Report No.: 69116-BF Burkina Faso Non-Monetary Poverty and Gender Inequalities 1993-2010 Trends A Policy Note 5 June 12, 2013 Poverty Reduction and Economic Management 4 Country Department AFCF2 Africa Region Document of the World Bank CURRENCY EQUIVALENTS (Exchange Rate Effective May 8, 2013) Currency Unit = CFA franc (CFAF) 1 US$ = CFAF 500 FISCAL YEAR January 1 – December 31 ABBREVIATIONS AND ACRONYMS DHS Demographic and Health Surveys EA Enumeration Areas EP Enquête Prioritaire FCFA Franc CFA FMG Female Genital Mutilation INSD Institut National de la Statistique et de la Demographie MCA Multiple Correspondence Analysis MDG Millennium Development Goals SCADD Stratégie pour une Croissance Accélérée et une Développement Durable WHO World Health Organization Vice President: Makhtar Diop Country Director: Madani M. Tall Country Manager: Mercy M. Tembon Sector Director /Sector Manager: Marcelo Giugale Task Team Leader: Andrew Dabalen ii Table of Contents 1. NON-MONETARY POVERTY AND GENDER INEQUALITIES IN BURKINA FASO: 1993-2010 TRENDS .................................................................................................................................... 1 A. INTRODUCTION.......................................................................................................................................... 1 B. QUALITY OF HOUSING .............................................................................................................................. 2 C. ACCESS TO BASIC SERVICES ..................................................................................................................... 4 D. POSSESSION OF ASSETS ............................................................................................................................. 7 E. HEALTH AND NUTRITION .......................................................................................................................... 9 F. EDUCATION ............................................................................................................................................. 13 G. SOCIAL PROTECTION ............................................................................................................................... 15 List of Figures: Figure 1.1: Housing deprivation rates between 1993 and 2010 .................................................................. 3 Figure 1.2: Gender differences in housing deprivation, 1993-2010 ........................................................... 3 Figure 1.3: Regional housing deprivations in 2010 and differences compared to 2003 ............................. 4 Figure 1.4: Deprivations rates in utilities, 1993-2010................................................................................. 5 Figure 1.5: Gender differences in utilities, 1993-2010 ............................................................................... 6 Figure 1.6: Regional utilities deprivations in 2010 and differences compared to 2003.............................. 6 Figure 1.7: Deprivation rates in assets ownership, 1993-2010 ................................................................... 7 Figure 1.8: Cumulative density curves of asset index, 1993-2010 ............................................................. 7 Figure 1.9: Proportions of individuals deprived in land between 1998 and 2010 ...................................... 8 Figure 1.10: Gender differences in land and house deprivations by region, 2010........................................ 9 Figure 1.11: Infant and child mortality rates, 1993-2010 ............................................................................. 9 Figure 1.12: Infant and child regional mortality rates, 2010....................................................................... 10 Figure 1.13: Gender differences in infant and child mortality rates, 1993-2010 ........................................ 10 Figure 1.14: Under-five nutritional status, 1993-2010................................................................................ 11 Figure 1.15: Under-five nutritional status by regions, 2010 ....................................................................... 12 Figure 1.16: Gender differences in under-five malnutrition, 1993-2010 .................................................... 12 Figure 1.17: Classification of women nutritional status, 1993-2010 .......................................................... 13 Figure 1.18: Deprivation in education and enrollment rates, 1993-2010.................................................... 14 Figure 1.19: Proportion of uneducated individuals aged 18 to 64 years, 1993-2010 ................................. 15 Figure 1.20: Gender differences in enrollment and education deprivation, 1993-2010 ............................. 15 Figure 1.21: Teenagers and social protection (cohort of women aged 15 to 24 years - 1993-2010 ........... 16 Figure 1.22: Social protection and teenagers deprivation by regions, 2010 ............................................... 17 Figure 1.23: Proportion of under-five children without birth certificate, 2003-2010 ................................. 17 Figure 1.24: Proportion of under-five children unregistered or without birth certificate by region, 2010 . 