AN ASSESSMENT OF EMPLOYMENT IN NIGER June 2017 Trading cowpeas on the weekly market Kargi Bangou, Dosso region © Thomas Bossuroy ACKNOWLEDGEMENTS The Jobs Assessment team was led by Thomas Bossuroy. Lead authors and contributors for each chapter are as follows: • Mauro Testaverde led Chapter 1 on Employment Structures in Niger. The re-analysis of results based on the last round of ECVMA was undertaken by Benedikte Bjerge. • Benedikte Bjerge and Jan von der Goltz led Chapter 2 on Employment Dynamics and Transitions. • Jonathan Kaminski and Thomas Bossuroy led Chapter 3 on Agricultural Employment, with contributions from Chand Mazumdar. • Bienvenue Tien and Dimitris Mavridis led Chapter 4 on Non-Agricultural Household Enterprises. • Thomas Bossuroy led Chapter 5 on Youth Occupational Aspirations, with contributions from Erwin Tiongson, Victoria Morse and Benedikte Bjerge. The team acknowledges the support of the National Statistical Institute of Niger, their willingness to integrate an original module on aspirations and psychology of the youth, and their critical role in collecting and preparing the data used in this report. Diane Steele (LSMS team, World Bank) provided critical support with the creation and implementation of a dedicated module on aspirations and psychology added to the last survey round of ECVMA, on which Chapter 5 is based. Michael Wild (DECRG, World Bank) supported the team with the creation of statistical weights for the cross-sections and the panel datasets. Richard Moussa (consultant) assisted the INS team with the data cleaning and curating of several modules of ECVMA 2014. CONTENTS ACKNOWLEDGEMENTS .......................................................................................................................... 3 CONTENTS.................................................................................................................................................. 4 Chapter 1 : EMPLOYMENT STRUCTURES IN NIGER ........................................................................... 1 Country context............................................................................................................................................. 2 Characteristics of the workforce ................................................................................................................... 6 Employment structures ............................................................................................................................... 10 Characteristics of the population by employment status............................................................................. 12 Characteristics of jobs by sector ................................................................................................................. 17 Transitions: school-to-work and occupational mobility ............................................................................. 22 Gender-specific constraints to participation................................................................................................ 25 Summary of main findings.......................................................................................................................... 27 Chapter 2 EMPLOYMENT DYNAMICS AND TRANSITIONS ............................................................. 29 Introduction ................................................................................................................................................. 30 Common transitions between employment statuses and sectors................................................................. 32 The dynamics of diversification .................................................................................................................. 38 Intergenerational rigidity restricts employment dynamics. ......................................................................... 42 References ................................................................................................................................................... 44 Chapter 3 AGRICULTURAL EMPLOYMENT ........................................................................................ 45 A nation of farmers ..................................................................................................................................... 46 Characteristics of Agricultural employment ............................................................................................... 49 Constraints to Agricultural employment ..................................................................................................... 53 Linkage between agricultural and non-agricultural activities ..................................................................... 58 What drives employment outcomes in agriculture? .................................................................................... 60 Chapter 4 : NON-AGRICULTURAL HOUSEHOLD ENTERPRISES .................................................... 65 Introduction ................................................................................................................................................. 65 Female-owned and rural HEs have much lower productivity levels .......................................................... 70 Constraints and drivers of productivity of Household Enterprises ............................................................. 72 Policies and Programs for the HE sector .................................................................................................... 74 References ................................................................................................................................................... 76 Appendix ..................................................................................................................................................... 78 Chapter 5 : YOUTH OCCUPATIONAL ASPIRATIONS ......................................................................... 83 Introduction ................................................................................................................................................. 84 Data and analytical method ......................................................................................................................... 85 Occupational aspirations ............................................................................................................................. 88 Determinants of aspirations ........................................................................................................................ 94 Internal constraints to fulfillment.............................................................................................................. 100 External constraints to fulfillment............................................................................................................. 102 References ................................................................................................................................................. 108 Boxes Box 2.1: Characteristics of the balanced panel ........................................................................................... 31 Box 3.1: Niger Livelihood Zone Map ........................................................................................................... 47 Tables Table 1.1: Population in Niger (%) ................................................................................................................ 8 Table 1.2: Labor Market Indicators for Different Groups, Population 15 to 64 ......................................... 11 Table 1.3: Employment Sectors by Selected Characteristics ...................................................................... 17 Table 1.4: Access to Different Employment Sectors by Selected Groups................................................... 18 Table 2.1: Mean consumption expenditure by economic activity ............................................................. 32 Table 2.2: Characteristics of the agricultural and non-agricultural sector, 2011-14 .................................. 32 Table 2.3: Labor market activity and transition matrixes, 2011-14 ............................................................ 34 Table 2.4: Economic activities and activity transition matrix, 2011-14 ...................................................... 36 Table 2.5: Primary economic activity and transition matrix by sector, 2011-14 ........................................ 37 Table 2.6: Impact of diversification out of agriculture on welfare ............................................................. 38 Table 2.7: Share of individuals shifting to exclusive agricultural work, by categories ............................... 39 Table 2.8: Determinants of the transition in and out of agriculture .......................................................... 40 Table 2.9: Absence from residence............................................................................................................. 41 Table 2.10: Characteristics of working migrants......................................................................................... 42 Table 2.11: Intergenerational reproduction of occupations: farm/non-farm odds-ratios ......................... 43 Table 3.1: Village level infrastructure in rural areas .................................................................................. 57 Table 3.2: Correlates of individual and household-level diversification of activities for agricultural workers and households ............................................................................................................................. 59 Table 3.3: Determinants of welfare income across agricultural workers .................................................. 61 Table 3.4: Crops and agricultural welfare ................................................................................................ 63 Table 3.5: Infrastructure and agricultural welfare .................................................................................. 63 Table 5.1: Demographics of Youth Sub-Sample (%).................................................................................... 86 Table 5.2: Medium-term aspirations by employment status (%) .............................................................. 92 Table 5.3: Determinants of no aspirations ................................................................................................. 96 Table 5.4: Internal constraints to fulfillment: Locus of control and self-esteem ..................................... 101 Table 5.5: Intergenerational reproduction of occupations: farm/non-farm odds-ratios ... Error! Bookmark not defined. Table 5.6: Regression analysis of transition from agriculture by gender ...... Error! Bookmark not defined. Figures Figure 1.1 GDP and Population Growth ...................................................................................................... 3 Figure 1.2: Population Pyramids ................................................................................................................... 6 Figure 1.3: Working Age Population by Education Level (2014)................................................................... 9 Figure 1.4: Distribution of education by gender, age and urban/rural divide among the working age population (2014) ......................................................................................................................................... 9 Figure 1.5: Distribution of Population in Niger ........................................................................................... 10 Figure 1.6: Inactivity Rates for different population groups ...................................................................... 13 Figure 1.7: Reasons for not looking for a job by gender - working age population ................................... 14 Figure 1.8: Labor Market Status of Population Not In Working Age .......................................................... 15 Figure 1.9: Distribution of Hours Worked per Week in Primary Jobs ......................................................... 16 Figure 1.10: Economic Sector of Primary and Secondary Jobs ................................................................... 17 Figure 1.11: Earnings by Economic Sector (2011)....................................................................................... 20 Figure 1.12: Monthly earnings (FCFA) and % of the workforce in the wage sector (2011) ........................ 21 Figure 1.13: Percentage of Nigeriens absent from their household in the last 12 months for work reasons .................................................................................................................................................................... 21 Figure 1.14: Percentage of age cohort by main activity (top panel) and school enrolment (bottom panel) by gender .................................................................................................................................................... 22 Figure 1.15: Percentage of age cohort by main activity (top panel) and school enrolment (bottom panel) by geographic area...................................................................................................................................... 24 Figure 1.16: Percentage of employed age cohort by sector of employment and gender (top panel) and geography (bottom panel) .......................................................................................................................... 25 Figure 1.17: Percentage of age cohort by gender and school enrolment (left panel) and marital status (bottom panel) ............................................................................................................................................ 26 Figure 1.18: School enrolment by age cohort and geography.................................................................... 27 Figure 2.1: Selected transition figures by gender ....................................................................................... 33 Figure 2.2: Women move in and out of inactivity ...................................................................................... 35 Figure 2.3: Characteristics of movers from unemployment to employment............................................... 35 Figure 2.4: Cross-country comparisons of farm/non-farm odds-ratios, men aged 20-69.......................... 43 Figure 3.1: Distribution of working age population, total and working age agricultural workers (primary or secondary activity), and farm size by agro-ecological zones.................................................................. 46 Figure 3.2: Share of individuals working in agriculture or livestock ........................................................... 48 Figure 3.3: Share of households with at least one member engaged in agricultural or livestock jobs ...... 48 Figure 3.4: Share of agricultural and livestock jobs (as percentage of active population’s primary occupation) in the workforce across gender and age cohorts ................................................................... 49 Figure 3.5: Acreage share, labor share, and market participation rate by crop ......................................... 50 Figure 3.6: Percentage of farm households who commercialize at least one of their crop ....................... 50 Figure 3.7: Share of agricultural employment by type of seasonality ........................................................ 51 Figure 3.8: Occupational rate of the working age population and average number of worked hours/week for the workers ........................................................................................................................................... 51 Figure 3.9: History of regional locally-produced millet markets................................................................ 52 Figure 3.10: Box plot of monthly millet price distribution by regional marketplace ................................. 52 Figure 3.11: Distribution of agricultural workers across land size classes (by land available and household net farm size – no double counts with intercropping) ............................................................................... 53 Figure 3.12: Distribution of plot sizes by crop for the main 8 crops in Ha ................................................. 54 Figure 3.13: Average net farm size and available land per worker (in hectares) by agro-ecological zones .................................................................................................................................................................... 55 Figure 3.14: Land ownership and land title by agro-ecological zones ........................................................ 55 Figure 3.15: Share of agricultural households using irrigation ................................................................... 56 Figure 3.16: Organic and inorganic fertilizer/pesticide use ........................................................................ 56 Figure 3.17: Median net farm size and technology coverage by technology (in hectares) ........................ 57 Figure 3.18: Share of agricultural workers with non-agricultural secondary occupation by age cohort, education level, and gender ....................................................................................................................... 58 Figure 3.19: Share of agricultural households with a mix of ag and non- occupations (processing, enterprise, other), by type of agriculture ................................................................................................... 58 Figure 4.1: Most Households do possess one non-agricultural enterprise or more .................................. 66 Figure 4.2: Most HEs operate throughout the year .................................................................................... 67 Figure 4.3: Most of HEs are not short lived ................................................................................................ 67 Figure 4.4: Source of inputs. Most inputs of HEs are from small local markets ......................................... 67 Figure 4.5: Client types. Most HEs products are sold in local markets ....................................................... 67 Figure 4.6: Type of activity of the non-agricultural micro-enterprises, by area of residence .................... 68 Figure 4.7: Most of non-agricultural household enterprises are not registered in any form ....................... 68 Figure 4.8: Most HEs do not keep books .................................................................................................... 69 Figure 4.9: Most HEs do not have fixed place to operate........................................................................... 69 Figure 4.10: Non-agricultural micro-enterprises owners are mostly young people................................... 69 Figure 4.11: Most of Non-agricultural micro-enterprises owners have no formal education ................... 69 Figure 4.12: Most Non-agricultural micro-enterprises owners are household heads ............................... 70 Figure 4.13: Most individuals in non-ag enterprises work alone................................................................ 70 Figure 4.14: Productivity dispersion by gender .......................................................................................... 71 Figure 4.15: Cumulative distribution of productivity by gender................................................................. 71 Figure 4.16: Productivity dispersion by residence ...................................................................................... 71 Figure 4.17: Cumulative distribution of productivity by residence ............................................................ 71 Figure 4.18: Output per worker by sector .................................................................................................. 71 Figure 4.19: Productivity premium compared to restaurant/hotel activities ............................................ 71 Figure 4.20: Most Non-agricultural micro-enterprises are financed by own savings................................. 72 Figure 4.21: Rural entrepreneurs rely more on own savings than urban counterparts ............................. 72 Figure 4.22: Main constraints and difficulties operating a non-agricultural household enterprise........... 72 Figure 4.23: Productivity premium compared to home based activities ................................................... 73 Figure 4.24: Years in operation and productivity ....................................................................................... 73 Figure 5.1: Employment status ................................................................................................................... 87 Figure 5.2: Primary Sectors of Employment ............................................................................................... 88 Figure 5.3: Did you aspire to a certain job as a child? ................................................................................ 89 Figure 5.4: If aspired to a certain job as child, which job?.......................................................................... 89 Figure 5.5: Aspired job by gender and educational level............................................................................ 91 Figure 5.6: Occupational aspirations vs. actual attainments ...................................................................... 91 Figure 5.7: Aspiration over next 5 years (restricted to out-of-school individuals) ..................................... 92 Figure 5.8: Aspired income by gender and geographical location ............................................................. 93 Figure 5.9: Income aspirations by educational level .................................................................................. 94 Figure 5.10: Aspirations by educational achievement................................................................................ 95 Figure 5.11: Parental Educational Background (%)..................................................................................... 97 Figure 5.12: Did your parents have a defined occupational aspiration for you? ....................................... 97 Figure 5.13: If parents have a defined aspiration, which job did they want for you? ................................ 98 Figure 5.14: Gender of role model .............................................................................................................. 99 Figure 5.15: Occupation of role model by geographical location ............................................................... 99 Figure 5.16: Perceived reason for role model’s success ........................................................................... 100 Figure 5.17: Do you feel you need to do the same job as your father/mother? ...................................... 103 Figure 5.18: Why do you feel you need to do the same job as your father/mother? .............................. 103 Figure 5.19: Need to move out of this place to find a good job? ............................................................. 104 Figure 5.20: Could you move out of this place to find a good job? .......................................................... 104 Figure 5.21: Parents' reaction to plan to move within 2 years ................................................................. 105 Figure 5.22: Desired support programs by gender ................................................................................... 105 Figure 5.23: Desired support programs by education levels .................................................................... 106 Figure 5.24: Do support programs exist in that area? .............................................................................. 107 Figure 5.25: Have you ever received any of these programs?.................................................................. 107 Chapter 1 : EMPLOYMENT STRUCTURES IN NIGER Highlight of the main results • 56 percent of the population in Niger is below 15 years old, the largest share of all comparator African countries. More than 500,000 youth enter the labor force every year. • 70 percent of Nigeriens have not completed any level of schooling; however young generations are more likely (40 percent of 15-24 year olds) to have been in school than old generations (7 percent of 55-64 year olds). • 34 percent of Nigerien women are out of the labor force; those working are employed on average for fewer hours (28 vs. 43) than men and receive lower earnings. • 0.5 percent: unemployment rate in the strict definition. But 43 percent of employed Nigeriens work less than 40 hours a week, and 32 percent work less than 24 hours a week. • 26 percent of employed Nigeriens report having a secondary occupation; 70 percent report being employed as seasonal worker. • 81 percent of employed Nigeriens work in agriculture, and 7-20 percent of people engaged in non-agricultural work report having a secondary occupation in agriculture; 42 percent of agricultural workers report earning zero income. • 4 percent of employed Nigeriens work in the non-agriculture wage sector, where earnings are more than 10 times higher than in agriculture. The extractive sector offers the highest wages, but only employs 1 percent of the wage workers. • 42 percent of children (5-14 years old) work, including 43 percent of 9 year-olds. • 60 percent of girls are have been married at age 18, against 2 percent of boys 1 Country context Due to frequent domestic and external shocks, Niger is one of the poorest countries in the world. Located in a sub-region repeatedly challenged by security threats such as the Libyan conflict, the Mali crisis and the Boko Haram armed attacks, the Nigerien economy is exposed to severe climate shocks that largely impact the country’s economic performance. Over the years, natural disasters have posed several challenges to Nigeriens’ lives and assets: the 2011 drought had serious impacts on half of the popula tion; in 2012, approximately 176,000 people were left without a home due to violent floods. As a result, with a Gross Domestic Product (GDP) per capita of US$895 (PPP, constant 2011)1, Niger in 2014 was the 6th poorest country in the world. Despite going through several episodes of political turbulence, Niger has experienced relative political stability, with beneficial effects for economic growth since 2011. Historically subject to frequent coups d’états since its independence from the French Community in 1960, the country experienced a period of relative national political stability between 1999 and 2010. However, the political environment worsened again in 2010, when, following a political crisis, a transition military regime was in power until 2011. Finally, in 2011 democracy was restored and a new president elected. Since then, the country has registered a good economic performance, with GDP growth as high as 11.8 percent in 2012 (the highest in the last 35 years) and 6.9 in 2014 (ibid). Traditionally driven by a rural sector frequently affected by climatic shocks and subject to the fluctuations in the price of its mineral exports, Niger’s growth rate is extremely volatile. Agriculture (36.7) and services (43.6) account for the largest shares of the GDP, while the industrial sector remains underdeveloped and only contributes to a small fraction (19.5) of domestic product (ibid). The country’s economic performance is highly dependent on the rainfall patterns. Following a poor harvest, GDP growth was only 2.3 percent in 2011, but it rocketed to 11.8 percent after a good harvest in 2012. Agricultural products and livestock account for half of Niger’s exports, while the other half of the exports is uranium. Since 2011, the country started its production in the petroleum sector, which together with uranium production is likely to be the main engine of growth in the medium term. Even in years of good economic performance, economic growth only partially translates into improvements in development outcomes due to extremely high rates of population growth and fertility. With an annual population growth of 4.0 percent since 2010, Niger has one of the most rapidly growing populations in the world, which at this pace is expected to reach 54 million people in 2050. Income per capita growth has severely suffered due to the persistence of these demographic trends, which has resulted in limited progress in the development agenda. Error! Reference source not found. show that even in a positive year such as 2012, when the county registered an annual GDP growth of 11.8 percent, per capita GDP growth was not as high as it could have been (7.4 percent). Even more worryingly, in years of slower GDP growth, such as 2011, GPD per capita growth was negative because of the high population growth (ibid). But the impacts of a fast growing population are not only detrimental for economic growth. With an average fertility rate of 7.6 children per woman, the resulting large population growth poses serious challenges to the already scarce infrastructures present in the country, especially health, nutrition and education services. As a result, maternal and child mortality as well as nutrition indicators are negatively impacted. Furthermore, high fertility has detrimental implications on the possibility for women to acquire education or participate in the labor force. Not surprisingly, as a result of these challenges, in 2014 the 1 World Bank World Development Indicators Database. 