Report No. 24422-YEM Republic of Yemen Poverty Update (In Two Volumes) Volume II: Annexes December 11, 2002 Middle East and North Africa Social and Economic Development Group (MNSED) u Document of the World Bank CUR= E1diTCy 1 l-QVJVALuZN¶P Unit of Currency = Yemeni Rial (YR) Period Average Exchange Rates (YR per US dollar) 1995 1998 1999 2000 2001 2002 127 1 141.7 159.7 161.7 169 8 178 3 IP'E$C˘AL YITAR January 1 - Decermber 31 ACRfl(DN'YM1$S AN::) i . 'HS AFPPF Agriculture and Fisheries Production PAEG Poverty Alleviation and Employment Promotion Fund Program BEEP Basic Education Expansion Program PCE Per capita expenditure CDD Community-driven development PER Public Expenditure Review CPI Consumer Price Index PG Poverty gap CSO Central Statistical Organization PIVMS Poverty Information and Morntoring EMIS Education Management Information System System ppp Purchasing power parity EOS Economics of scale PRSP Poverty Reduction Strategy Paper ESIP Education Sector Investment Project PWP Public Works Project FAO Food and Agncultural Orgamzation REFLECT Regenerated Freirean Literacy of the United Nations Through Empowerment Communuty FFYP First Five Year Plan Techniques GCC Gulf Cooperation Council SFD Social Fund for Development GDP Gross domestic product SFYP Second Five Year Plan GER Gross enrolment rate SWF Social Welfare Fund HBS Household budget survey TFR Total fertility rate IDA Intemational Development UCW Understanding Children's Work Association (UNICEF/ILO/WB) LDB Live database UNOP United Nations Development LFS Labor force survey Program LMIS Labor Marlcet Inforrmation System UNICEF United Nations Children's Fund MDG Millennium development goal VT Vocational traiing MENA Middle East and North Africa WDR World Development Report MICS Maternal and chlid survey WFP World Food Program MOE Minstry of Education WHO World Health Organization MOPH Ministry of Public Health YAR Yemen Arab Republic NA National accounts YDMCHS Yemen Demographic and Matemal NPS National poverty survey and Child Health Survey p.a. Per annum Vice President: Jean-Louis Sarbib Country Director: Mahmood Ayub Sector Director: YIustapha Nabli Task Team Leader: Setareh Razmara ACKNOWLEDGEMENTS This report has been prepared by a team of several people and is based on the findings of a mission to Yemen in January 2002. The team included Setareh Razmara (Task Team Leader), Giovanni Vecchi (Consultant), Dominique van de Walle (DECRG), Martin Ravallion (DECRG), Nadir Mohammed (MNCO3), and Takako Yuki (Consultant). The report was prepared under the supervision of Dipak Dasgupta (Sector Manager, MNSED). Inputs were also provided by Furio Rosati, coordinator of UCW project on child labor, and Lorenzo Guarcello (UNICEF Florence). Mohamed Al-Sabbry (MNCO3) has provided assistance in primary data collection and analysis of the surveys, and has greatly facilitated the collaboration with the Central Statistical Organization (CSO). Valuable comments and suggestions were received from Mustapha Nabli, Jacques Baudouy, Sameh El-Saharty, Jeffrey Waite, Keiko Miwa, Jean-Francois Barres. Peer reviewers were Margaret Grosh (HDNSP), Polly Jones (LCSHD), and Kalanidhi Subbarao (AFTH2). Christina Djemmal edited the report and Emma Etori was responsible for formatting. A special thanks is extended to all Govemment officials, particularly the Ministries of Planning and Development, and Social Affairs, and to the PRSP team for their support and active collaboration. The guidance provided by Mr. Abdel Rahman Tarmoun (Vice Minister of Planning and Development) and Dr. Yahya Al-Mutawakil (PRSP Coordinator) is highly appreciated. In particular, the CSO Chairman, Mr. Abdolraboh Gradah, facilitated the team's work and supported the preparation of the poverty analysis. His technical team kindly provided assistance and information from the household surveys (1998 HBS and 1999 NPS). REPUBLIC OF YEMEN POVERTY UPDATE VOLUMEE : ANNEXES TABLE OF CONTENTS ANNEXES 1. Household Surveys in Yemen I 2. New Poverty Estimates for Yemen 1998: Methodology and results 4 3. Economies of Scale 17 4. How to improve the quality of the 2003 HBS? 19 5. National Accounts Data 21 6. Poverty Incidence Forecasts 24 7. Education Incidence analysis 29 8. Poverty and Private & Public transfers in Yemen 68 Statistical Annex 98 ANNEX 1 HOUSEHOLD SURVEYS IN YEMEN Since the time of Yemen unification, the Central Statistical Organization (CSO) has implemented three household surveys: (i) the 1992 Household Budget Survey (HBS-92), (ii) the 1998 Household Budget Survey (HBS-98), and (iii) the 1999 National Poverty Phenomenon Survey (NPS-99). This annex illustrates the main features of the three surveys, and discusses the limits of their utilization. Table 1.1 compares the 1992 and the 1998 HBS survey designs.! Among the main differences between the two surveys are: (i) the 1992 sampling frame was replaced by the 1994 population census frame, (ii) although a two-stage stratified sample design is comrmon to both 1992 and 1998, the former was defined on 4 strata, 704 clusters, 9,152 households, as compared with the 1998 design which was based on 12 strata, 840 clusters, 15,120 households, and (iii) recall periods for expenditures on food and beverages are last week for the 1992 HBS, last month for the 1998 HBS. The use of the HBS-92 requires extreme caution in that the 1992 datasets lack the expansion factors (i.e., the reciprocal of the inclusion probabilities) needed to guarantee the representativeness of the estimates based on the 1992 sample. In fact, given the urban bias of the HBS-92 sample (urban observations represent 72% in the 1992 sample, whereas the share of the urban population is estimated at about 20%), the lack of weights prevents from any meaningful poverty comparisons with the 1998 HBS. For the above reasons, the HBS-98 represents - defacto - the first household budget survey for Yemen, and it will be used for establishing the baselne for measuring income poverty for Yemen. The HBS-98 was carried out during a full year (January-Decemnber 1998) to ensure that seasonal patterns in consumption expenditures and incomes are adequately taken into account. Households were sampled according to a two stage sampling plan: the first phase relates to the selection of clusters, and the second phase concentrated on the selection of the sample households. The sample consists of 15,120 households which were interviewed over 4 rounds, completing 3,780 households in each round. Rounds took place according to the following calendar: (1") January 1 - March 31, (2nd) April 1 - June 30, (31d) July I - September 30, and (4"') October I - December 31. The number of responding households is 13,641 households, corresponding to 97,544 individuals.2 The 1998 HBS collects the following information: 1. Basic characteristics of household dwelling and possessions 2. Socio-economic characteristics of household members 3. Local society services 4. Consumption expenditures 5. Incomes 6. Water and electricity consumption In fact, the data concerning sections 3 and 6 above have never been used by the CSO, with the consequence that no official datasets are available for public use. In contrast to the HBS-98, the NPS-99 was conducted during the course of only one month (September-October 1999), but covered a larger number of households (49,450, corresponding to 367,987 individuals). Such a remarkable sample size makes the NPS-99 sample representative at the govemorate level. The main purpose of the NPS-99 was primarily to provide detailed information on lThe scanty documentation available for the 1992 HBS did not allow us to provide more details than those shown in Table 1. 1. 2 Further details are provided in Republic of Yemen, Ministry of Planning and Development, Central Statistical Organization (1999), Summary of Final Results - Household Budget Survey 1998, Sana'a, July. 2 Annex ) access to services and other aspects of non-income living standards. Since the survey was conducted over one month, household expenditures are likely to be severely biased by seasonality effects.3 This may affect all analyses which rest on the NPS-99 expenditure and/or income variables, specially poverty analysis, benefit-incidence analysis and assessments about the targeting of transfer incomes. Overall, the HBS-98 seems to provide adequate information on incomes and expenditures. In this respect it can be regarded as the best candidate for providing the baseline for reaching the poverty reduction by 2015 MDG. However, the HBS-98 is severely deficient in information necessary to measure the non-monetary facets of well-being. Among the most important areas not covered by the HBS-98 are: (i) health conditions, (ii) access to basic services and social infrastructures, (iii) access to credit, and (iv) labor market conditions (see Annex 4). In contrast, the NPS-99 is best used in assessing the non-income poverty (specially access to social services) thereby supplementing and complementing the HBS-98. 3"Me food consumption recall period was only for the last week. Moreover, the NPS covers far fewer consumption items than the HBS. For example, the lBS collects consumption on 20 cereal products compared to 9 in the NPS. 3 Annex I Table 1.1. HBS 1992 versus HBS 1998, A Comparison 1992 1998 Sarnple firarne n.a. 1994 populatLon census Response rate n.a 90 2% Two-stage stratified: 4 strata, (1") 704 .Two-stage stratified: 12 strata, (1s) clusters, (2nd) 9,152 households. 840 clusters, (2nd 15,120 households 60,550 mdnviduals corresponding to 97,544 individuals corresponding to (Actual) sample size 8,405 households (6,052 urban, 2,353 13,641 households (8,626 urban, 5,015 rural) rural). Rounds Four: dates n.a. Four. Jan 1 to Mar 31, Apr 1 to Jun 30, Jul 1 to Sep 30, Oct I to Dec 31 Expendfiture on food and beverages: week Expenditure on food and beverages: Time reference/Recall period Expenditure on nonfood commoceties, month (forms filled in weekly). Tl refrenc/Reclloprtod Expndture on nonfoodExpenditure on non-food commodities. month last month, three months, year. 1. Characteristics of household . Socio-economic characteristics of 2 dwelling and possessions 1.hSolousehonold membe harscten of 2. Socio-economic characteristics of household members household members 2. Income 3. Local society services 3 Expenditures on food 4. Expenditures on food comnmodities.. Questionnaire, sections of 4. Expenditures on non-food commodities commodties ad servces 5.Expenditures on rapid circulation commodities and services commodities and services 5. Income 6. Expenditures on non-food 6. Characteinstics of household commodities and services dwelling 7. Income 8. Water and electricity consumption Expenditures on both food and non- Consumption food consumption includes in-kind Idem. consumption Items. FOOD FOOD 92-1) Grains & related products (15) 98-1) Cereals and by products (20) 92-2) Dry & canned vegetables (11) 98-2) Preserved and dned legumes (13) 92-3) Fresh and canned vegetables (17) 98-3) Fresh and preserved vegetables (22) 92-4) Fresh and canned fruits (20) 98-4) Fresh and preserved fruts (27) 92-5) Meat, poultry, fish and eggs (15) 98-5) Meat (14) 92-6) Daisy and subproducts (8) 98-6) Fish (5) 92-7) Oils and fats (8) 98-7) Milk, milk products and eggs (12) 92-8) Sugar, and sugar products (7) 98-8) Oils and fats (8) 92-9) Other food products (9) 98-9) Sugar and sweets (8) 92-10) Tea, coffee, cococa (6) 98-10) Spices and other foodstuff (10) 92-11) Mineral water, sodas and juices (7) 98-11) Tea, coffee and cocoa (5) 92-12) QAT and tobacco (6) 98-12) Mineral water and carbonated drnks Consumption aggregation plan NON FOOD 98-13) Tobacco and QAT (in parentheses the number of 92-13) Cloth materials and clothes (12) NON FOOD commodities which belong to 92-15) Personal beauy products (20) 98-14) Personal services (6) the aggregate). 92-16) Other tools, services (of personal 98-16) Domelslng (8) nature) (13) 98-16) Dwelling (8) 92-17) Furniture and equipment (14) 98-17) Lighting and fuels (10) 92-18) Entertainment equipment (18) 98-18) Fabrcs and ready-made clothes (46) 92-19) Household durable equipment (14) 98-19) Shoes (11) 922)Masfor transportation (4) 98-20) Personal care and cosmetics (29) 92-20) Means for 98-21) Furnture and domestic fittings (20) 92-22) Cleaning proucts ( 98-22) Leisure durables (8) 92-23) Educationuand spro rtuctivstLes (8) ) 98-23) Household durable equipment (19) 92-23) Educaton and spot actvties (14) 98-24) Commumcation and tansport (7) 92-24) Health services (10) 98-25) Health catre services (10) 92-25) Transportation and communication (6) 98-26) Heans of servies (4) 92-26) Other transfers and expenditures (11) 98-2) Means of transportation (4) 92-27) Fuel and electarity expenditures (10) 98-27) Educanton culture and sport (16) 92-28) Other products not mentioned above 98-28) Transferable payments + other (25) ANNEIlX 2 NEW POVERTY ESTMAAIE$ FOR YEN0N a990 METHODOLOGY AND LESUL7 Hntroductlom The methodology used in this report to estimate poverty lines for Yemen in 1998 is along the lines suggested by Ravallion (1994, Appendix 1). According to this method poverty lines are made up of two components: (i) a food poverty line, giving the cost of a bundle of foodstuffs attaining a pre- specified minimum food energy requirement, and (ii) an allowance for basic non-food goods. Depending on the specific definition of the two components poverty lines may vary not only between countries but also withm any one country. Three poverty lines were estimated for Yemen 1998: 1. A food povertv line, giving the cost of a bundle of goods necessary for attaining a pre- determined minimum food energy requirement 2. A total poverty line, making an allowance for basic non-food components. 3. An uyuerpove line, making a more generous allowance for the non-food component than the total poverty line. This annex discusses the choices made in calculating the poverty lines used throughout the report. IFood Poverty Line The food poverty line (ZF) can be defined as the cost of a food bundle deemed to assure that basic consumption needs are met. In order to identify "food basic needs", we followed common practice by fixing the food energy requirement threshold equal to 2.200 calories per person per day. The food bundle used in this report was estimated by Dr. Abdul Bald (Sana'a University), and adopted by the Central Statistical Organization (CSO). Table 2.1 shows how the food bundle was scaled down to reach 2,200 calories per person per day. The cost of the bundle was determined on the basis of two sources of information: 1) unit values (expenditures divided by quantities, from the HBS 1998 data) after purging these of quality and scale effects4 , and 2) 1998 price survey data. Source 1 entailed regressing the unit values for each commodity on a complete set of govemorate dummy variables (split urban rural) plus log household expenditure relative to the national median and log household size relative to the national median). The following regression was estimated by ordinary least squares: unitvaluea = 6,1,GOV1 +... + /,618GO V28 + , In(xh lm ) + y, In(hsizeh /hsizemod) + E where i denotes the i-th food item (i=l,...,28), and h the h-th household (h=l,...,13,641). GOV], ..., GOV28 are dummies for the 28 urban/rural govemorates in Yemen, xh is the h-th household per capita monthly expenditure, x,,d is the median household expenditure, hsizeh is the household size, hsizeed is the median household size, and E is the error term.5 The coefficients on the dumnmy variables give the quality-adjusted unit values at median expenditure and household size. Table 2.2 shows the food poverty lines by govemorate and urban/rural area. 4See Deaton (1988), Ravalion and Chen (1996). In 1998 the CSO conducted a national price survey, providing retail prices broken down by urban/rural governorates. This infornation has not been used to estimate the food poverty lines. Among the reasons for preferring unit values to the pnce survey data, is the discrepancy between the aggregation scheme of the Price Survey and the 1998 food bundle. 5Because of small sample size problems, the governorates of Al-Jawf and Mareb were aggregated, as well as were Hadramout, Al-Mahrah and Shabwah. S Annex 2 Setting Allowances for Non-Food Goods Food poverty lines (FPL) provide a measure of extreme poverty, in that individuals classified as 'poor' according to FPL cannot even afford the minimum food energy intake requirement. For this reason, an allowance for non-food basic consumption is commonly added to the FPL. In principle, one could set the allowance for non-food goods by designating a bundle of non-food goods according to the consumption pattern of a reference group (e.g., the poorest 20% of the population), and cost that bundle separately in each region and sector. However, two considerations militate against that approach in the case of non-food goods: (i) while food energy requirements are the obvious anchor for food consumption, basic non-food consumption has no analogous basis. (ii) non-food prices are difficult to estimate reliably. Following the approach outlined in Ravallion (1994), we estimated two other poverty lines which differ in terms of the allowance for non-food goods. For the overall povertv line the non-food spending is the mean spending of those households having total expenditure equal to the food poverty line, while for the upper poverty line, it is the estimated mean spending of households whose food spending equals the food poverty line. To illustrate, let us assume that food spending increases with total spending, with a slope less than unity. Let us also assume that there is a unique expenditure level needed to reach nutritional requirements, as indicated in Figure 2.1. This is the food poverty line, ZF. Among those households that can afford to reach their nutritional requirements, the lowest level of non-food spending is given by the distance NF in Figure 2.1, all of which displaces basic food spending. NF constitutes the minimum allowance for non- food spending one can think of It is always a good idea to consider more than one poverty line. An alternative approach to estimate the non-food allowance is to ask: what is the level of non-food spending found among those who actually reach the food poverty line, rather than those who can merely afford to do so if they cut all non-food spending? This allowance is given by N F in Figure 2.1, and can be considered the maximum reasonable amount for basic non-food needs, assuming that those who reach their food requirements will also have reached their basic non-food needs. Total Poverty Lines (Including Food and Non-Food) For the total poverty line, the non-food spending is the estimated mean for those whose total expenditure equals the food poverty line, while for the upper line, it is the estimated mean for those whose food spending equals the food poverty line. These means were estimated both parametrically and non-parametrically, as a check on robustness. The parametric estimator used a regression of the food budget share on log total spending relative to the food poverty line and its squared value: wh =a +±,lIn(xhlZF)+/+2 I n(xh /z F ) + e where wh denotes the h-th household food budget share, Xh is household per capita expenditure, zF is the food poverty line, and e is the error term. The intercept of this regression gives the estimated food share at the point where total spending equals the food poverty line. The parametric estimate of the total poverty line is then obtained as follows: zt = (2 - a). ZF The nonparametric estimator was a locally weighted mean formed by taking means of bins of varying sizes in an interval around the food poverty line and averaging these means: ZL = ZF + r1 Zi r 2, =E{x- |x E [ZF 7r,ZF +Y]} r=1 6 Annex 2 We defined y as belonging to the set (70, 140, 210, ..., 700), so that F=Jo. There is an issue specific to Yemen on how to treat qat. Should qat be classified as a basic need? Does qat positively contribute to the welfare of individuals? Should there be a qat-allowance in the estimation of poverty lines? The poverty estimates based on the above lines include qat in both x (total expenditure, i.e. welfare indicator) and zL (i.e. poverty line). In contrast, past work on poverty in Yemen has excluded qat from non-food basic needs, though expenditure on qat has typically been included in the expenditure aggregate used to measure welfare. This practice is questionable in that (i) qat is arguably an important social need in Yemen society; (ii) it is hard to see why if one deems qat to not be a basic consumption need, one would still include it in measuring welfare. There are three different approaches one can take to deal with qat when measuring poverty. One approach is to include qat in both the poverty lines and the welfare measure. A second one consists in including qat in total expenditure, but excluding it from the non-food allowance in the poverty line. The third is to exclude it from both. The following table summarizes these three options. Welfare Indicator Poverty line Method I Includes Qat Excludes Qat Method 2 Includes Qat Includes Qat Method 3 Excludes Qat Excludes Qat In method 1, qat is not a basic need, but it is deemed to raise welfare. Methods 2 and 3 treat qat more consistently, but in very different ways. Comparison of method 1 and 3 gives the poverty impact of Xa. Such a corparison allowed us to simulate what would happen to poverty estimates if all qat spending was re-allocated to other uses (food, clothing, etc.). Corresponding to methods 1-3, we obtained as many sets of estimates for the overall food poverty lines. Tables 2.4-2.5 show the results. Upper Poverty Lines Compared to total poverty lines, upper poverty lines make a more generous allowance for the non- food expenditure component of the poverty line. Overall poverty lines (called usually lower poverty lines) incorporate a minimal allowance for non-food goods (being the typical non-food spending of those who can just afford the food requirement). Upper poverty lines give a more generous allowance for non-food goods by incorporating the typical non-food spending of those who just attain the food requirement. Accordingly, the upper poverty line was defined as follows: ZU = ZF +rXz, X IZr =E{x-XFIXF E [ZF -Y,zF + rD r=1 where XF denotes the household per capita expenditure on food. The estimation procedure is the same as used for total poverty lines. Tables 2.5-2.6 show the non-parametric estimates for upper poverty lines, for methods 1-3. Poverty Estimates On the basis of the poverty lines shown in Tables 2.4-2.7 preliminary poverty estimates were obtained. For each set of poverty lines we estimated three poverty measures: * The headcount index * The poverty gap index * The squared poverty gap index All the above indices belong to the Foster-Greer-Thorbecke class of measures (Foster et alia, 1984), whose general formula is the following: 7 Annex 2 where a is a non-negative parameter. The Headcount Index gives the proportion of the population for whom expenditure yi is less than the poverty line z is obtained by setting a=0. The Poverty Gap Index (obtained when a=l) reflects changes in the degree of poverty among the poor. The Squared Poverty Gap Index (obtained when a=2) reflects the severity of poverty and is sensitive to inequality among the poor. The main results can be summarized as follows (Tables 2.8-2.9 provide more estimates): Headcount Poverty gap Squared poverty Index (%) Index (xl00) gap Index (xloo) Yemen 38.5 12.0 5.2 Method I Urban 28.6 7.5 2.8 Rural 41.4 13.3 6.0 Yemen 41.8 ' 132:' 5'.8 Metho'd2 .Urban 30.8 l8.2 32, Rural,' 45.0' _ 14:'7 6 7 Yemen 44.7 14.1 6.2 Method 3 Urban 33.2 8.9 3.4 Rural 48.1 15.7 7.1 Notes: The table reports non-parametric estimates of poverty measures are estimated at the overall poverty lines. The variable used throughout is household per capita expenditure per month. Observations were weighted using individual expansion factors. The sum of the expansion factors across households is 15,658,163. 8 Annex 2 Table 2.1. The food bundle fTor Yemen, 1998 Commodity category Spcific item OS d Consumpton k eEnergY (2522kcal) (2,22keal) (0) (1) (2) (3) (4) (5) (6) 1 Wheat Imported wheat 102 170 578 62.1 54.1 2 Sorghum Red sorghum, domesftc PS 53 178 19.3 16.9 3 Maize Maize, India PS 20 69 7.3 6A 4 Barley Barley, domestic PS 8 27 2.9 2.5 5 Millet Millet, domestic PS 10 32 3.7 3.2 6 White Flour 112 115 403 42.0 36.6 7 Rice Rice Pakistani basmati, 2 PS 24 82 8.8 7.6 8 Legumes Broad beans, domestic PS 38 129 13.9 12.1 9 Red Meat Fresh goat, mutton, sheep 501 8 9 2.9 2.5 10 Chicken Fresh poultry 511 12 17 4.4 3.8 11 Fish Emperor, frozen fish PS 8 11 2.9 2.5 12 Eggs 712 5.8 8 2.1 1.8 13 Milk (and Milk Products) Uquid milk Yemani° PS 39 139 14.2 12A 14 Yeast PS 7 24 2 6 2.2 15 Dark-colored Vegetables Cucumbers PS 60 21 21.9 19.1 16 Green Vegetables Okra, domestic PS 50 19 18 3 15.9 17 Tomatoes 301 13.6 6 5,0 4.3 18 Potatoes 30S 12.4 9 4.5 3.9 19 Carrots PS 10.5 4 3.8 3.3 20 Papayas PS 20 6 7.3 6A 21 Watemfelon PS 20 4 7.3 6A 22 Cantaloupes PS 10 3 3.7 3.2 23 Bananas 401 20 18 7.3 6A 24 Oranges PS 12 5 4.4 3.8 25 Dates Saudi dates PS 12 34 4.4 3.8 26 Mangoes PS 8 6 2.9 2.5 27 Grapes Black grapes PS 19 13 6.9 6.0 28 Vegetable OiUGhee 'Chair' Butter, local PS 50 445 18.3 15.9 29 Sugar 901 50 200 18.3 15.9 30 SALT PS 10 0 3.7 3.2 31 Spices Dried chillies PS 2 0 0.7 0.6 32 Tea 1101 4 0.04 1.5 1.3 33 Coffee Imported coffee, grains PS 3 0.03 1.1 1.0 34 Coffee Husks Yemeni coffee husks PS 8 0.08 2.9 2.5 35 Various Juices Yemeni mango Juice PS 30 24 11.0 9.6 Total 942.3 2523.2 Source: A. Bald (1999), "National Food Basket for 1998". Notes: The food bundle used to estimate the food poverty line is reported in the column 6. Quantities in column 6 were obtained by multiplying column 2 by 365, dividing by 1,000 (from grams to kilos), and scaling down the result by the ratio between 2,220 (food energy requirement) and 2,523.2 (actual food energy intake). Column I shows the actual commodities used to price the food categories in column 0. Column 2 shows the CSO code for items in column 1, "PS" denotes that the price was estimated on the basis of the 1998 price survey. 9 Annex 2 Table 2.2. Food poverty lines by governorate, urban/rural, 1998 (Rial/person/month) I Urban/rural Governorate I Rural Urban ---------------------------+----- __________ Ibb | 2039 2073 Abyan I 1928 1921 Sana'a City | 2095 Al-Baida | 2102 2171 Taiz 2082 2095 Al-Jawf - Mareb 2090 2139 Hajjah | 2259 2058 Al-Hodeida 2008 1921 Hadramout - Al-Mahrah - Shabwah 2257 2339 Dhamar 2179 2246 Sa'adah | 2222 2246 Sana'a j 2045 2110 Aden | 2089 Laheg | 2090 2197 Al-Mahweet 1990 2104 Source: World Bank esumates. Table 2.3. Poverty estimates: food poverty lines Headcount Poverty gap Squared poverty index (%) Index (xlOO) gap index (xlO01 Yemen 17.7 TS4.5 . 9 Urban 10.0 2.1 0.7 Rural 19.9 5.2 2.0 Source: 1998 HBS, World Bank estimates. Figure 2.1. Construction of the total and upper poverty lines NO food spending . ....... .. . . . . ...... ,"~~~~~~F 45° totmi sprmng ZL Zu Source: Ravallion (1994). Note: the figure shows a stylised relationship between food spending and total spending. The overall povcrty line ZL i defined as the food poverty line (zF) plus the non-food spending of households who can just afford zp. Note tat sZL corresponds to the upper poverty line, as defined in Yemem government documentation. The upper poverty line (zu) is the total spending level at which a household actually spend ZF on food. 10 Annex 2 Table 2.4. Total poverty lines (method 2) by governorate, uirban/rural (Rlal/person/month) I Urban/rural Governorate I Rural Urban ------------ ---______ -__________+___________-___ Ibb | 3268 3223 Abyan | 2772 2706 Sana'a City I 3352 Al-Baida 3147 3132 Taiz 3268 3194 Al-Jawf - Mareb 2531 3390 Hajjah | 3212 2904 Al-Hodeida | 2954 2803 Hadramout - Al-Mahrah - Shabwah | 3271 3278 Dhamar 3424 3404 Sa'adah | 3458 3392 Sana'a 3323 3303 Aden 3315 Laheg | 3280 3306 Al-Mahweet | 3019 3054 ------------------------------------------------ Note: in method 2 (see text) poverty lines mclude an allowance for qat consumption. Table 2.5. Total poverty lines (methods I and 3) by governorate, urban/rural (Rial/person/month) ------------------------------------------------ I Urban/rural Governorate I Rural Urban --------------------------------+--------------- Ibb | 3128 3099 Abyan | 2755 2695 Sana'a City I 3211 Al-Baida | 2904 2894 Taiz | 3166 3101 Al-Jawf - Mareb 2407 3036 Hajjah 2892 2614 Al-Hodeida | 2792 2662 Hadramout - Al-Mahrah - Shabwah | 3268 3276 Dhamar 3316 3307 Sa'adah 3212 3162 Sana'a I 3065 3063 Aden | 3251 Laheg | 3160 3194 Al-Mahweet | 2917 2954 ------------------------------------------------ Note: both method I and method 3 (see text) exclude qat from the non-food allowance. The two methods differ in the definition of total expenditure, but share the same poverty lines. 11 Annex2 Table 2.6. Upper poverty lines (method 2) by governorate, urban/rural, 1998 (RlaVperson/month) ------------------------------------------------ I Urban/rural Governorate I Rural Urban --------------------------------+--------------- Ibb | 5016 4897 Abyan | 3528 3423 Sana'a City I 5529 Al-Baida 4949 4831 Taiz 5237 5078 Al-Jawf - Mareb 4733 4605 Hajjah 4962 4571 Al-Hodeida 4126 3929 Hadramout - Al-Mahrah - Shabwah | 3921 3868 Dhamar 5121 5024 Sa'adah 4776 4622 Sana'a | 4671 4562 Aden 4778 Laheg | 4591 4534 Al-Mahweet 3533 3495 Note: in method 2 (see text) poverty lines include an allowance for qat consumption. Table 2.7. Upper poverty lines (methods 1 and 3) by governorate, urban/rural, 1998 (RialUperson/month) I Urban/rural Governorate I Rural Urban -------- ------------------------+--------------- Ibb | 4497 4411 Abyan | 3408 3313 Sana'a City I 4958 Al-Baida 4197 4124 Taiz 4851 4716 Al-Jawf - Mareb 4297 4201 Hajjah | 4284 3941 Al-Hodeida | 3694 3533 Hadramout - Al-Mahrah - Shabwah 3887 3842 Dhamar 4727 4654 Sa'adah | 4247 4120 Sana'a 4058 3988 Aden 4564 Laheg 4230 4206 Al-Mahweet 3317 3300 Note: both method I and method 3 (see text) exclude qat from the non-food allowance. The two methods differ in the definition of total expenditure, but share the same poverty lines. 12 Annex 2 Table 2.8. Poverty estimates: overall poverty lines Headcount Poverty gap Squared Index index poverty gap (%) (x 100) index (x 100) Yemen 38.5 12.0 5.2 Non-parametric Urban 28.6 7.5 2.8 Rural 41.4 13.3 6.0 Method 1 Yemen 34.4 10.1 4.3 Parametric Urban 29.2 7.7 2.9 Rural 35.9 10.9 4.7 Yemen 41.8 13.2 5.8 Non-parametric Urban 30.8 8.2 3.2 Rural 45.0 14.7 6.7 Method 2 Yemen 35.8 10.7 4.5 Parametric Urban 30.0 8.0 3.0 Rural 37.5 11.6 5.0 Yemen 44.7 14.1 6.2 Non-parametric Urban 33.2 8.9 3.4 Method 3 Rural 48.1 15.7 7.1 Yemen 40.2 12.1 5.1 Parametric Urban 33.8 92 3.5 Rural 42.1 13.0 5.6 Note: methods 1, 2 and 3 are defined in the text. Parametric estimate is obtained by zP = (2- &)* ZF Nonparametric estimate is calculated by: ZL ZF +r Z_ 7 i r = E{-xf |x E [ZF -.Y, ZF + r} where r=1 y belongs to the set (70. 140. 210, .., 700), so that r=1o. Table 2.9. Poverty estimates: uper poverty lnes Headcount Poverty gap Squared poverty _ ___________ Index (%) index (xlOo) gap index (xlo1) Yemen 60.4 22.8- , Method 1 Urban 51.3 16.7 7.4 Rural 63.2 24.7 12.5 Yemen 66.9 26.6 ' Method 2 Urban 57.8 19.8 9.1 Rural 69.6 _ 28.7 IS I Yemen 66.5 - -26JN - : Method 3 Urban 57.5 19.6 8.8 Rural 69.2 28.1 14.6 Notes: Non-parametric estimates only. Note: methods 1, 2 and 3 are defined in the text 13 Annex 2 Table 2.10. Overall poverty lines (method 1) by governorate Governorate I HEADCOUNT POVGAP POVGAP2 ___---- ____---- --- -- --- --_ _ +--- - -_-_ -___ -__-_ Ibb j 0.53731 0.21996 0.11582 Abyan 0.53235 0.15670 0.06161 Sana'a City | 0.20633 0.05315 0.01997 Al-Baida 0.12401 0.03015 0.01075 Taiz | 0.53142 0.18974 0.08881 Al-Jawf - Mareb 0.24058 0.05696 0.02161 Hajjah | 0.23646 0.05065 0.01836 Al-Hodeida 0.34612 0.09481 0.03490 Hadramout - Al-Mahrah - Shabwah 0.42638 0.12990 0.05619 Dhamar 0.45176 0.13666 0.05503 Sa'adah 0.20393 0.03513 0.00933 Sana'a 0.34840 0.08752 0.03487 Aden | 0.28818 0.06490 0.02199 Laheg | 0.49425 0.15934 0.06557 Al-Mahweet 0.28386 0.06951 0.02366 Table 2.11. Total poverty lines (method 2) by governorate Governorate I HEADCOUNT POVGAP POVGAP2 Ibb 0.55500 0.23375 0.12505 Abyan | 0.53371 0.15891 0.06273 Sana'a City 0.22884 0.06006 0.02300 Al-Baida | 0.15425 0.03871 0.01425 Taiz 0.55584 0.20081 0.09533 Al-Jawf - Mareb 0.26292 0.06718 0.02561 Hajjah | 0.30322 0.07215 0.02652 Al-Hodeida 0.39840 0.10987 0.04198 Hadramout - Al-Mahrah - Shabwah 0.42638 0.13018 0.05633 Dhamar 0.48510 0.14722 0.06038 Sa'adah I 0.26798 0.04956 0.01382 Sana'a 0.40503 0.10998 0.04440 Aden j 0.30192 0.06932 0.02372 Laheg 0.52092 0.17211 0.07276 Al-Mahweet j 0.29209 0.07680 0.02694 Table 2.12. Total poverty lines (method 3) by governorate Governorate I HEADCOUNT POVGAP POVGAP2 _--_-- __--- ---- -- -- -- --- - +------_-__ -____ -_ -____ - _ Ibb | 0.59319 0.24956 0.13504 Abyan - 0.55835 0.16661 0.06661 Sana'a City | 0.26025 0.06833 0.02599 Al-Baida 0.16086 0.03720 0.01303 Taiz 0.57552 0.20711 0.09755 Al-Jawf - Mareb | 0.27886 0.07382 0.02810 Hajjah | 0.36210 0.08447 0.03030 Al-Hodeida 0.44161 0.12656 0.04951 Hadramout - Al-Mahrah - Shabwah | 0.43154 0.13166 0.05679 Dhamar 0.52513 0.15802 0.06441 Sa'adah | 0.26369 0.04761 0.01325 Sana'a I 0.43879 0.12040 0.04846 Aden 0.31384 0.07402 0.02536 Laheg 0.54404 0.17800 0.07422 Al-Mahweet 0.31435 0.07499 0.02533 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - _ - - - - 14 Annex 2 Table 2.13. Poverty estimates: total poverty lines (method 1i by governorate, urban/rural I Urban/rural Governorate I Rural Urban ------------- --------------- +--- _______- ______ Ibb j 0.56622 0.36101 0.23944 0.10115 0.12827 0.03991 Abyan | 0.57143 0.36077 0.16996 0.09848 I 0.06724 0.03687 Sana'a City | 0.20633 0.05315 I 0.01997 Al-Baida 0.11626 0.16405 0.02439 0.05986 0.00751 0.02749 Taiz 0.57241 0.33032 0.20832 0.09855 I 0.09851 0.04123 Al-Jawf - Mareb | 0.25645 0.09007 0.06241 0.00534 0.02384 0.00048 Hajah 0.24345 0.16147 I 0.05201 0.03607 0.01900 0.01148 Al-Hodelda | 0.39262 0.25522 0.11315 0.05894 0.04314 0.01878 Hadramout - Al-Mahrah - Shabwah 0.41322 0.46372 0.12907 0.13225 0.05757 0.05227 Dhamar | 0.45928 0.38875 0.14041 0.10525 0.05679 0.04031 Sa'adah | 0.20820 0.16916 0.03676 0.02191 I 0.00971 0.00617 Sana'a 0.34644 0.38444 0.08724 0.09265 0.03481 0.03594 Aden | 0.28818 0.06490 I 0.02199 Laheg | 0.49809 0.42423 0.16168 0.11662 I 0.06681 0.04300 Al-Mahweet 0.26255 0.55561 0.06149 0.17179 0.01913 0.08143 Note: the first line in each cell is the Headcount Index, line 2 is the Poverty Gap, and line 3 is the Poverty Gap Squared) 15 Annex 2 Table 2.14. Poverty estimates: total poverty lines (method 2) by governorate, urban/rural I Urban/rural Governorate I Rural Urban Ibb | 0.58297 0.38449 0.25380 0.11153 I 0.13820 0.04490 Abyan I 0.57143 0.36809 0.17241 0.09960 I 0.06851 0.03739 Sana'a City | 0.22884 0.06006 0.02300 Al-Baida 0.14741 0.18955 0.03292 0.06858 0.01065 0.03286 Taiz 0.59850 0.34657 0.22022 0.10560 0.10565 0.04474 A1-Jawf - Mareb | 0.27417 0.15620 0.07223 0.01929 0.02801 0.00279 Hajjah I 0.31349 0.19292 0.07412 0.05090 0.02735 0.01758 Al-Hodelda I 0.45086 0.29584 0.13032 0.06990 0.05155 0.02326 Hadramout - Al-Mahrah - Shabwah 0.41322 0.46372 0.12939 0.13243 I 0.05773 0.05236 Dhamar | 0.49409 0.40974 0.15122 0.11370 0.06232 0.04419 Sa'adah | 0.27625 0.20072 0.05156 0.03328 0.01442 0.00898 Sana'a 0.40480 0.40918 0.10973 0.11451 0.04434 0.04544 Aden | 0.30192 0.06932 0.02372 Laheg | 0.52527 0.44160 0.17456 0.12745 0.07410 0.04829 Al-Mahweet d 0.26956 0.57950 0.06833 0.18485 0.02217 0.08765 Note: the first line in each cell is the Headcount Index, line 2 is the Poverty Gap, and line 3 is the Poverty Gap Squared. 16 Annxex 2 T[able 2.1L5. lPovertly e5tEmntes: totdI poverRy Unaez (mnethodf 3 by governsorate, arban/ruirŁ1 I Urban/rural Governorate I Rural Urban - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -_ - -_ _ - - - - -_ _ - - - - - - Ibb |0.62222 0.41623 0.27022 0.12354 0.14903 0.04975 Abyan 0.60168 0.36809 0.18139 0.10172 0.07322 0.03756 Sana' a City |0.26025 0.06833 0.02599 Al-Baida |0.14959 0.21905 0.03051 0.07172 0.00923 0.03266 Taiz |0.61829 0.36570 0.22657 0.11164 0.10782 0.04717 Al-Jawf - Mareb 0.29682 0.10852 0.08017 0.01368 0.03077 0.00277 Haj jah |0.37482 0.22553 0.08696 0.05772 0.03122 0.02038 A1 -Hodeida |0.50224 0.32308 0.15188 0.07704 0.06161 0.02585 Hadramout -Al-Mahrah - Shabwah |0.41952 0.46563 0.13118 0.13302 0.05832 0.05246 Dhamar |0.53204 0.46730 0.16216 0.12335 0.06638 0.04784 Sa 'adah |0.27271 0.19034 0.04913 0.03524 0.01360 0.01037 Sana'a 0.43801 0.45317 0.11972 0.13291 0.04818 0.05357 Aden |0.31384 0.07402 0.02536 Laheg 0.54616 0.50538 0.18020 0.13802 0.07546 0.05179 Al-Mahweet 0.29058 0.61750 0.06585 0.19165 0.02023 0.09028 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Note: the first line in each cell iS the Il[endcount Landen, line 2 is the Poverty Gap, and line 3 is the PoverVl Gap Squnired. ANNEX 3 ECONOMIES OF SCALE By using per capita expenditure as an indicator of well being, no allowance is made for economies of scale in the household, nor for differences in needs between household members.6 By ruling out economies of scale it is assumed that the needs of a family of eight are exactly twice the needs of a family of four. In contrast, by allowing for some degree of economies of scale a family of eight having the same per capita income as a family of four would be judged better off than the family of four. The hypothesis is that large households may have a distinct advantage over smaller ones as they can benefit from sharing conmmodities or purchasing products in bulk at a discounted rate. Size economies in household consumption is likely to be an important phenomenon in Yemen, where both the average household size (7.09 members nationally, 7.11 in rural areas, 7.08 in urban areas) and the range of the sizes are remarkably high (see Table 3.1). This annex provides an assessment of (i) the extent to which the existence of size economies affects the conclusion that larger families tend to be poorer in the case of Yemen, and (ii) the impact of size economies on poverty measures in 1998. The analysis is based on the 1998 HBS. Table 3.1. Distribution of households by size, 1998 Household size Urban Rural Yemen (%) (%) 1 2.0 0.8 2.8 2 4.4 1.2 5.6 3 5.2 1.6 6.8 4 7.4 2.1 9.5 5 7.9 2.4 10.3 6 8.7 2.6 11.3 7 9.4 2.7 12.1 8 8.7 2.5 11.1 9 7.0 2.0 9.1 10 4.9 1.7 6.6 11 3.7 1.1 4.8 12 3.7 1.2 4.9 13 0.9 0.3 1.1 14 0.9 0.2 1.2 15 or more 2.2 0.7 2.9 Total 77.0 23.0 100.0 Source: World bank calculation on the basis of the HBS 1998. There is no single agreed methodology for the estimation of economies of scale. One way of assessing their importance is to assume that the needs for each good do not expand with the number of people in the household, but less rapidly, for example in proportion to n raised to the power of 0 (0<0 :), where n denotes the household size, and 0 measures the extent of economies of scale. Accordingly, household expenditures (E) were transformed as follows: Ee=E/(n6) If 0=1 there are no economies of scale, and each person gets an nth of the total expenditure; for 0<0<1, there are economies of scale, and each person receives more that his/her share of the total. To investigate the relevance of economies of scale to poverty in Yemen we transformed the poverty lines as follows: PL8=PL/(7.0901) where PL denotes the poverty lines and 7.09 is the average household size 6 Note that economies of scale are independent from the age structure of the household and thus quite distinct from adult equivalency scales which dernve from differng needs of different household members. It may be useful to investigate the effects of equivalence scales on poverty outcomes. To this end it is recommended that Yemen develop a country specific equivalence scale. 