18 iii ACKNOWLEDGMENTS This policy note was prepared by Andrew Dabalen and Yele Batana (World Bank). Judite Fernandes provided excellent assistance with document preparation. The report was prepared under the guidance of Marcelo Giugale (Sector Director) and Miria Pigato (Sector Manager, AFTP4). We are grateful to the following for very helpful suggestions: Mercy M. Tembon and Azedine Ouerghi. The Team would also like to thank Macro International Inc. for sharing Demographic and Health Surveys. iv 1. Non-Monetary Poverty and Gender Inequalities in Burkina Faso: 1993-2010 Trends A. INTRODUCTION 1.1 The monetary measure remains the tool that is mostly used for analyzing poverty profile, yet it has some weaknesses. In general, the monetary poverty profile is established using the household per capita expenditure. Thus, as Lanjouw (2012) stresses, this approach does not permit to adequately measure individual welfare within the household. First, it does not take into account the existence of economies of scale within the household, which generally results in over-estimating poverty mainly in large-sized households. Actually, the existence of public goods within the household (such as television set, radio, stove, refrigerator, etc) and the possibility for large-sized households to do wholesale purchases, which is likely to reduce the unit costs of products, reduce the household per capita cost of expenditure. Next, the per capita approach is founded on the assumption of fair expenditure allocation within the household. This is not generally the case since the intra-household allocations strongly depend on each household member’s negotiation capacity. Thus, instead of being a homogeneous entity, households generally remain characterized by expenditure inequalities across the members. 1.2 Multidimensional poverty is a good complementary approach to analyze the household poverty profile. Since Sen’s studies (1976, 1985), welfare is increasingly characterized as being a multidimensional phenomenon. While monetary measure is a good proxy, it does not enable to grasp all the aspects of welfare. For example, a household may have as enough resource as to be regarded as non-poor, but not be in position to provide health care to one of its sick member just because it does not have access to a health center. It is therefore obvious that poverty issues are sometimes a matter of access to basic services particularly in developing countries. Formally the World Bank and the United Nations define poverty as a phenomenon that goes beyond a simple shortage of income as it also means an absence of some basic services which allow it for people to properly carry out their lives in the society. Another advantage of using the multidimensional analysis is that price changes are not taken into account under the monetary approach when poverty and welfare comparisons are conducted over the time. 1.3 The theoretical framework of the multidimensional approach needs to be strengthened. Measuring multidimensional poverty is generally multi-challenging. First is the selection of the dimensions to be taken into account in studying welfare. In practice, several studies draw upon the Millenium Development Goals (MDG) to select welfare indicators. But obviously there is no universal procedure for selecting the dimensions and this exercise often rests with each researcher. Another challenge is about the choice of a method for aggregating the various dimensions in order to obtain a composite measure of poverty. Several methods are suggested in the literature but, as Ravallion (2012) stresses it, they generally display a weak conceptual framework and do not enable to consider possible correlations between the various dimensions. Other difficulties are linked to the weights to be allotted to each dimension. There is actually no consensual method for assigning weights. The “dashboard� approach proposed by Ravallion (2011) consists in analyzing poverty separately in each dimension rather than seeking -1- to aggregate them. This means that several indices should be considered instead of only one composite index. This approach is however rather restrictive insofar as a composite index which takes into account the joint distribution of all the dimensions contains more information than several indices (Ferreira, 2011). 1.4 The methodology adopted in this report combines the “dashboard� approach and the multidimensional approach. The analysis of non-monetary poverty is carried out using the deprivation indices in each dimension. The dimensions considered include mainly the quality of housing, access to basic public services, possession of assets, health and nutrition, education, quality of employment and social protection. However, the possession of assets is derived using the multiple correspondence analysis (MCA), one of the methods often used in multidimensional poverty analyses. The study draws on four Demographic and Health Surveys (DHS) which were conducted respectively in 1993, 1998-99, 2003 and 2010. Three questionnaires were used: the household questionnaire which collects information on the main socio-economic and demographic characteristics of the households; the individual women focused questionnaire for women aged 15-49; and the individual men focused questionnaire for men aged 15-59. The DHS are generally based on a two-stage stratified sampling. While the enumeration areas (EA) defined as part of the 1985 census were used as frame for the 1993 survey, those of the 1996 census served for the 1998/99 and 2003 surveys, whereas the 2006 census EA are used for the 2010 survey. However, all the surveys are representative at the national level. B. QUALITY OF HOUSING 1.5 The right to a decent housing is stipulated in article 25 of the Universal Declaration of Human Rights. Housing remains an important component of welfare and the right to a decent housing is recognized by several Organizations and countries including Burkina Faso. Burkina Faso subscribed to the African Charter on the Rights and Welfare of the Child which clearly states the right to housing. In line with inter alia the MDGs, the 1996 Istanbul Declaration, and the Poverty