2 country ranked 187th out of 187 countries on the United Nation Development Program (UNDP) Human Development Index. Figure 1.1 GDP and Population Growth 20 15 10 5 0 -5 -10 -15 -20 -25 GDP growth (annual %) GDP per capita growth (annual %) Source: World Bank World Development Indicators Database Poverty reduction is closely linked to growth in agriculture, the sector that employs the majority of the population. With approximately 8 out of 10 workers employed in agriculture and 9 out of 10 poor individuals residing in rural areas, poverty reduction in Niger largely depends on the performance of the agriculture sector. Between 2005 and 2011, both the incidence and the depth of poverty fell by approximately 10 percentage points, from 51.3 and 20.5 percent to 40.8 and 10.4, respectively. However, the economic shocks that affect the country have severe impacts on the poorest segment of the population, with significant long-term repercussions. To illustrate, nearly one quarter of rural households reported to have decreased consumption of food to cope with the food crisis in 2009 and one third reported to have reduced the number of meals per day consumed by children. Another common coping strategy reported by rural household is depletion of productive assets, i.e. selling of livestock or consumption of seeds needed for planting. While sometimes necessary as an immediate reaction to catastrophic economic shocks, these behaviors can have long-term consequences on the country’s potential to further decrease poverty. Although investments in agriculture productivity have potentially large pro-poor impacts, households tend to make limited use of modern but more risky inputs because of their high exposure to shocks and food insecurity. Despite employing a large share of the workforce, agriculture is characterized by remarkably low levels of productivity and the majority of jobs in this sector are not sufficient to lift people out of poverty. Because of their high exposure to economic risks, workers employed in agriculture tend to choose low-yielding but less risky inputs. As a result, agricultural productivity in Niger is low compared to other countries in the Sahel region. Except for rice, average yields per hectare of agricultural products such as millet, sorghum and groundnuts are well below the Sahelian average (Poverty Assessment, 2012). 3 On the other hand, the mining sector is characterized by high productivity levels, but only employs a small fraction of the population. According to estimates based on 2011 data, the mining sector only accounts for less than a percentage point of total employment, but is characterized by high levels of productivity. Despite being capital intensive, oil and uranium productions are likely to generate large government revenues in the years to come. If used for strategic investments such as the modernization of the agriculture sector and the diversification of the economy, these resources could translate into significant improvements in the living standards of the Nigerien population. In a context in which employment income is the major source of livelihood for the majority of the population, a revamped jobs agenda is needed to improve the country’s development perspectives. The government is fully aware that GDP-growth-as-usual, based on low productivity agriculture and a capital intensive mining sector, is not sufficient to lead to significant improvements in living standards for the Nigerien population. As such, the Plan for Economic and Social Development 2012-2015 (PDES), launched in August 2012, identifies economic diversification and inclusive growth as two of the five main axes of the Poverty Reduction Strategy. In order to achieve the targets set in the PDES, the Government of Niger implicitly identifies three priorities for the new jobs agenda. To keep pace with the high population growth, the country needs to create more jobs. In line with this, the PDES aims to create 50,000 jobs per year between 2012 and 2015. But more jobs are not going to be a panacea for Niger’s development. In order to sustain a new model of economic growth, jobs also needed to be of a better quality, in terms of both earning potentials and development impact. For this reason, the PDES prioritizes employment in i) labor intensive infrastructure works that can help the country take full advantage of its mining and hydraulic resources and support economic diversification; ii) green jobs in the primary sector, in line with the country’s efforts on climate change mitigation and reforestation, and iii) jobs that put emphasis on skills development. Finally, the PDES highlight the importance of an economic model based on inclusive, balanced and sustainable growth. As such, going forward it is paramount for the Nigerien economy to create jobs that are more inclusive. In a country in which: i) more than two thirds of the population are below the age of 25, ii) women participation in the labor force is limited due to early marriage and high fertility rates, and iii) recent episodes of political unrest were fueled by perceived regional disparities, the jobs agenda needs to pay particular attention to population groups such as youth and women as well as to the promotion of a balanced regional development. But to successfully embark on the new jobs agenda, the country needs to address a number of constraints. Countries that in the recent decades have experienced successful structural transformations such as China, Brazil and Turkey were driven by the rise of a vibrant private sector (WDR, 2013). While in the short to medium term agriculture is still going to be dominant in Niger, going forward it is important for the country to create an enabling environment for private sector development. However, at present Niger’s formal private sector has one of the smallest bases in Sub-Saharan Africa. According to the 2013 Doing Business country ranking, Niger was among the bottom 10 economies in terms of easiness of doing business. These results reflect a number of challenges, including: i) an underdeveloped financial sector, with levels of depths well below the regional average and where rural sector and Small and Medium Enterprises (SME) have very limited access to credit; ii) a workforce that for the largest part is uneducated; iii) poor infrastructures and high transport costs, which, given the landlocked location of the country, sensibly hamper competitiveness; and iv) exposure to potential security threats, resulting from frequent conflicts and instability in neighboring countries. Nevertheless, Niger is currently experiencing a number of positive economic, social and political changes that could help the country make substantial progress with the new jobs agenda. Despite the numerous challenges, there are also some opportunities that the country is currently experiencing and could 4 take advantage of. The return to political stability and the announcement of the PDES are important signs that show both the Government’s commitment in leading the efforts towards a structural transformation and the popular support necessary to sustain the new agenda. Furthermore, following the Tuareg revolt in 2007-2009 due to claims on the unequal regional distribution of natural resources’ revenues, the country has started a progressive decentralization process, with beneficial effects on social cohesion. Finally, the start of oil production in 2011, together with the continuous demand for agricultural products from Nigeria and other neighboring countries, represent additional opportunities for the main economic sectors of the country. To sum up, Niger is a country that faces the jobs challenges of an agrarian and increasingly resource-based economy, and in which a large part of the population is young, poor and exposed to external and internal shocks. The 2013 World Development Report on Jobs provides a framework to help understand countries’ jobs challenges based on their level of development, economic resources, demographic patterns and institutional capacity. In Niger, most population is poor and lives in rural areas, implying that the jobs with the highest development impact are in agriculture. However, as discussed, jobs per se do not guarantee sufficient income opportunities to escape poverty; as such, increases in productivity are crucial for improving living standards. Moreover, giving the high exposure of poor and rural households to climatic shocks and food insecurity, the new jobs agenda needs to encourage the use of modern and inputs and decrease reliance on rainfall patterns. The increasing importance of the extractive sector offers both opportunities and challenges. Relying mainly on capital rather than labor, jobs opportunities in the natural resource sector could be limited; nevertheless, if strategically invested, the large revenues from the exports of uranium and petroleum could play an important role for the jobs agenda in Niger. In particular, potential investments to support the currently small manufacturing and handcrafts sectors in urban centers could be important to stimulate the development of a diversified economy. At the same time, given past episodes of political unrest due to the perceived lack of redistribution of wealth from the extractive sector, it will also be extremely important to ensure that all the different groups benefit from the newly available sources of prosperity. As the vast majority of the population is below 25 and only a minimal fraction has levels of education beyond primary, investments in skills development could improve youth’s access to better jobs thereby reducing the potential for conflict. Concurrently, interventions aimed at improving the business environment, especially access to credit, infrastructure and land rights, would attract more private investments and benefit the demand side of the labor market. The objective of this chapter is to: i) identify potential constraints that prevent the country from successfully implementing a jobs strategy based on more, better and inclusive jobs, and ii) discuss potential opportunities and policy options to overcome these barriers. The analysis of the workforce is based on data collected from the Household Living Condition Survey (Enquête Nationale sur les Conditions de Vie des Ménages, ECVMA) household survey in Niger conducted in 2011 and 2014. The main findings are based on the unbalanced panel and focus on the most recent survey from 2014, while employment dynamics using the balanced panel are postponed to Chapter 2. Hence, unless anything else is stated the statistics in the rest of the chapter refers to 2014. 5 Characteristics of the workforce Niger is the country with the youngest fraction of under 15 years old. According to the United Nations World Population Projections, approximately half of the Nigerien population in 2010 was younger than 15 years old. While young populations are common in western and central African countries, Niger is the country with the youngest population (Error! Reference source not found.). Figure 1.2: Population Pyramids Niger Male Female 80-84 70-74 60-64 50-54 40-44 30-34 20-24 10-14 49.8% 0-4 -15.0% -5.0% 5.0% 15.0% Western Africa Burkina Faso Male Female 80-84 80-84 Male Female 70-74 70-74 60-64 60-64 50-54 50-54 40-44 40-44 30-34 30-34 20-24 20-24 10-14 10-14 43.8% 46.0% 0-4 0-4 -15.0% -5.0% 5.0% 15.0% -15.0% 5.0% 6 Chad Mali Male Female Male Female 80-84 80-84 70-74 70-74 60-64 60-64 50-54 50-54 40-44 40-44 30-34 30-34 20-24 20-24 10-14 10-14 48.8% 46.8% 0-4 0-4 -15.0% -5.0% 5.0% 15.0% -15.0% -5.0% 5.0% 15.0% Nigeria Senegal Male Female Male Female 80-84 80-84 70-74 70-74 60-64 60-64 50-54 50-54 40-44 40-44 30-34 30-34 20-24 20-24 10-14 10-14 0-4 44.0% 43.6% 0-4 -15.0% -5.0% 5.0% 15.0% -15.0% 5.0% The vast majority of the Nigerian population lives in rural areas. Less than two of each ten people lived in urban centers in 2014. As for the gender distribution of the population, there seem to be only a slight imbalance with around 2 percent more women in the country. The geographic distribution of the population reflects the fact that only some areas in the country are habitable. For instance the deserted regions of Agadez and Diffa account for respectively 50 and 13 percent of the country land mass, but for only 2.7 and 3.6 percent of the population. On the other hand, the regions in the south, which benefit from easier access to underground water, are home to the majority of the population. Despite the limited land area of the capital Niamey, 6 out of 100 people in Niger live there. The Nigerien population can be divided into nine ethnic groups. Out of these, three ethnic groups account for almost 90 percent of the total population. More than half of the population belongs to the Haoussa ethnic group, while 20 percent are Djema and 11 percent Touareg. Less than one percent of the population reports to be foreigner. This group mainly resides in Niamey. 7 Table 1.1: Population in Niger (%) Unweighted Weighted Total population 22,756 18,613,744 Gender: Female 51.5 51.2 Male 48.5 48.8 Geographical location: Rural 64.3 84.1 Urban 35.7 15.9 Region: Agadez 10 2.6 Diffa 8.9 3.5 Dosso 11.6 12.2 Maradi 13.3 20.8 Tahoua 11.3 18 Tillaberi 11.6 15.4 Zinder 13.7 21.3 Niamey 19.7 6.2 Ethnicity: Arab 0.4 0.2 Djema 25.4 20 Goumantche 0.4 0.2 Haoussa 40.9 57.4 Kanouri-Manga 7.9 6.5 Peul 6.2 3.8 Touareg 15.4 11 Toukou 2.1 0.7 Other Nigerien ethnicity 0.1 0 Foreigner 1.2 0.4 The working age population account for 52 percent of the population and is poorly educated. According to the ECVMA, this corresponds to 7.9 million people. As seen from Error! Reference source not found., the level of education among the working age population are remarkably low, as approximately three-fourth of all people between the age of 15 and 64 report to have no levels of education. 8 Figure 1.3: Working Age Population by Education Level (2014) 3% 10% 11% 76% None Primary Secondary first cycle Secondary second cycle and above Nigeriens educational level varies by gender, geographical location and age. Error! Reference source not found. shows that levels of education are heterogeneous across different population groups. For instance, men are more likely (32 percent) to have completed some education than women (18 percent). On the other hand, eight in each ten individuals living in rural areas do not have any education, as opposed to four in each ten in urban areas. Finally, younger generations seem to be more likely to pursue some levels of education. Figure 1.4: Distribution of education by gender, age and urban/rural divide among the working age population (2014) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% None Primary Secondary first cycle Secondary second cycle and above 9 Employment structures Most people in Niger are engaged in some form of economic activity. In line with the rapid population growth described in the introduction to the chapter, the population has increased almost 11 percentage points between the two survey rounds. Out of the more than 18 million people living in the country in 2014, approximately 7.9 million (or 43 percent of the total) are between 15 and 64 years of age, i.e. belong to the working age population. Of these, approximately 3 out of 4 report to have worked sometime during the last 12 months before the time of the survey or to have a job to go back to. However, unemployment in the country is virtually non-existent, as less than 1 percent of the working age population reports to be jobless even if looking for a job in the 12 months preceding the survey (Error! Reference source not found.).2 Figure 1.5: Distribution of Population in Niger While most people in Niger are employed, quality of employment emerges as a major issue. The fact that 76 percent of the total working age population reports to be employed and only a minimal fraction 2 The unemployed are those who in the last 12 months did not work, do not have a job to return to and did not look for a job. Given the incidence and severity of poverty in Niger, it is not surprising to find the only a very small number of people look for a job without finding one in the 12 months preceding the survey. A shorter recall period is available in the data, however the shorter recall period for the survey undertaken in 2011 and 2014 differ and thus are not comparable across survey years. 10 can afford to be unemployed suggest that employment income is a crucial source of livelihood for the Nigerien population. The vast majority of Nigeriens between 15 and 64 works in agriculture, a sector that, as discussed, is highly subject to climate shocks and in most cases does not provide any source of social protection to its workers. Similar conditions exist in the household enterprise sector, which employs approximately 9 percent of the working age population. Finally, only nearly 6 percent of the Nigeriens in working age report to work as wage workers. Approximately half of them are employed in the private sector, but even in these instances only 25 percent workers have formal contracts, as opposed to approximately 55 percent in the public sector. All the groups in the Nigerien population are generally employed, while unemployment in the country is low; however, some groups face higher barriers to access the labor market. Unemployment rates across the population in Niger are remarkably low for all the different groups; as such labor force participation and employment to population ratios are largely similar. Among different age groups, the youth tend to be less employed, but this reflect schooling decisions rather than difficulties in finding a job, as only 0.5 percent of the active population is unemployed. Women participate less than men in the labor force and are less likely to look for a job, as they report to be mainly involved in household chores. Labor force participation and unemployment is higher in urban centers, as they generally register lower levels of poverty and have a more educated population who can afford to spend some time in search of good quality employment. This finding is reinforced by the labor market indicators for people with different education levels: in fact, people with higher levels of education experience lower participation rates and higher unemployment rates. Again in line with these findings, similar labor market patterns are found in the two regions with the highest urbanization rates, i.e. Agadez and Niamey. The economic active population has declined since 2011. As a consequence of the large population growth and thus young Nigerien population the percentage of active people has declined from 81 percent in 2011 to 76 percent in 2014. This is also confirmed by the labor force participation rate which fell by 5 percentage points between the two survey rounds. The decline is particularly reflected in the participation rate for women and young people between 15 and 24 years old. Moreover, compared to the 2011 survey, an increasing fraction of the active population worked in agriculture in 2014, while the share of the population working in the non-agricultural wage sector declined by 1 percentage point. This suggests that the Nigerien employment transition towards industrialization has not yet begun. Table 1.2: Labor Market Indicators for Different Groups, Population 15 to 64 Labor Force Participation Employment to Unemployment Rate Population Ratio Rate Total 76% 76% 0.5% Gender: Female 66% 66% 0.3% Male 90% 89% 0.7% Geographical location: Rural 81% 81% 0.1% Urban 56% 55% 2.6% Age group: 15-24 64% 63% 0.5% 25-34 79% 79% 0.7% 11 35-44 83% 83% 0.3% 45-54 85% 85% 0.2% 55-64 79% 79% 0.8% Region: Agadez 57% 54% 4.7% Diffa 76% 76% 0.0% Dosso 89% 88% 0.4% Maradi 88% 88% 0.1% Tahoua 67% 67% 0.0% Tillabéri 75% 74% 0.4% Zinder 80% 80% 0.0% Niamey 52% 50% 4.2% Ethnicity: Djema 76% 75% 0.9% Haoussa 79% 79% 0.2% Touareg 70% 69% 1.0% Others 72% 71% 0.4% Educational level: None 79% 79% 0.1% Primary 75% 74% 0.8% Secondary first 64% 62% 1.9% cycle Secondary second 51% 48% 6.0% cycle and above Characteristics of the population by employment status People without employment: inactive and unemployed Understanding the barriers faced by individuals without a job is of crucial importance to inform policies aimed to create more, better and more inclusive jobs. While some individuals voluntarily decide to not engage in any economic activity, in other instances, social, economic and cultural factors prevent people from joining the labor force. This section of the chapter focuses on two different groups of the “jobless� in Niger, i.e. those out of the labor force and the unemployed. The analysis sheds light on the main characteristics of these groups and highlights the main reasons behind these labor market outcomes. Out of all those not in the labor force, most individuals are either engaged in housework or in school. Individuals out of the labor force are those who do not have a job and do not look for one. This form of disconnection from the labor market could be the result of factors such as lack on willingness to work, discouragement, social norms, disability, schooling, etc. For individuals between 15 and 64 years old, being out of the labor force because of schooling is not a source of worry. In fact, this decision may reflect the willingness to join the labor market at a later stage with higher levels of human capital and higher potential earnings. On the other hand, the group of individuals reporting to be out of the labor force for other reasons may face the challenges that result from a malfunctioning labor market. As such, this group, 12 referred to as economically inactive, is generally subject of a deeper analysis. As already shown in Error! Reference source not found., 81 percent of those out of the labor force are economically inactive. The remainder of this section will delve deeper into the characteristics of the inactive population. Figure 1.6: Inactivity Rates for different population groups 35.00 30.00 25.00 20.00 15.00 10.00 5.00 - Touareg Total Diffa Urban Agadez Niamey Haoussa Others Rural 25-34 Female 15-24 35-44 45-54 55-64 Dosso Maradi Djema Secondary first cycle Secondary second cycle+ Male Tahoua Tillabéri None Zinder Primary Inactivity does not seem to be a transitory phase in the life of those who do not work and are not looking for a job. Some individuals could decide to stay out of the labor force for a certain period due to some contingent situations such as perceived lack of suitable employment opportunity, lack of willingness to work or seasonal unemployment. However, among the inactive population only 3.3 percent reports to have had a job in the 12 months before the time of the survey. As such, inactivity in Niger appears to be a long term labor market status. The groups more likely to be inactive are women, youth and people with low levels of education and residents in urban centers or in the wealthier states. Error! Reference source not found. shows that people aged 15 to 24, women, people with no or primary education as well as individuals living in urban areas, and in the states of Agadez, Tahoua and Niamey face higher than average likelihood of inactivity (red bars). Societal norms are the main factors influencing inactivity. A closer look at the differences in the reasons for inactivity between various groups sheds light on potential barriers to employment faced by the population in Niger. Eight out of ten inactive women report to not engage in economic activities or looking for a job because of housework (Figure 1.7). On the other hand, less than 0.4 percent of male indicates housework as a reason for inactivity. In addition, when asked questions about time use in the seven day before the survey, half of the inactive men report to have not spent any time on activities such as gathering firewood, fetching water, cooking, doing the laundry, ironing clothes, cleaning the household of shopping for the households. 13 Figure 1.7: Reasons for not looking for a job by gender - working age population Male Female Other 6% 1% Don't know how to look 4% 0% Lack of employment 19% 1% Waiting response to employment demand 7% 1% Waiting to start own business 4% 0% Too old/retired 12% 6% Housework 0% 82% Sickness/handicap 22% 4% Too young 1% 1% Does not want to work 12% 1% -60% -40% -20% 0% 20% 40% 60% 80% 100% Wealth and education levels also influence labor market choices. Across different age groups, housework is mentioned as the main reason for not looking for a job, but this is less so for older age cohorts (55+), for which retirement becomes the main reason. While being also the predominant reason for inactivity in urban areas, housework is even more so in rural areas. In fact, in urban areas, larger shares of the working age population seem to be able to afford inactivity, as they are more likely to report that they are waiting for future (better) employment opportunities. A similar pattern is visible in Agadez and Niamey and for people with secondary education and above. The unemployed are a very small fraction of the working age population, and are concentrated among the youth in search of first employment. Only about 29 thousand people, out of a total working age population of almost 8 million people, report to not have a job and to not have found one despite looking for it in the 30 days before the survey. Of the unemployed, 65 percent are below the age of 35, while 16 percent is above 55 years old. Based on the survey it is not possible to get a reliable picture of the reasons for the unemployed not to be looking for a job as only 10 percent in 2011 and none in 2014 responded to the question regarding non-work during the last 30 days. Similarly, to inactivity, unemployed is a choice influenced by societal norms and wealth/education status. As already discussed and shown in The economic active population has declined since 2011. As a consequence of the large population growth and thus young Nigerien population the percentage of active people has declined from 81 percent in 2011 to 76 percent in 2014. This is also confirmed by the labor force participation rate which fell by 5 percentage points between the two survey rounds. The decline is particularly reflected in the participation rate for women and young people between 15 and 24 years old. Moreover, compared to the 2011 survey, an increasing fraction of the active population worked in agriculture in 2014, while the share of the population working in the non-agricultural wage sector declined by 1 percentage point. This suggests that the Nigerien employment transition towards industrialization has not yet begun. 14 , women tend to register slightly lower unemployment rates than men. This does not imply that the female population has higher chances to finding a job when looking for one; rather, this is the result of the fact that women are less likely to look for a job. Generally, unemployment is almost entirely an urban phenomenon. Despite urban areas are home of only 20 percent of the working age population, 77 percent of the people who report to look for a job without finding one live in urban areas. A similar pattern is true for individuals with higher education and residents in Agadez and Niamey. These results suggest that people with higher income or better job prospects can afford to wait until more suitable job opportunities become available. People in Employment: underemployment and multiple activities The largest part of the working age population in Niger has a job, but quality of employment is an issue. The previous section shows that the inactive population is mainly composed by women who, due to societal norms, do not participate in the labor market, and by a small number of individuals who can afford to be without employment while waiting for better job opportunities. Nevertheless, almost eight out of each ten Nigeriens have a job. But in a country in which approximately 50 percent of the population lives with less than $1.90 a day and 82 percent with less than $3.10 a day having a job is clearly not enough to escape poverty.3 The fact that 46 percent of the population aged 5 to 14 (more than 2.3 million children) and 57 percent of the population 65 years and older (280 thousands elderly) are employed confirms that existing jobs in Niger are not sufficient to ensure adequate living standards for local households ( ). Figure 1.8: Labor Market Status of Population Not In Working Age Child (5 to 14) Elderly (65+) 46% 42% 54% 57% 0% 1% Employed Unemployed Out of the Labor Force Employed Unemployed Out of the Labor Force Box 1.1. Child Labor In Niger As a result of the insufficient earnings from employment, a large number of children between 5 and 14 years of age are involved in some forms of economic activities to contribute to family income. As shown in this chapter, 42 percent of 5 to 14 year olds report to be employed. The majority of children in employment are boys (60 percent). 3 World Bank Indicators. Poverty headcounrt ratios reported for year 2011. 15 Children are mostly employed as family aide in the agriculture sector . Child labor is mostly common in rural areas (97 percent of the total), where children work to support their families (97 percent) in the agriculture sector (98 percent). For children who work, employment takes a large part of their time, with potential implication for human capital accumulation. At age 7, more than 30 percent of Nigerien children are engaged is some form of economic activity. Some 16 percent of those between age 5 and 14 combine school and employment. As working children are employed on average for 27 hours per week, it is likely that work obligations may have an impact on their school progression. The negative impact on learning is going to be even more severe, given that slightly more than two thirds (68 percent) of children in employment also reports to have spent time helping with the household chores in the 7 days preceding the survey. Most Nigeriens are primarily employed as seasonal workers and a large share of them work below their full potentials. The structure of the Nigerien workforce reflects the characteristics of an economy mainly driven by an agriculture sector subject to seasonality and frequent shocks. As such, 70 percent of those with a job in 2011 report to be primarily employed as seasonal workers. On average a Nigerian worker was employed for 37.5 hours a week in 2011. This compares to 36 hours a week in 2014. The decrease in the number of hours worked may partly be explained by the fact that the labor supply increase more rapidly than the labor demand as a consequence of the constantly rising population density. According to Error! Reference source not found., half of the primary jobs in 2014 employ people for 35 hours a week or less, suggesting that workers still have some time to dedicate to other activities. Using a 40 hours’ threshold in line with the ILO criteria and a less stringent classification, 55 percent of the Nigerien workers are classified as underemployed in 2011, i.e. working less than 40 hours a week.4 This share of the employed increased to 62 percent in 2014. Figure 1.9: Distribution of Hours Worked per Week in Primary Jobs 4 According to the ILO definition, a worker is classified as underemployed is s/he is working less than 40 hours and willing to work more. The ECVMA 2011 does not collect information on whether workers would be willing to work more, but given the high poverty status in the country it is expected that most workers would meet this additional condition. 