WPS4954 Policy Research Working Paper 4954 The Mauritanian Labor Market through the Lens of the 2004 National Household Survey Tania Rajadel Nicola Pontara Maria Laura Sanchez Puerta The World Bank Poverty Reduction and Economic Management Network--Africa Region Africa Technical Families & Human Development Network Social Protection Team June 2009 Policy Research Working Paper 4954 Abstract This paper provides a snapshot of Mauritania's labor mar- high. Women systematically earn less than men inde- ket using data from the 2004 national household survey. pendently of their sector and type of employment and The results show that the labor market is characterized by controlling for other factors, such as education. Young lower participation rates, lower employment-to-popula- adults face considerable difficulties in entering the labor tion rates, and relatively higher unemployment rates than market: more than half of the population aged 15­24 is in neighboring countries. The non poor fare better in neither studying nor participating in the labor force. As the labor market than the poor. Although the labor force gender disparities remain important for similar levels of participation of the poor is higher than that of the non education, more work is needed to understand whether poor, the poor display a higher unemployment rate and a cultural factors may prevent women from entering the lower employment rate than the non poor. The data also labor market. Concerning young adults, future poverty suggest a negative correlation between wage employment reduction strategies need to pay more explicit attention and poverty. Substantial differences in labor market indi- to the promotion of employment through informed labor cators emerge when disaggregating the analysis by gender market policies. and age-group. Female non-participation is extremely This paper--a product of the Africa Technical Families , Poverty Reduction and Economic Management Network (Africa Region) and the Social Protection Team, Human Development Network--is part of a larger effort in the networks to analyze the labour markets of low-income countries. Policy Research Working Papers are also posted on the Web at http:// econ.worldbank.org. The authors may be contacted at npontara@worldbank.org and msanchezpuerta@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team The Mauritanian Labor Market through the Lens of the 2004 National Household Survey 1 Tania Rajadel World Bank, Washington DC, USA Nicola Pontara World Bank, Washington DC, USA Maria Laura Sanchez Puerta World Bank, Washington DC, USA and IZA, Bonn, Germany JEL Classification: O-12, J-7, J-21, J-31 Key words: Mauritania, Labor Market, Wage Determination, Gender, Youth, Poverty. 1 We are grateful to the Mauritanian authorities for their support, notably the National Institute of Statistics, (ONS) for the availability of the 2004 National Household Survey data, and the Ministry of Economic and Finance (MEF). We also thank for insight and comments Alexandre Kolev and Paul Cichello. The findings, interpretations, and conclusions are our own and should not be attributed to the World Bank, its Executive Board of Directors, or any of its member countries. The usual disclaimers apply. ii Table of Contents INTRODUCTION 1 1. DATA SOURCES 3 2. MAURITANIA'S MAIN LABOR MARKET INDICATORS 4 2.1 Non-Participation 4 2.2 Unemployment 7 2.3 Employment 9 3. SPECIFIC ISSUES 14 3.1 Poverty and Labor Market Outcomes 14 3.1.1 Non-participation 14 3.1.2 Unemployment 15 3.1.3 Employment 17 3.2 Gender Gaps in the Labor Market 20 3.2.2 Gender gaps in rural labor markets 21 3.3.3 Gender gaps in urban labor markets 23 3.3 Youth 25 3.4 Earnings Determination 28 3.4.1 Earnings of the self-employed 29 3.4.2 Wage earners 30 CONCLUSIONS 35 REFERENCES 36 ANNEX 1. LABOR MARKET INDICATORS: DEFINITIONS 37 ANNEX 2. ADDITIONAL TABLES 38 iii Figures Figure 1. Non-Participation by Age Group and Education Level (in %)................................................ 7 Figure 2. Unemployment by Age Group and Education Level (in %) ................................................... 9 Figure 3. Employment-to-Population by Age Group and Level of Education (in %) .......................... 10 Figure 4. Types of Employment and Employer (in %) ......................................................................... 11 Figure 5. Types of Employment and Education (in %) ........................................................................ 11 Figure 6. Sectors of Employment ......................................................................................................... 12 Figure 7. Main Labor Market Indicators by Education and Poverty Status (in %)............................... 14 Figure 8. Types of Employment and Education by Poverty Status (in %) ........................................... 19 Figure 9. Sectors of Employment by Poverty Status (in %) ................................................................. 20 Figure 10.Types of Employment in Rural Areas by Gender (%).......................................................... 22 Figure 11. Types of Employment in Urban Areas by Gender (in %) ................................................... 24 Figure 12. Main Labor Market Indicators by Education Level for Youth (in %) ................................. 27 Figure 13. Types of Employment for Youth (aged 15 to 24) and Adults (aged 25 to 64) .................... 28 Figure 14. Adjusted Wage Differentials by Level of Education and Gender ....................................... 32 Figure 15. Wage Earners - Kernel Density in the Formal and Inf. Sectors (in thousands of MRO) .... 33 Tables Table 1. Main Labor Market Indicators ­ Comparison with Neighboring Countries (in %) 4 Table 2. Main Labor Market Indicators (in %) 5 Table 3. Non-Participation Rate by Gender, Living Area, and Age Group (in %) 6 Table 4. Reasons for Not Participating in the Labor Force by Gender and Living Area (in %) 6 Table 5. Share of Wage Earners in the Informal Sector (in %) 12 Table 6. Non-Participation Rate by Poverty Status, Gender, Living Area, and Age Group (in %) 15 Table 7. Unemployment Rate by Poverty Status, Gender, Living Area, and Age Group (in %) 16 Table 8. Employment-to-Population by Poverty Status, Gender, Living Area, and Age Group (in %)17 Table 9. Types of Employment and Poverty (in percentages) 18 Table 10. Main Labor Market Indicators by Gender and Education (in %) 20 Table 11. Out of the Labor Force ­ Motives for Women in Rural Areas (in %) 21 Table 12. Rural Labor Force Participation by Gender and Family Status* (in %) 22 Table 13. Out of the Labor Force ­ Motives for Women in Urban Areas (in %) 23 Table 14. Urban Labor Force Participation by Gender and Family Status* (in %) 24 Table 15. Main Labor Market Indicators for Youth (in %) 25 Table 16. Youth Out of the Labor Force ­ Motives (in %) 26 Table 17. Types of Employment for Youth by Gender and Living Area (in percentages) 28 Table 18. Self-Employed ­ Monthly Earnings by Living Area and Gender (in thousands of MRO) 29 iv Table 19. Self-Employed ­ Monthly Earnings by Gender and Education (in thousands of MRO) 30 Table 20. Wage Earners ­ Monthly Earnings, by Living Area, Gender, and Type of Employer (in thousands of MRO) 31 Table 21. Wage Earnings ­ by Gender and Education (in thousands of MRO) 31 Table 22. Wage Earners - Econometric Estimates of Wage Determinants 34 -- 0 -- Annex Table A Education Level by Gender and Age Group (in percentages) ..................................... 38 Annex Table B. Types of Employment and Employer by Poverty Status (in percentages) ................. 38 Annex Table C. Sector of Employment ­ Males in Rural Areas (in percentages) ............................... 38 Annex Table D. Sector of Employment ­ Females in Rural Areas (in percentages) ............................ 39 Annex Table E. Sector of Employment ­ Males in Urban Areas (in percentages)............................... 39 Annex Table F. Sector of Employment ­ Females in Urban Areas (in percentages) ........................... 39 Annex Table G. Main Labor Market Indicators ­ Youth and Poverty (in percentages) ....................... 40 Annex Table H. Self-Employed ­ Monthly Earnings and Poverty (in thousands of MRO)................. 40 Annex Table I. Incidence of Poverty by Region and Relative Contributions (2000-2004) .................. 40 1 Introduction Endowed with natural resources and home to 2.9 million people, Mauritania is a low-income country, with gross domestic product (GDP) of US$ 2.7 billion and per capita gross national income (GNI) of $938 (World Bank, 2008). The country's economy has been dependent on natural resources since the 1950s: iron ore first; then combined with fisheries; and more recently oil, gold and copper. The country has gone through massive structural changes. Whereas at independence (1960) barely one- fifth of the population was settled in towns and villages, nomads today constitute less than 5 percent of the population. The severe drought of the late-1960s and early-1970s caused many Mauritanians to seek refuge in the cities and the sharp increase in aid flows made it possible for most of them not to return to their rural homelands. 2 As a result, the importance of agriculture and ­ to a lesser but still significant extent ­ livestock, has diminished overtime, due to the demise of traditional livelihoods (Auty and Pontara 2008). 3 Mauritania has not been able to substantially improve the well-being of the population in recent years. Poverty is still pervasive, affecting 46.7 percent of the population, and mainly a rural phenomenon, as the rural poor made up approximately 75 percent of all of Mauritania's poor. Between 2000 and 2004, the overall poverty incidence declined from 51 to 46.7 percent, the net effect of declining poverty in rural areas (from 66 to 59 percent) and increasing poverty in urban areas (from 28 to 29 percent, Table I). Over the same period, the Gini coefficient remained at 0.39, indicating that the moderate decrease in poverty did not make a dent on inequality. 4 The 2007 Human Development Index (HDI) ranked Mauritania 137th of a total of 177 countries (UNDP 2007). In addition, recent work conducted by the World Bank concluded that with current resources and policies Mauritania is highly unlikely to reach most of the Millennium Development Goals (see Magnoli-Bocchi et al., 2008). This paper represents the first attempt to analyze the Mauritanian labor market, on the basis of the data of the 2004 National Household Survey (EPCV 2004), which is the most recent and reliable source of labor market data in Mauritania. While the authorities have aimed in recent years to reduce the incidence of poverty by including the promotion of labor-intensive non-oil growth in their poverty reduction strategies (GIRM 2000, GIRM 2006), the country has so far not fully managed to: (i) diversify the sources of growth besides the exploitation of natural resources and the tertiary sector; 2 The post-colonial capital, Nouakchott, dominates the urban settlement hierarchy and hosts over 50 percent of the urban population and around a quarter of the national population (GIRM 2000). 3 In 2006, services accounted for 40 percent of GDP, agriculture and livestock 30 percent (with livestock making up just less than two-thirds of this magnitude), mining 12 percent, oil 12 percent, and fisheries 6 percent (IMF 2007). 4 Concerning the working-age population (15-64 years), 44 percent lives in a poor household and 26.6 percent in an extremely poor household (Table 2). 2 and (ii) attract investments by creating an enabling environment for private sector development. In the labor market policy arena, the paucity of reliable analyses of the labor market has resulted in an unfocused attention on employment generation and promotion (EGP) in both the Poverty Reduction Strategies (PRPSs) and the domestic policy debates. 