough, tetanus, polio, measles and tuberculosis. Immunizations are given either in the MCR clinics, PHC clinics or through mobile teams. The central EPI unit in the MOE coordinates all activities. In the seven NCB clinics visited by the mission there were arrangements for immunization and the cold-storage facilities were functioning. It is difficult to make an accurate estimate of the extent of coverage, but for the Mogadishu area coverage is estimated to be 40Z. The coverage in rural areas is substantially lower. The official NHP target is to increase coverage to 80% by 1987 in the regions covered by PHC, and to 70Z by 1987 in non-PEC regions. Rowever, UNICEF officials estimate the current coverage within PHC regions to be below 10% and expect a coverage of only 30% by 1987. Inadequate supervision and logistics support are major constraint to expanded program coverage.. Weighted by the proportion of population in Nogadishu (40% coverage), PHC regions (10 coverage) and non-PHC regions (2% coverage), the national level of coverage is estimated to be 11Z. Nevertheless, considering the high mortality caused by preventable childhood diseases in Somalia, the EPI project, even with only 30Z coverage by 1987, should have a significant impact on childhood mortality and morbidity over the near future. 3.37 Other Vertic&.l Services. In addition to the above services, there are other separate project activities designed to control specific disease conditions or attain specific health objectives. Among the service-related projects are schistosomiasis control (in four regions), leprosy control (in six regions), TB control (in six regions), sexually transmitted disease control (in three regions), malaria control (in all regions), prevention of blindness (in three regions), diarrheal disease control and nutrition programs through PHC. All of these projects are operated as vertical programs, in many cases with heavy expatriate involvement in operations and separately org.nized logistics systems. The NOR intends that all the above project activities, except the schistosomiasis control, leprosy control and sexually trar,smitted disease control, vill be integrated with PHC in the near future. It is important that adequate administrative and logistics capacity be developed ti ensure the success of integration (see para 3.38). E. Health Support Systems Logistics and Pharmaceuticals 3.38 The operating efficiency of district bospitals and other health facilities at the periphery is seriously impeded by a lack of maintenance  29 support from the central ministry and an inadequate flow of supplies and equipment. Government officials cited problems with transportation due to a shortage of vehicles, spare parts and fuel. Donor support of the capital cost of transport has been sufficient in the past, and the current problems with transportation are due to inadequate maintenance and management of the transportation fleet. Problems with logistics are also attributable to a poorl) 'rganized and managed inventory and distribution system. 3.39 Acute problems with the central logistics system have led the sponsors of major vertical programs to provide independent logistics support. USAID maintains a motor pool that provides transportation of contraceptives and other MCH supplies to the regions in which it supports PEC. UNICEF similarly provides logistics support for UNICEF supported projects. -In the short-run direct participation of donor agencies will continue to be crucial to provide an assured flow of supplies to rural projects. Ultimately, however, it will be essential to develop a centralized and unified logistics capacity serving the entire country. A UNICEF project has been proposed to proviue technical assistance to identify- for the total health programme, the transportation and storage requirements at the central, regional and district levels, to identify distributional constraints, and to develop a comprehensive logistics system. The UNICEF project should be followed by coordinated donor support for the technical and material assistance to build the government logistics and maintenance capacity. 3.40 The official drug agency (ASPIMA) is responsible for procurement and supply of drugs. Essentially all drugs are imported either by ASPIMA or directly by donor supported projects. On the regional and district levels as well as national level, the supplies for the vertical programs, for example EPI, tuberculosis control and PHC projects, are generally satisfactory. Outside the donor supported projects, the supply and availability of drugs procured by ASPIIA is severely limited by the shortage of foreign exchange and the low level of NOE recurrent cost support. Encouragement of a larger private sector role in importation and distribution of drugs could provide greater availability and coverage. Private and Traditional Medicine 3.41 Traditional medicine remains the major source of health care in rural areas and is used by over 80Z of the population. Traditional birth attendants (TBAs) coumonly provide maternity services. The heavy use of traditional practitioners is partially due to the limited accessibility of modern health services, but is also due to the strength of cultural tra- ditions and the perceived effectiveness of traditional medicine. The NOR has established a division of indigenous medicine under the department of drugs and medical supplies. The MOR should consider the potential for capital- izing on the strength of traditional medicine, possibly through retraining and IEC activities to develop TBAs as a vehicle to carry selected aspects of basic FP and MCH care to nomads and the settled rural population. 3.42 At the beginning of 1984 the Government legalized the private practice of medicine by physicians. There is no data yet on the percentage of the country's 350 physicians that have actually started private practice,  30 but the opinion of NOR officials is that the number in Mogadishu and other urban areas is large and many Government doctors interviewed during field trips participated in private practice on a limited scale. Private clinics have been started in the larger urban areas and newly started private labo- ratories have the capacity to carry out basic blood and urine analyses. There is no information on fees or hospital admission privileges. It is too early to assess the ultimate effect of the legalization of private prac- tice. On the positive side the removal of the government monopoly on health services may increase the availability and quality of care. On the negative side allowing MOB physicians to engage in private practice may divert physician time from government provided care, create a conflict of interest, and reduce the accessibility of care to lower income groups. It may also aggravate the maldistribution of physicians. F. Health Personnel and Trainins Current Health Personnel 3.43 The main categories of health personnel in 1983 are summarized in Table 4 . The ratios of population per health worker are less favorable than for moderate income East African countries but are, nevertheless, better than in low income neighboring countries. For example the population per nurse in Somalia is 1,600 compared to 2,300 in Uganda, 28,000 in Ethiopia and an average of 9,700 for countries classified as low income in the World Development Report (WDR). The population per doctnr is 15,600 in Somalia compared to 25,000 in Uganda, 77,000 in Ethiopia and 15,900 for other WDR low income countries. Table 4: PRINCIPAL CATEGORIES AND NUMBER OF HEALTH PERSONNEL FOR 1983/ Population per Cate2orv Number Health Worker Doctor 326 15566 Registered Nurse 1534 3308 Auxiliary Nurse 1632 3109 Midwife 158 32114 Sanitarian 303 16748 a/ WHO, "Health Manpower Planning for the Somali Democratic Republic," Basu Ghosh, (January 1984), WHO/EMIHMD/474. 3.44 The uneven distribution of health personnel across the country (Annex 2, Table 2.6) is the major health manpower problem. Mogadishu (Banadir area) with 10% of the country's population has 74Z of the physicians, 39% of the registered nurses, 45% of the auxiliary nurses, 68Z of the sanitarians, 762 of the pharmacy assistants, 732 of the laboratory assistants, and overall,  31 50% of all medical and paramedical personnel in the country. This unfavorable distribution of health manpower with high concentration in the Mogadishu area appears to have been worsening in the last fev years, with the proportion of total medical and paramedical personnel in the Mogadishu area increasing from 43% in 1981 to 502 in 1983. During the same period, concentration of doctors in the Mogadishu area increased from 71% to 741, auxiliary nurses increased from 27X to 45Z, sanitarians from 42% to 68%, pharmacy assistants from 671 to 76X and laboratory assistants from 671 to 731, auxiliary nurses increased from 271 to 45%, sanitariane from 421 to 68%, pharmacy assistants from 671 to 76% and laboratory assistants from 671 to 73%. Registered nurses are the only health manpower category whose distribution improved during the above period, with the proportion concentrated in the Mogadishu area declining from 48% in 1981 to 391 in 1983. 3.45 Maldistribution of manpover is a major constraint to strengthening the referral network to complement the PHC system. Many district hospitals have no doctor and are severely understaffed in other medical categories. A careful examination of incentives, quality of housing, and other factors leading to this situation is needed. The examination should be folloved by an estimate of the cost of bringing about a more efficient distribution of personnel. Training Programs 3.46 The institutions that produce health manpower in Somalia are the faculty of medicine under the National University of Somalia, two Nursing schools, the Health Personnel Training Institute and two Rural Health Training Centres. Except for the Faculty of Medicine, the training institutions are under the administrative control of the Department of Training in the Ministry of Health. 3.47 Physicians. The faculty of Medicine in the National University of Somalia is under the Ministry of Higher Education and conducts undergraduate medical education and, in collaboration with an Italian university, confers a medical degree equivalent to the MD. Medical students are admitted after the 12th grade of basic educaticn and after one year of a primary school teaching assignment and one year of military service. The total duration of the medical curriculum is five years. The present annual intake is about 75-80, and about 30 to 35 doctors graduate every year. The present rate of output is sufficient for the short run but over the next five years the attrition rate should be reduced and the number of graduates increased to 65-75 per year to meet the long-term needs of the health system. The language used in teaching is Italian, while the language used in primary school is Somali and the language used in secondary school is English. This has caused some difficulties for the students and may be a factor contributing to the high dropout rate. A recently revised curriculum devotes a reasonable amount of time to preventive health care. The faculty of medicine conducts post- graduate degree courses in the specialities represented by the five Departments of the Faculty-community health, medicine, surgery, MCR and pediatrics, pathology. Short duration courses providing selected retraining in adminis- trative skills for doctors currently involved in management are needed.  