Reduction Strategy Paper (PRSP), the housing and urban development policy has set the goal to create conditions for improving the populations’ living environment through leading the cities to contribute to the fight against poverty (Ministry for Housing and Urban Planning, 2008). Three specific objectives are defined: (i) change the cities into growth and development poles; (ii) ensure access to decent housing for the most deprived people; (iii) fight against urban poverty. The history of the housing department has been linked to decentralization since 1995, and underwent some renewal in 2006 with the creation of a Ministry for Housing and Urban Planning, as well as of the Housing Bank. Beyond being just a component of welfare, living in a precarious house increases the risks for people to contract diseases (WHO, 2006). Based on the relevance of this reality, the World Bank has allocated more than 16 billion dollars, to 90 countries and throughout 278 projects, to support actions aiming to improve housing conditions over the three decades before 2005. -2- Figure 1.1: Housing deprivation rates between 1993 and 2010 80 Percentage of population 60 40 20 0 Overcrowding Poor floor Poor housing 1993 1998 2003 2010 1.6 Deprivation of decent housing seems to have dropped in Burkina Faso between 1993 and 2010. Figure 1.1 shows a reduction in the proportions of individuals living in insecure housings. Housing deprivation is defined following precariousness of the floor and/or overcrowding in the household. A household is said to be overcrowded when a room bears more than five members. The decline noted results only from a decline in floor precariousness. From nearly 71% in 1993, the floor deprivation incidence increased to more than 73% in 1998, before gradually decreasing to stabilize at around 57% in 2010. On the other hand, the overcrowding rate increased during the same period, going up from 6% in 1993 to nearly 8% in 2010. Figure 1.2: Gender differences in housing deprivation, 1993-2010 4 Gender differences in percentage points 3 2 1 0 Overcrowding Poor floor Poor housing 1993 1998 2003 2010 1.7 The gap between women’s deprivation and men’s deprivation of housing narrowed between 1993 and 2010. While gender specific overcrowding difference remains marginal, there exists on the other hand a relatively important difference with regard to deprivation in floor (see -3- Figure 1.2). The difference between housing deprivation rate of women and the one of men, which ranged at more than 3 percentage points in 1993 and 1998, came down to about 1 point in the years 2003 and 2010. 1.8 Housing deprivation displays wide regional disparities. Thus, while the housing deprivation rate stands below 20% in the Centre region, it exceeds 90% in the Sahel region in 2010 (see part (1) of Figure 1.3). The Boucle du Mouhoun and the Sud-Ouest regions have also high rates which go beyond 80%. The Centre-Est region also displays a relatively low rate, with a level below 30%. Part (2) of Figure 1.3 shows differences in deprivation rate between 2010 and 2003 in each region. The rates in some regions (Boucle du Mouhoun, Centre and Centre-Sud) increased over the period by less than 10 percentage points. Inversely, in the other regions deprivation of housing dropped, especially in the Centre-Nord where the reduction ranges between -20 and -30 percentage points. Figure 1.3: Regional housing deprivations in 2010 and differences compared to 2003 (1) Deprivation rates in housing by regions in 2010 (2) Differences within regions between 2003 and 2010 Sahel Sahel Nord Nord Centre-Nord Centre-Nord Boucle du Mo Boucle du Mo Plateau Cent Plateau Cent Centre Centre Est Est Centre-Ouest Centre-Ouest Centre-Est Centre-Sud Centre-Est Centre-Sud Hauts-Bassin Hauts-Bassin Percentage of population (%) Percentage points (%) (90,100] (0,10] Sud-Ouest (80,90] Sud-Ouest (-10,0] Cascades (60,80] Cascades (-20,-10] (40,60] [-30,-20] (20,40] [0,20] C. ACCESS TO BASIC SERVICES 1.9 Access to basic services is an important development objective. Access to basic services such as electricity, drinking water and sanitation is not just a human right issue, but also a public good matter which can generate positive externalities (Brown, 2009; Hailu and Tsukada, 2009). The importance of this issue is entrenched in the Millennium Development Goals and one of its targets, specifically the target of the 7th objective, is to reduce by half the proportion of individuals who lack access to drinking water and basic sanitation services. Some of the SCADD (Strategy for Accelerated Growth and Sustainable Development) objectives also comply with the MDGs, particularly those set in its 2011-2015 action plan. One of SCADD’s main targets is to open access for isolated rural areas through extending the electrical power grid with the view to achieve a 60% electrification rate at the national level in 2015. Other reforms involve the intensification of reforms defined in the National Programme for Water Supply and Sanitation -4- (PN-AEPA), which mainly focus on building sanitation and water supply infrastructures and on fostering demand for improved sanitation in rural areas (SCADD, 2011). Figure 1.4: Deprivations rates in utilities, 1993-2010 100 Percentage of population 80 60 40 20 0 No access to No electricity No toilet Poor utilities drinking water 1993 1998 2003 2010 1.10 While a large proportion of the Burkinabe population has access to drinking water, relatively few people have access to electricity and sanitation. Figure 1.4 shows that some improvement was recorded in the field of access to basic services between 1993 and 2010. The drinking water deprivation rate declined from 22% to 17.5%, which is still inadequate