18 Annex 3 for Yemen in 1998.7 Figure 3.1 shows the distribution of poverty risk by household size for three economies of scale (EOS) regimes: (i) no EOS corresponding to 0=1, (ii) large EOS, corresponding to 9=0.6, and (iii) snmll EOS corresponding to 0=0.8. As expected, adjusting for econornies of scale has a flattening impact on the poverty risk -household size curve: once economies of scale are accounted for, the probability of being poor increases with household size up to the medium household size, but less rapidly than in the absence of size econornies. For households sized 10 or more the poverty risk trend tend to reverse (see figure 3.1). Figure 3.1. Econoimies of scale, 1998 1.40 1.20- 1 100 -- - ~ 0.80 o0.60 ~-0.40- 0.20 0.00 I I I I I 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Household slzo -no EOS -large EOS - small EOS Note. EOS is for economies of scale. "No EOS" coresponds to 0=1, 'large EOS" to 8=0.6, "small EOS" to 8=0.8. Source: World Bank estimates based on 1998 HBS. Table 3.2 shows the extent to which the allowance for economies of scale impacts on poverty measures. Allowing for economies of scale reduces poverty incidence, depth and severity. This happens because the allowance for econoniies of scale corrects, at least partially, the over-estiration of the negative impact on poverty of the large number of children and infants. In other words, larger households, which usually include many children, are likely to benefit most from economies of scale. Thus, measuring the living standard by neglecting economnies of scale leads to overstate the number of large households that are poor. Taking this effect into account would reduce the estimate of poverty incidence in Yemen from 41.8% (no EOS) to either 3 8.8% (corresponding to 0=0.8) or 36.8% (corresponding to 0=0.6). Table 3.2. Impact of economies of scale on poverty measures, 1998 Headcount index IPoverty Gap Squared Poverty Gap 8=1 (no economies of scale) 41.8 13.2 5.8 0=0.8 38.8 12.1 5.3 0=0.6 36.8 11.4 4.9 7See P. Lanjouw and M. Ravallion (1995), "Poverty and household size", The Economic Journal, 105, 1415-34. ANNEX4 HOW TO IUPROVE THE QUALITY OF THE 2003 HOUSEHOLD BUDGET SURVEY? Accurate and up-to-date data from household surveys are essential for governments to make effective anti-poverty policies. As discussed elsewhere in this report, the Yemeni 1988 Household Budget Survey (HBS) provides reliable information on household consumption expenditure, which constitutes the single most important welfare indicator for both identifying the poor and tracking poverty over tume. However, to improve upon poverty measurement, monitoring and targeting in Yemen, additional information is needed. Non-monetary measures of poverty are often used to complement and supplement income-based poverty measures, and in this respect the information collected by the HBS- 98 is largely incomplete. In particular, social indicators, which identify access to and effectiveness of social services cannot be adequately estimated on the basis of the HBS-98. Looking ahead, specially to the HBS scheduled for the year 2003, this Annex provides a set of key suggestions aimed at improving the statistical base for poverty analysis in the years to come. The main recommendations for improving the HBS-03 are as follows: * As a general recommendation, the HBS-03 should be designed in such a way as to both (i) mninimize problems of comparability with respect to the HBS-98, and (ii) significantly enlarge the scope of the HBS-03 questionnaire. Every effort should be made to accomplish the above tasks simultaneously. The first task can be achieved by adopting the same sample design used for the HBS-98, as well as by defining the main variables of interest as in the HBS-98 questionnaire. Enlarging the number of topics covered by the HBS-03 can be achieved by adding a selection of new questions to the questionnaire, along the lines of the World Bank's multi-topic Living Standards Measurement Study (LSMS).8 * Given the binding time constraints for the HBS-03, it is worth suggesting a selection of essential information which should be collected by the HBS-03. In order of priorities: 1. Regarding household members' incomes, the HBS-03 should collect information on the receipt of money or in-kind assistance from key government or NGO programs (e.g., cash welfare payments, unemployment insurance, food stamps, food rations, school feeding programs, scholarships, free or subsidized textbooks). 2. Price information should be collected at the community level (the primary sampling unit) for a selection of at least 30-40 of the most commonly consumed food items, and 10-20 of the most commonly purchased nonfood items. 3. In addition to information on durable goods and housing, the HBS-03 should collect information on basic household assets such as land and capital equipment used for agncultural activities and non agricultural enterprises. 4. The health data collected in the HBS-98 are of limited usefulness for policy analysis. The HBS-03 should collect information on (i) all visits made by household members to medical facilities during the reference period, (ii) immunization status of children 0 - 5 years old, (iii) insurance coverage, and (iv) knowledge of health providers. 5. The HBS-98 does not allow the assessment of the nutritional status of the Yemeni population, since it lacks an anthropometry module. The 2003 survey should collect data on (i) weight, and (ii) height for all household members (or, at least, for children 0 - 5 years old). 6. Additional information should be collected on the use of social services and programs: beside access to government health facilities and schools, the HBS-03 should collect data on use of agricultural extension services and social assistance programs. 7. The collection of data on employment and labor force participation can be improved by collecting data on (i) whether household members are looking for work, (ii) hours of work during the previous seven days, (iii) average hourly wage earnings (if employee), (iv) non See M. Grosh and P. Glewwe (eds.) (2000), Designing Household Survey Questionnaires for Developing Countries - Lessonsfrom 15 Years of the Living Standards Measurement Study, World Bank, Washington. 20 Annex 4 wage benefits (such as paid sick leave, and pensions), and (v) main characteristics of workers' employer. Additional recommendations are as follows: o An effective improvement on the information relevant to the many aspects of living standards requires the involvement of a team of experts which should include local and international researchers and policy analysts, policymakers and staff from the CSO. We strongly recommend that such a team be set up as a "permanent poverty committee". This would help build and strengthen the capacity for poverty analysis in Yemen. o The CSO should provide adequate documentation accompanying the HBS-03, containing the details of the sample and survey design. In order to encourage the international conmmunity to use the Yemeni data in their policy analysis, it could be advantageous to Yemen to make the HBS-03 documentation available in English. • Statistical abstracts should emphasize any problems that may have arisen during a survey, in order to allow analysts to devise adequate solutions to the survey shortcomings. Q The CSO should reduce the time lag between collection and dissemination of the data. We believe that a sensible length of time to make the data available is between 6 and 12 months after collection. The dissemination of the data should consists in the release of a CD-ROM in which the cleaned datasets are stored together with the necessary documentation. We suggest that the datasets be provided in both ASCII format and in at least one of the formats used in popular statistical packages such as SPSS and Stata. The datasets should be accompanied by an exhaustive documentation (in electronic format) describing all aspects of the data (labels, codes, format, range of admissible values, etc.). In particular, the values taken on by the categorical variables should be fully explained. ANNEX 5 NATIONAL ACCOUNTS DATA A. National Accounts Statistics in Yemen The National Accounts Division of the Central Statistics Organization (CSO) compiles national accounts statistics in Yemen. The estimates are based on the concepts and classifications of the System of National Accounts 1968. The CSO produces and disseminates the following data on an annual basis covering the calendar year: (i) GDP at current market prices by 11 major industrial activities and 17 sub-activities; (iii) GDP at current prices by the following expenditure categories (private and public final consumption expenditures, gross capital fonnation (no distinction between private and government components), changes in inventory for the total economy, exports and imports; (iii) GDP at constant prices (base year 1990) by 11 major industrial activities and 17 sub- activities without expenditure categories. In addition, data are also produced and dissemnnated on gross national income, gross disposable income and domestic and national saving. GDP estumates are also complied annually for the whole economy using the production approach at current and constant prices following the same industrial classification. The definition and concepts of the Yemen national accounts follow primarily the 1968 SNA, but a number of features of the 1993 SNA are being implemented whenever possible. Value-added by industrial activity is calculated on gross production less intermediate consummation. GDP by expenditure category is calculated as the sum of the final use of goods and services through final consumption, gross capital formation and exports less imports. Industrial activity is classified and estimates published according to the International Standard Industrial Classification (ISIC of all Economrc Activities, Revision 3) except for services and government activities. The 15 ISIC categories are used with sub-activities as follows: agriculture, hunting and forestry (excluding Qat); fishing; other agriculture, hunting and forestry (including Qat); mining and quarrying; crude oil and gas; other mining and quarrying; manufacturing, oil refining; other manufacturing; electricity, gas and water supply; buildings and construction; wholesale and retile trade; hotels and restaurants; transport, storage and commnunication; financial intermediation; real estate and business services; community, social and personal services; producers of government services, and private non-profit intuitions servmg households. B. Sources of Data on National Accounts Used in this Report This Report utilizes mainly the new set of revised national accounts prepared by the Central Statistical Organization (CSO). The new set of national accounts, which was released in September 2000 (again updated in March 2001 and March 2002), updated national accounts in the old set, which was released in October 1999. The major differences have been: (i) an increase in the estimates of construction activity by an annual average of 8%; (ii) addition of maintenance to the national accounts and higher estimates for restaurant and hotels (as a result wholesale and retail trade estimates mcreased); (iii) slight revision in the estimates for transport, communication and storage in 1997 and 1998; (iv) major revision in the estimation of financial institutions and real estate; (v) reduction in the estimates of community and social services; (vi) major changes in the estimation of government services particularly in the late 1990s; (vii) and higher estimates for private non-profit services. As a result of these revisions, real GDP estimates were increased by about 19% on average (and higher in the late 1990s). Most of the revisions resulted in higher estimates of non-oil GDP. Thelable below provides annual percentage changes in the estimates of national accounts between the new set and the old set. 22 Annex 5 Table 5.2. Percentage Changes betweem New and Old Sets of NationaE Accounts Item 1990 1991 1992 1993 1994 1995 1996 1997 1998 Construction 4% 6% 7% 6% 8% 7% 13% 10% 11% Wholesale and Retail Trade 14% 14% 14% 14% 17% 15% 15% 14% 15% Restaurants and Hotels 44% 51% 50% 51% 53% 53% 53% 46% 47% Maintenance N/A N/A N/A N/A N/A N/A N/A N/A N/A Transport, Storage & 0% 0% 0% 0% 0% 0% 0% 2% 12% Conmnumcations Financial Institutions & Real 51% 50% 53% 54% 51% 53% 62% 62% 52% Estate Real Estate& Business 111% 104% 101% 100% 97% 93% 95% 93% 92% Services Community Social & Personal -32% -29% -31% -31% -32% -34% -30% -32% -30% services Govermnent Services 0% 26% 52% 84% 156% 233% 316% 325% 307% Private Non -Profit services 105% 160% 115% 300% 453% 188% 397% 416% 463% GDP at Market Prices 4% 8% 11% 26% 22% 26% 29% 29% 29% Non -Oil GDP _ 4% 9% 13% 18% 27% 32% 37% 37% 37% On the demand side, the revisions in the national accounts provides for higher estimnates in the final consumption and expenditures, in particular in private consumption. Gross domestic investmnents were also revised upwards m the late 1990s. C. Reliability of the Data Although significant progress has been made in revising national accounts in Yemen, there is still a long way to go until the country adopts the 1993 SNA. The current system suffers from a number of deficiencies in terms of coverage, estimation, reconciliation, compilation and classification. Currently, the national accounts cover the whole territory of Yemen and cover, in principle, the economic activities of all Yemen residents in conformity of the 1968 SNA. However, no adjustment is made to impute the value of infornal activities. With regard to transaction coverage, at present estimates are limited to value added by economic activity in current and constant prices and expenditure in current prices for the total economy. Estimates are produced for GNP, gross disposable income, gross domestic saving and gross national saving. Fixed capital formation includes construction, machinery and equipment. There are no separate accounts showing instructional or consolidated transactions on income and outlay and capital finance. Most transactions are recorded on the accrual basis in conforrnity with the 1968 SNA but government and financial sector transactions are largely on a cash basis. Output is valued at producers' prices and domestic uses are valued at purchasers' prices. Imports are valued c.i.f and exports are valued f.o.b. GDP by production approach is based on a variety of sources, censuses surveys and administrative records that are available annually or at some periodic intervals. Ad hoc data sources through surveys or special studies are also used as benchmark estimate or as basis for deriving estimation parameters for indirect estimation. The CSO has an ambitious program to address the current deficiencies in the system of national accounts A project for improving national accounts of Yemen (during 1999-2003) was developed in collaboration with the EIF Statistics Departments. The Dutch government is also providing assistance in data sources and training though the project of "Strengthening the Institutional Structure and Capacity of the CSO". In the short-term, the CSO hope to rebase the CPI using the 1998 HBS weights and revise the coverage of the index, revise the national accounts and rebase them to 1998 or a more recent year, improve the coverage of GDP, initiate work on intuitional sector accounts and use international classification, focusing on the household, government and financial corporation, enhance computerization and develop constant estimates for GDP by expenditures. In the medium term, the CSO plan to develop a producer price index and foreign trade indices, improve the methodology to 23 Annex 5 estimate gross capital fornation and restructure production sector according to ISIC Revision 3 as recommended in the 1993 SNA, conduct new surveys to enhance use of direct methods for GDP estimation, and start the implementation of the 1993 SNA. D. Divergence In the Estimates of Private Consumption As an example for the reliability of the national accounts, there has been a big difference in the estimates for private consumption expenditures in the national accounts and the 1998 HBS. In Yemen's national accounts, private consumption expenditure consists of the value of final consumption expenditures on goods ad services by households and private non-profit institutions serving households. Private consumption is first estimated as the residual taking GDP by production as control total, separate direct estimates are made using data from the 1992 household budget survey. Furthermore, as private consumption is determined as a residual in the current and constant price accounts; the estimates from the production and expenditures approaches are not independently determined and reconciled by definition. Therefore, any underestimation in the national accounts will undoubtedly be reflected in low private consumption expenditures levels. Taking 1998 for comparison, the HBS reveals that per capita expenditures in Yemen were about YR 4,436 (YR 5,396 for urban population and YR 4,148 for rural population). Estimates of the national accounts for the same year show per capita private consumption expenditure about YR 3,201. There could be different reasons for the variation but the underestimation of national accounts for GDP (e.g., due to exclusion of informal activities) would be one of the most important factors. However, more investigation is required for understanding these discrepancies. ANNEX 6 POVERTY INCIDENCE FORECASTS This section provides the results of a set of simulations aimed at forecasting poverty measures in Yemen for three benchmark years, 2001, 2005 and 2015, on the basis of: (i) the 1998 household budget survey (HBS) (ii) projections of the value added for Yemen 2001-2015 (iii) projections of the population growth for 2001-2015 The first part of this annex presents the results for years 2001 and 2005. Forecasts for 2015, estimated to allow an assessment with respect to the Millennium Development Goals, are presented in the second part of this annex. Poverty forecasts for 2001 and 2005. Table 6.1 shows the projections for the GDP growth rates by activity sector. Table 6.1. National Accounts in constant prices (% growth rates) 1999 2000 2001 2002 2003 2004 2005 1 Agriculture & Forestry 0.31 6.60 2.95 3.88 4.63 5.53 6.10 2 Fishing -3.44 8.62 4.93 5.87 6.62 7.84 9.12 3 Mining and Quarrying 7.82 7.21 -0.06 -0.39 -7.60 -2.62 -4.97 4 Non-refining related Manufacturing 0.24 6.45 2.95 3.58 3.93 4.83 5.60 5 OilRefining 1.16 1.15 6.05 7.00 13.99 5.95 9.13 6 Electricity,Water and Gas 687 6.13 3.94 4.87 5.63 6.83 8.11 7 Construction 0.41 7.86 3.44 4.38 4.93 5.83 6.61 8 Wholesale and Retail Trade 2.43 4.68 3.03 3.72 4.30 5.17 6.03 9 Transport, Storage & Communicatons 1.69 0.68 6.54 7.73 9.11 11.18 9.82 10 Financial Institutions & Real Estate 12.42 4.46 3.55 4.37 5.00 6.00 6.92 11 Community Social & Personal serv. 4.50 4.82 5.29 5.84 4.85 7.75 7.51 12 Government Services 5.57 5.52 3.83 5.60 4.83 4.62 4.39 GDPAtMarketPrices 3.70 5.14 3.30 4.10 3.58 5.20 5.42 Non -Oil GDP 2.90 4.71 4.00 5.00 5.70 6.50 7.00 Source: Govemment PRSP projections for 2002-2005. In order to obtain GDP per capita growth rates, two set of estimates are available for population growth rates: (i) IMF projections, (ii) Government projections Consequently, the simulations for assessing the impact of growth on poverty were carried out under two population growth regimes, corresponding to the alternative estimates of the population growth rates (see Table 6.2). 25 Annex 6 Table 6.2. Population growth rates (%) 1999 2000 2001 2002 2003 2004 2005 IMF estimates 3.6 3.3 3.1 3.0 2.9 2.9 2.8 Govermment estimates 3.6 3.5 3.4 3.3 3.2 3.1 3.0 Source: Government and IMF. Table 6.3. Impact of growth on pover: Yemen 2001 and 2005 v2001 r 2005i Urban Rural Yemen Urban Rural Yemen v-. -: iF Population growth rateis Headcount ratio 29.3 44.6 41.0 23.5 38.3 35.0 PG 7.8 14.6 13.0 5.9 11.9 10.5 PG2 2.9 6.6 5.7 2.1 5.2 4.5 ..l '1,:! i' .j: <. GbVinmentpopuladongroth rates-. Headcount ratio 29.6 45.0 41.5 24.4 39.3 35.9 PG 7.9 14.7 13.1 6.1 12.3 10.9 PG2 3.0 6.6 5.8 2.2 5.4 4.7 Source: WB estimates. Note: the poverty line used is the "lower poverty line", as defined in the report. The results shown in Table 6.3 can be summarized as follows: (i) Between 1998 and 2001 poverty measures do not show major changes. Under the IMF estimates of the population growth rates, poverty incidence at the national level decreases from 41.8% in 1998 to 41.0% in 2001 (corresponding to a 0.26% reduction per year) . Using government estimates poverty hardly changes (from 41.8% in 1998 to 41.5% in 2001, corresponding to 0.1% points per years). (ii) The rate of poverty reduction between 2002 and 2005 is substantial (see Table 3a in the Annex). During this period, poverty incidence at the national level decreases by 1.2-1.3% per year. (iii) Poverty incidence is estimated to decrease from 41.8% in 1998 to 35.0-35.9% (depending on the population growth scenario), which corresponds to an average reduction of 0.9% points per year. The analysis of the yearly forecasts shows that the trend in poverty reduction is not linear: more than one third of the estimated poverty incidence reduction during 1998-2005 is due to the reduction in poverty estimated to take place between 2004 and 2005. (iv) The above results suggest that the positive effects of economic growth on poverty measures is negatively influenced by high population growth rates, such as those used in Table 6.3. This indicates the need to consider policies aimed at affecting fertility decisions in order to control the population dynamics. (v) Table 6.4 shows that the reduction in poverty incidence is more pronounced in urban areas than in rural areas: between 1998 and 2001 urban poverty incidence decreases by 4-5%, as compared to less than 1% in rural areas. This holds true for the period 1998-2005, during which poverty incidence in urban areas decreases by 21-24%, as compared to 13-15% in rural areas. This results shows that the pattern of growth (i) will benefit more-than-proportionally urban areas where only 13% of the poor lived in 1998, as opposed to rural areas (according to the 1998 HBS, 83% of the poor were concentrated in urban areas), and (ii) will benefit the relatively richer group of the poor (in 1998, on average, the per capita expenditures (PCE) of urban poor was YR 2,337 per capita per month, as compared to YR 2,170 in rural areas). 26 Annex 6 Table 6.4. Impact of growth on poverty (% -C/G _ , 1998-200 s -2005 Urban Rural Yemen Urban Rural Yemen IMF scenario Headcount ratio -4.9 -0.9 -1.9 -23.7 -14.9 -16.3 PG -4.9 -0.7 -1.5 -28.0 -19.0 -20.5 PG2 -9.4 -1.5 -1.7 -34.4 -22.4 -22.4 Government scenano Headcount ratio -3.9 0.0 -0.7 -20.8 -12.7 -14.1 PG -3.7 0.0 -0 8 -25.6 -16.3 -17 4 PG2 -6 3 -1.5 0.0 -31.3 -19.4 -19.0 Source: WB estimates. (vi) The pro-urban bias shown by the poverty incidence forecasts holds true also for the poverty gap (PG) and poverty gap squared (PG2) measures. In fact, the latter shows the highest degree of urban-bias: during 1998-2001, under the IM population growth regime, PG2 decreases by 9.4% in urban areas, as opposed to 1.5% in rural areas. This implies a reduction in the severity of urban poverty which is more than 6 times as much the reduction in the rural poverty severity. However, because poverty severity in rural areas is more than double than poverty severity in urban areas, the conclusion is that the pattern of economic growth projected by the Govemment will widen the gap in inequality measures between the poor in urban and rural areas. Table 6.5. Impa t of growth onU povey icidenee Ie adwcunt uo %), 1 998-2SZ§ IMF population growth rates rOVCnii:t unpaktes i growth Urban Rural Vame;l Urbam IRural Yemen 1998 (actual) 30.8 45.0 41.8 30.8 45.0 41.8 1999 30 8 46.5 42.9 30.8 46.5 42.9 2000 29.8 44.8 41.3 29.9 45 1 41.6 2001 29.3 44.6 41.0 29.6 45.0 41.5 2002 28.0 43.7 40.1 28.6 44.2 40.6 2003 27.0 42 7 39.0 27.7 43.3 39.7 2004 25.4 40.9 37.3 26.1 41.6 38.0 2005 23.5 38.4 35.0 24.4 39.3 35.9 Source: WB estimates. Note: the poverty line used is the "lower poverty line", as defined in the report. Table 6.6. Impact of growth an poverty g n?G), 202982G.§ IMF population growth rates GovcxnUeat o'ulatioa growth rates Urban Rural Yemern Urhbnn LEinral Yemen 1998 (actual) 8.2 14.7 13.2 8.2 14.7 13.2 1999 8.3 15.5 13.8 8.3 15.5 13.8 2000 7.9 14.6 13.1 7.9 14.7 13.1 2001 7.8 14.6 13.0 7.9 14.7 13.1 2002 7.4 14.2 12.6 7.5 14.4 12.8 2003 7.0 13.6 12.1 7.2 13.9 12.4 2004 6 4 12.8 13 6.7 13.2 11.7 2005 5.9 11.9 10.5 6.1 12.3 10.9 Source. WB estimates. Note: the poverty line used is the "lower poverty line", as defined in the report. 27 Annex 6 Table 6.7. Impact of growth on poverty gap squared (PG2), 1998-2005 RAF populaton growth rates Government population growth rates Urban Rural Yemen Urban Rural Yemen 1998 (actual) 3.2 6.7 5.8 3.2 6.7 5.8 1999 3.2 7.1 6.2 3.2 7.1 6.2 2000 3.0 6.6 5.8 3.0 6.6 5.8 2001 2.9 6.6 5.7 3.0 6.6 5.8 2002 2.8 6.4 5.5 2.9 6.5 5.7 2003 2.6 6.1 5.3 2.7 6.3 5.4 2004 2.4 5.7 4.9 2.5 5.9 5.1 2005 2.1 5.2 4.5 2.2 5.4 4.7 Source: WB estimates. Note: the poverty line used is the "lower poverty line", as defined in the report. Table 6.8. Impact of Orortb on per capita expenditure dlstrib, iion (Ginl Index, %), 1998-2005 Mff population growth rates Government population growth IMF _ population growth _ ratesrates Urban Rural Yemen Urban Rural Yemen 1998 (actual) 35.7 33.3 34.4 35.7 33.3 34.4 1999 35.8 33.4 34.6 35.8 33.4 34.6 2000 35.8 33.3 34.5 35.8 33.3 34.5 2001 35.8 33.3 34.5 35.8 33.3 34.5 2002 35.8 33.4 34.6 35.8 33.4 34.6 2003 35.8 33.4 34.6 35.8 33.4 34.6 2004 35.7 33.4 34.6 35.7 33.4 34.6 2005 35.7 33.5 34.6 35.7 33.5 34.6 Source: WB estimates. Note: the poverty line used is the "lower poverty line", as defined in the report. Poverty forecasts for 2015 The estimates reported in column "2015" (Tables 6.9 - 6.11) assume that during 1999-2005, GDP and population grew according to the government estimates. Between 2006-2015 the GDP is assumed to grow at 5% per year, and population at 2.5% per year. The poverty forecasts results for 2015 can be summarized as follows: (i) The rate of poverty reduction between 1998 and 2015 is substantial: poverty incidence at the national level decreases by almost 4% points per year (using the national poverty line, total poverty incidence drops from 41.8% in 1998 to 21.3% in 2015). The estimates in Table 6.9 appear consistent with both Target 1 of the Millennium Development Goals (i.e., halve, between 1990 and 2015, the proportion of people whose income is less than one dollar a day), and Target 2 (halve, between 1990 and 2015, the proportion of people who suffer from hunger).9 (ii) The poverty forecasts for the poverty gap (PG) and poverty gap squared (PG2) measures (Tables 6.10 and 6.11) show marked declines between 1998 and 2015, which are also consistent with Target 1 of the MDGs. 9As far as Target 2 is concerned, the HBS 1998 does not provide information on either the prevalence of underweight children (under-five years of age), or on the level of dietary energy consumption. The headcount ratio calculated using the food poverty line, however, may be expected to provide a fair approximation of the proportion of people below the minirnum level of dietary energy consunption. 28 Annex 6 Table 6.9. Poverty incidence forecasts: Yemen, 2005 and 2015 Lower Pover3 I ;1998 2005 2015 Poverty Rie PercentaRes Lower poverty line 41.75 35.86 21.34 Food poverty line 17.65 14.54 6.70 1 USD/dayPPP 10.72 8.21 3.28 2 USD/day PPP 47.04 40.81 25.22 .______,_____ ._____.____ Num ber of poor _ Lower poverty line 6930500 5952760 3542440 Food poverty line 2929900 2413640 1112200 I USD/day PPP 1779520 1362860 544480 2 USD/day PPP 7808640 6774460 4186520 Source: WB estimates. Table 6.10. Poverty gap (PG) forecasts: Yemen, 2005 znd 2015 s overty lnoe . 1998 2005 2015 Lower poverty line 13.21 10.90 5.74 Food poverty line 4.46 3.45 .1.42 I USD/day PPP 2.42 1.82 0.69 2 USD/day PPP 15.43 12.87 7.01 Source: WB estimates. Table 6.11. Poverty gap squared (PG2) forecasts: Yemen, 2005 and 2015 3"over-i ine . 198 2005 2015, Lower poverty line 5.84 4.70 2.26 Food poverty line 1.66 1.25 0.48 I USD/day PPP 0.85 0.63 0.23 2 USD/day PPP 7.02 5.69 2.81 Source: WB estimates. ANNEX 7 EDUCATION INCIDENCE ANALYSIS Brief description of the education system After the unification in 1990, the Government endorsed the Education Law in 1994 and unified the education system. This change has been followed by vanous education policy reforms for improving and expanding the system in coordination with the macro-policy reforms such as civil service reform and decentralization (See Table 7.1). Table 7.1. Selections Information on Policy Context, Education Reforms and Household Surveys, 1990-2005 Political events Nabonal and macro- Educaton reforms Higher educabon Household strategies reforms surveys 1990 Unifity of North and South 1991 End of Gulf- War 1992 .HBS-92 .DHS-92 _ 9_ _ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 1994 .CMI War Education Law-94 Expansion of public Population Constitution unified the system universites (4 new Census-94 universities) 1995 New .Reduction of subsidies Govemment for wheat, petroleum products, etc. 1996 .1st 5 Year Plan (1996- Start of decentration of 2000) education .Unity of exchange administration rates 1997 New Implementabon of the Replace of MOHER DHS-97 Govemment first education With HCU comarehensive survey _______ 1998 - - - - - - - - - - - - - - - - - - ,New Teacher Law-98 .HBS.98 1999 Reelection of Start of civil service NPS-99 President reforms Saleh 2000 Revision of a budget New curriculum for circular and Grades 1-6 2001 -New CabInet 2nd5Year Plan(2001- .Finalized uniflaton of Creaton of MOHER. Local 2005) religious insitutes to Opening the first govemment MOE schools Yemeni community law college 2002 PRSP Start of decentralization of budget to local councils 2003 HBS-2003 (planned) 2004 Populabon Census-2004 (planned) Source: Prepared by author, Nofe MOHER Dailstiy of Hiher Education and Scientric Research HCU HKgher Coundl for Universites HBS: Household Budget Survey; DHS DemograpNc Health Survey, NPS. Nabonal Poverty Phenorrena Survey 30 Annex 7 The formal unified education system in Yemen consists of 9 years of compulscry basic education, 3 years of secondary education, and 2 to 6 years of higher education (see figure 7. 1).'o Vocational and technical training" (VTT) is available after basic education and after secondary education. Provision of pre- primary education is very limited. Informal education and training is also available at literacy centers'2 and Koran schools. In 1999/2000, the Yemeni education system enrolled more than 3.8 million students, roughly equivalent to 42% of population aged 6-24. Among the total students, more than 80% are enrolled in basic education, about 10% in secondary, 4% in university, and less than 1% in vocational and technical training. Most students are enrolled in public institutions. The private provision is gradually expanding but still limited. The share in enrollments is around 1% in basic and secondary education and 8% in university education. The central government was the major fmancer with a centralized budget allocation mechanism (see figure 7.2). But no single government authority is responsible for preparation and execution of government budgets allocated to the education sector as a whole. 3 The Ministry of Education (MOE) is responsible for basic and secondary education, while the General Authority for Vocational/echnical Training (GAVIT) handled vocational training. Each public university is fiscally independent and overseen loosely by the Higher Council of Universities (HCU) chaired by the Prime Minister.'4 Households play a limited role in direct financing of public education. In principle, the government did not provide financial supports to private education. lO Children are officially entitled to start basic schooling at age 6. To promote from basic to secondary education, students have to pass a regional standard exanmnation and a national standard examination from secondary to higher education. " VTr is available after basic education at VIT centers (2- or 3-years) and teacher training institutes (3- or 4-years) VTT is also available after secondary education at V1T centers (2 or 3 years), teacher training institutes (2 years), and a few professional institutes. Pre-service teacher training is phasing out as teacher education at universities is expandirg. 12 Literacy prograrn is consisted of two programs: 2-year basic level program for a certification equivalent to completion of the first four years of basic education and I-year follow-up level program for a certification equivalent to the first 6 years of basic education. 3 Many changes are expected in 2002 due to new laws for decentralization and higher education. 14 In addition, some other agencies manage govemment investment budgets and donors' funds for construction of schools (e.g., public works). There are also several public higher education institutes budgeted and supervised by other ministries such as ministries of health and defense. However, in this analysis, we focus on expenditures for the main educational authorities only. 31 Annex 7 Figure 7.1. Education System in Current Yemen Kindergarden Basic education Secondary Education ! Higher Education -_[ 3I 11 21 31 41 5[1 71 1791212 University Al\ <\ \ . Art \ . DlplAta , \ I Science \\ Law Undergraduate Literacy Education Graduate Basic stage 1 2 Follow-up level Community colleges Vocational education & Vocational education & technical training technical training 2 year training 1 1 3 yeareducaUon I (Teacher Training * I (Higher institutes for TT¶ 3-4 Teartrlni'na 2yerbtaining & rrl 21 31~(4) 12 Koranic recital education Source Prepared by author with informabon from ministnes' officials In 2002. Notes * The duration depends on the fields of speciality. "This system was abolished in late 1990s All teacher training insitutes are no longer used for pr-semce teacher training They are used for in-service training, with a very few exceptions for female secondary graduates in a few Govemorates 77 Educabon is not formal Thus, no official years of educaton durabon is required 32 Annex 7 Figure 7.2. Overview of the Flow of Fund on Public Education, 1998* Society International Donors (e.g., ctzens, companies. oranizatons) General revenue _i Budgetaray support | (tax and non-tax) Sector specific foan and grants Central Ministry of Finance_ CentralW Mlnistry of Education General Authority for Vocational Tralining Govemorate/ District Offices Insitutes Baslc schools Univo mitsi rVTT centers Secondary schools Teacher training institutes Literacy centers Klndergartens Ouran redtal schools _ ._ _ - ~~~~fees donations Beneficlals Students (Family) donations Students abroad Felilows abroad Indirect users Society Notes: I/ Due to ongolng education reforms, some changes have been made since 1998. For example, ministry of higher educaton was reestablished in 2001 and local govemment law became in 2002. [ Independent budgetary authorities --H'. Dotted line means that the funds do not necessarily go through the authorities. 33 Annex 7 Trends and Progress in Access to Education and Literacy In Yemen enrollments have increased substantially at all levels. However, the coverage is still low, access to education services is uneven between boys and girls, and gaps between rural and urban areas is large. During the past five years (1995-2000), enrollments had increased by some 30% in basic and 50% in secondary education; higher education enrollments were more than doubled; and vocational and technical training enrollments increased by about 40%. Despite these progress, the literacy rates, which increased from 37.3% to 43.8% for adult and from 60.4% to 65.2% for youth (ages 15-24), are still low as compared to the average of low-income countries (74.9% of youth literacy rate).15 Although the opportunities for girls improved gradually, the girls' share in total enrollments in basic education was 33% in 1998, among the lowest in the world. The girls' GER in basic education (49%) was much lower than the boys' GER (90%). The GER was also lower in rural area (64%) than in urban areas (95%). Only 33% of rural girls aged 6-14 were in school as compared to 78% of urban girls and 73% of rural boys.16 The youth illiteracy rate was also extremely high for rural females (73%), compared to urban females (18%) and rural males (15%) in 1998 (see Figure 7.3). Although children from better-off families tend to have more access to school, particularly in urban areas and for secondary education, the gap between the poor and rich does not sound as large for basic-school aged children. Based on 1998 HBS, 56% of children aged 6-14 in the poorest decile households were enrolled in school compared to 67% of children in the richest decile. In urban area, the gender gap within a decile is not very large and the richer girls have better access than poor boys. However, in rural area the gender gap at all level of education is serious regardless of family welfare: the enrollment rate for girls in the richest 10% households (31%) is much lower than that for the poorest boys (67%); and the youth illiteracy rate is very high among rural females at all level of income. Figure 7.3. Enrollment and Illiteracy rate in urban and rural areas by decile, 1998 100%- . ndn 90% ne.as 80% -6 U rd 100% 70% nolfr 0 60% -Bdraute ag 0% 50% l 6ec4 fende 60% h d 30%9 -~Uit 40%-- 20% 15- 24. nEde 20%-- 10%6Uitrc 0% 1 23 456 7 89 10 15.24. Uzban houshO21 decibfnd Rural household decile Source: World Bank staff estimates based on 1998 HBS. In addition, NPS-99 shows a large regional gap among Governorates, especially among girls. The girls' enrollment rate ranges from 17% in Sa'ada to 84% in Aden. The regional gaps in youth and adult literacy rates show the similar patterns. Govemorates where more females are illiterate have lower coverage of basic education for girls. (See figure 7.4). 5 Average for 1999 from World Bank Development Indicator database (http //devdata.worldbank.org/idg/) The analysis of NPS-99 also shows the similar large gender and urban-rural gaps in enrollment rates, but overall the enrollment rates are lower than the results of HBS-98. 34 Annex 7 Figure 7.4. Enrollment rate for aged 614 (%),1999 100 90 80 70 Rld11En_r 60 5o 40 30 20 10 ,S, oD sofesOR P, RP Factors Explaining the Gaps in Access to Education Education services do not reach a large number of children, especially rural girls regardless of family welfare. The main factors explaining the low enrollment of girls in rural areas are: (i) late start of schooling; (ii) various impediments faced by rural girls to leave school; (iii) long distance from the school; and (iv) lack of single-sex school in rural areas. Late start of schooling and early leave from school. Enrollment rate reaches the peak at age 10 and starts declining at age 14 (see figure 7.5). This suggests that children tend to start schooling a few years later than the official age (age 6)17 and among basic school-aged children, younger children are more likely to be out-of-school than older children. While a quite number of out-of-school boys and urban girls would catch up later, the most of out-of-school rural girls would not. Even if rural girls were enrolled, they are more likely to leave from school earlier. The dropout rates are estimated at 6% for basic school-aged children and 13% for rural girls in 1999.18 More than 60% of dropouts did not reach grade 5. 