16 25 20 50% % of workers 15 10 5 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 Weekly Hours worked As existing jobs are not sufficient to ensure minimum living standards, a large number of Nigeriens are employed in a secondary economic activity, often different from their primary occupation. Almost three in each ten workers report to have had a secondary employment during the last 12 months. According to Error! Reference source not found., the share of Nigerien with a secondary job has declined over time, independent of sector of occupation. As a form of protection against risk, most workers with a secondary job choose to differentiate sources of income. For instance, in 2011 only 11 percent of those with a primary job in agriculture are also employed in this sector for their secondary job ( Error! Reference source not found.). Workers employed in agriculture, however, were less differentiated in 2014 as compared to 2011. In 2014, 16 percent of those with a primary job in agriculture were also employed in this sector in their secondary job. Figure 1.10: Economic Sector of Primary and Secondary Jobs 100% 90% 80% 70% 64% 60% 73% 71% 85% 84% 89% 50% 40% 30% 3% 1% 8% 20% 22% 6% 1% 4% 5% 2% 2% 10% 20% 1% 3% 11% 10% 16% 12% 0% 7% Agriculture Non-Ag Self- Non-Ag Wage Agriculture Non-Ag Self- Non-Ag Wage Employed Employed 2011 2014 Agriculture Non-Ag Self-Employed Non-Ag Wage No secondary Job 17 Characteristics of jobs by sector In order to fully understand challenges and identify opportunities in the Nigerien labor market, it is important to go beyond analysis of aggregate trends and explore differences across the three main employment sectors, i.e. agriculture, non-agriculture self-employment and non-agriculture wage sectors. As such, the remainder of this chapter will highlight the typical features of employment in these sectors as well as the characteristics of their workforce. The analysis will have a particular focus on the characteristics that can be associated to good quality and inclusive jobs. Table 1.3: Employment Sectors by Selected Characteristics Agriculture Non-Ag Self- Non-Ag Employed Wage Total 4,885,242 892,857 248,740 (%) Public 0.0 0.0 57.5 Private 95.5 97.7 34.0 Household 4.5 2.3 8.6 Permanent 13.2 52.3 76.6 For a specific period 33.8 26.5 17.7 Temporary(seasonal) 53.0 21.1 5.7 Formal 0.4 2.2 48.3 Informal 99.6 97.9 51.7 Underemployed 59.6 45.8 26.2 Have secondary Job 28.6 15.6 10.8 Received health benefits 0.3 0.4 31.6 Despite only accounting for a small share of total employment, jobs in the Non-Agriculture Wage sector are those of highest quality. As already discussed, the working age population in Niger is mainly employed in Agriculture (81 percent), while the remaining 19 percent works in Non-Agriculture Household Enterprises (15 percent) or in Non-Agriculture Wage sector (4 percent). The latter provides better jobs opportunities to its workers. Error! Reference source not found. shows that the wage sector is characterized by higher levels of formality (defined as existence of an employment contract), lower levels of underemployment (therefore fewer workers with secondary jobs), and higher shares of the workforce covered by medical care. Non-Agriculture Wage employment is also the only sector in which a state owned or public enterprise can potentially be the main employer. In addition, it is worth noting that, differently from agriculture, the other two sectors provide longer-term job opportunities for their workers. Finally, while the number of weekly hours worked on average in agriculture and non-farm self-employment are 18 very similar (34 hours), individuals employed in the wage sector work on average more hours per week (46 hours). Table 1.4: Access to Different Employment Sectors by Selected Groups Agriculture Non-Ag Self- Non-Ag Employed Wage Total 81.1 14.8 4.1 Gender: Female 80.8 17.0 2.2 Male 81.3 12.9 5.9 Geographical location: Rural 90.8 8.5 0.7 Urban 19.6 54.7 25.7 Age group: 15-24 86.3 10.8 3.0 25-34 78.9 17.2 3.9 35-44 79.0 16.1 4.8 45-54 79.4 15.1 5.5 55-64 81.9 14.3 3.8 Region: Agadez 34.6 45.9 19.5 Diffa 88.8 8.6 2.6 Dosso 92.3 6.9 0.8 Maradi 85.2 13.3 1.4 Tahoua 86.7 11.2 2.1 Tillabéri 89.1 9.6 1.3 Zinder 81.0 17.0 2.0 Niamey 3.2 53.2 43.6 Ethnicity: Djema 81.5 12.0 6.5 Haoussa 81.6 15.5 2.9 Touareg 80.0 15.8 4.2 Others 78.3 15.6 6.1 Educational level: None 86.8 12.2 1.1 Primary 66.4 27.0 6.5 Secondary first cycle 61.4 24.4 14.3 Secondary second cycle+ 11.7 11.9 76.4 Access to quality jobs is not equal across different groups of the Nigerien population. While jobs in the non-agriculture wage sector offer better quality jobs, Error! Reference source not found. shows that jobs in the non-agriculture only accounts for a small share of the workforce. In order to pursue a jobs agenda based on inclusive jobs, it is extremely important to understand whether there are barriers that prevent certain groups to access these quality jobs. With this objective, Table 1.4 shows the employment distribution of selected population sub-groups across the three different employment sectors. The main message that emerges is that some groups have a significantly easier access to the non-agricultural wage 19 sector. These groups include: (i) male workers, (ii) those living in urban areas, (iii) in Agadez and Niamey and (iv) with secondary education and above. On the other hand, youth, ethnic Haoussa and Touareg as well as individuals with no education, living in rural areas and in Dosso, Maradi and Tillaberi are more likely to face barriers that prevent them from accessing the non-ag wage sector. Women have limited access to good work opportunities. Female workers are less likely than men to find employment in the non-ag wage sector, while they are more likely to be non-ag self-employed, where hours worked are usually fewer. As a result, while on average a Nigerian male worker is employed for 43 hours a week, women average weekly hours worked are 28. This finding reinforces the evidence according to which societal norms play an important role in women’s labor market decisions. A large share of workers in agriculture works as family aide and as such does not receive any remuneration; even among those receiving earnings from their work, earnings in agriculture are sensible lower than in the other two sectors.5 Out of all the workers primarily employed in agriculture and reporting remuneration for their work in 2011, more than 40 percent of the respondents report to receive remuneration equal to 0 FCFA (Error! Reference source not found. – left panel). This phenomenon is less common in the other two sectors, as only 6 and 7 percent of the total workforce report to earn zero FCFA in non-farm self-employment and wage sector, respectively. The vast majority of those receiving no remuneration from work are individuals working for their families (88 percent in agriculture and 83 percent in non-ag self- employment) or trainees/apprentices (90 percent in non-ag wage). Even among those who report a positive remuneration in 2011, earnings in agriculture a nearly one third lower than in non-ag self-employment and more than one tenth lower than in the non-agriculture wage sector. These differences are even larger if it is taken into account that 35 percent of the wage workers report to receive at least some form of additional benefits,6 while only one percent in the other two sectors receive such additional compensations. Across all sectors, regression analysis shows that women consistently earn less than men. Furthermore, education ensure higher levels of remuneration only beyond primary levels in agriculture and non-farm self- employment, i.e. there is not much difference in earnings levels between comparable individuals with none and primary education. Figure 1.11: Earnings by Economic Sector (2011) A: Workers Receiving No Remuneration B: Median Earnings 5 These findings are exclusively based on information about earnings collected in 2011. The information about earnings in 2014 is imprecise due to a substantial number of missing observations. 6 Additional benefits include compensations for: housing, clothing, fuel or transport, domestic worker, communication, water/electricity, school fee, and family grants. 20 42% 62000 18000 6% 7% 5968 Agriculture Non-Ag Self- Non-Ag Wage Agriculture Non-Ag Self- Non-Ag Wage employment employment % workers with no earnings Median months earnings in FCFA (only earnings>0) The public sector generally offers best earning opportunities for a larger share of the (small) workforce in wage employment; private sector based activities offer income opportunities at both ends of the remuneration spectrum. The sub-sectors offering highest and lowest earning opportunities in wage employment are mainly led by the private sector. While the most attractive private sub-sectors (extractive activities and finance/real estate) employ a small fraction of the wage workforce, the least attractive employ larger shares. For instance, private sector based activities such as transport, manufacturing, wholesale/retail and personal services usually do not take advantage of economies of scale and provide lower earnings, while workers employed in extractive activities, utilities and a finance/real estate (overall only 4 percent of the wage workforce) register top earnings. On the other hand, public-led sub-sectors such as public administration, education and health offer good earnings opportunities and employ the educated workforce (Error! Reference source not found.), typically in urban areas. In the wage sector, wages are higher for workers employed in the public sector and for those with a formal contract. Regression analysis shows that, even when jobs characteristics such as gender, education, age, geography and sector of occupation are accounted for, workers employed in the public sector are more likely to have higher wages than those employed in the private sector. Similarly, workers with a formal contract systematically show to have higher wages than informal workers. High levels of education also appear to be strong predictors of high earning opportunities, while women and younger workers tend to receive lower wages. 21 Figure 1.12: Monthly earnings (FCFA) and % of the workforce in the wage sector (2011) 350,000 25% 300,000 20% 20% 250,000 15% 200,000 15% 11% 14% 15% 150,000 10% 100,000 8% 6% 5% 50,000 1% 3% 4% 1% 2% 0 0% 0% Median monthly earnings in FCFA (only earnings>0) % workforce in the wage sector Despite the large differences in economic opportunities between rural and urban areas, only a small percentage of Nigeriens report to have been away from their household for work reasons in the last year. Among all the working age population, only 10 percent report to have been absent from their household for work reasons (Error! Reference source not found.). Temporary migration is a phenomenon common exclusively among men, as virtually no women report to have been absent from their households in the 12 months preceding the survey. Other individual characteristics are correlated with the propensity to temporarily migrate for work: individuals between 25-54 years old as well as Nigeriens with some education and residents in Zinder and more likely to be absent from home for work reasons. While more than half of those who temporarily move stay in the same region, urban Nigeriens are more likely to move to other regions within Niger or to foreign non-neighboring countries. There is also a vast geographic heterogeneity in destinations: individuals from Diffa are most likely to stay in Diffa (77 percent), while individuals living in Zinder and Marandi are relatively more likely migrate temporarily to a neighboring country. Figure 1.13: Percentage of Nigeriens absent from their household in the last 12 months for work reasons 20% 16% 12% 13% 12% 10% 11%11% 10%11%11% 10% 9% 8% 9%10%9% 9% 8% 7% 6% 7% 7% 5% 5% 2% Secondary… Secondary first… Agadez Haoussa None Primary Rural Niamey Male Maradi 15-24 25-34 35-44 45-54 55-64 Diffa Tillabéri Touareg Total Dosso Tahoua Zinder Djema Female Urban Others 22 Transitions: school-to-work and occupational mobility Understanding the main constraints faced by the Nigerien population when transitioning from inactivity to employment or between different economic activities is crucial to identify policies focused on creating more, better and more inclusive jobs. This sub-section will focus on school-to-work transitions, transitions between sectors and family formation. A large part of the Nigerien youth starts to work at an early age, with negative implication for school progression and work transition. As discussed in the previous sections, educational levels in Niger are remarkably low and a large number of youth in young age cohorts is employed in some forms of economic activities. Nearly 15 percent of all five-year-old children in the country is employed and does not go to school, and the percentage rises to 48 percent for 14 years old. These patterns are similar for both boys and girls, but while by age 10 most boys work and/or goes to school, the percentage of girl/women between age 10 and 35 not in school nor in employment is never lower than 20 percent. Generally, boys are more likely to attend school, but between age 10 and 17 they are less likely to only focus on schooling as a large number of them mix work and formal education. As a result, full transition to work happens slowly (after age 26 for women and after age 28 for men, Figure 1.14 – top panel). In most cases, slow transition to work is the result of a slow school progression rather than of attendance of higher levels of education. Figure 1.14 (bottom panel) shows that 12 percent of boys and 8 percent of girls aged 15 are still in primary school, while 21 percent of men and 16 percent of women aged 18 attend the first cycle of secondary school. Figure 1.14: Percentage of age cohort by main activity (top panel) and school enrolment (bottom panel) by gender Men Women 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 Only work Only work Work and Formal Schooling Work and Formal Schooling Only Formal Schooling Only Formal Schooling Not Employment or Formal Schooling Not Employment or Formal Schooling 23 Men Women 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% 12 13 14 15 16 17 18 19 20 21 22 23 24 25 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Secondary second cycle+ Secondary second cycle+ Secondary first cycle Secondary first cycle Primary Primary In urban centers, schooling is the main focus for the majority of children up to age 15. 85 percent of 8-year-old children only goes to formal schooling, while 7 percent combines school with work (Error! Reference source not found.). Overall, only a minority of youth in urban areas works and study at the same time; young Nigeriens focus mainly on school in the early years and by age 29 virtually all Nigeriens in urban centers either work (64 percent) or are out of both the labor force and formal education (33 percent). Those who remain in school after age 25 generally pursue higher level of education: out of the 8 percent of the 26 year olds who attend formal schooling in urban centers, the vast majority of them (83 percent) were enrolled in tertiary education and above. Differently from urban areas, only a minority of children in school age are in formal education in rural areas. In rural centers, where the majority of the Nigerien population lives, only a small fraction of youth can afford to exclusively attend school. According to Error! Reference source not found., the age cohorts most exposed to formal schooling are the 8 and 9 years olds: among the 8 year olds 27 percent only attend formal schooling and 17 percent combines schooling with work; among the 9 year olds 26 percent only focus on formal schooling and 22 percent combines formal education with employment. After age 16, almost no Nigerien in rural areas is exclusively focused on studying: the small numbers of youth who continue with their studies do so by combining education with work. The education levels attended by youth in rural areas are almost never higher than primary (up to age 16) or the first cycle of secondary (between age 16 and 24). 24 Figure 1.15: Percentage of age cohort by main activity (top panel) and school enrolment (bottom panel) by geographic area Urban Rural 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 5 8 11 14 17 20 23 26 29 32 35 5 8 11 14 17 20 23 26 29 32 35 Only work Only work Work and Formal Schooling Work and Formal Schooling Only Formal Schooling Not Employment or Formal Schooling Only Formal Schooling Urban Rural 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 12 14 16 18 20 22 24 26 28 12 14 16 18 20 22 24 26 28 Secondary second Secondary second cycle+ cycle+ Except in urban areas, transition out of the agriculture sector is limited for both men and women, but women have less opportunities to access better job opportunities with age. As shown in Error! Reference source not found. – top panel, virtually all boys below age 13 were employed in agriculture, while by the same age 9 percent of girls was in non-agricultural self-employment. However, only very few girls eventually move to the wage sector: only 1 percent of the 28 year old employed women were working in the wage sector, compared to 4 percent of the employed men in the same age cohort. Access to the private wage sector, however, is only slightly easier for men compared to women across all the age cohorts. Urban centers offer more opportunities to transition into the wage sector, especially after age 16. As expected, ony a small share of rural employed youth in each age cohort is employed outside of agriculture (bottom panel). 25 Figure 1.16: Percentage of employed age cohort by sector of employment and gender (top panel) and geography (bottom panel) Men Women 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 5 8 11 14 17 20 23 26 29 32 35 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 Non-Ag Wage Household Non-Ag Wage Household Non-Ag Wage Private Non-Ag Wage Private Non-Ag Wage Public Non-Ag Wage Public Non-Ag Self-Employed Non-Ag Self-Employed Agriculture Agriculture Urban Rural 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 5 7 9 11131517192123252729313335 5 8 11 14 17 20 23 26 29 32 35 Non-Ag Wage Household Non-Ag Wage Household Non-Ag Wage Private Non-Ag Wage Private Non-Ag Wage Public Non-Ag Wage Public Non-Ag Self-Employed Non-Ag Self-Employed Agriculture Agriculture Gender-specific constraints to participation Early family formation and societal norms represent significant barriers for women to access more and better jobs. The evidence presented in this section suggests that in all the age cohorts women are less likely 26 to start their transition to the labor market. Figure 1.17 (left panel) shows that this is not due to schooling decisions, as enrolment rate in each age cohort is almost consistently lower for girls/women than for men. On the other hand, the right panel of Figure 1.17 shows that women are disproportionately more likely than men to enter in a marriage at very early stage of their life. As fertility rates in Niger are extremely high, early marriages often result in teenage pregnancy with clear implications on women’s availability to wo rk. Even when available to work, the presence of large families and children to take care of limit the employment choices of women. As a result, female workers face higher barriers to access the wage sector. Figure 1.17: Percentage of age cohort by gender and school enrolment (left panel) and marital status (bottom panel) % enrolled in school % ever married 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% 12 14 16 18 20 22 24 26 28 30 32 34 12 14 16 18 20 22 24 26 28 30 32 34 Men Women Men Women At the same time, the low levels of enrolment of children in rural areas pose severe challenges for the future employment prospects of these groups. One of the most striking differences between urban and rural areas is the employment transition of youth across different sectors. While the vast majority of young children in school age attend formal education in urban centers, not even half of the children in rural areas are in school (Error! Reference source not found.). In order for rural areas to be able to create more and better jobs, access to education for children in school age needs to be substantially improved. 27 Figure 1.18: School enrolment by age cohort and geography 100% 80% 60% 40% 20% 0% 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Urban Rural Summary of main findings The Nigerien economy is characterized by a workforce with low levels of education, largely concentrated in rural areas. Because of levels of productivities in agriculture well below the regional average, jobs in this sector are not productive and as a result do not ensure enough income to lift households out of poverty. As a result, a large number of children in rural areas spend substantial parts of their time contributing to family income. This creates a vicious circle, where low income imposes limited school attendance, which in turn creates low human capital accumulation and low income for the future generations. As such, in order for the country to be able to create better jobs, it is extremely important to increase the productivity levels in agriculture so as to create incentives for families to send their children to school. The recent upward trends in school enrolment are a reassuring step in the right direction. Societal norms and expectations limit the full potential of women in the labor market. Early family formation limits women’s possibilities to acquire education and/or enter the labor market. A large number of women are out of the labor force and even if interested in re-entering the labor market, their low levels of education would limit their job opportunities. Women in employment are mainly concentrated in a low- productive and low-paying farm and non-farm self-employment, where they only work for a limited number of hours. Furthermore, there is some evidence of gender gaps in earnings even in the wage sector. Better employment is more common in urban centers, but there are some challenges associated with expansion of employment in these jobs. Urban locations are usually characterized by a more educated workforce and a larger presence of jobs in the wage-sector, which offer better pays, more stability and additional benefits. However, the jobs offering higher salaries are either in the extractive sector, which being capital intensive has limited capacity to absorb large number of workers, or in sectors such as public administration, utilities, education and health, which are largely public sector based. Strengthening the links of the extractive sector with the other sectors in the economy is likely to bring high pay offs for the country. Taking advantage of the booming extractive sector to develop a more dynamic manufacturing and service sector in cities could lead to the creation of a large number of good jobs, especially for women. In fact, connecting the non-farm self-employment sector (largely populated by female workers) to the larger markets created by the influx of revenues in the extractive sector could result in new and better employment opportunities for the Nigerien workforce. In line with this and to benefit from the opportunities in urban centers, efforts to improve the accessibility of cities from rural areas could improve living standards in remote areas. 28 29 Chapter 2 EMPLOYMENT DYNAMICS AND TRANSITIONS Highlight of the main results • About half of the workers change their employment status over three years, moving in and out of employment, inactivity, underemployment and unemployment. Agricultural work is particularly volatile. • More than 70% of people moving in and out of inactivity are women, due to childbearing and lower productivity. • About 40% of unemployed workers find full employment after 3 years, reflecting their higher level of education and aspiration to protected jobs. • Multi-activity is common and adjusts in response to fluctuations in agricultural labor demand. Two thirds of the workers drop their secondary non-agricultural activities during the agricultural season. • Workers also move across non-agricultural sectors such as retail or transport. • Nearly one in three workers previously active in manufacturing shifted to agriculture during the 2014 harvest time, reflecting the casual nature and low productivity of many manufacturing activities. • In 2014, 11% of men (and very few women) migrated for work over the year preceding the survey. • Poorer workers have the most volatile jobs portfolios. Education and higher productivity allows workers to stabilize employment sectors over time. • Expenditure of Nigeriens with diverse jobs portfolios was 20-25% higher than expenditures of farmers in 2014. Expenditures of those working entirely outside of agriculture were about twice as high. • Moving out of agriculture is likelier for older people and is associated with short- term expenditure gains on the order of 6%. • The odds are steep against a transition out of farm work for Nigeriens whose fathers were farmers, one to twelve for women, and one to nine for men. 30 Introduction As in most low-income countries, many jobs in Niger are best thought of as portfolios of activities that may be carried out in different sectors of the economy, with different intensity, and perhaps in different places. Maintaining portfolios of activities can be a risk-management strategy. It can also be a way to adjust to changes in the productivity of different activities; for instance, changes in agriculture productivity with the harvest season or in the productivity of mining labor in response to price changes. This chapter seeks to look beyond static numbers of labor market indicators such as underemployment or sectoral employment shares, and to understand how workers in Niger adjust their job portfolios. Jobs portfolios in low income countries are fluid along different dimensions. Where risks and opportunities for productive work change over time, people adjust their jobs portfolios in various ways to build a livelihood. First, intensity changes – from full-time work, to underemployment, to idleness. Secondly, people carry out activities in different types of employment, and alternate between sectors, from self-employment in agriculture and outside of agriculture, to wage employment, and to diverse combinations of these as primary and secondary activities. Third, workers migrate, including temporarily, to seek out opportunities. Finally, over the long run, people make choices on whether to continue the activities their parents engaged in – especially in farming – or to rebalance toward other occupations. Shifts in jobs portfolios may be temporary – for instance, seasonal – or more durable. Fluidity has different sources, and manifests over different time horizons. In the short run, opportunities for productive work change within the year; with the harvest calendar, but also in less predictable ways through shocks, and changes in networks and information. In the medium and longer term, shifts in the economy – whether business cycles, structural changes, or shocks with a more prolonged impact – impact on the way workers can be productive. Due to differences in the timing of enumeration, we can say little about the permanence of observed shifts. However, our results likely reflect seasonal changes to a significant degree. The 2011 survey was collected between mid-July and mid-September, largely during the post-planting season in Niger’s different agro-climatic zones. By way of contrast, in 2014, data collection took place between early September and early November, mostly during the early harvest season. Because of this discrepancy in timing, our data does not provide us with many opportunities to separate seasonal changes from other effects. However, we can make two observations. One, because the two survey rounds were collected at very different times in the cropping calendar, it is a priori plausible that changes reflect to a reasonable degree of seasonal variation in jobs portfolios. The patterns we observe are consistent with this assumption. Two, we study inter- generational transition to look for longer-term structural changes. This analysis indicates a strong persistence of work in agriculture, in line with the lack of macroeconomic indications of structural transformation. Between the two survey rounds, many workers changed employment status. Transitions in employment status are very common, with only about half of the employed, the underemployed, and the inactive maintaining their status over the two survey rounds. Much of this churn is related to changes in employment status in agriculture; through shifts between full employment and underemployment among men, and, between inactivity and employment among women. These shifts are likely to be seasonal. Queueing among educated urban workers further contributes to fluidity in employment status. Workers also alternate between forms of employment, shift their primary and secondary activities, and move temporarily for work. Seasonality in agriculture was a powerful source of fluidity. Workers who in 2011 reported that they were active in agriculture as well as a non-farm activity were in 2014 – when the survey was taken at harvest time – more likely than not to have been employed only in agriculture. 