when considering the progression as compared to the target objective of reducing by half stated in the 7th MDG. In fact, the reduction occurred primarily between 1993 and 1998 since some slight increase in the deprivation rate was noted during the 2000s. The improvement rate for water supply was higher in urban areas than in rural areas. There was some improvement in the access to electricity during the 2000s. The deprivation rate for electricity thus stagnated to around 94% in the 90s, before declining in 2003 and 2010 when it stood at 88%. This improvement is also rather noted in the urban areas. Rural areas remain very disadvantaged with a proportion of the population living without electricity reaching 99% there. Rural areas are also disadvantaged in the field of access to sanitation with relatively high rates, despite some improvement over the period. From nearly 72% in 1993, the deprivation rate for sanitation increased first to 77% in 1998 before gradually declining to finally range at 64% in 2010. Deprivation of access to basic services is defined by the fact of being deprived of at least two of the three services. An analysis of the resulting deprivation rate also indicates some improvement during the period. 1.11 Women appear to be slightly disadvantaged in the access to basic services as compared to men. Concerning access to drinking water, the difference between women’s deprivation rate and men’s deprivation rate outlines a gap of nearly 2 percentage points in 1993 (see Figure 1.5). But this percentage became insignificant over the whole period. The difference in the field of access to electricity seems rather constant during all the period with a rough value of 2 percentage points. The gaps are relatively more stressed for access to sanitation. The difference exceeded 4 percentage points during the 90s whereas it dropped to 3 or less during the 2000s. -5- Figure 1.5: Gender differences in utilities, 1993-2010 5 Gender differences in percentage points 4 3 2 1 0 No access to No electricity No toilet Poor utilities drinking water 1993 1998 2003 2010 Figure 1.6: Regional utilities deprivations in 2010 and differences compared to 2003 (1) Deprivation rates in utilities by regions in 2010 (2) Differences within regions between 2010 and 2003 Sahel Sahel Nord Nord Centre-Nord Centre-Nord Boucle du Mo Boucle du Mo Plateau Cent Plateau Cent Centre Centre Est Est Centre-Ouest Centre-Ouest Centre-Est Centre-Sud Centre-Est Centre-Sud Hauts-Bassin Hauts-Bassin Percentage of population (%) Percentage points (%) Sud-Ouest (80,100] Sud-Ouest (0,10] Cascades (60,80] Cascades (-10,0] (40,60] (-20,-10] (20,40] [-30,-20] [0,20] 1.12 Significant regional disparities are noted in both the level and the trends of access to basic services. In light of Figure 1.6, the Centre region seems to be the least disadvantaged with a deprivation rate under 20%, followed by the Hauts-Bassins region with a rate close to 40%. The rates posted by the other regions are higher to 60% and five of them (Centre-Est, Centre- Sud, Est, Sahel and Sud-Ouest) display deprivation rates that exceed 80%. Between 2003 and 2010, all the regions experienced an improvement except from the Sahel and the Boucle du Mouhoun regions where the deprivation rate increased. The Center Plateau region appears to have recorded the most significant improvement, with a decrease in rate of more than 20 percentage points, followed by the North region where the rate fell by at least 10 percentage points. A relatively weak decline was observed in the other regions regarding deprivation of basic services. -6- D. POSSESSION OF ASSETS 1.13 The deprivation of assets constantly declined between 1993 and 2010 in Burkina Faso. The possession of assets per household member is a good welfare indicator. Thus, not only, the possession of assets reduces the probability to fall into monetary poverty but also this indicator is less volatile than income and could therefore be a good proxy for permanent income (McKay, 2009). The wealth index is estimated drawing on information about the possession of durable goods (radio, television set, car, bicycle, motorbike, refrigerator), and using the multiple correspondence analysis method with a global sample including the data obtained from the four surveys. The deprivation rate is then defined by the proportion of individuals whose wealth index is below a relative poverty line equivalent to the average wealth index in the total sample. Figure 1.7 shows that deprivation constantly decreased between 1993 and 2010, from about 45% to slightly above 20%. The trend is the same for both men and women even if the rate for women is slightly higher to that recorded for men. This downward trend is also observed in rural areas while the situation is different in urban areas. Indeed, if the rate declines in urban areas between 1993 and 2003, from 15% to 8%, it increases during the 2000s as it ranges at around 12% in 2010. The robustness analysis illustrated in Figure 1.8 shows that the ranking between surveys does not necessarily remain the same when a different poverty line is applied. Figure 1.7: Deprivation rates in assets ownership, 1993-2010 50 40 Percentage of population 30 20 10 0 Female Male National Rural Urban 1993 1998 2003 2010 Figure 1.8: Cumulative density curves of asset index, 1993-2010 1 .8 Cumulative density curve .6 .4 1993 1998 2003 2010 .2 0 0 .2 .4 .6 .8 1 Asset index -7- 1.14 The proportion of individuals owning land decreased both within men and women between 1998 and 2010. As Figure 1.9 highlights it, while the proportion of individuals deprived in land slightly increased during the period, a different trend is observed according to the residence area. In