Report No. 89837-NE REPUBLIC OF NIGER TRENDS IN POVERTY, INEQUALITY AND GROWTH, 2005-2011 August 2014 Poverty Reduction and Economic Management 4 Africa Region Document of the World Bank __________________________ CURRENCY AND EQUIVALENT Currency Unit = West African CFA Franc (CFA) US$1.00 = 492.00 CFA (August 6, 2014) GOVERNMENT FISCAL YEAR January 1st-December 31st WEIGHTS AND MEASURES Metric System ACRONYMS AND ABREVIATIONS (Franc) CFA West African CFA Franc (Franc de la Communauté Financière Africaine) ECVMA Household Living Conditions and Agriculture (Enquête Nationale sur les Conditions de Vie des Ménages et l’Agriculture) ENBC National Survey on Household Budgets and Consumption (Enquête Nationale sur le Budget et la Consommation des Ménages) GDP Growth Domestic Product FCFA (West African CFA Franc) Franc de la Communauté Financière Africaine INS National Institute of Statistics (Institut National de la Statistique) IUTS Impôt Unique sur les Traitements et Salaires (Labor Income Tax) Vice President: Makhtar Diop Country Director: Paul Noumba Um Sector Director: Marcelo Giugale Sector Manager: Miria Pigato Task Team Leader: Johannes Herderschee ii Republic of Niger: Poverty Note Trends in Poverty, Inequality and Growth, 2005-2011 Table of Contents ACKNOWLEDGEMENTS .................................................................................................................. IV INTRODUCTION ............................................................................................................................... 1 CHAPTER 1: ECONOMIC GROWTH AND POVERTY REDUCTION ..................................................... 4 CHAPTER 2: INEQUALITY AND THE STRUCTURE OF HOUSEHOLD CONSUMPTION ........................ 8 CONCLUSION ................................................................................................................................. 12 ANNEX A: PUBLIC HEALTH AND POVERTY IN NIGER ................................................................. 14 Tables TABLE 1: HOUSEHOLD CONSUMPTION DATA AND GDP GROWTH RATES ........................................................ 4 TABLE 2: TRENDS IN POVERTY INDICATORS, 2005-2011(*) ........................................................................ 5 TABLE 3: GDP GROWTH BY SECTOR, 2005-2011 ..................................................................................... 7 TABLE 4: DISTRIBUTION OF HOUSEHOLD CONSUMPTION BY DECILE ............................................................. 10 TABLE 5: INEQUALITY INDICATORS BASED ON ANNUAL AVERAGE CONSUMPTION PER CAPITA ........................... 11 Figures FIGURE 1: POVERTY/GROWTH ELASTICITY ................................................................................................. 6 Box BOX 1: INEQUALITY INDICATORS ............................................................................................................. 9 iii Acknowledgements This report was prepared by Prospere Backiny-Yetna and Diane Steele. The authors would like to acknowledge Sebastien Dessus, Janet Owens and Sean Lothrop for their important contributions to this analysis. Judite Fernandes provided assistance with document preparation. The estimates reported in this paper were sent for information to the Niger authorities. iv Introduction 1. The ability to accurately monitor poverty trends is crucial to ensure the adoption of effective antipoverty policies and to assess progress toward the achievement of national development goals. Unreliable data or inconsistent measurement techniques can obscure important changes in the nature and characteristics of poverty. But while credible statistics are necessary to evaluate poverty trends, how those statistics are interpreted by policymakers is equally essential, as interpretation largely determines how poverty data will inform future policy decisions. 2. In Niger, efforts to assess poverty dynamics between 2005 and 2011 are complicated by methodological differences in the three household surveys conducted over the period, in 2005, 2007/08 and 2011. A methodological analysis prepared concurrently with the present report1 examines how different survey techniques impact observed poverty trends and attempts to rectify the incompatibility between surveys. Survey-to-survey imputation techniques based on the 2011 survey reveals that the poverty headcount rate fell by 5.5 percentage points between 2005 and 2011, from 53.7 percent to 48.2 percent. However, indicators of the depth and severity of poverty (i.e. the poverty gap and squared poverty gap) exhibited less strong decline the period. 3. While Niger’s overall poverty rate has dropped significantly between 2005 and 2011, changes in the poverty incidence are highly uneven across location types2. Poverty rates have fallen more rapidly in urban centers than in rural areas. As for nonmonetary poverty indicators, urban areas did register the same type of improvements while living conditions in the countryside were left largely unchanged. In addition, the rural economy’s reliance on inherently unpredictable rain-fed agriculture makes it difficult to determine whether the drop in the rural poverty rate reflects underlying structural productivity gains, or merely a temporary increase in annual yields due to strong harvest seasons. Nevertheless, the decrease of 4 percentage points in the rural poverty rate over the 2005-2011 period is clearly linked to the agricultural sector’s annual average growth rate of 6.7 percent. 4. Despite the moderate drop in rural and national poverty rates since 2005, poverty in Niger remains extremely pervasive and severe. Among the major causes of persistent poverty are the country’s minimal economic diversification and extremely limited agricultural infrastructure, which leave the majority of Nigerien households dependent on highly vulnerable farming and livestock production. In addition, the country’s extremely high rate of population growth presents a serious obstacle to sustainable poverty reduction. Not only does Niger have one of the highest population growth rates in the world, but fertility correlates inversely with income level. In other words, the fastest-growing segments of the population are also the poorest, and as a result, the declining trend in the national poverty rate is continuously offset by a steady increase in the absolute number of Nigeriens living in poverty. 1 World Bank (2014) “Measuring Poverty in Niger: Methodological and Analytical Issues.� Washington DC: The World Bank 2 The decline in poverty headcount between 2005 and 2011 is statistically significant at the national level and in urban and rural areas. 