17 HBS-98 also shows the similar pattem. DHS-92 and 97 also show the similar pattem, the share of out-of-school children is higher for younger children (aged 6-10) than elder children (aged 11-15) by around 10 percentage points. 18 The dropout rate was measured as the ratio of children who have ever been enrolled but not currently enrolled to the children who have ever been enrolled or currently enrolled. Dropout rates by grade or level of education were not estimated in this study with two reasons: (i) NPS-99 does not provide information on the grade attended in the previous year; and (ii) the population by single age needs smoothing as samples were heavily biased to a few specific ages. 35 Annex 7 Figure 7.5. Enrollment rate by age, 1999 100% 90% 80% 760% - 4-- U±bnmab 50% -R- U3n imab 40%- -X Rumlfmab 40%6_ _> RuzalmaS 30% 20% 10%. 0% 6 7 8 9 10 11 12 1314 15 16 17 18 19 20 21 Age Reasons for never enrolled in school and leaving school. Figure 7.6 shows the reasons why children had never been enrolled in 1999.19 Rural girls appear to face all impediments: supply-side issues (no school nearby, difficulty in transportation, and lack of teachers), economical issues (work to support family, not afford school expenses), and family attitudes towards girls (family does not want to send girls to school). Even for younger girls (aged 6-11), family's attitude was an important constraint, especially in rural area. While 22% of rural girls who have never been in school reported the main reason was family's attitude, 8% of their urban counterparts provided the same reason. Supply-side factors are the main impediments both for rural boys and girls. The reasons for leaving school have similar patterns. 20 Figure 7.6. Reasons for never enrolling in school (%),1999 80 70 60 50 _ l Supply-side M Economic 40 _ _ O Attitidue for gprls 30 0 Other 20 10 i - 0 Rural Urban Rural Urban Rural Urban Rural Urban femal female male male femal female male male Ages 6-11 Ages 12-14 '9 Since expenditures are not adequately measured in the 1999 NPS, the survey data do not allow the analysis by decile. 20 It should be noted that further analysis is needed to understand why children were not in school given that the majority of children did not provide a specific reason, especially among urban and younger children. 36 Annex 7 Distance from schooL The 1998 HBS shows that rural students need to walk more to access to basic school than urban students, and in rural area families appear to allow boys to walk more than girls. While only 1 % of urban students take more than 30 minutes to reach school, 10% of rural male students and 7% of rural female students do so.21 In rural area, poor students tend to walk more than the rich. Lack of single-sex school in rural areas. In Yemen, it is often claimed that single-sex school (either boys' or girls' schools) is important for families to send girls to school, especially in upper grades. Single-sex schools are widely seen in urban area, especially in secondary education: more than 90% of urban students and 35% of rural students were in single-sex schools in 1998. However, due to lack of single-sex school in rural areas, among rural female students, only 16% are in girls' schools, 10% are in boys' schools, and the rest in co-educational schools. Public Spending on Educatiom Trends in public education expenditures Based on previous public expenditure reviews, the aggregated education expenditures are at the adequate level (See figure 7.7). Govermnent expenditures on education grew from 5.1% of GDP in 1996 to 6.1% in 2000, which is high compared with most Arab countries and lower-income countries. The increase resulted mainly from rises in the wage bill. Recurrent education expenditures as a share of GDP increased from 4.6% in 1996 to 5.3% in 2000, while the share of education investment expenditures to GDP fluctuated between 0.5% and 1.5 % during this period. Figure 7.7. Government education expenditures 6 100 '4 -Cunent -4~ 0c.)o 5 U r ; ; U 90 - g Capitalexperxlkms 80 0 N X.57cMOE 3 sham r tnta1 70 ! W 2 Z MOEb sbai n h4 = 1 -6 n 4) ~~~~~~~~60 ci a iecunent c -0 0 - X MOEbshamin C6 capiEtzLiwest w 1996 1997 1998 1999 2000 Source: Author's estin ate using M O F dat (ee Tabb A4 1) However, a glance of sub-sectoral allocations shows a worrisome signal in investment expenditures. In the late 1990s, the government, on average, allocated about 85% of total education expenditures to MOE, 13% to universities, and the remaining to vocational training and research activities.22 For recurrent expenditures, the sub-sectoral distribution pattem did not change much during the period and the share of MOE was nearly 90%. However, the share of MOE in investment expenditures declined and the share of higher education and vocational/technical training increased from 34% in 1996 to 46% in 2000. Although 21 The analysis of the NPS-99 shows students have to walk longer, for example about 30% of rural male students walk for 30 mninutes and more. A large number of households answered at "30 mmnutes" or "one hour. " The validity of the data on the distance to school sounds low. Yet the analysis shows the simnilar pattern on the urban-rural difference and gender difference in rural area to findings of HBS-98. 22 There is no systematic information on MOE expenditures disaggregated for basic and secondary education or disaggregated for rural-urban areas. 37 Annex 7 some donor funds for investment in basic education are not included in this analysis,23 this sub-sectoral allocation pattern is not adequate taking into consideration the implications for future current expenditures as well as the need of provision of basic education services to millions of children. The breakdowns of expenditures within each sub-sector also suggest that Yemen needs to improve efficiency in the allocation and use of public funds.24 For MOE, key issues include: (i) the bulk of recurrent expenditures are allocated to salaries and wages and most of non-salary spending is for printing textbooks, leaving very small amounts to operations and maintenance at schools; (ii) the increase of salaries and wages reflected a large number of new hiring that did not match with teacher needs (by region, subject, and gender) as well as the wage increases by the new teacher law in 1998 that was not linked to the improved teacher deployment and performance; and (iii) the allocation of investment budgets had not been adequate or well-prioritized even to complete ongoing projects. For universities, key issues include: (i) the rapid increase in investment expenditures was not accompanied by a system-level strategy; (ii) the scholarships abroad have increased and correspond to about 14% in recurrent expenditures in 2000;25 (iii) public universities had reduced the dependency on expensive expatriate teachers but their share in the total wage bill was still substantial, 34% in 1999; and (iv) Yemeni teachers had increased gradually their salaries and wages but no clear system to monitor their performance has been introduced. Unit cost per student Table 7.2 shows the estimated unit cost as yearly recurrent spending per student by level and type of education: YR 10,973 for basic, YR 16,397 for secondary, YR 41,270 for university education, and YR 61,420 for vocational/technical training in 1998.26 For disaggregating MOE expenditures by level, the analysis assumes that MOE spent 85% of the recurrent expenditure on basic education and the remaining on secondary education on the basis of the difference in student-to-teacher ratios. Spending on literacy programs is ignored as the amount is estimated as marginal.27 For basic, secondary, and university education, this analysis excludes spending on scholarships/fellowshlps abroad which is a major expenditure item but it is not easy to say who benefit from. The analysis also excludes the benefit from capital/investment expenditures (i.e., opportunity cost of the stock of physical capital such as classrooms). 23 Donor contributions to educational authorities had increased since 1996 and amounted to 45% of education investment budgets in 1998. In addition, some other authorities such as the Social Funds for Development also receive donor funds for basic education and literacy programs. However, the government does not have a systematic data on expenditures of such authonties for the education sector. 24 See annex tables for trends in the breakdown of education expenditures by transaction item, 1994-2001. 25 The share becomes 39% if the analysis includes scholarships abroad under the MOE budget into university budgets. The study abroad program consisted of study programs under MOE budget for both undergraduate and graduate studies, while under university budgets are mainly allocated to graduate studies for prospective university teaching staff. 26 Given a difference between a fiscal year (January-December) and a school year (October - June), the analysis simply assumes that actual expenditures in 1998 were spent for students enrolled in 1997/1998. 27 The estimate is due to the scale of enrollments and inputs. There is no officially recruited teacher only for literacy education. Most are teachers from basic and secondary schools, with a monthly allowance of YR 1050 for 40 teaching hours (2 hours per day) or voluntary teachers. 38 Annex 7 Table 7.2. Uiit Cost as ?er Student aRecurrent zpeaditure, 1998 UC UC excludig sPendingSn scholarships abroads YR YR US$ % of Ratio to GDP pc Basic Basic 11,680 10 973 77 21 1.0 Secondary 17,453 16,397 116 32l 1.5 University 50,353 _41 270 j 80 3.81 Vocational 61420 433,4201191 5.6 Source: Author's calculation with data from the Government. Notes: UC: unit cost. n.a Not applicable. Exchange rate US$1=YR141.7 in 1998 (may be revised with a new macroeconomic report) GDP per capita 5 1565 YR in 1998 (may be revised with a new macroeconormic report) Vocational training covers both pre- and post-secondary training. Level of unit cost is overall adequate. To assess whether the level of unit costs is adequate, accurate understanding of the composition is required. Yet, a quick review of unit costs as a share of GDP per capita implies that Yemen allocates a fair level of public budgets per student in basic and secondary education as compared to other low- or lower-middle income countries.28 It spends 21% of GDP per capita for a student in basic education and 32% in secondary education. Although the spending increases, as expected, with the level of education, the differences are not very wide. Higher education costs about 4 times as much as basic education and vocational training costs 5.6 times. Great regional variation in unit cost. Per student current spending varies greatly among Governorates, and differences seem to remain fairly constant over time.29 In 1998, the ratio of the highest to lowest Governorate per student spending is 2.6 (Abyan to Sana'a City). As differences in personnel expenditure are large, it is apparent that they are not necessarily caused by high transportation costs due to low population density, or economies of scale possible in denser areas. While Sana'a City is at the extreme in high population density, Governorates with fairly low density (Sa'adah and Al-Baida) have very low per student spending. There is no evidence that Govemorates of higher poverty incidences (Tais, lbb, Abya, and Lahej) receive lower or higher per student spending than others. Variation in unit cost by field of university study. The unit cost is similar among public universities, except for two universities (Hadramout and Aden), where low student-teacher ratios contribute to high costs. A study also estimated that the unit cost for science fields would be four times as much as ftat for art fields due to the difference in student-teacher ratios.30 Distributon of pubeic ed&cst!om speadAng: ulnit-incieence SME2ySIS 31 Estdsazdtag benefit-incirdence Figure 7.8 shows the distnbution of public spending on education across household deciles (ten groups of households ranked by per capita monthly household expenditure). The figure suggests that public spending on education as a wholefavors the poorest households. The poorest 10% households gain 12% of the total public education subsidies, the largest share among deciles. However, the distribution pattern 20 Compared with the data available in WD1 2001 and World Bank (2001). 29 See PER (1999). 30 See higher education rationalzation study for details. 31 In the education sector, standard benefit-incidence analysis combines the unit cost of education with the infornation on enrollments across the population (households or individuals), often ranked and/or grouped by a welfare indicator. For details, see Demery (2000). 39 Annex 7 becomes neutral among the next four deciles, as the share is almost constant, around 10-1 1%. Then, the distribution pattem becomes weakly pro-poor among the top five deciles, as a richer decile tends to gain a slightly less share. The richest 10% households gain 7% of the subsidies.32 The degree of equity in the benefit-incidence differs by level of education. While public spending on basic education favors the poor, spending on universities and vocational training strongly favors the rich. The poorest 10% households receive 13.3% of the spending on basic education, while the richest receive 5.3%. The pattern is opposite in university spending, from which the poorest benefit 6.5% but the richest gain 15.1%. Spending on secondary education does not show a clear pattern. Figure 7.8. Benefit-incidence of education subsidies (%), 1998 25 20 - -4 BaSiz --w- Secarxiary Unir 10...V --*-Total 1 2 3 4 5 6 7 8 9 10 Houehol decib When the benefit-incidence is assessed in relation to the distribution of household expenditures, public education subsidies are more equitably distributed than household expenditures regardless of level of education. Even the share of the poorest decile in public higher education subsidy (6.5%) is larger than the share in total household expenditures (3.9%). In per capita terms, a household in the poorest decile, on average, gains per capita education subsidy of YR 2,995, 17.8% of the mean per capita yearly household expenditure. This is a much larger proportion than the rich, who gains at per capita subsidy of 1.6% of the per capita household expenditure (YR 2,762). However, benefit-incidences across households might mislead the judgment on equity. On average, the poor households tend to have more household members, especially younger ones. Although the poorest households gain the largest share of the education subsidy, the targeting might not be considered enough. As expected, the distribution of education subsidies across individual deciles does not favor the poorest. The poorest 10% individuals gain 9.7% of the total education spending while the richest 10% gains 10.8%. The distribution pattern is not pro-poor but modestly pro-rich as each decile receives around 10%, with a little favor towards the rich deciles. The analysis disaggregated for urban and rural areas shows that public subsidies are more reaching to the poor in rural areas. The poorest 10% individuals receive 10.5% of the total subsidies in rural areas but 9.0% in urban areas. Overall, Yemen's results are consistent to findings in previous benefit-incidence studies in other developing countries that public education spending is more equally distributed than household expenditure, but in absolute terms public spending on education is not distributed progressively but frequently regressively across individuals. To assess the degree of inequity in Yemen compared to other countries, Table 7.3 shows the poorest individual quintile's shares in the education subsidy using results 32 See annex tables for details. 40 Annex 7 of 23 studies. It is found that Yemen's result is ranked higher among the studies in targeting to the poorest. Table 7.3. Poorest quintile's share in public education spending in seRective studies Country Year of Poorest quintile's share Source data T P S H Romania 1997 24 31 26 10 Tesliuc, et. al (2000) Indonesia 1987 22 25 21 8 Van de Walle (1992) Peru 1997 21 28 16 6 WoRd Bank (2001b) South Africa 1993 20 26 19 6 Castro-Leal (1996b)/Demery (2000) Yemen 1998 ¶0 21 17 10 Yuki (2002) Macedonia, FYR 1996 19 25 13 7 World Bank (1999b) Ghana 1989 17 21 17 8 Demery, et.al. (1995) Kenya 1992 17 22 6 2 World Bank (1995a), Castro=Leal et. al (1999) Ghana 1992 16 22 15 6 Demery, et.al. (1995) Malawi 1995 16 20 9 1 Castro-Leal (1995a) Indonesia 1989 15 22 5 0 Demery (2000) from WB (1993) Cote d'lvoire 1995 14 19 5 12 Demery (2000) Morocco 1999 12 21 7 2 World Bank (2001) Vietnam 1993 12 20 5 0 World Bank (2000c) Lao 1993 12 18 7 0 World Bank (1995b) Ecuador 1998 11 24 9 1 WDR 2001 from World Bank (2000d) Malawi 1991 10 15 7 3 Castro-Leal (1996a) Madagascar 1994 8 17 2 0 Demery(2000)/World Bank (1996c) Albania 1996 ... 27 7 8 World Bank (2000b) Ghana 1998 ... 20 19 13 Canagarajah and Ye (2001)* India 1994 ... 14 ... ... Demery (2000) from Lanjouw and Ravallion (1999) Indonesia 1998 ... 25 13 ... Lanjow, Peter et. al (2001) Vietnam 1998 ... 26 9 1 World Bank (2000c) Notes: Unit of analsis for quiniles is individual. T, Total education or education as a whole; P, Primary (in some countries, referring to basic education); S, Secondary (for Vietnam and Indonesia, the average for lower and upper secondary educabon); H. Higher education. (...) Informnation is not available. (") forthcoming. However, due to data limitations, some measurement issues should be noted in the interpretation of results in Yemen. Firstly, the analysis does not capture quality variances in education subsidies. Secondly, the analysis assumes that all students were enrolled in public schools, as HBS-98 does not provide information on whether students are enrolled in public or private institutions and the share of private provision is known as small. Thirdly, the analysis does not include benefits from public spending on scholarships abroad. These issues might cause under- or over-estimate of the benefit-incidence for the poor. For example, if students who receive scholarships are more likely to be rich, the benefits for the rich were underestimated above. Private spending on education Some benefit-incidence studies also aim to analyze households' spending on education across socioeconomic groups with two approaches. First, some relate the benefit-incidence of public spending to the incidence of taxes with the focus on the redistributional effects (either across households or individuals over a recipient's life time). However, such a fiscal-incidence analysis might not be very useful for Yemen as there is no education-earmarked tax and the share of income tax is small in the total revenues.33 33 In addition, there is no data that enable to undertake the incidence of taxes across households groups by wealth indicator. 41 Annex 7 Second, some studies aim to relate the benefit-incidence to the incidence of private costs of schooling to asses the extent to which households have to pay to gain access to subsidized public education services, beyond the cost-recovery contributions. While public education is free from tuitions at all levels of education in Yemen, students still pay some fees such as community contribution fees, school activity fee, examination fee, and for university registration fees. In NPS-99, many Yemeni households responded "the difficulty to pay school expenses" as the main reason for leaving school or not sending children to school. Therefore, it may be useful to see the extent to which households have to pay and whether there is any substantial difference between the poor and rich households. On average, the yearly household expenditure on education is YR 2,272 or 0.6 % of the total household expenditure. It is also estimated that households spend about YR 250 on government school fees and expenses and YR 300 on stationeries for sending one child to school. Although the magnitudes do no sound substantial, the cost of sending children to school is not negligible for the poorest, especially in urban areas (see figure 7.9). Overall, urban households spend 3.7 times as much as rural households. In urban areas, the yearly household expenditure on education for the richest household decile is the highest in amounts (YR 13,000) and as a share of the total expenditures (1.5%), while the poorest also spends 1.3% of the total expenditures. In rural areas, the private cost of education is regressive in relation to household expenditures, as the poorer tends to spend a higher share of the expenditure on education. On average, a household of the rural poorest decile spends 0.8% of the total expenditure on education while a household of the rural richest decile spends 0.2%. Figure 7.9. Yearly household expenditure on education 14o000 1 6 12 00 1,4 ~~~~~~~~~1 2 AI 4CM67 91 10,000 iUzban 1R 8,00 X ~Ruxal OCR) 4 ,000 _ 0, AI2~L 6 2~O Po A --E-- Rum1al 0 OD~~~~~~~ ___ _ __ 1 2 3 4 5 6 7 8 9 10 Househofl decil Sou3e:Aut±ors estinat usng HBS-98 The urban-rural difference in private education expenditures among the rich is related to their use of private provision and lessons (see Table 7.10). The urban riches households spend nearly 40% of their education expenditures on pnvate schools and lessons while their rural counterparts spend only 6% and the poorest households spends about 1% either in urban or rural area. The urban richest households also out-spend on university education, YR 3550 or 27% of their education expenditures. 42 Annex 7 Figure 7.10. Household spending om private schools znd lessons 40- 'a 35 30 *10 (V5 co 0 1 2 3 4 5 6 7 8 9 10 Householl decil Souma:Authorb estanate USD9 HBS -98 Targeted Government programs to underserved Grotnps Government does not have specific program explicitly targeted to the poor as defined by family welfare level. However, some programs such as free-tuition and free-textbooks for all might have more positive impacts on the poor's enrollments. In addition, the govemment has some programs broadly targeted to underserved groups, and more recently targeted to rural girls. However, the impact of the programs are not evaluated. Informal education program targeting for underserved boys. Many policy makers address that Yemeni people want education and in particular the poor families are keen to get children educated. When there is no school nearby, it is known that families try to have at least incomplete schools in villages such as school under tree, tent and at renting rooms. Informal literacy education was also more available for underserved boys at various places such as mosques and military camps. Until 2001, all boys aged 16 and above had the obligation for about 1- to 2-year military service. Military camps often provide educational programs for eradicating illiteracy. Reduction in direct costs of schooling for poor girls. For the poor girls, MOE has started the exemption of community participation fees since 1998. In urban areas, schools stopped emphasizing the obligation to wear uniforms, taking into consideration the financial burden on the poor families. In rural areas, school uniforms have not been compulsory. Incentive programs for sending girls to school. There are some donor-supported programs to provide incentives for girls' schooling such as provision of free learning kits (support from UNICEF and IDA) and foods (support from WFP). The limited experiences tell that programs also have to benefit boys in the same communities. Although there are regional differences in communities' attitudes, material incentives targeting exclusively girl students are not accepted by communities. SFD also started a pilot special program for girls in five sub-districts (Uzra) in 2002 to test various means. Increasing availability of teachers for girls and children in remote rural areas. Key policies include: providing additional allowances for teachers in remote rural areas defined in the Teacher Law of 1998; giving the priority to fermale candidates in hiring new teachers; and improving the management in allocating teachers to meet additional needs of existing and new schools before starting a school year. To increase the number of female teachers in rural areas, especially for girl students in grades 5 and above, the Government 43 Annex 7 is hiring female secondary school graduates as new teachers, reducing the entrance criteria for females to the Faculties of Educatlon (FOE), and piloting a scholarship program for rural females to enroll in the FOEs. Targeting and improving sustainability of investment projects for expanding the access for girls and rural children. Since the late 1990s, MOE has introduced the following policy measures to increase the efficiency of school construction, i.e. to built more classrooms of a lower cost at a more needy site in a shorter period, and enroll more girls in the new places: (a) basing school location decisions on analysis of target children with school mapping tools and on consultations with the communities; (b) placing small schools (e.g., 3- classroom school) closer to girls' homes; (c) changing the physical design of schools to include sanitary facilities and boundary walls; and (d) obtaining the community's commitment to enrolling girls as a prerequisite for school construction in the community area. MOE also has adopted measures that might encourage parents to keep their daughters in basic education, for example (a) providing separate classrooms for girls (not includes boys) in grades 7-9 and (b) building girls' secondary schools. Piloting targeted site selections for basic education. MOE has started strengthening the capacity for implementing new procedures on targeted site selections under the Basic Education Expansion Project (BEEP) through leaming experiences under other agencies, mainly Social Funds for Development (SDF). The current pilot procedure starts with a preliminary selection of priority underserved-areas on the basis of simple indicators such as girls' enrollment rate and female illiteracy rate using the school-mapping database.34 This selection is then assessed by field visits of a team comprising an engineer and a community participation staff from the Govemorate Education Office, and a representative of the District Education Office. MOE is undertaking a review of their initial experience but the results are not yet available. The SFD's site selection procedure starts with office-screening to assess communities' proposals for new projects, followed by field-screening through visiting communities. Education programs to children with special needs. Govemment provides boarding schools for nomads, e.g. in Sabwa and Hadramout. There are a few schools for orphans, e.g. in Sana'a City. Some Governorates also provide transportation for secondary students in scattered population areas In higher education and vocational training, there is no explicit targeting program such as scholarships. Only dormitories are in-kind subsidies to students and implicitly supporting the poor as the most of dormitory users are from rural area. While universities tend to charge fees for dormitories (e.g., $6 dollar per month in Aden university, 2002), all vocational training centers provide free dormitories including meals. In 2001/2002, 42% of vocational training students (3638 students) reside at dormitories.35 Recommendations for monitoring the implementation of PRSP The education sector, especially basic education, is one of themes composing the Poverty Reduction Strategy (PRSP). The PRSP is expected to be a prioritized policy and expenditure framework for the government and some donors over the next-three years. Thus, the PRSP should also provide a basis for monitoring the progress towards improving educational opportunities for the poor and underserved population. The PRSP should enable a broad discussion among various stakeholders (e.g., parliamentarians, local council members, central and local administrators, school managers, teachers, communities, and donors) about the causes of and solutions to educational attainment of the poor, and enable policy makers to make decisions. Therefore, the data required for monitoring the PRSP should be: relevant to policies; reviewed on a regular basis (timely and consistent); and easy to understand. A few key indicators should be identified for policy makers to judge the progress. 36 34 The school mapping database, available for 8 Governorates, 35 Source: Ministry of Vocational Education and Technical Training. 36 [c.f PRSP Source (2001)] 44 Annex 7 Table 8 shows indicators that the Government may want to consider as key monitoring indicators and supplementary indicators for Yemen PRSP. These indicators are identified as they are: (i) relevant to internationally agreed goals (such as the Intemational Development Goals agreed at the OECD Development Assistance Committee and the Education for All 2000); (ii) relevant to Yemeni targets and policies as described in the second five-year plan (SFYP), the human development indicators in SFYP, issues and solutions in draft PRSP matrix (January 2002 version), and policy monitoring indicators for the Basic Education Expansion Project (supported by the World Bank); (iii) relatively feasible to collect data required for calculation of indicators on a regular basis; and (iv) easy to understand. It is aimed to differentiate between "key indicators" and "supplementary indicators," and between "indicators" and "data required for calculation of indicators" explicitly in order to meet different demands of various stakeholders in monitoring. The key indicators should be the fundamental indicators that everyone should know and that should be calculated for each Governorate and possibly District. The supplementary indicators could be for sector-oriented policy makers and specialists need to know more details on the progress of the education sector. Therefore, supplementary indicators include not only quantitative indicators but also qualitative indicators that could be judged based on various information from the fields and specialists for questions, for example "to what extent, the implementation of teacher redeployment is satisfactory?" Table 7.4. Ibdicators that may be considered as monitoring indicators for the education sector of PRSP Frequency Target Name of indicators of Measure 1998 1998 1999 1999 2000 2005 199711998 (HBS-98) 1998/1999 (NPS-991 1999/20002004/2005 Core Indicators EP- Illiteracy rate, adult (age 15 and over) (%) selectve 58 2 55.7 ,, D Illiteracy rate, youth (age 15-24) (%) selectUve ... 34 8 D Rato of young literate females to males (age 15-24) (%) selective 40 4 P Grass enrollment rate, basic (%) every year (70 6) 61.4 Gross enrollment rate, basic, girls (%) every year Gross enrollment rate, grades 1-8, girls (%) every year Rabo of girls to boys in basic educabon (%) every year Govemment spending on educabon (% of GDP) every year Rato of operabons and maintenancs budgets to MOE recurrent budget (%) every year Supplementary Indicators Illiteracy rate, adult (ages 15 and over), females selective 77 8 Illiteracy rate, youth (ages 15-24), rural females selectve 73.0 Gross enrollment rate, basic, boys (%) every year Gross enrollment rate, basic, rural (%) selective Gross enrollment rate, basic, urban (%) selectve Gross enrollment rate, basic, rural giris (%) selectve Gnoss enrollment rate, grades 1-6, rural girls (%) selectve DEP Net enrollment rate, grades 1-0 (%) selective DE Ratio of girls to boys in basic and secondary education (Y4very year P Apparent intake rate, grade 1, (% of population aged 6) every year P Complebon rate, basic (% of population aged 14) every year Number of classes, basic every year Ratio of evening/night-shiftng classes to total, basic (%) every year Number of teachers, basic and secondary every year Ratio of female teachers to males, basic and secondary (1Jery year Implementabon of redeployment of teachers to rural areaeVery year Implementation of new hiring based on school needs- every year Dropout rate, basic (%) selective RepebUon rate, basic (%) every year ? Ratio of teachers with at least post-secondary diploma (04very year ? Perception of teachers on the new cumculum, basic'- every year ? Improved capacity of MOE and Govemorate educabon offices for planning, budgetng, and implementation... every year ? Increased community particpation for constructing and operating schools- every year ? Notes * D Deveiopmnent Goal indicators E Indicators agreed in ihe 2000 EFA conference P PRSP sourcebook. QualitaUve indicators QualitaUve indicators based on vanous inforrnrUon (reference, BEEP Key Perforrranco Indicators for Serri-Annuil Montionng of Basic EducatUon Polides) HBS-98, Household Budget Survey 1998, NPS-99 Natonal Poverty Phenomena Survey 1999.. not avadable. 45 Annex 7 Table 7.5. Education and training indicators in the second five-year development plan (This summarv was received from Ministry of Planning in January 2002) 2000101 2005106 1995/96 (estimate) (target) Students in literacy program (total of 5years,000) na 290.0 345.0 Enrollment rate, basic (%) 56.6 61.4 69.3 Enrollment rate, basic, boy (%) 71.6 77.2 82.4 Enrollment rate, basic, girl (%) 39.3 43.9 55.0 Intake in basic education (000) 426.9 513.7 691.2 Intake in basic education, boy (000) 254.9 295.8 372.8 Intake in basic education, girl (000) 172.0 217.9 318.4 Enrollment, basic (000) 2,600.0 3,348.0 4,489.0 Enrollment, basic, boy (000) 1,763.0 2,203.0 2,787.0 Enrollment, basic, girl (000) 837.0 1,145.0 1,702.0 Graduate, basic (000) 113.6 196.0 265.0 Graduate, basic, boy (000) 81.2 144.1 191.1 Graduate, basic, girl (000) 32.4 51.9 73.1 Enrollment rate, secondary (%) 26.1 34.7 41.3 Intake in secondary education (000) 114.6 175.0 255.1 Intake in secondary education, boy (000) 91.7 123.3 168.8 Intake in secondary education, girl (000) 22.9 51.7 86.3 Enrollment, secondary (000) 288.0 444.0 663.0 Enrollment, secondary, boy (000) 231.0 315.0 436.0 Enrollment, secondary, girl (000) 57.0 129.0 227.0 Graduate, secondary (000) 46.2 109.9 126.0 Graduate, secondary, boy (000) 27.8 80.2 88.5 Graduate, secondary, girl (000) 18.4 29.7 37.5 Enrollment in vocational and technical training (000) 7.8 11.1 17.0 Enrollment in post-secondary technical education (000) ... 6.0 8.0 Enrollment in public university (000) 87.0 184.0 282.0 Enrollment in public university, male (000) 71.0 139.0 206.0 Enrollment in public university, female (000) 16.0 45.0 76.0 Ratio of university graduates in science to total graduates (%) 12.0 12.2 16.0 Teachers, basic (000) 116.1 142.7 179.7 Teachers, basic, male (000) 95.5 100.1 114.0 Teachers, basic, female (000) 20.6 42.6 65.7 Teachers, secondary (000) 15.4 25.5 37.8 Teachers, secondary, male (000) 12.3 20.3 30.0 Teachers. secondary, female (000) 3.1 5.2 7.8 46 Annex 7 TabRe 7.6. Inl tercy rate arad earoRllgment rate, 11994 nad 1998 Reoublic Ur an 199 8 1994 19981 1994 1998 Illiteracy rate youth (ages 15-24 1 9 Tota 39.6 34.8 15.8! 10.61 49.3 43.7 Male 17.1 11.9 7.7 3.8 21.4 14.9 Female 64.4 58.2 26.5 17.51 78.0 73.0 Illiteracy rate adult ( aes 15 and above) Total 62 56' 40. zu _ O2 MAale _3, Female 7-7 IL 57 4! Enrollment rate gaes 6-14) 1 Total 5 . _0 79. 8i. 48.- ' 5fit Male 70. 75. 83. 84.9 6-1 -M Female 43. 75 .4 77.! _283:4 Sources: CSO (1996) for data in 1994. Authors estimation using HBS-98. Table 7.7. Edunc&tion stazt ,Toir popuBati2on aged 1S amd above, 1998 M% _ Total I Urban Rural Male Female Total I Male | Female Total Male Female Total Illiterate 34.1 77.8 56.21 18.21 48.61 33.3 39.4 87.2 63.7 Read & write 30.8 11.3 20.9 28.0 19. 23.6 31.8 8.7 20.0 Lower basic (2rimary) 5.1 2.5 3.8 7.0 5.9 6.4 4.5 1.4 2.9 Upper basic (unified, - preparatorY, basic) 14.7 4.8 9.7 17.61 1321 15.5 13.8 2.1 7.8 Pre-secondary diploma 1.3 0.2 0.8 2.2 0. 1.4 1.0 0.1 0.5 Secondary 9.6 2.5 6.0 17.1 9.1 13.2 7.0 0.4 3.6 Post-secondary diploma 1.8 0.3 1.0 2.51 1.7 1.5 0.0 0.8 Undergraduate and above 2.6 0.7 1.6 7.3 2.41 4.9 1.0 0.1 0.6 No answer 0.0 0.0 0LJ 21 0 0.0 0 0.0 Total 100 100 100 1 100 10 100| 100 Source: Authoes estimation using HBS-98. IWA 7I. tpt 1n Urban I Rural Male Female Total I Male mal Total Male emale Total lilitrac1y rate Illiteracy rate, youth (ages 15-24) 11.9 58.2 34.8 3.8 17.5 10.6 14.9 73.0 43.7 Rato of literate females to males (ages 15-24) na na 46.4 no na 8 n na 31.1 Enrollment rate Ape enrollment rate, ages 6-14 75.4 43.4 601 84.9 77.7 72.8 33.3 54.1 Gross enrollment rate, basic (grades 1-9) 90.4 49.0 70.6 100.9 89 2 95.2 87.5 37.3 63.7 Gross enrollment rate, secondary (grades 10-12) 60.2 17 7 39.3 | 64 4 49.1 56.9 58.8 6.9 33 2 Gross enrollment rate, university and highar 17.2 5.0 11.1 30.6 16.9 23.8 12.5 0.8 6.6 Share of temale students Share of females, basic (grades 1-9) na na 331 na na 45.9 na na 27.7 IShare of females, secondary (grades 10-12) I nal na 22.21 na nal 42.3 nal nal 10.2 IShare of females, vocational training I nal na 27 91 na nal 35.3 nal nal 13.7 [Share of females, university and higher I nal nal 22.21 nal nal 3541 nal nal 5.7 Source: Authors estmation using HBS-98. Note. na.not applicable. 47 Annex 7 Table 7.9. Selective indicators by decile, 1998 Decile (1= the poorest 10%) 1 9 3 4 l R 7 A g 1X Household Docile 1 (Based on Per Capita Household Monthly Expenditure) giare in total households () 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 glare in total population (%) 12.3 11.5 11.3 11.1 10.7 10.4 9.6 8.8 8.0 6.4 Illiteracy rate, adult (ages 15 and 63% 60% 60% 58% 58% 55% 55% 53% 51% 45% Illiteracy rate, ages 15-24 39% 34% 40% 39% 35% 33% 34% 33% 30% 28% Illiteracy rate, adult (ages 15+), male 43% 39% 39% 36% 35% 32% 33% 30% 27% 23% Illiteracy rate, adult (ages 154-) female 83% 79% 80% 80% 80% 78% 76% 75% 75% 67% Illiteracy rate, ages 15-24, male 15% 12% 20% 17% 14% 7% 9% 8% 7% 7% Illiteracy rate, ages 15-24, female 65% 57% 61% 62% 58% 59% 58% 56% 53% 49% Enrollment rate, ages 6-14 56% 55% 56% 60% 63% 61% 65% 64% 62% 67% Enrollment rate, ages 15-17 45% 46% 44% 46% 50% 53% 52% 53% 55% 60% Enrollment rate, ages 18-23 25% 26% 22% 24% 29% 27% 22% 28% 32% 28% Enrollment rate, ages 6-11 53% 53% 56% 59% 61% 58% 63% 63% 59% 66% Enrollment rate, ages 12-14 65% 60% 58% 63% 66% 68% 68% 65% 69% 68% Enrollment rate, ages 6-14 , male 70% 70% 73% 76% 77% 77% 80% 83% 78% 83% Enrollment rate, ages 6-14, female 41% 39% 39% 43% 47% 44% 48% 45% 47% 52% Enrollment rate, ages 6-11, male 64% 65% 69% 72% 73% 73% 75% 78% 73% 77% Enrollment rate, ages 6-11, female 40% 40% 41% 43% 48% 43% 49% 47% 44% 55% Enrollment rate, ages 15-17, male 69% 67% 64% 67% 75% 79% 77% 74% 77% 77% Enrollment rate, ages 15-17, female 20% 26% 24% 25% 23% 28% 29% 31% 32% 37% Enrollment rate, ages 18-23, male 37% 38% 35% 35% 42% 43% 34% 48% 49% 44% Enrollmentrate, ages 18-23, female 11% 12% 8% 11% 12% 12% 11% 10% 14% 15% Urban Household Decile 1 (Based on Per Capita Household Monthly Expenditure) glare in total population (%) 13.6 12.5 11.8 11.0 10.5 9.8 9.2 8.3 7.8 5.5 Illiteracy rate, ages 15-24 13% 13% 11% 11% 9% 11% 8% 9% 7% 9% Illiteracy rate, ages 15-24, male 6% 5% 5% 3% 3% 3% 2% 3% 2% 2% Illiteracy rate, ages 15-24, female 21% 20% 18% 20% 15% 18% 14% 15% 11% 17% Enrollment rate, ages 6-14 73% 77% 78% 81% 84% 87% 87% 87% 91% 87% Enrollment rate, ages 6-14 , male 77% 81% 81% 86% 89% 91% 87% 89% 93% 93% Enrollment rate, ages 6-14, female 69% 72% 75% 76% 79% 84% 86% 86% 88% 82% Enrollment rate, ages 15-17 61% 63% 69% 71% 70% 74% 77% 75% 77% 73% Enrollment rate, ages 15-17, male 69% 68% 80% 81% 80% 84% 84% 82% 87% 87% Enrollment rate, ages 15-17, female 53% 58% 56% 62% 59% 64% 70% 67% 66% 58% Enrollment rate, ages 18-23 31% 32% 39% 38% 36% 35% 41% 39% 36% 42% Enrollment rate, ages 18-23, male 35% 38% 44% 46% 44% 44% 53% 47% 45% 55% Enrollment rate, ages 18-23, female 27% 26% 33% 29% 28% 27% 30% 32% 27% 28% Rural Household Decile 1 (Based on Per Capita Household Monthly Expenditure) clare in total population (%) 12.03 11.39 10.87 11 10.97 10.28 9.78 9.05 8.16 6.46 Illiteracy rate, ages 15-24 44% 40% 47% 49% 46% 42% 43% 43% 38% 43% Illiteracy rate,ages 15-24, male 17% 13% 21% 25% 17% 13% 8% 13% 8% 10% Illiteracy rate, ages 15-24, female 74% 68% 74% 78% 75% 72% 75% 73% 70% 73% Enrollment rate, ages 6-14 54% 53% 51% 52% 59% 55% 57% 57% 52% 53% Enrollment rate, ages 6-14 , male 69% 70% 70% 70% 78% 72% 74% 82% 74% 77% Enrollment rate, ages 6-14, female 37% 33% 31% '30% 37% 36% 36% 29% 29% 31% Enrollment rate, ages 15-17 41% 42% 39% 38% 39% 48% 46% 43% 47% 46% Enrollment rate, ages 15-17, male 68% 68% 63% 58% 72% 75% 82% 70% 74% 72% Enrollment rate, ages 15-17, female 14% 16% 14% 13% 11% 18% 15% 17% 19% 13% Enrollment rate, ages 18-23 23% 24% 21% 19% 23% 24% 20% 21% 27% 22% Enrollment rate, ages 18-23, male 37% 39% 36% 32% 38% 42% 39% 40% 48% 43% Enrollment ratei ages 18-23. female 7qu A% AOh w96 7q A A O Ah AOL AO 7 OA 48 Annex 7 Table 7.10. Selective indicators by decile. 