urban areas, the proportion of women who do not possess land decreased, from 95% to 88%, and the proportion of men also dropped from 93% to 76%. This is an evidence of the growing interest of townspeople for agricultural land. Conversely, the proportions rather increased in rural areas, rising from 56% to 61% among women, and from 23% to 33% among men. Figure 1.9: Proportions of individuals deprived in land between 1998 and 2010 (1) Women (2) Men 100 100 Proportion of women Proportion of men 80 80 60 60 40 40 20 20 0 0 Urban Rural National Urban Rural National Gender differences in percentage points (3) Women minus men 30 20 1998 2003 2010 10 0 -10 Urban Rural National 1.15 Gender inequality regarding land access in rural areas is still significant in the late 2000s. With the view of improving inter alia women access to land, a rural land tenure policy was adopted in October 2007 by the Government. But this action is still far from ensuring gender equality with regard to land possession. Figure 2.9 enlightens that though a slight decline is recorded in gender inequality between 1998 and 2010, especially in rural areas, the difference between women’s deprivation rate and men’s deprivation rate remains rather high. Actually, in 2010 this gap is 28 percentage points in rural areas and around 23 percentage points at the national level. The regional differences observed in gender inequality with regard to land access are also very important. The Sahel region thus appears to have the largest gaps concerning land access inequality with a difference exceeding 60 percentage points (see Figure 1.10). This region is followed by the Centre-Est and the Nord regions where the differences stand above 40%. On the other hand, the Centre, the Plateau Central and the Est regions seem to be the least affected with differences below 10 percentage points. Concerning housing possession, significant inequalities are also recorded between men and women. The Sahel region once again holds the highest record regarding gender inequalities and shows a difference of more than 60 percentage points. Then follows the Centre-Est and the Centre-Ouest regions where the difference exceeds 50 percentage points. The Plateau Central is the region which experiences the least marked inequality with a value standing at less than 10 percentage points. -8- Figure 1.10: Gender differences in land and house deprivations by region, 2010 (1) Ownership of land (2) Ownership of house Sahel Sahel Nord Nord Centre-Nord Centre-Nord Boucle du Mo Boucle du Mo Plateau Cent Plateau Cent Centre Centre Est Est Centre-Ouest Centre-Ouest Centre-Est Centre-Sud Centre-Sud Centre-Est Hauts-Bassin Hauts-Bassin Sud-Ouest Sud-Ouest Cascades Percentage points (%) Cascades Percentage points (%) (60,70] (60,70] (50,60] (50,60] (40,50] (40,50] (30,40] (30,40] (20,30] (20,30] (10,20] (10,20] [0,10] [0,10] E. HEALTH AND NUTRITION 1.16 Infant and child mortality significantly decreased since late in the 80s until the end of the 2000s. The objective of reducing infant mortality, mainly by reducing the child mortality rate by two thirds between 1990 and 2015, remains included in the fourth MDG. Figure 1.11 shows that infant and child mortality rates dropped during the period between 1993 and 2010. Actually, from an average of 98‰ during the decade preceding 1993, infant mortality declined to 75‰ on average during the 2000s, after having increased to around 108‰ during the 1990 decade. The same trend is observed for infant mortality which falls from 180‰ to 138‰. The most significant improvements seem to have been recorded in rural areas where mortality rates dropped by 28% against 18% in urban areas. The decline however appears to be similar for both areas concerning child mortality, for which a 26% decline rate is reported. It is worth stressing however that the decline observed over the last two decades does not mean that Burkina will achieve the planned objective of reaching a 67% reduction by 2015. Figure 1.11: Infant and child mortality rates, 1993-2010 (1) Infant mortality (2) Under-five mortality 150 250 200 Mortality rate (‰) Mortality rate (‰) 100 150 100 50 50 0 0 Urban Rural National Urban Rural National 1993 1998 2003 2010 -9- 1.17 There are significant regional disparities and the infant mortality rate among boys seems to be slightly higher than the rate recorded among girls. Figure 1.12 shows that the Sahel and Sud-Ouest regions hold the highest mortality rates with respectively 113‰ and 109‰ for infant mortality, and 221‰ and 189‰ for child mortality. On the other hand, the Centre-Est and the Centre regions have the lowest mortality rates with respectively 46‰ and 52‰ for infant mortality, and 75‰ and 89‰ for child and infant mortality. Figure 1.13 rather shows the difference between girls’ mortality rates and boys’ mortality rates. It clearly appears that mortality rates among boys are higher than rates among girls even though the gap has narrowed since the mid-90s. Figure 1.12: Infant and child regional mortality rates, 2010 200 Mortality rate (‰) Infant mortality Under-five mortality 150 100 50 0 d l t ns st tre ud l t un d s st tra he Es s or de or ue ue -E si ho -S en Sa en N -N ca as tre O -O tre ou C C tre as d- -B en tre en M u Su en C ts ea C en C du au C at C H e Pl l uc Bo Regions ranked by infant mortality level Figure 1.13: Gender differences in infant and child mortality rates, 1993-2010 0 Girls' rate minus boys' rate (‰) -5 -10 -15 -20 Infant mortality Under-five mortality 1993 1998 2003 2010 1.18 In general, child malnutrition prevalence dropped between 1993 and 2010. Malnutrition is generally seen as one of the major causes of infant mortality. Besides being a direct cause of mortality, it