32 3.48 Nurses. The nursing schools are in Mogadishu and Hargeisa. Another nursing school, recently completed in Kismayo, is expected to begin functioning by end of 1984. However, the hospital in Kismayo, because of the poor state of the facilities and low standard of care, does not currently provide a suitable environment for nursing residencies and the Bank mission recomends that the opening of the Kismayo training unit should be delayed. Nurse administrators, tutors and midvives are given advanced training at the Post Basic Institute in Mogadishu. The duration of training for a regis- tered nurse is three years. Primary education up to eighth grade with aptitude for science subjects is required for entry into the program. The total capacity of the two existing nursing schools is 750 with a current annual output of about 575. The capacity may be overestimated if quality is taken into account. A recent evaluation (WHO) notes that the tutor/student ratio, currently about 30, is too high and that hostel facilities are overcrowded. The study recommended that intake of nursing schools be reduced to a level consistent with the facilities, that clinical field training and hostel facilities be increased, and that faculty development be carried out. The curricula for registered nurses does not cover managerial subjects and, in the past, has emphasized curative care rather than preventive services. Managerial and administrative subjects should be introduced. Nursing curricula were revised in late 1983 to add a community medicine and MCH/PP content consistent witb PHC. It is too early to evaluate the success of the changed curriculum. 3.49 Other Health Service Personnel. The Realth Personnel Training Institute trains sanitarians, laboratory technicians, pharmacy assistants and midwives. The annual output, summarized in Annex 1, Table 2.7, has been satisfactory for the current scale of services. As for the other personnel categories, the major problem is one of distribution of personnel rather than numbers trained. Training of nurses, laboratory technicians and pharmacy assistants equalled or exceeded the National Health Plan target for the four years of the plan period. 3.50 Training for the Primary Egalth Care Protram. Two rural health training centers have been established with USAID support as part of the primary health care program, one in Burao in the North and the other in Baidoa in the South. These institutes provide in-service training of four months duration for nurses, sanitarians and midwives making up the PHC project teams. The institutes are intended to provide unified PEC training for all of the regional programs, including those funded by other donors. Construc- tion of the training centers was marred by operational, logistical, and managerial problems. The curriculum and program design are adequate but the substantial problems with hardware as well as problems in execution of the training program on schedule have remained serious deficiencies that are hurting the quality and quantity of training. Up to mid 1984, 65 nurses, 56 sanitarians and 58 midwives had been trained. Consolidation of PEC training was a desirable step towards greater coordination of MOB programs across regions and donors, but because of consolidation output from the training program affects a number of different projects and it has become especially important that the program implementation is of bigh quality.  33 3.51 Community Health Workers (CEW) and Traditional Birth Attendants (TBAs) are trained at the village level by the nurses and sanitariuns who have received training at the institutes. The courses for CIWs and TBAs are of four to six months duration, including intensive course work for two months, folloved by four months of field work for on-the-job training, and then followed by an additional one to two months course. Up to mid 1984, 311 CEWs and 246 TBAs had been trained. Manpower Planningt 3.52 Conceptually, a significant part of the current shortage of staff at the district level could be relieved through a better distribution of personnel, but in practice in many countries as well as in Somalia redistrib- ution of staff already in urban institutions has not proven feasible. As the referral system is strengthened over the future to provide the appropriate back up for PlC, additional personnel requirements will develop and redistrib- ution alone will not be sufficient. There is currently no trained manpower planner at the MOE. Manpower planning is needed to forecast personnel requir- ements and plan for the changing needs arising from the reorientation of the health services towards rural basic health services. G. Health Expenditures Total Expenditures and Sources of Funds for Health Care 3.53 Estiaates of total expenditures on health care in 1982 by source of funds are presented in Table 5. Total health care expenditures (public plus private and recurrent plus capital) were So. sh. 431 (US$31) million or 2.3% of gross domestic product (GDP). Private capital expenditures are not available but are thought to have been negligible. Annual in kind or monetary private expenditures on traditional and modern care have been estimated at So. oh. 128 aillion and So. sh. 91 respectively, or 1.5Z of estimated total private consumption expenditures. The estimates of private consumption expenditures are based on household expenditure surveys for Mogadishu and the Sudan and are thus only approximate as applied to Somalia as a whole. Of the total sources of funds, the Government provided 262, donors provided 242, and the private sector was estimated to provide the remaining 51S. Including public and private expenditures, the total per capita expenditure of So. sh. 90 (US$6.3) is lower than for most other East African coratries where estimates have been made (in 1982 prices: Zambia, US$38; Zimbabwe, US$38; Lesotho, US$12; Malawi, US$10; Uganda, US$9.5; Rvanda, US$7). This intercountry comparison uses the official exchange rate of 13.9 So. sh.JUS$ and may thereby overestimate the Somali expenditure.  34 Tabl, 5: TOTAL EXPENDITURE AND SOURCES Of REALTH CARE FUNDS8/ 1982 (000,000 So. oh.) Source of Funds Service grovider Government Private Sector Total Government 109.9 109.9 Foreign Donorsb/ 102.1 102.1 Private SectorC/ a. Modern 90.7 90.7 b. Traditional 128.1 128.1 Total 212.0 218.8 430.8 &/Including recurrent and capital expenditures by the Government and foreign donors. There is no basis for estimating capital expenditures by the private sector* b/Foreign donor funds includes both direct donor expenditures and funds disbursed under government agencies. Estimates are based on UNDP records. Donor reporting to UNDP is incomplete and some donor expenditures may be omitted. c/In the absence of a comprehensive household survey, private sector expend- itures are roughly estimated. Modern sector expenditures are calculated as 0.8X of urban private consumption expenditures plus 0.21 of rural private consumption. Expenditure percentages are based on an analysis of the 1977 Mogadishu household survey by income class. Private consumption is derived from World Bank estimated national accounts. Based on the distribution of population between sectors, urban private consumption is estimated to be 27Z of total private consumption. Traditional expenditures are estimated to be 0.75% of rural private consumption based on estimates for Rwanda, Zambia and Zizbabwe. Because modern health service coverage in Somalia is lover than for the three countries cited, traditional expenditures may be underestimated. The estimated total of private sector expenditures equals 1.5Z of private consumption expenditures. Trends in Ministry of Bealtb Expenditures 3.54 In nominal terms (Annex 3, Table 3.1) annual Ministry of Health (MOB) total expenditures have increased from So. oh. 41.3 million in 1975 to So. *b. 109.9 million in 1982, but the nominal increase has hidden a large decline in the real value of health ezpenditures. Deflated by a price index for gross domestic product and measured in 1982 shillings, total MNE expend- itures in real terms (Annex 3, Table 3.2) bave fallen by 29Z, from So. uh. 154 (US$11.1) million in 1975 to So. oh. 110 (Us$7.9) million in 1982. The extent of the decline is emphasized when the effect of population growtb is included (Annex 3, Table 3.3); from 1975 to 1982 real ezpenditures per capita fell by 411, from So. oh. 37 (US$2.5) to So. sh. 22 (US$1.5).  35 3.55 The trend in KOB expenditures has not paralleled the path of total government expenditures. As a percent of GDP (Annex 3, Table 3.4), total government expenditures increased from 15.3% in 1975 to 23.9% in 1982. In contrast, MOB expenditures as a percentage of total government expenditures have declined from 6.1% in 1975 to 2.5% in 1982. The combined effect of the increase in government expenditures and declining proportion going to health has been to reduce NOR expenditures as a percentage of GDP from O.9Z in 1975 to 0.6Z in 1982. 3.56 Examination of the Government recurrent expenditures by functional category (Annex 3, Table 3.5) shows that the social and economic sectors have received a declining share of total government expenditures while defense and finance have received an increasing share. A part of expenditures unier finance goes to support general government salaries, but the largest pjart of this category is used for supplementary support of defense. The total share received by the social sectors in 1975 was 40%, while defense received 26% and finance received 12Z. By 1984 the share received by the social and economic sectors had fallen to 19Z while that for defense had risen to 32% and finance to 37Z. Among the social sectors the health expenditure share has fallen by the greatest amount, by over 70%, since 1975. The falling share of total recurrent expenditures going to health clearly reflects a shift in emphasis from basic social service needs to defense. 