31 Temporary specialization in agriculture came partly at the expense of a shift of nearly one in every three workers previously active in manufacturing to agriculture. While likely seasonal, this shift does speak to the casual nature and low productivity of many manufacturing activities in Niger. Economic activity also showed significant fluidity outside of the agriculture sector. For instance, a significant number of workers switched commerce, transport, hospitality, and trade. Temporary migration is part of jobs portfolios for men, but remains limited. In 2014, eleven percent of men (and very few women) reported having been away from their residence for work over the year preceding the survey. Jobs portfolios that were not limited to agriculture were associated with higher expenditures. However, it is poorer workers who have the most fluid portfolios. The poor are more likely to make any switch in activities. Workers without any formal education were more likely than others to have moved into the agricultural sector, and women were more likely than men to have set aside another primary activity for work in agriculture. While being active outside of agriculture correlates with higher expenditure, the short run effect of moving out of agriculture within the three years between our survey rounds is only weakly associated with expenditure gains. Box 2.1: Characteristics of the balanced panel The analysis undertaken in this chapter is entirely based on the balanced panel, which is composed of all individuals interviewed in the first survey wave in 2011 and re-surveyed in 2014. The total sample consists of 13,596 individual observations each year. The two household surveys were administrated at different times in the crop calendar, during the post- planting period in 2011 and the early harvest season in 2014. The 2011 survey was completed from July 18 to September 17, 2011, during the late planting season and post-planting season. The 2014 data collection started in the beginning of September and lasted through November 10, 2014, overlapping with much of the harvest season. Consequently, intertemporal comparisons are likely to reflect seasonal shifts to an important degree. The characteristics of the population in the balanced panel and in the two cross-sections are very similar. However, the analysis of the balanced panel does not use statistical weights so it does not correct for the oversampling of urban areas. This does not affect most of the results in this chapter which are based on individual transitions rather than structural descriptive. An assessment of intergeneration transitions in occupation shows no evidence of significant structural shifts. Structural change features prominently in the debate on growth in Africa (Go and Page, 2008; Page, 2012; McMillan and Rodrik, 2011; McMillan et al, 2014). Niger’s Plan for Economic and Social Development 2012-2015 (PDES) identifies economic diversification as one of the five main axes of the Poverty Reduction Strategy. It would be attractive to trace such transitions in our micro data – but the opportunities to do so are narrow. As noted, seasonal shifts are likely a strong force behind changes observed between the survey rounds; even if this were not the case, one could not draw strong conclusions as to the permanence of shifts that occur over the three years between the two survey rounds. The one feasible approach to studying longer-term shifts in our data is to analyze whether young workers enter into the same occupations as their parents. Under this approach, we find that the odds are against a transition out of farm work for both women (an odds ratio of twelve) and men in Niger (an odds ratio of nine) whose fathers were farmers. For men, the odds decrease somewhat with age, perhaps reflecting the importance of savings and social standing in diversifying activities. 32 Common transitions between employment statuses and sectors Nigeriens work in a labor market where job quality and welfare outcomes diverge sharply between agriculture and the non-agriculture sector. Chapter 1 has shed light on the large differences in welfare outcomes between those employed in agriculture and other sectors. Thus, median consumption expenditure of those active in agriculture only was 44 percent below that of Nigeriens in non-agriculture self- employment, and 63 percent below that of Nigeriens in wage employment.7 ( ) Job and worker characteristics between the agriculture and non-agriculture sectors also diverge in intuitive ways, with the former offering shorter hours, and more temporary employment to a younger and less educated workforce. Table 2.1: Mean consumption expenditure by economic activity Mean consumption by activity (constant CFA) Total Ag only Non-Ag Self Non- Ag & Wage Ag & Self Self & Ag Wage Wage 2011 213,704 178,453 269,898 425,876 183,678 195,531 418,004 2014 251,758 192,243 343,126 517,927 240,367 231,400 394,406 Mean consumption as a multiple of the mean in agriculture Non-Ag Self Non- Ag & Wage Ag & Self Self & Ag Wage Wage 2011 151% 239% 103% 110% 234% 2014 178% 269% 125% 120% 205% Table 2.2: Characteristics of the agricultural and non-agricultural sector, 2011-14 Agriculture Non-agriculture (self and wage employment) 2011 2014 2011 2014 Total 70.4 73.3 29.6 26.7 Individual characteristics: Female 41.8 43.8 49.3 43.7 Rural 92.5 92.9 29.4 21.3 7 This measure of well-being is our preferred measure as measurement errors are likely to be less severe than in alternative measures. Moreover, due to data inconsistency in the income variables across the two survey rounds we are unable to report meaningful estimates of income in 2014. 33 Age 28.5 30.1 37.0 38.7 Non-working age (<15 or >65) 33.5 32.8 10.4 8.1 No education 75.9 77.1 54.7 47.4 Primary education 20.2 17.4 21.0 22.9 Secondary first cycle 3.7 5.2 14.6 18.2 Secondary second cycle+ 0.2 0.3 9.7 11.6 Employment characteristics: Public 0 0 12.0 15.8 Informal 99.9 99.7 89.0 85.0 Permanent 25.9 19.5 83.7 70.4 Temporary (seasonal) 72.5 59.8 7.5 10.2 Health insurance 0.1 0.4 5.3 10.4 Hours worked per week 38.4 33.7 42.5 46.1 Secondary job 27.8 21.6 15.9 11.5 Employment vulnerability predominantly affects farmers, women and the youth Transitions in employment status are common. Transitions in employment status are very common among workers in Niger. As the transition matrix in Table 2.3: Labor market activity and transition matrixes, 2011-14 shows, only about half of the employed, underemployed, and inactive maintained their status over the two survey rounds. Conversely, as Error! Reference source not found. illustrates, one-third of all men and two-thirds of all women employed full-time in 2011 were either underemployed or inactive in 2014. But equally, nearly half of all men and 40 percent of women who were inactive in 2011 reported some level of activity in the following survey, and large shares of the underemployed were able to work longer hours in 2014. Some of the apparent movement between full-time employment and underemployment may be due to measurement error in hours worked. However, there is no doubt that to an important degree, the data reflects significant variation in how much time Nigeriens are able to work productively in their jobs. Figure 2.1: Selected transition figures by gender Men Women 34 Table 2.3: Labor market activity and transition matrixes, 2011-14 Active (employed) Unemployed Out of labor force Active (employed) 81.6 0.3 18.1 Unemployed 56.9 6.9 36.2 Out of labor force 32.1 0.5 67.4 Employed Underemployed Unemployed Inactive Student Employed 54.2 32.0 0.4 11.0 2.5 Underemployed 26.9 49.9 0.2 18.2 4.8 Unemployed 39.7 17.2 6.9 31.0 5.2 Inactive 12.6 29.0 0.5 45.5 12.4 Student 6.3 13.8 0.5 11.7 67.7 Alternation between full employment, underemployment, and inactivity is in particular a reality for those working in agriculture. Error! Reference source not found. shows the characteristics of workers changing employment status. Those alternating between employment and underemployment are overwhelmingly men living in rural areas, employed in the agricultural sector, and with no formal education – suggesting that variation in hours worked is likely related to seasonal changes in labor productivity and demand. Rural women without formal education are the workers most likely to shift between (under-) employment in agriculture and inactivity. Figure 2.1 shows how women are much more likely than men to decrease their labor force participation over time. Between both survey rounds, men overall experienced an increase in the intensity of employment, with more transitions from inactivity to employment and from underemployment to full employment, in line with the greater intensity of agricultural work during the harvest season. However, women were more likely to experience a decrease in labor force participation, with large numbers shifting to inactivity or from full employment to underemployment (Figure 2.2). This reflects the fact that women go through repeated periods of childbearing (with a fertility rate of 7.6 children per woman), are responsible for the vast majority of household work, and are the first ones to drop out of the labor force when there is a contraction in labor demands. These transitions clearly illustrate how much more precarious employment is for women than for men. 35 Figure 2.2: Women move in and out of inactivity 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% (under)employment → inactivity → inactivity (under)employment Female Male Figure 2.3: Characteristics of movers from unemployment to employment Gender Locality Sector reached 1 1 1 13% 22% 17% 2 2 3 26% 2 61% 83% 78% Transition patterns show that of the few workers who are unemployed, many are educated young urban men queueing for good jobs, who do not stay out of work for long. Of the very small number of unemployed workers in the 2011 sample, only seven percent reported that they remained unemployed in 2014, while 57 percent and 36 percent transitioned into the labor market and out of the labor force, respectively (Table 2.3: Labor market activity and transition matrixes, 2011-14). Many of those moving from unemployment to employment are men in urban areas who are between 25 and 34 years old, and have at least primary education. A large share of them enter non-agriculture wage employment, and a smaller share, self-employment outside of agriculture: their unemployment spell can therefore most usefully be 36 interpreted as queueing for good jobs. Similarly, former students encounter inactivity and underemployment upon completing their studies, but it is a temporary stage in their transition into the labor market. Among respondents who completed their studies between the survey rounds, 43 percent become underemployed and 36 percent inactive, while 20 percent found full-time employment. Taken together, temporary unemployment, underemployment and inactivity among educated workers speak to the scarcity of desirable jobs (and perhaps, to a skill mismatch). Evolutions in number of jobs and sectors of employment Multi-activity is common in Niger, with on average 21 percent of workers having a secondary occupation. Table 2.4: Economic activities and activity transition matrix, 2011-14 shows how respondents transitioned between different activities between survey rounds, allowing for different combinations of primary and secondary activities. Seasonality in agriculture is a powerful source of fluidity in employment type and sector of activity. The 2014 survey, taken during the harvest season, reflects a shift into agriculture as the sole activity. While we cannot be certain that this shift is seasonal, it is highly plausibly to interpret it as such, given the timing of the surveys and relatively short time elapsed between the two rounds. The share of individuals engaged only in farm work at survey time increased by more than seven percentage points. Workers who in the 2011 survey had reported that they combined agriculture work with self- employment outside the sector or with wage work were more likely than not to report that they now only worked in agriculture. A smaller share of those who previously had only been active in non-agricultural self-employment also switched into agriculture, as did more than one in every five previously inactive respondents. By implication, the harvest season is a time of less diversified portfolios of activities: in 2011, around 8 percent combined agricultural activities with wage employment. This share fell to less than 3 percent in 2014. As shown in the diagonal of Table 2.4: Economic activities and activity transition matrix, 2011-14, the largest share of non-changers is either inactive or specialized agricultural workers. We have discussed above the movements in and out of inactivity, in particular among women, which are reflected in the large transition flows between non-agricultural self-employment and inactivity. Even though wage employment is generally more stable, transition patterns point to precariousness and vulnerability in the wage sector too. Non-agricultural wage workers are more likely to have become self-employed between the panel rounds (21 percent) than the other way around (4 percent), a transition that mostly affects men living in urban areas with no education. Similarly, those engaged in multiple activities including wage employment are more likely to give up their wage job overtime than to keep it. Table 2.4: Economic activities and activity transition matrix, 2011-14 2014 Ag Non Non- Ag & Ag & Self & No only -Ag Ag Wage Self Wage activity Self Wag only e only 2011 34.7 11.4 4.5 8.1 1.3 0.3 39.8 2014 41.9 10.6 4.4 2.7 2 0.4 38 Transition matrix Ag only 72.5 2.7 0.4 2.5 2.7 0 19.2 201 1 Non-Ag Self 13.4 48.3 3.5 5.4 1.3 1.3 26.9 37 Non-Ag 3.3 21 57.7 1 2.6 1.8 12.7 Wage Ag & Wage 64.5 7.8 0.5 11.1 8 0.1 8 Ag & Self 61.9 5.2 9.8 5.2 7.5 0.6 9.8 Self & Wage 4.9 29.3 39 2.4 0 17.1 7.3 No activity 22.7 6.2 2.4 0.6 0.3 0.2 67.6 Except from a few protected activities, there are sizeable movements between sectors. Table 2.5: Primary economic activity and transition matrix by sector, 2011-14 reflects sectoral shifts in primary activities only. First, the important role of seasonal work in agriculture is apparent. The share of workers employed in agriculture increased by three percentage points (the discrepancy to the shift observed in Table 2.4: Economic activities and activity transition matrix, 2011-14 is due to the fact that the latter accounted for the movement of the inactive into agriculture). Secondly, by way of contrast, the share of workers employed in manufacturing declined from 2011 to 2014, by three percentage points. Of those employed in manufacturing in 2011, 30 percent shifted their activities to agriculture. While these shifts are most likely seasonal, rather than reflecting an adverse structural change, they do speak to the casual nature and low productivity of many manufacturing activities in Niger. Thirdly, it is striking how low are the observed probabilities of workers remaining in any of their initial primary activities. As the diagonal elements in the transition matrix show, workers are often about as likely as not to switch activities. In addition to moving into agriculture, there is also a significant amount of switching among service categories, transport, hospitality, and trade. The only relatively stable occupations beyond agriculture appear to be the extractives industry, and the education sector, both of which appear to be the main providers of stable, formal jobs. Table 2.5: Primary economic activity and transition matrix by sector, 2011-14 2014 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Agricultu Transpor Extractiv Activities Construc Wholesal Administ Hospitali Educatio Manufac Personal Finance/ Services Services Utilities e/Retail Health Public turing Estate ration Other Real tion re ty n e t 2011 70 0 9 0 1 8 2 1 0 1 1 1 2 3 2014 73 0 5 0 1 9 1 2 0 1 2 1 2 2 Transition matrix 1 Agriculture 94 0 1 0 0 2 0 0 0 0 0 0 1 0 2 Extractive Activities 15 77 8 0 0 0 0 0 0 0 0 0 0 0 3 Manufacturing 30 0 35 0 1 19 4 1 0 1 0 0 6 2 4 Utilities 0 14 0 43 0 0 0 0 0 29 0 0 14 0 2011 5 Construction 18 2 9 0 48 2 0 6 0 3 0 0 5 9 6 Wholesale/Retail 26 0 7 0 1 56 3 3 0 0 0 1 2 1 7 Hospitality 32 0 7 0 0 31 26 0 0 1 1 0 1 2 8 Transport 10 1 1 0 2 8 0 57 1 3 1 3 5 9 9 Finance/Real Estate 13 0 0 0 0 0 0 0 31 25 0 0 6 25 38 1 Public 7 0 0 1 2 5 1 4 4 55 11 5 1 5 0 Administration 1 Education 5 1 0 1 0 2 0 0 0 8 83 0 0 0 1 1 Health 6 2 2 0 0 2 0 0 2 6 2 58 14 6 2 1 Personal Services 22 0 7 0 2 13 1 4 0 1 4 3 36 7 3 1 Other Services 7 2 5 2 3 9 1 7 1 4 1 2 11 44 4 The dynamics of diversification Diversification is associated with better welfare outcomes Jobs portfolios that go beyond specialization in agriculture are associated with higher expenditure. In the economy at large, diversification away from agriculture into more productive work in other sectors underpins structural transformation. In the panel sample, it is associated with higher expenditure. We noted above that median expenditure is higher among wage workers and those self-employed outside of agriculture than in other occupational groups. More importantly, Error! Reference source not found. shows that expenditure is also meaningfully higher among those who are able to combine work in agriculture with other activities. In this sense, a more diverse jobs portfolio is associated with better welfare outcomes than agriculture only and maintaining a diversified portfolio during the peak agriculture season is a correlate of better welfare outcomes. Diversification is associated with higher expenditure in the short run. The bivariate regression in Column 1 in Error! Reference source not found. shows that expenditure levels were some 13 percent higher, compared to agriculture, among workers who diversified or moved out of agriculture within the three years between the survey rounds had expenditure levels in 2014. This observed correlation is consistent both with wealthier workers moving out of agriculture, or an effect of diversification on expenditure. Column 2 adds to the regression lagged expenditure and individual characteristics by way of accounting for selection of wealthier workers into work outside of agriculture; the estimated effect suggests that for a given level of expenditure in 2011, a switch out of agriculture is associated with weakly higher expenditure in 2014. The individual fixed effects model in Column (3) similarly shows weakly higher short- term income among those active outside of agriculture. In balance, the regression results reflect that diversification is associated with previous wealth (as seen above), and suggest that the short-run effects of diversification are weakly positive. This is consistent with a labor market where advancement is slow and hard-won: diversification is open to those who have already been able to build better livelihoods, and while it may lead to further improvement if welfare, the effect is not overwhelmingly strong. Table 2.6: Impact of diversification out of agriculture on welfare (1) (2) (3) Transition out of agriculture 0.126** 0.063 (0.042) (0.033) Non-Ag Self 0.038 (0.054) 39 Non-Ag Wage 0.005 (0.046) Ag & Wage 0.118** (0.037) Ag & Self 0.032 (0.093) Self & Wage 0.021 (0.075) Inactivity 0.015 (0.038) Lag log(real consumption) 0.584*** (0.044) Individual fixed effects No No Yes Lagged individual characteristics No Yes No Observations 13383 13383 13383 Note: OLS. Dependent variable is log of real consumption expenditures per capita. Standard errors are clustered at the regional level. Individual characteristics include gender, age group, schooling level, literacy, geographical location, person is head, person is spouse, married, and ethnicity dummies. The models in Columns (1) and (2) include a full set of indicator variables for the activities of non-transition individuals. ***, ** and * indicates significance at the 1%, 5% and 10% level, respectively. Whose portfolios shrink when agricultural labor demand peaks? Maintaining diversified portfolios is desirable for income smoothing and resilience, but is not possible for all. To shed light on the characteristics of workers who shift – presumably seasonally – into agriculture, Error! Reference source not found. reports on the probability of transitioning into agriculture for different sub-groups of workers, disaggregating across gender, location and education level. Each column corresponds to results for a sub-group. Each row corresponds to a particular occupation in 2011, and reports the share of workers who switched into agriculture in 2014 (for instance, the second row reports the share of those initially in self-employment outside of agriculture who were working in agriculture in 2014). Women, those without formal education and rural workers are the first to shed a secondary occupation to focus exclusively on agriculture. While gender differences are generally low, women employed outside the agricultural sector are three times more likely than men to shift entirely to agriculture (11 percent as opposed to 3 percent). Secondly, those without formal education were much more likely than those with some schooling to focus on agriculture rather than maintaining a secondary activity in self- employment. These patterns are consistent with lower productivity of non-agriculture activities among the less educated and women than among other groups; indeed, while most ‘inactive’ women undoubtedly lead busy lives of caring for their families, the observed shifts into and out of the labor force may also speak to a lack of any alternative productive activities for some Nigeriens. Finally, while it is trivial that shifts are greatest in rural areas, it is instructive to consider that twenty to thirty percent of those maintaining diversified portfolios in urban areas also focus exclusively on agriculture during the busy season. Table 2.7: Share of individuals shifting to exclusive agricultural work, by categories 40 All Men Wome No Some Rura Urba n educatio educatio l n n n Occupation Ag & Wage 64.5 65.1 63.5 65.0 62.6 68.1 29.0 s of origin Ag & Self 61.9 62.0 60.0 73.7 39.2 77.2 19.6 Self & Wage 4.9 3.1 11.1 6.7 4.2 20.0 2.8 Poverty is associated with volatility of employment portfolios. Table 2.8 enriches the assessment of sectoral moves with a regression analysis, and more formally explores the characteristics of individuals that transition into and out of agriculture. The dependent variable in column (1) takes the value one for respondents who transitioned into specialized agriculture between the two surveys, and zero otherwise. Column (2) models the probability of any transition out of specialized agriculture, and Columns (3)-(5), specific transitions. The first message is clear: poorer workers are more likely to switch in any way. The coefficient on log expenditure is not always significant, but it is consistently negative with relatively high magnitude. While diversified portfolios are associated with better welfare outcomes, volatility of employment status is not. The youth are more likely to drop diversified portfolios to focus on agriculture. The age coefficient in the first column is highly significant and negative. Being married, on the other hand, is highly correlated with more diversified portfolios in the agricultural season, which reflects how larger economic units are able to engage in multiple activities to which all members might end up participating. Table 2.8: Determinants of the transition in and out of agriculture (1) (2) (3) (4) (5) (6) Transitioned Transitioned Move Diversified Diversified Became into out of into Non- to Ag & to Ag & inactive agriculture agriculture Ag Self. Wage Self Lag -0.218*** -0.128 -0.165** -0.209** -0.075 -0.087 log(consumption) (0.068) (0.117) (0.079) (0.092) (0.080) (0.122) Gender -0.073 0.049 0.302*** 0.225 0.586*** -0.073 (0.083) (0.107) (0.080) (0.141) (0.128) (0.122) Lag age -0.006*** 0.004*** 0.007** 0.003 -0.003** 0.006*** (0.001) (0.001) (0.003) (0.003) (0.002) (0.002) Lag any education -0.094 -0.087 -0.084 0.088 -0.336*** -0.069 (0.090) (0.105) (0.150) (0.137) (0.072) (0.133) Lag literate -0.036 -0.024 0.010 0.002 0.057 -0.059 (0.062) (0.099) (0.092) (0.094) (0.063) (0.119) Lag urban -0.933*** -0.507** 0.090 -0.317 -0.306 - 0.605*** (0.229) (0.202) (0.147) (0.224) (0.197) (0.201) Lag household head 0.303*** -0.215 -0.050 0.094 0.072 - 0.601*** (0.097) (0.154) (0.231) (0.079) (0.126) (0.148) Lag household -0.007 0.009 0.187 -0.024 0.172 0.025 41 spouse (0.072) (0.148) (0.198) (0.109) (0.169) (0.158) Lag married -0.012 -0.006 0.151 0.482*** 0.786*** -0.250 (0.084) (0.185) (0.215) (0.153) (0.176) (0.198) Ethnic group controls Y Y Y Y Y Y Observations 13,383 13,383 13,383 13,383 13,383 13,383 Pseudo R-squared 0.107 0.050 0.056 0.092 0.128 0.074 Note: Probit. Standard errors are clustered at the regional level. ***, ** and * indicates significance at the 1%, 5% and 10% level, respectively. Temporary migration is part of jobs portfolios for men About ten percent of men leave home for work in each survey year; very few women do. Among all panel respondents, five percent reported having been away from their usual residence for work-related reasons within the twelve months preceding the 2011 survey. (Error! Reference source not found.) In 2014, the share of work migrants had increased to six percent of the survey respondents, an 18 percent increase in relative terms. The median working migrant reported having been absent for 60 days a year, consistent with seasonal work and other short forays. Error! Reference source not found. shows characteristics of work migrants across the two survey years. Men, unsurprisingly, account for nine out of every ten work migrants. Migration appears equally distributed between urban and rural households: two-thirds of the working migrants lived in rural areas in both survey periods, in line with the overall share of rural respondents. Table 2.9: Absence from residence 2011 2014 Freq. Percen Freq. Percen t t Work-related 688 5% 813 6% Seasonal work 332 2% 319 2% Travel for work 288 2% 407 3% Trip with animals to search 48 0% 77 1% pasture/water Temporarily called to help a 20 0% 10 0% household Not related to work 2,525 19% 2,507 18% Other family reason/vacation 1,258 9% 1,053 8% Attended a ceremony 805 6% 965 7% Other 227 2% 195 1% Health reasons 129 1% 164 1% Pilgrimage 78 1% 80 1% Attending school 21 0% 47 0% Military service 7 0% 3 0% 42 Total 3,213 24% 3,320 24% Table 2.10: Characteristics of working migrants 2011 2014 Work migrants Work migrants Male 91.3 91.2 Rural 66.1 66.7 Agadez 3.7 11.2 Diffa 16.6 6.6 Dosso 11 8.1 Maradi 16.5 11.6 Tahoua 8.2 12.7 Tillaberi 7.9 13.6 Zinder 17.3 21.2 Niamey 18.9 15 Intergenerational rigidity restricts employment dynamics. An assessment of intergenerational occupational mobility offers some perspective of more durable shifts in jobs portfolios. One angle on whether changes in portfolios are seasonal or more durable is to consider inter-generational mobility. In Error! Reference source not found., we assess whether respondents to the 2014 survey followed their parents into work in the agriculture sector.8 The odds ratios shown describe the probability that a respondent works in agriculture as opposed to another sector, given that their father or mother works in agriculture. There is little indication of an inter-generational shift out of agriculture. Fundamentally, the odds are very much against a Nigerien whose parents were farmers changing occupation – with odds ratios of nine to one for men, and twelve to one for women whose fathers were farmers. This is consistent with the continued dominance of agriculture in the economy. Still, it is noteworthy that there is such stasis in occupational choice despite the significant intergenerational educational mobility documented in Chapter 1. 8 Given the dominance of agriculture in the economy, we follow Bossuroy and Cogneau (2013) in considering only occupational mobility between agriculture and the non-agricultural sector. 