1998 Decile (1= the poorest 10X) 1 2 3 4 5 6 7 8 9 10 Individual Docile (Based on Per Capita Household Monthly Expenditure) ghare in total population (%) 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 ,hare in population aged 6-14 (%) 11.9 11.3 10.8 10.9 10.2 10.0 9.8 9.3 8.5 7.4 Illiteracy rate, adult (ages 15 and 64% 60% 60% 59% 58% 58% 55% 54% 52% 47% Illiteracy rate, youth (ages 15-24) 39% 35% 38% 38% 38% 35% 33% 34% 30% 29% Enrollment rate, ages 6-14 56% 56% 56% 57% 60% 64% 63% 65% 62% 66% Enrollment rate, ages 15-17 44% 45% 45% 47% 48% 50% 55% 51% 55% 57% Enrollment rate, ages 6-11 52% 52% 55% 57% 60% 62% 59% 63% 61% 64% Enrollment rate, ages 12-14 64% 66% 57% 59% 62% 68% 70% 67% 65% 71% Share in total enrollment (%) 10.4 10.0 9.6 9.8 10.1 10.5 10.2 10.1 10.0 9.4 Share in enrollment basic (%) 10.8 10.6 10.0 1 0.2 10.2 10.5 10.1 9.9 9.4 8.3 Share in enrollment secondary (%) 9.3 7.8 8.3 8.0 10.1 10.4 10.6 10.6 11.6 13.2 Share in enrollment university (%) 5.5 4.9 5.4 7.0 6.7 10.2 10.6 12.9 15.0 21.7 Share in enrollment vt (°L0 4 4 4 A R2 n n 1 rj 1in 1jn7 1R n 99 9 1; rn Source: Author's calculation using HBS-98. Table 7.11. Il iteracy rate and enrollment rate by governorate, 1999 _ _ Age en rollmen for 6-14 ) Illiteracy rae (% Illiteracy ra _%) I Youth ( 5-24) Adult (1 5+) - Male Femal Total * Male Femal Total * Male Femal Total * San'a City 85.. fD 2 80 1.. 7AW.2 13AI 4..2- ..27 Taiz 7J. 4 6 41 :5 - 28.' 2 6 8 - ]i AI-Mahrah 5..~ .53. 4.,5 - 22-. 39-a -a A4 4 -~ Hadramout 68.l !5 50 _ 35. 5 > Laheg 77- 49. 67 - 5 i 27 SO 1 Ibb 7. 43, i 79 7.1 57, 3 8 3. 8i5 1 Abvan 70' 4?2 lo. I_6 42 - 25.. 6 - Al-Baida Z Q 4 , 164. 7A9. 61- Mareb _ 7 _1 1 14.11 64- Al. -i_s Aldhlea 71i. 1.4..~ .1 ... (I .57- 32~.- ...2i- 77-A 53..~ Al-Hodeida '5re 1A. 44.1 _ 1 _. 41 48: 77 63- Shabwah 71- i 1:LL53 7.1 6Q i q F7t I9 57- Ai-Mahweet 467.4 1 3- 71 _ .A _ Aa-Jawf 5 2 L 26 4 2 40-. 8 o- 1 A L 89 62. Amran 74. i 26- 1Si lf 9: 69 A -38; 341L8 5 1 Sana'a 1 25 _ 43 -1 40.' _7 3 Haiiah 54I 2LL@q- 76- X 2 6.: 877. Dhamar -44 11. 863:_ Sa'adah - 17. 40 23 8 5 - 7 91 6 t TOTAL 408A. 454 1 336 74. 55. 8 Source: Authors estimation using the NPS-99. Notes * Rank for female indicators. From the better to worst Governorate. 49 Annex 7 Table 7.12. Distribution of population by enrollment status In 1998/1999 and 1997/1998 (%) Total Male Female Urban Rural Urban FemaleRural Female 6-11 12-146-11 12-146-11 12-146-11 12-146-11 12-14 6-11 12-14 6-11 12-14 Enrollment status in 1998/1999 Enrolled currently 51 63 63 81 39 43 73 86 45 55 71 82 30 29 Enrolled previously 1 9 1 8 2 10 1 7 2 10 1 9 2 11 Never enrolled 47 28 36 11 59 47 26 7 54 35 29 10 68 60 100 100 100 100 100 100 100 100 100 100 100 100 100 100 For enrolled previously Enrolled in 1997/98 30 20 33 20 29 20 32 16 30 21 34 16 28 20 Enrolled before 1997/98 70 80 67 80 71 80 68 84 70 79 66 84 72 80 100 100 100 100 100 100 100 100 100 100 100 100 100 100 Source: Authoes estimation using NPS-99. Table 7.13. Dropout rate by age group, 1999 Urban Rural Total Urban Rural Male Female Female Female age 6-11 3 1 3 2 4 1 5 12-14 12 8 15 9 20 10 28 6-14 6 4 8 5 10 5 13 Source: Author's estimation using NPS-99. Table 7.14. Distribution of children who left school by the highest grade attended and the main reason for leavine school (aeed 6-14) Total Male Female Urban Rural Female Male Urban Rural Urban Rural Highest level of grade attended 1 6.7 6.2 7.0 5.3 7.0 5.2 7.4 5.5 6.4 2 15.7 13.7 17.3 13.0 16.3 12.2 18.5 14.0 13.6 3 19.5 16.5 22.0 17.9 19.9 17.0 23.2 19.0 15.9 4 20.2 19.2 21.1 21.8 19.9 23.5 20.5 19.5 19.1 5 15.6 15.5 15.7 20.4 14.5 20.7 14.5 19.9 14.6 6 14.9 17.4 12.8 13.5 15.2 14.7 12.4 11.7 18.6 7 4.0 5.8 2.5 4.3 3.9 4.5 2.0 4.0 6.2 8 1.9 3.3 0.8 1.9 1.9 1.0 0.8 3.1 3.3 9 0.7 1.4 0.1 0.8 0.7 0.3 0.1 1.5 1.3 10 0.3 0.3 0.3 0.4 0.3 0.2 0.4 0.7 0.2 11 0.0 0.1 0.0 0.1 0.0 0.0 0.0 0.2 0.1 12 0.2 0.3 0.2 0.7 0.1 0.6 0.1 0.8 0.2 No answer 0.2 0.3 0.1 0.1 0.2 0.1 0.1 0.0 0.4 Total 100 0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Before grade 5 62.1 55.6 67.4 58.0 63.1 57.9 69.7 58.0 55.0 Grade 5 and after 37.7 44.1 32.5 42.0 36 7 42.0 30.3 42.0 44.6 Main reason why they left school Completion of the level (basic or higher) 3.4 4.7 2.5 4.9 2.9 3.8 2.0 6.5 4.2 Suaply-side factors o schoo f nearby 12.9 12.5 13.2 6.0 15.2 7.5 15.2 3.7 15.3 Transportation difficulties 4.1 4.9 3.5 3.0 4.4 3.4 3.5 2.3 5.8 Lack of teachers 4.6 3.9 5.1 2.7 5.2 3.8 5.6 1.2 4.8 Demand-side factors Work to support the family 14.0 19.3 10.3 11.6 14.8 6.1 11.7 19.7 19.1 Family can't afford school expenses 21.7 25.6 18.9 26.8 20.0 24.8 16.9 29.7 24.3 Family doesn't want to send girls to school 13.7 1.3 22.6 13.3 13.8 21.8 22.8 0.7 1.5 Marriage 0.7 0.5 0.8 0.7 0.7 0.8 0.8 0.5 0.5 Others 24.8 27.2 23.2 31.2 22.8 28.1 21.5 35.7 24.5 No answer 0.1 0.1 0.0 0.0 0.1 0.0 0.1 0.2 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Author's estimation using NPS-99. 50 Annex 7 guu : ~~~~~~~~~~~~~~998 ( ___ ___ __ ___ ___ __ ___ ___ nmp Y Rumi nswr ota00m F T F T Disotnco to echoolo fcz thcoo cnrollod in bacidc oduction Less than 15 minutes 73. 80.6 .9 89,. .0 8 68 74.9 705 1.6-30 minutes 86--4 85 2. 31-60 minutes 6 4.1 Q An hour and more 1.6 032 __ 02 ___Oj ___ J .._9. I l No answer 09 00_ 0. 0 Q I --JOA 01 0 Total Ji-l0 10 1001 0t .JQ -JQ 10Q t 9 J..tO Distknco to achoilt fa7 thoolo n7011cd in Coco duc2 on. _.__qRss th.a 49. 71. 75ir t7.0 7$JE6.2 31k5 56c L 1M6-r b minke/bi *7 0. 0, _0.3 0.1 02 _M 0 0_ .3t-60 minutst .21OP9 3. . 2 An heur 0n mnne __U _0 09 0.0 0 QJ - No answer 0.2 ° ° 2 200 0.1 0.2 O.C 0.. Total I no 1 QQ 1O 1 no I no 1 oJ o 1 9 no1 JWAlking _ 7 a1 -21 MO 783& Car/bus 2.91 2.9 1 J4. J4 J2j)1 Motor bike/bike 0.3 0.0 _ 0 oa 0.1 0. 0. O.C 0. Other 0,1 O O 0.0 06 0.1 O.t 0.1 No answer 721.& 220 1J Z5 248 JjQ33 203 Total 1 Qi JQ9 JQP J M J~QF ~p opnon of ema .nRchoolq bnei .dotion YVakin- fit Q 52 7 _60n . 91 .9- 3 Qar/hbu 16.5 12? 4 _15_ _.C _129 _14 9 -L E1S- Mover hiko /beker 1.5 I 0.2 12 ___. Q. O.' -9. 1 -f oQthd 0 7 00 021 0. 0.0 00m 0Q 0 0Q No answer L%J 2Q0 03 20.7 18.8 0 03 Totaco 100 00 0 00 00 OOC Loontion of achool. aco aducotion -in illase .- L678 4 A 713_ in ediuc-nt city -RM 1 7 RAMoveorate creter Jj1 ! 16 outside gover I ...-0. 0 00 QO No answer f .L1 3 0 3 0.3 Total lQa 9 100 io 01U1OJ 100 lOt Loction of ochol, bccondtio oducnton in villazs 51.S| 74 2l66 7.1794 7. 111r . 32 in adlucOnt CitY ! M _4Q7. 1- govemorate center L_J_ _ 436 boyts& ids Iaa _A _58 q _~ .iIJ2-2 Nionswer __0 Ttal I loQ.P1 Q I00~jIfSI~ 1009P100 10. Tvl of Cahoolcndic oducntion bovs & aiWl ! 64.4 58.43 -D hboysonly , a .s L _2 3 j3 8 airlsonls 2c7M -?-I __U 1.i 4 No answer 0 91 0 - QJ0 __ fl 1 Total 1 00 l.0 I= QPiIf~ I 1g- inn i00nnn~0 inn.0 STaou of Achoolr cluconduao oducasin h bvsnv lhdr 48.2 24.61 4 3 7.9 47 66 6.17 6 | irls only -L I Ng ainswer L Q9 1 0 J_ T Source: Author's calculation using HBS-98 51 Annex 7 Table 7.16 - Access to Schools for Those (aged 6 and above) Enrolled in 1998/1999 (%) Name Total Urban I Rural IM I F I T I M I i- I T I M I F I T Distance to basic schools Less than 10 minutes 34.40.4 45. 47. 46.t 29. 34A 31. 10-29 41: 43(1 41 44 4. 44A 44 40. 42A 4A2 30-59 minutes 6 12, 14. 8. 7 816r 18. An hour and more _ 4. 7 1 0.. .11 11. 6 10. No answer 0.1 0.( 01 Total .J.10tJ 10i 10t 10 10i 1 Distance to secondary schools (academic and vocational) Less than 15 minutes _ 2 27.1 2 2 16. 9 16-30 minutes 3 5 42. 49.A 1 5.. 2.. 33.1 45 !i 35A. 31-60 minutes 20~ Q .20-413 1 7L .23 223 ..23. An hour and more 4Ag 14. 1 26 12.24 No answer - Q.t O. O.t 0.( Total I10 101 10( f 1 10( Type of school, baslc education* Govemment 98.3 6 97.1 96.t 99.: Government&non-government 0.I O.' 0. 1.o.d 0 1. O..' 0 .1 0 .'.O.' Non-government . 1 1. 3 . 0 ! 0.i 0 0.f Total 10_ 10 10( 101 100 10( 10Q 1! 1 1 Type of school In detail basic education Govemment 6.1 97-1 1 96 .21 a 1 9 .2 Private sector 1 .42 .3 .7 2. ..1. 0.36 4A .3, Cooperative association 1 02d 0.. i 0...lA . 0.21 0.2 0.-2 Gov & private Q ° °4 °.I °.0 0.11 0.01 0.11 Gov & cooperative 0.0 ° °° Private & cooperative fQ0 l.o 0 04 0..Q 0.0& 0.0:. No answer 0 o Q.o Q oa I I.Q 0 .0 Total 11o 1 loO Type of school, secondary educatlon* Govemment 8 97.7 7.4 97.2 99. .1 98. Government&non-government 0. o 0. 0.1*Q. . o-d flo d ° Non-government 1 2.4 J. 2.0 2.L 0.Q 1.. 1.1 Total JQ 01 m0 ..oa4Q ...1Q. .1.Q 10J Type of school, vocational training (pre- and post- secondary)* Govemment 98.9 90. 97.J nl - nl _nl_ _nl nl ni Governmentnon-government 0. 5. 1 n n nl Non-government 114...n.j.nl Total r 1 od 1 od 1o nl nl l nl n r Type of school, higher education (academic diploma, university, and higher)* Govemment |92. 92.6 92.1 n a n Governmentnon-government 0. . .7 nl Non-government&n 6.n 7.d 6.7 nln n n n Total 11 1o 1o nl nl n Source: Authors estimation using NPS-99 Notes: * Non-government includes: private organization or cooperative associations. 52 Annex 7 Table 7.17 - PublHc Provision of Education for Households. 1999 Republic Rurau Urba vd _ ilabilitv of school in the aLga B3asic school YES 77.2 91-6 8.7 No ??-a 8.4 t9.3 Secondarv school . . Yes 36.1 81 4 _ No 64.0 18.1 5. _tance to the neairest schol Rasic school Very near 60.6 83A 6. Far 30.6 15-.5 26.9 Verv far 8.8 1s] 7, Secondary school Vrvna 26l.1 70.3 36.t Far ea 46-8 27D 42.0 _Verv far 27.1 217 21.2 Basic school 19.0 5 5 10-4. Car/bus/van 0.4 0.1 0.4 Animal 86.9 94.2 887 On foot 0.3 0.1 Q.3 Roscyrle Q.1 0.1 0Q1 Other 0,°2 I L0 0X No answer Secondary school Car/bus/van 42.9 15.8 36.3 Animal 0.4 0.2 0-4 On foot 54.3 83.4 6t. Bisvole ~~~ ~ ~~1.2 0.5 1,0 _ No answer I 0.6 Q0 il ry reo r t re4 ..ach the trans ot servic. .( in minu . as) Rasicr school I ess than 29 minutes 58.5 9?2I 66.6 30 minutes 17.t8 5.tl 14-5 31-59 minutes 4.6 1. 1 3.7 60 minutAes I 05 0.' 8.1 60 minutes and more 8.6 0.4 61i Secondary school Less than 29 minutes 319 7 44,' 30 minutes 18. R. 15.t 31 -59 minutes 6.8 53 7. 60 minutes 16.5 1.1 12 I 60 minutes and more 27.6 o3 21.{ On oo school R asic school Publdc 98.2 97.7 98.1 Private 0.6 2.3 1 5 Coonerative 0.9 0.0 0.1 Other 0.1 0.0 QJ No answer 0.2 _ _0 __ 2 Secondarv school Public 98=5 98.7 9IB 5 Private 0.6 1.2 r n7 Cooneratl Q.4 0.1 Other 0 Q. _ No answer 0.6 I 0.0 Source: Author's estimate using NPS-1 999. 53 Annex 7 Table 7.18 - Distance to schools for those enrolled in basic education. 1998 Decile (1= the poorest 10%) 1 2 3 4 5 6 7 8 9 10 Total Rural female Less than 15 minutes 68 72 76 71 79 72 82 73 78 88 75 16-30 minutes 18 18 19 22 17 21 16 26 14 11 18 31-60 minutes 14 9 5 6 3 6 2 1 6 1 6 An hour and more 0 0 0 1 0 1 0 0 1 0 0 No answer 0 0 1 0 1 0 0 0 0 0 0 Total 100 100 100 100 100 100 100 100 100 100 100 Rural male Less than 15 minutes 62 64 68 73 69 73 76 64 71 74 69 16-30 minutes 21 26 21 20 22 19 18 27 18 20 21 31-60 minutes 16 10 9 6 6 6 4 7 7 4 8 An hourand more 2 0 2 1 2 2 2 1 3 1 2 No answer 0 0 0 0 1 0 0 1 1 0 0 Total 100 100 100 100 100 100 100 100 100 100 100 Urban female Less than 15 minutes 87 87 87 89 89 91 90 92 89 90 89 16-30 minutes 11 10 12 10 9 8 10 6 10 9 10 31-60 minutes 2 2 1 1 2 1 0 2 1 0 1 An hour and more 0 0 0 0 0 0 0 0 0 0 0 No answer 0 0 0 0 0 0 0 0 0 0 0 Total 100 100 100 100 100 100 100 100 100 100 100 Urban male Less than 15 minutes 87 87 89 90 89 91 91 93 88 87 89 16-30 minutes 10 12 10 8 11 8 8 6 11 12 10 31-60 minutes 2 2 1 1 1 1 0 1 1 1 1 An hour and more 0 0 0 1 0 0 0 0 0 0 0 No answer 0 0 0 0 0 0 0 0 0 0 0 Total 100 100 100 100 100 100 100 100 100 100 100 Source: Author's calculation using HBS-98. 54 Annex 7 1Table 7.19 - Overview of Public Education E atures, 11996-2001 1996 19!7 1998 1999 2000 Actua5 AcluaI Actual Actual Budget 0verall Public Education Expenditures Total expenditures (current YR billion) 37.3 46.2 56.9 67.4 91 .2 Total expenditures (YR billion, 1999 price) 47.8 52.7 68.1 67.4 70.E as share of total public expenditures (%) 16.0 15.0 18.9 21 .0 23.2 as share of GDP (%) 5.1 5.2 6.6 5.8 6.1 Current expenditures as share of public current expenditures (%) 20.1 16.0 20.6 23.5 24.6 as share of GDP 4.6 4.2 5.2 5.0 5.3 Capital expenditures as share of public capital expenditures (%) 8.7 15.7 19.9 12.0 l6.8 as share of GDP (%) 0.5 1.0 1.4 0.7 0.8 Subzectoral Alloczaltonz Total expenditures by sub-sector (% of GDP) MOE 4.4 4.4 5.4 4.8 5.1 Universities 0.6 0.7 1.0 0.8 0.8 MOUGAVTT 0.0 0.1 0.1 0.1 0.2 Research institutions O.o C.0 0.0 0.0 0.0 Share in total expenditures (%) MOE 87.7 85.1 81.9 84.2 83.G Universities 11.3 13.3 15.9 13.5 13.2 MOUGAVTT 0.7 1.3 2.1 2.0 2.6 Research institutions 0.2 0.2 0.2 0.4 0.3 Share in recurrent expenditures (%) MOE 90.2 88.6 87.6 87.9 88.A Universities 9.0 i 0.0 11.0 10.5 9.8 MOUGAVTT 0.6 1.1 1.2 1.1 1.3 Research institutions 0.2 0.2 0.2 0.4 0.4 Share in capitalVinvestment expenditures (%) MOE 66.0 71.2 60.9 58.1 53.7 Universities 32.2 26.6 33.8 33.9 34.E MOUGAVTT 1.8 2.1 5.2 7.9 11.E Research institutions 0.1Q 01 0 0.1 0.2 Source: See Table 7.20 Notes: 1/ If university expenditures were calculated including scholarship abroad under MOE budget and community colleges, the total university spending increases to 1.3% in 2000. 55 Annex 7 Table 7.20 - Overview of Public ucatlon Expenditures. 19962001 1996 19971 1998 1999 2000 Actual Actual I Actual Actual Budget (YER million YR. Nominal) GDP (Market prices)' 736,414 889263 861,357 1,172,474 1,503,038 General State Expenditure 232,754 307,568 301,431 342,932 422,249 Total Recurrentllnvestrnet Expenditures MOE 32,693 39,303 46,580 56,786 _ 6,5R6 Universities 4,223 6,161 .9,036 9,071 I12,009 MOUGAVTT 271 615 1,170 1,319 2,407 Other Education 77 88 99 265 . 16 Total Education Sector 37,264 46,167 56,885 67,440 91,317 Total Government 232,754 307,568 301,431 321,453 .3944 1 12 Recurrent -Salaries & Wages MOE 24,260 26,104 30,438 42,787 55,642 Universites 1,992 2,889 3,879 4,f22 MOLUGAVTT 243 295 336 I 45 Other Education 58 61 156 179 Total Education Sector 28,397 33,683 47,158_ 60,968 Total Govemment 46,098 82,150 93,635 119,113 _1 4, '4b Recurrent -Non-Salaries MOE 5,956 6,650 8,694 9,107 14,3/'6 Universities 1,724 2,013 2,337 3,058 MOLUGAVTT 182 235 319 587 Other Education 23 28 100 113 Total Education Sector -8,579 10,970 11,863 18,134 Total Govemment 120,233 149,347 123,546 132,280 172,173 Recurrent-Total _ _ _ MOE 30,216 32,754 39,132 51,894 70,018 Universities 3,017 3,716 4,902 6,216 7,78C MOUGAVTT 205 425 530 655 994 Other Educaton 75 81 89 256 292 Total Education Sector 33,513 36,976 44,653 59,021 73,084 Total Govemment 166,331 231,497 217,181 251,393 321,471 Investment-Total MOE 2,476 6,548 7,448 4,892 6,t68 Vniversities 1,207 2,445 4,134 2,855 4,229 MOUGAVTT 66 190 640 664 1,413 Other Education 2 7 10 9 Total Education Sector 3,751 9,191 12,232 8,419 1 t2 233 Total Government 43,181 58,681 61,378 70,060 72, Sources: For budget, 1996 data from MOF budget book 1997; 1997-1999 data from the MOF final account book 1999; 2000 data from MOF budget book 2000. For GDP, data from the World Bank (received from Setareh, May 2002) 56 Annex 7 Table 7.21- Econemk Comaosltionof5 PubRiC lEductltc xneadt ures. 990-20 Actutal Budge Actual Budge Actual Budge Budget 1994 1995 1996 1997 1 998 1998 1999 1999 2000 2001 Ministry of Education (and GASI until 1996)_. _ _ Recurrent Expenditures 98 97 92 83 81 84 90 91 91 93 Capital &Investment 2 31 8 17 19 16 10 9 9 7 100 1001 100 100 100 100 100 100 i0 100 Recurrent by: Wages and Salaries 84 84 80 80 78 78 82 62 79 76 Operations and Maintenance 1/ 12 19 12 91 12 13 10 9 12 11 Scholarships/Fellowships Abroad 2 31 6 6 7 6 5 5 5 4 Transfers __2 212 3 3 3 3 3 4 9 100 100 100( 100 100 100 100 100 100 100 Universities (and Community Col egs since 2000) Recurrent Expenditures 90 81 71 61 46 54 64 69 651 66 Capital &Investment 1 0 1 9 29 39 54 46 36 31 35 34 100 100 100 100 100 100 100 100 100 100 Recurrent by: Wages and Salaries 61 63 55 54 58 59 62 62 60 59 Operations and Maintenance 1/ 33 26 27 26 23 21 23 21 23 22 Scholarships/Fellowships Abroadl 4 19 17 18 13 15 14 13 Transfers 2 2 i I 2 2 2 2 31 6 10t__ 100 10t 90t 100 100 100 100 100 100 General Authority for Vocational Trailn - Recurrent Expenditures 85 28 761 671 12 43 34 48 37 - Capital &Investment 15 72 24 33 88 57 66 52 63 10t 10o 100 10t 100 100 100 100 100 Recurrent by: Wages and Salaries 85 88 90 59 55 57 44 52 Operations and Maintenance 1 1 0 7 37 41 40 54 46 Transfers 3 3 3 3 4 3 2 2 1 -__ _ 10 0 100 100 100 100 100 100 90( Source: Authors calculation using data from MOF final account and budget books. See Table 7.22 for details Notes: 1/ Excluding fellowships and training abroad or scholarships abroad 57 Annex 7 T A3P f0 olnal, nil R Table 7.22- Public Education Expdiiturm 1990-2001 (nominal, ml RY) AcTuals Atutal Budgeted Acutal Budgeted Actutal Budget Budget 1990 1991 1992 1993 1994 1995 1996 1997 1998 1998 1999 1999 2000 2001 Ministry of Educaton (Plus GAIE untl 1996) Total Expenditures 5,535 7.498 9.293 12,104 15,096 20,565 32,693 39,303 49,905 46,580 57.172 56,786 76,586 89,403 Recurrent Expendiures 5,084 6,950 8,211 11,629 14,798 19,987 30,216 32,755 40,380 39,132 51,398 51.894 70,018 83,298 Recurrent Exding F&T 4,919 6,715 7,954 11,400 14,587 19,291 28,346 30,899 37,533 3S,764 48,572 49,055 66,844 80,020 Capital &Investment 472 546 1,082 475 300 598 2,476 6,548 9,545 7,448 5,775 4,892 6,568 6,105 Wages and Salais 4,463 6,026 7,120 10,133 12,441 16,682 24,260 26,104 31,441 30,438 41,976 42,787 55,642 63,355 Operations (Goods and Service) and Mait 379 550 623 998 1,800 2222 3,552 3,733 4,901 5,126 5,079 4,898 8,318 9,100 Felowships & Traning AtroadfSchlarship 145 235 257 230 229 676 1,870 1,858 2,826 2,369 2,826 2,839 3,174 3,278 Transfers 77 140 212 271 327 388 534 1,061 1,192 1,200 1,516 1,370 2,884 7,555 Universles Total Expenditures 558 927 1360 1375 1537 2080 4223 6144 11316 9048 10322 9074 12009 14150 Recurrent Expendiures 456 780 1074 1259 1388 1683 3017 3718 5210 4914 6580 6219 7780 9380 Recurrent Exduding Followship & Train 441 756 1043 1202 1339 1537 2513 3025 4309 4028 5891 5255 6872 8203 Captal &Investment 102 147 286 116 149 396 1207 2425 6105 4134 3762 2855 4229 4770 Wages and Saaies 264 483 641 747 850 1,0S9 1,674 1,994 2,998 2,891 4,091 3,879 4,722 5,613 Operations (Goods and Service) and Mair 161 246 371 422 456 442 806 976 1,218 1,056 1,484 1,279 1,745 2,057 Felowships & TraningAbroad/SctlarshipE 15 24 31 57 49 147 504 693 902 888 869 964 1,108 1,177 Transfers 16 27 31 34 33 26 33 55 93 80 115 96 204 532 Comnwnity Colleges Total Expenditures 0 O O 78 179 Recurrent Expendiures 78 124 Capital &Investment 0 54 --4n evst- Wages and Salares 10 39 Operations (Goods and Service) and Ma tenalce Excuding Feowslp(Train@rgfrca2ho1ArsrpA ad 38 50 Fellowships & Trarnng AbroadlSchlarshlp . (2-2 ur- i1CY, i-q-6-i frlt 2 ,) 0 29 Transfemrs 30 7 General Authority for Vocatona Trining Total Expenditures C CO 77 65 326 271 572 3,932 1,007 2,233 1,123 2,226 Recurrent Expendlures _ 48 55 90 205 384 456 433 761 537 826 Capital &Investmnent I _ 29 10 236 66 187 3.477 574 1,472 586 1,400 Wages and Salaries a M 40 47 79 185 228 251 248 332 280 Operations (Goods and Service) and Mai O 6 7 9 14 144 188 172 413 245 Transfers 0 - - 2 2 3 6 12 16 12 15 12 Foklowship & Training Abroad Sources: MOF Budget Books and Final Accounts 58 Annex 7 Table 7.23 - Students in Public Institutions, End Totas nn6 Per Student Recurrent Expenditures,1998 1998 (1i997/1998) Basic Students 2,847,94 Total recurrent expenditures (million YR) 1/ 33,26 Subsidy per student-1 5 i D@@ Recurrent expenditures (million YR), excluding scholarship abroad 31,24 Subsidy per student-2 - Secondary Students 336,32 Total recurrent expenditures (million YR) 1/ 5,87 Subsidy per student-1 7 4 Recurrent expenditures (million YR), excluding scholarship abroad 5,51 Subsidy per student-2 D3J00 University Students 97,59 Total recurrent expenditures (million YR) 4,91 Subsidy per student-I 0m3w Recurrent expenditures, excluding scholarship abroad (million YR) 4,02 Subsidy per student-2 6Q 27 Vocational Training, Pro- and Post-Secandary 2/ Students 7,05 Total recurrent expenditures (million YR) 433 Subsidy per student 3,42,2 Notes: 1/ It is assumed that 85% of MOE recurrent expenditures were for basic and 15% for secondary education. (see Table 7.24) for detai;) 2/ It is impossible to disaggregate unit subsidy for pre-secondary and post-secondary vocational training due to a lack of information. 59 Annex 7 Table 7.24 Computation to disaggregate MOE recurrent expenditures 1997/1998 MOE Schools 1/ Students (STU) Basic 2,522,92 Secondary 294,36 Teachers Basic and Some of Mixed Basic and Secondary 2/ 137,22 Secondary and Some of Mixed Basic and Secondary 2/ 23,38 Total 160,60 Teachers per Student (TS) Basic 0.05 Secondary 0.07 Share in Recurrent Expenditure 3/ Basic 85% Secondary 15% 100 Notes 1/ As CSO (1998) does not provide data on teachers for GASI, this table use data only for MOE schools 2/ As teachers are typically distinguished between general teachers for grades 1-4 and subject teachers for grades 5-12. In other words, subject teachers often teach both basic and secondary levels. The fore, it Is assumed that simply 5/8 of all teachers are for basic education while the rest (3/8) for secondary educaUon. 3/ The following assumptions were made: (a) No expenditure on public kindergarten and literacy centers as their shares are small In the total expenditures. (b) The difference In unit cost between basic and secondary education is due to the difference in teacher to student raUo as personnel costs are the major part of the total recurrent expenditures Based on these assumptions, the following formula were used for computation: UC-Basic (Unit Cost for Basic Education) = Total MOE Recurrent Expenditure I (STU-Basic + STU-Secondary * (TS-Secondary/TS-Basic)) Share of Basic Educafton in Total Recurrent Expenditure = (UC-Basic STU-BaslcyEXP = STU-Basic (STU-Basic + STU-Secondary ^ (TS-Secondary/TS-Basic)) 60 Annex 7 Table 7.25 - Unit Cost in Basic and Secondary Education by Re ion,1998 In YR Rato to the lowest Basic Secondary Basic Secondary Sana'a City 7,66 11,49 1.0 1.( Sa'adah 8,12 12,18 1.1 1. Al-Jawf 8,16 12,24 1.1 1. Al-Baida 8,36 12,55 1.1 1. Ibb 8,62 12,93 1.1 1. Taiz 9,27 13,91 1.2 1. Al-Mahweet 9,51 14,27 1.2 1. Sana'a 9,78 14,67 1.3 1. HaDah 9,94 14,91 1.3 1. Dhamar 10,49 15,74 1.4 1. Mareb 10,61 15,92 1.4 1. Al-Hodeidah 10,84 16,26 1.4 1. Hadramout 13,92 20,89 1.8 1. Shabwah 14,02 21,03 1.8 1. Laheg 14,09 21,13 1.8 1. Aden 14,82 22,23 1.9 1. Al-Mahrah 15,63 23,45 2.0 2. Abyan 18.09 27.14 2.4 2 Note: Unit cost as per student recurrent expenditure, excluding expenditure on scholarships abroad. 61 Annex 7 Tabhl 7 7iC - eleetiv.e Indlentnrs by h le'Dfle _QQR Docile (1= the poorest 10%) 1 7 :1 4 .r fi 7 R i tn TntMl Household Docile (Based on Per Capita Household Monthly Expenditure) Share in total households %) 10.0 100 10.0 100 100 10.0 100 10.0 10.0 10.0 100 Share in total population (%) 123 115 11.3 11.1 10.7 104 9.6 8.8 80 64 100 Share in population aged 6-14 14.4 127 12.5 115 10.7 105 9.0 7.7 63 4.7 100 Mean household size per household (persons) 8.7 8.1 8.0 7.8 7.6 7.4 6.8 6 2 5.7 4.6 Mean per capita monthly household expenditure (YR) 1,398 2,098 2.638 3,152 3.686 4,310 5,087 6,146 7,862 14.553 50,929 Mean monthly household expenditure (YR) 12,110 17,023 21,158 24,741 27,879 31.679 34,613 38.198 44,270 62,830 314,500 Share in total monthly expendrtures (%) 39 5.4 67 7.9 89 10.1 110 12.1 141 200 100 Share in enrollment basic (X) 13.3 118 11.5 11.3 11 2 10.7 9 5 8.4 7.0 5 3 100 Share in enrollment secondary 11.1 9.6 8.7 10.1 11.6 10.9 86 10.7 9.8 8.9 100 Share in enrollment university 6.5 67 6.1 86 87 10.6 10.0 132 146 151 100 Share in enrollment vt (X) 5.6 55 51 9.1 11 10.0 12.0 131 20.4 8.1 100 Share in total enrollment (%) 128 11 3 109 11.0 11.2 107 9.5 8.9 77 61 100 Mean yearly education subsidy per household (YR) 27,685 24.703 23,710 24,851 25,456 24.765 22,168 21.990 20,320 16,439 232,087 Share in total education subsidies () 11.9 10.6 10.2 107 11.0 107 96 9.5 88 7.1 100 Mean yearly education subsidy per household (X of yearly household expenditure) 19.1 12 1 9.3 84 76 65 53 48 3.8 22 Mean yearly education subsidy per capita (YR) 2,995 2,792 2,682 2,806 2,972 2,958 2,806 2,875 2,903 2,762 Mean yearly education subsidy per capita (% of per capita yearly householdexpenditure) 178 11.1 85 74 67 57 46 3.9 31 16 Individual Decile (Based on Per Capita Household Monthly Expenditure) Share in total population (%) 10.0 100 100 10.0 10.0 10.0 10.0 10.0 100 10.0 1000 Share in population aged 6-14 11.9 11.3 10.8 10.9 102 100 9.8 93 85 74 1000 Share in total monthly expenditures () 2.95 4.38 548 654 7.57 872 10.21 12.10 1528 2676 100.0 Mean per capita household expenditure (YR) 1,308 1.943 2,435 2,896 3,364 3,871 4,523 5,370 6,773 11.897 Share in total monthly expenditures (%) 2.9 4.4 5.5 65 7.6 87 10.2 121 15.3 268 1000 Share in enrollment. basic (%) 108 106 10.0 102 10.2 10.5 10.1 9.9 94 83 1000 Share in enrollment secondary 9.3 78 8.3 80 10.1 10.4 10.6 106 11.6 13.2 100.0 Share in enrollment university 5.5 49 5.4 7.0 6.7 10.2 106 129 150 21.7 1000 Shareinenrollmentvt(X) 4.8 44 5.3 3.0 11.5 102 107 13.0 222 150 1000 Share intotal enrollment(%) 104 10.0 9.6 98 101 105 102 101 10.0 9.4 1000 Mean yearly education subsidy per capita (YR) 3181 3038 2966 3033 3225 3429 3352 3431 3549 3539 Share in total education subsidies (%) 97 9.3 9.1 9.3 98 105 10.2 105 108 108 100 Mean yearly education subsidy per capita (X of per capita vearlvhousehold exanridtture) 20.3 13.0 10.1 87 80 7.4 6.2 5.3 4.4 275 62 Annex 7 Table 7.27 - Selective Indicators by Ilrhan-Rurflnl 1DVtH 1991 Docila (1= tho poor3st 1015) 1 2 3 4 5 7 a 1P Tatrl Urban Individual Decile (Based on Per Capita Houschold Monthly Expenditure) Mean per capita household expenditure (YR) 1654 2360 2890 3379 3910 4533 5314 6419 8290 15223 Share in total monthly expenditures(%) 3.1 44 54 63 72 84 9.8 11.9 154 282 100 Mean yearly education subsidy percapita (YR) 4445 4511 4898 4978 4987 4794 5125 5405 4945 5141 Share in total education subsidies (X) 90 92 99 101 10.1 97 104 110 100 10.4 100 Mean yeariy education subsidy per capita (% of per capita yearly household expenditure) 22.4 15.9 14.1 123 106 8.8 80 7.0 50 28 Rural Individual Decile (Based on Per Capita Housohold Monthly Expenditure) Mean per capita household expenditure (YR) 1243 1846 2317 2765 3220 3700 4297 5094 6352 10858 Share in total monthly expenditures () 30 44 5.6 6.7 78 89 104 123 153 257 100 Mean yearly education subsidy per capita (YR) 2931 2941 2568 2543 2728 2941 2831 2682 2977 2650 Share in total education subsidies (%) 105 106 9.2 92 98 106 102 97 107 95 100 Mean yearly education subsidy per capita (X of per capita yeerly household expenditure) 19.7 13 3 9 2 7 7 7 1 6.6 5.5 4 4 3.9 2.1 Urban Household Decilo (Based on Per Capita Houschold Monthly Expenditure) Share in total population (%) 136 12.5 11.8 11.0 10.5 98 9.2 8.3 7.8 5.5 100 Share in population aged 6-14 (SS) 16.0 14.3 12.6 11.8 10.5 94 8.7 7.0 6.3 3.5 100 Mean monthly household expenditure (YR) 17409 23268 27042 30001 33238 36659 40860 45135 56032 73902 Mean per capita monthly household expenditure (YR) 1809 2627 3233 3825 4482 5255 8262 7694 10140 19882 Mean yearly education subsidy per household (YR) 43266 41843 40980 39039 35970 35963 34020 29931 28588 20252 Share in total education subsidios (%) 12.4 120 11.7 11.2 10.3 10.3 9.7 8.6 8.2 58 100 Mean yearly education subsidy (15 of yearlyhouseholdexpendituro) 207 150 126 10.8 9.0 82 69 5.5 43 23 Mean yearly education subsidy per capita (YR) 4355 4,562 4619 4539 4402 4580 4509 4332 4428 4009 Mean yearly education subsidy per capita (% of per capita yearly household expenditure) 20.1 14.5 11.9 9.9 82 7.2 6.0 47 3.6 1.7 Rural Household Decilo (Based on Per Capita Household Monthly Expandituro) Share in total population (1) 12.0 114 10.9 11.0 11.0 103 9.8 9.1 8.2 6.5 Share in population aged 6-14(%) 142 12.7 117 11.6 11.0 10.0 9.4 82 65 4.7 Mean monthly household expenditure (YR) 11193 15924 19186 23205 27101 2°632 33105 36700 41722 56091 Mean per capita monthly household expenditure (YR) 1329 1980 2495 2979 3491 4057 4792 5729 7238 12573 Mean yearly education subsidy perhousehold (YR) 25798 22273 19751 19638 21955 21122 19720 18128 16652 11781 Share in total education subsidies (%) 13 11.3 100 10.0 11.2 10.7 10.0 9.2 8.5 6.0 100 Mean yearly education subsidy (% of yearlyhouseholdexpenditure) 192 117 86 7.1 68 59 50 41 33 18 Mean yearly education subsidy per capita (YR) 2870 2580 2286 2184 2521 2508 2415 2314 2297 1858 Mean yearly education subsidy per capita (X of per capita yearly hou-sehold exnenditure) 18i0 1QR9 72l6 f1 0 5l 4? 3 4 ? t.2 63 Annex 7 Tsihlp 7 R - Mpnn yenrly hnuuqphnid pYntndituirp nn pdiurvtinn (YR) Households by number Households by number of members who are of currently enrolled in any level of education r suof university students 0 1-2 3-4 .9 or morA n 1 or morp Fees and expenses on nurseries and kindergartens 18 87 33 90 31 426 Fees and expenses on govemment schools 13 501 1.012 1.895 548 841 Fees and expenses on private schools 11 214 564 694 192 1.48C Fees and expenses on university 7 234 562 763 38 4.57E School books 2 38 85 142 40 134 University books 6 115 158 434 15 1.96C Private tuition (expenses on private lessons) 5 19 49 182 32 98 Writing and drawing books and pens 33 657 1.435 2.559 710 1.952 Schoolbags 5 135 314 511 151 341 TOTAL n100 2Q00 4 211 7 270 1 757 11 An Source: Author's calculation using HBS-98. 64 Annex 7 Table 7.29 - Household Snendini, on Education hv Household Dleclie (Contfned) 1 2 a 4 5 a 7 8 P lfl Ave. Fees and expenses on nurseries and kindergartens 3 3 5 22 12 11 8 71 87 291 51 Fees and expenses on govemment schools 492 430 447 569 1.074 628 468 500 493 529 563 Fees and expenses on private schools 8 9 38 28 39 88 84 174 450 1,666 258 Fees and expenses on university 59 52 78 122 146 181 206 266 353 1.238 270 School books 27 29 32 47 46 56 44 39 48 79 45 University books 43 50 42 69 75 133 80 141 212 303 115 Private tuition (expenses on private lessons) 8 15 21 9 21 22 42 57 94 67 35 Writing and drawing books end pens 587 651 647 826 795 891 773 801 807 959 773 School bags 71 98 117 158 154 184 192 196 190 248 161 Total (YR) 1,299 1.336 1.425 1,849 2.384 2,195 1,898 2,243 2.733 5.382 2,272 Distribution ) 5.7 5.9 6.3 81 104 9.7 8.4 99 1Z0 23.7 Household education spending as proportion to household expenditure 09 07 06 0.6 0.7 0.6 05 0.5 05 07 06 Mean yearly education subsidy per household deducted household educaton spending (YR) 26,386 23,366 22,285 23,001 23.093 22,570 20,270 19,747 17.587 11,058 Distribution(%) 12.6 11.2 10.6 11.0 110 10.8 97 94 8.4 5.3 Per capita household spending on education (YR) 1424 149.8 1613 207.0 259.1 262.0 249.6 300.1 410.3 964.9 311 Urban area Fees and expenses on nurseries and kndergartens 12 14 51 19 44 81 275 215 275 1,133 212 Fees and expenses on govemment schools 662 677 724 696 867 735 789 919 793 767 763 Fees and expenses on private schools 19 24 128 157 318 300 248 1,176 2.845 4,716 993 Fees and expenses on university 128 199 320 228 476 350 365 418 1,396 3,550 743 Schoolbooks 27 32 67 69 160 76 90 101 146 114 88 University books 135 105 207 153 202 233 257 432 499 561 279 Private tuition (expenses on private lessons) 19 25 41 53 38 133 97 378 50 225 108 Writing and drawing books and pens 1.420 1,474 1.684 1,670 1,747 1.570 1.740 1,652 1,867 1,540 1,636 School bags 246 272 359 371 403 431 454 375 550 376 384 Total (YR) 2.669 2,821 3,581 3,417 4,254 3,910 4.314 5,666 8,420 12.983 5,203 Distribution N 5.1 5.4 6.9 66 8.2 7.5 8.3 10.9 16.2 25.0 Household education spending as proportion to household expenditure 13 1.0 11 0.9 1.1 09 0.9 1.0 1.3 15 Mean yearly education subsidy per household deducted household education spending(YR) 40,597 39,021 37,400 35,621 31,716 32.054 29,707 24,295 20,168 7,269 Distribution (X) 13.6 13.1 126 12.0 10.6 10.8 10.0 8.2 6.8 2.4 Per capita household spending on education CYR) 266 305 393 393 514 507 600 816 1.161 2.658 761 65 Annex 7 Table 730 - Household Soendine on Education hy Household DecUle I 7 R 4 5 R 7 R 9 in AvAraAm Rural area Fees and expenses on nursenes and kindergartens 2 2 2 5 20 1 0 0 1 0 3 Fees and expenses on government schools 446 416 394 463 601 1,232 350 462 360 309 503 Fees and expenses on private schools 7 8 43 9 23 15 1 84 123 72 38 Fees and expenses on university 58 27 57 53 71 165 164 156 297 239 129 Schoolbooks 32 19 37 30 41 37 27 32 25 35 32 University books 40 10 48 25 34 134 30 81 126 131 66 Private tuition (expenses on private lessons) 7 7 30 3 6 19 17 39 5 10 14 Writing and drawing books and pens 478 508 481 516 614 599 521 536 487 412 515 School bags 48 71 75 97 104 106 119 114 98 108 94 Total (YR) 1,118 1,068 1.168 1,202 1.515 2,309 1,230 1,504 1,522 1,316 1,395 Distribution (%) 80 77 8.4 8.6 10.9 16.6 8.8 10.8 109 9.4 Household education spending as proportion to household expenditure (%) 08 06 0.5 0.4 0.5 0.6 0.3 03 0.3 02 Mean yearly education subsidy per household deducted household education spending (YR) 24,680 21,205 18,583 18,436 20,440 18.813 18.490 16,625 15,130 10.465 Distribution (%) 13.5 11.6 10.2 10.1 11.2 10.3 10.1 9.1 8.3 5.7 Per capita household spending on education (YR) 128 125 133 137 175 253 155 193 230 229 176 Share of private school and private lessons in the total household expenditure an education (%) All 1 2 4 2 3 5 7 10 20 32 Urban 1 2 5 6 8 11 8 27 34 38 R,-rnl I 1 f 1 2 9 I A A A Table 7.31 - Data required for annual calculations of indicators Target 1998 1998- 1999 1999 2000 2005 1997/1998 (HBS-98) 1998E1999 (NPS-99) 1999/20002004/2005 Enrollment in basic educabon Enrollment in basic education, girls Enrollment In basic education, boys Enrollment in grades 1-6, girls Enrollment in grade 1 Enrollment In grade 9 Enrollment in secondary education, girls Enrollment in secondary education, boys Populabon of basic school-age (age 6-14) Populaton of basic school-age (age 6-14), girls Populabon of basic school-age (age 6-14), boys Populaton of lower basic school-age (age 6-11), girls Populabon of age 6 Populabon of age 14 Total education budgets (million YR) Total MOE budget (million YR) MOE recurrent budget (million YR) MOE operatons and maintenance budget (million YR) Number of classes, basic Number of evening/night-shifting classes, basic Number of teachers, basic and secondary Number of female teachers, basic and secondary Number of teachers with at least post-secondary diploma Number of repeters, basic Data for 1998-2000 need to be filled in 66 Annex 7 Table 732 - Per student household expenditure on education, average for households with at least one student Ratio of Dl Decile (D) 1 2 3 4 5 6 7 8 9 10 Total to DIO Share of households at least one student (%) Nationwide 76 71 69 68 70 67 63 60 54 44 64 Urban 85 83 83 80 78 75 72 65 61 45 73 Rural 75 73 64 61 71 63 61 56 52 40 61 Number of students per household for households with at least onte student (persons) Nationwide 3.0 2.8 2.8 2.9 2.8 2 8 2.7 2.6 2.5 2.5 2 8 Urban 3 9 3.7 3 6 3 5 3.3 3.2 3.2 3.0 3 0 2.5 3.4 Rural 2.8 2 6 2 6 2.7 2 5 2 6 2.5 2.5 2 3 2 2 2.6 Per student household expenditure on education (YR/Year): Nationwide deciles 1. Fees and expenses on government schools 218 239 246 326 520 372 278 350 351 523 334 2.4 2. Fees and expenses on nurseries/blndergartens I 1 2 15 6 6 12 44 80 470 47 446.5 3. Fees and expenses on private schools 5 6 15 10 11 34 50 118 265 1,497 149 281.2 4. Fees and expenses on umnversity 34 22 36 49 74 92 190 200 370 1,092 176 32.0 5. School books 14 16 19 29 22 23 24 23 32 74 26 5.1 6 University books 26 20 23 36 33 45 57 95 186 324 72 12.3 7 Pinvate tuition (expenses on private lessons) 5 8 8 4 9 12 20 22 33 54 15 10.7 8. Wnting and drawing books andpens 269 317 338 432 406 472 462 503 586 883 444 3.3 9.Schoolbags 32 51 60 80 72 93 111 121 137 222 91 70 TOTAL 605 681 746 981 1,155 1,149 1,205 1,476 2,041 5,139 1,354 8.5 TOTAL(excluding3.4.6.7) 534 624 664 882 1.027 966 887 1.041 1,186 2.172 942 4.1 Per student household expenditure on education (YR/Year): Urban Nationwide deciles 1. Fees and expenses on govemment schools 204 211 246 261 288 306 327 497 408 720 323 3.5 2. Fees and expenses on nurseries/hlndergartens 4 6 21 8 16 49 167 177 151 1,882 173 536.6 3. Fees and expenses on private schools 8 8 32 44 119 143 172 522 1,356 4,633 503 602.7 4. Fees and expenses on university 32 59 92 66 199 146 209 216 857 2,814 351 88 6 5.Schoolbooks 9 12 19 21 46 29 35 52 82 111 37 12.1 6.Universitybooks 42 33 59 50 83 103 119 304 348 568 144 13.6 7. Pinvate tuition (expenses on private lessons) 8 9 12 21 13 51 40 104 28 154 37 18.3 8. Wnting and drawing books and pens 421 478 592 604 689 647 750 849 989 1,464 702 3.5 9.Schoolbags 75 101 128 141 159 183 212 203 300 316 170 4.2 TOTAL 803 917 1,201 1,216 1,614 1,657 2,031 2,923 4,519 12,664 2,441 15.8 TOTAL(excluding3,4.6.7) 714 808 1.006 1,035 1.200 1.214 1,492 1,778 1,930 4.494 1.406 6.3 Per student household expenditure on education (YR/Year): Rural deciles 1. Fees and expenses on govemment schools 215 235 243 334 385 686 256 340 331 417 338 1.9 2. Fees and expenses on nurseries/kindergartens I I 1 2 15 0 0 0 2 0 2 0.0 3. Fees and expenses on private schools 3 7 19 4 7 7 1 40 102 89 23 27 4 4. Fees and expenses on university 33 22 23 35 54 84 200 103 364 410 114 12.4 5.Schoolbooks 18 11 24 25 25 21 20 24 16 41 22 2.3 6. University books 29 5 24 26 21 48 32 54 133 172 47 5.9 7. Private tuition (expenses on pnvate lessons) 5 7 9 3 3 11 10 14 2 16 8 3.3 8. Wnting and drawing books and pens 248 274 308 345 362 386 359 395 453 507 353 2.0 9. School bags 25 41 48 58 58 63 78 75 90 133 63 5.2 TOTAL 579 603 698 831 930 1,307 955 1,044 1,493 1,787 969 3.1 TOTAL (excluding 3.4.6.7) 508 562 624 765 845 1.156 712 834 892 1,099 778 2.2 67 Annex 7 Table 7.33 - Demographic