3.57 Development Prosram Expenditures and Donor Financins. The implemen- tation rate for health projects under the previous five-year plan (1977-81) was low; out of a planned capital expenditure of So. sh. 101 million, 14 million was spent by December 1981 giving an actual expenditure rate of less than 14%. The pace of capital expenditures in health has increased considerably since 1981. It is too early to assess the implementation rate for the 1982-86 PIP plan, but total donor expenditures in the health sector in 1982 were So. sh. 102 million (US$7.3 million, Annex 3, Table 3.6). Of the total donor expenditures, 43% came from multilateral sources, primarily UN agencies (USS3.0 million), and 57% came from bilateral sources, primarily USA (US$3.0 million) and Italy (US$0.8 million). Expected 1984 investment in PIP health projects totals So. sh. 268 million. Of this total only So. sh. 11 million or 4Z, is to come from domestic sources, the remaining So. sh. 257 million coming from donors. 3.58 WHO plays the major coordinating role in the health sector among donors and directly represents a uumber of donors, both multilateral and bilateral, in project supervision. A total of 32 projects were being admin- istered by WHO in 1934. In addition to the country representative, the WHO mission has professional staff dealing with the areas of health eervices, epidemiology and statistics, training and manpover. WHO have had some diffi- culty in providing accounting and administrative support and the capacity of the resident WHO mission is heavily taxed by the ongoing technical assistance that it it providing. 'UNICEF, and USAID are the major donors supporting health services. Their activities are concentrated in the area of PHC with a strong emphasis on NCR activites. UNICEF's organization of logistics and supervision for rural health services is particularly strong and their experience in these areas deserves further examination as a source of guidance  36 for other sector programs. USAID is promoting NCH/FP through the Family Health lait:itives (FBI) Project, and is funding the bulk of ongoing population activities either directly through the FRI project or indirectly through intermediaries (IPPF). UNFPA is giving technical support to the forthcoming census (early 1986) and is continuing to support the strengthening of demographic survey -and analysis capacity. UNFPA also provides funding for a part of the NCR program and, together with UhESCO, supports population education activities. The EEC has previously provided funding for the central hospital in Mogadishu, for a pharmaceutical production facility (not yet operational) and food assistance. The EEC programs suffer from design difficulties and the effectiveness of EEC funds would be greatly augmented if they were to collaborate in funding of more complex service oriented projects. Italian funding has been marked for the teaching hospital planned in Mogadishu. The Italian government and the Belgian Survival Fund are providing funding and technical assistance for MOR rural health programs. Italian technical assistance for financial management was started in 1984 and could provide an important base for furthur streutbening of government administrative and financial management capacity. 3.59 Absorptive Capacity. Discussions with donors and government offi- cials, analysis of such process data as is available, and on site aseessment indicates that the abrupt increase in capital expenditures since 1981 has exceeded the absorptive capacity of the health sector. Especially significant constraints to a rapid absorption of funds are a lack of trained management (pars. 3.20), inadequate planning capacity (para. 3.14-3.16), a limited information and monitoring system (para. 3.23), maldistribution of health personnel (para. 3.44 and 3.45) and insufficient recurrent funds (para. 3.60). An additional, and highly important constraint, is the inability to provide adequate financial planning, auditability and accounting (para. 3.61). It is important that either donor expenditures be reduced or the constraints released to avoid the long run institutionalization of cost-ineffective administrative and financial management procedures brought about by a continued input of greater funds than can be efficiently absorbed. 3.60 Recurrent Expenditures. Recurrent expenditures have not kept pace with capital expenditures over the last decade. Insufficient funds for supplies, including drugs, and for maintenance have reduced the effec- tiveness of health facilities including large urban hospitals and smaller regional and district facilities. For example Dikfa hospital, built with EEC sssistance in 1962, is in such a state of disrepair that one option under serious consideration is to tear it down and build anew rather than attempt to rehabilitate the present facility. Never, basic health care programs are also affected. Donor funded programs such as EPI, regional PEC, TB control and nutrition programs have added rapidly to health capital, but government recurrent expenditures remain far less than needed to sustain the programs. A demonstration of the underfunding of recurrent costs can be obtained by comparing 1984 annual donor development expenditures and recurrent government expenditures. The total disbursed 1984 donor development expenditures are expected to be US$7.4 million and 1984 government recurrent expenditures are budgeted to be US$8.6 million. With an estimated recurrent cost/capital cost ratio of .21 (based on the five year development plan estimates for PHC and referral system projects) the required incremental  37 recurrent expenditure to sustain the nev projects is US$1.6 million per annum or about 20X of current total recurrent expenditures, yet real MOB recurrent expenditures have fallen by approximately 7% per year over the last five years. Without an increase in recurrent expenditures, the current rate of development expenditures cannot be sustained. This point is elaborated in Chapter V (para. 5.16) and in Annex 3, Table 3.8, where it is estimated that requirercats for recurrent expenditures to maintain the current level of services and add the planned PIP projects will exceed available funds by a total of So. sh. 1300 million (US$ 76 million) over 1985 to 1989. 3.61 Financial Manaftement. Planning and financial management within the MOR is not adequate to give the information needed for the planning and control of health programs. The deficiency is due to both a lack of trained personnel and an inadequate control system. There is no regular reconciliation of budgeted and expended amounts. Budgets are formulated entirely at the central level without substantial input from district officers. Recurrent budgets are constructed incrementally on thle basis of past budgets, use highly aggregated line item categories and give almost no functional detail that can be used for monitoring and control. 3.62 Adoption of decentralized budgeting and planning with greater functional accounting detail would allow improved central management. Coor- dination of budgeting and accounting reforms with strengthened planning and data gathering would promote the rationalization of recurrent costs and donor capital expenditures. It would also facilitate improved donor coordination through reconciliation of donor programs with medium term and annual plans. Improved planning and management would be promoted by knowledge of how much is floving to specific functional areas (PHC, MCH/FP, Hospitals) and to specific geographic areas. Improved management control would also come about through knowledge of the relationship between actual expenditures and services in a given region or district and program. 3.63 Cost Recovery. Government health services are currently provided free of charge. The provision of all health services involves the use of national resources of infrastructure, manpover and funds. Ultimately, there are no free services. All costs, if not met through foreign financing or gsants, must be met by the Somali people either indirectly in the form of inflationary financing, semi-directly in the form of taxation or directly in the form of user charges. The difficulties in funding recurrent costs call for an examination of possible means of recovering a part of operating cost directly from users. For cost recovery to succeed a creditable quality of referral services are necessary and future introduction of charges should be paced to accompany the strengthening of services.  38 IV. NUTRITION A. Current Nutritional Status 4.01 The evidence on nutritional status is fragmentary and, in most cases, based on samples that are limited in both place and time. On all issues except the extent of protein-calorie malnutrition, hovever, the surveys present a coherent picture. They suggest that the main nutrition related health problems are iron deficiency anemia, goitre, riboflavin deficiency, diarrhea, and night blindness, and that small children and pregnant or lactating women are particularly vulnerable. In the case of protein-calorie malnutrition, however, there are vide discrepancies among the surveys. Protein-calorie Malnutrition 4.02 The Government estimates that 19% of the child population suffer from moderate, and 7% from severe malnutrition. These levels are similar to those found in other African countries: slightly higher than Zimbabwe, Lesotho, Cameroon, and much higher than Togo and Egypt, but less than Sierra Leone. Support for the view that protein-calorie malnutrition is a serious problem, at least among children, is provided by some, but not all, of the available nutrition surveys. The results of eight suzveys are summarized in Annex 4, Table 1. The urban surveys show a high level of malnutrition, but the figures from the rural surveys are much lower. Moreover, the National Morbidity Survey (1982) indicated a prevalence rate of 0.3Z for severe proteincalorie malnutrition among children 0-4 years old. (Annex 4, Table 4.2). There are several possible explanations for the discrepancies among tbe surveys: (a) The four surveys which show the highest rates of malnutrition vere all taken at MCR centers (two in 1975, two in 1984). One view is that these data represent merely the tip of an iceberg composed of children beyond the reach of the MCR system, or suffering from lesser degrees of protein-energy malnutrition. An alternative view is that they exaggerate the picture because the NCR population is atypical of the whole since the centers attract sick children; (b) The two high prevalence surveys that carry most weight in terms of the circumstances under which they were carried out, took place in 1975 at the end of a prolonged drought. The surveys showing the lowest prevalence rates were conducted recently in climatic conditions that were generally more favorable; economic conditions may also have been more Favorable (see pars. 4.11); (c) Higher rates of prevalence were found in the urban surveys than in the rural ones. This may reflect income differences between the two areas; and (d) The high rates were all recorded against weight for age measurements; the lower rates reflect height for weight. In a  39 society vhere births are not usually registered, the weight for age calculations may simply be inaccurate. 