5 Building on the work of Pontara (2007) and Rosanvallon (2006), this paper aims to redress the lack of information on the labor market in Mauritania ­ focusing in particular on gender issues, youth's access to jobs, and earnings determinants. The aim is to provide policy makers with a starting analytical basis, which can be enriched by further analysis and work, as new data become available. Section 1 briefly reviews the data sources for this analysis. Section 2 presents the data analysis of the main labor indicators, focusing on non-participation, unemployment, and employment patterns. Section 3 focuses on specific labor market issues including poverty, gender gaps, youth's integration in the labor market, and earnings determination. The final section summarizes the results and highlights issues for further investigation. 5 According to Rosanvallon (2006), until 1997, statistics on the main labor market indicators were not collected in Mauritania. Historically, therefore, policy makers have had scant data to formulate a sound labor market policy. Data scarcity on the labor market is not something specific to Mauritania, but to Sub-Saharan Africa (SSA) as a whole. Sender et al. (2005) maintain that data on labor supply and employment are difficult to obtain in Africa, and are less reliable than other socio-economic indicators. 3 1. Data Sources The analysis conducted in this paper is based on the 2004 National Household Survey, EPCV 2004. The EPCV 2004 is the last in a series of nationally representative surveys that the Mauritanian authorities have carried out since 1992. It was conducted to update the main household well-being indicators, with a particular focus on poverty, and support the preparation of Mauritania's second PRSP, as well as the elaboration of regional poverty reduction strategies in the wilayas (provinces) of Adrar, Brakna, Nouadhibou, Inchiri and Trarza. Four types of questionnaires were elaborated for the EPCV 2004: (i) a core welfare indicators questionnaire (CWIQ) covering basic social indicators6; (ii) an expenditure questionnaire, focusing on current and occasional expenditures, transfers, revenues, credit and savings; (iii) a price questionnaire, focusing on the price of principal products in the different zones covered by the survey; and (iv) a community questionnaire, focusing on the local infrastructure in the areas covered by the survey (GIRM 2006a). At the household level, the EPCV 2004 provides information on households' assets, income, and expenditures. At the individual level, the data includes basic demographic characteristics (e.g. gender, age) as well as data on education, health, and employment. In order to better capture the important effects linked to seasonality, the EPCV 2004 was carried out during two separate phases: (i) a first 3- month phase ­ between end-August and end-November 2004 ­ during which information was collected on social indicators and household expenditure; and (ii) a second phase ­ between April- June 2005 ­ during which information was collected on education (which was not covered in the previous phase under the CWIQ) and household expenditure. The EPCV sampled a subset of households by using two-tier random sampling in each wilaya of the country. The sampling base during the first phase (sampling of primary units) was established on the basis of a list of Census Districts (CD) extracted from the 2000 Census. Subsequently households were randomly selected within each CD. The surveyed population was in the order of 52,609 individuals, of which 29,038 are of working age (15 to 64 years old), belonging to 9,385 households. The sample comprised 95 percent of sedentary people. Poverty was determined at the household level based on per capita expenditures using a 1$ a day per capita poverty line. After taking into account inflation and exchange rates, this poverty line in the EPCV 2004 translates into 94,600 Mauritanian Ouguiyas (MRO) per year and per capita for 2004. The poverty line for extreme poverty was set at 71,550 MRO (see GIRM 2006a). 6 The CWIQ is a package which aims at collecting information to measure the access to, utilization of, and levels of satisfaction concerning the main social and economic services. The CWIQ sheets can be scanned, reducing the data entering, cleaning, and analysis phases. 4 2. Mauritania's Main Labor Market Indicators The Mauritanian labor market appears to be less `dynamic' than that of its neighbors. As shown in Table 1, 49.3 percent of the Mauritanian working-age population participates in the labor market, a lower figure than in nearby countries. The employment-to-population ratio is 42.1 percent, lower than in Senegal (62 percent), Burkina Faso (81.5 percent) and Mali (70.2 percent). Besides, 14.7 percent of the labor force is unemployed in Mauritania, which is above unemployment rates displayed by neighboring countries (5.8 percent for Senegal, 8.8 percent for Mali, 11.2 percent for Morocco). Female participation, which stands at 28 percent, is also well below that displayed by neighboring countries. Table 1. Main Labor Market Indicators ­ Comparison with Neighboring Countries (in %) Mauritania b Mali c Morocco c Burkina Faso c Senegal d Labor force 49.3 participation a 79.3 55.4 85.0 70.7 Employment-to- 42.1 population ratio 70.2 46.9 81.5 62 Unemployment rate a 14.7 8.8 11.2 ... 5.8 Female participation in 28.0 the labor force 47.5 28.7 46.5 42.5 Figures are displayed in percentage; a Standard ILO definition. Sources: b: EPCV 2004 ; c: World Banks 2005, 2006 ; d: "Analyse des Conditions du Marché du Travail dans les Pays Sous-Développés ­ Cas du Sénégal", World Bank 2007. 7 2.1 Non-Participation The overall non-participation rate for Mauritania stands at 50.7 percent and is similar for the poor and the non poor (Table 2 and Table 3). Female non-participation drives this statistic, as 72 percent of working-age women do not participate in the labor market. Striking discrepancies between female and male non-participation arise within the prime-age group (25-54). This same group reports family or household obligations, which fall upon women, as the chief reason for being out of the labor force (Table 4). This finding is dramatically at odds with the corresponding averages found in Sub-Saharan Africa (SSA), where around 70 percent of women are considered active participants in the labor market; but it is comparable with those found in the Middle-East and North Africa region (MENA), where around only 40 percent of women participate actively in the labor market (Kapsos, 2007). 7 The definitions of the labor market indicators used in this paper are provided in Annex 1. 5 Table 2. Main Labor Market Indicators (in %) Share of the working Labor force Employment-to- Unemployment Relaxed Relaxed labor force age population participation population ratio Rate unemployment rate* participation* All 49.3 42.1 14.7 25.3 56.3 Gender Male 46.8 73.4 65 11.4 18.4 79.7 Female 53.2 28 21.8 22.3 38.7 35.6 Living area Rural 57.1 49.8 42.7 14.2 23.1 55.6 Urban 42.9 48.6 41.2 15.3 28.1 57.2 Age group 15-24 36.4 32.7 21.6 33.8 52.4 45.5 25-34 22.5 52.5 43 18.1 27.4 59.2 25-54 33.5 63 60 4.9 8.4 65.4 55-64 7.6 56 54.9 1.9 3.7 57 Poverty Non poor 55.9 48.7 42.5 12.6 23.6 55.6 Poor 44.1 50.1 41.5 17.2 27.4 57.1 Extremely poor 26.6 51 40.9 19.8 29.1 57.6 Male by age group 15-24 17.2 45 31.3 30.3 45.4 57.3 25-34 9 85.6 74.6 12.8 18.3 91.4 25-54 16.3 94.5 91 3.7 5.4 96.2 55-64 4.3 78.6 77.4 1.5 3.3 80 Female by age group 15-24 19.2 21.6 12.9 40.4 63 34.8 25-34 13.5 30.3 21.8 28.2 42.2 37.7 25-54 17.2 33.4 30.6 8.1 16 36.5 55-64 3.3 26.5 25.6 3.4 5.2 27 Male by living area Rural 26.4 71.3 68.8 10.5 15.9 68 Urban 20.4 67.7 60 12.7 21.9 64.2 Female by living area Rural 30.7 28.6 20.3 23.3 38.4 32 Urban 22.5 32.3 23.9 20.8 39.2 35.8 Source: EPCV 2004. Figures are displayed in percentage. * The relaxed definition for unemployment includes all the unemployed according to the ILO definition as well as all individuals who did not work in the reference week and did not look for a job because they thought there were no jobs available. 6 Table 3. Non-Participation Rate by Gender, Living Area, and Age Group (in %) Living Area Age Group Rural Urban 15-24 25-34 35-54 55-64 All 50.7 50.2 51.4 67.3 47.5 37 44 Net of individuals attending school 45.1 46.6 42.9 55.1 46.4 37 44 By gender: Male 26.6 23.1 31.2 55 14.4 5.5 21.4 Male, net of those attending school 16.9 16.1 18.1 35.5 11.8 5.5 21.4 Female 72 73.5 69.8 78.3 69.6 66.6 73.5 Female, net of those attending school 69.1 71.9 65 71.3 69.1 66.6 73.5 Source: EPCV 2004. Figures are displayed in percentage. Non-participation is also high among younger groups and decreases gradually with age before going up again for people aged 55 or more. Over two-thirds of the population aged 15 to 24 is out of the labor force compared to less than 40 percent of those aged 35 to 54. Part of the youth is in school. However, over half of the young people aged 15 to 24 and not attending school remains out of the labor force (55.1 percent). Strikingly, the 25 to 34 age group displays a substantially higher non- participation rate than the 35 to 54. This gap appears to be starker for men than women, which is unsurprising given that the latter's overall participation in the labor force is low. These figures suggest a late integration in the labor market for males in particular. Table 4. Reasons for Not Participating in the Labor Force by Gender and Living Area (in %) Male Female All Rural Urban Rural Urban No work 15.5 26.1 27.7 10.1 14.4 Slack season 1 3.7 2.6 0.3 0.5 Student 20.6 36.5 51.5 8 20 Family/household obligations 50.3 11.4 2.8 70.2 56 Too young/old 9.9 16.2 11 9.3 7.7 Disabled 2.2 5.4 3.9 1.5 1.2 Other 0.5 0.8 0.6 0.5 0.3 Total 100 100 100 100 100 Source: EPCV 2004. Figures are displayed in percentage. Total might not always exactly amount to 100 due to rounding. Rural and urban non-participation vary across gender groups. Although differences between rural and urban areas are small when considering the whole sample (Table 3), men living in urban centers are more likely to opt out of the labor force than those in rural regions, whereas the opposite is true for women. For men, this observed discrepancy seems mostly due to the high proportion of students among non-participants in the labor market in urban centers. The gap between rural and urban areas narrows when considering non-participation rates net of those attending school (respectively 16.1 and 18.1 percent, see Table 4). While roughly similar proportions of men residing in rural and urban areas 7 report being out of the labor force because of the lack of work (respectively 26 and 27.6 percent), some 51.5 percent of urban males, against 36.5 percent of rural males, mention `school' as a reason for non-participation. For women, family or household obligations weigh heavily on their decision to remain out of the labor force in rural areas (70 percent) more so than in urban centers (56 percent), where they are also more likely to be studying (20 percent in urban areas versus 8 percent in rural areas, Table 4). Figure 1. Non-Participation by Age Group and Education Level (in %) 15-24 25-34 None Primary Secondary 35-54 Higher 55-64 0% 20% 40% 60% 80% 100% Source: EPCV 2004. Non-participation decreases with education, except for young adults. As shown in Figure 1, for prime age workers and seniors, education appears to be positively correlated to labor force participation. The reverse is observed for the youth, who have benefited from more schooling than their elders thanks to the considerable efforts Mauritania deployed in the 1990s to improve access to education but still experience difficulties in entering the labor force. This is only partly due to the fact that some of the youth is still in school (see section 3.3). 