43 Older men are somewhat more likely to be able to shift out of their family’s activity in agriculture. Among men, there is a pronounced gradient of odds ratios in age: younger men are much more likely to join their parents in working in agriculture than older men. The odds against change among young men are very steep, at 36 to one among men younger than 25, and 13 to one for men aged 25-34. Among men aged 35 and older, they are much lower, at eight or nine to one (though they remain steep, and are not significantly different, individually, from those among younger cohorts). This pattern mirrors the finding shown above that older men are more likely to have diversified their activities between our survey rounds. It is consistent with an economy with little access to finance and low market integration, where diversification into new activities requires a potentially very long process of saving and building relationships and reputations. Results for women are harder to interpret, given that women shift in and out of inactivity with marriage and childbearing. Overall, the odds that a woman whose father was active in agriculture is able to work in another sector are worse than for men. Table 2.11: Intergenerational reproduction of occupations: farm/non-farm odds-ratios Father's occupation Mother's occupation Son Daughter Daughter All 9.1 11.9 6.7 [7.20;11.63 ] [9.33; 15.19] [5.47; 8.17] Working-age (age 15-64) 9.1 11.9 6.7 [7.20; 11.63] [9.32; 15.19] [5.47; 8.17] Age 15-24 35.9 6.0 4.9 [8.68; 169.87] [2.54; 13.95] [2.35; 10.29] Age 25-34 13.0 15.3 5.7 [7.70; 22.17] [10.04; 23.35] [4.04; 8.04] Age 35-44 8.7 16.8 8.7 [5.63; 13.64] [9.90; 29.18] [5.76; 13.26] Age 45-54 8.3 10.6 6.3 [5.13; 13.69] [5.93; 19.44] [3.98; 10.11] Niger’s intergenerational occupational mobility corresponds to other low-income African countries. Error! Reference source not found. show simple comparisons of odds-ratios for the sample of men aged 20-69 between different developing countries across the globe. Niger share a similar level of intergenerational reproduction of occupations as other low-income African countries did in the 1990’s, though higher than Uganda and Ghana. Figure 2.4: Cross-country comparisons of farm/non-farm odds-ratios, men aged 20-69 44 35 30 25 20 15 10 5 0 Note: For India the sample includes a representative sample of male electorate. Source: Own calculations for Niger. The additional country results are from other authors’ results. Brazil: direct computation from PNAD 1996 survey (see also Cogneau and Gignoux, 2008); China: from table 3 in Wu and Treiman (2006); India: from tables 2 and 3 in Kumar et al. (2002a, 2002b); South Africa: direct computation References from the NIDS 2008 survey (Wave 1); Madagascar, Uganda, Ghana, Guinea, Côte d’Ivoire are from table 5 in Bossuroy and Cogneau (2013). Banerjee, A. V. and E. Duflo (2007). The economic lives of the poor. Journal of Economic Perspectives 21 (1), 141–167. 1. de Vries, G. M. Timmer and K. de Vries (2015) “Structural Transformation in Africa: Static Gains, Dynamic Losses�, The Journal of Development Studies, 51, 674-688. Go, D. S. and J. Page (eds) (2008) “Africa at a Turning Point? Growth, Aid and External Shocks�, Washington, D.C.: The World Bank Kuznets, S. (1955) “Economic Growth and Income Inequality�, American Economic Review, 45, 1–28. Lewis, W. A. (1954) “Economic Development with Unlimited Supplies of Labour�, Manchester School, 22, 139–91 McMillan, M. S. and D. Rodrik (2011) “Globalization, structural change and productivity growth�, Cambridge MA, National Bureau of Economic Research, NBER Working Paper No. w17143. McMillan, M.S., D. Rodrik, and I. Verduzco-Gallo (2014) �Globalization, Structural Change, and Productivity Growth, with an Update on Africa�, World Development, 63, 11-32. Page, J. (2012) “Can Africa industrialise?�, Journal of African Economies, 21, 86–124. 45 UNU-WIDER (2015) “Growth, Structural Transformation and rural Change in VietNam: A rising Dragon on the Move�. Forthcoming as a book. 46 Chapter 3 AGRICULTURAL EMPLOYMENT Highlight of the main results • 91% of all households have at least one member engaged in agricultural activity, including 46% in urban areas. • Farm work is poorly diversified, with more than 85% of agricultural employment dedicated to four crops (cowpeas, millet, sorghum and peanuts). • 74% of farm households who do not commercialize any crop, and a minimal share of the agricultural labor force is involved in market-oriented crops. • Agricultural employment is intense but highly dependent on the season and on the large fluctuations of food prices. • Farms are small. Almost 50% of the agricultural labor force works on less than one hectare of land per worker, and 85 percent have less than 2 hectares of cropping land per labor unit. • Almost 90% of farm households own their plots, but only 12 percent have at least one titled plot. • Irrigation is used by around 10% of farmers, and chemical agricultural inputs are used by less than one in five farmers. When used, they are applied on small plots. • Agricultural employment is plagued by poor access to goods and capital markets. Only about 10% of villages have a permanent market or a credit institution. • One third of agricultural households diversify activities across farm and non-farm sectors. Diversification at the household level is associated with an 8% increase in consumption. • Agricultural productivity and non-agricultural activities are complementary rather than substitute. Education and access to inputs foster diversification and raise welfare. • Village infrastructures and specific crops such as onions, cowpeas and sesame provide significantly higher incomes. 47 A nation of farmers Nigerien farmers work in five different agro-ecological zones with different environmental characteristics and livelihoods patterns (Box 3.1). Geographical and climatic conditions, availability of land, soil quality and access to irrigation vary across zones. The zones include urban, consisting of Niamey and other urban areas, agricultural, agro-pastoral and pastoral regions. The total population (based on survey respondents and their household sampling weight, thus comparable to 2011 demographic census data) of the four zones and the distribution of plots and agricultural workers across the zones are shown in Figure 3.1. The available farm area is highest in the agricultural region, followed by agro-pastoral, pastoral and urban zones. Figure 3.1: Distribution of working age population, total and working age agricultural workers (primary or secondary activity), and farm size by agro-ecological zones 8,000,000 7,000,000 6,000,000 5,000,000 4,000,000 3,000,000 2,000,000 1,000,000 0 Urban Agricultural Agro-pastoral Pastoral Working age population (16-65) Agricultural workers (main or second activity) Working age ag workers Plots (total farm size) in Ha 48 Box 3.1: Niger Livelihood Zone Map [Source: USAID/FEWS NET] In the absence of a map that precisely outlines the 5 agro-ecological zones used in the household survey data analyzed in this chapter, we refer to the above map and explain how livelihood are aggregated for our purpose. The valley areas of zone NE02 (Agadez region) promote some irrigated vegetable gardening, but in the absence of sufficient rainfall, has primarily pastoral lands. Livestock rearing is the primary form of rural livelihood in zones NE03, NE06 and NE14, which makes them the primary pastoral belt of the country. These zones constitute parts of Agadez, Diffa, Tahoua, Tillaberi, Zinder and Maradi regions. Zone NE04 is the primary agro-pastoral belt of Niger, and is situated between the densely populated rain-fed agricultural zone and the sparsely populated southern limits of pastoral zone. Parts of Diffa, Dosso, Maradi, Tahoua, Tillaberi and Zinder regions constitute this zone. Zones NE05, NE07, NE08, NE09, NE10, NE11, NE12 and NE 13 constitute the densely populated major agricultural belt of the country, and includes rain-fed cultivation, Niger River and Komadougou River irrigated lands and Lake Chad flood-retreat cultivation. These zones constitute parts of Diffa, Dosso, Maradi, Tahoua, Tillaberi and Zinder regions. Agriculture and livestock are the primary source of employment in Niger, with almost 80 percent of the workforce engaged in the sector as a primary occupation and nearly 5 percent as secondary occupation. As shown in Figure 3.2, among active Nigeriens, 74 percent are engaged in agriculture (excluding livestock) as the main occupation, and 4 percent as secondary occupation. Livestock, hunting, forestry and fisheries is the primary and secondary occupation for 6 percent and 1 percent of the active population respectively. Livestock jobs are more prevalent in pastoral areas at almost 12 percent. Interestingly, agriculture is the secondary occupation for 7 percent of the urban workforce. 49 Figure 3.2: Share of individuals working in agriculture or livestock 87% 80% 81% 74% 20% 6% 2% 6% 12% 7% 7% 4% 4% 2% 1% 1% 2% 1% 0% 0% National Urban Agricultural Agro-pastoral Pastoral Primary = agriculture Primary = livestock/fishing/hunting Primary = non ag, secondary = agriculture Primary=non ag, secondary = livestock/fishing/hunting The importance of agriculture is even more striking when considered at the household level. More than 91 percent of all households have at least one member engaged in agriculture or livestock at the national level. Numbers are over 95 percent in rural areas. In urban areas, almost half of all households have at least one member engaged in agricultural or livestock activities. Figure 3.3: Share of households with at least one member engaged in agricultural or livestock jobs 98.9% 99.4% 96.5% 91.4% 97.2% 98.3% 95.3% 89.5% 46.2% 35.2% National Urban Agricultural Agro-pastoral Pastoral Agriculture: Primary or Secondary Agriculture or Livestock: Primary or Secondary All age groups and both genders are highly involved in agriculture, with a slightly higher share for men and younger workers, including children. Men are more engaged in agriculture than women but both genders are equally engaged in livestock activities. In agriculture, the female labor share is at 68 percent compared with over 77 percent of men, as shown in figure 3.4. Around 6 percent of men and women alike work on livestock raising. The agricultural sector accounts for 82 percent of all child and adolescents’ 50 labor (persons under 15). The share of agricultural jobs is also a bit lower among the senior (above 65) at 63 percent. Livestock employs 10 percent of all child labor and 6 percent of older workforce. Child labor is more concentrated on agricultural and livestock activities since those are mostly low-skill family- operated activities. Figure 3.4: Share of agricultural and livestock jobs (as percentage of active population’s primary occupation) in the workforce across gender and age cohorts 6% 4% 6% 4% 4% 78% 68% 74% 66% 69% Men Women 15-30 30-50 50-65 Gender Working age cohorts Primary = agriculture Primary = livestock/fishing/hunting Characteristics of Agricultural employment A poorly diversified agricultural production focused on subsistence Niger agriculture is poorly diversified with four crops (cowpeas, millet, sorghum and peanuts) making up the bulk of overall production. As shown in Figure 3.5, those four crops represent 85 percent of the total labor involved in agricultural activities, and about 93 percent of the total acreage of agricultural land. Sesame and rice are the next more important crops, but represent a very small fraction compared to the big four. Most of agricultural work focuses on crops with very small commercialization rates, and a minimal share of workers are involved in market-oriented crops. Millet and sorghum are pure subsistence crops, with commercialization rates under 1 percent. Cowpeas and peanuts offer some market opportunities, with 15 to 18 percent of production sold on the market. Among the four major crops, those that are somewhat more marketed such as cowpeas and peanuts are generally more labor-intensive (and likely profitable) with their labor share significantly higher than their acreage share, while the reverse applies to the non-marketed subsistence crops such as millet and sorghum. Sesame, the fifth biggest production in terms of labor involved, has a much higher commercialization rate (37 percent). Only vegetables or tubers, including onions, have market participation shares that are significantly above 50 percent, but taken together they represent only a few percent of the agricultural labor. 51 Figure 3.5: Acreage share, labor share, and market participation rate by crop 80% 73% Labor share 70% Acreage share 60% Commercialization rate 50% 38% 38% 37% 40% 36% 30% 27% 24% 20% 18% 20% 15% 15% 12% 10% 7% 4% 4%3% 1% 1% 2%0% 1%0% 1%0% 0% cowpeas Millet Peanuts Sorghum Sesame Paddy rice Onion Squash Overall, market participation for agricultural workers is very low, especially in pastoral regions. Nationally, just above a quarter of all agricultural households commercialize at least one of their crop (Figure 3.6). The other three quarters produce only for self-consumption. Probability of market participation of smallholders depends on localization and agro-ecological areas. The most productive and possibly surplus-producing agricultural regions are also those where market participation rates are higher (for at least one crop) and concern one third of farming households, while rates are lower in agro-pastoral areas and marginal in pastoral areas (5 percent). This highlights the issue of being connected to local and regional markets and to production and productivity potential being still highly vulnerable to agro-ecology. Figure 3.6: Percentage of farm households who commercialize at least one of their crop 100.0% 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 31.9% 30.0% 26.5% 24.1% 20.0% 13.1% 10.0% 5.3% 0.0% National Urban Agricultural Agropastorale Pastoral Agricultural work is highly seasonal In a mostly rain-fed agriculture, labor is highly dependent on the season. Only 13.2 percent of agricultural work is permanent, in line with the share of irrigated agriculture (see next section). The rest of the work is 52 seasonal, or only limited to a specific period, which means that it does not even take place every year like seasonal work does (Figure 3.7). Figure 3.7: Share of agricultural employment by type of seasonality Permane nt, 13.2 Seasonal, For a 53.0 specific period, 33.8 Strikingly, agricultural work is very intense during the main season, which points to its low levels of productivity. As Figure 3.8 shows, a vast majority of individuals are employed, albeit only seasonally, and the number of working hours is high, especially in pastoral areas where less fertile soils but larger plots make work less productive, and because of the necessary combination of both agricultural and livestock activities. Figure 3.8: Occupational rate of the working age population and average number of worked hours/week for the workers 84.5 84.5 77.6 59.3 55.8 48 48 49.7 Urban Agricultural Agropastoral Pastoral Employment rate of the working age population in pct Weekly hours of the workers Note: Employment status has been defined as positive as long as a minimum of 10 hours of work have been undertaken by a single individual. Working age population is defined between 15 and 60 years old. 53 When commercialized, agricultural production is subject to the volatility of prices on local food markets, causing significant risks to the profitability of agricultural work. Average millet monthly prices varying by 15 percent on average on the long term, and month-on-month average price change of 6 percent (Figure 3.9). As the boxplot in Figure 3.10 shows, Agadez and Tillaberi have the highest millet prices, while Zinder and Agadez are the most volatile regional markets, owing to their remoteness and logistical costs as being more distant from surplus-producing regions. Other market analyses (GIEWS 2017) performed for the two other most significant staples which are sorghum and cowpeas also show significant and comparable volatilities. Imported commodities which are important for consumers such as maize and rice (competing with local crops) are less volatile. Figure 3.9: History of regional locally-produced millet markets Local millet wholesale nominal prices, CFAF/100 Kg 40000 35000 30000 25000 20000 15000 10000 5000 0 Jul-08 Jul-13 Jun-06 Apr-07 May-09 Mar-10 Jun-11 Apr-12 May-14 Jun-16 Apr-17 Nov-06 Nov-11 Nov-16 Jan-06 Sep-07 Feb-08 Dec-08 Jan-11 Mar-15 Jan-16 Aug-10 Sep-12 Feb-13 Dec-13 Aug-15 Oct-09 Oct-14 Agadez Dosso Maradi Niamey Tillaberi Zinder Figure 3.10: Box plot of monthly millet price distribution by regional marketplace Source : FAO Global Information and Early Warning System (2017) 54 Constraints to Agricultural employment The workforce is concentrated on small plots Cultivated land size per family in Niger tends to be small. As shown in Figure 3.11, the median farm size is 5.2 hectares per household while the average farm is 7.5 hectares. While this is standard for smallholder’s agriculture in the region (see for example Kaminski and Thomas 2011 for Burkina Faso), it tends to be higher than other Living Standards Measurement Study - Integrated Surveys on Agriculture countries (see Deininger 2016) where land can be more concentrated by some estates with much less left to smallholders (especially in Nigeria and Malawi) but with more agricultural wage jobs opportunities. But in terms of income generating potential and food self-sufficiency, this number is still low in absolute terms, also when compared with other agricultural economies. Almost 50 percent of the agricultural labor force works on less than one hectare of land per worker, while 85 percent have less than 2 hectares of cropping land per labor unit. Figure 3.11: Distribution of agricultural workers across land size classes (by land available and household net farm size – no double counts with intercropping) Distribution of nationally-representative number of agricultural workers across classes of available land 41.1% 30.7% 32.1% 27.6% 22.5% 13.1% 13.6% 5.1% 4.0% 4.3% 3.6% 2.2% [0-1] [1-2] [2-3] [3-4] [4-5] [5-10] >=10 [0-2] [2-5] [5-10] [10-20] >=20 Classes of available land / ag worker in Ha Classes of household farm size in Ha Subsistence crops take up the largest plots, but for any crop there is a large concentration of very small plots. The median plot size is around 1 Ha but varies between 0.5 for the irrigated/counter-season and small crops (rice, onion, squash, peanuts) and 1.5 Ha for the biggest field crops (sorghum, millet, cowpeas), according to Figure 3.12. This is rather standard for smallholder agriculture in Africa and in those agro- ecological areas. Of equal if not more importance, there is sizable variation in plot sizes across households for each crop, most likely reflecting differences across agro-ecological zones. About 50 percent of plots are smaller than 1.5 Ha for the main subsistence crops. 55 Figure 3.12: Distribution of plot sizes by crop for the main 8 crops in Ha Note: The above chart is a box plot of plot size distributions by crop, which shows the median value in the median line of each box, the average value displayed with a cross, the values of the bottom and top box limits correspond to the first and third quartile values (25 and 75 percentiles), and the bottom and top extreme limits to the 5 and 95 percentiles. Outlier values are not displayed to keep this chart self-contained. Note also that plot sizes do not account for whether plots are mono or multi-cropped. Millet and cowpeas which are among the top 3 crops are often grown together or inter-cropped on same plots of same size. Onions or peanuts are more often mono-cropped though. That relativizes the differences shown above. Farm size also varies across regions, but the amount of available land per worker is quite stable, from about 3 hectares in urban areas to 8 hectares in agro-pastoral areas, as shown in Figure 3.13. Households in both agricultural and pastoral areas farm own or operate just over 5.5 to 6.5 hectares of land, which is closer to the national average. Interestingly, differences across regions are smaller when the available land per worker is considered, as the size of households and the number of workers employed are also higher where farms are bigger. 56 Figure 3.13: Average net farm size and available land per worker (in hectares) by agro-ecological zones 8.10 6.09 6.37 5.62 3.15 2.51 1.96 2.11 1.60 1.75 National Urban Agricultural Agro-pastoral Pastoral Farm size Available land per ag worker Almost 90 percent of farm households own their plots, but only 12 percent have at least one titled plot (formal or informal certificate or any other document). Figure 3.14 shows that land ownership is more difficult to access in urban areas (69 percent) compared with other zones, where households own over 90 percent of their plots. However, titling is slightly more common in urban areas with 15 percent households owning at least one titled plot. About 12 percent of households in the agricultural and agro-pastoral zones, and only 7 percent households in pastoral zones own any titled plot. Figure 3.14: Land ownership and land title by agro-ecological zones 10.8% 9.3% 7.5% 10.5% 31.2% 77.3% 78.5% 80.5% 82.9% 53.6% 11.9% 15.2% 12.2% 12.0% 6.6% National Urban Agricultural Agro-pastoral Pastoral Owned and titled Owned and no title Not owned 57 The very limited access to inputs stifles agricultural productivity Irrigation is used by 12 percent of agricultural households, inorganic fertilizers are used by 18 percent of farm households and pesticides are used by 8 percent. As shown in Figure 3.15, irrigation use is almost at par in the urban and agricultural zones at 13.2 percent and nearly 14 percent respectively. As described earlier in the Niger livelihood zone map (Box 1.1), the agricultural zone of the country, along with the major urban area of Niamey, is straddled by the Niger River, and thus allows substantial irrigated rice cultivation as well as garden produce cultivation during the dry season. These zones also include Komadougou River irrigated lands and Lake Chad flood-retreat cultivation, which facilitate cash crop produce. The relatively higher level of irrigation usage in the pastoral zone at over 9 percent compared with the agro-pastoral zone (6 percent) is attributed to the Agadez pastoral zone. The water table in this region allows crop irrigation from wells, and farmers are able to produce both cereals and cash crops, primarily onions, throughout the year. Figure 3.15: Share of agricultural households using irrigation 11.7% 13.2% 13.8% 6.0% 9.2% National Urban Agricultural Agro-pastoral Pastoral Chemical fertilizers or pesticides are used on just above 20 percent of all farms. Farms that do not have access to those modern inputs are either using organic inputs (mainly manure) or are not using any kind or soil enhancement or pest control product. Figure 3.16 shows that access to inputs is easier in urban areas, with 27 percent of farm households in the region using inorganic fertilizers and pesticides compared with 21.6 percent nationally. While the agricultural areas also have a higher than average inorganic input use (23 percent), more than 50 percent of the farm households in the region use only organic fertilizers. Input use is significantly lower in pastoral areas, with 67 percent of households not using any organic or inorganic inputs. Figure 3.16: Organic and inorganic fertilizer/pesticide use 35.8% 30.8% 26.7% 43.2% 67.0% 42.6% 42.0% 50.4% 36.3% 18.3% 21.6% 27.3% 23.0% 20.5% 14.7% National Urban Agricultural Agro-pastoral Pastoral Inorganic fertilizer and/or pesticides Only organic fertilizer No fertilizer/pesticide Irrigation and fertilizers, where used, are applied to small plots. As shown in figure 3.17, farms using irrigation are smaller than average, and the area that receives irrigation is just about 1 ha on average. 58 Similarly, households who use inorganic inputs on their farm only apply them on a quarter of their land, corresponding to just above 1 Ha. In farms where they are used, even organic inputs are only applied to about half of the land. Figure 3.17: Median net farm size and technology coverage by technology (in hectares) 5 4.62 4.3 4 3.33 3 2.61 2.25 2 1.1 1 1 0 No inputs Manure or compost Chemical inputs Irrigation Total size of the farm using input Area receiving input Note: the median values of farm sizes and areas under each of the above technologies have been calculated in net terms, which mean that plots were only counted once, avoiding double counts when plots were intercropped with two or more crops at the same time or during two different seasons during the crop year. Limited access to goods and capital markets Agricultural employment outcomes may be strongly affected by the presence or lack of infrastructure for accessing inputs, commercializing production or receiving support services. Facility surveys provide information on the availability of a number of infrastructure components at the village level. As shown in Table 3.1, less than one in ten communities have a permanent market for buying and selling goods, and less than a third has a periodic market. Access to capital is made difficult by the absence of formal credit facilities, with just above 10 percent of communities with a bank or MFI. An even lower share of villages (6.5 percent) have an agricultural extension center, which means that support services are inexistent for the vast majority of workers. Cereal banks are more common. The lack of good transportation infrastructure also hinders agricultural productivity and commercialization. Less than half of communities in Niger are connected to stabilized roads, of which only a third is made of asphalt. Table 3.1: Village level infrastructure in rural areas Infrastructure Component (in percentage) Permanent market 8.87% Periodic market 28.52% Bank or microfinance institution 11.22% Agriculture Extension Center 6.46% Cereal Bank 59.48% Food stores for animals 9.05% Asphalt roads 12.54% Laterite roads 30.78% 59 Linkage between agricultural and non-agricultural activities Who has a secondary occupation out of agriculture? One third of agricultural workers has a secondary non-agricultural occupation and this holds roughly true across all agro-ecological regions, with slightly less for pastoral areas. This proportion is close to 50 percent for the 30-50 years old more active age cohort and declines after 50 years old while also being significantly higher among male workers than female ones (35 against 23 percent). Being involved in a non-agricultural activity is much more frequent among uneducated agricultural workers than among the educated ones (33 versus 10 percent). Figure 3.18: Share of agricultural workers with non-agricultural secondary occupation by age cohort, education level, and gender 100.0% 80.0% 60.0% 48.8% 47.1% 42.1% 30.2% 34.5% 33.0% 30.0% 40.0% 23.4% 6.1% 12.7% 9.3% 20.0% 0.0% >65 Female 15-30 30-50 50-65 Primary Male No formal Preschool Secondary Age cohort Gender Education National Diversification of farm and non-farm activities is more common for households with more available land and better access to inputs. Figure 3.19 suggests that access to technology may be related to more sizable market opportunities and higher farm profitability which is in turn correlated with non-agricultural income opportunities (e.g. agro-processing or agro-trading) or less labor-intensive farming freeing up labor time available for other activities. Farm-non-farm linkages seem to be mutually reinforcing when agricultural systems are modernizing and diversification and intensification seem to go hand in hand. Figure 3.19: Share of agricultural households with a mix of ag and non- occupations (processing, enterprise, other), by type of agriculture 35.0% 32.6% 14.8% 29.2% 32.0% 20.3% Irrigation Manure only Nothing Irrigation Manure only Nothing and/or and/or inorganic inorganic inputs inputs Well available land per ag worker Low available land per ag worker Note: Household occupational status relies on whether one member or more have an agricultural or a non-agricultural activity as their main ones. 60 Non-agricultural activities within farm households are dominated by agricultural processing, non- farm household businesses, housing construction, and other services. While one fourth of all farm households have at least one active member in the agro-processing, this proportion is around 45 percent for farm households who are more involved in market exchanges and commercialization of their crops. The proportion only slightly increases for other job categories and sectors. This reflects how integration to agricultural markets raises agricultural productivity but also expands the set of non-farm economic opportunities available for agricultural workers. Market participation and integration seems to provide more job opportunities to the non- agricultural sector, such as transportation, commerce, agro-processing. There seems to be a minor but significant share of the rural population that is engaged in a more productive and dynamic portfolio of agricultural-related activities that entail more returns and more investment incentives, with possibilities to engage into more lucrative markets and with higher value-chain prospects. Determinants of diversification Better understanding the conditions under which households engage in diversified activities is key for charting pathways toward more productive employment. A simple regression looks at probabilities that an agricultural household has at least one member engaged into a non-agricultural activity. Results are displayed in Table 3.2. Table 3.2: Correlates of individual and household-level diversification of activities for agricultural workers and households Household level Probability to have one non-agricultural worker Coef. Std. Err. Total operated area/ag worker -1.42E-07 1.25E-07 GENDER Female (resp. % of female) -0.075 0.009 *** Age (resp. average intra hh) 0.000 0.000 EDUCATION Preschool -0.013 0.163 Primary 0.081 0.015 *** Secondary first cycle-general 0.175 0.031 *** Secondary first cycle technical 0.383 0.211 *** Secondary second cycle general -0.014 0.138 Secondary technical and Superior AGRICULTURAL INPUTS Using irrigation technology 0.128 0.014 *** 61 Using organic matter 0.054 0.009 *** Using inorganic fertilisers 0.002 0.012 Using pesticides -0.003 0.017 INFRASTRUCTURE Permanent or periodic market -0.028 0.010 *** Paved or laterite road 0.009 0.010 GEOGRAPHY Agricultural -0.330 0.018 *** Agro-pastoral -0.406 0.018 *** Pastoral -0.343 0.022 *** Constant 0.502 0.022 *** Access to education and inputs increases the likelihood that agricultural households have at least one member engaged in non-agricultural activities. Each additional level of schooling is associated with a higher probability of farm-non-farm diversification at the household level. Better access to agricultural inputs also drives up the likelihood to work outside of agriculture, especially such farm technologies as irrigation and organic inputs. This result highlights that enhanced agricultural productivity may not be a substitute to diversification, but