4.03 In an important sense, the absolute size of the problem is not the main issue. Food supply in Somalia is not abundant at most times, and may be disrupted by a range of factors, of which drought is the most obvious. The main issue is to identify the situations and groups that have the highest degree of nutritional vulnerability: which social, economic or geographical groups are most vulnerable even when supplies are adequate; what kinds of supply disruptions are most likely given medium-term trends, and on which groups are they likely to impinge most severely. Nutritional planning in Somalia should be built around these issues. Nicronutrient Deficiencies 4.04 Anemia. The evidence is much more conclusive that anemia is the most common nutritionally-related problem. The National Morbidity Survey found a prevalence rate of 40.7 per thousand. There is dietary evidence to indicate that one cause of the anemia is insufficient iron intake. The problem is particularly acute in pregnant and lactating women and also for weaning children. The incidence is also related to poverty; apart from housewives, the occupational groups most at risk according to the National Morbidity Survey were fishebrme (generally considered one of the poorest groups) and the unemployed. 4.05 Goitre (iodine deficiency). The National Morbidity Survey recorded an incidence of 1.4 per thousand, with the rate- for men being 3 times that of women. Sixty per cent of cases were in people over 20 years of age. The incidence among nomads is nearly tvice the national average. The incidence is largely confined to adults. B. Causes of Malnutrition 4.06 Although the data are sparse and poor, it appears that a wide stratum of society depend upon purchases for a part of their food supply. Nutritional vulnerability is linked, therefore, to purchasing pover as well as to capacity to withstand drought, to position in the social hierarchy and to custom. Settled farmers, fishermen and some of the urban population appear to be most vulnerable and, within these groups, pregnant and lactating women and weaning children are most at risk. Overall Availability of Food 4.07 The per capita food supply in Somalia has fallen over most of the last fifteen years. Production of staples fell from 81 kilos per capita in 1970 to 60 kilos per capita in 1980. During the 1970s, the food trade was tightly controlled and official prices kept low. These low prices probably acted as a disincentive to production, although this effect may have been moderated by increased auto-consumption and sales on the unofficial parallel market. Since 1980 the food trade has been liberalized progressively. The effects of the liberalization on nutritional status vary among economic groups. Those farmers able to increase production stand to  40 benefit. But it is likely that not all farmers vill be able to increase production significantly. Marginal farmers may benefit indirectly, however, from a movement in the domestic terms of trade favorable to agriculture, through increased opportunities for employment. The impact on rural and urban consumers will depend upon the extent to which subsidies actually prevailed in the past, and the effect of liberalization on total food supply. Food forms a major part of the expenditure of the poorest groups. In the past, the actual availability of food has had a greater effect on prices than has the official price level. If the reform stimulates a greater supply to the market it could have favorable nutritional effects. 4.08 Imports, especially concessional imports, have become an increasingly important element in food supplies. In the period 1980-82, concessional imports accounted for 27%, and commercial imports for 11% of total grain supply; together they provided over half of marketed grain supply. This trend has two possible nutritional implications. First, the level of total food supplies is dependent upon the availability of foreign exchange or of food aid. This may lead to supply instability. In mid 1984, for example, urban food prices were rising rapidly; the food component of the Mogadishu cost of living index rose by 80X between December 1983 and March 1984. This has been attributed, in part, to a sharp drop in maize imports8- Second, the increased imports may facilitate changes in consumer tastes towards foods that are nutritionally less satisfactory than are the traditional domestic staples. The obvious example is rice, which traditionally has been part of the Somali diet, but vhich may be increasing in popularity at the expense of maize and sorghum. Particularly serious is the apparent tendency to give polished rice as a weaning food to young children. Inports provide a major share (67% in 1982) of rice supplies. 4.09 Given the importance of purchases in household food supply, the sources and distribution of income have a direct bearing upon nutrition. The main sources of income are livestock exports and remittances from migrant workers primarily in Saudi Arabia. Because part (probably a substantial part) of this income flows outside of official channels, it is impossible to obtain a reliable estimate. Migrant worker savings alone could be as great as $3 million a year. If the level of protein-energy malnutrition is as low as some of the evidence suggests, one plausible explanation is that these earnings are providing an income supplement channeled through Somalia's tight kinship system to families who might othervise be vulnerable allowing them to maintain adequate food consumption. Any fall in the level of livestock and labor exports (or weakening of the kinship system) could have profound nutritional consequences. It is a matter of concern, therefore, that a question mark hangs over both export cosmmodities. The oil-induced economic slow-down in Saudi Arabia is affecting the market for both labor and meat. This decline has been exacerbated by a Saudi ban on imports of Somali cattle imposed in May 1983. A fall in livestock exports would probably result in an increase in meat supplied to the domestic market and a fall 8M. I. Asser et el, Agricultural Incentives and Grain Marketing in 80mali8. COS and World Bank, January 10, 1984, P15.  41 in price. But the nutritionally favorable consumption effects of such a shift would be limited by the fact that the poorest groups consume little meat, and would probably be outweighed by the nutritionally unfavorable income effects. Availability of Food Amonf PoDulation Sub-Groups 4.10 It is misleading to make a hard and fast distinction between nomads, settled farmers and the urban population. The groups merge into each other and differences within a group may be greater than differences between groups. Nonetheless, it is helpful to distinguish between their relative nutritional vulnerability because this highlights the causes of malnutrition. 4.11 Nomads. A high protein diet, based on milk and cereals, and access to -cash income from livestock exports make the nomads as a group relatively invulnerable to malnutrition. Also, their greater mobility renders them better able to survive short droughts than are other population sub-groups. Nonetheless, there are sources of danger. The current reduction of livestock exports to Saudi Arabia coupled with prolonged drought is creating the potential for a nutritional crisis among nomads over the near future. Moreover, particular social groups are disfavored. Pregnsnt women suffer from the custom of reducing their food intake sharply during the third trimester of pregnancy. While custom is more favorable to lactating women, their already weakened state makes it likely that malnutrition will continue. For children, the early availability of animal milk as a supplementary food provides some protection, but problems occur at weaning (which often begins very late) since the diet is normally cereal- -based. In all occupational groups, consumption of fruit and vegetables is low. Also, all groups suffer from infections that inhibit the absorption of nutrients. One survey found that 30% of pediatric inpatients at Benadir Hospital had diarrhea as the main cause of admission. While the nomad diet is satisfactory for proteins and for calories in good years, it is deficient in some micronutrients, especially of iron and vitamin C. 4.12 Settled Farmers are more vulnerable than nomads to drought. There is no evidence on- the size of cereal stocks held by farmers. It is known that traditional storage methods are effective, and that some farmers store grain for several years. But small farmers are unlikely to be able tc afford to hold large stocks and, hence, have to rely heavily on current production. They are also liable to lose their source of animal protein during a mild drought as the nomads move away in search of better pasture. Their children are more vulnerable than are the children of nomads because of the more limited availability of milk as a food supplement. Pregnant ard lactating women face similar problems in both communities. Settled farmers tend also to be poorer than nomads, except perhaps in the South where farming traditions are better established: a household survey in Brava, for example, showed that "nomads" and "settled farmers" both cultivated land and herd livestock; the main difference between them was that the farmers owned fewer livestock.  42 4.13 Urban Ponulation. There are a priori reasons for expecting that the poorest groups in the urban areas are the most vulnerable nutritionally. Evidence for the greater nutritional vulnerability of peripheral urban groups is provided from the nutrition surveys. The two main towns, Mogadighu and Hargeisa, have grown extremely rapidly and, to some extent are a refuge for those who have lost their livelihood in the rural areas. One estimate is that 10,000 people who live on the periphery of Mogadishu keep 50,000 head of cattle; daily output of milk from the cattle is less than 1 liter per head. Another factor contributing to urban malnutrition is poor environmental health infrastructure. C. Nutritional Activities and Policies 4.14 The nutrition-related activities currently supported by the Government are having only a moderate impact, in spite of the large volume of concessional food imports. The most substantial activity outside of refugee assistance is a WFP-assisted program of supplementary feeding to pregnant and lactating mothers and pre-school children through the MCR centers. Apart from this, there exists a very limited amount of nutrition education. Malnutrition is considered to be a serious health problem and after several false starts in the past nutrition has been identified as a priority in the National Health Plan and the 1984-86 Public Investment Program. 