also makes children vulnerable to other diseases (Heltberg, 2009). - 10 - Therefore it is necessary to fight child malnutrition in order to reduce mortality in children. Regarding malnutrition or severe malnutrition, the results presented in Figure 1.14 show that compared with 1993, malnutrition worsened in 1998 and 2003, before improving in 2010. Thus, the stunting incidence, which was 40% in 1993, increased to 45% in 1998 before dropping to 35% in 2010. A similar trend is observed for the incidence of severe stunting which declined from 19% in 1993 to 15% in 2010, after a rise in 1998. Concerning the underweight rate, it experienced a slight decline over the period, from approximately 30% to 26% and after having increased in 1998 and 2003. The same trend is recorded for severe underweight. On the other hand, there seem to be some stability concerning wasting and severe wasting except in the year 2003 during which the prevalence rates significantly increased. The wasting incidence remained close to 16% while the rate for severe wasting was around 6%. Figure 1.14: Under-five nutritional status, 1993-2010 (1) Malnutrition rates (2) Severe malnutrition rates 50 25 40 20 Proportion of children (%) Proportion of children (%) 30 15 20 10 10 5 0 0 Stunting Underweight Wasting Stunting Underweight Wasting 1993 1998 2003 2010 1.19 The nutrition status in children was marked by regional disparities and a slight gender imbalance in favor the girls. The regional analysis presented in Figure 1.15 highlights Sahel as the region where the highest stunting prevalence is recorded with a rate close to 46%. This situation thus seems to keep up with the high infant mortality rate reported in this region. Though occupying the 4th place, stunting prevalence is also high in the Sud-Ouest region with a rate close to 40%, which is nearly the double of the rate prevailing in the Centre region. Stunting remains the most relevant malnutrition indicator of the level of welfare. Actually, the literature often recommends using this indicator (Pradhan et al., 2003; Wamani et al., 2004; Sahn and Younger, 2005) because stunting is related to the antenatal and postnatal conditions which include underweight at birth, poor health care, poor quality nutrition, as well as recurring infections. The two other indicators (underweight and wasting) however remain important for analyzing child short-term malnutrition. Concerning underweight, relatively low rates are recorded in the Hauts-Bassins, Centre, Boucle du Mouhoun, Centre-Sud and Centre-Ouest where they range between 20 and 23%. On the other hand, the Est and Sahel regions hold the highest rates with respectively 37 and 36%. Figure 1.16 shows that malnutrition is slightly accentuated for boys as compared with girls. Indeed, regarding stunting, the differences between the rate for boys and the rate for girls vary a little less than 3 to nearly 5 percentage points according to the - 11 - year. The gap remains the same for severe stunting, with always a difference of more than 3 percentage points. This trend at the disadvantage of boys is also observed for the other two malnutrition indicators even if the differences in absolute values are smaller. Figure 1.15: Under-five nutritional status by regions, 2010 50 Proportion of children (%) underweight Wasting Stunting 40 30 20 10 0 d l st ns st tre d l un t d s st tra he Es or de Su or ue ue -E si ho en Sa en N -N ca as e- tre O -O ou C C tre as r d- -B en t tre en M u Su en C ts ea C en C du au C at C H le Pl uc Bo Region ranked by stunting level Figure 1.16: Gender differences in under-five malnutrition, 1993-2010 (1) Malnutrition Severe malnutrition 0 0 -1 Gender gap in malnutrition (%) Gender gap in malnutrition (%) -1 -2 -2 -3 -3 -4 -5 -4 Stunting Underweight Wasting Stunting Underweight Wasting 1993 1998 2003 2010 1.20 While thinness in women is a concern in rural areas, obesity and overweight appear to be an urban phenomenon. The body mass index was used to rank women according to the WHO standards. As a result, four groups were identified based on the nutritional status of women: the group affected by thinness, the group of those women with a normal weight, the group of women affected by overweight, and that of women suffering from obesity. Thinness is a risk factor for a woman particularly during pregnancy, and therefore raises a maternal health issue. Moreover, the WHO acknowledges obesity and overweight as a risk factor leading to chronic diseases such as diabetes, cardiovascular diseases, as well as some cancers and muscular - 12 - disorders. This issue is a major concern since the fifth MDG is to improve maternal health, by decreasing inter alia by three quarter the maternal mortality rate between 1990 and 1995. Figure 1.17 shows the ranking of Burkinabe women according to their nutrition status, and its development since 1993. At the national level the proportion of women with a normal weight declined from approximately 80% in the 90s to a little less than 74% in the 2000s. This reduction kept pace with an increase in the number of women suffering from thinness and of women suffering from overweight or obesity. Actually, the proportion rose from 13% to nearly 15% in the first case, and from 8% to 12% in the second case between 1993 and 2010. In rural areas, the proportion of thinness cases overall decreased between 1993 and 2010, from 17% to 14% after a peak to 23% in 2003. This proportion however increases in urban areas where it goes up from 7% to 9%. While the obesity and overweight issue remains relatively insignificant in rural areas where only 5% of women are concerned in 2010, the proportion appears rather high in urban areas with an estimated rate