2.2 Unemployment The unemployment rate for Mauritania stands at 14.7 percent (see Table 2). The relaxed unemployment rate ­ also used in this study ­ stands at 25.3 percent for the population as a whole, thus suggesting a high level of discouragement among participants in the labor market. Unemployment is slightly higher in urban areas than in rural ones, but the poor, who live predominantly in rural regions, are more likely to be unemployed than the non poor. Rural and urban unemployment rates are fairly close (respectively 14.2 and 15.3 percent, Table 2). However, the gap widens when considering the relaxed definition for unemployment (23.1 percent in rural areas and 28.1 percent in urban areas), thus suggesting that discouragement is higher in urban centers. While 8 17.2 percent of the poor are unemployed, 12.6 percent of the non poor are. Interestingly, unemployment rates for the poor are the same in rural and urban areas, standing at 17.2 percent, whereas they are higher for the non poor in urban centers (14.6 percent compared to 10.2 percent in rural regions, see Table 7). The male unemployment rate is lower than the overall population's but discouragement is significant, in particular in urban areas. On average, 11.4 percent of men are unemployed (Table 2). Those living in urban areas display higher unemployment rates than their counterparts in rural regions (respectively 12.7 percent and 10.5 percent). Moreover, for men, discouragement appears to be higher in urban centers, with a relaxed definition for unemployment standing at 21.9 percent. Women are twice as likely as men to be unemployed and exhibit significantly higher levels of discouragement than men. The overall female unemployment rate stands at 22.3 percent (Table 2). Women living in urban centers seem to fare somewhat better than their counterparts in rural areas, displaying a slightly lower standard unemployment rate. However, although female relaxed unemployment rates are similar in rural and urban areas, the gap between the relaxed and standard rates is higher for urban centers (respectively 18.4 and 15.1 percentage points), suggesting that discouragement is more widespread in urban than rural regions. In addition, women aged 15 to 24 face substantially higher unemployment rates than any other age and gender group. It is important to keep in mind that female non-participation is significant across all age groups. Women might indeed decide to opt out of the labor market when they are unemployed, thus artificially decreasing observed unemployment rates. Young people face substantially higher unemployment rates than the rest of the working-age population. While 33.8 percent of the 15 to 24 year-old group is unemployed, less than 5 percent of the 25 to 54 are unemployed (Table 2). Similarly to what was observed for participation in the labor market, the 24 to 35 age group experience difficulties in accessing jobs, displaying an 18.1 percent unemployment rate. Discouragement also appears to be a major issue for youth whose relaxed unemployment rate towers at 52.4 percent. As for the overall population, the gender gap among young people is significant: young women exhibit higher unemployment rates than young men, using both standard and relaxed definition. Unemployment appears to increase with education for young adult workers in particular (Figure 2). Among individuals over the age of 25, unemployment rates are higher for those with primary and secondary schooling. Those with higher education display lower unemployment rates, although these figures should be interpreted with caution given the small size of the samples. As far as youth is concerned, those with primary education are less likely to be unemployed than the others whereas 9 those with secondary education exhibit a strikingly high unemployment rate. This could suggest that secondary schooling does not provide its students with the skills required to enter the labor market, although further analysis ­ including on the determinants of labor demand ­ would be required to investigate this finding. Figure 2. Unemployment by Age Group and Education Level (in %) 15-24 25-34 None Primary Secondary 35-54 Higher 55-64 0% 10% 20% 30% 40% 50% Source: EPCV 2004. 2.3 Employment The employment-to-population ratio is low in Mauritania. Table 2 shows that only 42.1 percent of working-age Mauritanians are employed. The poor and non poor display similar employment-to- population ratios. Indeed, 42.5 percent of the non poor are employed compared to 41.5 percent for the poor and 40.9 percent for the extremely poor (Table 2). This results from the fact that, although the poor are slightly more likely to participate in the labor market, when they do, they face higher unemployment rates. A substantial gender gap emerges when comparing male and female employment-to-population ratios. Sixty-five percent of males aged 15 to 64 are employed against only 21.8 percent of females of the same age group (Table 2). Men's employment-to-population ratio is higher in rural areas than in urban centers, which is consistent with previously discussed results showing that both unemployment and non-participation were lower in rural areas for men. The reverse is true for women: urban women appear to participate more actively in the labor market than those in rural areas. This finding, however, should be interpreted with caution as the frontier between domestic chores and agriculture- related work might be somewhat blurred in rural households. Women in rural areas might actually be employed while reporting being inactive. 10 Employment-to-population is particularly low for youth. It stands at 21.6 percent compared to respectively 43 percent and 60 percent for the 25 to 34 and the 35 to 54 age groups (Table 2). Young women in particular display the lowest employment-to-population ratio of all categories, standing at 12.9 percent. The employment gap with older women is, however, less marked than it is for men, mostly because female labor force participation is low throughout all age groups. Although 31.3 percent of all young men are employed, the employment-to-population ratio goes up to 74.6 percent and 91 percent for males aged 25 to 34 and 35 to 54. This differential appears to be due to both substantially higher labor force participation and lower unemployment rates among the latter groups. Employment-to-population increases with education for prime-age workers. The employment-to- population ratio is lowest among youth, whose levels of education are higher than the overall population's. For the 15-to-24 year-old group, employment is highest among those with no education and lowest for those with secondary and higher education, which is consistent with the fact that some are still in school. As expected, employment for prime-age workers increases with education. Differences between levels of education seem to become more marked as workers gain experience. For instance, differences in employment ratios between people with secondary schooling and with primary schooling are strongest for the 35-to-54 age group. The benefits of having some schooling, as opposed having no education, is starker for elder workers (55 to 64 years of age, Figure 3). Figure 3. Employment-to-Population by Age Group and Level of Education (in %) 15-24 25-34 None Primary Secondary 35-54 Higher 55-64 0% 20% 40% 60% 80% 100% Source: EPCV 2004. Self-employment is predominant in Mauritania, gathering over half of the employed population (Figure 4). As shown in Figure 5, three-quarters of the self-employed have received no education. 14 percent went to primary school, 9 percent to secondary school and virtually none have a higher 11 level of education. Most of the self-employed are farmers or livestock breeders (46 percent of the self- employed) while 17.6 percent are sales-persons, and 15.5 percent are traders or shopkeepers. Wage earners ­ 30.4 percent of the employed ­ form the second major group of employed people (Figure 4). Figure 5 shows that, on average, wage earners exhibit higher levels of education than the overall population: only 35 percent of them have no schooling, while 33 percent attended secondary school, and 11 percent studied beyond high school. According to Figure 4, less than a fifth of wage earners report working for a private company. They are mostly employed by the public sector (40 percent of all wage workers) or smaller structures such as private individuals or households (35 percent). Figure 4. Types of Employment and Employer (in %) 52.1% Self-employed Unpaid family 9.1% Public sector worker/apprentice Private company 8.4% Individual/household Day laborer Other 30.4% Wage earner 0% 10% 20% 30% 40% 50% 60% Source: EPCV 2004. Unpaid family workers and apprentices represent nearly a tenth of the employed population (Figure 4). They work mostly for private individuals or households. They exhibit low levels of education: 70 percent never went to school whereas 25 percent attended primary school (Figure 5). Nearly two- thirds of unpaid family workers and apprentices are aged 15 to 24. Figure 5. Types of Employment and Education (in %) Self-employed Unpaid family None worker/apprentice Primary Secondary Day laborer Higher Wage earner 0% 20% 40% 60% 80% 100% Source: EPCV 2004. 12 Day laborers, who account for 8.4 percent of total employment, work mainly for private individuals or households (40 percent of them) or private companies (20 percent). Two-thirds of all day laborers received no education, over 20 percent went to primary school, and 10 percent to secondary school. Figure 6. Sectors of Employment Source: EPCV 2004. The labor market is dominated by the agriculture and trade sectors. As shown in Figure 6, over half of the employed population is engaged in activities related to agriculture or trade. The agricultural sector alone accounts for 27 percent of total employment in Mauritania, 34 percent when including livestock activities. Trade, with 24 percent of the employed, is the second biggest sector. Services and the administration represent respectively 10 percent and 8 percent of total employment. The fishing and mining sectors, which provide virtually all of the country's export revenues, together employ only 4 percent of all workers. However, on aggregate, the tertiary sector of the economy absorbs the greatest number of people in Mauritania. Table 5. Share of Wage Earners in the Informal Sector 8 (in %) Gender Living Area Gender by Living Area All Male Female Male Female Rural Urban Rural Urban Rural Urban All 68.6 69.1 66.2 81.8 61.2 82.1 61 79.6 61.7 Poor 80.2 78.7 87.3 86.7 72 86.1 67.9 90.7 85 Non 62.9 64.3 55.9 76.7 57.9 78 59.1 67.6 53.3 Poor Source: EPCV 2004. Figures are displayed in percentages. 8 A wage earner is considered as working in the informal sector if she does not benefit from social security, which here refers to old-age pension rights. 