4.15 The National Health Plan outlines an appropriate but ambitious seven point strategy aimed primarily at pregnant and lactating women and pre-school children. The strategy, which overlaps with other NCR objectives, includes (1) nutritional surveillance (growth charts), (2) nutritional education for mothers, (3) home based preparation of weaning foods, (4) promotion of breast feeding, (5) treatment of diarrhea and i munization against measles, (6) encouragement of child spacing, (7) nutrition training of health personnel. The financial allocation for nutrition in the National Health Plan was modest, but in the Public Investment Program 1984-86 it is much larger: 13% of total capital expenditure on health and 25% of domestically financed capital expenditure on health is earmarked for unspecified nutrition programs, presumably related to the seven point strategy as well as food distribution. The earmarking of scarce domestic resources is an indication of the higher priority now being accorded to nutrition programs (narrowly defined), as is the recent appointment of the first medical nutritionist in the Ministry of Health. 4.16 However, it is not clear that the government has yet recognized the staffing implications of the seven point strategy, which emphasizes labor-intensive solutions. The National Health Plan adopts the position that "No additional personnel are needed" to implement the strategy because nutrition services will be integrated within the MCI and PHC services. In principle, the integration of nutrition with other primary health care activities is welcome. But the view that no additional staff will be required assumes that personnel recruited under the umbrella programs will be sufficient for all the tasks loaded on to them. This may be an optimistic view (see para. 3.32-3.33). If field health staff are overburdened, the tasks that remain unfulfilled are likely to be those that  43 are perceived to be of lesser urgency. And, given current perceptions, this is likely to include nutrition activities. 4.17 A major gap in the government's plans appears to be anemia. This is a serious problem yet is accorded a lower priority in the National Health Plan. Whereas protein-energy malnutrition is ranked as the sixth priority, anemia falls outside the core of highest priority diseases. 4.18 The distinction drawn in the National Health Plan betveen "Malnutrition" and anemia is illustrative of-the narrowly defined parameters of the current approach to nutrition issues. The seven point strategy is a sensible approach to protein-energy malnutrition among mothers and children, but it does not indicate a wider concept of nutrition planning to deal vith the causes of undernutrition among other groups, to protect the vulnerable from the consequences of drought, and to identify the nutritional implications of developments outside of the health sector. Vulnerable Group Feeding: 4.19 The WFP assisted program of vulnerable group feeding using food aid commodities, is currently operating in 89 of the 92 NCR centers. The ionthly average number of recorded food recipients attending the centers is approximately 25,000. This represents some 3% of the relevant population groups. Not only is this proportion small, but coverage is uneven over the country. The great majority of the MCI centers are located in cities and districc headquarter towns, so that the coverage of the rural population is negligible. 4.20 For those malnourished mothers and children who do receive it, the food ration is potentially very valuable. In terms of energy, the ration covers approximately 55X of the daily requirements of the children and 165% of the average additional requirements of the mothers; in terms of protein, the coverage is 200X and 260X respectively. There is a significant difference between the standards at the 12 MCI centers in Mogadishu, which are wellequipped and staffed, and those in the rest of the country. In theory, all clinics are staffed by up to three PlN/midvives who provide basic nutrition monitoring and guidance, and identify, from among the attenders, those that require a food supplement. In practice, a 1979 WFP evaluation found that at any one time about one-third of the nurses were pregnant or were absent from work for other health reasons, and that during such periods those centers staffed by only one PEN were practically closed down. A survey of operations at nine centers in 1981 by a WRO team found that in 45% of cases nutritional status vas not checked and in no cases was hemoglobin checked. Selection of child beneficiaries of the food ration is based, in theory, on nutritional status, but in practice is haphazard. Nut- rition education and cooking demonstrations are a part of the NCR curriculum, but attendance is below average on these occasions, and MCR staff do not always understand the nature of some of the food. 4.21 However, there are standard problems with this kind of food distribution program which may reduce its nutritional impact. The rations are distributed at monthly intervals and taken home by the beneficiary.  44 There is thus the likelihood that they are not consumed solely by the intended beneficiary but are shared among the family. Given the evidence that pregnant women and very younS children are disfavored by custom, the probability is that their actual increase in consumption vill be less than these figures suggest. Moreover, vithout adequate records, it is impossible to ascertain either whether actual rations are the same size as official rations, or the extent to vhich there is an overlap between the intended and actual recipients. A mismatch can occur either deliberately (because food is sold fraudulently, or because relatively affluent mothers obtain rations intended for the malnourished) or by accident (because of incorrect identification of the nutritionally vulnerable, or simply because there are insufficient rations). The absence of adequate record keeping and the haphazard selection of beneficiaries must make such accidental mismatch more likely. Figures obtained from the El Gab center in Mogadishu by the Mission for March 1984 show that 1069 children were recorded as suffering from second and third degree malnutrition, but only 724 received rations, together with 53 pregnant and 13 lactating women. Given that the Mogadishu centers tend to be better supplied than the average, the mismatch in rural areas could be expected to be much greater. 4.22 In some feeding programs, such problems are reduced by distributing an inferior food of liziAted attraction to the relatively affluent for consumption or resale, so that there is some degree of self-selection of the malnourished as beneficiaries. In Somalia, hovever, this is not the case. Apparently there is no so al stigma attached to receiving the ration, and there is evidence that all the foods distributed are readily acceptable. 4.23 In February 1984 for the first time, supplementary feeding was instituted at three primary schools in the poorer areas of Mogadishu at Yagshid, Karan and Wadajir. This is a small part of a much wider primary school feeding program which was planned with WFP assistance, but which has been held up by the absence of cooking facilitieb- A primary school feeding program is limited as a tool for combating malnutrition by the fact that school-age children are past the period of greatest vulnerability. However, it can be a useful educational supplement, making the children more attentive and receptive; and, in a country where cyclical undernutrition is a problem, it can be seen as a useful (but untargeted) household food supplement. It would not be cost effective to sustain this program if the recurrent cost were derived from monetary resources, but given that foreign support is in kind and earmarked for the program, it should be continued as an important part of the overall nutritional policy. Nutrition Education 4.24 Like other nutrition-oriented activities, nutrition education has seen some false starts. Some nutrition courses that used to exist have been discontinued, and the present level of nutrition education within the formal education system is very limited. Only the Home Economics Teacher Training Center provides a substantial course on nutrition, for teachers who have qualified to teach at Family Life Centers and as community development agents. Elsewhere, human nutrition forms a small part of biology, physiology and home economics courses.  45 4.25 This situation may improve in the near future. In the formal system, nutrition is to be added to the syllabus of the University of Mogadishu Faculty of Medicine in the 1934185 academic year. The development and transmission of appropriate nutritional messages is included in the PUC program, and also in the activities of the National Adult Education Centre, which uses outreach teaching methods to reach a wide spectrum of adult society. Vegetable and fruit gardens as a vehicle for nutrition education are also being developed at some educational institutions and in the regional integrated social development programs. V. ISSUES AND RECOMNENDATIONS A. Population 5.01 At the current fertility levels combined with declining mortality the rapidly growing population of Somalia will have nearly doubled by the year 2005 and tripled by 2015. This high rate of population growth threatens to absorb most or all of the gains from future economic growth. The fixed rangeland cannot support the projected growth in nomad population. Accommodation of the growth in nomad population in the urban and agricultural sectors would require a continuing annual investment of more than US$90 million in the development of irrigation and urban employment possibilities (para. 2.16-2.21). Yet, high rates of population growth also spur consumption and deflect national productive capacity and foreign exchange from the investment needed for agricultural and industrial development. Additionally, the high rate of population growth in all sectors will lead, ultimately, to urban growth that risks outpacing the Goverument's capacity to provide health, education and other social services at acceptable levels of quality and coverage. The transformation of attitudes and institutional development necessary to bring about a substantial decline in fertility would take time, and a high rate of population growth is unavoidable in the short run. But in the medium and long terms, lower fertility can substantially reduce population growth and be a major factor in determining the rate at which Government can bring social services to the people and generate investment for agricultural and industrial development. Population Policy and Family PlanninR Services. 