composition of households by expenditure decile, 1998 1 2 3 4 5 6 7 8 9 10 Total National decile No. of infants (ages 0-5) 1.7 1.6 1.6 1.5 1 4 1.4 1.3 1.1 1.0 0.7 1.3 No. of children (ages 6-14) 3.0 2.7 2.6 2.4 2.2 2.2 1.9 1.6 1.3 1 0 2 1 No. of adults (ages 15+) 4.0 3.9 3.8 3 9 3 9 3.8 3.7 3.5 3.4 2.9 3.7 Total (houshold size) 8.7 8.1 8.0 7.8 7.6 7.4 6.8 6.2 5.7 4 6 7.1 Urban decile No. of infants (ages 0-5) 1 5 1.4 1.4 1.3 1.3 1.2 1.1 0.9 0.9 0.5 1.1 No. ofchlddren (ages 6-14) 3.2 2.9 2.5 2.4 2.1 1.9 1.7 1.4 1.3 0.7 2.0 No. of adults (ages 15+) 5.0 4.6 4 5 4.2 4.1 3.9 3.7 3.5 3.4 2.7 3 9 Total (houshold size) 9.7 8.9 8.4 7.8 7.4 7.0 6.5 5.9 5.5 3.9 7 1 Rural decile No. of infants (ages 0-5) 1.7 1 6 1.6 1.6 1.6 1.4 1.3 1.1 1.0 0.7 1.4 No. of children (ages 6-14) 3.0 2.7 2.5 2.5 2.3 2.1 2.0 1.8 1.4 1.0 2.1 No of adults (ages 15+) 3.8 3.8 3.6 3.7 3.9 3.8 3.6 3.5 3.3 2 9 3.6 Total (houshold size) 8.5 8.1 7.7 7.8 7.8 7.3 6.9 6.4 5.8 4.6 7 1 Table 7.34 - Household expenditures on education, 1998 Urban Rural Among all households Total spending In absolute amount Progressive Progressive % of household's expenditure Progressive Regressive Spending excluding private schoolllessons and university In absolute amount Progressive Regressive % of household's expenditure Regressive Regressive Among households with at least one student Per student spending in absolute amount Progressive Progressive Per student spending excluding private school/lessons and university in absolute amount Progressive Progressive Notes: Progressive if households from the richest decile, on average, pay more than households in the poorest decile ANNEX 8 POVERTY AND PRIVATE AND PUBLIC TRANSFERS IN YEMEN Following a brief description of private transfers and their incidence, this annex concentrates on public transfers, though excludmng pension schemes (which will not be considered part of the safety net in this analysis). It provides a description of the main public programs and critically assesses their coverage and targeting to the extent possible. In particular, the annex exploits recent household level data to see what they reveal about the incidence and targeting of transfers to poor areas and poor people. Unfortunately, this exercise is limited as the surveys collected very little information on participation in public programs. Furthermore, little can be concluded concerning program impacts on poverty since no impact evaluations have been conducted. Methodology of Incidence Analysis As explained elsewhere, there are two household surveys: the 1998 Household Budget Survey (BBS) and the 1999 National Poverty Survey (NPS). There are a number of differences between these surveys. The HBS is a traditional consumption survey which collects exhaustive expenditure and income data over a full year but little inforrnation on other aspects of well-being. In contrast, the NPS has a wealth of information on non-income facets of living standards, but is not ideal for measuring expenditures or incomes. The NPS was completed during the course of one month only, so that it does not capture welfare variability due to seasonality. Also the consumption data in the NPS is based on last week's expenditures compared to the HBS's focus on the last month with specific questions on each of the last 4 weeks. Finally, the NPS covers far fewer consumption items than the HBS. For example, the HBS collects consumption on 20 cereal products compared to 9 in the NPS. We would thus expect the total expenditure variables to be less well measured in the NPS. This may influence the ranking of individuals in the distribution of welfare with implications for conclusions about incidence and targeting of transfer incomes. Against that, comparisons of transfer amounts that can be identified in both surveys indicate remarkably simnilar totals as well as distribution. One would also expect more regular payments such as from the government's Social Welfare Fund to be more accurately collected in the NPS than irregular income sources. In this annex, use is made of both surveys, with preference given to the HBS when information is available in both surveys. However, caveats about the NPS should be kept in mind. Below the population is ranked by household per capita expenditures into national deciles. Hence the deciles are comparable whether reference is being made to the rural, urban or national populations. However, the deciles are not strictly comparable across the two surveys. Unless otherwise noted monetary amounts are expressed in annual per capita riyals (YR). The key to determining whether transfers reach the poor is to assess what their welfare would have been without those transfers. Only then can we know the distributional impact - by seeing the incidence of transfers according to how poor people would have been without them. First, an appropriate indicator is needed to identify the poor. Outcomes may depend on that choice. Studies of the incidence of public spending often subtract the entire amount of government transfer receipts from household income or consumption to approximate pre-intervention welfare, and so rank the population into deciles (say). In the following analysis we will follow this common practice. However, this assumes that there is no replacement through household behavioral responses. That assumption is implausible. The opposite assumption - treating post-transfer consumption as the welfare indicator for assessing incidence - is just as questionable. Ideally, one would like to 69 Annex 8 subtract the intervention amount but add in the replacement income households would have achieved through their behavioral responses had they not benefited from the intervention. In the few studies for other countries (Hungary and Viet Nam), the estimated marginal propensities to consume out of transfer income have been around 0.5 (van de Walle et al. 1994 for Hungary, van de Walle 2002 for Viet Nam). Jalan and Ravallion (2002) also estimate about 50% income replacement for public transfers in Argentina. We will test transfer incidence sensitivity by using this estimate as well as the two others to determine the net gain to consumption from public and private transfers and to construct alternative counterfactual consumption levels. Poverty and Transfers Table 8.1 outlines the sources of private and public transfers that are available to households. Remittances from relatives abroad are by far the largest source of private transfers in Yemen. Although a large drop in remittances occurred after the massive repatriation of Yemeni workers from the Gulf in 1991, these have slowly picked up once again. Inter-household transfers from relatives and friends are also significant. This includes outlays from individuals (primarily male relatives) to their dependents (usually 'unsustained' female relatives and their children) whom they are legally responsible for supportmg. It is unclear how well this system works and what the compliance rate is. Zakat (and Sataqa) are religiously stipulated charitable contributions. Public transfers come from both government programs and from donor-initiated and financed projects. These are detailed below. The only government transfers that can be specifically identified in the HBS are from pension and retirement accounts, while the NPS allows an identification of Social Welfare Fund receipts. Table 8.2 provides an overview of the net per capita monthly money amounts of those transfers that can be identified in the 1998 HBS data - a small subset of the transfers listed in Table 8.1. The amounts represent mean total income from Zakat, domestic and foreign cash and in-kind remittances, transfers from other government organizations and pension and retirement income. Netted from this total are the amounts paid out by the household for Zakat duty, donations and gifts to friends and family and transfers to dependents.3 Table 8.2 shows the sensitivity of the incidence of mean per capita net transfers across deciles under different assumptions about the propensity to consume out of transfer mcome, namely fully excluding, including half only and fully including transfer incomes when assigning households to pre-intervention deciles. Conclusions about targeting and incidence clearly depend on how the counterfactual is defined. Concentrating on deciles defined on per capita expenditures net of transfers in the first 3 colurmns of Table 8.2, the results suggest that net transfers are rather well targeted to the very poorest households. The bottom decile benefits from the largest mean per capita amounts and interestingly, gets five times more than the next poorest decile. All other deciles receive much less. Mean transfers equal 10% of the lower poverty line nationally, ranging from 9% in rural areas tol3% in urban areas. By contrast, transfers to the national, rural and urban population in the poorest decile respectively equal 45, 38 and 86% of the value of poverty lines. In general, there is evidence of an urban bias: each decile's urban population receives a larger absolute amount. 37 Unfortunately, the HBS lumps income from local remittances and transfers from 'other government organizations' together. We do not know what is contained in 'other govermment organizations' but this could well include payments from the government's social welfare fund, for instance. We cannot tell. 70 Annex 8 Focusing instead on deciles defined on the basis of post-transfer expenditures (the last 3 columns), the incidence pattern across deciles is strikingly different with transfers rising with welfare, and a significant concentration of transfer mcome in the richest decile. When deciles are formed netting out half of transfers, there is still a concentration of transfers incomes in the poorest decile though the amounts are lower, and more is going to the higher deciles. A possible explanation for the reversal of transfer concentration from bottom to top decile depending on the decile definition is measurement error. For example, if some transfers have been erroneously inflated they could be dwarfing other expenditures so that when they are subtracted (included) from total expenditures, their recipients all tend to fall in the lowest (highest) decile. To test for this possibility, we trim off the top 2 percent of transfers under the assumption that these reflect errors in measurement. This results in an attenuation of the amounts going to the top and bottom deciles but little change in pattern across deciles. This seems to indicate that the general pattern is correct. Table 8.3 presents information on the percentage of the population living in households that received these income transfers in 1998. Nationally, 33 percent of the population received transfers though that ranged from less than a third of the rural population to almost half of the urban population. Coverage is highest in the poorest decile at 57 and 80 percent of the rural and urban populations respectively when deciles are defined based on pre-transfer expenditures. There is a monotonic decline as welfare rises, but coverage is still high at 19 and 38 percent of the rural and urban populations in the richest decile. When the post-transfer deciles are used, the distribution of beneficiaries is much flatter with only small differences across deciles. Table 8.4 presents total public and private transfers as a percentage of household expenditures. It shows that together, transfers make up around 8 percent of total expenditures for the average Yemeni household. However, transfer incidence is highly progressive, exhibiting an almost monotonic decline as welfare rises. Transfers account for 38 percent of total expenditures for those in the poorest pre-transfer decile nationally. This rises to 56 percent for those in urban areas and drops slightly to 35% for those in rural areas. They are clearly a quite substantial source of livelihood for the poorest Yemenis. As can be observed in Table 8.5 - which shows the share of individual transfer sources in the total - the bulk of these transfers come from private sources, primarily foreign and local remittances. The government transfers that can be identified in the surveys are small and account for a maximum of 41 percent (if there are no local remittances) and a minimum of 10 percent of total transfers (no transfers from 'other government' organizations). However, retirement and pension transfers account for a much larger percent of total transfers in urban areas at a mean of 16%. Table 8.6 also shows that while 8% of the urban population lives in households that receive income from this source, only 2% of the rural population does so. The other significant difference between urban and rural areas is in the incidence of Zakat transfers that clearly favors the urban population with 27% receiving them versus 8% in rural areas.38 Finally a word can be said about transfers to and from dependents. The HBS does not specifically identify such transfers on the income side as they are lumped together with other local remittances and donations from 'other government organizations.' However, it does identify them on the expenditure side. Subtracting the mean amounts paid out from mean income from the HBS category 'local remittances and transfers from other government organizations' reduces the total amount by 7% nationally, 3% and 15% in rural and urban areas respectively. Thus, these transfers are either quite small or underestimated in the HBS's expenditure data. 38 UNDP 1998 notes that Zakat is increasingly treated as general taxation and used to finance general infrastructure and services. It is not clear how this works or what it means for interpretation of the incidence data. 71 Annex 8 In conclusion, the above results suggest that the urban population is favored in terms both of coverage and absolute transfer amounts, and that foreign remittances are the most significant transfers among those that we can identify in the data. I. Government Programs Intent on minimizing the social impacts of an adjustment and reform program, the government instituted a number of social programs in 1995/96. Thus, the programs are relatively recent. Below we discuss the government's key social assistance programs: the Social Welfare Fund and the Agricultural and Fisheries Production Promotion Fund. There continues to be a subsidy on diesel in Yemen which is claimed to be pro-poor. This is also discussed briefly. A number of other small schemes including the Martyr's Welfare Fund (also known as the War Veteran's Fund) which provides assistance to veterans of the 1962 war, and the Tribal Authorities Fund which transfers resources to tribal leaders, are difficult to get information about and are therefore not discussed. i) The Social Welfare Fund (SWF): The SWF is the government's main targeted social assistance program. It was originally conceived in 1996 as a way to compensate the poor for the removal of subsidies. Having little prior experience with this kind of program, the country has struggled with how to best select and reach beneficiaries. An assessment by the World Bank in 1997 found that the SWF suffered from a lack of clear norms, followed extremely bureaucratic procedures with very little follow-up on targeting and beneficiaries, and that too little spending on the program's administration rendered the program particularly weak in rural areas. A field visit revealed poor targeting, widespread ignorance about the program and problems in the distribution of transfers with evidence of middle men pocketing part of the (already small) payments. The SWF's overall budget for beneficiary transfers and administration has also been low. Description of the SWF Under the SWF, cash transfers of YR1000 per beneficiary, plus 200 for each additional dependent up to a maximum of YR2000 per household per month are available to 15 different target groups. The payments are made on a three monthly basis, so that a household receives a maximum of YR24,000 a year. For a family of 6 this translates to about YR333 per person per month or only about 10% of the 1998 national poverty line. Over YR 8 billion were spent on the SWF in 2001 to the benefit of 450,000 households. There is a first stage of geographic targeting. The SWF's Board of Directors decide how many cases they can afford each year and allocate case shares to each governorate based on the incidence of poverty, the share of the country's population and cases of pre-SWF assistance.39 The govemorates are in turn responsible for distributing the cases to the districts on the basis of lists of the eligible. These allocations are likely to be influenced by political considerations. Given the geographical coverage, the program targets the following groups: (1) orphans; (2) widows with children; (3) widows without children; (4) divorced women with children; (5) divorced women without children; (6) single women; (7) the fully disabled; (8) the partially disabled; (9) the poor; (10) the elderly; (11) the temporarily fully disabled; (12) the temporanly partially disabled; (13) families with a missing head of household; (14) families with an 39 The SWF replaced a smaller social assistance scheme run by the Ministry of Insurance and Social Affairs. 72 Annex 8 imprisoned head of household; and (15) families with a head of household recently discharged from prison. In addition to falling into one of these groups, recipients must also be deemed to be without income (income must be the benefits from the SW1F) or income earning potential. This means that those already receiving assistance from the Martyr's Fund, or receiving a pension, for example, are not eligible. There is an attempt to respond to shocks by allowing for both temporary and permanent cases. Beneficiaries in groups I through 10 above are identified as permanent, while those in 11 to 15 are eligible only for temporary assistance for which yearly renewals cannot exceed three years. The law also provides for lump-sum assistance to households who experience personal emergencies or are affected by covariate disasters. According to SWF records, in 2000 the largest number of direct beneficiaries were either the poor, or widows with children. This was the case in all governorates. To receive transfers, potential beneficiaries must fill out applications and provide proof of status and lack of income or earning potential in the form of documentation and various certificates (birth, marriage, age, disability, police testimonial, etc.). Typically, they must come to branch offices to submit their applications. This procedure clearly penalizes the illiterate, elderly, disabled and remote. Certificates can be hard to come by and applications difficult to complete. In principle, efforts are also made to search out the eligible. Social workers from the local offices verify applications but also help identify potential beneficiaries and fill claims. Local NGOs, local councils and Sheiks may also be active in identifying candidates, drawing up lists and informing those in their communities about the program. Even with these additional efforts, it is not clear that those in isolated rural areas and arguably the most vulnerable, are being adequately reached. Local offices do not have the capacity (manpower, vehicles, informnation technology, etc) for exhaustive outreach. Yet, reliance on local Sheiks brings about its own worries including the possibility of capture by friends and clients, with outcomes that do not necessarily favor the highest priority cases. Indeed, the Yemen Voices of the Poor provides some evidence that this is occurring (PRSP 2002). In some areas, poor men and women felt that the real poor were not benefiting, that friends of the Sheik were most likely to do so and that in some cases, bribery and corruption were used to get on the rolls. This is clearly region specific as in other areas, respondents felt that the program reached the right households. At any rate, the SWF budget is generally too low to cover all of the potentially eligible. Even existing applications far outweigh the supply of transfers. (In table 8.7 below, we use the HBS to make an estimate of the target group and existing coverage.) A huge backlog of applications and long delays in dealing with applications is common and caused by a lack of personnel and funding together with a burdensome application and substantiation process. All requests must be checked by local office staff before being forwarded for a second check by the govemorate and then on to Sana'a for final confirmation or rejection. In 1998, the time between a govemorate's approval and a beneficiary's first payment took between 6 and 12 months (World Bank 2000d). Many of those who apply are found to be ineligible. Applicants are disqualified for incomplete or misleading applications. Applications are also rejected if a household member is found to have been begging. In principle, local staff are required to conduct follow-ups every 3 months, as well as yearly when beneficiaries are left or taken off the lists. It is not known how rigorously this is done. Once an application is approved, there is still the issue of getting payments to beneficiaries. Payments are in the form of checks 4 times a year that can be cashed by local cashiers at the local offices or at post offices. The SWF recently started to rely on agents of the Ministry of Finance as local cashiers in residence in every district. This cashier directly pays the transfers to the beneficiaries. This appears to have improved the payment system. The SWF is also increasingly 73 Annex 8 relying on mail delivery of transfers. This is limited to the larger population centers where there are post offices, but should increase in the future as more post offices open. Despite these changes, many beneficiaries clearly continue to have difficulty in receiving and cashing their checks. Those in remote areas, or who have trouble getting around, are likely to continue to have to rely on middle men who take their cut. The SWF has instituted a number of other administrative changes in the past few years. It now has branch offices in all governorates, and in 127 districts. Since 2001, the program has acquired and trained 620 'researchers' who are based at SWF sub-branches throughout the country and help people fill applications, verify applications and update lists of the eligible. The number of researchers is growing though they are not always well trained and tend to be overworked. The researchers spend considerable time checking up on the status of those who are already on the lists to see whether they are still eligible. One of the WB recommendations was to raise the share of total spending going to program administration to 15%. SWF staff state that operational costs are low at 4% of total costs. Others have estimated the administrative costs to be more on the order of 13%. The latter seems more plausible. For one, the SWF's figure consists solely of the amount it spends and does not include the costs faced by the post offices and most importantly, the Ministry of Finance. Although they work for the SWF, the cashiers, for example, are both appointed and paid for by MOF. Only their non-salary costs are covered by SWF. Second, given the complex identification and substantiation process one would expect the administrative cost per applicant to be quite high relative to the benefits. In 1997, the direct costs of identifying and monitoring a beneficiary in rural areas were calculated to be as high as YR1400 for the first year and YR600 in follow-up years (World Bank 1997b). Although these costs may have declined over time as the SWF has become more efficient, the process remains complex and time consuming, so that costs are unlikely to have declined that much. The SWF management feels that there have been many improvements recently but that follow-up remains too slow and should be continuous. A system of fool proof picture ID cards is planned. Effort is also needed to improve the documentation and linkages between the regions, along with coordination with other social welfare programs. Assessment of the SWF In Table 8.7, an attempt is made to estimate the target population from the NPS and match this up with current participation. Our estimates are rough since we can not exactly identify all members of the target groups and can only approximate the way in which the income and asset tests are applied. We define the target group as the population who are very poor (as identified by being in the bottom per capita expenditures decile defined net of SWF payments) and the population who are both poor (as defined as being below the poverty line) and living in households with a severely disabled adult, an elderly man or woman, headed by a widowed, divorced or never married woman. . Table 8.7 shows that 4.2 percent of the target group, or 0.88% of the population, received SWF transfers. 57 percent of those who got the program were not in the target group. Of those who were not in the target group but who got the program, 41percent were non-poor and 16 percent were poor. Table 8.7 strongly suggests that the complicated targeting mechanisms used by the SWF are not working particularly well. A survey which collects information on program participation provides an independent test of public program coverage. The NPS indicates far fewer participants than are claimed by the government. According to their records, the SWF spent YR2.4 billion on 290,000 families in 74 Annex 8 1998 and YR5.2 billion on 350,000 beneficiaries in 1999. The NPS identifies only 51,841 households covering a population of 360,029, as receiving payments in the year from September 1998 to October 1999. Around 5 percent of beneficiary households are found to have more than one direct beneficiary. Payments averaged YR5139 per person per year in recipient households for a total of YR1.85 billion identified in the NPS. There are a number of discrepancies here between the survey and SWF data which we are unable to explain. While, it is possible that survey respondents did not record getting SWF payments, it is unclear why they would not. Table 8.8 provides a picture of the distribution of SWF payments and of participation across deciles defined net of SWF payments. Coverage and amounts received are negligible. The average per capita amount received is worth less than 0.3 percent of the national poverty line at YR 105 annually and only 2 percent of the population lived in households who received payments from the SWF in 1999. Following the pattern of other transfers, there is a concentration of payments and recipients in the poorest population decile, and a subsequent reduction in both as welfare rises. However, there are beneficiary households in every decile. The average amount received in the bottom decile is equal only to 1.2% of the poverty line and only 5% of the poorest ten percent of the population live in households that received SWF benefits. Again, there is an advantage to being in an urban area where per capita payments for the poorest decile are nearly three times higher and 9% of the population received transfer payments compared to 4% in rural areas. Yet, some payments clearly reach the poorest, particularly in urban areas. From a policy perspective, one conclusion could be to simply spend more on the SWF. We can test for the impact of expanding the SWF by increasing the amounts received by current beneficiaries. We increase the total spent by ten times. We find that this would reduce the poverty rate by 0.7 of a percentage point only. This is due both to the low level of payments and the leakage. Of course, such a policy would also need to be financed. Zero cost is assumed above, as might be the case if the expansion were externally financed or entirely financed by the non-poor. But we also experiment with two internally financed scenarios. In the first case, the financing comes from a tax which is proportional to income. This would result in a slightly reduced impact on poverty of 0.6 of a percentage point. If the cost were borne equally by everyone instead, there would be no change in the poverty rate. It is clear that much more would need to be spent on the SWF to lift those who currently receive it out of poverty. The SWF relies primarily on eligible beneficiaries knowing about the program and applying for it. Have Yemenis heard about the program? The NPS asked households this question. The results indicate that only 31% of the population had heard about it in 1999. And this number may overstate the number who understand the eligibility criteria and whether they can apply for the SWF. The NPS also indicates that only 18% of Yemenis (5% in rural and 56% in urban areas) live in an area with a post office (see Table 8.9).40 So, 82% of the population would probably face some difficulty in obtaining benefits, and it is likely that many of these will be among the most needy given that poor people live in areas without a post office. There are also large geographical differences with only 1.9% of those living in Sana'a governorate to 62% of those in Sana'a city having access to a post office. Access to banks is even lower: only 3 percent of Yemen's rural population live in an area serviced by a bank. The poorest groups are likely to have even lower access. This makes one quite pessimistic about using post offices or Bank branches to speed up and facilitate delivery of SWF payments. On the other haind, in some areas, there may be 40 It is not clear how 'area' is defined in 1999 NPS. 75 Annex 8 possibilities of working through schools to which 78 percent of the population have access in their areas of residence. Indeed, only 8 percent say they are 'very far' away from a basic education school. The identification and selection of beneficiaries for the SWF follows a protracted and cumbersome process. There are a number of reasons militating against the current heavy administrative side of identifying beneficiaries in a poor and rugged country like Yemen. The procedures place a heavy burden on the target groups who may be disqualified simply because they can not get the right certificates together or live too far away or have not heard about the program. And as is apparent in table 8.8, the process does not prevent errors of inclusion. Another implication is that the SWF is an ineffective instrument for dealing with shocks since its response rate is much too sluggish. In addition, reliance on local Sheiks may lead to program capture and other problems. There is evidence of this. The program will tend to reinforce the concentration of power and the Sheiks' local control. It would be better to rely on a countervailing institution instead. Another issue concerns the administrative costs involved m the identification and selection of beneficiaries. This appears to be large relative to the benefits. This does not make sense. Given the stringent eligibility criteria, it is probably impossible for the program to check every case each year, as well as add new cases and not make a lot of errors. Serious thought should go into how beneficiary selection and final approval could be decentralized to the governorates or even to the district level. This would speed up the application process. The SWF should also consider simplifying its targeting rules. There should be much finer geographical targeting coupled with the status indicators already used. The income and asset tests should be dropped. These are easy to manipulate, hard to ascertain and highly variant over time, and at any rate, anyone who truly passed them would be barely surviving. ii) Diesel Subsidies Diesel is the only consumer good that continues to be significantly subsidized. It appears to be used primarily to run irrigation pumps, electricity generators and fishing boats. Some in the government argue that the subsidy benefits the poor. Indeed, this underlies the rationale for the government's Agriculture and Fisheries Production Promotion Fund (see below). But others argue that the subsidy has encouraged excessive pumping and inefficient use of water to the detriment of the rural poor. Are the poor benefiting from these subsidies directly or indirectly? Would they bear the brunt of subsidy cuts? Unfortunately, neither data set directly identifies diesel consumption or the ownership or productive use of irrigation pumps and fishing boats. However by looking at the ownership of power generators, the household use of generators for lighting, and whether irrigation is achieved by means of artesian wells - all of which may require diesel - it is possible to indirectly get a sense of who the direct consumers and hence beneficiaries of the diesel subsidy are, Of course, there may also be indirect effects through employment and consumption benefits which we cannot ascertain with the available data. Table 8.10 presents population percentages across deciles who own an electric generator, for whom a generator is the main lighting source at home, and whose land is irrigated by means of an artesian well as opposed to a lace dam, a spring, floods or 'other'. The last two columns show the total percent of the population this implies may be directly using diesel. In all cases, the data suggest that the poor probably consume far less diesel than wealthier households. The population percentages consistently rise with the welfare indicator. For example, while 4 percent of the rural population in the poorest decile live in families whose land is mostly irrigated and is irrigated 76 Annex 8 through an artesian well, 20 percent of those in the top decile do so. Two percent of the rural poorest own a generator versus 11 percent in the top decile. Although as many as 6 percent of the rural population in the poorest decile live in households who may directly consume diesel, 28 percent in the richest decile do so. This evidence is not conclusive but it does suggest that the direct beneficiaries of the diesel subsidy are by and large not the poor. There will no doubt be some poor people who benefit indirectly from the subsidy, though they and other poor people would also benefit (directly and indirectly) from the extra public spending on (inter alia) schools and health clinics that eliminating the diesel subsidy could finance. iii) The Agriculture andi Fisheries Production Promotion Fund (AFPPF) The AFPPF was launched in 1995 m light of worries that increases in diesel prices and eventual elimination of the diesel subsidy would affect the poorest population groups in rural and coastal areas who, both as consumers and producers, are highly dependent on agriculture and fisheries. The Fund aims to promote agriculture, livestock and fisheries production through a wide range of activities in these sectors. These include schemes that subsidize the cost of agricultural inputs and equipment (seeds, fertilizer, tractors etc.), water projects such as dam and construction of smaller works to reduce the risks of drought and recharge aquifers, and production marketing schemes. The AFPPF is financed through a system whereby YR 2.5 (increased from YRI since 1995) is deposited for every liter of diesel sold in the country. Resources also come from the general budget and foreign grants. The yearly budget is around YR 4.5 billion (US$25 to 27 million). The Fund's role is essentially to appraise, approve and finance projects that are formulated by others - namely the agricultural cooperatives, the Agriculture Cooperative Union, the private sector or the local Ministry of Agriculture offices. Once a project is endorsed, these counterparts supervise them, receiving four percent of costs to cover overheads. The AFPPF thus follows a demand-driven format though this is channeled through intermediaries. It is not clear how widespread knowledge about the Fund is or whether all who could productively use it have the opportunity to do so. The terns of financing differ a lot across projects. Some receive full or partial grants, soft loans or loans with subsidized interest rates. The Fund's administration and operation has been reviewed elsewhere (El-Ganmual et al. 1999). This evaluation found that the AFPPF did not meet its objectives in part due to design deficiencies that prevented efficient implementation. Much of the money was found to have gone to parastatals and government who probably have access to other forms of credit, and to building large dams. Projects were found to be badly designed and formulated. The evaluation may now be out of date and did not address the issue of poverty targeting or impact. Unfortunately, data on participation or a more recent evaluation are not available and so the brief comments below are based on conversations with the Fund's head. The program leaders claim that the fund has become more geared to tackling poverty over time. There are currently 34 activities.41 All are seen to indirectly benefit the poor through their employment creation and other benefits, such as increasing the number of crop cycles and the availability of water. Some of the smaller programs are also geared to directly generating income for the poor. Yet, due to a lack of capacity and absorption, expansion of these sorts of programs is limited. One such program emphasizes home-based animal raising. Poor families are given 10 41 No infornation is available on the number of beneficiaries. 77 Annex 8 goats or 5 goats and a cow (for a maximum of YR70,000 per family), taught to raise them and expected to pay back 60% of their value within 2 years. Such schemes entail heavy risks for households who must pay back loans even if the animals die, and so may end up worse off. Another activity encourages the cultivation of date palms by providing credit incentives to farmers to plant trees. These types of schemes may be better at reaching poorer households than some of the other schemes such as building dams but their cost-effectiveness still needs to be looked at carefully. Resources are currently allocated to governorates on the basis of population and poverty indicators; governorates where qat is grown are considered rich and are excluded. However it is not clear if this allocation applies to all of the Fund's resources and where the poverty indicators come from. Within the governorates, the strategy appears to be to focus on one district each year. For the smaller programs, lists of poor people and potential participants are provided by the Sheiks and checked by workers of the local agriculture offices. One of the objectives of the AFPPF is to make non-collateral based credit more easily available to poor farmers. The only private source of credit in rural areas is the Cooperative Agricultural Credit Bank (CACB) which tends to reach richer farmers and agricultural cooperatives. A program like AFPPF needs to ensure that it doesn't squander its resources on those who already have access to sources of credit while missing those who don't and could make productive use of credit. Experience in other countries shows that rich farmers are very adept at cornering the benefits from schemes like the AFPPF and at failing to repay outstanding loans. In making a judgment about this program, more needs to be known about who the direct participants in AFPPF are, what its costs and benefits are and its longer term impacts on poverty. I Donor -Assisted Programs A large number of international agencies and bilateral donors are active in poverty-related development interventions in Yemen. In common with government programs, many of the donor-financed projects were started in the mid 1 990s with the objective of mitigating the adverse effects of adjustment and reform. There appears to be increasing cooperation between many of the donors such as in forming partnerships to better meet their objectives. Below, the main donor financed interventions are described and discussed. There are others that may well have high impacts on poverty, though they are small. Unfortunately, we were not able to cover all donors and interventions. i) The Social Fund for Development (SFD)42 Established in 1997 as a World Bank financed autonomous entity, the SFD was conceived as a demand driven Social Fund aimed at raising living standards and promoting income earning opportunities for the poor. To meet these objectives it has emphasized community development, capacity building and micro-finance programs in poor areas. The community development activity has largely focused on building small scale infrastructure projects to improve access to education, health and water harvesting services using labor intensive techniques. (Feeder road, environment and cultural heritage micro-projects are also allowed but less popular.) This is complemented by support to NGO, government, private sector and community projects that promote the delivery of services. Income generation is supported through providing micro-credit, savings and income-generating programs to the poor through intermediary institutions. An early 42 Further discussion can be found in World Bank (1997) and (2000). This section draws on the latter documents as well as interviews with SFD staff in January 2002. 