13 The informal sector is sizeable in Mauritania. Wage earners are regarded as working in the informal sector when they do not benefit from social security. According to this definition, two-thirds of Mauritania's wage earners work in the informal sector (Table 5). Unsurprisingly, informality is more widespread in rural areas: 81.8 percent of rural wage earners hold a job in the informal sector against 61.2 percent of urban ones. The gap between the poor and the non poor is wider in urban centers: 72 percent of poor wage earners in urban areas are employed in the informal sector compared to 58 percent for the non poor (these figures stand at 86.7 percent and 76.7 percent for the rural poor and non poor respectively). Women appear less likely than men to work in the informal sector, especially in rural areas. However, a different pattern emerges when the data is disaggregated by poverty status: among the poor, a greater proportion of female wage earners are employed in the informal sector than male workers (respectively 87.3 percent compared to 78.7 percent) whereas the opposite holds for the non poor (55.9 percent for women against 64.3 percent for men). The gap between being poor and being non poor is hence substantial for women: 87.3 percent of poor women work in the informal sector compared to 55.9 percent of non poor women. All in all, poor wage earners are more likely to hold a job in the informal sector than non poor ones (respectively 80.2 percent and 62.9 percent). 14 3. Specific Issues 3.1 Poverty and Labor Market Outcomes Over half of the working-age population lives in a poor household (Table 2). As expected, the non poor fare better on the labor market than the poor. Even though labor force participation of the poor is higher than that of the non poor, the poor display a higher unemployment rate and lower employment rate than the non poor. Figure 7. Main Labor Market Indicators by Education and Poverty Status (in %) Poor Non Poor 17.2 12.6 41.5 41.5 All Poor All Non Poor 48.7 50.1 16.6 8.7 45.2 46.2 None 54.3 None 50.6 24.3 18.1 30.7 35.2 Primary Primary 43 40.6 26.3 20.1 28.5 35.7 Secondary Secondary 44.7 38.7 8.4 9.7 56.9 64.3 Higher* Higher* 62.1 71.2 (in %) (in %) 0 20 40 60 80 0 20 40 60 80 Unemployment Rate Employment-to-Population Ratio Labor Force Participation Source: EPCV 2004. * Only 1 percent of the poor and 4.5 percent of the non poor received some higher education. 3.1.1 Non-participation Overall the poor and the non poor display similar non-participation rates, but differences emerge when educational attainments are accounted for. 9 As shown in Table 6, non-participation rates among the poor and extremely poor stand at 49.9 percent and 49 percent, slightly below that observed for the non poor (51.4 percent). The poor with no education appear more likely to participate in the labor market than the non poor with participation rates standing at respectively 54.3 percent and 50.6 percent (Figure 7). On the contrary, the poor with primary or secondary education participate less in the labor market than the non poor with the same level of education. Youth, seniors, and, to a lesser extent, women living in non poor households appear more likely to opt out of the labor market than those living in poor households. Non-participation is highest for women (72 percent), all the more so when they belong to a non poor household (73.5 percent). Hence, 9 Educational attainments for the poor are as follows: 66.5 percent received no education, 21.7 percent some primary schooling, 10.9 percent some secondary schooling, and 1 percent a higher education. For the non poor, these figures stand respectively at 44 percent, 23.9 percent, 27.6 percent, and 4.5 percent. 15 females from wealthier households seem slightly more likely to opt out of the labor market than their counterparts from poorer families. The same trend is observed for youth: while 67.3 percent of all individuals aged 15 to 24 are out of the labor force, the proportion increases to 71.4 percent for those from non poor families. Finally, 38.4 percent of the poor aged 55 to 64 do not participate in the labor market compared to 48.2 percent of the same age group among the non poor. Table 6. Non-Participation Rate by Poverty Status, Gender, Living Area, and Age Group (in %) All Non Poor Poor Extremely Poor All 50.7 51.4 49.9 49 By gender: Male 26.6 26.5 26.7 26.9 Female 72 73.5 70.1 68.5 By living area: Rural 50.2 52.3 48.5 48.3 Urban 51.4 50.5 53.4 51.4 By age group: 15-24 67.3 71.4 62.6 61.8 25-54 41.2 40.2 42.5 41.7 55-64 44 48.2 38.4 34.6 Source: EPCV 2004. Figures are displayed in percentage. While women's labor market participation seems to be affected by their poverty status, wealth does not appear to influence men's decision to participate in the labor market. As shown by Table 8, male non-participation rate remains relatively constant across all wealth groups amounting to 26.7 percent for the poor and 26.5 percent for the non poor. Although rural non-participation is slightly lower than the urban one for the overall population, when it comes to the non poor, the opposite holds (Table 6): a higher proportion of the rural non poor is out of the labor force than of the urban non poor (respectively 52.3 percent and 50.5 percent). Moreover, in rural areas, the poor appear more likely to participate in the labor market than the non poor. The converse is observed in urban centers. This might suggest that low returns on income-generating activities contribute to poverty in rural areas whereas exclusion from the labor market is more of an issue for the poor in urban centers. Further analysis would however be required to confirm this point. 3.1.2 Unemployment The poor display substantially higher unemployment rates than the non poor. As shown by Table 7, 17.2 percent of the poor and 19.8 percent of the extremely poor are unemployed compared to 12.6 percent for the non poor. This is mirrored by differences within gender groups. Hence, the 16 unemployment rate stands at 25.1 percent for women living in poor households against 19.5 percent for women from non poor families. Yet, the gender gap does not increase with poverty: female unemployment rate amounts to roughly double of that of males whether they belong to poor or non poor households. Differentials between the poor and the non poor narrow to some extent with education, but the poor with some schooling exhibit significant unemployment rates (Figure 7). Unemployment among the poor with no schooling is roughly twice as high as that observed for the non poor (16.6 percent compared to 8.7 percent). The gap between the poor and the non poor with some schooling is somewhat smaller: 24.3 percent of the poor with primary education and 26.3 percent of those with secondary education are unemployed compared to 18.1 percent and 20.1 percent respectively for the non poor. The gap observed in unemployment between rural and urban areas is mainly due to the non poor. Indeed, similar proportions of the poor and extremely poor report being unemployed in rural and urban areas. However, for the non poor, the difference widens: 10.2 percent of the rural non poor are jobless compared to 14.6 percent of the urban non poor. Table 7. Unemployment Rate by Poverty Status, Gender, Living Area, and Age Group (in %) All Non Poor Poor Extremely Poor All 14.7 12.6 17.2 19.8 By gender: Male 11.4 9.8 13.5 15.2 Female 22.3 19.5 25.1 29.4 By living area: Rural 14.2 10.2 17.2 20 Urban 15.3 14.6 17.3 19.4 By age group: 15-24 33.8 30 37.1 42.8 25-54 9.6 9.1 10.3 11.6 55-64 1.9 1.8 2 2.1 Source: EPCV 2004. Figures are displayed in percentage. Finally, correlation between poverty and unemployment is starker for the 15 to 24 age group. As shown in Table 7, while 37.1 percent of the young from poor households and 42.8 percent of those from extremely poor families report being unemployed, this figure stands at 30 percent for the non poor. In contrast, other age groups display similar unemployment rates across wealth groups. 17 3.1.3 Employment The poor and the non poor display similar employment-to-population ratios, but differences emerge when taking into account education levels. As shown in Table 8, 42.5 percent of the non poor are employed compared to 41.5 percent for the poor and 40.9 percent for the extremely poor. Employment-to-population ratios are close for the poor and the non poor with no education, standing at respectively 45.2 percent and 46.2 percent (Figure 7). But for individuals with some schooling the gap widens. Indeed, 30.7 percent of the poor with some primary schooling and 28.5 percent of those secondary schooling are employed, whereas these figures stand respectively at 35.2 percent and 35.7 percent for the non poor. Table 8. Employment-to-Population by Poverty Status, Gender, Living Area, and Age Group (in %) All Non Poor Poor Extremely Poor All 42.1 42.5 41.5 40.9 By gender: Male 65 66.2 63.4 61.9 Female 21.8 21.4 22.4 22.2 By living area: Rural 42.7 42.8 42.7 41.3 Urban 41.2 42.2 38.5 39.1 By age group: 15-24 21.6 20 23.5 21.8 25-54 53.2 54.4 51.6 51.6 55-64 54.9 50.9 60.4 64 Source: EPCV 2004. Figures are displayed in percentage. While employment-to-population does not seem linked to poverty for females, it is lower for poor males as compared to non poor ones (Table 8). Indeed, 66.2 percent of non poor men are employed compared to 63.4 percent for the poor and 61.9 percent for the extremely poor. As men's labor force participation is roughly similar for the poor and the non poor (Table 6), this difference is mainly due to higher unemployment rates among the poor (Table 7). Female's employment-to-population rate appears to be only weakly correlated to poverty: 22.4 percent of poor women are employed against 21.4 percent of the non poor. Yet, poor women tend to participate more in the labor market than the non poor, as they can probably not afford not to work or look for a job (Table 6). Their higher participation rate is counterweighed by the higher unemployment rate they face compared to non poor women (Table 7). 18 Poor prime-age workers display lower employment-to-population ratios than the non poor, whereas youth and seniors living in poor households are more likely to be employed than those from non poor families (Table 8). While 51.6 percent of the poor aged 25 to 54 are employed, this figure stands at 54.4 percent for the non poor. The poor work through older age than the non poor. As shown in Table 8, 60.4 percent of poor individuals aged 55 to 64 are employed compared to 50.9 percent for the non poor, which suggests that the poor cannot afford to stop working as early as the non poor. Results not included here also show that 30.9 percent of the poor aged 65 or more are employed against 21.2 percent for the non poor of the same age group. The non poor exhibit similar employment-to-population ratios in urban and rural areas, whereas the poor are less likely to be employed when living in urban centers (Table 8). Indeed, 42.8 percent of the rural non poor and 42.2 percent of the urban non poor are employed, whereas these figures stand at 42.7 percent and 38.5 percent for the rural and urban poor respectively. This gap appears to be mostly due to the fact that the poor are less likely to participate in the labor market when they live in urban areas as compared to rural ones as unemployment rates are roughly equal for the poor wherever they live. Table 9. Types of Employment and Poverty (in percentages) Non Poor Poor Extremely Poor Wage earners 66.7 33.3 19.3 Day Laborers 60.4 39.6 22.5 Unpaid family labor / apprentices 39.9 60.1 36.9 Self-employed 47.5 52.5 28.4 Source: EPCV 2004. Figures are displayed in percentage. Similarly to what was observed for the overall population, self-employment is the main form of employment for both the poor and the non poor. Indeed, 56.7 percent of workers living in poor households and 48.5 percent of those from non poor households are self-employed (Figure 8). Hence, over half the self-employed are poor and 28.5 percent are extremely poor (Table 9). The vast majority of self-employed people who are poor received no education (85 percent, see Figure 8). The same stands for the non poor self-employed (66 percent never went to school) although they include more individuals with primary and secondary education (respectively 17 percent, and 18 percent), thus suggesting that the latter are somewhat more likely to engage in higher value-added independent activities. 