5.02 The incipient family planning progran within the MOB and the SPHCA in the non-governmental sector provide a good basis for building population policy. In the past there may have been undue caution in introducing FP considering the increasing freedom of women and signs (para. 2.27 and 2.31-), albeit limited, of changing attitudes. With sufficient donor support, expansion of FP could occur at a faster rate than the currently planned extension from 31 to 92 HCH centers over five years. The recommendations to achieve faster expansion of FP focus on IEC efforts and improved delivery infrastructure. Information. Education and Communication 5.03 A major constraint in proceeding with population programs is the past pronatalist policy of the Government and the pronatalist traditions of the Somali culture. Key government officials are aware of population  46 problems and interested in taking steps to overcome them but implementation of family planning programs requires firm support at all levels, from service to management, within the NOR as well as the Ministries of Finance and Planning. The general awareness of officials about the relationship of population growth to health and nutritional status, external structural adjustment, public investment in infrastructure, ag-iculture and energy, and required funds for economic development needs to be increased. In addition, the rural population, and much of the urban population, are culturally oriented towards large families. There is very little knowledge of family planning methods or the role of family planning in promoting greater health and welfare. 5.04 A prototype program needs to be developed for a countrywide IEC campaign to promote child spacing. Suggested elements of the program include the following: (a) SWDO and the SFECA should be used at the forefront of IEC activities and should be given expanded staff and resources for this purpose. The design of the IEC strategy should reflect a careful consideration of the socio-cultural and religious background of the population and favorable Islamic religious interpretation, should be utilized as has been done in Egypt, Bangladesh, Malaysia and Indonesia. At the household level, the IEC strategy should highlight the health benefits of child spacing and breastfeeding, with reference to the household effects of population growth. (b) The population education activities of the MOE being supported by UNESCO should be continued and expanded, and the incorporation of family planning into the activities of the Women's Education Service should proceed rapidly (para 2.29). (c) The implicatious of population growth should be clarified among top level policy makers and middle level staff involved in program implementation. This could be carried out through seminars and policy papers developed by the MNP and MOR. 5.05 Infibulation. The practice of infibulation and female circumcision poses a serious health problem. A vigorous IEC campaign as a part of MOR MCR/FP activities and MOE programs should be mounted to make the public aware of the hazards. The GOS should give material and moral support to the SWDO's efforts to curtail the practice. In formulating the IEC campaign the experience of neighboring countries should be reviewed to derive culturespecific strategies. Delivery Infrastructure 5.06 The population and healtb strategies are closely interlinked. The urban bias if the health system and extremely low coverage of the population with basic health services make it difficult to use the system of rural health centers as the delivery vehicle for family planning services. In tandem with the IEC effort a major emphasis of the population strategy vould need to be placed on the strengthening and expansion of the national MCR/FP program and rural health delivery system. Important areas for  47 strengthening are physical facilities, logistics, management, and training. The delivery of FP services through the MOB will be enhanced by improvement of the health services infrastructure as vill be discussed in the health section below. However, specific programing of FPP/MC activities, staff training and IEC efforts should accompany (ratbe than await) the health service improvements. 5.07 As a supplementary vehicle for service delivery, consideration should be given to the establishment of a social marketing program. Also the incorporation of traditional birth attendants (TBAs) into the strategy for both IEC and service delivery, perhaps by allowing TBAs to market contraceptives, should be considered. The development of TBAs as an agent for contraceptive sales could serve to provide a substitute for income lost from reduced infibulation and deliveries. The CRS and TBA marketing programs would need to be kept under the surveillance, although not necessarily under the direct control, of the SFHCA and MOE to be sure that the promotional efforts are culturally compatible. B. Health 5.08 There are serious deficiencies in services that could be remedied by efficient use of capital, but there are clear indications that currently the MOB cannot adequately absorb the available donor funds (para 3.56). The first priority in the health sector is to eliminate the constraints to greater absorptive capacity. The constraints include inadeq,iate planning capacity, inefficient management and monitoring, and inadequate financial planning and accounting. Only after relaxation of these constraints, can an effort be made to bring about the substantive, durable improvements to health infrastructure required for greater coverage with health and family planning services. Long term technical assistance is required in each of these areas to build up the absorptive capacity of the MOE. The problems in organization, management, planning and finance are all closely interrelated and remedial steps for all tbree areas need to be coordinated. Preparatory studies need to be carried out to: (a) outline the basic requirements in the areas of health program management, financial management, planning, and manpower; (b) develop administrative, planning and financial management procedures; and (c) identify long term technical assistance and training needs necessary to carry out the suggested improvements. Recommended foci for change in the three constraining areas are set out belov. PlannmR 5.09 The National Health Plan is an important initial step towards the developsent of an operational strategy to reorient the health services to the problems of rural areas, but the MOE and NNP will need assistance to translate the NiP program into consistent medium term and annual implementation plans. The current procedure for project formulation is  48 haphazard and does not fully reflect the long term strategy for health. As a consequence of planning deficiencies, there is a lack of donor coordination and the Government is unable to assure that donor ezpenditures cover identifiable segments, horizontal or vertical, of the NiP in accordance vith national priorities. Further, project formulation should be made an integral part of the overall development process and HOP, MOF and MOE efforts should be complementary and consistent. 5.10 An inter-ministerial mechanism for a project approval process coordinated by the Ministry of Planning is necessary to ensure that health planning objectives and other sectoral development objectives are complementary to and consistent with each other, and also to ensure that a uniform development strategy and standard of resource allocation are applied to all sectors. The Social Service Sector of the HNP and the planning unit vithin the MOB need to be strengthened through training of staff in health planning and development of planning procedures. 5.11 Team work between the Department of Planning in the MOB and the concerned sections in the HNP should be promoted, and coordinated long and short term planning carried out. A medium term plan and budgeting framework (para. 5.15) would serve as a link between broad long term objectives and specific short term progjects and achievements. The medium term plan should be dynamically revised on a continuing basis as project implementation proceeds and experience vith the NEP strategy accumulates. Finally, an annual planning process should be developed to reconcile resource avail- ability with the long term strategy and update the medium tern plan on a rolling basis. 5.12 The planning process should involve a detailed reconciliation of planned capital expenditures with recurrent cost capacity. Thus, improved planning will require concomitant improvements in budgeting and financial management (see para. 5.15). Modest revision in the budget categories used by the MOE would contribute significantly to the planning process. Although more detailed revision might be important in the long tenm, the immediate revision in budgeting should not be more elaborate than necessary to allow the identification of expenditure by institution and major programs. It should be possible, following a study, to develop budget headings that would conform to specific MOB needs and be implemented with a minimum of disruption by the accounting units, yet be subsumed within the budget headings required by the MOF. Such a revision in budget categories would be of greatest benefit if it was carried out with strengthened surveillance and measures of program performance related to planning objectives. Together the surveillance and cost information would make it possible to monitor cost effectiveness and distribution of services to determine the best allocation of resources to achieve planning objectives. OrManization and Manatement 5.13 The organizational structure of the MOB (summarized in Figure I, and discussed in para. 3.17-3.22) is based on functional lines and conceptually compatible with efficient management, but in practice the structure does not function as designed. Probleams with management are affecting all areas of service delivery including logistics, XCR/PP programs and referral services. The only programs that are relatively unaffected are  49 those where donor support is being given to vertical management. In the short run, to avoid damage to ongoing programs, vertical management will have to be retained, but in the long run vertical maagement is an inefficient use of limited financial and managerial resources. The impediments to efficient management are a lack of clearly defined lines of authority, insufficient training of top and mid level staff in management, and inadequate information and surveillance systems. The management study recommended above (para. 5.08, item a) should include an analysis of the managerial requirements needed to carry out NOR programs within the existing organizational structure. As part of this analysis, the responsibilities of senior and mid level officials should be better defined. Training in health management is needel1 for a few senior staff who could then be instrumental in carrying out short duration retraining of mid level staff. To complement the effort in management, the ongoing NOR/WHO actions to initiate a reporting system should be further supported to include process data for all programs as well as periodic epidemiological impact data. Financial Issues 5.14 The constraints on absorptive capacity together with the underfunding of recurrent costs mandate a reexamination of MOR financial management. The shortage of operating revenues is expected to grow over the next five years and allocational choices will become increasingly difficult. Improved financial information is needed to analyze the cost effectiveness of NOR alternatives and develop budget priorities. The shortage of recurrent funds also calls into question the government policy of providing health services without charge. The first section below gives some recommendations for improved financial management. The second section gives the results of projections of NOR recurrent costs and available funds and draws implications for future expansion. The third section considers the possibilities for health cost recovery. 