78 Annex 8 program to help banks develop the capacity for lending to small enterprises has been abandoned. The SFD works through partnerships with third parties such as NGO's, government, cooperatives and other community entities that work closely with communities. This aspect of the project requires the SFD to engage in a fair amount of capacity building geared both to NGOs and CBOs as well as the communities themselves. As of May 2001, the SFD had provided US$90.3 millien to 1,465 projects which it is estimated have benefited 3.4 million Yemenis. By far and away the most common projects are in the education sector - 788 projects with a total commitment of US$46.2 million. Water projects follow at 279 projects and US$17 million committed. On the employment creation side, 9343 permanent jobs and 3,027,097 person days of temporary work were expected to be created from 1997 through 2000. Of those, only 19 percent of the permanent jobs were expected to be filled by women, and even fewer at 0.3 percent of the temporary work days (Republic of Yemen 2000). The SFD is now in its second phase. Over time, it has capitalized on early experience and altered its activities in various ways. For example, the second project scales up assistance to the particularly vulnerable and disadvantaged-notably destitute women, abandoned children and the handicapped. This social protection component promotes informal education and training, support to rehabilitation centers, orphanages and other centers that cater to these groups, child care and literacy training for female pnsoners and so on. The SFD has also attempted to correct for weaknesses in the demand-driven concept of project identification by taking a more active role in targeting marginal groups and the poorest communities who are less well organized or inadequately represented by intermediaries. In these areas, the SFD identifies special needs and works with the communities to address them. Central among these supply-driven special programs are basic education for girls, water harvesting, and integrated development schemes. Sub-districts with the most dismal indicators are first identified. Further targeting is achieved based on village level indicators and probability of an intervention's success. For example, among the sub-districts where 40 or more percent of the population are dependent on rain water collection, water harvesting projects are targeted to the villages where 95 percent do so, and the population is greater than 250. Among the 50 sub-districts with the lowest girls' school enrolments, four were chosen to receive a girls' basic education project based on demand, population density and the proximity of secondary level colleges, and hence, expected effectiveness. Schools are built or rehabilitated, female teachers are trained and communities sensitized to the importance of educating girls. These activities may be accompanied by a water project so as to relieve water collection pressures on girls. In its second phase the SFD is also broadening emphasis from building infrastructure to building the complementary capacity necessary to make a success of the infrastructure. For example, a water harvesting project is always accompanied by efforts to constitute a water user's association that will ensure sustainability and efficient usage. Finally the SFD has also augmented its activities in more remote locations and worked to improve the targeting of its interventions. From the start the SFD has aimed to reach poor communities. Under the first project, resources were allocated across governorates based on a formula combining an index of umnet basic needs constructed with data from the 1994 census (with 0.25 weight) and population density (with 0.75 weight). That has been altered in the second project phase: 30 percent of total resources now go to the supply driven special programs and to target socially marginalized groups. The remaining 70 percent are allocated to govemorates and districts using an index constructed from data on access to basic needs from the 1994 census and 79 Annex 8 data on income poverty from the 1999 NPS, with equal weight. Section m below provides an assessment of the SFD's regional targeting performance. The SFD allocates 87 to 90% of the budget goes to the communities, 5% to overheads and 5 to 7% for consultancies, supervision and capacity building in communities. Importantly, the communities are also required to contribute to the sub-projects in the form of materials, labor time or cash. There are six regional offices and 85 staff (including drivers, messengers etc) who deliver over $20 million a year to 640 locations including extremely remote ones. Past evaluations show that the SFD delivers these services much more cost-effectively than the government or other donor assisted projects. Due to its excellent reputation and success at delivery and implementation, the SFD is often considered the only institution that can address problems and support activities that have otherwise fallen through the cracks in Yemen. It now addresses a plethora of diverse needs and activities including capacity building and setting up a data information system at the SWF, helping the government develop a national social protection strategy, the new social protection component and the special supply-driven interventions. It is not clear that all these activities should be taken on board by the SFD. It will need to be careful not to spread itself too thin and to dilute its effectiveness and impact. However, increased attention to better targeting and its new emphasis on supply-driven interventions to those areas it deems the most disadvantaged is applauded. The 1999 NPS includes a question on whether households have heard of the SFD. Two years after the SFD's inception, only 9% of the rural population lives in households where a member has heard about the SFD and 13% in urban areas. Thus many more are aware of the SWF. This reinforces the need for renewed information campaigns and the emphasis that the SFD now places on non-demand driven special programs targeted to particularly marginalized communities. The SFD claims to use labor-intensive techniques for its small-projects building component. However, this appears to get relatively little emphasis. There certainly seems to be scope for reinforcing the labor intensity of the projects and thereby creating employment for the unemployed rural poor. The SFD is currently embarking on some rigorous evaluations of the impact of its projects. These will be the first such studies in Yemen and should be of great benefit not only to the SFD's own work but to that of other poverty programs in Yemen as well. ii) Public Works Project (PWP) The Public Works Project was established in June 1996 with World Bank funding. It aims to create jobs, provide the poor with small development projects, enhance community participation and develop local contracting firmns. By the end of Phase I in June 2000, US$30 million had been spent on demand driven small scale infrastructure projects such as education and health facilities (mostly rehabilitation or extension of existing facilities), water supply or collection, sanitation, road rehabilitation, vocational training and social security. Community contributions totaled an additional US$2.4 million. Under Phase I, 435 sub-projects were completed, of which 54% were education (accounting for 57% of total spending), 18% on health, 11% on water and 9% on roads. The project covers the entire country and about 70 to 75 % is spent in rural areas. 80 Annex 8 A second project phase began in 1999 and will run to 2003, with $60 million already committed by IDA and the Government. A number of other donors have also come forward with funds to participate in certain sectors in the PWP. Education continues to be the most popular intervention with water overtaking health in second place. Throughout, the works have focused attention on women and children with their highest priority being girls' schooling and reducing the heavy water collection burden shouldered by women in Yemen. The PWP is unable to meet demand. Up till now there have been 12,000 requests, though only 1000 could be provided. In principle, the PWP closely coordinates with the Ministries of Health and Education and the SFD, who undertake to equip and operate the units built by the project. The relationship with the Ministry of Health has been difficult. For example, 16 health units built under phase 1 were non-operational for a long time due to the Ministry's failure to furnish and equip them. On occasion, the PWP also finances the supply of classroom furniture for the schools built, relying on MOE to provide teachers. PWP funds are allocated across governorates according to a formula which distributes 50 percent according to population, 30 percent according to the poverty headcount (using the 1999 NPS) and the remaining 20 percent according to remoteness. This reflects a change from the first phase allocation which was made by the steering committee based on population and remoteness only. The govemorates then allocate funds across districts according to the selection of proposals made by communities. Communities or local councils are expected to identify their needs and submit proposals to the PWP. In some cases, they are aided by local NGOs while in others the PWP sub-branches themselves work to seek out the poorest communities. Participation is clearly very dependent on communities knowing about the project. A new approach whereby requests must come from local councils is under discussion. This would help with the conmmon run-ins with local sheiks who often make demands and want to take over the projects. They frequently apply for the contracts themselves, and make trouble for the contractor - trying to extract money from him or the people - once they are refused. Arrangements can usually be made but 8 projects have been stopped because of disputes with the sheik over placement within the community. Although the PWP is in principle demand driven, exceptions occur when money is accepted from donors who are keen to invest in agency pet projects such as girls' schools in specific regions. Proposals are reviewed and subject to selection criteria including a lower limit of 30% labor content, the sector of intervention, costs below $250,000, community participation, provisions for sustainability and the promise of improving living standards for the beneficiaries. The project has increasingly emphasized sustainability through requiring community cost sharing and setting up local maintenance and operating committees for the infrastructure. Communities are required to pay a minimum of five percent of estimated sub-project costs upfront in cash or in-kind. The average size of projects is $50,000. Although the creation of jobs is said to be the project's main objective, the focus has been on building facilities rather than employment. The PWP was initially meant to be implemented in high unemployment areas (World Bank 2000a). But, unemployment does not appear to be considered in the targeting of resources or the selection of projects. Against that, capital intensive projects are rejected, and smaller contractors who lack heavy machinery are said to be favored. However, only 30 percent of the cost of each sub-project actually goes to labor and 20 percent goes to skilled labor brought in by contractors from outside the communities and districts (World Bank 2000c). Only 10 percent thus goes to locally recruited unskilled labor. As of January 2002, 81 Annex 8 over 2.2 million man-days has been created of which 60% were unskilled labor days. School and health units which account for most of the budget, are less labor intensive than other works. The Bank's annual review for 2000 argues for reducing the focus on employment generation even further because this focus prevents financing equipment and hardware and forces a rejection of many water projects where investment costs are too high. The review essentially argues that providing access to water will improve living conditions (through effects on consumption and production, and time savings) more so than employment generation, and similarly for spending project money on furnishing schools. However, this puts no weight on the potentially important role of the PWP as a short-term safety-net for the poor. By shifting as far as possible toward relatively unskilled labor intensive techniques of production, the PWP would be able to play a more prominent role in dealing with the problem of uninsured risk and transient poverty facing the poor. There may be a tradeoff with longer-term poverty reduction though this can be overstated. Frequently, highly capital intensive techniques are no more efficient in a low wage economy with abundant labor but are chosen for other reasons. iii) Poverty Alleviation an Employment Program (PAEG) The UNDP's PAEG program is an ambitious and complex program that has been under implementation since 1998. It compnses a number of sub-components. An important part concerns the setting up of data information systems (including the Poverty Information and Monitoring System (PIMS) and the Labor Market Information System (LIMS)) which resulted in the 1999 National Poverty Survey (NPS) and the Labor Force Survey (LFS). The objective here is to improve poverty monitoring, targetmg and policy impact assessment. The PAEG also includes the creation of a National Committee for Social Safety Net (NCSSN) that would have responsibility for the major poverty alleviation programs in Yemen. This is as yet not fully operational. Of greater relevance to the present discussion are three sub-components of the PAEG that are more directly aimed at reaching the poor. * The National Programme for Productive Families (NPPF) intends to increase the employment and income earning potential of poor families, and in particular, deprived women, through vocational training and skills development. Training has focused on literacy and women's craft work such as sewing, embroidery, etc. There are 41 training centers around the country with some working through local NGOs and cooperatives. These centers are not all functioning well. The staff is often under-trained, markets do not always exists for the goods created and the ability to integrate graduates into the labor market has tended to remain weak. Finally, as there is no follow-up system for judging the impacts on graduates, longer-term impact is unknown. * The Micro-Start program is a small enterprise and micro-finance development initiative being implemented through four local NGOs, three of which exclusively serve women. Hence, the scheme is heavily focused on women. The four NGOs are also all urban based. Each NGO was given $150,000 of which one third is for operating costs and the rest for starting a revolving fund. Small loans are granted to individuals with prior experience in the area of investment, a good reputation and a guarantee. Past fully paid-up borrowers can take further and larger loans. The focus has been on the sustainability of the revolving fund rather than on small enterprise and livelihood development and sustainability per se. A UNDP evaluation mission deemed the program to have good repayment 82 Annex 8 rates but not a sustainable impact on enterprises. Beneficiaries appear to be poor but far from the poorest. 0 The Regional Development initiative (RegDev) is a community based development scheme designed to empower local communities to develop and help themselves. 'Demonstration' pilot schemes in each of the five major geographical regions involve setting up the institutional framework for community based organizations to flourish, training them in various skills, providing them with social services and technical and financial assistance so that they can engage in income generating and wealth creating small projects. Direct employment generation and livelihood benefits are expected for the entire communities. RegDev is being implemented in partnership with other UN agencies, local NGO's, WFP and others active in the chosen areas. iv) The World ]Food Program (WFP): The World Food Program has long been active in promoting food security in Yemen. In its latest program (2002-2006), it aims to do so by concentrating attention on women and girls as development change agents and by carefully targeting interventions geographically. Targeting follows a three stage prioritization. First, the most vulnerable (as measured by FGT poverty indices based on the 1999 NPS) and the most food insecure (using nutrition indicators) districts are identified. Next, a prioritization is made among those according to local needs in the planned WFP areas of intervention -such as, for example, MCH, girl's education etc. The third level of prioritization considers more practical matters such as accessibility, the security situation, and the presence of complementary activities by other organizations. This results in a concentration of WFP activities in 77 districts in 15 governorates. Forty million US dollars worth of food commodities have been committed so far (50 million more is being sought) to be used to implement the new program's three components and directly benefit 260,310 beneficiaries in the targeted regions. o The first component is nutrition support to malnourished women and children. This is implemented by providing food assistance to pregnant and nursing mothers and children under 5 who are identified by mid-wives as malnourished when they attend health care centers. Chosen beneficiaries are given monthly take-home rations for a set number of months, provided they regularly attend the clinic and stick with the treatment. Within the WFP targeted Districts, health centers are selected for participation based on their ability to provide MCH services and health and nutrition education which is also part of the program. T o The second component of WFP's program is aimed at promoting access to primary education for girls through a food-for-education style scheme. Parents receive a wheat and vegetable oil ration for every three months a daughter attends primary school and for as many daughters as do so. Within the geographically targeted regions, further targeting aims to identify schools in communities with particularly low girl's enrolments, high levels of girl child labor, and the capacity to take in more pupils. There is no targeting within a selected school so that all attending girls receive food transfers. These are distributed each quarter by parents' associations, teachers and head masters with supervision from WFP. In an earlier version of this scheme, the response rate far exceeded WFP's expectations leading to teacher and school size constraints and insufficient food stocks. To avoid running into the same problems, WFP is now 83 Annex 8 working more closely with the Ministry of Education, the Social Fund, donors, NGOs and local communities to expand and rehabilitate schools so that all girls can be accommodated. * The third smaller activity in WFP's program is aimed at supporting the economic empowerment of women by using food transfers as incentives for women to participate in skills training, credit and income-generating activities. WFP helps identify micro-projects that help reduce the burdens, such as water and fuel collection, on women's time This will be coupled with small food-for-work projects open to both genders but which will create assets that directly benefit women. Here too, WFP is working closely with partners including UNICEF, IFAD, the Dutch and the Social Fund. * Finally, WFP continues to be responsible for assisting close to 20,000 Somali refugees who have been in Yemeni camps for the last 11 years. At first sight at least, there are many things to commend in WFP's program of interventions. The focus on nutrition and children is important, given the very high, and possibly rising rates of child malnutrition in Yemen. Fine targeting to certain particularly disadvantaged districts helps to focus resources. Each intervention also includes a degree of self-targeting whereby participants must bear a cost to receive benefits--they must attend clinics and health education briefings, or ensure that their girls attend school or they must provide labor. International experience shows that this helps guarantee that those who do not need the assistance do not participate and hence improves targeting outcomes. It also reduces the administrative costs of identifying participants. Other programs could learn from studying WFP's approach. However, there is still a need to evaluate cost-effectiveness and impacts on living standards. Without that information, it is difficult to be sure that the interventions are fulfilling their aims. v) The Southern Governorates Program (SGP) Former South Yemen implemented a land reform program that was revoked at the country's reunification in 1993. The SGP was initially designed to compensate the small farmers who lost their lands then by allocating new land. From the beginning this was a politically charged project which eventually found that there was, at any rate, no land to give and very limited water in these areas as well. The SGP was reoriented from targeting the land dispossessed to targeting the poor. Implemented with IFAD, the project has become more like a demand driven social fund focusing on community development in 40 communities in the provinces of Hadramout, Abyan, Shebwa and Lahej. Capacity is built through a community development fund to help finance small development projects. Efforts are currently underway to restructure the SGP by working through local councils and by emphasizing small agricultural infrastructure projects. m. Assessing the regional targeting performance of poverty programs For many of Yemen's poverty programs, an analysis of program incidence at household level is not feasible because information on household level participation is not available. However, information on cross-governorate budget allocations is available for the SFD, PWP, and SWF. In addition, there is also information on district level allocations for the SFD. These data, together with provincial and district level poverty measures, allow an analysis of inter-govemorate and, for the SFD intra-governorate, targeting performance. Given recent efforts to better target program allocations to poorer regions, it is of interest to examine geographical targeting performance. Following Ravallion (2000) the 'targeting differential'- interpretable as the mean difference in spending between the poor and non-poor - is estimated. Targeting performance is measured by 84 Annex 8 exploiting the spatial variances in both spending and poverty incidence across govemorates. The inter-governorate targeting differential is estimated by regressing program allocation across governorates on the governorate poverty measure (given by the percent of poor households based on the 1999 NPS). This provides a measure of how well program allocations match the governorate poverty map. The intra-governorate targeting performance for the SFD is estimated based on ordinary least square (OLS) regression of program spending on poverty across all districts within each governorate. Table 8.11 gives the mean estimated amounts going to poor and non-poor households across govemorates and the difference in the two - the targeting differential - for a number of different programs. The SFD's first phase allocations across governorates were not pro-poor. The targeting differential is not significantly different from zero. Although it is clear that it wasn't pro-poor, it is hard to say whether none went to the poor or roughly the same went to the poor as went to the non-poor. Yet, the second phase targeting turns this around completely. Poor households now benefit by about US$90 which is $62 more than the non-poor get on average. The SFD's increased efforts at better geographical targeting have demonstrably paid off. This does not appear to be the case for the PWP. For both phases of the PWP, one can not reject the null hypothesis that the amounts going to the poor and the non-poor are the same; equally well, one can not reject the null that the poor got nothing. These results suggest that the program allocations across governorates are biased against the poor. No improvement is discernible in the second phase of the program after its targeting criteria were revised. The inter-govemorate distribution of SWF allocations shows no signs of pro-poor targeting. Again the amount going to the poor is not significantly different from zero. These results concem allocations across governorates and the extent of pro-poor geographical targeting. However, they say nothing about how the money was spent within each governorate. It is certainly possible that the programs are reaching the poor in richer governorates. Next, Table 8.12 presents the same information for intra-governorate targeting performance for phase I and II of the SFD. This uses data on district level allocations and poverty rates. As noted earlier, in its first phase resources were targeted according to population density with 75% weight and an index of unmet basic needs with 25% weight. Unfortunately, we do not have that composite index and so can not test the degree to which targeting accorded with it. In its second phase, the SFD targets districts on the basis of poverty and an index of unmet basic needs. A test of the allocations against the actual index used shows practically perfect targeting to the poor. However, it is also of interest to ask how well the budget allocations in the two phases were distributed from the point of view of the narrower poverty indicator. Table 8.12 shows a very mixed picture for phase 1. Out of the 12 governorates for which there are sufficient district observations to estimate the targeting differential, for 7 the differences between the amounts going to the poor and non-poor are not significantly different from zero, three have negative differentials indicating that the non-poor are favored and only two have significant and positive differentials. Hadramout appears to perform particularly badly in reaching its poor. Targeting performance changes markedly in the case of the second phase allocations. Table 8.12 reveals consistently positive and significant targeting differentials for phase II except for Hadramout and Abyan where the difference in the amounts going to the poor and non-poor can not be considered significantly different from zero. In all cases, except Mareb, targeting has significantly improved in phase D[. However, there is also quite a lot of variance in these targeting differentials across governorates. Some clearly perform much better in reaching 85 Annex 8 their poor districts. In a few cases (Taiz, Aden, Dhamar and Al-Baida) the entire allocation appears to reach the poor. IV. Overall assessment and recommendations It has been argued that the SWF looks after those who are unable to work and look after themselves, the PWP provides employment to the able-bodied who can work, while the SFD provides long term development opportunities for the poor and that this makes for a complete safety net program in Yemen (World Bank 2000b). Yet a review of these programs indicates that this is not the reality even when they are coupled with other existing poverty and safety net schemes. First, the coverage of these programs is extremely limited compared to the needs and is far from able to work to meet these different needs in the same areas. But even if the programs vastly increased their coverage, they would still not combine to fulfill these aims. The most striking aspect of the programs reviewed is in the similarity of their objectives, format, methods and benefits delivered. The PWP is in many ways indistinguishable from the SFD. With the exception of the SWF, these and practically all other interventions resemble demand-driven social fund type programs where, to varying degrees, the operative emphasis is on 'providing long-term opportunities for the poor.' There's a common focus on women and children, women's education, community development, building small infrastructure projects and bringing services to communities. In many ways these efforts are making up for the social services and physical assets that different government ministries are failing to deliver. Such programs are clearly necessary and many appear to be working well. However, many of the benefits are likely to be longer term. Together the schemes do not fulfill an insurance or traditional safety net role that helps prevent destitution and asset depletion by helping people through shocks and short-term difficulties. They may well also fail to reach the poorest and most needy. Children. For example, may not be sufficiently well protected from poverty and its lifelong consequences. This is a clear deficiency of Yemen's current safety net. In theory, the SWF is an exception. It attempts to provide protection to the worse off among those who are unable to work. Yet, the only program that can address shocks and prevent permanent destitution is administratively cumbersome, slow to respond'and of extremely limited coverage. In addition, the evidence points to extremely weak targeting efficiency of current SWF payments at both govemorate and household level. Poverty impacts are also questionable. The SWF should be reformed to better meet these objectives in practice. It would be advisable to develop this instrument (with the appropriate design changes) to better reach more of the non able-bodied poor and to be more responsive to idiosyncratic shocks and vulnerability issues. As discussed earlier, this requires that, together with more stringent geographical targeting to designated poor areas, the identification and targeting of participants be simplified and broadened. The current effort at fine targeting with both income and asset tests is of doubtful value and much too burdensome administratively. Instead the SWF should work to establish indicators on the basis of poverty data that are more easily verified and transparent as well as difficult to manipulate. Many of these correlates of poverty will be public knowledge in the communities. To better tap into this knowledge and avoid relying solely on Sheiks, one might think about setting up a local women's council in the targeted communities who would be responsible for identifying the eligible based on set criteria, including ones which identify those who may not normally be poor but are hit by some temporarily or permanently debilitating shock. The fact that women figure so highly in the groups of the disadvantaged that are unable to work provides an important rationale 86 Annex 8 for relying on them. They are also likely to have intimate knowledge about the living standards of community members. The SVWF should also add a school attendance requirement for the school- aged children of recipients as a condition for receiving payments. (Obviously, this can and should only be enforced if a school is accessible.) The women's council should be required to consult closely with teachers to ensure that children of beneficiaries are attending school. If not, payments should be cut off. The women's council's recommendations would go to the local council or district office where final decisions on beneficiaries would be made. Following a first stage of fine geographical targeting, more decentralization of this sort for the next stages would be highly desirable and reduce the response time needed to reach beneficiaries. More resources should also be devoted to the SWF. With some redesigning the SWF could perform an extremely valuable role in Yemen's safety net but the targeted amounts are currently far too small to make much difference to the most needy. In addition to these changes to the SWF, thought needs to go into how other instruments can be better developed to serve an insurance role and address vulnerability for the able-bodied poor. Yemen is a country with considerable permanent and seasonal unemployment. Agriculture, from which most of the population derives its livelihoods is subject to harsh enviromnental conditions, precanous access to water and considerable variability. This is an enviromnent in which households are subject to multiple risks that will affect their ability to escape poverty. Many countries have successfully addressed issues of vulnerability and income variability with self- targeted workfare programs. A key player could be the PWP or a new public works program. Either way the scheme would need to increase its self-targeting aspect and be designed to better address vulnerability to seasonal and other income earning and living conditions variability. Again, there should be much better targeting to geographical areas based on levels of unemployment and underdevelopment, as well as targeting on the basis of year round employment variation focusing on the lean periods. It is important to recognize that private assets are important for employment and well-being as does the current PWP. However, public works can also better serve the short-term employment and consumption smoothing function. The program must be willing to face some tradeoffs, for example by building assets that are not as good or sustainable as they might be if less of the project fumds went to employment generation. Other programs such as the SFD appear to be going in the right direction with its increased focus on better targeting and its attempts to reach the most disadvantaged areas. However, it could do more to increase the labor intensity of its small-scale infrastructure projects. A final focus that appears to be missing from Yemen's current safety net has to do with protecting children. Other than efforts at improving schooling, particularly for girls, current programs appear to place little emphasis on child nutrition and other conditions that can cause irreversible damage. The WFP's programs provide an exception, and thought could go into replicating and expanding on some of their health and nutrition and food for education schemes that aim to protect children from the future consequences of current poverty. It is gratifying to see the extensive use of household survey poverty data for targeting fundmg allocations across programs. The SFD also makes good use of census data to target more finely geographically. In theory, this should be further encouraged. However, current practice needs to be vastly improved. As discussed, geographical budget allocations are not pro-poor for most programs for which we have the information. The SFD's phase II is the big exception. The other schemes' targeting criteria clearly need to be revised. it will also be important for the next household survey to focus on collecting a robust measure of expenditures so that the currently used poverty numbers can be checked. The programs should then use the updated regional 87 Annex 8 poverty numbers (adjusted for inflation) to retarget allocations. It will be extremely important to ensure that the next household survey collects the necessary data for effective targeting, as well as information that allows analysis of program participation and incidence. The criteria currently used to target funds geographically appear questionable in some cases. This may explain the lack of pro-poor targeting revealed above. For example, most programs use population density or levels as a criteria, along with poverty rates. This makes little sense: the population weight may simply cancel out the weight given to poverty. It would be better to make per capita allocations a function of the poverty rate only. In addition, there may be a 'fixed cost' argument for allocating more to certain regions, where due to difficult terrain, remoteness and so on, interventions simply cost more. The solution then is to make the total allocation proportional to the total number of poor as long as that allocation does not fall below some amount which represents the fixed cost. This would also do away with the need to use 'remoteness' as a targeting criteria. There is clearly quite poor representation at local level. One recent important change in Yemen is the election of local councils. Many of the programs discussed above are aiming to rely and work much more closely with the councils. This is a promising direction. However, it will be important to monitor these developments and evaluate the role they are playmg in reaching the poor and additional potential for using them or regulating them. Recent moves towards greater decentralization and local councils elections may lead to more effective service delivery but also more local capture. Another promising avenue would be to make better use of radio for advertising demand driven programs and the SWF. A relatively large percentage of the rural population (51%) has access to a radio according to the NPS, though only 31% of those in the lowest decile. There are also pronounced variance across states with 71 percent of Hadramout's rural population having a radio compared to only 31 percent of Al-Hodeidah. Televisions are much less common. Many more impact evaluations and cost-effectiveness studies need to be completed before definitive assessment of the existing transfer programs and recommendations for reform can be made with reasonable certainty. The SFD has taken the lead here, and it is hoped that other programs will follow suit. We recommend that much more emphasis be placed on understanding the costs and benefits of the transfer programs to poor people. 