19 Figure 8. Types of Employment and Education by Poverty Status (in %) Self- Non Poor 48.5% Employed 56.7% Poor Unpaid Family Non Poor 6.5% Workers/ None Poor 12.5% Aprentices Primary Non Poor 9.1% Day Secondary Laborers Poor 7.7% Higher Non Poor 36% Wage Earners Poor 23.1% 0% 10% 20% 30% 40% 50% 60% Source: EPCV 2004. The non poor rely more on wage earning than the poor: wage earners represent 23.1 percent of employment among the poor and 36 percent among the non poor (Figure 8). Two-thirds of the wage earners are thus non poor (Table 9). Although over half of the wage earners who are poor have no schooling (55 percent), 24 percent did go to primary school and 17 percent to secondary school (Figure 8). Non poor wage earners generally display, as expected, higher levels of education than the poor: only 27 percent have no education whereas 20 percent have some primary education, 39 percent have some secondary education, and 14.3 percent higher education. As shown in Table B, non poor wage earners work predominantly for the public sector (46.2 percent) and private individuals or households (23 percent). The converse is observed for poor wage earners: 46.6 percent are employed by private individuals or households compared to 26.4 percent by the public sector. Similar proportions of poor and non poor wage earners work for private companies (18.2 percent for both). Unpaid family workers and apprentices are twice as more likely to live in poor households than in non poor ones: they represent 12.5 percent of employed people who are poor against 6.5 percent of those who are non poor (Figure 8). As is the case for the overall population, 70 percent of unpaid workers and apprentices received no education whether they come from poor or non poor households. Finally, a greater proportion of employed individuals living in non poor households relative to those living in poor households report being paid on a daily basis (respectively 9.1 percent and 7.7 percent, Figure 8). The first display higher levels of education than the latter: 56.9 percent of the non poor who are paid by the day received no education, 26.4 percent went to primary school, and 13.9 percent to secondary school whereas these figures stand respectively at 79.5 percent, 16 percent, and 4.2 percent for the non poor. This calls for a deeper investigation to disentangle high value-added activities from low ones. 20 Figure 9. Sectors of Employment by Poverty Status (in %) Poor Non Poor 39% 40% 30% 28% Administration Agriculture 35% 25% Construction 30% Fishing 18% 20% Industry 25% 20% Livestock 20% 15% Mining 11% 10% 10% Services 15% 11% 9% 10% 7% Trade 10% 7% 5% Transport & communication 4% 4% 3% 3% 5% 3% 5% 2% 2% 3% 2% Other 1% 0% 0% Source: EPCV 2004 Employment of the poor appears to be concentrated on a somewhat smaller number of sectors than that of the non poor (Figure 9). The poor are predominantly engaged in agriculture (39 percent) or trade (20 percent), and, to a lesser extent, in services (9 percent) and livestock breeding (7 percent). In contrast, the non poor are employed in trade (28 percent), agriculture (18 percent), the administration (11 percent), services (10 percent) and livestock (7 percent). 3.2 Gender Gaps in the Labor Market As highlighted throughout this paper, gender gaps in the Mauritanian labor market are substantial (Table 2). Differences in levels of education between men and women 10 can only partially explain gender gaps. At the same levels of education, labor market indicators for women are systematically worse than men's (Table 10). Hence, females with no education or primary schooling are around 2.3 times less likely to participate in the labor force than men and three times less likely to be employed than them. These differences are slightly less marked for women with secondary and higher education, but remain high. Furthermore, female unemployment is roughly double that of men, regardless of the level of education. Table 10. Main Labor Market Indicators by Gender and Education (in %) Labor force Employment-to- Unemployment participation* population ratio rate* Male Female Male Female Male Female None 85.5 32.2 76.7 26.4 10.3 18.1 Primary 62.8 22.3 53.4 14.3 15 35.8 Secondary 56.9 25.1 47 16.5 17.3 34.3 Higher 75.9 48.6 69.5 40.8 8.4 16 Source: EPCV 2004. Figures are displayed in percentages. * Standard definition 10 See Table A 21 Women seem to fare better in urban areas than in rural areas (Table 2). Female non-participation rates for instance are higher in rural areas than they are in urban centers (respectively 73.5 percent and 69.8 percent). The opposite is observed for employment-to-population: the female urban employment rate stands at 23.9 percent whereas the rural one stands at 20.3 percent. Unemployment is also lower for women living in urban areas (20.8 percent compared to 23.3 percent for rural areas). Hence, a better understanding of the gender gaps observed in the Mauritanian labor market requires distinguishing rural and urban areas. 3.2.1 Gender gaps in rural labor markets Labor market indicators are somewhat worse for females in rural areas than they are for the overall population. Indeed, labor force participation for women in rural areas stands at 26.5 percent compared to 28 percent for all working-age women (Table 2). The rural employment-to-population ratio is also slightly lower (20.3 percent against 21.8 percent). The reverse is true for males though: both their rural labor force participation and employment rates are approximately three percentage points higher than those exhibited by all working-age men. Prime-age women in rural areas mainly opt out of the labor market for family reasons. As shown in Table 11, rural females across all age groups report family or household obligations as the chief motive for their non-participation in the labor force. Table 12 also provides evidence that women aged 25 to 49 decrease their participation in the labor market when living in a household with children (31.6 percent for those with no children against 27.7 percent for those with children). It should be noted however that the participation rate for females without children remains strikingly lower than that of males, which stands at 92.1 percent for those aged 25 to 49. Hence, gender gaps in participation among the 25-to-49 working-age groups cannot solely be attributed to childcare motives. Table 11. Out of the Labor Force ­ Motives for Women in Rural Areas (in %) Age group 15-24 25-49 50-64 No work 16.4 8.1 1.7 Slack season 0.5 0.3 0 Student 21.8 0.6 0 Family/household obligations 52.8 85.7 60.5 Too young/old 6.3 3.4 35.7 Disabled 1.7 1.2 2.1 Other 0.5 0.7 0 Total 100 100 100 Source: EPCV 2004. Figures are displayed in percentage. 22 The presence of children in the household is positively correlated with younger and older women's participation in the labor market. Family reasons are reported as the chief motive for non- participation among both females aged 15 to 24 and aged 50 to 64, even if the proportion is smaller than that observed for the 30-to-49 age group (Table 11). However, contrary to prime-age women, they seem to increase participation when living with children (Table 12). In rural areas, this is probably explained by the fact that their labor is needed on the farm to compensate for labor losses due to women of childbearing age. Young women's participation in the labor market appears to be affected by the presence of children to a lesser extent though. Other motives, such as schooling, might play a part in their non-participation decision. Indeed, as shown in Table 11, 21.8 percent of out-of- the labor women aged 15 to 24 are in school. Table 12. Rural Labor Force Participation by Gender and Family Status* (in %) Female Male With Without All children children Age group 15-24 52.5 24.6 25 23.3 25-49 92.1 28.5 27.7 31.6 50-64 85.8 24.1 28.2 20 Source: EPCV 2004. Figures are displayed in percentage.* Whether or not there are children under 10 living in the household. Nearly two-thirds of employed women in rural areas are self-employed, a proportion which is similar to that of employed men (Figure 10). However, wage earners only accounts for 9 percent of female employment compared to 22 percent for males. Women are also much more likely than men to serve as unpaid family labor (respectively 26 percent and 10 percent). Figure 10.Types of Employment in Rural Areas by Gender (%) Females Males Self- Self- employed employed 63% 64% Unpaid family Wage earner worker/ 9% apprentice Unpaid family Wage earner 26% worker/appre 22% Day laborer ntice 10% Day laborer 3% 4% Source: EPCV 2004. 23 Agriculture is by far the main sector in which rural women are engaged. Indeed nearly two-thirds of self-employed women work in agriculture-related activities (Table D). But trade is also important among this same group (19.2 percent). Self-employed males are more evenly spread across agriculture, livestock breeding, and trade (respectively 47 percent, 13.7 percent, and 24.2 percent, see Table C). The overwhelming majority of female unpaid labor works in the agricultural sector (86.3 percent). A similar pattern is observed for unpaid male labor, although livestock breeding is more sizeable (71.1 percent in agriculture and 21.4 percent in livestock). Finally, female wage earners are employed in the services sector (28.9 percent), by the administration (25.4 percent), or in other activities. Male wage earners are more likely to be involved in trade (26.8 percent), livestock breeding (16.8 percent). Administrative activities and services come next (14.4 percent and 11.6 percent). 3.2.2 Gender gaps in urban labor markets Labor market indicators for women in urban centers are somewhat better than those for women living in rural areas (Table 2). Both participation in the labor force (30.1 percent compared to 26.5 percent) and employment (23.9 percent compared to 20.3 percent) are higher. Besides, unemployment for women is lower in urban centers than it is in rural areas (respectively 20.8 percent and 23.3 percent). This is at odds with males' unemployment, which is in fact higher in urban areas than in rural ones. Table 13. Out of the Labor Force ­ Motives for Women in Urban Areas (in %) Age group 15-24 25-49 50-64 No work 20.9 11.6 1.7 Slack season 0.1 0.9 0.1 Student 44 2.6 0 Family/household obligations 28.9 80.8 59.2 Too young/old 5.1 2.3 37.1 Disabled 0.5 1.6 1.8 Other 0.4 0.2 0 Total 100 100 100 Source: EPCV 2004. Figures are displayed in percentage. Although overall urban women report family obligations as their main non-participation motive, those aged 15-24 mostly opt out of the labor market to study (Table 13). Indeed, most young women in urban areas are out of the labor force because they are in school (44 percent). Family reasons come second (28.9 percent) and the lack of work third (20.9 percent). Young women's participation seems to be less affected by the presence of children in the household than that of women aged 25 to 49. But contrary to what was observed for rural women, the presence of children is negatively correlated with their likelihood to participate in the labor market. 