of about 20% during the same year. Figure 1.17: Classification of women nutritional status, 1993-2010 (1) Rural (2) Urban 80 80 Proportion of women (%) Proportion of women (%) 60 60 40 40 20 20 0 0 Normal weight Obesity Overweight Underweight Normal weight Obesity Overweight Underweight (3) National 80 Proportion of women (%) 60 1993 1998 2003 2010 40 20 0 Normal weight Obesity Overweight Underweight F. EDUCATION 1.21 Burkina Faso remains characterized by a low educational level within adults, by relatively low net enrollment rates even if some significant progress was made over the period. Education is not only an excellent means for achieving other development goals, but also stands alone as an important dimension of welfare. The right to education is recognized as such in the Universal Declaration of Human Rights and the second MDG aims at providing primary education for all. Figure 1.18 shows the enrollment rate trends and the proportion of children who have never attended school in Burkina Faso. The net enrollment rate in primary school seems to have remained stagnant from 1993 to 2003, at around 27%, before jumping to 46% during the year 2010. This progress was mostly realized in rural areas where the rate nearly doubled between 2003 and 2010, or from 21% to 39%. At the same time the enrollment rate in urban areas rose from 68% to 76%. This progress is likely to have resulted from the implementation of the ten year basic education development plan in 2002/2003 the objective of which is to improve education access and coverage, through inter alia the introduction of a - 13 - tuition-free primary education as from 2005. The net enrollment rate in secondary education also improved even if the level remains relatively very low. Actually, in 2010 the net enrollment rate in secondary education accounts for approximately 15% at the national level, with 8% in rural areas against 37% in urban areas. The proportion of children aged 6-17 who have never attended school sharply decreased during the 2000s, which confirms the progress realized. Figure 1.19 highlights education deprivation rates among adults aged 18-64. At the national level, this rate dropped between 1993 and 2010, that is to say from 84% to 74%. But the decline was more marked in urban areas with a variation of 52% to 41% against 92% to 87% in rural areas. Figure 1.18: Deprivation in education and enrollment rates, 1993-2010 (1) Net primary school rates (2) Net secondary school rates 80 40 Enrollment rates (%) Enrollment rates (%) 60 30 40 20 20 10 0 0 Urban Rural National Urban Rural National (3) Proportion of children (6 to 17 years) with no education 80 Pourcentage d'enfants 60 1993 1998 2003 2010 40 20 0 Urban Rural National 1.22 Gender inequality regarding education slightly decreased among the adult population while it significantly decreased regarding children access to education. Part (2) of Figure 1.19 shows that the difference between women’s deprivation rate and men’s deprivation rate is rather high in urban areas with a difference in percentage points reaching 15 and remains stable over the period. Gender inequality is less marked in rural areas, but kept on constantly increasing between 1993 and 2010 as the difference rises from 7.5 to 10. On the other hand a reduction in gender inequality is noted for access to education among children. Indeed, as Figure 2.20 illustrates it, inequality between girls and boys significantly decreased as a matter of net enrollment rate in primary education, and the same also confirmed for deprivation of education among children aged 6-17. Concerning the net enrollment rate in secondary education, gender inequalities are relatively insignificant in terms of absolute differences, but are relatively high when considering the rate which is low itself. - 14 - Figure 1.19: Proportion of uneducated individuals aged 18 to 64 years, 1993-2010 (1) Deprivation rates (2) Women deprivation rates minus men rates 100 15 80 Gender gap in percentage points Percentage of population 10 60 40 5 20 0 0 Urban Rural National Urban Rural National 1993 1998 2003 2010 Figure 1.20: Gender differences in enrollment and education deprivation, 1993-2010 10 Gender gap in percentage points 5 0 -5 -10 Net secondary Net primary Children (6 to 17 years) enrollment enrollment with no education 1993 1998 2003 2010 G. SOCIAL PROTECTION 1.23 While some significant improvement is recorded since the 1990s, Female Genital Mutilation (FGM) and early marriage continue to be a serious issue in Burkina Faso, particularly in rural areas. Practices like FGM, as well as sexuality and early marriage, are likely to curtail human rights. Figure 1.21 enlightens that the occurrence of FGM significantly declined in the country as, among women aged 15-24, the proportion of women having undergone a genital mutilation decreased from 85% in 1998 to nearly 65% in 2010. The improvement is more significant in urban areas where the rates fell from 97% to 56% against - 15 - 82% and 69% in rural areas. Early sexuality prevalence also decreased between 1993 and 2010, but still remains high. Indeed, the proportion of women having had sexual intercourse before being 16 still accounts for more than 20% in the country in 2010, namely 13% in urban areas and 26% in rural areas. Early pregnancy, which only slightly declined over the period, also remains a serious concern especially in rural areas where the proportion of affected women accounts for 28% against 12% in urban areas. Early marriage also remains a primarily rural phenomenon though the incidence in cities is not negligible. Actually, the proportion of women having got married precociously decreased between 1993 and 2010 from 34% to 21%. In rural