5.15 Financial ManaRement. Improved financial management and accounting is required to give the information needed to achieve the optimal allocation of resources to health programs. Accounting and budgeting headings need to be established to coordinate with NOR organization and provide knowledge of the flow of funds to specific functional areas (for example PlC, NCHIPP, Inpatient facilities, outpatient facilities) and to specific geographic areas. Reconciliation of budgeted and actual flows would add greater control over programs. Appropriate financial management would also improve donor coordination through rationalization of donor capital expenditures and recurrent cost capacity. The mission recomends that a study be carried out to examine the current financial management system and recommended procedures for improved financial planning, management and budgeting. Such a study would be most efficiently done in conjunction with the planning and management studies identified above (para 5.08 and 5.12). Long term technical assistance and training program would be needed to implement the recomuendatious. 5.16 Recurrent Cost Fundint. Minimum estimates of the supplementary recurrent expenditures needed for maintenance and the Public Investment Programme over the period 1985 to 1989 have been added to a projection of baseline expenditures in order to obtain a conservative estimate of required recurrent funds. The projections are conservative because of the low level  50 of per capita baseline expenditures and the low maintenance expenditure per hospital bed and facility used in their formulation (see Annex 3, Table 3.8 for details). When recurrent funding needs are compared to a projection of available funds over the period from 1985 to 1989, the result is a total shortfall of So. sh. 1330 (US$ 76) million over the five year period. If the planned hospital projects in the PIP are omitted, the cumulative savings in recurrent cost over a five year period would be approximately 3o. sh. 260 (US$ 15) million, using a conservative recurrent cost ratio of ,21. Even if the hospital investments are avoided, however, the recurrent cost deficit will remain large. A potential source to cover a part of the shortfall is a greater government commitment to the health sector. If the share of the government budget going to the NOR were to recover to levels reached at the end of the 1970s over the next five years to reach 52 of recurrent government expenditures by 1989, the shortfall would be reduced to So. sh. 1090 (US$ 62) million. Even larger government support is needed but given the fiscal constraints that exist and the underfunding of recurrent costs in other sectors it may not be forthcoming until a larger tax base can be developed. In the interim period it will be necessary to continue to rely on donors to carry some of the burden of recurrent costs, especially those involving foreign excbatge. In addition, the urgency of finding additional sources of recurrent funds requires that the government reexamine its current policies on cost recox7ery for government health services. 5.17 Cost Recovery. The use o.f charges for selected health servi:es can be justified (1) as a means of raising additional revenue to cover operating costs; (2) as a means of improving the efficiency of resource allocation; and (3) as a means of improving the equity of health service use. The major justification fo- fees in Somalia is as a revenue measure but equity and efficiency are also considerations. Preventive services that have substantial benefits accruing not only to the individual receiviag the services but to society at large, or services used primarily by the poor, should be provided without user charges. Services of this type include (1) most disease control programs such as malaria, communicable childhood diseases, -.ually transmitted diseases, leprosy and tuberculosis; (2) antenatal care; (3) nutritional programs; (4) well baby services (growth monitoring, immunizations, instruction in oral rehydration therapy); and (5) family planning services. 5.18 Beyond the core group of basic services, there are considerable possibilities for cost recovery. A schedule of charges for drugs and supplies and charges for non-referred use of physicians ana higher level care could be drawn up. Pharmaceuticals accompanying preventive care could continue to be provided free but other pharmaceuticals could be provided at full or subsidized costs according to government schedules. Nominal fees could be levied for first visits to outpatient clinics to discourage unjustified use of services. Also, referred use of physicians and higher level care could be allowed without charge but patients who desired to enter the health system at levels above primary health care and who are not emergency cases  51 would be required to pay fees. This would help prevent misuse of higher level facilities and would shift a part of the cost of higher level care to upper income groups that are aore apt to try to circumvent entry into the health system at the primary health care level. 5.19 It is proposed that a study be conducted to examine the costs and benefits and possible design of a system of cost recovery. The proposed system of charges should be developed with reference to the costs of different services and facilities, and the study should evaluate the likely effects of the charges on the demand for services. Equity effects should be carefully evaluated. Finally means for assuring that collected fees accrue as closely as possible to the level of the collecting agency should be examined. One efficient possibility would be for health centers and hospitals to use collected fees to directly replenish a district level revolving fund for drugs and supplies. Auditing and monitoring would be carried out at higher administrative levels but the actual administration of the fund would be carried out at the facility level. The effectiveness of a system of cost recovery and the efficiency of collection would be seriously jeopardized if the fees accrued at the central level to the MOP. If cost recovery started in 1987 and 10% of private expenditures could be diverted to fees for government services, So. Sh. 132 (US$ 7.5) million, or approximately 10X of the projected recurrent cost deficit, could be recovered between 1985 and 1989 (Annex 3, Table 3.9). Health Services. 5.20 After relaxation of the constraints identified above, extension of PHC services to a broader population base and inproved quality of basic referral services is needed. Additionally, special efforts should be made in affected regions to control tuberculosis and schistosomiasis. The strategy for strengthened health services should expressly concentrate on cost effective programs of basic care to achieve a higher coverage. Construction of large scale central facilities should be avoided as it would further divert recurrent funds from wider coverage. 5.21 Primary Health Care. Coverage with the PHC program should only be extended as managerial and planning capacity is built and experience with the existing projects has been used to design improved supervision, training and coverage. The program should continue to rely on paid workers and strategies to promote broad community support, while concentrating on a minimum number of interventions - immunization, ORT, nutrition, family planning, antenatal care, and first aid. Development of community support could include incorporation of TBAs into PHC activities. A study with inputs from an anthropologist and sociologist may be valuable to identify the best way of using TBAs and also to find means of extending services to nomads. In the near future, the PlC program vill probably retain its current structure, with individual regions assigned to specific donors. The current structure allows donor funded programs to retain vertical control over management and logistics for PEC in order to ensure effective program implementation. Rowever as NOR management and logistics capacity is strengthened the PEC programs should be gradually incorporated into a unified KOR system including logistics, transport, and information. In additiou, the NC/EFP program, which is not uniformly accessible throughout  52 PEC regions, should be incorporated vith PRC and extended as rapidly as possible. 