24 Prime-age women in urban areas with children display lower participation rates than those who have no children (Table 14). Indeed, 41.4 percent of women living in households with no children participate in the labor force, against 37.3 percent of those with children. Although participation rates are higher than those observed in rural areas, the gap imputable to childcare appears to be slightly wider in urban centers. The overall differences in participation between men and women in urban areas cannot however be attributed to childcare only: 89.9 percent of males aged 25 to 49 are in the labor force compared to 41.4 percent for women of the same age without children. Finally, females aged 50 to 64 are more likely to work when living in a household with children than not (respectively 37.5 percent and 27.4 percent). This is probably to compensate for the income loss generated by prime-age women's decision to opt out of the labor force when they have children. Table 14. Urban Labor Force Participation by Gender and Family Status* (in %) Female Male With Without All children children Age group 15-24 35.8 17.8 17 19.5 25-49 89.9 38.4 37.3 41.4 50-64 84.6 32.4 37.5 27.4 Source: EPCV 2004. Figures are displayed in percentage. * Whether or not there are children under 10 living in the household. Half of the women working in urban centers are self-employed, which is substantially higher than the proportion of self-employed men among male urban workers (50 percent compared to 29 percent, see Figure 11). Wage earners amount to 31 percent of female employment (53 percent for males). Unpaid family labor or apprenticeship is comparable in urban centers for men and women (3 percent). It is however much lower than that observed in rural areas. Figure 11. Types of Employment in Urban Areas by Gender (in %) Females Males Self- employed Self- 50% employed 29% Unpaid family worker/ Unpaid family Wage earner apprentice worker/appre 31% Wage earner 3% ntice 53% 3% Day laborer Day laborer 15% 16% Source: EPCV 2004 25 The main sectors in which urban women are employed are trade, the administration, and services, although differences emerge across employment statuses. Indeed, urban self-employed women are mainly engaged in trade (76.2 percent, see Table F). As for urban female wage earners, their main sectors of employment are the administration (36.7 percent), and services (31.2 percent). Over half of all female day laborers work in trade and a quarter of them in services. As mentioned above, urban men are more likely to be wage earners than urban women. They are also engaged in a slightly wider range of sectors (Table E). For instance, although male wage earners ­ like female ones ­ are mainly employed by the administration (29.2 percent) or engaged in trade (13 percent), they also work, even if to a lesser extent, in transport, mining, construction, and fishing activities (respectively 9.6 percent, 6.4 percent, 5.1 percent, and 4.8 percent). As for urban self-employed men, they are mainly engaged in trade and services (42.2 percent and 12.8 percent). 3.3 Youth The labor market participation of youth ­ i.e. all individuals aged 15 to 24 ­ is roughly half that of people aged 30 or more (Table 15). It stands at 25.9 percent for individuals aged 15 to 19 and at 41.6 percent for those aged 20 to 24. Employment-to-population is also low: only 21.6 percent of the youth are employed (this ratio further falls to 16.2 percent for the 15-to-19 population) whereas 56.3 percent of all adults over 30 are employed. Moreover, the unemployment rate of the 15-to-24 age group is five times that of the 30 to 64-year-old population. Finally, gender gaps are substantial: males' labor force participation and employment ratios are double that of females' for instance. Table 15. Main Labor Market Indicators for Youth (in %) All youth Age Group Gender (15 to 24) 15-19 20-24 25-29 30-64 Male Female Labor force participation* 32.7 25.9 41.6 49.9 60.2 45 21.6 Employment-to-population 21.6 16.2 28.7 38.7 56.3 31.3 12.9 ratio Unemployment rate* 33.8 37.4 30.9 22.5 6.4 30.3 40.4 Source: EPCV 2004. Figures are displayed in percentage. * Standard ILO definition Youth's labor force participation and employment-to-population indicators are highest among the poor. Young people aged 15 to 19 and 20 to 24 living in poor households display labor force participation rates approximately 10 percentage points higher than the non poor (Table G). Moreover, 18.3 percent of the poor aged 15 to 19 and 31.7 percent of those aged 20 to 24 are employed as opposed to respectively 14 percent and 26.6 percent for the non poor. The young poor also experience higher unemployment rates than the non poor, especially those aged 15 to 24 whose unemployment rate is nearly 10 percentage points above that of non poor people of the same age. However, the gap 26 between the poor and the non poor grows narrower after the age of 25, especially regarding labor force participation rates which converge to 50 percent for both groups. Low participation and employment rates among youth are only partially due to schooling. Table 16 displays the reasons for which youth are out of the labor force. Although nearly 50 percent of the non- participants aged 15 to 19 are in school, 18.1 percent of them report being out of the labor force because there is no work and 20 percent because of family or household obligations. As far as the 20 to 24 age group is concerned, non-participation is mainly due to family reasons (41.2 percent), as the proportion of students among them only stands at 27.2 percent. Table 16. Youth Out of the Labor Force ­ Motives (in %) 15-19 20-24 25-29 All Male Female All Male Female All Male Female No work 18.1 19.9 16.6 25.9 37.7 20.9 20.9 44.9 16 Slack season 0.6 0.9 0.3 0.7 1.9 0.2 2 7.4 0.9 Student 49.9 60.4 41.5 27.2 45.4 19.4 8.6 27.8 4.7 Family/household 20 5.8 31.4 41.2 7.4 55.7 64 9.5 75.1 obligations Too young/old 9.9 11.6 8.5 2.8 4.4 2.1 1.8 2.3 1.7 Disabled 1.3 1.3 1.3 1.3 2.1 1 2.1 7.2 1 Other 0.2 0.1 0.3 0.8 0.9 0.8 0.6 0.9 0.6 Total 100 100 100 100 100 100 100 100 100 Source: EPCV 2004.Total might not always exactly amount to 100 due to rounding. Youth's participation in the labor market decreases with education. As illustrated by Figure 12 youth with no education exhibit a 47.9 percent rate of labor force participation and a 31.6 percent employment rate, whereas these figures stand respectively at 31.4 percent and 22.5 percent for young people with primary school education. This is partly due to the fact that 46.7 percent of the 15 to 24 age group is in school. However, labor force participation among unschooled people of the same age stands at 48 percent only, in other words 52 percent of the youth who are not in school are not in the labor force either (results not shown here). Finally, for those who do participate in the labor market, unemployment rates remain high, regardless of their level of education. They hover between 30 and 35 percent for all groups and peak at 49.7 percent for people with secondary schooling. Once employed, young people are either self-employed or work as unpaid family labor or apprentices. Figure 13 shows that among those aged 15 to 24, 33 percent are self-employed and 31 percent are unpaid family labor or apprentices. Compared to their elders, young workers are more likely to serve as unpaid labor or apprentices (31 percent compared to only 4 percent for the 25-to-64 employed age group) and fewer of them are self-employed (33 percent compared to 56 percent for the 27 25 to 64). However, the proportion of wage earners among young employed people is very close to that of adults (respectively 28 percent and 31 percent). Figure 12. Main Labor Market Indicators by Education Level for Youth (in %) 20.8 Labor Force 17.9 31.4 47.9 Participation 13.7 Employment-to- 9 Higher 22.5 Population 31.6 Secondary Primary 34.1 49.7 None Unemployment 28.4 34 0 10 20 30 40 50 (in %) Source: EPCV 2004. Types of employment vary across living areas. Self-employment among youth is indeed more widespread in urban centers than it is in rural areas (Figure 13). The contrast is even more striking for unpaid family labor and apprenticeship: almost 40 percent of all young men residing in rural areas against 14.5 percent for those in urban areas. For young women, these figures stand at 43.6 percent and 11.4 percent, respectively. Hence, young people employed in urban areas are mostly wage earners whereas those employed in rural areas are mainly unpaid workers or self-employed. Youth's participation in the labor market features substantial gender gaps, as does the overall population. As evidenced by Table 17, young women's labor force participation and employment rates are less than half of that of men of the same age. Although 41.5 percent of women aged 15 to 19 are not taking part in the labor force because they are in school, 31.4 percent of them report family reasons as a main motive for non-participation (compared to only 5.8 percent of males of the same age, Table 16). The differences are even more striking when considering the 20-to-24 age group: 55.7 percent of these women are out of the labor force for family reasons, whereas only 19.4 percent of them are in school (these figures stand respectively at 7.4 percent and 45.4 percent for men aged 20 to 24). Furthermore, females engaged in the labor market experience an unemployment rate of 40.4 percent, i.e. 10 percentage points above that displayed by males. 28 Figure 13. Types of Employment for Youth (aged 15 to 24) and Adults (aged 25 to 64) Unpaid family 15-24 25-64 w orker/ Unpaid apprentic family Self- e Self- worker/ employed employed apprentice 56% 31% 33% 4% Day laborer 9% Day laborer Wage Wage 8% earner earner 28% 31% Source: EPCV 2004. Differences between those living in urban and rural areas are striking for young women who are employed (Table 17). In rural areas, 43.6 percent of them work as unpaid family labor or apprentices, compared to only 11.4 percent in urban centers. Self-employment is also less common for urban young women than it is for rural ones (respectively 45.3 percent and 28.3 percent). They are indeed more likely to be wage earners (45.7 percent). Table 17. Types of Employment for Youth by Gender and Living Area (in percentages) Male Female Rural Urban Rural Urban Wage earner 22.4 49.5 8.6 45.7 Day laborer 3.7 16.8 2.5 14.6 Unpaid family worker / apprentice 38.8 14.5 43.6 11.4 Self- employed 35.1 19.1 45.3 28.3 Total 100 100 100 100 Source: EPCV 2004. Total might not always exactly amount to 100 due to rounding. 3.4 Earnings Determination 11 This section presents an analysis of earnings of the self-employed and wage-earners. Unfortunately, as the EPCV 2004 data does not include information on the number of hours worked, this section does not investigate underemployment issues. 11 Data on individual earnings in the EPCV 2004 survey are provided either by day, week, or month. However, given inconsistencies observed in the orders of magnitude between these three categories, this section is based on monthly earnings only. The present analysis on earnings thus includes 78 percent of the self-employed and 93 percent of all wage earners. Finally, earnings were not adjusted to take into consideration regional differences in standards of living. 29 3.4.1 Earnings of the self-employed 12 Over half of the working population is self-employed in Mauritania (Section 2.3.). Earnings from self- employment vary across living areas (Table 18). Mean earnings of the self-employed are higher in urban centers than in rural regions (37,800 MRO compared to 26,800 MRO). Median monthly earnings of all the self-employed, however, stand at 20,000 MRO, very similar to those observed in rural and urban areas. Mean earnings of self-employed people engaged in agriculture or livestock in rural areas amount to 22,700 MRO (median at 15,000 MRO), whereas mean earnings of those engaged in trade or services in urban areas amount to 38,600 MRO (median at 25,000 MRO). 13 Earnings of the self-employed in Nouakchott are approximately 50 percent higher than those of the overall self-employed population. 