areas it fell from 57% to 50% over the same period. Figure 1.21: Teenagers and social protection (cohort of women aged 15 to 24 years - 1993-2010 (1) Urban (2) Rural 100 80 Proportion of women Proportion of women 80 60 60 40 40 20 20 0 0 Sex Sex Genital Early Early Sex Sex Genital Early Early before 18 before 16 mutilation pregnancy marriage before 18 before 16 mutilation pregnancy marriage (3) National 80 Proportion of women 60 1993 1998 40 2003 2010 20 0 Sex Sex Genital Early Early before 18 before 16 mutilation pregnancy marriage 1.24 The regional analysis reveals strong disparities at the disadvantage of a number of regions such as the Sahel and the East, while the Center appears to be less affected than the others. Indeed, in light of Figure 1.21 while less than 20% of women experience early marriage in the Center region, a figure of more than 60% is recorded in the Sahel and East regions. The Sahel is also mostly affected by early sexual intercourse with a percentage of more than 50% women concerned. Regarding early pregnancies, 4 regions (Cascades, East, Sahel, and South- West) are top with proportions standing at more than 30%. Concerning FGM, the North and Center-East regions appear to be mostly affected with a proportion of more than 80% women. On the other hand, the Center-West and Center-South regions are the least affected and display a proportion of less than 50% women. - 16 - Figure 1.22: Social protection and teenagers deprivation by regions, 2010 (1) Early marriage (2) Early pregnancy Sahel Sahel Nord Nord Centre-Nord Centre-Nord Boucle du Mo Plateau Boucle du Mo Centre Cent Est Centre Cent Plateau Est Centre-Ouest Centre-Ouest Centre-Est Centre-Sud Centre-Est Centre-Sud Hauts-Bassin Hauts-Bassin Proportion of women (%) (60,80] Proportion of women (%) Cascades Sud-Ouest Cascades Sud-Ouest (50,60] (30,40] (40,50] (20,30] (30,40] (10,20] (20,30] [0,10] [0,20] (3) Early intercourse (4) Female genital mutilation Sahel Sahel Nord Nord Centre-Nord Centre-Nord Boucle du Mo Plateau Boucle du Mo Centre Cent Est Centre Cent Plateau Est Centre-Ouest Centre-Ouest Centre-Est Centre-Sud Centre-Est Centre-Sud Hauts-Bassin Hauts-Bassin Proportion of women (%) Proportion of women (%) Cascades Sud-Ouest Cascades Sud-Ouest (50,60] (80,90] (40,50] (70,80] (30,40] (60,70] (20,30] (50,60] (10,20] [40,50] [0,10] 1.25 Although some efforts are being deployed since 2003, there still remains a large proportion of children who do not have a birth certificate. As Figure 1.23 shows it, 30% of children under 5 did not have a birth certificate in 2010 while this rate nearly reached 56% in 2003. Though the situation is not gender specific, the phenomenon seems inversely more stressed in rural areas with a proportion of 35% children for 7% in urban areas. Figure 1.24 reveals many regional disparities. Consequently, while more than 60% children do not have a birth certificate in the Sahel region, their number account for less than 10% in the Center, Hauts- Bassins and Center Plateau regions. The rates in the other regions range between 10% and 30%, except for the Center-West region where it stands at around 36%. Figure 1.23: Proportion of under-five children without birth certificate, 2003-2010 60 Proportion of children 40 20 0 Female Male National Rural Urban 2003 2010 - 17 - Figure 1.24: Proportion of under-five children unregistered or without birth certificate by region, 2010 Sahel Nord Centre-Nord Boucle du Mo Plateau Cent Centre Est Centre-Ouest Centre-Sud Centre-Est Hauts-Bassin Proportion of children (%) Sud-Ouest (60,70] Cascades (50,60] (40,50] (30,40] (20,30] (10,20] [0,10] - 18 - References Brown, A. C. (2009), "Poverty Issues in Infrastructure Regulation: Why, Who and How?," Poverty in Focus, 18: 13–15. Ferreira, F. H. G. (2011), "Poverty is multidimensional. But what are we going to do about it?," Journal of Economic Inequality, 9: 493–495. Hailu, D. and R. Tsukada (2009), "Equitable Access to Basic Utilities: An Overview," Poverty in Focus, 18: 3–5. Heltberg, R. (2009), "Malnutrition, Poverty, and Economic Growth," Health Economics, 18: S77-S88. Lanjouw, P. (2012), "Counting Women in Poverty: Potential Pitfalls in Conventional Poverty Analysis," PREM Knowledge and Learning Weeks, World Bank, May 8. McKay, A. (2009), "Assets and chronic poverty: Background paper," Working Paper no 100, Chronic Poverty Research Centre. Ministère de l'Habitat et de l'Urbanisme (2008), Politique nationale de l'habitat et du développement urbain, Burkina Faso, Mars 2008. Pradhan, M., Sahn, D. E. et S. D. Younger (2003), "Decomposing world health inequality," Journal of Health Economics, 22: 271-293. Ravallion, M. (2011), "On Multidimensional indices of poverty," The Journal of Economic Inequality, 9: 235–248. Ravallion, M. (2012), "Mashup Indices of Development," The World Bank Research Observer, 27: 1–32. Sahn, D. E. et S. D. Younger (2005), "Improvements in children’s health: Does inequality matter?," Journal of Economic inequality, 3: 125-143. SCADD (2011), Stratégie de Croissance Accélérée et de Développement Durable 2011-2015, Burkina Faso. Sen, A. (1976), "Poverty: An Ordinal Approach to Measurement," Econometrica, 44, 219–231. Sen, A. (1985): Commodities and Capabilities, Amsterdam: North-Holland. Wamani, H., Tylleskär, T., Åstrøm, A. N., Tumwine, J. K. et S. Peterson (2004), "Mothers’ education but not fathers’ education, household assets or land ownership is the best predictor of child health inequalities in rural Uganda," International Journal for Equity in Health, 3: 9. WHO (2006), Report on the WHO Technical Meeting on Quantifying Disease from Inadequate Housing, Bonn, Germany, 28-30 November 2005, Copenhagen: WHO Regional Office for Europe - 19 -