may well be a complement. It might allow households to grow crops other than subsistence crops, which they may then process and sell. It might also improve incomes and enable productive investments in non-agricultural activities over time. Conversely, a steady stream of income through non-agricultural activities may enable investments in farming technologies, which boosts agricultural productivity. What drives employment outcomes in agriculture? The characteristics of agricultural employment impact the welfare of workers. A welfare analysis of agricultural workers allows to identify which of the main characteristics discussed in this chapter account for variations in consumption levels, a more accurate measure of welfare than income. Household and individual level determinants of welfare The first set of determinants relates to the characteristics of the farm and the individuals or household that work on it. We first regress the consumption aggregate at the individual agricultural worker level on the agricultural technology categorical variables, available land per agricultural worker in the household, diversification of activities, and a set of individual characteristics. We then perform the same regression design to sub-samples of cash crop and subsistence crop agricultural workers. Results are presented in Table 3.3. Farm-non-farm diversification at the household level is associated with higher welfare. While coexistence of agricultural and livestock activities does not trigger additional income (is even associated with lower incomes for cash crop growers), that with non-agricultural activities entails additional 14,500 FCFA per capita on average. It only adds 9,500 FCFA per capita for households and workers growing cash crops and 62 commercializing them likely because their income level is initially higher and that their cash income is more concentrated and focused on cash crop activities. Since diversification and agricultural technology intake and use are positively correlated (see above section), additional cash income from diversified labor portfolio likely helps households to invest cash back into their farming systems. Irrigation, fertilizers, and use of pesticides are strong welfare determinants at the household level as they allow agricultural workers to allocate time to non-agricultural activities. Inputs are associated with additional income of 5,000 (for subsistence farmers) to 14,000 FCFA per capita for the cash crop growers, when used separately. Households who can afford combining irrigation and chemicals do make average income increase of 60,000 FCFA more per capita of welfare, which is substantial (25 percent of the poverty line and 4 times more than when using those technologies separately). Additional analysis not shown here finds that those technologies may not necessarily improve farm incomes but do raise farm productivity, which frees up work time and enables individuals to diversify into the non-agricultural sector, improving their overall productivity and employment outcomes across sectors. Table 3.3: Determinants of welfare income across agricultural workers Cash crop workers Whole sample and market Subsistence farmers participants Coef. Std. Err. Coef. Std. Err. Coef. Std. Err. GENDER (base category = male worker) Female 654 1,505 -3,669 2,798 2,093 1,758 Age 226 43 *** 196 80 *** 227 50 *** EDUCATION (base category = no formal school attended) Preschool 18,284 12,334 -23,215 28,467 24,199 13,654 * Primary -8,134 2,352 *** -5,775 4,353 -8,786 2,753 *** Secondary first cycle general 8,084 5,685 -19,674 10,310 ** 17,910 6,700 *** Secondary first cycle technical 137,737 33,132 *** 42,813 80,229 151,451 36,355 *** Secondary second cycle general 55,071 19,667 *** 98,042 80,213 57,187 20,513 *** Secondary technical and Superior Not enough observations - no significance FARM AND HH CHARACTERISTICS Total operated area/ag worker 1,798 169 *** 1,273 251 *** 2,425 221 *** Household share of workers 64,892 3,549 *** 61,353 7,374 *** 67,568 4,051 *** At least one livestock worker in hh -9,406 1,802 *** -29,377 2,986 *** 1,493 2,223 At least one non-agricultural worker 14,576 1,841 *** 9,509 3,144 *** 14,216 2,249 *** Using only organic inputs -2,861 1,876 3,490 3,870 -4,057 2,159 *** Using irrigation or inorganic inputs 7,611 1,963 *** 14,631 4,091 *** 4,785 2,272 ** Using both 58,770 6,646 *** 31,252 8,577 *** 67,568 10,788 *** 63 At least one crop is commercialized 8,414 1,863 *** INFRASTRUCTURE Permanent or periodic market 3,170 1,675 ** -5,591 3,148 * 8,236 1,968 *** Laterite or paved road 21,140 1,724 *** 21,552 3,837 *** 17,627 1,957 *** AGRO ECOLOGICAL AREA (Urban=base category) Agricultural -46,243 2,769 *** -71,822 6,822 *** -42,952 3,092 *** Agro-pastoral -71,531 2,867 *** -101,053 7,140 *** -67,847 3,179 *** Pastoral -46,472 3,383 *** -33,225 7,477 *** -56,619 3,928 *** Constant 181,758 3,712 0 225,845 8,488 *** 175,063 4,216 *** Land access, farm size and market participation also have positive effects on employment outcomes. Households with more land available per individual worker have better welfare levels of around yearly additional 1,800 FCFA per capita and per Ha per agricultural worker. This varies between 1,200 FCFA per ha for cash crop growers and two times more for subsistence farmers, but note that subsistence farmers have less available land per worker on average from base. Market participation seems to have direct positive effects with additional 9,000 FCFA per capita for those households who commercialize at least one crop. Additional analysis shows that commercialization affects welfare through more employment since marketing of one crop is associated with more employment of both agricultural and non-agricultural workers of the household. Community and crop-level determinants of agricultural productivity and welfare Tables 3.4. and 3.5. present results from similar regressions as Table 3.3 but with additional variables. They display the coefficients and standard errors corresponding to these new variables only. Households that grow onion, niche cereals and fruits, and to a lesser extent sesame and cowpeas, have higher employment outcomes. Growing onions is associated with a much higher level of consumption. For the biggest crops, we note the significant additional income effect of cowpeas -6,000 FCFA per capita per Ha- and sesame -11,000 additional FCFA per capita per Ha. Small-scale niche cereals, starches as well as fruits and vegetables do often provide more income on a per Ha basis too, but with small plots in most cases, like what is observed for onions. All in all, cowpeas seems to be the most profitable crop among the top five crops, then followed by sesame and onions among the cash crops of lower acreage. Niche markets seem profitable too even with small plots but one may wonder how and whether they could be scaled up. Infrastructures are confirmed as being significantly positive on household welfare. Having a permanent market, MFIs, agricultural extension centers, food stores for animals and asphalt road have a significant welfare-improving effect. Those infrastructures are indeed associated with more rural non-agricultural employment as well as more agricultural employment for the MFIs and food stores. Permanent markets are associated with additional 30,000 FCFA per capita on average, while agricultural extension, food stores, microfinance institutions and asphalt roads contribute to an additional 10 to 12,000 FCFA per capita. 64 Table 3.4: Crops and agricultural welfare Table 3.5: Infrastructure and agricultural welfare Coef. Std. Err. Coef. Std. Err. VILLAGE CROPS: area per INFRASTRUCTURE capita Permanent Cowpeas 5,702 1,698 *** market 30,177 3,151 *** Gombo -16,398 3,703 *** Periodic market -13,502 1,855 *** Millet 374 1,833 Bank or microfinance Onion 230,778 21,935 *** institution 10,476 2,593 *** Peanuts 1,478 2,847 Agriculture Sesame 11,470 3,425 *** Extension Center 13,817 3,825 *** Sorghum 297 1,180 Cereal Bank 1,409 1,658 Food stores for Sorrel -11,636 3,327 *** animals 10,187 3,152 *** Vouanzou 7,040 9,069 Asphalt roads 12,047 2,521 *** Other cereals and starches 161,979 9,159 *** Laterite roads 3,305 1,896 * Other fruits and vegetables 31,988 9,880 *** Individual characteristics Y Individual Farm characteristics Y characteristics Y Farm Household characteristics Y characteristics Y Household Constant 177,460 3,739 *** characteristics Y Constant 184,820 3,693 *** 65 66 Chapter 4 : NON-AGRICULTURAL HOUSEHOLD ENTERPRISES Highlight of the main results • Almost all households own one or several non-agricultural household enterprises, even though they only represent 15% of primary occupations. • 87% of HEs only have one worker. • Close to 60% of household enterprises operate throughout the year, which makes them a useful income complement to highly seasonal agriculture. • HEs operate at a very local level: 72% of HEs source their inputs locally, and 73% have individuals as their primary customers. • 70% of HEs do not keep books and only 2% have a dedicated operating facility • Women own 40% of HEs, but those are 40% less productive than men’s, all else being equal. • In 80% of the cases, HEs are financed by the household’s own savings. Another 10% comes from other households. Only 0.7% are financed by loans from formal or informal credit institutions. • Access to customers and capital are the main constraints to HE productivity and development. Introduction Non-agricultural self-employment and household enterprises (HE) constitute 15% of primary occupations in Niger. Yet even when agriculture is the primary occupation, HE are ubiquitous as an important source of income diversification. Furthermore, in the context of a declining share of agriculture in employment, and in the absence of growth in formal wage employment, HE are growing as a share of total employment and activity. A national jobs strategy must thus understand the diversity of this sector of employment, its determinants, as well as the constraints it faces, if it wants to leverage this sector for job creation and poverty reduction. Alleviating the constraints to the creation and growth of HE can have large effects in improving jobs, productivity and incomes, especially for households in poverty. As shown in Fox and Gaal (2008), and in Fox and Sohnesen (2012), most non-farm jobs are created in this sector in Sub-Saharan Africa, even during high-growth periods. Facilitating the income-generating activities of HE is thus one way of promoting inclusive growth, especially given that household in poverty heavily rely on HE as either their main or complementary source of income. 67 Characteristics of Household Enterprises and Non-agricultural self-employment Household enterprises can be defined as production units owned and operated by individuals and operating in the non-agricultural sector. Although there is no unique and clear definition of HEs, Benjamin et al. (2012) and Hussmann (2004) show that most firms and workers in the sector are informal. Household Enterprises operate consistently over time but at a very small scale HE are a standard feature of households in Niger and many households host more than one. Most households (67 percent) have one household enterprise and most household enterprises are not SMEs. Just a handful of them tend to possess more than four household enterprises (Figure 4.1). Most household enterprises operate year-round. As shown in Figure 4.2, 58 percent operate from 10 to 12 months a year, while about 30 percent are in operation less than six months. In that respect, household enterprises are in clear contrast with agricultural work, which is generally seasonal and concentrated in just a few months every year. This difference highlights the benefits for households to combine both types of activities in order to generate a steady stream of income throughout the year. Figure 4.1: Most Households do possess one non-agricultural enterprise or more 67.0 22.7 7.1 2.0 0.7 0.4 One HE Two HEs Three HEs Four HEs Five HEs Six HEs Household enterprises generally operate over several years, and a significant share lives 10 years or more. The most common age for a HE is around 3 years, but the average life span is 10 years and a large segment of the distribution has been in operation for more than a decade, as is shown in Figure 4.3. They form a relatively stable pool of economic activities over the years. 68 Figure 4.2: Most HEs operate throughout the Figure 4.3: Most of HEs are not short lived year Kernel density estimate .08 58.2 .06 Density .04 18.6 11.5 11.7 .02 0 0 10 20 30 40 50 60 70 80 Up to 3 months 4-6 months 7-9 months 10-12 months Number of years in operation Note: Red line is average number of years in operation kernel = epanechnikov, bandwidth = 1.7669 Most household enterprises operate within the realm of local markets. Figures 4.4 and 4.5 present the source of input and the client types for HE. It can be observed from these figures that most of the production of HE is oriented to the local market, and HE acquire most of their inputs from a local market. Their clients are mostly local households (73 percent) or the public sector (21 percent), while their inputs are also mostly local, from either small companies (44 percent of the inputs) or directly from other households/individuals (28 percent of their inputs). Only 11 percent of their inputs are coming from large enterprises or imports. Figure 4.4: Source of inputs. Most inputs of HEs Figure 4.5: Client types. Most HEs products are are from small local markets sold in local markets 44.4 73.4 27.6 17.2 21.4 7.3 3.6 3.6 1.5 0.1 Petite ent. Menage/Particulier Secteur public Importations Grande ent. Menage/Particulier Secteur public Petite ent. Grande ent. Exportations directes directes Most of the non-agricultural individual enterprises are in small-scale services. At the national level, some 37.1 percent are in retail services, followed by other services (21.4 percent). Other common activities are in agro-business (16.3 percent), hotel and restaurant services (6.1 percent), and personal care services, such as shoe repairs (5.9 percent) (Figure 4.6). 69 Figure 4.6: Type of activity of the non-agricultural micro-enterprises, by area of residence 45.1 37.1 Rural 35.1 Urban National 21.2 22.3 21.4 17.1 16.3 13.1 7.7 6.2 5.6 6.1 5.4 5.9 6.4 5.6 5.7 4.9 2.4 2.9 2.1 2.7 1.6 Commerce detail Services Agro-alimentaire Restaurant/hotel Confection/Chaussures Autres industries Commerce de gros BTP Nearly all non-agricultural household enterprises are informal. Indeed, less than 1 percent of the operating household enterprises are registered in any form (either having a tax ID or registered in the registry of commerce etc.) (Figure 4.7), and more than 70 percent do not keep any formal books (Figure 4.8). Moreover, given the nature of these enterprises, most of them do not have a stable operation location. For instance, nearly 40 percent of HE operate at home and another 46 percent operate either on the street or in a mobile unit such as a car or motorcycle (Figure 4.9). Figure 4.7: Most of non-agricultural household enterprises are not registered in any form No Yes 0.5 0.6 0.9 2.6 6.3 32.0 99.5 99.4 99.1 97.4 93.7 68.0 CNSS NIF RC Tap water Electricity Telefon 70 Figure 4.8: Most HEs do not keep books Figure 4.9: Most HEs do not have fixed place to operate 70.1 39.7 25.7 19.8 29.5 7.5 4.9 2.3 0.4 Oui, transmis à la DGI Oui, non transmis à la DGI Non, pas de comptabilité Women, the youth and the non-educated are well represented in HEs Women represent a significant share of owners of HE. In the context in which labor participation is higher for men (90 percent) than for women (66 percent), it is important to note that women make up more than 40% of the owners of HE. In that regard, HE are an important way for women to participate in the labor market. HE are thus a crucial pathway for women to contribute to the household’s portfolio of activities. Similarly, the youth and people without education represent a large share of HE owners. Figures 4.10 and 4.11 show that the distribution of HE ownership is quite balanced across age groups, and those without education represent more than 80 percent of owners. Figure 4.10: Non-agricultural micro-enterprises Figure 4.11: Most of Non-agricultural micro- owners are mostly young people enterprises owners have no formal education 17.6 81.2 14.1 12.5 13.0 11.3 9.2 6.9 5.9 6.0 12.6 3.5 5.9 0.4 No Edu Primary Sec. 1st cycle Sec. 2nd cycle 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 and above Owners of HE are often the head of the household (Figure 4.12), but a significant share is owned by the husband or wife (30.5 percent), or sons and daughters of the head. More than 87 percent of the individuals 71 in household enterprise work alone (Figure 13), with few non-family members. Most of the family help comes from young adults working with their parents. Figure 4.12: Most Non-agricultural micro- Figure 4.13: Most individuals in non-ag enterprises owners are household heads enterprises work alone 56.1 87.4 30.5 8.7 9.3 3.6 0.4 2.5 1.6 Self-employed Self-employed With 1-4 5 + employees Head Husband/Spouse Son/Daugther of Direct relative of Others with family employees of Head Head Head helpers Female-owned and rural HEs have much lower productivity levels There are substantial variations in productivity in the sector, reflecting its heterogeneity. Productivity is defined as in Nagler and Naude (2014). Figure 4.14 to Figure 4.15 present different facets of this heterogeneity. A few of stylized facts emerge from these Figures. The first and most salient fact about the productivity differences is that there are on average many individuals that are just surviving in the sector, while a lot of them that are doing very well. The sector has a continuum of productivity: low, average and high-productive operators co-exist in the sector, as can be observed in all four figures presented (Figure 4.14 to Figure 4.17). Gender differences in productivity may reflect specific additional constraints faced by women. A second important stylized fact is that male owners work in more productive HE, while women are on average observed in less productive HE (Figure 4.14 and 4.15). The differences in productivity by gender are extremely large: the median woman has earnings that are equivalent to the 15th percentile man. These differences in productivity are not due to a composition effect (that could arise if women worked in less productive sectors). Even when comparing similar HE and otherwise similar owner characteristics (in terms of educational attainment, skills and age), women earn almost 40 percent less than men, as is shown in the regression analysis presented in table A2 in the Appendix. Rural areas and second tier cities are less productive with the same level of inputs and owners’ background. A third, and common finding in most countries, concerns the productivity difference between the capital, other urban areas, and rural areas. Productivity is higher in Niamey, followed by the other urban areas, while the rural areas have the HE that are on average less productive. (Figure 4.15). Finally, capital-intensive sectors tend to be more productive than labor-intensive sectors. The heterogeneity in household enterprises is reflected in the wide productivity variation across sector type. Sales-related activities (e.g., retails of personal articles or wholesale) are the most productive sectors. These sectors are generally more capital intensive (Figure 4.18, Figure 4.19, and Table A2 in the Appendix). Personal services such as shoe confection, requiring less capital, have lower output per worker. 72 Figure 4.14: Productivity dispersion by Figure 4.15: Cumulative distribution of gender productivity by gender 1 Kernel density estimate .25 Male Female .8 .2 Cumulative Probability .6 .15 Density .4 .1 .2 .05 Male Female 0 0 5 6 7 8 9 10 11 12 13 14 15 16 17 18 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Log(productivity per worker) Log(productivity per worker) kernel = epanechnikov, bandwidth = 0.3696 Figure 4.16: Productivity dispersion by Figure 4.17: Cumulative distribution of residence productivity by residence 1 Kernel density estimate .25 Niamey Other Urban .8 Rural .2 Cumulative Probability .6 .15 Density .4 .1 .2 .05 Niamey Other urban Rural 0 0 5 6 7 8 9 10 11 12 13 14 15 16 17 18 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Log(productivity per worker) Log(productivity per worker) kernel = epanechnikov, bandwidth = 0.4359 Figure 4.18: Output per worker by sector Figure 4.19: Productivity premium compared to restaurant/hotel activities Adjusted mean of productivity by sector Contrasts of adjusted means to owners in restaurant/hotel activities Commerce de gros 1.5 Commerce de detail 1 Contrasts of Linear Prediction BTP Services .5 Agro-alimentaire 0 Restaurant/hotel -.5 Confection/Chaussures Autres industries 11 11.5 12 12.5 13 13.5 -1 Log (productivity) (1 vs 3) (2 vs 3) (4 vs 3) (5 vs 3) (6 vs 3) (7 vs 3) (8 vs 3) Note : 1=Autres industries ; 2=Confection/chaussures ; 3=Restaurant/hotel ; 4=Agro-alimentaire; 5=Services; 6=BTP; 73 7=Commerce de detail; 8=Commerce de gros Source : ECVMA 2014; Bank staff’s calculation Constraints and drivers of productivity of Household Enterprises In Niger, individuals face multiple constraints to create and operate household enterprises. These constraints include having access to capital, the difficulty of reaching sufficient customers, or the amount of competition they face in their sector. Individuals have limited access to capital, and therefore often rely on own savings and other informal forms of lending (Figure 4.20), and especially for rural entrepreneurs (Figure 4.21). Access to capital is among the most binding of constraints in creating a HE. The lack of access to capital is especially pronounced in Niger, when compared to other countries of the Sahel and West Africa region (Table A3 in the Appendix). To overcome this constraint, more than 80 percent of households finance their HE through savings. Figure 4.20: Most Non-agricultural micro- Figure 4.21: Rural entrepreneurs rely more on enterprises are financed by own savings own savings than urban counterparts 80.7 Rural Urban 81.6 77.4 5.2 5.1 8.3 10.4 8.9 5.8 0.7 3.9 5.2 4.6 0.4 1.7 Epargne du Cadeau d'un Pret d'un autre Prets (tontine, Autres Epargne du Cadeau d'un Pret d'un autre Prets (tontine, Autres menage parent menage IMF, ONG, menage parent menage IMF, ONG, cooperative) cooperative) Figure 4.22: Main constraints and difficulties operating a non-agricultural household enterprise 74 No Yes 6.2 6.0 4.7 4.7 3.0 1.4 13.0 20.6 29.9 38.4 93.8 94.0 95.3 95.3 97.0 98.6 87.0 79.4 70.1 61.6 Manque clientele Concurrence Acces au credit Approvisionnement Manque de place, Acces aux Diff. techniques de Diff. techniques de Trop de Recrutement de local adapte equipements fabrication gestion reglementation, personnel qualifie impots et taxes Market reach and access to finance are among the largest constraints enumerated by HE. While in operation, lack of customers and the difficulty of accessing rolling capital represent key constraints for individuals in the sector, as shown in Figure 4.22. The lack of a fixed and safe working place represents a key constraint for having a productive non-agricultural household enterprise. As described earlier, close to 40 percent of the owners do operate from home, followed by 25.7 percent on the street. Yet, estimates indicate entrepreneurs operating from home have significantly lower productivity (Figure 4.23). Figure 4.23: Productivity premium compared to Figure 4.24: Years in operation and home based activities productivity Contrasts of adjusted means to owners operating at home Predictive Margins with 95% CIs 12.4 1.5 Contrasts of Linear Prediction 12.2 Linear Prediction 1 12 .5 11.8 0 0 2 4 6 8 10 12 14 16 18 20 (2 vs 1) (3 vs 1) (4 vs 1) (5 vs 1) (6 vs 1) Number of years since establishment Note: 1=A domicile ; 2=Domicile du client ; Source: ECVMA 2014; Bank staff’s calculation 3=Ambulant; 4=Autres ; 5 =Voie publique ; 6=Local professionnel Source: ECVMA 2014; Bank staff’s calculation 75 The productivity of HE sharply increases with the owner’s ski lls. The drivers of productivity are analyzed in Table A2 in the Appendix, where productivity is decomposed using a linear regression in which both the owner’s and firm’s characteristics are examined. One of the most important variables predicting the productivity level of income of a HE is whether the owner can calculate (in any language). The results indicate that owners responding that they can calculate have earnings around one third higher than those that cannot, even when comparing within the same sectors of activity, and even after taking into account other personal factors such as age, level of education and gender. More importantly, this measure of skills seems to be a better predictor of earnings than the educational attainment of the owner. When it comes to formal education, once the question on skills is controlled for, there is very faint evidence of positive returns of schooling on productivity in household enterprises, especially for those with secondary and upper level of education (Table A2 in the Appendix). The small returns to schooling disappears when other factors such as experience (proxied here with number of years in operations) are controlled for (Table A2, in the Appendix). The positive and significant effect of experience is shown in Figure 4.24. Policies and Programs for the HE sector HE need support from tailored policies in order to grow. Most policies designed to boost employment and reduce poverty have been limited in their scope. More importantly, national employment strategies in Niger have not targeted specifically the needs of the HE. The particular needs of HE are related to access to credit, access to market, addressing the skills and market information of the owners, as well as addressing the specific constraints faced by women. Access to credit can be a major constraint for both starting HE and for developing existing ones. The current low level of banking coverage and lack of micro-credit availability, compared even to countries in the region, is a major issue preventing the development and growth of HE. Given that most HE are very small, rural, or have no accounting, it follows that access to formal banking credit seems unrealistic even in the medium run. The development of micro-credit institutions and other innovative financial intermediaries is more realistic as a source of capital for the growth of most HE in Niger. Micro-credit institutions have shown to have large effects on the expansion of income-generating activities and self- employment (Crepon, 2015). In most countries and areas where micro-credit becomes available, families either create or expand their HE, they increase their business assets, and therefore observe an increase in their incomes (Banerjee, 2015). Education and skills are a strong correlate to productivity in HE. Developing a national strategy for improving the skills that HE owners have is also necessary. As shown in the previous section, the productivity of those with calculating skills is at least 30 percent higher compared to the one with no calculating skills. Those with a finished primary school degree have higher earnings than those without. Ensuring that children are going through school and ensuring good skills-acquisition in school may have large positive impact on the productivity of future HE sector. In that sense, reforms that increase both school quality and attendance is a recommended policy that can be part of both the jobs and the national skills strategy. Focusing scarce public funds on earlier schooling interventions may be a very effective and inclusive employment strategy [Adams et al, 2013). Training and skill-building programs for the youth are seldom very successful (Blattman and Ralston, 2015), and are often difficult to design in a cost-effective way. However, mobilizing the private sector for the delivery of training programs can be a way of ensuring that the training supplied is in line with market demands. In providing the training options, particular attention can be transmitted to raise awareness of the earnings obtained by HE in overcrowded, low-returns sectors with easy entry. 76 Finally, focusing efforts on the specific constraints faced by women may have large returns. Although most HE owners are men, women still own a large share of the HE, but earn considerably less, even for HE with the same characteristics in the same sector. Identifying and addressing the constraints they face may have large impact on their earnings. While studying the effect of micro-credit for HE in Kenya, Dupas and Robinson (2013) found that women were much more likely to benefit it than men, pinpointing that access to capital was more of a constraint for them than for men. In a similar manner, programs that specifically aimed at fostering entrepreneurship appear to be more effective for women than for men, in a meta-analysis of randomized controlled trials (Cho and Honorati, 2013). 77 References Adams, Arvil V.; Silva, Sara Johansson de; Razmara, Setareh. 2013. Improving skills development in the informal sector : strategies for Sub-Saharan Africa. Directions in development ; human development. Washington DC ; World Bank. Augeraud, P. 1988. « Exploitation de l’Enquête Secteur Informel Niger 1987/1988 pour la Comptabilité Nationale » http://www.dial.prd.fr/dial_publications/STATECO/pdf/65/65_3.pdf Banerjee, A. V. (2013). Microcredit Under the Microscope: What Have We Learned in the Past Two Decades, and What Do We Need to Know? Annual Review of Economics (Vol. 5). https://doi.org/10.1146/annurev-economics-082912-110220 Beck, T., & Cull, R. (2014). Small-And Medium-Sized Enterprise Finance in Africa. Africa Growth Initiative Working Papers. Benjamin, N. and A. A. Mbaye. 