14 Table 18. Self-Employed ­ Monthly Earnings by Living Area and Gender (in thousands of MRO) All Rural Urban Male Female Mean Median Mean Median Mean Median Mean Median Mean Median Monthly 29.2 20 26.8 20 37.8 21 35.1 21 14.6 10 Earnings Source: EPCV 2004. Unsurprisingly, earnings from self-employment are lower among the poor. Indeed, the poor earn on average 40 percent less than the non poor engaged in self-employment (Table H). The gap is wider in urban areas where earnings of the self-employed non poor amount to nearly the double of that of the poor (in rural areas, the differential is of approximately 40 percent). As evidenced by Table H, earnings of the extremely poor do not differ substantially from those observed for the poor. Investigating this point would require deeper analysis which is beyond the scope of this paper. At least one element could account for this result: extremely poor households exhibit higher dependency ratios than the poor. 15 Hence, extreme poverty could result from the fact that earnings are not sufficient to cover all household members' needs.16 12 Earnings reported by the self-employed should be interpreted with caution. It is indeed unclear whether respondents reported their actual earnings or the sales figures of their businesses. Moreover, as the EPCV 2004 does not include information on the number of hours worked, we were unable to adjust these earnings for unpaid family labor. Individual earnings for the self-employed might thus be somewhat overestimated. 13 Results not included here. 14 Mean earnings of the self-employed in Nouakchott amount to 51,000 ouguiyas, with a median at 30,000 ouguiyas. 15 Dependency ratios are computed here as the number of household members below 15 or above 64 for each working-age member. The average dependency ratio for the poor is equal to 1.33 compared to 1.43 for the extremely poor. For households with at least one member engaged in self-employment, the average dependency ratios stand at 1.34 for the poor and 1.45 for the extremely poor. 16 We also checked whether extremely poor households relied more on unpaid family labor than poor households. But poor households in which at least one member is engaged in self-employment exhibit on average 0.215 unpaid family worker compared to 0.216 for extremely poor households. 30 Table 19. Self-Employed ­ Monthly Earnings by Gender and Education (in thousands of MRO) All Male Female Mean Median Mean Median Mean Median All 29.2 20 35.1 21 14.6 10 (4,403)* (3,141)* (1,262)* None 22.3 17 27.5 20 13.3 10 (2,330) (1,474) (856) Primary 31.1 20 36.4 20 14.6 10 (526) (391) (135) Secondary 53.2 30 56.3 35 35.1 20 (314) (263) (51) Higher 114.1 50 116.4 50 4.9 7 (45) (43) (2) Source: EPCV 2004. Numbers in parenthesis correspond to unweighed observations.* The number of observations is greater than the sum of observations over all education levels, due to missing data on education. The earnings of the self-employed feature substantial gender gaps. Indeed, as evidenced by Table 19, male earnings are approximately 2.5 times higher than females'. Most self-employed women are engaged in agriculture and trade. Their earnings in the former sector are nearly half what they are in the latter: 11,300 MRO on average for agriculture compared to 19,600 MRO for trade (medians are respectively 8,000 MRO and 15,000 MRO). These figures stand at 22,600 MRO and 44,000 MRO for males, i.e. double that of females (medians at 20,000 MRO 30,000 MRO). 17 The gender gap is also highest among the non poor: non poor men earn approximately 60 percent more than non poor women whereas for the poor the differential comes to 50 percent (Table H) Besides, at the same level of education, women's earnings in self-employment are still lower than men's (Table 21). Females with no education earn approximately half of what males with no education earn. Women with primary schooling fare even worse when compared to men, as their average earnings only amount to 40 percent of men's with primary education. 3.4.2 Wage earners Wage earning is the second form of employment in Mauritania (Section 2.3). Mean wages in the public sector are lower than those observed in the private sector (Table 20). Average wages amount to 34,400 MRO in the public sector compared to 39,200 MRO in the private sector (with medians respectively at 28,000 MRO and 30,000 MRO). On average men earn more in the private than in the public sector, but in both sectors, median wages stand at 30,000 MRO. This is confirmed by the results obtained in the wage regression (see Table 22), as the coefficient for the private sector variable is significant and positive for males. Female wage workers are predominantly employed in the public sector and by individual households and are almost absent from the private sector despite the 17 Results not shown here. Comparisons between earnings by gender for other sectors are rendered impossible due to a too small number of observations. 31 relatively higher wages offered there. As shown in the regression in Table 22, women working for individual households earn significantly less than those employed in the public sector. Table 20. Wage Earners ­ Monthly Earnings, by Living Area, Gender, and Type of Employer (in thousands of MRO) All Rural Urban Male Female Mean Median Mean Median Mean Median Mean Median Mean Median All 30 23 25.3 20 32.7 25 32.2 25 19.5 15 (3,531)* (1,141)* (2,390)* (2,904)* (627)* Public 34.4 28 30.5 25 35.4 30 36.4 30 26.2 23 (1,567) (302) (1,265 (1,265) (302) Private 39.2 30 35.5 26 40.8 30 40.7 30 25.1 30 (556) (130) (426) (505) (51) Individual/ 20.1 15 20.4 15 19.8 15 22.4 20 10.1 8 Household (1,137) (581) (556) (901) (236) Other 30.8 20 25.5 20 36.4 20 33 20 16.1 10 (243) (114) (129) (212) (31) Source: EPCV 2004. Numbers in parenthesis correspond to unweighed numbers of observations. * The number of observations is greater than the sum of observations over all employer types. This is due to missing data on employers. Wages appear to be higher in urban centers than in rural areas, although the effect is weak. Average urban wages are approximately 30 percent higher than rural ones (see unadjusted wages in Table 20). Wages in the public sector are also greater in urban centers than in rural regions. The same is true for the private sector. But, wage earners employed by an individual or a household (e.g. maids) are offered similar wages regardless of whether they live in urban or rural areas. The coefficient of the rural variable included in the male and female wage regressions in Table 22 is not significant. This indicates that once other characteristics are controlled for, such as education or the sector of activity, the mere fact of living in a rural region has no impact on wage levels. Table 21. Wage Earnings ­ by Gender and Education (in thousands of MRO) All Male Female Mean Median Mean Median Mean Median All 30 23 32.2 25 19.5 15 (3,531)* (2,904)* (627)* None 21.5 18 24.1 20 10.8 9 (946) (740) (206) Primary 25.1 20 25.4 20 23.5 18 (620) (517) (103) Secondary 32.9 28 35.4 30 23.4 23 (972) (758) (214) Higher 56.1 39 60.2 40 33.9a 27a (380) (323) (57) Source: EPCV 2004. Numbers in parenthesis correspond to unweighed observations. * The number of observations is greater than the sum of observations over all education levels. This is due to missing data on education. 32 Observed gender gaps among wage earners are significant. Overall, women earn on average 40 percent less than men (see unadjusted wages in Table 20). The unadjusted wage differential is slightly lower in the public sector, where women nevertheless earn nearly 30 percent less than men. When women are employed by individual households, the gender gap seems to be much wider as their wages just amount to 60 percent of men's. At the same level of education, female earnings tend to remain lower than males' (see unadjusted wages in Table 21). This is especially true for women with no education whose earnings amount to less than 50 percent of men's. Figure 14. Adjusted Wage Differentials by Level of Education and Gender Male Higher over Secondary Secondary over Primary Primary over None 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 in % Female Higher over Secondary Secondary over Primary Primary over None 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 in % Source: EPCV 2004. Wages increase significantly with education. Indeed, as illustrated by Table 22, most coefficients on the variables for education in the wage regressions are significant and positive. Figure 14 shows adjusted wage differentials by level of education and gender. Controlling for other characteristics, male wage workers with primary schooling earn 5 percent more than those with no education. 18 As 18 The coefficient is only significant at a 15 percent level though. 33 for male wage workers with secondary schooling, they earn approximately 20 percent more than those with primary education. Finally, a male worker with higher education earns about 37 percent more than what he would have earned if he had only been to secondary school. Overall, a similar trend is observed for women, although the orders of magnitude differ and the impact of secondary schooling appears to be weak. Female workers with primary education earn approximately 15 percent more than those having received no schooling. But those with secondary education only earn 3 percent more than those having attended to primary school. The impact of higher education on earnings is similar to that observed for men: female wage earners with higher education earn 34 percent more than those with secondary education. Figure 15. Wage Earners - Kernel Density in the Formal and Informal Sectors 19 (in thousands of MRO) .04 .03 Density .02 .01 0 4.312 21 50 100 Monthly Wages Formal Sector Informal Sector Source: EPCV 2004 The 2004 minimum wage, which stood at 4,312 MRO, was too low to be relevant. Figure 15 displays the kernel density distribution of monthly wages for wage earners. It distinguishes the formal sector, where minimum wage regulations are normally applicable, from the informal sector, where they are not. It should be noted here that 70 percent of all wage workers are employed in the informal sector (Table 5). As illustrated by Figure 15, wages in the informal sector are notably lower than that in the formal sector. Table 22 also shows that the coefficients on the informal variable is significant and negative in the female and male wage regressions, thus providing evidence that wages in the informal sector are lower than those in the formal sector. According to the wage distribution for formal sector employees exhibited in Figure 15, few people report earnings close to the 2004 minimum wage. Indeed, less than 1 percent of all formal wage earners were paid below 4,312 MRO in 2004. Even in 19 A wage earner is considered as working in the informal sector if she does not benefit from social security. 