88 Annex 8 Table 8.1 - Poverty and safety net programs in Yemen Public sector Intended beneficiaries Form of benefit Poverty targeting Diesel subsidies Not clear Diesel fuel consumption subsidy No Social Welfare Fund The poor & unable to work w/out income Cash transfers Geographical; means-test + status sources & their dependents. indicator War Veterans Fund 1962 war veterans. Cash transfers No Tribal Authorities Fund Tribal groups Cash transfers n.a. Agriculture and Fisheries Production Poor farmers, pastoralists, & fishermen. Ag production promotion; lower input prices; Geographical; Promotion Fund employ't creation. Social Fund for Development Poor communities; girls & women; vulnerable Community developm't; facilities/unfrastructure Geographical; & disadvantaged women & children. assets; social services; credit; capacity building/ training. Public Works Project Poor communities; the unemployed; girls & Employm't; facilities/infrastructure assets. Geographical; work requirem't. women. World Food Program Girls & women Food transfers; training, credit, income Geographical; clinic/school generation, infrastructure assets. attendance or work requirem't Poverty Alleviation Program Women; poor communities. Credit, training, enterprise promotion; Geographical community developm't. Southern Govemorates Project Rural poor in Southern governorates Community developm't; ag infra assets. Geographical Pension schemes Retired contributors & employees of army, Cash payments no police & gov't Private sector Remittances from relatives workmg abroad Remittances from relatives in Yemen (transfers to dependents) Religious charity donations (Zakat & Sataqa) Other traditional community and kinship- based systems and organizations National non-govenmmental charitable associations 89 Annex 8 Table 8.2 - Distribution of net public and private transfers in 1998 under different assumptions about the propensity to consume out of transfers (annual YR per capita) Welfare Per capita expenditures net of transfers Per capita expenditures net of 0.5* transfers Per capita expenditures with tranfers fully included indicator: net mean per capita transfers net mean per capita transfers net mean per capita transfers 1998 National Rural Urban National Rural Urban National Rural Urban National deciles 1 14757 32942 17347 8879 15924 9707 1181 1651 1233 2 3169 5482 3552 2292 4242 2618 1625 2055 1696 3 2290 4165 2671 2796 4203 3088 1650 2468 1818 4 2158 3925 2528 2639 3519 2826 2331 2311 2327 5 2237 2718 2346 2012 3323 2293 1985 3200 2252 6 985 2601 1352 2353 3524 2620 3246 3693 3350 7 1777 3153 2106 2994 4004 3248 3039 4658 3443 8 1294 3172 1780 2453 4227 2914 5138 4948 5090 9 1475 3987 2146 2766 5167 3419 4860 6400 5288 10 1749 2023 1851 3698 6952 4933 10777 11915 11217 Total 3358 5139 3770 3358 5139 3770 3358 5139 3770 Source: 1998 HBS Note: Deciles are formed by ranking the population by household per capita expenditures under different assumptions about the propensity to consume out of transfers. Net transfers are calculated from income and expenditure on household transfers that can be identified in the HBS-namely, income from zakat, retirement and pensions, local and foreign remittances and payments from government organizations minus transfers given on Zakat, aid to dependents and other gifts and donations. The total household expenditure variable includes expenditures on transfers, so that only transfer income needs to be netted out to get at the "net" amounts. 90 Annex 8 Table 8.3 - Percent of population living in households who received public and private transfers in 1998 Welfare Per capita expenditures net of transfers Per capita expenditures net of 0.5* transfers Per capita expenditures with tranfers fully included indicator: % of population in households who % of population in households who % of population in households who received received transfers received transfers transfers 1998 National Rural Urban National Rural Urban National Rural Urban National deciles 1 57.1 80.3 60.4 43.8 65.6 46.4 30.5 51.1 32.8 2 34.1 63.5 39.0 29.8 59.5 34.8 30.4 52.0 34.0 3 30.7 55.7 35.8 30.8 54.5 35.7 25.7 53.0 31.3 4 25.1 53.5 31.1 26.8 51.4 32.1 27.5 47.9 31.7 5 22.6 47.4 28.2 23.6 48.3 28.9 22.8 49.7 28.7 6 18.8 46.7 25.1 23.1 48.7 28.9 28.9 49.3 33.7 7 25.8 42.2 29.7 28.2 44.2 32.2 27.8 46.7 32.5 8 21.2 45.7 27.5 24.6 48.6 30.9 29.1 49.9 34.4 9 22.3 40.5 27.1 25.7 42.5 30.3 31.5 45.9 35.5 10 19.4 38.1 26.3 22.8 42.2 30.2 29.0 46.8 35.8 Total 28.3 48.8 33.0 28. 3 48.8 33.0 28.3 48.8 33.0 Source 1998 HBS Note: Deciles are formed by ranking the population by household per capita expenditures under different assumptions about the propensity to consume out of transfers. Net transfers are calculated from income and expenditure on household transfers that can be identified in the HBS-namely, income from zakat, retirement and pensions, local and foreign remittances and payments from government organizations minus transfers given on Zakat, aid to dependents and other gifts donations. The total household expenditure variable includes expenditures on transfers, so that only transfer income needs to be netted out to get at the "nef' amounts. 91 Annex 8 Table 8.4 - Public and private transfers as a share of household expenditures Net national Transfers as a percentage of total household expenditures population deciles Rural Urban National 1 35.1 56.0 38.1 2 8.3 12.9 9.0 3 5.8 8.8 6.4 4 4.4 7.8 5.1 5 4.0 5.2 4.3 6 2.2 4.7 2.8 7 3.2 4.6 3.5 8 2.2 4.2 2.7 9 2.0 4.2 2.6 10 1.6 2.7 2.0 total 7.3 8.7 7.7 Source: 1998 HBS. Note* Transfers include income from zakat, retirement and pensions, local and foreign remittances and payments from government organizations. 92 Annex 8 Table 8.5 - Incidence of the percentage of individual transfers in total public and private transfer income in 1998 (YR per year per capita) National Total Zakat Local remittances Foreign Retirement & population transfers & other gov't orgs remittances pension Net decile % of total % of total % of total % of total 1 17520 3.0 26.8 63.6 6.6 2 3688 7.9 28.8 52.2 11.1 3 2816 6.5 34.2 48.8 10.5 4 2752 6.8 35.6 47.3 10.3 5 2659 7.6 30.7 48.9 12.9 6 1747 12.5 42.2 31.0 14.4 7 2532 9.0 34.1 45.2 11.7 8 2207 9.4 34.5 42.4 13.7 9 2745 8.6 33.7 46.7 11.0 10 3557 7.4 36.9 45.7 9.9 total 4224 6.0 31.1 53.5 9.5 Rural 1 14909 3.2 26.0 65.7 5.2 2 3314 6.8 31.3 53.8 8.1 3 2445 4.0 36.9 53.9 5.2 4 2381 4.9 39.6 52.1 3.5 5 2574 5.4 31.7 54.7 8.2 6 1414 12.4 47.0 30.1 10.5 7 2235 7.5 32.2 52.0 8.4 8 1739 7.4 37.5 46.3 8.7 9 2036 8.7 35.0 50.1 6.1 10 2699 5.9 39.6 48.3 6.1 total 3726 5.1 31.5 57.2 6.2 Urban 1 33238 2.6 28.9 58.0 10.5 2 5570 11.3 21.3 47.4 20.0 3 4268 11.9 28.3 37.4 22.4 4 4148 11.1 27.0 37.0 24.9 5 2951 14.1 27.7 31.5 26.7 6 2882 12.5 34.2 32.4 20.8 7 3478 12.2 38.0 31.3 18.5 8 3547 12.2 30.4 36.7 20.7 9 4689 8.3 32.3 42.6 16.9 10 5013 8.9 34.3 43.4 13.4 total 5879 8.0 30.1 45.6 16.3 Source: 1998 HBS Note. Individuals are ranked mto national population deciles based on household per capita expenditures net of transfers receipts. 93 Annex 8 Table 8.6 - Incidence of transfer incomes (%/o ofpopulation) National % or population living in households wbo received: po ulationt Net decile Zakat Local Foreign Retirement charity remittances remittances and pension 1 22.2 35.2 20.9 7.2 2 15.7 20.5 8.4 5.0 3 14.9 17.9 9.1 2.9 4 12.4 16.5 7.1 2.7 5 11.4 13.6 5.7 3.1 6 10.7 13.9 3.2 2.7 7 10.4 17.5 6.7 3.2 8 10.8 15.1 6.0 3.2 9 9.1 15.1 6.4 3.4 10 7.8 15.5 5.9 2.8 total 12.6 18.1 7.9 3.6 Rural' .' - a ; Net decile 1 20.1 35.5 20.1 4.8 2 10.2 21.4 8.4 3.3 3 9.5 17.9 9.6 1.5 4 7.2 15.6 7.1 1.1 5 6.1 13.0 5.7 1.8 6 5.9 12.6 2.4 1.6 7 6.5 17.2 7.2 2.4 8 6.2 12.7 6.1 1.9 9 5.1 13.6 6.3 2.7 10 3.6 12.6 5.2 1.8 total 8.3 17.6 8.0 2.3 Urban - : . . . - Net decile 1 35.2 33.1 26.2 21.8 2 43.3 16.1 8.0 13.5 3 36.0 17.9 7.2 8.3 4 32.0 19.7 7.0 8.6 5 29.8 15.6 5.6 7.3 6 26.9 18.3 5.8 6.6 7 23.0 18.7 5.3 6.0 8 24.1 22.1 5.9 7.1 9 19.9 19.0 6.9 5.5 10 15.0 20.3 7.1 4.5 total 26.7 19.8 7.7 8.0 Source. 1998 HBS. Note: Individuals are ranked into national population deciles based on household per capita expenditures net of transfers receipts. 94 Annex 8 Table 8.7 - Estimated SWF target population and coverage in 1999 (%) Target Non-Targeted Poor Others SWF yes 0.88 0.33 0.84 2.05 SWF no 20.04 18.76 59.15 97.95 total 20.91 19.10 59.99 100.00 Source: NPS 1999 Note: We define the target group as the very poor (as defined as those in the lowest decile net of SWF payents); the population living in households with a severely disabled adult, an elderly man or woman, and headed by a widowed, divorced or never married woman and who are poor (as defined as being below the poverty lIme). The above is a table of individual level obs (n=368001) with sample-weighted percentages for belonging to a household with at least one member receiving SWF transfers against belonging to a target/non-target household. A household is defined as 'targeted' by the program if the per capita expenditures net of SWF payments place it in the bottom decile, or if it is in deciles 24 and has one or more of the following members: a disabled adult beyond school age; an elderly man (over 60) or woman (over 55); a widowed, divorced, married or never married female head (in a household) that is not receiving income from pension and insurance, private domestic remittances or private external remittances. Of the non- target group, the poor are those in deciles 2 to 4, and all else are in 'other.' Table 8.8 - Incidence of SWF payments in 1999 ( YR per year per capita and % ofpopulation) 1999 net Mean per capita SWF transfers % of population in households receiving national SWF deciles Rural Urban National Rural Urban National 1 340 930 449 4.2 8.9 5.1 2 78 143 92 2.2 3.8 2.6 3 68 61 66 2.1 2.3 2.1 4 128 71 114 2.2 2.2 2.2 5 41 82 51 1.6 1.9 1.6 6 46 57 49 1.5 2.3 1.7 7 53 54 53 1.3 1.7 1.4 8 34 37 35 1.4 1.1 1.3 9 52 37 47 1.3 1.2 1.2 10 115 46 89 1.3 0.8 1.1 total 98 124 105 1.9 2.3 2.0 Source: 1999 NPS. Note: Deciles are defined based on per capita household expenditures net of social welfare fund payments. 95 Annex 8 Table 8.9 - Percent of population with amenities and living in an area with various services in 1999 Rural Urban Total % pop with radio 51.4 71.1 56.6 % pop with TV 19.3 80.3 35.3 % pop living in area with: post office 4.8 56.2 18.2 Bank 3.1 47.2 14.6 basic education school 77.8 91.2 81.3 secondary school 36.5 81.7 48.3 prinaryhealth care center 23.8 67.1 35.1 hospital 7.6 61.9 21.8 public transportation 10.9 73.3 27.2 cooperative association 6.6 47.4 17.3 Source. 1999 NPS. Table 8.10 - Population possibly consuming diesel through use of generators for lighting or irrigation in 1999 National % population with % population with % population who total % population population electric generator private generator irrigate with who may use diesel deciles as main lighting artesian well source at home rural urban rural urban rural urban rural urban 1 2.2 1.2 1.1 0.3 4.1 0.7 6.4 2.0 2 2.2 2.1 0.9 0.1 5.4 1.2 7.7 3.4 3 2.8 3.2 1.5 0.2 6.5 1.4 9.2 4.6 4 4.4 1.7 2.2 0.2 7.6 1.8 11.9 3.5 5 3.6 3.6 2.1 0.2 7.9 2.5 11.1 6.0 6 4.0 3.6 2.3 0.3 9.2 1.9 13.0 5.2 7 4.0 2.6 1.8 0.4 12.2 3.2 15.8 5.7 8 6.0 4.5 3.5 0.6 12.4 3.1 17.8 7.6 9 7.0 4.9 4.5 0.4 15.5 3.5 21.9 8.3 10 11.0 8.5 5.3 0.7 19.5 5.8 28.1 13.4 total 4.6 3.9 2.4 0.4 9.7 2.8 13.9 6.5 Source 1999 NPS Note Deciles are defined based on total household per capita expenditures 96 Annex 8 Table 8.11: Program performance In targeting the poor across governo0rates Actual mean per Estimated mean Estimated mean Estimated targeting h'hold allocation amount going to amount going to non- differential poor poor SFD phase I ($US/per 32.7 -66.8 81.2 -148.0 household) (1.0) (2.2) (1.4) SFD phase II ($US/per 47.9 90.2 28.2 62.0 household) (6.1) (3.6) (2.8) PWP phase I ($US/per 12.9 -19.8 28.2 -48.0 household) (0.8) (2.0) (1.2) PWP phase I 33.8 -62.8 79.1 -141.9 ($US/per household) (0.6) (1.4) (0.9) SWF (YR/per 5060.9 -8831.4 11566.9 -20398.3 household/yr) (0.8) (1.8) (1.2) Note: T-ratios in parentheses are based on standard errors corrected for heteroskedasticity. The targeting differential is the difference between the per household amounts gomg to the poor rmnus that going to the non-poor. When an amount is not significantly different from zero, it is set to zero when calculating the targeting differential. 97 Annex 8 Table 8.12: The SFD's perfonnance in targeting the poor across govemorates under phases I and II ($US/per household) Mean per h'hold Mean amount going Mean amount going to Targeting differential allocation to to poor non-poor govemorate Phase I Phase II Phase I Phase II Phase I Phase II Phase I Phase II Ibb 22.1 51.0 14.2 100.5 26.8 29.0 -12.6 71.5 (4.9) (19.9) (12.1) (I1.9) (2.6) (10.0) Abyan 33.4 48.7 -- 71.0 -- 41.9 -- 29.1 (2.2) (3.3) (0.7) Sana'a City 20.5 34.3 -- 90.7 -- 6.8 -- 83.8 (11.2) (2.9) (8.4) Al-Baida 32.6 54.8 73.7 121.9 -0.03 6.3 73.7 115.6 (3.7) (20.4) (0.0) (0.6) (2.1) (7.3) Taiz 19.8 52.2 27.4 108.7 25.9 16.8 1.6 91.9 (2.2) (8.7) (3.3) (1.4) (0. 1) (3.9) Al-Jawf 20.3 55.7 -- 94.8 -- 31.4 -- 63.4 (21.8) (8.1) (7.9) Hajja 22.5 57.2 16.6 93.9 23.2 34.8 -6.6 59.1 (5.5) (48.0) (17.2) (21.3) (1.7) (19.4) Al-Hodeidah 18.3 52.8 539.1 88.1 -212.9 38.0 752.0 50.1 (1.2) (17.7) (1.0) (10.8) (1. 1) (6.7) Hadramout 23.5 51.2 -251.4 35.3 316.4 77.9 -567.8 -42.6 (2.2) (2.2) (2.8) (5.7) (2.5) (1.4) Dhamar 22.2 49.5 16.8 140.7 27.3 13.9 -10.6 126.8 (0.8) (5.6) (3.2) (1. 1) (0.4) (3.3) Shabwah 35.5 50.4 -- 101.6 -- 23.9 -- 77.7 (9.2) (5.6) (5.2) Sa'adah 29.5 50.6 51.8 86.8 24.8 39.2 27.1 47.5 (4.3) (17.6) (5.3) (12.9) (1.7) (6.3) Sana'a 37.0 50.1 15.0 105.1 25.4 26.9 -10.4 78.2 (2.8) (21.4) (10.7) (8.1) (1.5) (9.9) Aden 24.1 32.7 -- 108.1 -- -3.08 -- 111.1 (56.3) (1.7) (31.8) Lahj 26.3 50.2 -- 90.9 -- 36.2 -- 54.7 (7.0) (11.4) (3.4) Mareb 47.6 45.5 151.3 105.1 30.1 33.6 121.2 71.5 (4.3) (8.9) (4.1) (6.1) (3.0) (4.2) Al-Mahwit 31.0 50.6 2.2 104.1 43.1 28.9 -40.9 75.2 (0.3) (8.3) (15.6) (7.5) (4.6) (4.6) Al-Mahrah 122.4 34.7 -- -- -- -- -- -- Amran -- 47.0 16.5 111.4 25.3 27.7 -8.8 83.7 (1.4) (5.1) (7.7) (3.4) (0.6) (2.8) Dhaleh -- 47.5 -- 78.6 -- 37.1 -- 41.5 (8.4) (24.6) (3.8) Yemen 37.9 51.4 87.7 94.3 13.3 29.7 74.4 64.7 Republic (1.4) (36.5) (0.6) (21.6) (0.9) (17.3) Note: T-ratios in parentheses are based on standard errors corrected for heteroskedasticity. The targeting differential is the difference between the per household amounts going to the poor minus that going to the non-poor. When an amount is not significantly different from zero, it is set to zero when calculating the targeting differential. STATISTICAL ANNEX 99 StatisticalAnnex Table I - Household Budget Survey 1998: Sample Size by Governorate (Households) -------------------+-_______-__-___ -+- ___-___ Governorate I Rural Urban I Total --------------------__+_________________-----_+--------- Ibb | 612 530 1142 Abyan 138 208 | 346 Sana'a City 0 2210 2210 Al-Baida 210 269 | 479 Taiz 758 843 | 1601 Al-Jawf - Mareb 95 107 202 Hajjah 416 275 | 691 Al-Hodeida | 480 1414 1894 Hadramout - Al-Mahrah | 614 631 | 1245 Dhamar 407 250 | 657 Sa'adah | 142 210 | 352 Sana'a j 668 327 J 995 Aden | 0 1094 | 1094 Laheg | 342 191 | 533 Al-Mahweet | 133 67 j 200 ----------------------+_______---------------+---------- Total 5015 8626 13641 Source: 1998 HBS. Table 2 - Household Budget Survey 1998: Sample Size by Governorate (Individuals) ____--_______--_--4-------------------- _-----_--_ Governorate I Rural Urban I Total --------------------__+_________________-----_+--------- Ibb 4214 4258 | 8472 Abyan | 1140 1549 | 2689 Sana'a City I 0 15865 I 15865 Al-Baida 1667 2006 3673 Taiz | 4919 5795 10714 Al-Jawf - Mareb | 795 778 | 1573 Hajjah | 3204 1997 5201 Al-Hodeida | 2767 9306 12073 Hadramout - Al-Mahrah | 5369 4985 | 10354 Dhamar 2721 1920 | 4641 Sa'adah 1286 1574 2860 Sana'a 4914 2535 I 7449 Aden | 0 7132 | 7132 Laheg | 2317 1268 3585 Al-Mahweet | 816 447 | 1263 --------------------__+_________________-----_+--------- Total 36129 61415 | 97544 Source: 1998 HBS. 100 Statistical Annex Table 3 - Average household size by PCE decile 1998 Declle Rural Urban National 1 8.55 10.25 8.71 2 8.06 9.42 8.26 3 7.73 9.12 7.98 4 7.71 8.53 7.86 5 7.78 8.26 7.88 6 7.42 7.85 7.52 7 7.10 7.38 7.17 8 6.65 6.97 6.73 9 5.95 6.20 6.01 10 4.73 4.93 4.80 Aggregate 7.08 7.11 7.09 Source: 1998 HBS. Table 4 - Distribution of Population by Deciles Decile Rural Urban Yemen 1 1394083 173241 1567324 2 1307389 259750 1567139 3 1241445 321971 1563416 4 1245238 323623 1568861 5 1220099 343980 1564079 6 1198887 365853 1564740 7 1176683 391324 1568007 8 1167547 398180 1565727 9 1132307 434956 1567263 10 958286 604022 1562308 Total 12041964 3616900 15658863 Source: 1998 HBS. 101 Statistical Annex Table 5- Distribution of Population by Governorates Governorate I Rural Urban Total -------------------------------+-____-______________-________ Ibb j 1639316 268859 1908175 Abyan 350958 79928 430886 Sana'a City 1015248 1015248 Al-Baida 389223 75386 464608 Taiz 1829193 372910 2202102 A1-Jawf - Mareb j 367978 38796 406773 Hajjah I 1243609 115809 1359418 Al-Hodeida j 1103472 564477 1667949 Hadramout - A1-Mahrah - Shabwah 927609 326994 1254603 Dhamar 936286 111763 1048050 Sa'adah 487381 59882 547263 Sana'a 1827019 99317 1926336 Aden j 428039 428039 Laheg 603661 33124 636785 A1-Mahweet 336261 26369 362630 Total j 12041964 3616900 15658863 Table 6- Poverty measures by governorate (Food poverty lne) Yemen Govarnorate I ADCOUNT POVOA FOVGAP2 ------ -------------------------------+- Ibb I 0.32138 0.10517 0.04655 Abyan I 0.27307 0.05391 0.01558 Sana'a City j 0.06114 0.01214 0.00374 Al-Baida 0.05092 0.00913 0.00292 Taiz j 0.28310 0.07565 0.02764 A1-Jawf - Mareb f 0.15176 0.03387 0.01383 Hajjah I 0.08764 0.02162 0.00782 Al-Hodeida 0.14061 0.03031 0.00934 Hadramout - A1-Mahrah - Shabwah 0.20145 0.05134 0.01970 Dhamar 0.18947 0.03794 0.01141 Sa'adah I 0.03502 0.00421 0.00082 Sana'a J 0.11954 0.02637 0.00944 Aden 0.06062 0.00978 0.00258 Laheg I 0.23469 0.04802 0.01387 A1-Mahweet 0.08886 0.01454 0.00408 Table 7- Poverty measures by governorate (Food poverty line) - Urban areas Governorate I awO am1 a=2 Ibb 0.11137 0.02838 0.01006 Abyan I 0.17604 0.03097 0.00935 Sana'a City 0.06114 0.01214 0.00374 Al-Baida j 0.10394 0.03170 0.01277 Taiz I 0.13930 0.03347 0.01160 A1-Jawf - Mareb 0.00000 0.00000 0.00000 Hajjah I 0.07600 0.01362 0.00347 Al-Hodeida I 0.09001 0.01435 0.00344 Hadramout - A1-Mahrah - Shabwah j 0.20983 0.04758 0.01573 Dhamar 0.12884 0.03090 0.01023 Sa'adah I 0.03044 0.00385 0.00101 Sana'a I 0.13056 0.02988 0.00989 Aden 0.06062 0.00978 0.00258 Laheg I 0.15833 0.03059 0.00950 A1-Mahweet 0.29034 0.08581 0.03460 Source: World Bank staff estimates based on 1998 HBS data. Note: poverty measures are calculated at lower poverty lines. 102 Statistical Annex Table 8 - Poverty measures by governorate (Food poverty line) - Rural areas Governorate I a=0 a-1 a,2 ___ ----------------------------+-_______-_-___-______-______-_____ Ibb 0.35583 0.11777 0.05253 Abyan | 0.29517 0.05913 0.01700 Sana'a City I Al-Baida 0.04065 0.00475 0.00101 Taiz 0.31241 0.08425 0.03091 Al-Jawf - Mareb 0.16776 0.03744 0.01528 Hajjah | 0.08872 0.02237 0.00823 Al-Hodeida | 0.16650 0.03847 0.01236 Hadramout - Al-Mahrah - Shabwah 0.19850 0.05267 0.02110 Dhamar 0.19671 0.03879 0.01155 Sa'adah 0.03558 0.00425 0.00080 Sana'a 0.11894 0.02617 0.00942 Aden Laheg 0.23888 0.04898 0.01411 Al-Mahweet 0.07307 0.00895 0.00169 Table 9 - Poverty measures by governorate (Lower poverty line) Governorate HEADCOUNT POVGAP POVGAP2 Ibb 0.55500 0.23375 0.12505 Abyan | 0.53371 0.15891 0.06273 Sana'a City I 0.22884 0.06006 0.02300 Al-Baida 0.15425 0.03871 0.01425 Taiz 0.55584 0.20081 0.09533 Al-Jawf - Mareb 0.26292 0.06718 0.02561 Hajjah | 0.30322 0.07215 0.02652 Al-Hodeida | 0.39840 0.10987 0.04198 Hadramout - Al-Mahrah - Shabwah 0.42638 0.13018 0.05633 Dhamar 0.48510 0.14722 0.06038 Sa'adah 0.26798 0.04956 0.01382 Sana'a j 0.40503 0.10998 0.04440 Aden 0.30192 0.06932 0.02372 Laheg 0.52092 0.17211 0.07276 Al-Mahweet 0.29209 0.07680 0.02694 Table 10 - Poverty measures by governorate (Lower poverty line) - Urban areas Governorate I EADCOUNT POVGAP POVGAP2 ------------------ ----------------------------------------------- Ibb | 0.38449 0.11153 0.04490 Abyan 0.36809 0.09960 0.03739 Sana'a City | 0.22884 0.06006 0.02300 Al-Baida 0.18955 0.06858 0.03286 Taiz 0.34657 0.10560 0.04474 Al-Jawf - Mareb 0.15620 0.01929 0.00279 Hajjah | 0.19292 0.05090 0.01758 Al-Hodeida 0.29584 0.06990 0.02326 Hadramout - Al-Mahrah - Shabwah 0.46372 0.13243 0.05236 Dhamar | 0.40974 0.11370 0.04419 Sa'adah 0.20072 0.03328 0.00898 Sana'a 0.40918 0.11451 0.04544 Aden 0.30192 0.06932 0.02372 Laheg 0.44160 0.12745 0.04829 Al-Mahweet | 0.57950 0.18485 0.08765 Source: World Bank staff estimates based on 1998 HBS data. Note: poverty measures are calculated at lower poverty lines. 103 StatisticalAnnex Table 11 - Poverty measures by governorate (Lower poverty line) - Riarni areas Governorate I HEADCOUNT POVGAP POVGAP2 --------------------------------+-----------------------------______ Ibb 0.58297 0.25380 0.13820 Abyan 0.57143 0.17241 0.06851 Sana'a City Al-Baida | 0.14741 0.03292 0.01065 Taiz 0.59850 0.22022 0.10565 Al-Jawf - Mareb 0.27417 0.07223 0.02801 Hajjah | 0.31349 0.07412 0.02735 Al-Hodeida | 0.45086 0.13032 0.05155 Hadramout - Al-Mahrah - Shabwah | 0.41322 0.12939 0.05773 Dhamar 0.49409 0.15122 0.06232 Sa'adah | 0.27625 0.05156 0.01442 Sana'a 0.40480 0.10973 0.04434 Aden Laheg | 0.52527 0.17456 0.07410 Al-Mahweet 0.26956 0.06833 0.02217 Table 12 - Poverty measures by governorate (UPer poverty line) - Yemen Govornorato I EAIDCOUNT POVWAP POVGAP2 Ibb 0.77652 0.38773 0.23656 Abyan 0.69199 0.25867 0.11964 Sana'a City I 0.54345 0.18932 0.08935 Al-Baida | 0.43804 0.12413 0.05315 Taiz 0.81383 0.39050 0.22505 Al-Jawf - Mareb 0.65476 0.25941 0.12801 Hajjah | 0.66782 0.23269 0.10217 Al-Hodeida | 0.63251 0.22805 0.10562 Hadramout - Al-Mahrah - Shabwah | 0.54977 0.18940 0.08733 Dhamar 0.75213 0.30497 0.15564 Sa'adah | 0.55121 0.14436 0.05320 Sana'a j 0.66963 0.23670 0.10963 Aden | 0.57884 0.18326 0.07781 Laheg 0.72885 0.30313 0.15606 Al-Mahweet 0.37218 0.11342 0.04516 Table 13 - Poverty measures by governorate (UDper poverty line) - Urban areas Governorate I HEADCOUNT POVGAP POVGAP2 Ibb 0.69362 0.26218 0.12770 Abyan 0.57024 0.17789 0.07552 Sana'a City I 0.54345 0.18932 0.08935 Al-Baida 0.40089 0.14771 0.07651 Taiz | 0.63829 0.25158 0.12893 Al-Jawf - Mareb 0.41154 0.08280 0.02427 Hajjah j 0.49412 0.16327 0.07152 Al-Hodeida 0.57463 0.17496 0.07227 Hadramout - Al-Mahrah - Shabwah | 0.57098 0.19272 0.08369 Dhamar | 0.69625 0.26119 0.12478 Sa'adah | 0.41772 0.10034 0.03558 Sana'a 0.64720 0.23573 0.10889 Aden | 0.57884 0.18326 0.07781 Laheg | 0.71696 0.25168 0.11565 Al-Mahweet 0.66678 0.24020 0.11737 Source. World Bank staff estimates based on 1998 HBS data. Note: poverty measures are calculated at lower poverty lines. 104 Statistical Annex Table 14 - Poverty measures by governorate (Unper poverty line) - Rural areas Governorate I HEADCOUNT POVGAP POVGAP2 _--_--------------------------+-_____-- ____-____-_-__-______-_____ Ibb | 0.79012 0.40832 0.25441 Abyan 0.71972 0.27706 0.12969 Sana'a City I Al-Baida 0.44524 0.11956 0.04863 Taiz 0.84961 0.41882 0.24464 Al-Jawf - Mareb | 0.68040 0.27804 0.13894 Hajjah 0.68399 0.23915 0.10502 Al-Hodeida 0.66213 0.25521 0.12268 Hadramout - Al-Mahrah - Shabwah | 0.54230 0.18824 0.08861 Dhamar 0.75880 0.31020 0.15932 Sa'adah | 0.56761 0.14977 0.05537 Sanala j 0.67085 0.23675 0.10967 Aden | Laheg 0.72950 0.30595 0.15828 Al-Mahweet | 0.34907 0.10348 0.03950 Table 15 - Poverty shares by governorate (lower poverty line) - Yemen Governorate I HEADCOUNT POVGAP POVGAP2 __-- ---------------------------_ --------------------. --------__ --__ Ibb 0.16199 0.21568 0.26069 Abyan 0.03518 0.03311 0.02953 Sana'a City j 0.03554 0.02948 0.02551 Al-Baida 0.01096 0.00870 0.00724 Taiz 0.18722 0.21382 0.22935 Al-Jawf - Mareb 0.01636 0.01321 0.01138 Hajjah | 0.06305 0.04742 0.03938 Al-Hodeida 0.10164 0.08861 0.07649 Hadramout - Al-Mahrah - Shabwah 0.08182 0.07898 0.07721 Dhamar 0.07776 0.07461 0.06914 Sa'adah 0.02243 0.01311 0.00826 Sana'a 0.11934 0.10244 0.09344 Aden 0.01977 0.01435 0.01109 Laheg 0.05074 0.05300 0.05062 Al-Mahweet 0.01620 0.01347 0.01067 Table 16 - Distribution of poor and non-poor by governorate - Yemen Governorate I non poor poor Total Ibb 849132 1059043 1908175 Abyan | 200918 229968 430886 Sanala City 782914 232334 1015248 Al-Baida 392942 71666 464608 Taiz 978092 1224010 2202102 Al-Jawf - Mareb 299824 106949 406773 Hajjah 947218 412201 1359418 Al-Hodeida 1003440 664509 1667949 Hadramout - Al-Mahrah - Shabwah 719665 534938 1254603 Dhamar 539642 508408 1048050 Sa'adah 400606 146657 547263 Sana'a 1146115 780220 1926336 Aden | 298805 129233 428039 Laheg | 305074 331711 636785 Al-Mahweet | 256708 105922 362630 Total 9121094 6537770 15658863 Source. World Bank staff estimates based on 1998 HBS data. Note poverty measures are calculated at lower poverty lines. 105 Statistical Annex Table 17 - Distribution of poor and non-poor by governorate - Urban areas Governorate I non poor poor Total --------------------------------+-____-_____-______________ Ibb 165486 103372 268859 Abyan | 50507 29421 79928 Sana'a City I 782914 232334 1015248 Al-Baida 61096 14289 75386 Taiz | 243672 129238 372910 Al-Jawf - Mareb 32736 6060 38796 Hajjah 93467 22342 115809 Al-Hodeida j 397484 166993 564477 Hadramout - Al-Mahrah - Shabwah 175361 151633 326994 Dhamar | 65969 45794 111763 Sa'adah | 47863 12020 59882 Sana'a 58678 40639 99317 Aden 298805 129233 428039 Laheg | 18496 14628 33124 Al-Mahweet | 11088 15281 26369 Total | 2503623 1113277 3616900 Table 18 - Distribution of poor and non-poor by governorate - Rural areas Governorate I non poor poor Total -------------------------------+-_____-_____-________________ Ibb | 683646 955670 1639316 Abyan | 150411 200548 350958 Al-Baida 331846 57377 389223 Taiz 734420 1094772 1829193 Al-Jawf - Mareb 267088 100889 367978 Hajjah | 853750 389859 1243609 Al-Hodeida j 605955 497516 1103472 Hadramout - Al-Mahrah - Shabwah 544304 383305 927609 Dhamar j 473673 462613 936286 Sa'adah | 352744 134638 487381 Sana'a j 1087437 739581 1827019 Laheg | 286578 317084 603661 Al-Mahweet 245620 90641 336261 Total j 6617471 5424493 12041964 Source: World Bank staff estimates based on 1998 HBS data Note. poverty measures are calculated at lower poverty lines. 106 Statistical Annex Table 19 - Distribution of PCE by decile, Yemen 1998 Mean Expenditure Cumulative (%) Average food budget share (%) Decile (YR/person/year) Urban Rural National Urban Rural National Urban Rural National 1 16656 15578 15697 3.07 3.00 2.95 50.3 64.1 62.6 2 23575 23272 23322 7.44 7.44 7.34 50.4 62.9 60.9 3 29238 29223 29226 12.80 13.04 12.82 50.1 61.8 59.4 4 34898 34717 34754 19.06 19.70 19.36 49.4 60.0 57.8 5 40284 40394 40370 26.31 27.46 26.93 49.7 59.6 57.5 6 46503 46438 46453 34.70 36.38 35.65 48.0 57.6 55.3 7 54285 54281 54282 44.56 46.74 45.86 46.7 58.1 55.2 8 64611 64386 64443 56 44 59.01 57.96 46.0 57.9 54.8 9 81224 81292 81273 71.81 74.34 73.24 44.2 55.4 52.3 10 151183 137462 142767 10.0 100.00 100.00 39.3 51.2 46.6 Total 64752 49779 532371 - - 46.5 59.2 56.2 Source: 1998 ]HBS Table 20 - Distribution of Population and Mean PCE by Governorate - Yemen Mean per capita Mean of per capita Governorate Pop. expenditure expenditure of the (YR) poor [bb 0.12186 3646 1889 Abyan 0.02752 2969 1940 Sana'a City 0.06484 6719 2472 Al-Baida 0.02967 6728 2356 Taiz 0.14063 3698 2082 Al-Jawf- Mareb 0.02598 4450 1927 Hajah 0.08681 4945 2435 Al-Hodeida 0.10652 4040 2110 Hadramout - Al-Mahrah - Shabwah 0.08012 4429 2274 Dhamnar 0.06693 4207 2384 Sa'adah 0.03495 4994 2814 Sana'a 0.12302 4546 2420 Aden 0.02734 5310 2554 Laheg 0.04067 3711 2197 Al-Mahweet 0.02316 4910 2228 Total 1 Source- 1998 HBS. 107 Statistical Annex Table 21 - Distribution of Population and Mean PCE by Governorate - Urban Areas Governorate Pop. % Mean PCE (YR) Mean PCE of the poor Ibb 0.07433 4658 2288 Abyan 0.02210 3349 1974 Sana'a City 0.28070 6719 2472 Al-Baida 0.02084 6776 1999 Taiz 0 10310 5181 2221 Al-Jawf- Mareb 0.01073 6313 2972 Hajjah 0.03202 5767 2138 Al-Hodeida 0.15607 4666 2140 Hadramout - Al-Mahrah - Shabwah 0.09041 4236 2342 Dhamnar 0.03090 4569 2459 Sa'adah 0.01656 5799 2830 Sana'a 0.02746 4393 2379 Aden 0.11834 5310 2554 Laheg 0.00916 4018 2352 Al-Mahweet 0.00729 3833 2080 Total I Source. 1998 HBS. Table 22 - Distribution of Population and Mean PCE by Governorate - Rural Areas Governorate Pop. % Mean PCE Mean PCE of the poor (YR) Ibb 0.13613 3480 1845 Abyan 0.02914 2883 1936 Sana'a City Al-Baida 0.03232 6719 2445 Taiz 0.15190 3396 2065 Al-Jawf - Mareb 0.03056 4253 1864 Haliah 0.10327 4868 2452 Al-Hodeida 0.09164 3720 2100 Hadrarnout - Al-Mahrah - Shabwah 0.07703 4497 2247 Dhamar 0.07775 4164 2376 Sa'adah 0.04047 4895 2812 Sana'a 0.15172 4554 2423 Aden Laheg 0.05013 3694 2190 Al-Mahweet 0.02792 4995 2254 Total I Source: 1998 HBS. 108 Statistical Annex Table 23 - Household Expenditures by Deciles (current YR, per capita, per year) Decile Food Housing Clothing Health Education Transport and Leisure Other __________ __________ ____________ ~~com m unications _ _ _ _ _ _ _ _ _ _ _ 1 9814 2608 1137 196 150 361 1225 206 2 14191 3725 1715 450 166 536 2148 392 3 17374 4470 2190 587 188 700 3058 659 4 20101 5259 2566 654 229 977 4135 833 5 23197 5753 3006 902 274 1237 4781 1220 6 25696 6600 3491 1082 374 1643 6085 1482 7 29973 7459 3921 1232 313 2125 7504 1755 8 35347 8263 4945 1494 361 2686 9117 2231 9 42471 10246 5874 2707 476 4370 11683 3447 10 63922 19269 9899 5834 1175 9981 23087 9600 Total 28200 7362 3873 1513 370 2460 7279 2180 Source: World Bank staff estimates based on 1998 HBS data. Note: Food does not mclude 'Tobacco and Qat'. Table 24 - Household Expenditures by Deciles - Urban Areas (current YR, per capita, per year) Decile Food Housing Clothing Health Education Transport and Leisure Other communications Lesr Otr 1 8417 5405 721 101 251 271 1025 465 2 11853 6460 1263 380 302 579 1984 753 3 14662 7326 1701 460 317 725 3000 1046 4 17244 8345 2299 543 379 917 3796 1375 5 20013 9055 2707 704 494 1297 4554 1460 6 22292 9840 3217 831 528 1782 6016 1998 7 25314 11512 3997 963 635 2393 7018 2453 8 29689 12608 4999 1515 625 3268 9117 2789 9 35827 15677 6463 1979 904 4156 12326 3893 10 56814 25887 11673 5678 2374 10904 25957 11895 Total 28064 12820 4775 1730 839 3444 9420 3661 Source: World Bank staff estinates based on 1998 HBS data. 109 Statistical Annex Table 25 - Household Expenditures by Deciles - Rural Areas (current YR. per capita, per year) Decile Food Housing Clothing Health Education Transport and Ldisure Other communications 9987 2261 1189 208 138 373 1250 173 2 14655 3181 1804 464 139 528 2181 320 3 18077 3730 2317 620 155 693 3073 558 4 20844 4456 2636 683 190 993 4223 692 5 24095 4822 3090 957 212 1221 4845 1152 6 26735 5612 3575 1159 327 1601 6106 1324 7 31522 6112 3895 1322 206 2036 7666 1522 8 37277 6781 4926 1486 270 2487 9117 2041 9 45023 8160 5648 2986 311 4452 11436 3276 10 68402 15098 8780 5932 420 9399 21278 8153 Total 28241 5723 3602 1447 230 2164 6635 1736 Source: World Bank staff estimates based on 1998 HBS data. Table 26 - Distribution of Household Budget (%) by Deciles Decile Food Housing Clothing Health Education communications Leisure Other 1 62.6 16.9 7.2 1.2 1.0 2.2 7.6 1.3 2 60.9 16.0 7.4 1.9 0.7 2.3 9.2 1.7 3 59.4 15.3 7.5 2.0 0.6 2.4 10.4 2.3 4 57.8 15.2 7.4 1.9 0.7 2.8 11.9 2.4 5 57.5 14.3 7.4 2.2 0.7 3.1 11.8 3.0 6 55.3 14.2 7.5 2.3 0.8 3.5 13.1 3.2 7 55.2 13.7 7.2 2.3 0.6 3.9 13.8 3.2 8 54.8 12.8 7.7 2.3 0.6 4.1 14.2 3.5 9 52.3 12.6 7.2 3.3 0.6 5.3 14.4 4.3 10 46.6 13.5 7.1 3.7 0.8 5.9 16.0 6.4 Total 56.2 14.4 7.4 2.3 0.7 3.6 12.2 3.1 Source: World Bank staff estimates based on 1998 HBS data. I10 Statistical Annex Table 27 - Distribution of Household Budget (%) by Deciles - Urban Areas Decile Food Housing Clothing Health Education Transport and Leisure Other communications Lesr Otr 1 50 3 32.9 4.4 0.6 1.6 1.6 5.9 2.7 2 50.4 27.4 5.3 1.6 1.3 2.4 8.3 3 2 3 50.1 25.1 5.8 16 11 25 10.2 3.6 4 49.4 23.9 6.6 1.6 II 2 6 10.9 3.9 5 49.7 22.5 6.7 1.7 1.2 3.2 11.3 3.6 6 ~ 48 0 21.2 6.9 1,8 1 3,8 12,9 4.3 7 46.7 21,2 7.4 1.8 1.2 4.4 12.9 4.5 8 46.0 19.5 7.7 2.3 1.0 5.0 14.1 4 3 9 44.2 19.3 8.0 2 4 1.1 5.1 15.1 4.8 1 0 39.3 17.5 8.2 3.2 1.5 6.4 16.9 7.1 Total 46.5 21.9 7.0 2 0 1.2 4.1 12.7 4.5 Source World Bank staff estimates based on 1998 HBS data. Table 28 - Distribution of Household Budget (%) by Deciles - Rural Areas Decile Food Housing Clothing Health Education Transport and Leisure Other 1 64.1 14.9 7 5 1.3 1.0 2 3 78 1.1 2 62.9 13.7 7.8 2.0 0 6 2.3 9 3 1.4 3 61.8 12.8 8.0 2.1 0.5 2.4 10.5 1.9 4 60.0 12.9 7.6 2.0 0.5 2.9 12.2 2 0 5 59 6 11.9 7.6 2.4 0.5 3.0 12.0 2.9 6 57.6 12.1 7.7 2.5 0.7 3.4 13.2 2.8 7 58.1 11.2 7.2 2 4 0.4 3.7 14.1 2.8 8 57 9 10.5 7.7 2.3 0.4 3.8 14.2 3.2 9 55.4 10.1 7.0 3.7 0.4 5.4 14.1 4.1 10 51.2 10.9 6.5 4.1 0.3 5.6 15.5 5.9 Total 59.2 12.2 7.5 2.4 0.6 3 4 12 1 2.7 Source World Bank staff estimates based on 1998 HBS data. 111 StatisticalAnnex Table 29 - Household Expenditures by Governorates - Yemen (current YR, per capita, per year) Governorate Food Housing Clothing Health Education Transport and Leisure Other communications Lesr Otr Ibb 23798 5478 3473 1066 303 1689 5975 1972 Abyan 22482 4719 2523 784 182 1477 2554 908 Sana'a City 30999 16301 5992 1874 1410 5854 13447 4744 Al-Baida 42500 6441 5548 2283 244 5685 13519 4517 Taiz 23167 6760 3868 2014 456 1598 4899 1615 Al-Jawf - Mareb 27884 7733 4007 1498 83 2679 8190 1322 Hajjah 31879 5954 4282 2122 263 2041 10936 1860 Al-Hodeida 27591 6513 2491 1109 195 1114 8052 1416 Hadramout - Al-Mahrah - Shabwah 32024 7746 3524 1098 222 3232 3354 1947 Dhamar 28344 7188 4230 994 274 1223 5357 2878 Sa'adah 34272 7096 4220 876 136 1910 7409 4005 Sana'a 28638 6702 3508 1110 293 3595 8787 1912 Aden 27137 14607 4733 2614 683 4088 6842 3019 Laheg 24832 5778 3699 1668 437 1581 5166 1373 Al-Mahweet 33951 6317 5086 3735 286 2124 5798 1626 Total 28200 7362 3873 1513 370 2460 7279 2180 Source: World Bank staff estimates based on 1998 HBS data. 112 Statistical Annex Table 30 - Household Expenditures by Governorates - Urban Areas (current YR, per capita, per year) Governorate Food Housing Clothing Health Education Tmnsport and Leisure Other communications _____ Ibb 22600 10039 4488 3137 625 2523 7702 4783 Abyan 22911 8886 4448 345 402 614 2095 483 Sana'a City 30999 16301 5992 1874 1410 5854 13447 4744 Al-Baida 37793 10214 5540 3213 298 2614 16298 5344 Taiz 25558 11289 5467 2157 1306 3322 8720 4354 Al-Jawf -Mareb 36653 14134 6174 1290 346 2133 12219 2809 Hajah 30432 13221 5309 875 821 2363 13087 3099 Al-Hodeida 26719 11806 3010 955 482 1658 9383 1974 Hadramout - Al-Mahrah - Shabwah 29301 9326 3401 828 301 2317 2536 2824 Dhamar 24657 10322 4768 1610 479 1064 6535 5399 Sa'adah 33658 11921 5567 659 252 2791 11046 3691 Sana'a 23765 9386 3837 1080 298 2067 9516 2769 Aden 27137 14607 4733 2614 683 4088 6842 3019 Laheg 23176 10010 4214 666 447 668 5176 3854 Al-Mahweet 23537 9575 3394 319 314 1428 5624 1802 Total 28064 12820 4775 1730 839 3444 9420 3661 Source: World Bank staff estimates based on 1998 HBS data 113 StatisncalAnnex Table 31 - Household Expenditures by Governmorates - IRural Areas (current YR, per capita, per year) Governorate Food Housing Clothing Health Education Transport and Leisure Other communications LIsr OTe Ibb 23995 4730 3307 727 250 1552 5692 1510 Abyan 22384 3770 2084 884 132 1673 2659 1005 Al-Baida 43412 5710 5549 2103 234 6279 12981 4356 Taiz 22679 5836 3542 1985 283 1247 4119 1057 Al-Jawf - Mareb 26960 7058 3779 1520 55 2737 7765 1166 Hafah 32014 5277 4187 2238 211 2011 10735 1744 Al-Hodeida 28037 3806 2226 1188 47 837 7370 1131 Hadrarnout - Al-Mahrah - Shabwah 32983 7189 3567 1193 194 3555 3643 1638 Dhamnar 28784 6814 4165 921 250 1242 5217 2577 Sa'adah 34347 6503 4055 902 122 1801 6962 4044 Sana'a 28903 6556 3490 1112 293 3678 8748 1866 Laheg 24923 5545 3670 1723 436 1631 5165 1237 Al-Mahweet 34768 6061 5218 4003 284 2179 5812 1613 Total 28241 5723 3602 1447 230 2164 6635 1736 Source: World Bank staff estimnates based on 1998 HBS data. 114 Statistical Annex Table 32 - Distribution of Household Budget (%) by Governorates - Yemen Governorate Food Housing Clothing Health Education Transport and Leisure Other communications Ibb 56.8 13.7 8.2 1.7 0.9 3.1 12.3 3.3 Abyan 65.2 13.2 7.1 2.1 0.6 3.9 6.5 1.5 Sana'a City 40.6 23.2 7.1 1.8 1.5 5.9 15.5 4.6 Al-Baida 56 4 9.4 6.1 2 5 0.4 5.5 15.2 4.6 Taiz 55.0 16.1 9.1 3.7 1 0 2.8 9.4 2.9 Al-Jawf - Mareb 58.6 13.5 7.6 1.3 0 1 3.3 14.1 1.5 Hajah 54.7 10.9 7.9 3.0 0.5 3.2 17.1 2.8 Al-Hodeida 60.4 13.1 4.9 2.1 0.4 1.9 15.4 1.9 Hadramout - Al-Mahrah - Shabwah 63.8 14.5 6.3 2.0 0.5 4.7 5.1 3.3 Dhamar 58.1 14.9 8 1 1.6 0.6 2.2 10.0 4.4 Sa'adah 58.6 12.2 6.7 1.5 0.2 3.1 11.7 6.0 Sana'a 55.3 13.2 6 4 1.6 0.6 4.8 15.5 2.6 Aden 44.1 24.9 7.5 2.6 1.2 5.7 10.0 4.0 Laheg 58.4 13.7 8.7 2.5 1.1 2.8 10.3 2.6 Al-Mahweet 58.5 10.6 10.0 6.0 0.5 3.3 9.3 1.9 Total 56.2 14 4 7.4 2.3 0.7 3.6 12.2 3.1 Source- World Bank staff estimates based on 1998 HBS data. 1 15 Statistical Annex Table 33 - Distribution of Household Budget (%) by Governorates - Urban Areas Governorate Food Housing Clothing Health Education Trnunicand Leisure Other communications Ibb 43.1 20.4 8.3 3.8 1.3 3.0 12.5 7.6 Abyan 57.7 22.5 11.1 0.7 1.1 1.3 4.6 0.9 Sana'a City 40.6 23.2 7.1 1.8 1.5 5.9 15.5 4.6 Al-Baida 49.8 15.0 6.7 3.1 0.6 2.8 18.0 4.0 Taiz 43.7 21.5 8.3 3.1 2.1 4.3 11.7 5.4 Al-Jawf- Mareb 52.3 19.7 7.6 1.6 0.4 2.1 14.0 2.2 Hajjah 47.4 19.2 6.9 1.3 1.1 2.5 18.5 3.0 Al-Hodeida 50.6 22.3 4.9 1.4 0.8 2.4 14.6 3.0 Hadramout - Al-Mahrah - Shabwah 60.5 19.1 5.8 1.4 0.6 3.6 4.0 5.1 Dhamar 46.6 21.4 8.0 2.3 1.0 1.6 11.1 8.1 Sa'adah 49.7 18.1 7.4 1.0 0.4 3.8 14.6 4.9 Sana'a 46.9 19.6 6.9 1.7 0.7 2.9 17.3 4.0 Aden 44.1 24.9 7.5 2.6 1.2 5.7 10.0 4.0 Laheg 49.4 21.6 8.4 1.2 1.1 1.2 10.1 7.2 Al-Mahweet 53.5 25.0 7.4 0.8 1.0 1.7 8.3 2.3 Total 46.5 21.9 7.0 2.0 1.2 4.1 12.7 4.5 Source: World Bank staff estimnates based on 1998 HBS data. 116 Statistical Annex Table 34 - Distribution of Household Budget (%) by Governorates - Rural Areas Governorate Food Housing Clothing Health Education Transport and Leisure Other communications Ibb 59.1 12.6 8.2 1.4 0.8 3.1 12.3 2.6 Abyan 66.9 11.1 6.2 2.4 0.4 4.5 6.9 1.6 Al-Baida 57.7 8.3 6.0 2.3 0.3 6.1 14.6 4.7 Talz 57.3 15.0 9.3 3.8 0.8 2.5 8.9 2.4 Al-Jawf- Mareb 59.2 12.8 7.6 1.3 0.1 3.4 14.1 1.4 Hajjah 55.4 10.1 7.9 3.2 0.4 3.2 17 0 2.8 Al-Hodeida 65.4 8.4 4.9 2.4 0.1 1.7 15.8 1.4 Hadramout - A1-Mahrah - Shabwah 65.0 12.9 6.4 2.1 0.4 5.0 5.5 2.7 Dhamar 59.5 14.2 8.1 1.5 0.6 2.3 9.9 3.9 Sa'adah 59.7 11.4 6.6 1.6 0.2 3.0 11.3 6.2 Sana'a 55.8 12.8 6.4 1.6 0.6 4.9 15.4 2 5 Laheg 58.9 13.3 8.7 2.5 1.1 2.9 10.3 2.3 Al-Mahweet 58.9 9.5 10.2 6.4 0.4 3.4 9.4 1.9 Total 59.2 12.2 7.5 2.4 0.6 3.4 12.1 2.7 Source: World Bank staff estimates based on 1998 HBS data. 117 Statistical Annex Table 35 - Age Composition of the population by gender in 1998 I ----------- Rural ---------- ----------- Urban ---------- ----------- Total ---------- Age group I female male Total female male Total female male Total -----------+-___ _______ ____________________________________-______-______-_______-_____________-_____ 0 - 4 895565 944712 1840277 229905 241159 471064 1125469 1185871 2311341 5 - 9 1045802 1146464 2192266 281119 284096 565214 1326921 1430560 2757481 10-14 906616 1001268 1907883 275855 294766 570621 1182470 1296034 2478504 15-19 680794 698789 1379582 235208 243648 478856 916002 942437 1858438 20-24 417882 421228 839110 162951 167488 330439 580833 588716 1169549 25-29 398942 349067 748009 131085 120173 251258 530027 469240 999267 30-34 302394 216428 518822 98254 87597 185851 400647 304025 704672 35-39 339808 310274 650082 104434 97862 202296 444242 408136 852377 40-44 242475 193065 435541 67526 66379 133906 310002 259445 569446 45-49 j 200078 189579 389657 57365 61850 119215 257443 251428 508871 50-54 168021 154384 322404 47791 51442 99233 215812 205825 421637 