2012. The Informal Sector in Francophone Africa : Firm size, Productivity and Institutions, World Bank. Benjamin, Nancy, K. Beegle, F. Recanatini, and M. Santini. 2014. “Informal Economy and the World Bank,� Policy Research Working Paper, WPS6888, The World Bank, Washington, DC. Blattman, C., & Ralston, L. (2015). Generating Employment in Poor and Fragile States: Evidence from Labor Market and Entrepreneurship Programs. Crépon, B., Devoto, F., Duflo, E., & Pariente, W. (2015). Estimating the Impact of Microcredit on Those Who Take It Up: Evidence from a Randomized Expirement in Morocco. American Economic Journal: Applied Economics, 7(1), 123–150. https://doi.org/10.1257/app.20130535 Cho, Y., & Honorati, M. (2013). Entrepreneurship Programs in Developing Countries: A Meta Regression Analysis. Discussion Paper Series, IZA, 7333, 63. https://doi.org/10.1007/s10273-011-1262-2 Dupas, Pascaline, and Jonathan Robinson. "Savings constraints and microenterprise development: Evidence from a field experiment in Kenya." American Economic Journal: Applied Economics 5.1 (2013): 163-192. Filmer, Deon and L. Fox . 2014. Youth Employment in Sub-Saharan Africa. The World Bank, Washington, DC. Fox, Louise et al. 2013. “Africa’s Got Work to Do: Employment Prospects in the New Century.� IMF Working Paper. WP/13/201. IMF. Washington, DC Fox, Louise and Melissa S. Gaal (2008). Working out of Poverty: Job Creation and the Quality of Growth in Africa. The World Bank, Washington, DC Fox, L., & Sohnesen, T. P. (2012). Household Enterprises in Sub-Saharan Africa Why They Matter for Growth , Jobs , and Livelihoods. Policy Research Working Paper Series, The World Bank, (August), 53. Guenther, Isabel and A. Launov. 2012. “Informal Employment in Developing Countries: Opportunities or Last Resort?� Journal of Development Economics 97 (1): 88-98. Hussmanns, R. 2004. Defining and Measuring Informal Employment. International Labor Organisation, Geneva http://www.ilo.org/public/english/bureau/stat/download/papers/meas.pdf Kanbur, R. 2009. “Conceptualizing Informality�: Regulation and Enforcement, Cornell University, Dep t of Applied Economics and Management, Working Paper 09-11. Loayza, Norman V. and J. Rigolini. 2011. “Informal Employment: Safety Net or Growth Engine?� World Development 39 (9): 1503-15. Maloney, W. F. 2004. “Informality Revisited.� World Development 32 (7): 1159-78. Ministère des Finances. 2013. Enquête Nationale sur l’Emploi et le Secteur Informel au Niger (2012). Nagler, Pauler and W. Naude. 2014. “Non-Farm Enterprises in Rural Africa: New Empirical Evidence.� Policy Research Working Paper, WPS7066, The World Bank, Washington, DC. 78 OECD. 2008. Rapport Afrique de l’Ouest 2007-2008. https://www.oecd.org/fr/csao/publications/42358563.pdf Perry, E. Guillermo et al. 2007. Informality: Exit and Exclusion. The World Bank, Washington, DC. Steel, W. and D. Snodgrass. 2008. “Raising Productivity and Reducing Risks of Household Enterprises�: Diagnostic Methodology Framework, World Bank. UNECA. 2009. Étude sur la Mesure du Secteur Informel et de l’emploi informel en Afrique. http://repository.uneca.org/bitstream/handle/10855/3351/bib.%2027408_I%20F.pdf?sequence=1 79 Appendix Table A1: Probit Estimates for Selection into Self-employment Specification I Specification II All Urban Rural All Urban Rural (1) (2) (3) (4) (5) (6) Male 0.408*** 0.251*** 0.525*** 0.534*** 0.370*** 0.666*** (0.031) (0.050) (0.040) (0.033) (0.054) (0.044) Age 0.159*** 0.204*** 0.138*** 0.119*** 0.158*** 0.098*** (0.007) (0.012) (0.009) (0.008) (0.014) (0.011) Age2/100 - - - - - - 0.172*** 0.224*** 0.149*** 0.132*** 0.178*** 0.111*** (0.009) (0.015) (0.011) (0.010) (0.017) (0.013) Education (Ref.: No Education) Primary 0.260*** 0.212*** 0.185** 0.246*** 0.198*** 0.187** (0.049) (0.068) (0.077) (0.050) (0.069) (0.077) Secondary 1st cycle 0.098 0.075 0.071 0.062 0.045 0.049 (0.064) (0.082) (0.117) (0.065) (0.083) (0.119) Secondary 2nd cycle and - - -0.975** - - -1.028** more 0.897*** 0.886*** 0.922*** 0.900*** (0.107) (0.118) (0.405) (0.108) (0.120) (0.406) Can calculate in a language -0.059 - 0.122** -0.017 - 0.132** (Yes) 0.329*** 0.287*** (0.043) (0.064) (0.062) (0.044) (0.065) (0.063) Region (Ref.: Zinder) Agadez - - 0.008 - - 0.057 0.208*** 0.522*** 0.191*** 0.550*** (0.062) (0.125) (0.074) (0.063) (0.126) (0.075) Diffa - - - - - - 0.878*** 0.645*** 0.759*** 0.890*** 0.611*** 0.765*** (0.075) (0.172) (0.088) (0.076) (0.174) (0.089) Dosso - - 0.032 - - 0.040 0.235*** 0.825*** 0.248*** 0.842*** (0.059) (0.130) (0.069) (0.059) (0.131) (0.070) Maradi 0.316*** 0.008 0.510*** 0.310*** -0.020 0.518*** (0.056) (0.104) (0.069) (0.057) (0.105) (0.070) Tahoua -0.021 - 0.202*** -0.026 - 0.220*** 0.496*** 0.519*** (0.059) (0.110) (0.072) (0.059) (0.111) (0.073) Tillaberi - - -0.064 - - -0.036 0.323*** 0.810*** 0.325*** 0.833*** (0.060) (0.149) (0.071) (0.061) (0.149) (0.071) Niamey - - - - 0.211*** 0.742*** 0.149*** 0.721*** (0.053) (0.079) (0.054) (0.080) # of female between (12-49 - - - -0.020* -0.031* - 80 years) 0.032*** 0.051*** 0.047*** 0.049*** (0.011) (0.017) (0.016) (0.012) (0.018) (0.017) Marital status (Ref.: Single) Married mono 0.541*** 0.557*** 0.551*** (0.057) (0.083) (0.082) Married poly 0.601*** 0.561*** 0.684*** (0.065) (0.098) (0.094) Widow 0.828*** 0.643*** 0.953*** (0.094) (0.145) (0.130) Divorced/separated 0.859*** 0.699*** 0.957*** (0.102) (0.144) (0.149) Constant - - - - - - 3.833*** 3.899*** 3.703*** 3.529*** 3.439*** 3.466*** (0.139) (0.229) (0.180) (0.144) (0.240) (0.187) Observations 9,767 3,980 5,787 9,767 3,980 5,787 Pseudo-R2 0.144 0.209 0.135 0.156 0.220 0.147 Log Lik -4758 -1855 -2774 -4694 -1829 -2736 Note: Standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1 Source: ECVMA 2014; Bank staff’s estimates Table A2: OLS estimates for Determinants of Household Enterprises Productivity (Dep. var.: ln(productivity), results are from the second stage of Heckman Selection Model) All Urban Rural (1) (2) (3) (4) (5) (6) Male 0.074 0.913*** 0.341 1.075*** -0.022 0.813** (0.257) (0.245) (0.356) (0.338) (0.353) (0.329) Age -0.286*** 0.104 -0.183 0.165 -0.362*** 0.038 (0.096) (0.093) (0.131) (0.125) (0.135) (0.127) Age2/100 0.312*** -0.117 0.207 -0.180 0.390*** -0.045 (0.105) (0.101) (0.142) (0.136) (0.147) (0.137) Education (Ref.: No Education) Primary -0.719*** 0.034 -0.483* 0.152 -0.984*** -0.193 (0.176) (0.169) (0.253) (0.237) (0.245) (0.235) Secondary 1st cycle -0.030 0.199 -0.081 0.108 0.477** 0.721*** (0.139) (0.130) (0.187) (0.176) (0.215) (0.200) Secondary 2nd cycle and 2.956*** 0.338 2.578*** 0.216 3.593*** 0.789 81 more (0.681) (0.645) (0.858) (0.816) (1.260) (1.268) Can calculate in a language 0.386*** 0.305*** 0.131 0.089 0.340*** 0.221** (Yes) (0.095) (0.086) (0.141) (0.128) (0.124) (0.111) Sector (Ref.: restaurant/hotel) Autres indutries -0.605*** - -0.151 -0.200 -0.884*** -1.039*** 0.743*** (0.189) (0.173) (0.367) (0.336) (0.250) (0.217) Confection/chaussures -0.423** - 0.022 -0.152 -1.036*** -1.266*** 0.608*** (0.188) (0.168) (0.244) (0.224) (0.264) (0.220) Agro-alimentaire 0.222 0.184 0.125 0.207 0.224 0.064 (0.166) (0.147) (0.224) (0.210) (0.232) (0.191) Services -0.318* - -0.112 -0.378* -0.516** -0.824*** 0.577*** (0.163) (0.147) (0.219) (0.206) (0.235) (0.200) BTP -0.294 -0.308 -0.131 -0.098 -0.517 -0.678** (0.241) (0.216) (0.337) (0.306) (0.347) (0.293) Commerce de detail 0.404*** 0.243* 0.569*** 0.440** 0.178 -0.053 (0.152) (0.136) (0.199) (0.187) (0.221) (0.184) Commerce de gros 1.325*** 1.046*** 0.742* 0.441 1.446*** 1.147*** (0.229) (0.215) (0.389) (0.408) (0.285) (0.257) Place of operation (Ref.: home) Domicile du client 0.087 0.113 0.250 0.235 -0.053 -0.029 (0.114) (0.111) (0.159) (0.159) (0.168) (0.163) Voiture/moto/ambulant 0.354*** 0.389*** 0.319* 0.412** 0.270** 0.269** (0.100) (0.092) (0.175) (0.161) (0.124) (0.113) Autres 0.370** 0.354** 0.419 0.450 0.275 0.186 (0.177) (0.171) (0.366) (0.319) (0.199) (0.194) Voie publique 0.735*** 0.750*** 0.726*** 0.742*** 0.607*** 0.573*** (0.088) (0.081) (0.126) (0.117) (0.120) (0.110) Local professionnel 1.306*** 1.304*** 0.997*** 0.979*** 1.598*** 1.488*** (0.180) (0.162) (0.200) (0.182) (0.354) (0.313) Bookkeeping (Ref.: No) Yes, and registered w/ 0.369 0.423 0.456 0.470 0.910 0.502 DGI (0.367) (0.316) (0.363) (0.308) (0.770) (1.002) Yes, but not registered w/ -0.050 -0.086 0.113 0.072 0.019 -0.005 DGI (0.068) (0.062) (0.117) (0.106) (0.084) (0.075) Has a Fiscal ID (Yes) 0.544* 0.788*** 0.550* 0.759*** 0.334 0.457 (0.309) (0.262) (0.309) (0.247) (0.753) (0.995) # of workers in the HE - - -0.180*** 0.148*** 0.124*** (0.009) (0.011) (0.014) 82 # of years in operations 0.018*** 0.028*** 0.009** (0.004) (0.006) (0.004) Inverse Mills ratio -3.853*** -0.066 -3.084*** 0.267 -4.456*** -0.495 (0.825) (0.799) (1.143) (1.085) (1.163) (1.097) Constant 21.469*** 9.473*** 18.056*** 7.418* 24.059*** 11.773*** (2.999) (2.890) (4.126) (3.911) (4.214) (3.941) Observations 2,557 2,557 1,145 1,145 1,412 1,412 R-squared 0.360 0.463 0.356 0.463 0.397 0.506 Adj. R-sq 0.353 0.456 0.339 0.448 0.384 0.495 Note: Robust standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1 Source: ECVMA 2014; Bank staff’s estimates Table A3. World Bank Doing Business indicators, Access to Credit coverage (% of adults) Country Strength of Depth of credit Credit Credit legal rights information registry bureau index (0-12) index (0-8) Cameroon 6 1 8 0 Mauritania 2 3 6.6 0 Chad 6 0 2.3 0 Benin 6 0 0.6 0 Togo 6 0 0.5 0 Niger 6 0 0.3 0 Mali 6 0 0.1 0 Nigeria 7 6 0.1 7.7 Source: Doing business indicators database, retrieved online April 14th 2017. 83 84 Chapter 5 : YOUTH OCCUPATIONAL ASPIRATIONS Highlight of the main results • As children, girls have a lower capacity to aspire than boys, and so do rural children compared to urban children. • More than 90% youth aspire to non-agricultural jobs. The mismatch with the reality of employment structures might fuel frustration and disillusion. • Education widens the gap between aspirations and attainments, as it strongly reinforces aspirations for public sector jobs which are only available to 2% of the population. • Young men have better defined occupational aspiration for the medium term. Young women are 2.5 times more likely not to know what they want to achieve over the next 5 years. • Two thirds of inactive young women want to pursue occupational objectives. Continued inactivity does not appear as a positive choice for women. • Men’s aspired income is three times higher than women’s. Financial aspirations are much higher than attainments. • Education drives aspirations up, and cancels out the gender gap: educated girls have the same levels of aspirations as educated boys. • More than 85% youth have a male role model, including almost 80% of young women. Less than 5% have a role model working in agriculture. • Almost 50% of youth perceive success as mostly determined by luck. Initial endowments and hard work are distant second and third. • Almost half the youth feel obligated to do the same job as their parents. This sense of obligation is much lower in urban areas. • Migration is seen as a pathway to better jobs, but a third of young women think they can’t migrate because of family pressure as opposed to 3% of young men. • 45% of young women say soft skills trainings would help them fulfill their goals, while 35% young men mention management training. 85 Introduction Youth constitute an increasing proportion of the population in Niger. Ensuring that youth have opportunities for productive employment is critical to promote the political and economic development of the country. Aspirations are linked to educational and employment outcomes of youth and shaped by the context and environment in which children grow up. In addition, youth aspirations can explain economic behavior including savings, investments in human capital, and occupational decisions. Therefore, a successful development strategy to promote jobs and economic inclusion must take into account the nature of youth aspirations to maximize impact and effectiveness. According to the literature, many demographic, socioeconomic, and environmental factors are correlated with aspirations. These determinants include gender, ethnicity, location, socioeconomic status, educational level, and the presence and influence of successful role models (e.g. Jacob, 2002; Kritzinger, 2002; Heckman et al 2006; Benner and Mistry 2007; Macours and Vakis 2009; Leavy and Smith, 2010; Ainley et al. 2011; Kirk et al. 2011; Kosec et al., 2012; Beaman el al 2012; Bernard et al 2014). Thus, aspirations are determined by both exogenous and endogenous influences, meaning that they may simultaneously influence aspirations and be a result of aspirations among youth. Given the central role of aspirations in the transition from school to work, it is important to understand the particular factors that influence occupational aspirations of youth. Aspirations might explain why some youth take advantage of existing opportunities more than others, including employment opportunities and social assistance programs. It might also help explain why some youth are more motivated to succeed in their jobs and improve their social and economic conditions. To design effective policies it is therefore central to analyze to what extent the occupational aspirations of youth align with the current and longer-term job opportunities in Niger. The analysis of youth aspiration is based on data collected from ECVMA household survey in Niger conducted in 2014. In partnership with the National Statistical Office (INS Niamey) and the World Bank’s LSMS team, an innovative survey module dedicated to aspirations and non-cognitive skills was designed and integrated in the first wave of the 2014 national household survey. It was administered to a sub-sample of youth between 15 and 25 years old. The entire sample of youth in the ECVMA survey amounts to 3,276 individuals, including the full roster of youth identified by the surveyed households and those who did not complete the survey. The present chapter is the first study of aspirations in a large, nationally representative sample of youth in an African country, and one of the first globally.9 This chapter uses the definition of aspirations put forward by P. Macbrayne (1987): “an individual’s desire to obtain a status object or goal such as a particular occupation or level of education�. Following the structure of the ECVMA 2014 Survey, we analyze youth aspirations along three dimensions: (i) Occupational Aspirations (Brunello and Schlotter, 2011), (ii) Short and medium-term practical aspirations, and (iii) Financial Aspirations. This analysis is followed by a brief discussion of some of the potential determinants of aspirations including educational achievements, parental aspirations and finally a sub- section on the importance and relevance of role models. Internal and external constraints may hinder fulfillment of youth aspirations. In this chapter we focus on one main internal factor that may constrain fulfillment, namely non-cognitive skills. Non-cognitive skills are argued to influence the extent to which youth find and take advantage of educational and employment 9 Exceptions globally include the UNDP study in Armenia (UNDP, 2012) and the USAID study of aspirations in rural Pakistan (Kosec et al 2012). 86 opportunities (e.g. Jacob, 2002; Heckman et al 2006). Thus, low non-cognitive skills may hinder youth in overcoming barriers to entry in education or formal sector employment. The discussion on non-cognitive skills is following by a section on external factors that are likely to constrain occupational dreams. These factors include intergenerational rigidities, family pressure and constraints to migration, as well as access to support programs that may help improve youth’s skillset in order to meet their desired future occupational goal. Data and analytical method The analysis is based on a sub-sample of 3,222 youth observations. Youth is defined as an individual between 15-25 years old. The sub-sample was identified based on answers to the most central questions in the survey module on aspiration and non-cognitive skills (i.e. sections 15 and 16). Thus, individuals with missing information for central measures of aspiration and non-cognitive skills are not considered as part of the sub-sample. This ensures comparability across the different sections of the survey module analyzed here. The chapter focuses on descriptions of youth aspirations and simple correlations. The cross- sectional data and the endogenous nature of socioeconomic characteristics such as education, occupational choice and skill level do not allow us to identify causal impacts between youth aspirations and socioeconomic status. Thus, results throughout this chapter should be interpreted with caution and in terms of simple correlations. Instead we focus on between-group comparisons including gender, educational level and rural-urban divide. The sample includes more young females than males. Error! Reference source not found. shows demographic characteristics for the sample both with and without the use of sample weights. 47 percent of the sample is male and 53 percent is female, suggesting that females are slightly over-represented in the youth sub-population. Using weights, the distribution between men and women become even more skewed towards females. The respondents in the youth sample are split between urban and rural areas. Specifically, approximately half of the youth live in rural areas and the other half in urban centers. Related to the discussion provided in Chapter 1 the sample disproportionately represents urban youth, as two out of ten people live in urban centers of Niger. For this reason, the results reported in the rest of this chapter is weighted using representative population weights. The weighted numbers reveal that almost three out of every fourth youth live in rural areas. Compared to the total population a larger share of the youth population resides in urban areas. The ethnic distribution of the sample is similar to the lager sample of households surveyed. The sample is 43 percent Haoussa, 26 percent Djema, 15 percent Touareg, and 16 percent “Others� which includes six ethnic minorities (Arab, Goumantche, Kanouri-Manga, Peul, Toukou, other Nigerien ethnicity, and foreigners). The ethnic distribution of the sample slightly underrepresents the proportion of Haoussa, who make up 51 percent of the Nigerien population, while it slightly over-represents the proportion of Djema and Touareg, who make up 23 percent and 13 percent of the Nigerian population, respectively. The populations weights help adjust for some of these biases. 87 Table 5.1: Demographics of Youth Sub-Sample (%) Weighted (W) Unweighted (UN) Gender Female 53.2 51.8 Male 46.8 48.2 Area of residence Rural 72.4 49.5 Urban 27.6 50.5 Ethnicity Djema 20.0 25.6 Haoussa 56.1 43.4 Touareg 13.0 14.8 Others 10.8 16.2 Educational level None 54.0 41.1 Primary 15.7 16.8 Secondary first cycle 30.3 42.1 Household position Head of household 3.4 2.8 Spouse of head 20.0 15.4 Child 55.7 59.9 Other 20.9 21.9 Youth sub-sample (15-25 years old) 2,243,804 3,222 Slightly more than half of the Nigerien youth have some education. The numbers reported in Table 1 include youth that are still in school, as well as youth that are no longer in school. Considering the unweighted numbers, 41 percent of the respondents have no education compared to 61 percent in the total population. 17 percent have primary education, 42 percent have secondary education, suggesting that the youth are generally better educated. However, it is important to keep in mind that these numbers do not say anything about youth educational abilities due to high rates of grade repetition, high absentee rates, and generally low schooling quality. Male youth have greater access to education than female youth. Of the youth with no education that completed the survey, 61 percent are females and 39 percent are males. Similarly, 47 percent of female youth have not received any education, compared to 34 percent of males. The disparity in access to education among males and females in Niger is consistent with the findings of Chapter 1 as well as other studies of youth employment in Sub-Saharan Africa (Filmer and Fox, 2014). However, the results suggest that the proportion of youth with no education in Niger is higher than the continental average, in which 25 percent of females and 15 percent of males have received no education (ibid., 69). Of the youth that participated in the survey, only a minority are the head of their household. A larger proportion of the youth are the spouse of the household head, and the majority of the youth who 88 participated in the survey are identified as children less than 15 years old in the household in which they live (60 percent). Youth in Niger have low levels of unemployment, similar to the population as a whole. As shown in Figure 5.1: Employment status 1, unemployment among youth is one percent. However, more than half of the active youth population is underemployed, suggesting that despite the low unemployment rate, youth may still not be earning enough money to meet their needs, or not working even though they desire employment. In addition, nearly half of the youth population is out of the labor force (47 percent), which includes students as well as youth who are not looking for jobs. Those who are not looking for jobs could be engaged in work within the household that they are not directly compensated for. Figure 5.1: Employment status Youth Sub-Sample W: N=2,243,804 UW: N= 3,222 Out of the Labor Force Active W: 816,495 (36%) W: 1,427,309 (64%) UW: 1,528 (47%) UW: 1,694 (54%) Inactive Student Employed Unemployed W: 482,595 (59%) W: 333,900 (41%) W: 1,418,399 (63%) W: 8,910 (0.4%) UW: 666 (44%) UW: 862 (56%) UW: 1,671 (52%) UW: 23 (0.7%) Underemployed W: 768,091 (54%) UW: 825 (49%) The majority of employed youth work in the agricultural sector. As shown in 89 , more than three-quarters of the active youth population work in agriculture, 10 percent is self- employed, and a minority is receiving a formal wage. These figures are consistent with the population in Niger as a whole. Since agricultural work in Niger includes small-scale and subsistence farming, youth may not receive a consistent income or associated employment benefits. 90 Figure 5.2: Primary Sectors of Employment 4% 10% Agriculture 86% Self-Employed Non-Ag Wage Occupational aspirations Aspirations are measured along multiple dimensions. First, aspiration is measured as occupational/employment aspirations, measured both at childhood as well as at young. Second, we can measure aspirations as medium-term practical aspirations. This measure is related to the aspirations of youth to complete short-term projects. Finally, this section considers financial aspirations based on youth aspired income and expected future income levels. Each of the aspiration measures are linked to socio-demographic characteristics including gender, educational level and geographical location. Aspired employment type in childhood Aspired employment type is measured as youth aspirations in childhood. The survey asked youth “Which job did you want to do when you were a child?� The reported aspired occupational categories were then grouped into seven categories: (i) Agriculture, (ii) Trade or craft work, (iii) Teacher or doctor, (iv) Government official, (v) Senior executive, (vi) No plan, and (vii) Do not remember. As children, girls have a lower capacity to aspire than boys. Out of the 3,222 youth in the sub- sample, 60 percent remembered their aspired occupation as a child (Error! Reference source not found.3). Male youth are more likely to remember their occupational aspirations as children, while a larger share of the female youth either had no aspirations as children or do not remember their occupational aspirations while in childhood. The lack of direction in term of occupational aspirations for females could be linked with higher rates of labor market inactivity among females or the lower rates of educational attainment. A similar pattern is found looking at the difference between youth living in rural and urban areas: children in urban areas are more likely to have well-defined occupational aspirations. 91 Figure 5.3: Did you aspire to a certain job as a child? Urban Rural Does not remember No aspiration Male Yes Female 0 20 40 60 80 Youth aspire for high-skilled formal sector employment. Conditioning on aspiring for a certain job as a child, Figure 5.4 shows the distribution of aspired jobs by gender. More than two-thirds of the female youth aspire to work as a teacher or a doctor. This compares to only one-third of the male youth. Rather male youth aspire to work in the public sector as government officials and senior executives. Societal norms regarding gender roles and what occupations are considered acceptable for females and males may explain this variation. While we find sectoral differences in aspirations, there is no substantial gender difference in the type of jobs aspired to: youth aspire for high-skill, formal sector jobs. Figure 5.4: If aspired to a certain job as child, which job? 70 60 50 40 30 20 10 0 Agriculture Trade/craft Teacher/doctor Government Senior Exec official Female Male 92 A larger share of the rural youth aspires to work in the informal sector. One-third of the rural youth aspire to work in agriculture, trade or craft. This compared to less than 12 percent of the youth living in urban areas. While we find no difference in the share of rural and urban youth that aspire to work in middle- level skill occupations as teachers or doctors, urban youth are substantially more likely to aspire to work as government officials in the public sector (figures not shown). Another way to look at youth aspirations is to consider what job youth would choose if they had the opportunity to change occupation. The categories of employment include the following four categories: (1) Agriculture, (2) Non-agricultural self-employment, (3) Formal private employment, and (4) Formal public employment. Similar to childhood aspirations, youth current aspirations reveal that youth value formal employment. In fact, more than 50 percent on average aspire to work in the public sector. Aspirations are at odds with the reality of employment structures. The desire to work in the public sector is likely to go unfulfilled as less than 1 percent of the Nigeriens currently work in the public sector. More than 90 percent of youth aspire to work in non-agricultural sectors, yet about 80 percent will most likely work in the agricultural sector. This shift away from agriculture is consistent with other recent studies on youth employment preferences and aspirations, despite limited opportunities for non-agricultural wage or formal employment (e.g. Nwagwu 1976; Leavy and Smith, 2010; Filmer and Fox, 2014). For instance, Nwagwu (1976) reports that, in spite of awareness of tight labor markets and limited economic opportunities, students in Nigeria and Kenya maintain high aspirations and high expectations for their future employment. Current occupational preferences are correlated with educational level and geographical location but not gender. Figure 5.5 report employment sector aspirations by educational level and gender. Overall, education increases aspirations for formal sector employment. Thus, employment in agriculture and non- agricultural self-employment is more preferred by youth with no education. Considering the difference in gender across aspired employment by education, we find no significant gender gaps. In terms of the rural- urban divide, we find that more urban youth would move into self-employed if they had the opportunity to change job while rural youth are relatively more likely to value formal public sector employment (figures not shown). Not surprisingly, rural youth are generally more likely to aspire for employment in the agricultural sector compared to urban youth. 93 Figure 5.5: Aspired job by gender and educational level 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Female Male Female Male Female Male No edu. Primary edu. Secondary edu. Agriculture Self-employment Formal private Formal public The aspirations gap increases with educational level (Figure 5.6). Assuming that those who aspire to be farmers are currently farmers, between 70-80 percent of the youth employed in the agricultural sector would change occupation if they could. Considering aspirations by educational level it is evident that aspirations rise faster than opportunities: The gap between occupational aspirations and actual attainment increases with education meaning that those with secondary education have higher aspirations for formal sector jobs relative to those with no or only primary education. Figure 5.6: Occupational aspirations vs. actual attainments 100% 80% 60% 40% 20% 0% Aspired Actual Aspired Actual Aspired Actual No edu. Primary edu. Secondary edu. Agriculture Self-employment Formal private Formal public 94 Medium-term aspirations Young men have better defined medium-term aspirations than young women. Medium-term aspirations are measured as occupational aspirations within the next 5-year period. In general, according to Figure 5.7 youth aspire to find a good or better job in the near future or resume studies. Yet men are more likely to express positive aspirations for their work, than women. More than one-fifth of the female youth do not know what they would like to do in the next 5 years. This compares to less than 10 percent among the male youth. Figure 5.7: Aspiration over next 5 years (restricted to out-of-school individuals) 50 45 40 35 30 25 20 15 10 5 0 1. Find a 2. Continue 3. Resume 4. Grow 5. Does not Other good or current studies activity know better job work Female Male Young Nigeriens aspire to upward mobility through better jobs and better education, even when out of the labor force, with men reporting higher aspirations than women. As shown in Table 5.2, the majority of employed youth aspire to find a better job or resume studies, but women are more likely to have undefined aspirations than men. Among those out of the labor force, male youth are almost twice as likely to aspire to continue or resume studies, reflecting higher school participation in the age bracket considered. Even though women have less defined goals than men when out of the labor force, a clear majority still expresses their willingness to pursue occupational objectives. Continued inactivity does not appear as a positive choice for women. Table 5.2: Medium-term aspirations by employment status (%) Employed Out of labor force Female Male Female Male 1. Find a good or better job 34.8 43.7 27.9 32.4 2. Continue current work 10.5 8.5 2.8 2.8 3. Resume studies 15.5 17.6 29.8 53.5 4. Grow activity 6.2 7.6 1.2 0.9 5. Does not know 18.5 9.1 23.6 8.4 6. Other 14.5 13.5 14.7 2.0 95 Financial aspirations Financial aspirations are measured by youth aspired income and expected future income. Youth were asked “Which income level would you like to achieve?