34 the informal sector, the proportion is low (1.6 percent). The 2004 minimum wage was hence all but a binding constraint in the Mauritanian labor market. It was increased in January 2005 (after the EPCV 2004 data was collected) to 21,000 MRO. Although this new minimum wage was not applicable at the time of the survey, it is interesting to note that approximately 30 percent of the employees from the formal sector received wages below this threshold. These calculations should be carried out again once new household data becomes available. Table 22. Wage Earners - Econometric Estimates of Wage Determinants Regression of the log of wages on individual characteristics Male Female Standard Standard Coefficient Significance Coefficient Significance Error Error Education None (omitted) Primary 0.052 ns 0.036 0.143 * 0.085 Secondary 0.244 *** 0.034 0.171 ** 0.089 Higher 0.615 *** 0.045 0.508 *** 0.116 Age 15 to 24 (omitted) 25 to 49 0.287 *** 0.035 0.132 *** 0.067 50 to 64 0.414 *** 0.045 0.042 ns 0.111 Region Nouakchott 0.154 *** 0.039 0.348 *** 0.076 Dakhlet Nouadhibou 0.029 ns 0.069 0.194 ns 0.147 Senegal River Valley 0.027 ns 0.033 0.105 ns 0.073 Other (omitted) Rural 0.053 ns 0.035 -0.098 ns 0.073 Sector Agriculture (omitted) Livestock -0.258 *** 0.092 -0.515 * 0.301 Fishing 0.370 *** 0.100 0.242 ns 0.275 Mining 0.439 *** 0.105 0.125 ns 0.341 Industry 0.211 * 0.130 0.625 * 0.361 Construction 0.145 ns 0.099 0.536 * 0.300 Transport & Communication 0.023 ns 0.092 0.333 ns 0.301 Trade 0.187 ** 0.086 0.232 ns 0.213 Services -0.018 ns 0.088 -0.206 ns 0.208 Administration -0.007 ns 0.091 -0.281 ns 0.216 Other 0.044 ns 0.088 -0.127 ns 0.210 Type of employer Public (omitted) Private 0.233 *** 0.043 -0.089 ns 0.117 Individual household -0.072 ns 0.046 -0.685 *** 0.111 Other 0.046 ns 0.062 -0.609 *** 0.172 Informal -0.174 *** 0.037 0.081 ** 0.081 Constant 2.747 *** 0.102 0.241 *** 0.241 Observations 2,274 Observations 564 R-Squared 0.306 R-Squared 0.504 Source: EPCV 2004.OLS regression. ns: not significant, ***: significant at the 1 percent level, **: significant at the 5 percent level, *: significant at the 10 percent level. 35 Conclusions This paper represents the first attempt to analyze the Mauritanian labor market using data from the latest national household survey. On the basis of the analysis conducted in this paper, the following salient conclusions emerge. First of all, the Mauritanian labor market is characterized by lower participation rates, lower employment-to-population rates and relatively higher unemployment rates than the corresponding values in neighboring countries. Over half of the working-age population lives in a poor household. The non poor fare better on the labor market than the poor. Even though the labor force participation of the poor is higher than that of the non poor, the poor display a higher unemployment rate and lower employment rate than the non poor. The data also suggest a negative correlation between wage employment and poverty, that is, wage employment increases with wealth. Substantial differences emerge between gender labor market indicators. Women systematically earn less than men, independently of their sector and type of employment, and controlling for other factors (e.g. education). The youth experience considerable difficulties in entering the labor market. The employment-to-population ratio for the 15 to 24-year-old group is roughly half the overall employment rate for Mauritania. If Mauritania is to attain the ambitious poverty-related goals set out in the PRSPs, policy makers need to focus more squarely on labor market policies. The analysis conducted shows that the Mauritanian labor market is not vibrant. As evidenced by the PRSP-2, the country has taken a passive approach to employment generation and promotion to date, regarding it as a mere bi-product of policies aimed at fostering growth. Moreover, there is the need to ensure that labor market data are collected and analyzed on a regular basis, through a sound mix of quantitative and qualitative techniques, which is currently not the case. As mentioned, gender disparities remain important for similar levels of education. Therefore, policy makers need to pay attention to the binding constraints outside the education sector, which deserve further investigation, including cultural factors that may prevent women to enter the labor market. Concerning the poor labor market indicators for young adults, ensuring ­ within a reasonable timeframe ­ that the current generation has access to jobs should undoubtedly be a key priority for the authorities. Better informed labor market policies should be a prominent feature of future poverty reduction strategies. 36 References Auty, R., and N. Pontara, (2008), "A Dual-track Strategy for Managing Mauritania's Projected Oil Rent", Development Policy Review, Vol. 26, No.1, pp. 59-77. Magnoli-Bocchi, A., Pontara, N., Tejada, C., Fall, K., and P. Gallego-Cuervo, (2008), "Reaching the MDGs: Mauritania Should Care", World Bank Policy Research Working Paper 4674, Washington DC: The World Bank. GIRM, (Government of the Islamic Republic of Mauritania), (2000), National Census, Nouakchott: National Institute of Statistics. ______ (2006a), Poverty Profile, Nouakchott: National Institute of Statistics. ______ (2006), Poverty Reduction Strategy Paper II, Nouakchott: Mauritania. IMF, (International Monetary Fund), (2007), Mauritania: Article IV, Washington DC: IMF. Kapsos, S., (2007), "World and regional trends in labor force participation: Methodologies and key results", Economic and Labor Market Paper, International Labor Office. Pontara, N., (2007), A Preliminary Analysis of the Mauritanian Labor Market, mimeo. Rosanvallon, A. (2004), "Rapport de Mission, Bureau International du Travail, Programme d'Appui a la Mise en Oeuvre du CLPS-1", mimeo, Nouakchott: Mauritania. ______ (2006), "Situation et Bilan de L'Emploi 2000-2005, Bureau International du Travail, Programme d'Appui a la Mise en Oeuvre du CLPS-2", mimeo, Nouakchott: Mauritania. Sender, J., C. Cramer, and C. Oya, (2005), "Unequal Prospects: Disparities in the Quantity and Quality of Labor Supply in Sub-Saharan Africa", World Bank Social Protection Discussion Paper 525, Washington DC: The World Bank. UNDP, (2007), Human Development Report, New York: USA. World Bank, (2004), World Development Indicators, Washington DC: The World Bank. World Bank, (2005), World Development Indicators, Washington DC: The World Bank. World Bank, (2007), "Sénégal, à la Recherche de l'Emploi: Le Chemin vers la Prospérité", Country Economic Memorandum, Report No. 36750-SN, Washington DC: The World Bank. 37 Annex 1. Labor Market Indicators: Definitions Reference Period The last seven days prior to the survey. Working-Age Population All individuals between 15 and 64 years of age are classified as belonging to the working-age population. Labor Force Participation Share of the working-age population that is either employed or unemployed (ILO definition). Labor Force Participation ­ Relaxed Definition Share of the working-age population that is either employed or unemployed (relaxed definition). Employment-to-Population Ratio Share of the working-age population that is employed. An individual is classified as being employed if (i) the individual worked or (ii) had a job or enterprise but was temporarily absent from work during the reference period. Unemployment Rate ­ ILO Definition Share of the labor force that is unemployed. An individual is unemployed if (i) the individual did not work during the reference period and (ii) was looking for a job and ready to work during the reference period. Unemployment Rate ­ Relaxed Definition The relaxed definition for unemployment includes all the unemployed according to the ILO definition as well as all individuals who did not work in the reference week and did not look for a job because they thought there were no jobs available. Non-Participant A working-age individual is classified as a non-participant (or out of the labor force) if the individual was neither employed nor unemployed during the reference period. 38 Annex 2. Additional Tables Annex Table A Education Level by Gender and Age Group (in percentages) None Primary Secondary Higher Total All working-age individuals 54.2 22.9 20 2.8 100 Male 45.8 24.2 25.1 5 100 Female 61.2 21.8 15.8 1.1 100 By age group: 15-24 34.8 36.5 27.4 1.3 100 25-34 50.3 23.5 21.7 4.5 100 35-44 64.8 11.3 16.3 4.7 100 45-54 81.3 6.6 8.8 3.2 100 55-64 91 4.2 3.3 1.5 100 Source: EPCV 2004. Total might not always exactly amount to 100 due to rounding. For working-age individuals only. Annex Table B. Types of Employment and Employer by Poverty Status (in percentages) Self- Non Poor 48.5% Employed Poor 56.7% Unpaid Family Non Poor 6.5% Workers/ Public sector Aprentices Poor 12.5% Private company Day Non Poor 9.1% Individual/household Laborers 7.7% Poor Other Wage Non Poor 36% Earners Poor 23.1% 0% 10% 20% 30% 40% 50% 60% Source: EPCV 2004 Annex Table C. Sector of Employment ­ Males in Rural Areas (in percentages) Unpaid family Day Wage earners labor/ Self-employed laborers apprentices Agriculture 4.18 8.5 71.1 47 Livestock 16.8 3 21.4 13.7 Fishing 3.7 2.2 0.7 1.1 Mining 0.6 0 0.1 0.1 Industry 0.8 2.6 0.2 0.9 Construction 3.7 28.6 0 2.15 Transport 5.4 3.3 2.3 1.7 Trade 26.8 9.6 2 24.2 Services 11.6 12.5 0.5 3.2 Administration 14.4 0.3 0 0.2 Other 11.3 29.3 1.6 5.6 Total 100 100 100 100 Source: EPCV 2004. Figures are displayed in percentages. Total might not always exactly amount to 100 due to rounding. 39 Annex Table D. Sector of Employment ­ Females in Rural Areas (in percentages) Unpaid family Day Wage earners labor/ Self-employed laborers apprentices Agriculture 5.1 17.2 86.3 64.9 Livestock 0 0 2.1 0.6 Fishing 0 0.6 0 0 Mining 0.1 0 0.3 0.1 Industry 0 1.8 1 5 Construction 1.2 0 0 0 Transport 0.8 0 0 1.2 Trade 10.8 34 1.7 19.2 Services 28.9 33.2 2.4 4.1 Administration 25.4 0 0.7 0.1 Other 27.7 13.1 5.5 4.7 Total 100 100 100 100 Source: EPCV 2004. Figures are displayed in percentages. Total might not always exactly amount to 100 due to rounding. Annex Table E. Sector of Employment ­ Males in Urban Areas (in percentages) Unpaid family Day Wage earners labor/ Self-employed laborers apprentices Agriculture 1 1.3 25.6 5.7 Livestock 0.7 0.2 1.4 1.6 Fishing 4.8 9.3 2.2 6.2 Mining 6.4 0.1 0 0 Industry 2 1.5 4.9 5 Construction 5.1 18.3 3.1 6.5 Transport 9.6 6.1 7.9 5.4 Trade 13 16.3 14.5 42.2 Services 14.8 20.1 15.6 12.8 Administration 29.2 0.7 0.3 0.6 Other 13.4 26.1 24.5 14 Total 100 100 100 100 Source: EPCV 2004. Figures are displayed in percentages. Total might not always exactly amount to 100 due to rounding. Annex Table F. Sector of Employment ­ Females in Urban Areas (in percentages) Unpaid family Day Wage earners labor/ Self-employed laborers apprentices Agriculture 0.5 0 22.2 1.9 Livestock 1.4 0 0 0.1 Fishing 2.2 1 0.4 1 Mining 1 0.5 0 0 Industry 3.4 2.1 0 5.4 Construction 0.9 0.4 0 0 Transport 1.5 0.8 1.7 1 Trade 6.5 55.6 12.4 76.2 Services 31.2 26.2 15.7 5.9 Administration 36.7 0.7 6.7 0.2 Other 14.6 12.6 41 8.2 Total 100 100 100 100 Source: EPCV 2004. Figures are displayed in percentages. Total might not always exactly amount to 100 due to rounding. 40 Annex Table G. Main Labor Market Indicators ­ Youth and Poverty (in percentages) 15-19 20-24 25-29 Poor Non Poor Poor Non Poor Poor Non Poor Labor force participation (standard) 31 20.5 47.3 37.4 50.1 49.8 Employment-to-population ratio 18.3 14 31.7 26.6 37.8 39.2 Unemployment rate (standard) 41 31.7 33.1 28.9 24.5 21.3 Source: EPCV 2004. Figures are displayed in percentages. Annex Table H. Self-Employed ­ Monthly Earnings and Poverty (in thousands of MRO) All Non Poor Poor Extremely Poor Mean Median Mean Median Mean Median Mean Median 29.2 20 36.9 20 20.9 16 20.6 15 Rural 26.8 20 33.7 20 20.8 16 20.7 15 Urban 37.8 21 44.6 30 21.8 18 19.8 15 Male 35.1 21 44.5 30 24.8 20 24.6 20 Female 14.6 10 17.4 12 11.9 10 11.4 9 Source: EPCV 2004 Annex Table I. Incidence of Poverty by Region and Relative Contributions (2000-2004) 2000 2004 Overall poverty incidence (P0) (%) National 51 47 Rural 66 59 Urban 28 29 By region Rural­river 77 66 Rural­other 60 57 Urban­Nouakchott 29 26 Urban­other 27 33 Overall contribution to poverty (C0) (%) Rural 80 75 Urban 20 25 By region Rural­­river 35 17 Rural­­other 45 58 Urban­­Nouakchott 12 13 Urban­­other 8 12 Inequality (Gini coefficient) National 0.39 0.39 Rural 0.37 0.35 Urban 0.35 0.39 Source : National Statistics Office (ONS), "Enquête Permanente sur les Conditions de Vie des Ménages" (EPCV) (Nouakchott, 2000 and 2004) wb208439 C:\Documents and Settings\wb208439\My Documents\Nicola\Papers\Lab Market MRT\Fine\MRT -- Labour Market Paper -- Final Version -- Changes Accepted -- June 03 2009.doc 03/06/2009 15:05:00