5.22 Referral Services. Improvement of referral services should concentrate on improvement of quality rather than increased numbers of facilities. The large scale hospitals included in the PIP should be reconsidered in light of more pressing needs for preventive and basic curative care. There are selected regions where small scale facilities are needed, but the greatest current referral problems are bringing about a more equitable distribution of staff (especially doctors), inadequate provision of drugs and supplies, lack of running water and sanitation, and inadequate standards of care in existing facilities. Activities to attack these problems include improved housing and incentives for staff relocation, improved quantity and availability of drugs and supplies, upgrading and repair of facilities, and staff retraining. The highest priority with regard to facilities should be given to increased funding and retraining of staff for improved maintenance. Health Personnel 5.23 The main issues related to health personnel are: (1) personnel planning, (2) training for various types of personnel, and (3) distribution of personnel. The flow of personnel into the health system has not been coordinated with the new emphasis on rural health services and strengthening of lower levels of the referral system. Training for physicians and mid level health workers is mainly focused on curative medicine and does not have an adequate emphasis on prevention or the basic services needed in rural areas. Also, training of higher level personnel is deficient in administrative content. Problems with the management and implementation of the training for PHC has damaged the quality and quantity of basic health workers for the program. Finally, problems with staff housing, incentives and compensation have impeded an equitable distribution of manpower. 5.24 In order to respond to the new focus on rural health services, it is recommended that a trained manpower planner be employed by the MOE and that a task force should review the current manpower situation -to provide a basis for a start towards continued manpower planning. The amount and quality of training time devoted to maternal child health and family planning should be increased for all categories of staff. Retraining for improved msnagement is needed for upper and mid level staff. As a first step attention should be given to improved programs for retraining of trainers. Finally, efforts should be made to achieve a better distribution of personnel and reduce attrition rates within training programs and among professional cadres. Careful selection of program entrants, attention to the design and relevance of curriculum, and better management of training programs is needed to reduce attrition within training programs. It is recognized, hovever, that the primary requirement to reduce attrition and achieve better personnel distribution is salary rationalization, better rural staff housing and infrastructure and other incentives. Ultimately, personnel distribution and attrition are larger issues faced by several sectors and need a broader approach than can be achieved within the health sector alone.  53 C. Nutrition 5.25 The Government has appropriately assigned a high priority to combating malnutrition but the approach to nutrition programs has remained narrowly focused on reducing protein-calorie malnutrition (PCM) through food distribution schemes. The nutritional efforts need to be broadened without reducing the attention given to maternal and child PCM. The narrow approach has contributed, for example, to a relative neglect of anemia which is the most widespread nutritional disorder. Additionally, it has resulted in a neglect of the inter-sectoral issues linked to the causes of malnutrition. Intersectoral Issues and Nutritional Planning 5.26 Livestock exports and the remittances of migrant workers yield substantial income flows to a wide stratum of society; without this income, levels of PCM would be higher than they are. Demand prospects for both migrant labor and livestock exports are problematic. Action to develop new markets and contingency plans to cope with a drop in income from these sources lie within the portfolios of ministries other than the MOB. The MOR retains a responsibility to provide the impetus necessary to enunciate and underline the nutritional implications of migrant labor and livestock exports. Analysis of broader aspects of policies is, however, a shared intersectoral responsibility, and must involve ministries and donor programs outside of health. 5.27 Similarly, agricultural policies may have profound nutritional implications that need to be brought out. Because of growing pressure on the rangeland (see the population section) there will be a future need for some sedentarization of nomads. Given the greater vulnerability of settled farmers, and their wives and children, sedentarization could have adverse nutritional consequences. It should be accompanied, therefore, by measures to reduce the vulnerability of settled farmers. These include agricultural development strategies that enable small farmers in rainfed areas to grow and keep more food for home consumption: improved technology, more inputs, better pest control. Even with these measures, dryland farmers will remain vulnerable to drought and attention should be given to better planning of drought-related food distribution systems. 5.28 Monitorinu and Surveillance. Nutritional planning would be aided by improved nutritional information. There is a lack of clarity about the nutritional status of pre-school children, on whom most attention has been focused. Given the large size of the international assistance to nutritional programs, an additional effort is required to obtain both process and output indicators to monitor the functioning of programs and to assist in targeting and design. 5.29 Weight and height charts should be used more routinely in the MCR centers. These data could be collated and analyzed without a great deal of additional manpower or expense. Also, advantage should be taken of planned household expenditure or agricultural surveys to include questions on nutrition. Among the data that is most valuable for filling gaps in knowledge about nutrition are those that differentiate between  54 occupational and geographical groups, and that illuminate the seasonal dimensions to malnutrition. Technical information on incidence and clinical aspects of nutrition should be made regularly available by building on the nutritional division that has recently been started in the MOB. Nutrition Protrams 5.30 The major established nutrition program, the supplementary feeding scheme for mothers and children carried out through the NCR centers, is relatively successful from a logistical viewpoint but its effectiveness as a nutritional intervention is hampered by the limited coverage of the MCR program and the design of the delivery system. An important problem is the difficulty of identifying and reaching appropriate target groups. Targeting based on attendance at distribution centers gives an illusion of spurious accuracy. Because of the large number of people near the minimum adequate food consumption even in good years, consideration should be given to widening the range of the feeding program. A primary school feeding project has been planned for some time but implementation has been delayed by the lack of adequate cooking facilities. A food for work project could provide opportunities for people to earn income in kind while constructing the needed facilities. Other opportunities should be sought for enhancing the success of feeding projects by linking a source of food with ancillary support of services, non food supplies, and related infrastructure. 5.31 Poor dietary practices based on custom are an important cause of malnutrition, particularly among pregnant women. There is an important role for IEC to perform to combat such unnecessary malnutrition. The present level of nutrition education is low. The plans to develop and communicate nutrition messages as part of the PHC program and in the activities of the National Adult Education Center should be supported. 5.32 The feasibility of measures to overcome micronutrient deficiencies in the Somali diet should be investigated. A high priority should be given to the feasibility of iron fortification of sugar, which forms a major part of the diet of the poorest urban groups, and is used throughout Somalia. Iodine fortification of salt should also be investigated.  55 ANNEX 1A Page 1 of 8 pages Table 1.1: SIZE OF POPULATION BY SECTOR, 1975 CENSUS Sector Enumerated Population Adjusted Populationa/ Percentage Estimated Numbers Percent Numbers Percent Undercount Nomad 1,444,665 44.4 1,879,649 46.0 23.1 Rural Settled 1,015,791 31.2 1,172,089 28.7 13.3 Urban 792,568 24.4 1,037,465 25.4 23.6 TOTAL SCMALIA 3,253,024 100.0 4,089,203 100.0 20.5 a The adjusted population was derived by the Governument using procedures that resulted in an estimated undercount of the population by 30% for nomads, 15% for the rural settled population and 30% for the urban population.  56 ANNEX LA Page 2 of 8 pages Table 1.2: MORTALITY AND FERTILITY ASSUMPTIONS USED IN POPULATION PROJECTIONS 1985-90 1990-95 1995-2000 2000-05 2005-10 2010-15 Project Mortality (I assumption) Life expectancy, both sexes 47.8 50.0 52.4 54.5 56.8 58.9 Project Fertility:. Total Fertility Rate (3 assumptions) A. No fertility decline 6.70 6.70 6.71 6.74 6.75 6.77 B. Gradual fertility decline 6.70 6.64 6.35 5.87 5.32 4.71 C. Accelerated fertility decline 6.70 5.52 4.55 3.76 3.10 2.56  1ue 1.3: PR== SIZ "/ CF ME PW1I=N CF SOMJA BY HP(C SECflR: 1975-2015 (NO KaT1r AMXG SCRS) 1975 1980 1985 1990 1995 2000 2005 2010 2015 ,ed sector lb fertility decliU 1.89 2.06 2.31 2.63 3.01 3.46 3.99 4.63 5A0 Gradual fertility de1Ine 1.89 2.06 2.31 2.63 3,02 3.44 3.89 4.37 4.88 A de1erted fertility declm 1.89 2.06 2.31 2.63 2.93 3.22 3.49 3.75 3.97 Agricultural Sector lb fertility declip 1.18 1.33 1.53 1.79 2.11 2.50 2.99 3.60 4.35 Grdua fertility decline 1.18 1.33 1.53 1.79 2.11 2.47 2.89 3.34 3.83 kIearateI fertility declive 1.18 1.33 1.53 1.79 2.03 2.27 2.50 2.73 2.93 Urban aBctor tb ertiity wine 1.06 1.27 1.55 1.86 2.22 2.70 3.31 4.09 5.06 Gat1 fertility decline 1.05 1.27 1.55 1.86 2.20 2.60 3.07 3.61 4.17 kceeiraw fertility dcwlie 1.05 1.27 1.55 1.86 2.14 2.41 2.69 2.97 3.21 Ibtal SommllA b/ lb fertility decltm 4.12 4.67 5A0 6.28 7.35 8.67 10.30 12.31 14.81 Gadm&l fertility decline 4.12 4.67 5.40 6.28 7.32 8.51 9.85 11.32 12.89 m Acmlarted fertilty delne 4.12 4.67 5.40 6.28 7.10 7.91 8.69 9A4 10.12 a a/ W1ith,u nent aectord lpattai. Vfth etoral migration thl into of tio total population wlt be luer, a dgation ocazs frm lzwr fertilty rmv4ic region to hl r fettilty setad agriditural and urban are. as b/ Fm1 wt adsd ezictly to txals, due to mgdixg errow.  Tae 1.4: PK)JECI VTL RA1!ES CF THE FiJlAtN (a SOIA a/: 1980-2015 1980-85 1985-90 1990-95 1995-2000 2000-05 2005-10 2010-15 QUIZ BUM RAT (PER 1000) A. Nb Fertility Dcline 49.5 48.8 48.3 48.3 48.3 48.1 47.9 B. Gndual Fertility Decllne 49.5 48.8 47.6 44.9 42.3 39.4 36.1 C. kcelerated Fertility ecline 49.5 48.8 40.5 34.8 30.7 27.1 23.4 MMi1E 1M RA (MHR 1000) A. Nb Fertility ecUne 20.3 18.4 16.8 15.2 13.8 12.3 10.9 B. Gradual Fertiltyl Dwcm 20.3 18.4 16.7 14.8 13.1 11.5 10.1 C. kAlerated Fertility Declre 20.3 18.4 15.6 13.4 11.7 10.5 9.5 U' AVE= ARtL M (F PMiTIC GUM(PEIM) A. Nb Fertlity Dwc1ize 2.92 3.04 3.15 3.31 3.45 3.58 3.70 B. awl FertIlty elnI 2.92 3.04 3.09 3.01 2,92 2.79 2.60 C. AcIerated Fertility Dcline 2.92 3.04 2.49 2.14 1.90 1.66 139 _ I.athlut intersectoral mdgration 0 ,. 5 co