� and “Which income do you think you will have in 10 years?� Answers to questions regarding aspired income was only considered if the level stated were equal to or higher than the current reported income level. To ensure comparability between the different questions, the reported numbers of financial aspirations is limited to individuals that answered all income questions (i.e. current, aspired and future income). The following analysis on financial aspiration is based on 1201 youth observations. The aspiration gap is positive and large. Ray (2006) defines the aspiration gap as the difference between the aspired standard of living and the standard of living that one already has. He argues that is it not the aspirations per se, nor one’s standard of living that affects future -oriented behavior, but the size of the gap. Overall, 92 percent of the youth aspire to earn an income higher than their current income level, whereas 8 percent report no difference between their current and aspired future income. Out of those that aspire to earn a higher income, 50 percent aspire to earn at least 9 percent more compared their current income level, while 10 percent aspire to earn at least 55 percent more. Figure 5.8: Aspired income by gender and geographical location 35 35 30 30 25 25 20 20 15 15 10 10 5 5 0 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Female Male Rural Urban Males have higher income aspirations than females. Men’s median aspired income is 3 times higher than women’s. Specifically, Figure 5.8 show that 57 percent of the female youth aspire for an income in the lowest two quantiles. This compares to 36 percent among the male youth. By contrast, males are more than twice as likely to aspire for an income in the highest two quantiles compared to females. 96 Youth living in rural areas have lower income aspirations on average. Urban you th’s median income is 3 times higher than rural youth’s. According to Error! Reference source not found.8, a larger share of youth living in urban areas aspire for an income in the highest three quantiles compared to rural youth. More than 50 percent of the rural youth aspire for an income in the lowest two income quantile. This compared to 28 percent of the youth in urban centers. Financial aspirations generally increase with educational level. Figure 5.9 shows that youth with at least secondary education are more likely to aspire for higher income levels relative to youth with no or only primary education. As aspirations are shaped in part by perceived opportunities and experiences, the results suggest that youth with higher education may feel that they are more capable to achieve a more lucrative occupation in the formal sector or become more successful entrepreneurs. Surprisingly, no significant difference in the share of youth with primary and secondary education is found for aspirations levels in the 4th income quantile. Figure 5.9: Income aspirations by educational level 40 35 30 25 20 15 10 5 0 Q1 Q2 Q3 Q4 Q5 No edu. Primay edu. Secondary edu. Determinants of aspirations Youth aspirations are shaped by the aspirations window, i.e. the environment from which youth draw aspirations. An individual is said to draw aspirations from the lives, achievements or ideals of those that exist in the aspirations window (Ray, 2006). We analyze three important determinants of youth’s aspirations window in Niger: (i) Educational achievement, (ii) Parental educational level and aspirations for their children, and (iii) presence of role models. Educational achievements Educational achievements determine youth aspirations. Occupational aspirations were found to increase in educational achievements. Considering medium-term aspirations over the next 5 years we find that youth with at least secondary education would like to resume or continue their studies. In contrast the majority of youth with no or only primary education aspires to find a good or better job in the near future. 97 Figure 5.10: Aspirations by educational achievement a. No education 60 50 40 30 20 10 0 1. Find a 2. Continue 3. Resume 4. Grow 5. Does not Other good or current work studies activity know better job b. Primary education 50 40 30 20 10 0 1. Find a 2. Continue 3. Resume 4. Grow 5. Does not Other good or current work studies activity know better job c. Secondary education 60 50 40 30 20 10 0 1. Find a 2. Continue 3. Resume 4. Grow 5. Does not Other good or current studies activity know better job work Female Male Education is correlated with more clarity on future plans for everyone. Error! Reference source not found. presents determinants of no childhood aspirations as well as medium term aspirations.10 First, we find that females are more likely to have no plans for the future, i.e. no medium or childhood aspirations. Second, urban youth are less likely to have no occupational aspirations plans. Third, youth with primary or secondary education is positively associated with having future occupational plans both as children and as 10 It should be noted that educational achievement is likely to be endogenous to not having any defined occupational aspirations. Education is considered to be endogenous since the direction of influence is unclear: While education may influence the aspirations of youth, no aspirations could also affect youth’s academic success. 98 young. Finally, we interact the female dummy variable with educational attainment to allow for varying effects from schooling between females and males. The interaction is statistically significant for secondary education across both dependent variables, i.e. columns 2 and 4. However, the joint test is only statistically significant for primary education in column 4. This indicates that primary education leads to significant decrease in lack of occupational aspirations relative to not having any education. Table 5.3: Determinants of no aspirations (1) (2) (3) (4) No aspirations next 5 years No aspirations as child Female (=1) 0.065*** 0.146*** 0.096*** 0.228*** (0.011) (0.022) (0.018) (0.029) Primary education (=1) -0.073*** -0.015 -0.056** -0.019 (0.017) (0.020) (0.026) (0.035) Secondary education (=1) -0.088*** -0.017 -0.159*** -0.028 (0.016) (0.018) (0.025) (0.031) Urban (=1) -0.095*** -0.093*** -0.068*** -0.060*** (0.014) (0.014) (0.022) (0.022) Female*Primary education -0.104*** -0.050 (0.033) (0.049) Female*Secondary education -0.138*** -0.267*** (0.024) (0.037) Constant 0.162*** 0.122*** 0.460*** 0.399*** (0.017) (0.018) (0.026) (0.028) Controls Yes Yes Yes Yes Observations 3,222 3,222 3,222 3,222 Adjusted R-squared 0.102 0.110 0.078 0.093 P-value: Female + Female*Pri. edu. 0.113 0.000 P-value: Female + Female*Sec. edu. 0.532 0.103 Note: Linear Probability Model. Control variables include employment status, and ethnicity dummies. Robust standard errors in parenthesis. ***, ** and * indicates significance at the 1%, 5% and 10% level, respectively. Parental achievements and aspirations Nigerien youth are better educated than their parents. Parents are argued to be one of the most important factors influencing children as they may serve as reference points, mentors, and support systems. Hence, parents are likely to be at the center at youth’s aspiration window. If parents are one of the most important influences on their children, then a mismatch between parents and youth’s education level may anchor youth aspirations at relatively lower levels. The vast majority of youth’s parents in Niger have received no education, and the reported educational level of mothers is similar to fathers (Figure 5.11). Thus, the intergenerational education gap is – like many other African countries – large as almost 50 percent of youth have received some education, and more than a quarter is currently enrolled in school. 99 Figure 5.11: Parental Educational Background (%) 100 90 80 70 60 50 40 30 20 10 0 None Primary Secondary Mother Father Youth Parent’s aspirations influence and help shape their children’s aspirations. According to the youth reported answers shown in Figure 5.12, around half of the parents have a defined occupational aspiration for their children. Parents generally have lower levels of aspirations for girls: 51 percent of the parents have no aspirations for their daughters (i.e. no aspirations plans/or awareness by the daughter or no work). Out of those that had defined aspirations for their children, the majority would like them to get a job in the public sector as a teacher or a doctor (see Figure 5.13 ). In terms of gender, we find some gender differences in sectoral aspirations, but no substantial differences between high and low-skilled jobs. This pattern is consistent with youth own aspirations pattern by gender discussed above. Figure 5.12: Did your parents have a defined occupational aspiration for you? 100 60 50 40 30 20 10 0 Had defined No aspirations A job I like That I do not work aspirations Female Male Figure 5.13: If parents have a defined aspiration, which job did they want for you? 70 60 50 40 30 20 10 0 Agriculture Trade/craft Teacher/doctor Government Legislators and worker Senio Female Male Role models According to the literature, role models are correlated with high aspirations among youth. Despite the lack of obvious correlation between parent’s and youth educational level, other role models are likely to impact youth aspirations window. Two experimental studies point to the link between role models and an increase in aspirations, which translates into human capital outcomes. First, in a study conducted by Macours and Vakis (2009), interactions with leaders can increase aspirations and result in an increase in human capital investment and productive investment. Another experiment conducted in India found that female leadership increases educational attainment among female youth (Beaman et al. 2012). 101 Role models are predominantly male, especially in rural areas. 52 percent of the youth know somebody that is successful, i.e. have a role model. Having a role model varies by gender and geographical location. A greater proportion of males and urban youth have a role model, consistent with their higher aspiration levels. Figure 5.14 shows that role models are predominantly males, especially in rural areas and for boys. Role models are from the same community as the youth themselves. Thirty-seven percent of role models are members of the youth’s family, 11 percent are defined as friends, and 49 percent as members of the community. In addition, nearly half of youth with role models report that the role model shares a similar socio-economic status, while 29 percent state that the role model is from a wealthier family. Finally, a quarter of youth reported that the role model is from a poorer family compared to their own. 102 Figure 5.14: Gender of role model 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Female Male Rural Urban Male Female The occupation of role models differs by geographical location. While more than 90 percent are farmers in rural areas, only 5 percent of the role models main occupation is farming. Rather, rural youth role models are more likely to be self-employed in trade or craft work. A similar picture is found for urban youth where less than 1 percent is employed in the public sector as government workers while more than 35 percent of the role models are employed as public sector workers. Significantly less variation across occupational status of role models are found between female and male youth. Across both genders, non- agricultural self-employment is found to be a pathway to success. Figure 5.15: Occupation of role model by geographical location 50 45 40 35 30 25 20 15 10 5 0 Trade/craft Teacher/doctor Government Legislators and Agriculture Don't know worker Senio Rural Urban 103 However, according to Figure Error! Reference source not found.5.16, the main perceived driver of role model’s success is luck: Almost 50 percent of the youth highlight this reason. The perceived reasons of role models successes are very stable across gender, geographical location and educational levels, suggesting that youth might feel that change/luck is determining their success in life. This points to possible internal constraints to aspiration fulfillment. Figure 5.16: Perceived reason for role model’s success 60 50 40 30 20 10 0 Was lucky Worked Is Parental Good Education Initial Migrated hard intelligent support connections money Internal constraints to fulfillment Fulfillment of occupational aspirations is likely to be constraint by youth’s level of non -cognitive skills. Non-cognitive skills influence the extent to which youth find and take advantage of educational and employment opportunities (e.g. Jacob, 2002; Heckman et al 2006). Low non-cognitive skills may explain whether youth are able to overcome challenges or barriers to entry in education or formal sector employment. Therefore, understanding the types and nature of non-cognitive skills is important for designing effective interventions. Following the broader literature, non-cognitive skills are measures along two dimensions, both of which relate to youth’s attitudes towards their selves and the society (e.g., Heckman et al., 2006, Osborne-Groves, 2006). First, non-cognitive skills are measures as locus of control, which measures the degree of control individuals feel they possess over their life. Second, this chapter considers non-cognitive skills as perception of self-worth using a measure of youth self-esteem. Locus of control is measured according to three dimensions using a model defined by Levenson (1981). According to this model, locus of control is measured along three scales —Internality, Powerful Others, and Chance. A person's "locus" is conceptualized as either internal (i.e youth believe they can control their life) or external (i.e. youth believe their decisions and life are controlled by environmental factors which they cannot influence, or by chance or fate). Youth were asked 15 questions related to control that were adapted from Levenson’s model. Questions were divided into three categories depending on whether they were related to Internality, Powerful Others, or Chance. Responses were presented in a Likert scale format ranging from 0 as “Strongly disagree� to +3 as “Strongly Agree.� Index variables were created 104 for each of these sub scales. The values of each index ranges from 0-15, with a higher score indicating that the respondent have strong internal locus of control over events and that their life derive primarily from their own actions. Contrary, lower scores indicate that respondents have strong external locus of control, meaning that they praise or blame external factors for events or general life circumstances. The Internality index measures the extent to which people believe they have control over their own lives, while Powerful Others and Chance measure the extent to which people believe powerful others have control over their lives, or the extent to which their lives are controlled by chance, respectively. Mean values for each of the three measures are reported in Error! Reference source not found.. Youth perceive their life as controlled by chance as much as by themselves. On average, youth feel that powerful others have less external control over their lives, compared to chance. However, youth feel that their lives are determined by chance just as much as they feel they have control over their own lives. The picture is unchanged considering gender, the urban-rural divine and youth’s educational level. On average, urban youth, males and youth with more education have a greater sense of control. The average internality score for males is 10.3 while 9.3 for females, with higher values representing greater internal control over one’s life. Similarly, the internality score suggests that urban youth have more internal control than rural youth, but also feel that their lives to a larger extent is determined by chance. Considering locus of control by educational level, we find that reported internality increases with educational level, however chance is also a more important determinant for youth with some education (primary or secondary). Table 5.4: Internal constraints to fulfillment: Locus of control and self-esteem Locus of control Self-esteem Levenson scale (0-15) Rosenberg Chance Powerful others Internality scale (0-18) All Nigerien youth in sub-sample 9.7 6.1 9.8 11.9 Gender Male 9.7 6.0 10.3 12.1 Female 9.7 6.1 9.3 11.7 Area of residence Rural 9.4 6.0 9.5 11.9 Urban 10.6 6.1 10.5 12.1 Educational level None 9.1 5.9 9.5 11.9 Primary 10.4 6.8 9.6 11.9 Secondary cycle 10.4 5.9 10.4 12.1 Self-esteem is measured according to the Rosenburg self-esteem scale.11 Responses are ordered according to a Likert scale ranging from Strongly Agree to Strongly Disagree. Responses were recoded according to the scale such that “3� was awarded to the response indicating the highest level of self -esteem, and “0� was awarded to the response indicating the lowest level of self-esteem. The survey contains 6 out of the 10 questions of the Rosenburg self-esteem scale. An index variable was created for “self-esteem� 11 “Rosenburg’s Self-Esteem Scale,� http://www.wwnorton.com/college/psych/psychsci/media/rosenberg.htm (accessed June 23, 2015). 105 with values ranging from 0-18, with 18 indicating the highest level of self-esteem. Scores from 9-15 is defined as the normal range of self-esteem, which is proportional to the normal range defined by Rosenburg based on the number of questions. Mean values for self-esteem are reported in Error! Reference source not found. by gender, geographical location and educational attainment. Nigerien youth self-esteem is not abnormal. On average, the self-esteem of youth is 11.9 (SD 2.9), which constitutes a score in the “normal� range according to the Roseburg scale. Average self-esteem varies only marginally within demographic sub-groups. Males and urban youth have slightly higher self-esteem, consistent with their higher levels of occupational aspirations. Finally, self-esteem does not vary by educational attainment, though youth with secondary education tend to have a slightly higher self-esteem compared to youth with no education or only primary school. External constraints to fulfillment External constraints to fulfillment of aspirations goals are measured along three dimensions: family pressure, external constraints to migration, and access to support programs. Overall we find that family pressure is a considerable constraint to free movement and occupational choice, hindering youth fulfillment of their aspiration goals. However, the analysis on intergenerational rigidities suggests that education may be one of the ways to escape the social heritage to follow employment opportunities in the non-agricultural sector. Finally, support programs provide an opportunity to gain practical skills not necessarily obtained by school enrollment for more years. Family pressure toward intergenerational reproduction Almost half the youth feel that they have to do the same job as their father, mother or both. Males are more likely to feel obligations towards their father’s occupational discipline (27 percent), while 19 percent of the female youth feel that they should do the same job as their mother (Figure 5.17). In line with the higher aspirations levels is lower levels of intergenerational rigidity, the majority of urban youth do not feel obligated to follow their parent’s occupational choice (75 percent). This compared to 48 percent of the rural youth. Youth feel obligated to follow their parent’s occupational choice. Figure 5.18 reports reasons for why youth feel the need to do the same job as their father/mother. The main reason is a sense of obligation; parents wants me to or the youth feel that they cannot let their parents down. The second most important reason, especially for urban youth is choice, while lack of opportunities is the least cited reason – particularly by male and urban youth. This might explain the relatively lower intergenerational mobility found in the previous sub-section compared to other African economies. 106 Figure 5.17: Do you feel you need to do the same job as your father/mother? 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Female Male Rural Urban Neither Both Your father Your mother Figure 5.18: Why do you feel you need to do the same job as your father/mother? 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Obligations Choice Lack of Obligations Choice Lack of to parents opportunities to parents opportunities Female Male Rural Urban Constraints to migration Mobility may be a way to overcome intergenerational rigidities. Almost 50 percent on average think that migration is necessary to fulfill aspirations. In line with the higher aspirations, boys are more likely than females to state that a move out of their current place is necessary to find a good job (Figure 107 5.19). Not surprisingly, rural youth are more likely to think migration is necessary suggesting that urban centers are believed to provide youth with better job opportunities. Figure 5.19: Need to move out of this place to find a good job? 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Female Male Rural Urban No Yes No Yes Girls feel constrained by their parents or husbands. According to Figure 5.20, the most important reason for girls not to move out of their current place of residence is lack of approval from parents or husbands. For males, the main reason for not migrating is that they cannot leave their family or feel constraint by their parents or wife. These findings are also echoed in Figure 5.21, where parents’ reaction to migration within the next 2 years. 36 percent strongly discourage their children to migrate to a bigger city to find a job. This compared to 33 percent on average that strongly encourage movements. Again, parents are much more likely to strongly encourage their boys to migrate (48 percent) compared to their girls (19 percent). Figure 5.20: Could you move out of this place to find a good job? 108 100 90 80 70 60 50 40 30 20 10 0 Yes No: I can't leave No: my No: too costly No: other my parents/spouse personal Female Male Figure 5.21: Parents' reaction to plan to move within 2 years 60 50 40 30 20 10 0 Strongly Encourage Discourage Strongly Doesn't know encourage discourage Female Male Access to support programs Youth demand various types of training to fulfill their occupational goals. For girls, soft skills training is in particular demand (i.e. learn “how to behave in the workplace and talk to people such as clients, suppliers and employers�). This may reflect actual deprivation of these skills but also gender norms and aspiration levels, with girls less likely to project themselves in technical jobs. For boys, training in management and technical skills are in high demand. Surprisingly little demand for financial support and information about employment opportunities. In line with females, rural youth have a relatively higher demand for soft skills related to teamwork and customer service, while urban youth demand more technical skills. Figure 5.22: Desired support programs by gender 109 50 45 40 35 30 25 20 15 10 5 0 Literacy Training in Technical Teamwork Information Financial management skills and on support customer employment Female Male Youth with secondary education demand less soft skills. Youth with none or primary education demand considerably more soft skills, while youth with at least secondary education demand relatively more financial support technical skills and training in management (Error! Reference source not found.22). The break down by education level suggests that one of the skills obtained in secondary education is soft skills related to customer service and how to behave in the workplace. Figure 5.23: Desired support programs by education levels 50 40 30 20 10 0 Literacy Training in Technical skills Teamwork and Information on Financial management customer employment support None Primary Secondary Awareness of support programs raise in educational attainment. Error! Reference source not found.24 show youth knowledge about the existence of support programs by gender, geographical location and educational attainment. Slightly more than 30 percent of the youth state that support programs exist in the area in which they live. The knowledge about support programs increase in education attainment and males are slightly more aware of support programs than females. The latter may be explained by the higher labor market participation rate among boys. The majority of youth have never applied to a support program. Out of those that know that the program exists but have never applied, the majority have not applied between they do not know how to register. Females are more likely to apply and out of those that was accepted the vast majority state that the program was successful. While not reported in Error! Reference source not found.25 we also find that 110 rural youth and youth with no education are more likely to apply for support programs. Of those accepted to the program, more than 90 percent state that the program was very useful. Figure 5.24: Do support programs exist in that area? 50 45 50 45 40 45 40 35 40 35 30 35 30 30 25 25 25 20 20 20 15 15 15 10 10 10 5 5 5 0 0 0 Yes No Don't Yes No Don't Yes No Don't know know know Female Male Rural Urban None Primary Secondary Figure 5.25: Have you ever received any of these programs? 80 70 60 50 40 30 20 10 0 Yes, and it was Yes, but it was not No, my application No, I never applied useful very useful was rejected Female Male 111 References Beaman, Lori, Esther Duflo, Rohini Pande, and Petia Topalova. 3 February 2012. “Female Leadership Raises Aspirations and Educational Attainment for a Policy Experiment in India.� Science Magazine, vol 335. Bernard, Tanguy, Stefan Dercon, Kate Orkin and Alemayehu Seyoum Taffesse. 22 April 2014. “The Future in Mind: Aspirations and Forward-Looking Behavior in Rural Ethiopia. Centre for the Study of African Economies Working Paper Series 2014-16. University of Oxford: Oxford, United Kingdom. Filmer, D. and Louise F. (2014). “Youth Employment in Sub-Saharan Africa� The World Bank, Agence Francaise de Developpement: Washington DC. Furstenberg Jr., Frank F. and David Neumark. April 2005. “School-to-Career and Post-Secondary Education: Evidence from the Philadelphia Educational Longitudinal Study. Discussion Paper No. 1522. Institute for the Study of Labor IZA: Bonn, Germany. Gemici, Sinan, Alice Bednarz, Tom Karmel and Patrick Lim. 2014. “Young People’s Aspirations and Their Occupational Outcomes.� Australian Economic Review, Vol. 47, no. 1, p. 124-136. Heckman, J., Stixrud, N. and S. Urzua (2006). “The Effects of Cognitive and Noncognitive Abilities on Labor Market Outcomes and Social Behavior� Journal of Labor Economics 24 (3), 411-482. Jacob, B. A. (2002). “Where the boys are’t: non-cognitive skills, returns to school and the gender gap in higher education� Economics of Education Review 21, 589-598. Kosec, Katrina, Madeeha Hameed and Stephanie Hausladen. November, 2012. “Aspirations in Rural Pakistan: An Empirical Analysis.� International Food Policy Research Institute: Washington, DC. Kritzinger, A. 2002. “Rural Youth and Risk Society: Future Perceptions and Life Chances of Teenage Girls on South African Farms.� Youth and Society, vol. 33 (4): 545-572. Leavy, Jennifer and Sally Smith. June 2010. “Future Farmers: Aspirations, Expectations and Life Choices.� Discussion Paper 013. Future Agricultures Consortium: Brighton, United Kingdom. Levenson, Hanna. 1981.� Differentiating Among Internality, Powerful Others, and Chance.� In H. M. Lefcourt (Ed.), Research with the locus of control construct, (Vol. 1, p. 15-63). New York: Academic Press. MacBrayne, P. (1987) Educational and Occupational Aspirations of Rural Youth: A review of the Literature. Research in Rural Education, 4(3): 135-141. Macours, Karen and Renos Vakis. November 2009. “Changing Households’ Investments and Aspirations through Social Interactions: Evidence from a Randomized Transfer Program.� Policy Research Working Paper 5137, The World Bank: Washington, DC. Nwagwu, Nicholas A.1976. “The Vocational Aspirations and Expectations of African Students.� Journal of Vocational Education & Training, vol 28: p. 111-115. O’Rourke, Kevin H, and Richard Sinnot. 2006. “The Determinants of Individual Attitudes Towards Immigration.� European Journal of Political Economy, vol 22, p. 838-861. 112 Osborne-Groves, Melissa. 2006. How important is your personality? Labor market returns to personality for women in the U.S. and U.K. Journal of Economic Psychology (forthcoming). Page, Lionel, Louis Levy Garboua, and Claude Montmarquet. 2007. “Aspiration Levels and Educational Choices: An Experimental Study.� Economics of Education Review, no. 26: 748-758. “Rosenburg’s Self-Esteem Scale,� http://www.wwnorton.com/college/psych/psychsci/media/rosenberg.htm (accessed June 23, 2015). 113