55-59 I 87010 104043 191053 26364 33778 60142 113374 137822 251195 60-64 98584 130468 229052 26073 30990 57062 124656 161457 286114 65 or more 171494 226735 398228 40432 51312 91744 211926 278047 489973 Total 5955462 6086502 12041964 1784360 1832539 3616900 7739822 7919041 15658863 Table 36 - Distribution of poor and non-poor population by age in 1998 I ----------- Rural ---------- ----------- Urban ---------- --- Total ---------- Age group I non poor poor Total non poor poor Total non poor poor Total 0 - 4 971550 868727 1840277 331664 139400 471064 1303214 1008127 2311341 5 - 9 1078460 1113807 2192266 372385 192829 565214 1450845 1306636 2757481 10-14 966705 941179 1907883 365281 205340 570621 1331986 1146519 2478504 15-19 782963 596620 1379582 319144 159712 478856 1102106 756332 1858438 20-24 518776 320335 839110 236421 94018 330439 755197 414353 1169549 25-29 479421 268588 748009 188833 62425 251258 668254 331013 999267 30-34 288336 230486 518822 137532 48319 185851 425867 278805 704672 35-39 341661 308421 650082 149638 52658 202296 491298 361079 852377 40-44 j 242314 193227 435541 94782 39124 133906 337095 232351 569446 45-49 232430 157226 389657 84239 34976 119215 316670 192202 508871 50-54 196914 125490 322404 73024 26209 99233 269938 151699 421637 55-59 117293 73761 191053 43512 16630 60142 160805 90391 251195 60-64 145044 84008 229052 42571 14491 57062 187615 98499 286114 65 or more 255608 142620 398228 64598 27147 91744 320206 169767 489973 Total 6617471 5424493 12041964 2503623 1113277 3616900 9121094 6537770 15658863 Source: World Bank staff estimates based on 1998 HBS data. Table 37 - Distribution of poor and non-poor population by educational level of the head of the household Educational status ----------- Rural ---------- ----------- Urban ---------- ----------- Total ---------- of the hh's head non poor poor Total non poor poor Total non poor poor Total --------------------------+-____--_____--___----___________________________________________________-_____-______________ Illiterate 3506558 3336631 6843189 789268 523856 1313124 4295827 3860487 8156313 Read and Write 2166327 1494066 3660393 709375 315372 1024748 2875702 1809438 4685141 Primary l 186586 131045 317631 144914 79362 224276 331500 210407 541907 After primary | 49427 36286 85713 13297 8673 21970 62723 44959 107682 Preparatory/basic 241835 175473 417308 164221 56378 220598 406056 231851 637906 Pre-high school Diploma | 52264 13454 65718 32647 14723 47370 84911 28177 113088 High School 280306 139952 420258 293176 70045 363220 573482 209997 783478 Post-high school Diploma | 54552 48979 103531 76532 12550 89082 131084 61529 192613 Undergraduate and graduate 79617 48607 128224 280194 32319 312512 359810 80926 440736 Total 6617471 5424493 12041964 2503623 1113277 3616900 9121094 6537770 15658863 Table 38 - Distribution of the population by educational level of the head of the household and decile - Yemen decile I Illiterate Read & Write Primary Lower Secondary Secondary Higher Total _________-+---__------------------------------------------------__-----------__-------------------------------------_---_________ 1 I 1049306 376883 42430 45482 51220 2004 1567324 2 I 944212 444709 54001 58490 34829 30899 1567139 3 913402 424346 48507 66364 60245 50554 1563416 4 832593 457064 65121 107950 64909 41224 1568861 5 807834 525938 33691 94159 50985 51474 1564079 6 I 834051 457183 71807 79560 80388 41751 1564740 7 765866 488049 56097 82073 96340 79583 1568007 8 749908 496206 45887 97221 100956 75549 1565727 9 689980 489410 70970 116389 101919 98596 1567263 10 569164 525354 53398 110988 141687 161717 1562308 Total 8156313 4685141 541907 858676 783478 633349 15658863 Table 39 - Distribution of the population by educational level of the head of the household and decile - Urban Areas decile I Illiterate Read & Write Primary Lower Secondary Secondary Higher Total _________-+------------------------------------------------------__-----------__-----------------------------------------________ 1 102296 41114 11488 6866 10221 1257 173241 2 128525 74512 13835 20492 15243 7145 259750 3 I 152207 91012 27620 18883 19950 12300 321971 4 I 134771 100444 23284 28837 17990 18297 323623 5 139046 107055 21358 29560 26962 20000 343980 6 144935 100615 26347 34700 32118 27139 365853 7 133719 114377 25248 33957 38948 45075 391324 8 121054 112286 18362 40021 47948 58511 398180 9 I 113918 122115 25330 32676 64514 76404 434956 10 142653 161218 31407 43947 89328 135468 604022 Total 1313124 1024748 224276 289938 363220 401594 3616900 …-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Source World Bank staff estnmates based on 1998 HBS data. 119 Statistical Annex Table 40 - Distribution of the population by educational level of the head of the household and decOle - Rural Areas docile I Illiterate Read & Write Primary Lower Secondary Secondary Higher Total 1 947010 335769 30942 38616 41000 747 1394083 2 815687 370197 40166 37999 19586 23754 1307389 3 761195 333334 20887 47481 40295 38254 1241445 4 I 697823 356620 41837 79113 46920 22927 1245238 5 1 668788 418883 12333 64599 24023 31474 1220099 6 689116 356568 45461 44861 48270 14612 1198887 7 I 632147 373672 30849 48116 57392 34508 1176683 8 I 628854 383921 27525 57201 53009 17039 1167S47 9 I 576062 367295 45640 83713 37405 22192 1132307 10 I 426510 364136 21991 67041 52359 26249 958286 Total | 6843189 3660393 317631 568738 420258 231755 12041964 Table 41- Distribution of the population by occupational status of the head of the household, 1998 employment ----------- Rural ---------- ----------- Urban ---------- ----------- Total ---------- status non poor poor Total non poor poor Total non poor poor Total ---------------------+-____________________________________________________________________________________________ Employed, wage earner 1494803 1869709 3364511 1183128 566314 1749441 2677931 2436022 5113953 Self-employed 3884360 2621232 6505592 736703 273918 1010621 4621062 2895150 7516212 Employer | 210281 87844 298125 112791 22699 135489 323072 110542 433614 Employed, other 46386 46613 92999 3365 1159 4524 49752 47772 97523 Unemployed f 173456 121908 295364 65690 39196 104885 239146 161103 400249 Homemaker/Student 236444 233982 470426 143281 55865 199146 379726 289847 669572 Rentier 263498 166308 429806 169677 85365 255041 433175 251672 684847 Disabled 308243 276898 585141 88989 68763 157752 397232 345662 742894 Total j 6617471 5424493 12041964 2503623 1113277 3616900 9121094 6537770 15658863 …__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _- - - - _ _ _ _ _ _ _- _ _ _ _- - _ _- - _ _ _ _ _ _- - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _- - _ Source: World Bank staff estimates based on 1998 HBS data. 120 Statistical Annex Table 42 - Distribution of the population sector of employment of the head of the household and decile decile I Agriculture Manufactur Costructio Industry, oth. Merchandis Public Adm Services, oth. Total 1 626360 65083 196176 13565 254383 134543 33567 1323676 2 668602 75761 110660 18659 244865 137979 74371 1330895 3 621102 86921 94763 35252 274619 167634 69550 1349840 4 623670 60256 90258 20634 311304 161322 76474 1343917 5 634360 75540 84961 20413 283464 160023 70699 1329459 6 665400 60791 91917 25547 290809 142277 63138 1339879 7 1 606266 58610 75428 22391 347253 179471 68842 1358260 8 j 616104 66248 67017 17170 304936 180088 87455 1339018 9 527810 86037 73661 18208 389091 179612 69549 1343967 10 473581 69721 77117 11971 390627 211511 93268 1327797 Total 6063255 704966 961957 203809 3091350 1654458 706911 13386708 Source: World Bank staff estimates based on 1998 HBS data. Table 43 - Poverty by sector of employment of the head of the household mean PCE of Poverty Headcount Poverty Squared Household head activity sector pop. share (%) mean PCE the poor share index Gap Poverty the poor (%/) (%) Index Gap Agnculture/Forestry/Fishery 45.3 4111 2185 47.3 43.5 13.8 6.1 Industry, Manufacturing 5.3 4427 2275 5.7 45.2 13.7 5.7 Industry, Costruction 7.2 3773 2031 9.4 54.6 20.5 10.2 Industry, other 1.5 4011 2327 1.7 47.5 13.5 5.4 Services, Merchandise 23.1 4916 2244 19.9 36 10.8 4.6 Services, Public Admmnistration 12.4 4934 2239 11.1 37.4 11.4 4.9 Services, other 5.3 4895 2345 4.8 38 10.4 3.9 100.0 100.0 Source: World Bank staff estimates based on 1998 HBS data. 121 Statistical Annex Table 44 - Adjustments of 1?etroleunm ?LdEcts betvjeeE 29C-2G3 Adjustment during 19 0-1995 (YR/Liter) Ac rNE:stn2!r c_Lq 9 .-001 YR/L, 1990-1994 1995 Jn-96 Ma-561 Jz-97 Oct-97 1rn-98 May-9 J!-01 Benzene (Gasoline) 6.0 12.0 19.0 19.3 25.0 25.0 35.0 35.0 35 0 Diesel 3.0 3.0 9.0 6 0 6.0 10.0 10.0 10 0 17.0 Kerosene residential 3.0 3.0 7.0 8.5 13.0 13.0 15.0 16 0 16.0 Kerosene (Aviauon) 4.0 12.6 12.6 12.6 12.6 12.6 13.0 13.0 13.0 Mazut-cement (heavy fuel) 2.5 5 0 7.0 7.0 7.0 11.0 11.0 13 0 13.0 Mazut-Electricity 1.72 3 _ Gas (LPG) 50.0 73 0 140.0 150.0 150.0 150.0 200.0 200.0 220.0 Source: Althoubet Magazine. Table 45 - Average Electrkc!y C˝riff, L995-2Z0 Jan-95 Aug-95 Apr-96 Jul-97 Sap-97 3r.uI-93 lca-99 2CCD 20019 Electncity (per K.w.H.)j 2.07 2.07 5 40 5.40 5.80 8.70 8.70 8.70 9.5 Source: IMF and stafFs estimation Table 46 - Macroeconomic Indlentors Gn YYemen, t991-2=2 Year GDP Growth, Non-Oil GDP growth, ˘:l VaIL-Addfd :flniox., OK, IFisc˘ l36ace _________ (%) (%) GcwGth, (C%) f%) _ 1991 20 3 2 -5.6 44 9 -3.5 1992 8.3 11.7 -15.4 50.6 -11.9 1993 4.1 40 42 54.8 -12.8 1994 2.2 -23 427 71.3 -147 1995 10.9 94 19 9 62.5 -5.2 1996 59 45 135 400 -09 1997 8.1 8.2 7 5 4.6 -1 5 1998 4.9 5.4 25 11.5 -7 9 1999 3.7 29 78 80 0.1 2000 5 1 4.7 7.2 8.5 7 9 Source: World Bank, LDB (2001). Table 47 - Sectoral Contributlon to GDP I Yeme˝ 2 1S91I-2G Sector _ 199C-1992 I 1995-200 | 1990-2030 Total GDP Growth Rate (%) 48 6.4 5 9 Agnculture Growth Rate (%) 6.3 5 8 59 Share m GDP (%) 22.1 17.8 19 0 Contnbution to GDP Growth (%) 1.4 1 0 1.1 Share in GDP Growth (%) 28.9 16.0 19 3 Industry Growth Rate (%) 4 4 8.8 7.3 Share in GDP (%) 23 3 37.6 33 3 Contnbution to GDP Growth (%) 1 0 3 3 2 4 Share in GDP Growth (%) 21 0 51 9 41.4 Services Growth Rate (%) 4 4 5 4 5.0 Share in GDP (%) 54.7 44 6 47 7 Contribution to GDP Growth (%) 2.4 2.4 2 4 Share m GDP Growth (%) 50.1 37 3 41.0 Source: Computed by staffbased on data from the World Bank, LDB. 122 Statistical Annex Table 48 - Consumption in Yemen, 1990-2000 l 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Total Consumption CurrentYRBn 115.4 159.6 191.5 258.3 301 1 5024 580.9 688.6 700 7 9264 991 3 YR Private Consumption Current YR Bn 93.3 130.8 154.3 212.8 243.5 428.4 483.4 571.8 576.2 7702 7972 YR Private Consumption Current YR Bn 7,856 9,753 11,136 14,869 16,472 28,094 30,807 35,426 34,714 45,177 45,535 Per Capita YR TotalConsumption 1990YRBn 115.4 140.4 147.6 166.7 159.3 177.0 150.2 156.7 188.3 182.8 156.1 Public Consumption 1990 YR Bn 22.1 25.3 28.7 29.3 30.5 26.1 25.2 26.6 33.5 30.8 30.6 PrivateConsumption 1990YRBn 93.3 115.0 118.9 137.3 128.8 150.9 125.0 130.1 154.9 151.9 125.5 Population Millions 11.9 13.4 13.9 14.3 14.8 15.3 15.7 16.1 16.6 17.0 175 Private Consumption 1990 YR Bn 7,856 8,578 8,582 9,595 8,713 9,897 7,964 8,063 9,329 8,913 7,171 Per Capita I Table 49 - Labor Force Participation by Sector and Gender Labor Force Participation (%) Urban Area 40.0 Male 68.0 Female 11.5 Rural Area 483 Male 70.7 Female 25.9 Total 45.9 Male 69.9 Female 21.8 Source: 1999 Labor Force Survey. Table 50 - Trends In Employment in Yemen, 1990-2000 Year Economically Labor Employment Unemployment Unemployment Active Force Rate Population _gricultur Industry Services Total _ 1990 6,128,000 2,782,000 1991 6,383,000 2,914,000 1992 6,648,000 3,048,000 1993 6,934,000 3,240,000 1994 7,796,000 3,428,000 1,593,000 353,000 1,098,000 3,044,000 384,000 11.2 1995 8,095,000 3,552,000 1,650,000 363,000 1,130,000 3,143,000 410,000 11.5 1996 8,422,000 3,690,000 1,708,000 377,000 1,147,000 3,232,000 458,000 12.4 1997 8,753,000 3,841,000 1,805,000 389,000 1,200,000 3,394,000 447,000 11.6 1998 9,079,000 3,993,000 1,886,000 400,000 1,234,000 3,520,000 473,000 11.8 1999 9,431,000 4,091,000 1,959,000 403,335 1,259,000 3,621,335 469,000 11.5 2000 9,777,000 4,221,000 2,038,000 418,200 1,295,000 3,751,200 470,000 Ti1 Source: Ministry of Planning. Table 51 - Trends in Yemen's Civil Service Employment (thousands) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999e Number 168 203 235 267 295 322 328 334 336 352 % growtli 7% 14% 12% 9% 8% 2% 2% 0% 4% Civil administration, permanent and contract workers. Source: MOCSAR (Public Employment Survey) and WB staff calculations. 123 Statistical Annex Table 52 - Trend in Real Wages in Yemen's Civil Service Sector, 1990-2000 Total Wage Civilian Wage Civil Average Nominal CPI Real Wage Bill (YR Bill Servants Annual Wage Wage Inflation Increase (%/6) Billlon) (YR Bllllon) _ _ (YR) Increase (%/.) (%/6) 1990 19.0 10.4 168,000 62,153 .. 33.51 ._. 1991 24.2 13.3 203,000 65,607 5.6 44.90 -39.3 1992 31.2 17.0 235,000 72,398 10.4 50.63 -40.3 1993 39.3 22.8 267,000 85,489 18.1 54.85 -36.8 1994 46.8 26.6 295,000 90,101 5.4 71.28 -65.9 1995 63.7 33.6 322,000 104,458 15.9 62.51 -46.6 1996 74.0 47.8 328,000 145,642 39.4 39.96 -0.5 1997 82.2 54.4 334,000 162,919 11.9 4.64 7.2 1998 93.6 59.6 336,000 177,393 8.9 11.54 -2.7 1999 119.1 78.9 352,000 224,226 26.4 7.97 18.4 2000 141.6 95.2 365,000 * 260,921 16.4 8.50 7.9 * estimate. Source: (i) Ministry of Finance for the wage bill; (ii) CBY for CPI inflation rate; and Ministry of Civil Service for the number of civil servants. 124 Statistical Annex Table 53 - Public Spending in Yemen, 1992-2002 (YR Billions unless otherwise stated) 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 * 2002* Total Government 44 56 68 87 118 228 303 294 336 476 502, 532 Spending As % of GDP 29.2 29.3 28.4 283 23.1 30.9 34.0 34.2 28.7 31.7 32.71, - .28.i Current Expenditures 37 47 59 76 99 178 244 232 267 383 3.368, 382 As % of GDP 24.3 24.5 24.7 24.9 19.4 24.1 27.4 26.9 22.7 25.5 ',24.0 20.6 As % of Total Spending 83.3 83.7 87.1 87.9 84.1 78.0 80.7 78.6 79.3 80.5 -73.2 .71.8 Investment Expenditures 7 9 9 10 19 50 59 63 70 93 134 . 1150" As%ofGDP 4.9 4.8 3.7 3.4 3.7 6.8 6.6 7.3 5.9 6.2 8.8 8.1 As%ofTotalSpending 16.7 16.3 12.9 12.1 15.9 22.0 19.3 21.4 20.7 19.5 26.8 '28.2 Wages and Salanes 24 31 39 47 64 74 82 94 119 142 166, 178i. As%ofGDP 16.0 16.2 16.5 15.3 124 10.0 9.2 10.9 10.2 9.4 10.8l, 9X.6 As%ofTotalSpending 549 55.5 58.1 53.9 53.9 32.5 27.1 31.8 35.4 29.7 331 33.5 Subsidies and Transfers 4 5 7 8 9 53 89 69 54 122 ,38 30 As%ofGDP 2.4 2.6 2.8 2.6 1.8 7.3 10.0 8.0 4.6 8.1 225 1.6 As%ofTotalSpending 8.3 8.9 9.9 9.1 7.6 23.5 29.3 23.4 16.0 25.5 7.6 '5.6 Total Social Spending IS 23 30 34 47 115 163 164 164 272 212 1,2,3,4,5 and subsidies) As % of GDP 11.7 12.1 12.4 10.9 9.2 15.8 18.3 19.0 13.9 18.0 '13.7 As % of Total Spending 40.2 41.5 43.8 38.6 40.0 50.8 53.7 55.5 48.9 56.9 42.2 '.. Sofal Spending 14 18 23 26 38 62 74 95 '110'' 150I , 174" (excluding subsidies & transfers) 'As % of GDP 9.3 9.5 9.6 &4 7.5 &5' &3 11.0 9.3, 9.9' 'I.L, As % of Total Spending 31.9 32.5 33.9 29.6 324 2Z4 244 321 32.9' 31.4 1 '346 1. Education Spending 9 11 14 17 23 37 46 57 67 89 108 As%ofGDP 5.6 5.6 5.7 5.5 4.5 5.1 5.2 6.6 5.7 5.9 -7'00.. As % of Total Spending 19.3 19.0 20.1 19.3 19.5 16.4 15.2 19.3 20.0 18.7 ,21.5t.... 2.HealthExpenditures 2.1 2.5 3.3 3.2 45 9.1 10.0 13.9 14.4 20.1 23.6 As%ofGDP 1.4 1.3 1.4 1.0 0.9 1.2 1.1 1.6 1.2 1.3 1.4 As % of Total Spending 4.8 4.4 4 9 3.6 3.8 4.0 3 3 4.7 4.3 4.2 4.7 3. Social Secunty and 0.6 0.7 0.8 0.9 1.1 1.6 1.7 4.4 5.9 7.7 8.0'' Welfare As % of GDP 0.4 0.4 0.3 0.3 0.2 0.2 0.2 0.5 0.5 0.5 0.5 As % of Total Spending 1.4 1.3 1 2 1.1 1.0 0.7 0.6 1.5 1.8 1 6 1.6 4.HousingandUtilities 2 3 4 3 7 11 12 15 18 26 26 As%ofGDP 14 1.8 1.7 1.1 1.3 1.5 1.4 1.8 1.5 1.7 117 As%ofTotalSpending 4.6 6.2 6.1 3.8 5.8 4.9 4.0 5.1 5.3 5.4 5.1 .Cultural, Religious and 0.7 0 9 1.1 1 5 2.7 3.2 3.8 4.4 5.1 7.1 8.5 Entertainment Services As % of GDP 0.5 0.5 0.5 0.5 0.5 0.4 0.4 0.5 0.4 0.5 0.6 As%ofTotal Spending 1.7 1.6 17 1.8 23 1.4 1.2 1.5 15 1.5 1.7 *: Preliminary and **: Planned. Source- Ministry of Finance (October 2001). 125 Statistical Annex Table 54 - Percentage of Childrec aged 10-14, by sex and type off activlty Type of Activity Mate Female Fetal Workonly 7.17 11.63 9.29 Study only 79.71 43.98 62.68 Work and study 7.09 1.8 4.57 No activities 6.03 42.59 23.46 Total 100 100 1O10 Source: 1998 Household Budget Survey Table 55 - Total number of Children aged 10-14, by seX and type of activity Type of Activity Male Female Total Work only 92,942 137,437 230,379 Study only 1033824 519,778 1553602 Work and study 91,981 21,285 113,266 No activities 78,189 503,374 581,563 Total 1296936 1181874 2478810 Source: 1998 Household Budget Survey Table 56 - Percentage of Children aged 10-14, by sex, type off activity and area Type of Activity Area Male Female Total Work only Urban 1.7 0.9 1.3 Rural 8.8 14.9 11.7 Study only Urban 91.1 83.5 87.5 Rural 76.3 31.9 55.2 Work and study Urban 1.9 0.7 1.3 Rural 8.6 2.1 5.5 No activities Urban 5.3 14.8 9.9 Rural 6.2 51.1 27.5 Source: 1998 Household Budget Survey Table 57 - Percentage of Children aged 10-14, by sex and modality of employment Modality of employment Male Female Total Wage employ 13.8 5.94 10.17 Self employed 10.98 4.28 7.88 Unpaid Hh worker 74.54 88.85 81.15 Unpaid worker 0.68 0.94 0.8 Total 100 100 100 Source: 1998 Household Budget Survey 126 Statistical Annex Table 58 - Percentage of children aged 10-14, by poverty status and activity (Urban) activity I Non Poor Poor I Total ________-______+__________________----__+-_______ Work only | 0.84 2.16 1.32 Study only | 90.68 81.90 | 87.46 Work and study | 1.40 1.15| 1.31 No activities | 7.08 14.79 9.91 ________-______+_-________________-- - -+---------- Total I 100.00 100.00 I 100.00 Table 59 - Percentage of children aged 10-14, by poverty status and activity (rural) activity I Non Poor Poor I Total Work only | 11.05 12.34 11.68 Study only | 56.47 53.99 55.25 Work and study | 5.50 5.60 | 5.55 No activities 26.99 28.08 27.52 _______-_______+_-_-__________________+_________- Total I 100.00 100.00 I 100.00 Table 60 - Percentage of children aged 10-14, by poverty status and activity (national) activity I Non Poor Poor Total ----- ----+----------- - +------ Work only | 8.33 10.41 9.29 Study only | 65.70 59.17 62.68 Work and study | 4.38 4.79 i 4.57 No activities | 21.59 25.63 23.46 ________-______+__________________----__+-_______ Total I 100.00 100.00 I 100.00 127 StatisticalAnnex Table 61 - Percentage of children aged 10-14, by poverty status and modality of employment (urban) I povur mod_emp I Non Poor Poor Total ----------------+____-_-___-___-_--- - -+-___-_ Wage employ | 54.36 54.91 54.61 Self employed | 12.23 28.08 | 19.55 Unpaid Hh worker 31.25 17.01 24.67 Unpaid worker | 2.16 0.00 | 1.16 ____--__-------+___ ________-_-_--- - --+--------- Total I 100.00 100.00 I 100.00 Table 62 - Percentage of children aged 10-14, by poverty status and modality of employment (rural) I povur mod_amp I Non Poor Poor Total Wage employ I 7.23 9.00 8.13 Self employed | 8.25 6.49 | 7.35 Unpaid Hh worker | 84.04 83.44 | 83.73 Unpaid worker | 0.48 1.07 0.78 --------- -------- ---------+----------- - - ----------+ Total I 100.00 100.00 I 100.00 Table 63 - Percentage of children aged 10-14, by poverty status and modallty of employment (national) I povn mod_emp I Non Poor Poor | Total Wage employ I 9.45 10.86 I 10.17 Self employed I 8.60 7.19 7.88 Unpaid Hh worker 81.39 80.92 81.15 Unpaid worker 0.56 1.03 | 0.80 _______- --- ---_ + ---------------- +- _ ________ Total I 100.00 100.00 I 100.00 128 Statwsical Annex Table 64 - Percentage of Children aged 10-14 with health problems, by poverty status and activity (urban) I povur activity I Non Poor Poor Total ________-______+____-________________________ Work only | 10.7 9.9 10.2 Study only | 4.8 4.9 4.8 Work and study | 6.2 7.2 6.5 No activities 9.7 6.2 7.8 Total 5.2 5.2 5.2 Table 65 - Percentage of Children aged 10-14 with health problems, by poverty status and activity (rural) I povur activity I Non Poor Poor Total Work only | 12.2 6.0 9.0 Study only | 4.6 5.3 5.0 Work and study | 12.8 7.3 10.0 No activities 5.4 6.8 6.1 Total 6.1 5.9 6.0 129 Statistical Annex Table 66 - Percentage of Children aged 10-14 with health piroblems, by poverty status and modality of employment (urban) I povur mod_emp I Non Poor Poor Total Wage employ | 10.0 10.9 10.4 Self employed | 6.5 8.9 8.1 Unpaid Hh worker 5.3 3.0 4.6 Unpaid worker j 0.0 0.0 Total 7.9 9.0 8.4 Table 67 - Percentage of Children aged 10-14 with health problems, by poverty status and modality of employment (rural) I povur mod_emp I Non Poor Poor Total _______-_________+______-______________________ Wage employ | 10.1 13.4 12.0 Self employed | 4.0 3.2 3.6 Unpaid Hh worker 13.5 6.0 9.6 Unpaid worker 0.0 0.0 0.0 Total 12.4 6.4 9.3 130 Statistical Annex Table 68 - Total number of children in the sample, by sex and age Age Female Male Total 5 5385 5243 10628 6 6621 6805 13426 7 6637 6911 13548 8 6891 6967 13858 9 5753 5917 11670 10 7082 7648 14730 11 4398 5041 9439 12 6470 7288 13758 13 5497 6048 11545 14 5198 5466 10664 Total 59932 63334 123266 Source: Yemen 1999 National Poverty Survey. Table 69 - Total number of children in the population by sex and age Age Female Male Total 5 260,579 251,806 512,385 6 316,476 322,567 639,043 7 321,377 332,354 653,731 8 330,462 336,774 667,236 9 275,950 282,772 558,722 10 345,339 369,040 714,379 11 212,621 242,072 454,693 12 313,982 353,421 667,403 13 263,273 289,296 552,569 14 251,242 265,325 516,567 Total 2891301 3045427 5936728 Source: Yemen 1999 National Poverty Survey. Table 70 - Percentage of all working children aged 10-14, by sex and age Age Female Male Total 10 14.7 12.0 13.3 11 15.9 11.6 13.6 12 18.3 15.4 16.8 13 19.8 18.4 19.1 14 20.6 20.1 20.3 Total 17.7 15.4 16.5 Source: Yemen 1999 National Poverty Survey. Table 71 - Percentage of all working children aged 5-9, by sex and age Age Female Male Total 5 1.3 1.0 1.2 6 4.1 3.1 3.6 7 6.1 6.0 6.1 8 9.1 8.1 8.6 9 10.0 9.0 9.4 Total 6.2 5.6 5.9 Source: Yemen 1999 National Poverty Survey. 131 Statistical Annex Table 72 - Percentage of all worldng children, by sex and age Age Female Male Total 6 4.1 3.1 3.6 7 6.1 6.0 6.1 8 9.1 8.1 8.6 9 10.0 9.0 9.4 10 14.7 12.0 13.3 11 15.9 11.6 13.6 12 18.3 15.4 16.8 13 19.8 18.4 19.1 14 20.6 20.1 20.3 Total 12.8 11.3 12.0 Source: Yemen 1999 National Poverty Survey. Table 73 - Percentage of children enroUed, by sex and age Age Female Male Total 6 12.3 19.4 15.9 7 32 0 49.1 40.7 8 43.7 68.0 56.0 9 50.1 79.2 64.8 10 49.0 81.4 65.8 11 53.0 86.0 70.6 12 46.5 83.0 65.8 13 44.4 81.3 63.7 14 37.2 78.7 58.5 Total 40.4 68.9 55.1 Source: Yemen 1999 National Poverty Survey. Table 74 - Percentage of childreE working only, by sex and age Age Female Male Total 6 3.8 2.7 3.2 7 5.2 4.1 4.6 8 7.3 4.0 5.7 9 7.6 3.2 5.4 10 12.1 5.0 8.5 11 13.1 3.7 8.1 12 15.7 5.9 10.5 13 16.9 7.8 12.2 14 18.8 9.4 14.0 Total 10.8 5.0 7.8 Source: Yemen 1999 National Poverty Survey. 132 Statstical Annex Table 75 - Percentage of children studying only, by sex and age Age Female Male Total 6 12.1 19.0 15.6 7 31.0 47.2 39.2 8 41.9 63.9 53.0 9 47.7 73.4 60.7 10 46.4 74.5 60.9 11 50.2 78.1 65.1 12 44.0 73.5 59.6 13 41.5 70.7 56.8 14 35.4 68.0 52.2 Total 38.4 62.6 50.9 Source: Yemen 1999 National Poverty Survey. Table 76 - Percentage of Children Working and Studying, by sex and age Age Female Male Total 6 0.3 0.4 0.4 7 1.0 1.9 1.4 8 1.8 4.1 3.0 9 2.4 5.7 4.1 10 2.6 7.0 4.9 11 2.8 7.9 5.5 12 2.5 9.5 6.2 13 2.9 10.6 6.9 14 1.8 10.7 6.4 Total 2.0 6.3 4.2 Source Yemen 1999 National Poverty Survey. Table 77 - Percentage of Children Involved in no activities, by sex and age Age Female Male Total 6 83.8 77.9 80.8 7 62.8 46.9 54.7 8 49.0 28.0 38.4 9 42.3 17.6 29.8 10 38.9 13.5 25.8 11 33.9 10.3 21.3 12 37.7 11.1 23.6 13 38.7 10.8 24.1 14 44.0 11.9 27.5 Total 48.8 26.1 37.1 Source. Yemen 1999 National Poverty Survey. 133 Statistical Annex Table 78 - Percentage of Children aged 6-14, by sex and type of activity Type of Activity Female Male Total Work Only 10.8 5.0 7.9 Study Only 38.4 62.6 50.9 Work and Study 2.0 6.3 4.2 No activities 48.8 26.1 37.1 Total 100 100 100 Source: Yemen 1999 National Poverty Survey. Table 79 - Percentage of Children aged 6-14, by area and type of activity Type of Activity Rural Urban Total Work Only 9.9 1.3 7.9 Study Only 43.0 76.3 50.9 Work and Study 5.0 1.5 4.2 No activities 42.1 20.9 37.1 Total 100 100 100 Source: Yemen 1999 National Poverty Survey. Table 80 - Percentage of children aged 10-14, by sex and sector of activity Sector of Activity Female Male Total Government I 1 1 Private 98 98 98 Mixed 0 0 0 Cooperative 0 0 0 Others I I 1 Total 100 100 100 Source: Yemen 1999 National Poverty Survey. Table 81 - Percentage of Children aged 10-14, by area and sector of activity Sector of Activity Rural Urban Total Government 0.5 3.0 0.7 Pnvate 98.3 96.5 98.2 Mixed 0.1 0.1 0.1 Cooperative 0.2 0.2 0.2 Others 0.9 0.2 0.8 Total 100 100 100 Source: Yemen 1999 National Poverty Survey. 134 Statistical Annex Table 82 - All Working Children: daily hours worked, by sex and age Age Female Male Total 6 4.7 5.3 4.9 7 5.2 4.7 4.9 8 5.4 4.9 5.2 9 5.1 4.9 5 10 5.7 5.6 5.7 I 1 6 5.3 5.7 12 5.8 5.4 5.6 13 5.8 5.7 5.8 14 5.7 5.9 5.8 Total 5.6 5.4 5.5 Source: Yemen 1999 National Poverty Survey. Table 83 - Children Working Only: daily Hours Worked, by sex and Age Age Female Male Total 6 4.5 5.2 4.8 7 5.2 4.8 5 8 5.7 5.6 5.7 9 5.3 5.3 5.3 10 5.9 6 6 I 1 6.3 6.3 6.3 12 5.9 6.1 5.9 13 6 6.4 6.1 14 5.8 6.3 6 Total 5.8 5.9 5.8 Source: Yemen 1999 National Poverty Survey. Table 84 - Children and Studying: Daily Hours Worked, by sex and age Age Female Male Total 6 6.1 5.7 5.9 7 4.8 4.5 4.6 8 4.4 4.3 4.3 9 4.2 4.8 4.6 10 4.7 5.4 5.2 1 1 4.7 4.8 4.8 12 5.4 5 5.1 13 4.8 5.2 5.1 14 4.4 5.5 5.3 Total 4.7 5 5 Source: Yemen 1999 National Poverty Survey. 135 Statistical Annex Table 85 - Percentage of Children wit Health Problem Aged 6-14, by sex and type of activity Type of Activity Female Male Total Work only 13.1 11.3 12.5 Study only 11.4 11.1 11.2 Work and Study 14.6 13.4 13.7 No activities 11.9 13.1 12.3 Total 11.9 11.8 11.8 Source: Yemen 1999 National Poverty Survey. Table 86 - Percentage of Children with health problems aged 6-14, by area and type of activity Type of Activity Rural Urban Total Work only 12.6 11.5 12.5 Study only 11.4 10.9 11.2 Work and Study 13.9 11.8 13.7 No activities 12.5 11.2 12.3 Total 12.1 11.0 11.8 Source: Yemen 1999 National Poverty Survey. Table 87 - Percentage of Children aged 10-14, by sex and modality of employment Modality of Employment Female Male Total Employee 2.5 8.6 5.4 Self-employed 3.1 5.2 4.1 Employer 2.9 3.4 3.1 Works within Family 91.0 82.4 86.8 Unpaid work (apprentice) 0.6 0.5 0.5 Total 100 100 100 Source: Yemen 1999 National Poverty Survey. Table 88 - Percentage of Children aged 10-14, by area and modality of employment Modality of Employment Rural Urban Total Employee 4.6 18.9 5.4 Self-employed 3.5 13.4 4.1 Employer 3.2 1.8 3.1 Works within Family 88.2 65.1 86.8 Unpaid work (apprentice) 0.5 0.9 0.5 Total 100 100 100 Source: Yemen 1999 National Poverty Survey. 136 StatisticalAnnex Table 89 - Percentage of Children aged 10-14 with health problems, by sex and modality of employment Modality of Employment Female Male Total Employee 10.2 10.7 10.6 Self-employed 14.7 11.5 12.7 Employer 9.4 10.2 9.9 Works within Family 13.5 12.5 13.0 Unpaid work (apprentice) 8.5 12.2 10.2 Total 13.3 12.2 12.8 Source: Yemen 1999 National Poverty Survey. Table 90 - Percentage of Children aged 10-14 with health problems, by area and modality of employment Modality of Employment Rural Urban Total Employee 10.8 9.8 10.6 Self-employed 11.2 18.7 12.7 Employer 10.1 2.9 9.9 Works within Family 13.1 10.2 13.0 Unpaid work (apprentice) 11.4 0.0 10.2 Total 12.9 11.0 12.8 Source: Yemen 1999 National Poverty Survey. Table 91 - Percentage of Children working only, by household head's education and children's age Age Illiterate Read and Basic Secondary Above- Total Write secondary 6 4.3 2.0 1.0 1.7 1.0 3.3 7 5.7 3.6 2.9 1.5 1.2 4.6 8 7.2 3.9 3.2 1.4 0.4 5.7 9 6.5 4.0 3.7 2.7 0.6 5.3 10 10.4 6.4 4.4 2.5 1.7 8.5 1 1 10.3 6.5 3.5 1.4 1.8 8.1 12 12.6 8.9 6.0 2.6 1.5 10.5 13 14.7 10.4 5.4 2.7 2.5 12.2 14 16.9 10.9 9.6 3.6 2.8 14.0 Total 9.7 6.2 4.2 2.2 1.4 7.9 Source: Yemen 1999 National Poverty Survey. 137 Statistical Annex Table 92 - Percentage of children studying only, by household head's education and children's age Age Illlterate Read and Basic Secondary Above- Totan write secondary 6 12.7 17.4 20.0 24.9 27.0 15.5 7 33.1 42.2 50.0 56.9 69.5 39.3 8 44.6 60.6 70.5 74.4 86.9 53.0 9 51.9 69.4 77.5 80.1 92.5 60.8 10 53.3 68.5 77.7 81.3 90.0 60.9 11 57.0 71.7 79.9 86.0 90.6 65.0 12 52.7 64.5 77.8 82.1 87.3 59.6 13 49.2 63.1 74.9 82.0 88.3 56.8 14 44.9 57.4 68.4 76.8 85.3 52.2 Total 44.0 56.8 65.1 69.3 78.8 50.9 Source: Yemen 1999 National Poverty Survey. Table 93 - Percentage of Children worldng al d studying, by household head's education and children's age Age Illiterate Read and Basic Secondary Above- Total write secondary 6 0.3 0.5 0.1 0.6 0.1 0.4 7 1.5 1.6 1.4 1.0 0.4 2.4 8 3.2 3.5 1.4 2.2 0.4 3.0 9 4.3 4.3 2.3 5.8 1.3 4.1 10 4.7 6.3 4.5 3.4 2.9 4.9 11 5.7 6.0 4.1 5.0 2.2 5.5 12 5.9 8.4 4.0 5.1 5.1 6.2 13 6.9 7.9 6.8 5.7 3.5 6.9 14 6.1 7.9 5.8 4.5 3.5 6.3 Total 4.2 5.1 3.2 3.4 2.0 4.2 Source: Yemen 1999 National Poverty Survey. Table 94 - Percentage of children involved in no activities, by household head's education and children's age Age Illiterate Read and Basic Secondary Above- Total Write secondary 6 82.7 80.0 78.9 72.8 71.9 80.8 7 59.6 52.6 45.6 40.7 29.0 54.7 8 45.0 32.0 24.9 22.0 12.4 38.4 9 37.3 22.2 16.5 11.4 5.7 29.8 10 31.6 18.8 13.5 12.7 5.3 25.8 11 26.9 15.8 12.5 7.6 5.4 21.4 12 28.7 18.2 12.2 10.2 6.1 23.7 13 29.2 18.6 12.9 9.6 5.7 24.1 14 32.0 23.8 16.3 15.2 8.4 27.5 Total 42.1 32.0 27.5 25.1 17.8 37.1 Source: Yemen 1999 National Poverty Survey. 138 Bibliographical References Government publication Central Statistical Organization (CSO) [Yemen] and Macro International Inc (MI). (1998) Final Report: Yemen Demographic and Health and Maternal and Child Health Survey 1997. (DHS-1997) Calverton, Maryland: CSO and M. Central Statistical Organization (CSO) [Yemen] and Macro International Inc (Ml). (1994) Final Report: Yemen Demographic and Health and Maternal and Child Health Survey 1991/1992. (DHS-1991/1992) Calverton, Maryland: CSO and MI. Ministry of Planning and Development and Central Statistical Organization. (1994) Final Results of the Population, Housing and Establishments Census General Report. December 1994. Republic of Yemen. Central Statistical Organization. (2000) Statistical Year Book 1999. Republic of Yemen. Central Statistical Organization. (1999) Statistical Year Book 1998. Republic of Yemen. Ministry of Planning and Development. (2001?) "The Second Five Years Plan For Economic and Social Development in The Republic of Yemen 2001 - 2005." (Original in Arabic, translated to English) Republic of Yemen. Ministry of Education. (1999) Report on Education for All Evaluation for the Year 2000. Ministry of Education, Republic of Yemen October 2000. (www2.unesco.orgAvef/countryreportslyemen/contents.html) Republic of Yemen, "Teacher and the Education Professions." Law No. (37) of 1998. (Original in Arabic, translated to English) World Food Program. (2002) "County Brief-Yemen: Projected Needs for WFP Projects and Operations" (http://www.wfp.org/country brief]) UNDP, 2001, "Final mission report: Poverty Alleviation & Employment Generation Programme, mid-term evaluation, Sana'a, Republic of Yemen. UNDP, 1998, "An Operational Strategy for Community-Based Regional Development in the Republic of Yemen: Poverty Alleviation & Employment Generation Programme," Sana'a, Republic of Yemen, March. Republic of Yemen, 2001, "Public Works Project: June 1996 to September 2001,", Sana'a, Republic of Yemen. 139 Republic of Yemen, 2000, "Social Fund for Development: 2000 Annual Report," Sana'a, Republic of Yemen. Others Canagarajah, Sudharshan and Xiao Ye. (2001) "Public Health and Education Spending in Ghana in 1992-98: Issues of Equity and Efficiency." Policy Research Working Paper 2579. World Bank Country Director Groups Ghana Country Department and Africa Technical Families Macroeconomics 1. Castro-Leal, Florencia (1996b) "Poverty and Inequality in the Distribution of Public Education Spending in South Africa." PSP Discussion Paper Series No. 102, Poverty and Social Policy Department, The World Bank. Castro-Leal, Florencia, Julia Dayton, Lionel Demery, and Kalpana Mehra. (1999). "Public Social Spending in Africa: Do the Poor Benefit?" The World Bank Research Observer 14 (1): 49-72. Castro-Leal, Florencia. (1996a). "Who Benefits from Public Education Spending in Malawi? Results from the Recent Education Reform." World Bank Discussion Paper No. 350. Demery, Lionel (2000) "Benefit Incidence: A Practitioner's Guide," Poverty and Social Development Group Africa Region, World Bank. (http://wwwl.worldbank.org/publicsector/pe/benefitabstract.htm) Demery, Lionel, Shiyan Chao, Rene Bernier and Kalpana Mehra (1995) "The Incidence of Social Spending in Ghana." PSP Discussion Paper Series No. 82, Poverty and Social Policy Department, The World Bank (November). Lanjouw, Peter, and Martin Ravallion (1999) "Benefit Incidence, Public Spending Reforms, and the Tuning of Program Capture" World Bank Economic Review Vol. 13. 257-73. Lanjow, Peter, Menno Pradhan, Faida Saadah, Haneen Sayed, Robert Sparrow. (2001) "Poverty, Education and Health in Indonesia: Who Benefits from Public Spending?" Policy Research Working Paper 2739. December 2001. El-Ganmual, Yasser, S.R Katariya and Piet Goovaerts, 1999, Report on Review of the Agriculture and Fisheries Production Promotion Fund, Republic of Yemen, World Bank Jalan, Jyotsna and Martin Ravallion, 2002, "Estimating the Benefit Incidence of an Antipoverty Program by Propensity Score Matching," Journal of Business & Economic Statistics, forthcoming. 140 Tesliuc, Comelia, and Lucian Pop. (2000) "Poverty, Inequality and Social Protection." In Economic Transition in Romaina: Past, Present and Future, edited by Christof Ruhl and Daniel Daianu. van de Walle, Dominique, 2002, "The Static and Dynamic Incidence of Viet Nam's Public Safety Net," Policy Research Working Paper No. 2791 Development Research Group, World Bank, Washington, D.C., February 2002. van de Walle, Dominique, Martin Ravallion and Madhur Gautam, 1994, "How Well Does the Social Safety Net Work? The Incidence of Cash Benefits in Hungary 1987-89," Living Standards Measurement Study Working Papers, 102, 1994. van de Walle, Dominique (1992) "The Distribution of the Benefits from Social Services in Indonesia, 1978- 87." Policy Research Working Paper. No. 871, Country Economics Department, the World Bank. World Bank (2001b) Peruvian Education at a Crossroads Challenges and Opportunities for the 21' Century A World Bank Country Study 22357 May 2001 World Bank. (1 995a) Kenya Poverty Assessment. Population and Human Division, Eastern African Department, Africa Region, Report No. 13152-KE, Washington, D.C. World Bank. (1995b) Lao PDR Social Development Assessment and Strategy. Human Resource Operation Division, Country Department I, East Asia and Pacific Region, Report No. 13992-LA. Washington, D.C. World Bank. (1996c) Madagascar Poverty Assessment. Population and Human Division, South Central and Indian Ocean Department, Africa Region, Report No. 14044, Washington, D.C. World Bank. (1999b) Former Yugoslav Republic of Macedonia Focusing on the Poor. . Human Development Sector Unit, Country Department IV, Europe and Central Asia Region, Report No. 19411, Washington, D.C. World Bank. (2000a) World Development Report 2000/2001. World Bank, Washington, D.C. 141 World Bank. (2000b) Making Transitional Workfor Everyone. Poverty and Inequality in Europe and Central Asia. World Bank, Washington, D.C. World Bank. (2000c) Viet Nam Managing Public Resources Better: Public Expenditure Review 2000. Poverty Reduction and Economic Management Unit, East Asia and Pacific Region, Report No. 21021-VN, Washington, D.C. World Bank. (2000d). Ecuador: Crisis, Poverty and Social Services. Human Development Department, Latin America and Caribbean Region, Report No.19920-EC, Washington, D.C. World Bank. (2001 a) Kingdom of Morocco Poverty Update Morocco. Human Development Sector, Middle East and North Africa Region, Report No. 21506-MOR, Washington, D.C. Lockheed, Marlaine, Regina Bendokat and Takako Yuki. (1999) Yemen Education Sector Assistance Strategy. Human Development Group, Middle East and North Africa Region, World Bank: Washington, D.C. January 1999.* World Bank. (1998) "Yemen Public Education Expenditure Review FY1999" Paper submitted from the Human Development Group, Middle East and North Africa Region, World Bank to the Government of Yemen.* World Bank. (1999) "Yemen Public Expenditure Review: Education Sector Aide-Memoir." Document submitted from the Human Development Group, Middle East and North Africa Region, World Bank to the Government of Yemen. August 1999.* World Bank. (2000) "Draft Yemen: Budget and Institutional Reform in Support of the Five-Year Plan Education Sector Chapter." Submitted to Economic and Social Development Group, Middle East and North Africa Region, World Bank. December 2000.* World Bank. (2001) "Higher Education Rationalization Study." Paper submitted from the Human Development Group, Middle East and North Africa Region, World Bank to the Government of Yemen in March 2001.* World Bank. (2000) Project Appraisal Document of Republic Yemen Child Development Project. Human Development Group, Middle East and North Africa Region. Washington, D.C.: World Bank. World Bank. (2000) Project Appraisal Document: Yemen - Basic Education Expansion Project. Human Development Group, Middle East and North Africa Region. Washington, D.C.: World Bank. 142 World Bank, 2000a, Implementation Completion Report: Public Works project, December, report no. 21074, World Bank, Washington, D.C. World Bank, 2000b, "Project Appraisal Document on a proposed credit in an amount of SDR 56 million to the Republic of Yemen for a Second Social Fund for Development Project," Report No. 20291 YEM, World Bank, Washington, D.C. World Bank, 2000c, Public Works Project 1 & 2: Annual Review, May, World Bank, Washington, D.C. World Bank, 2000d, "Republic of Yemen, Comprehensive Development Review, Phase I: Poverty and Social Safety Nets Building Block," January, World Bank, Washington, D.C. World Bank, 1997a, "Project Appraisal Document on a proposed IDA credit in an amnount of SDR 21.7 million to the Republic of Yemen for a Social Fund for Development Project," Report No., 16301 YEM, World Bank, 1997b, "Yemen: Social Welfare Fund: Assessment and Recommendations," World Bank, Washington, D.C. i ~~~~~~~~~~~~~~~~IBRDI 303,17R1 42' 44' 4- 48- 50' 52- 5140 SAUDI ARABIA - REPUBLIC OF YEMEN S 0 TOWNS AND VILLAGES PRIMAARY ROADS ' '5dd =sh /qq \EM | (3 GOVERNORATE CAPITALS SECONDARY ROADS ,, , ' YE1EN ' NATIONAL CAPITAL WADIS ( AIRPORTS GOVERNORATE BOUNDARIES I PORTS - INTERNATIONALBOUNDARIES / - '- = / A 7 - SA< X = 5DAH 'AA .) *D,odnh ~~~~~~~AL-JOWF SoWd6I.S A Mr. /' DIOTI.".--' 5 3 5 2 _ U22 ~AMA i'. 18'~~~~~~~~~~~. '5' RI~~~~~ A 'A Jlkono H.,HhAB Hops IS,. ~ ~~~A-AL- D B . P AP > Gpuor AT Socoln Al\ .'- Ao S ' pO0S7L4PO9 0oT~W,dBo 25 50 75 IT0 KILOMETERS \ Io, Th,b0h / 11~~~~~~~~~~~~~~~T.5 b-dL......JI. ADEW 75 1 00 MIILESd ..o y,oAbd oPTE Wo,SM,E,ocp oIn - ETHIOPIA \- *' ' SoSoSn,0 2 0 7 T IE Ad D42OUT 40'SO' 52' 54 - Gep o, i ,N o.: 24422 YEM Type: SR