1 5. Nevertheless, Niger’s household surveys show modest but significant welfare improvements from 2005 to 2011. Monetary poverty (measured as household consumption) declined by 5.5 percentage points over the period, with poverty falling faster among urban households than among their rural counterparts. Nonmonetary poverty (measured in terms of household living conditions) follows the same patterns, improving more substantially for urban populations, as the rising consumption levels of rural households were not matched by a commensurate increase in the provision of essential public goods and services (see Annex A). However, despite these observed improvements the pace of poverty reduction remains well below the government’s development targets, and Niger is not on track to meet its poverty- related Millennium Development Goals by 2015. Understanding the evolution of poverty in Niger will allow policymakers to improve the scope and impact of their economic growth and poverty reduction agenda. 6. At least four overarching factors have driven recent poverty trends in Niger; the first is broad economic growth, which is boosting household consumption nationwide. From 2005-2011 Niger’s per capita GDP growth rate was low and unstable, yet average household consumption increased substantially during the period. However, while greater economic activity is responsible for the overall reduction in poverty incidence from 2005-2011, increases in household consumption have not been distributed evenly by region, location type, or income level. Moreover, many Nigerien households are highly vulnerable to economic shocks, especially rural and lower-income households, and broad improvements in overall consumption can obscure more ambiguous trends in the depth and severity of poverty. 7. The second, closely related factor is inequality, which seems to be on the rise3. Changes in inequality are being driven by the highest and lowest income groups, with consumption rising fastest among the former and not so much among the latter. Vulnerability to unpredictable shocks appears to be deepening the poverty of already low-income households. And while the increasing prosperity of the upper income groups is not problematic in itself, there is evidence that inequality may be weakening the link between growth and poverty reduction, as wealthy Nigeriens capture an increasingly large share of the returns to economic growth. This is reflected in a troubling decline in the poverty elasticity of growth over the period. 8. The third factor is production volatility in the agricultural sector. Niger’s largely rain-fed agricultural sector is highly susceptible to droughts, floods and pestilence. Agricultural shocks are the most important contributor to adverse changes in the depth and severity of poverty, and the economic damage they inflict on the rural sector slows the pace of poverty reduction and exacerbates inequality. Season-to-season changes in agricultural production also complicate poverty measurement. Farming households comprise the largest share of Niger’s poor population, and as a result poverty rates are highly sensitive to fluctuations in agricultural output. Separating the cyclical impact of variable harvests from the 3 The rising of inequality between 2005 and 2011 should be interpreted with caution. While the 2011 data are survey data, 2005 and 2007/08 are imputed values. Since it is difficult to impute the tails of the distribution, this could contribute to underestimate inequality in 2005 and 2007/08, thus overestimate the rise of inequality for the entire period. 2 underlying structural poverty trends is a major analytical challenge, but it is crucial in order to accurately identify the long-term trajectory of poverty rates. 9. Finally, the fourth factor is Niger’s extremely high rate of population growth, which is coupled with an inverse correlation between household size and income level. Extremely high fertility rates among the poor generate a steady increase in the total number of poor Nigeriens that significantly slows the decline in the national poverty rate. The rapid growth of the rural population is especially problematic, as dwindling access to quality farmland is prompting an unsustainable expansion in cultivated area, which presents a serious long-term risk to poverty reduction, food security and environmental integrity. 10. In addition, the relationship between population growth and rural-urban migration has important implications for poverty trends. Urban fertility rates are lower than rural rates and have been declining over time, while rural fertility rates remain both extremely high and relatively stable. Over the long run urbanization may have the added benefit of slowing nationwide population growth. However, this dynamic would be greatly accelerated by an independent improvement in conditions associated with lower birth rates in rural areas, including sustained increases in household incomes, broad improvements in education indicators, especially among women, and expanded access to healthcare facilities and family planning services (see Annex A). 3 Chapter 1: Economic Growth and Poverty Reduction 11. In principle, national accounts statistics should align with the findings of household surveys, as both are designed to reflect the same underlying economic reality. But this is not always the case, as each of these two data sources has its unique strengths and weaknesses. In the case of household surveys, data collection issues include potential sampling biases and survey errors, while in the national accounts the calculation of consumption per capita may vary depending on the methodology used. The perspective of each approach is also different: national accounts treat the entire country as a single aggregate, whereas surveys attempt to account for disparities in local economic circumstances. Differences in accounting techniques and in the baseline and final periods may also influence the results. 12. Bearing these caveats in mind, a summary of per capita household consumption and per capita GDP is presented below. The growth rate of per capita household consumption substantially exceeds the per capita GDP growth rate across the entire period. From 2005-2008/09 per capita household consumption grew by 5.8 percent, while per capita GDP rose by just 1.3 percent. The observed growth of per capita household consumption then slowed to 4.6 percent over the 2007/08-2011 period, while the per capita GDP growth rate accelerated to 2.5 percent. Despite this narrowing gap, the difference in observed growth rates remains striking. Table 1: Household Consumption Data and GDP Growth Rates 2005 2007/08 2011 Average per capita consumption (CFAF 2011) Urban 298689 361957 377095 Rural 184146 188153 193390 Total 203375 215138 225103 Growth rate for the period 5.8 4.6 Average annual growth rate 2.3 1.8 Per capita GDP (CFAF 2006) Growth rate for the period 1.28 2.52 Average annual growth rate 0.51 0.72 Source: Authors’ calculations based on CWIQ-2005, ENBC-2007/08 and ECVMA-2011, and INS National Accounts figures 13. According to the household surveys, during the 2005-2007/08 period per capita household consumption grew at an average annual rate of 2.3 percent. However, during the 2007/08-2011 period, growth slowed to an annual average of about 1.8 percent. Though modest, these growth rates were significantly higher than those obtained from the national accounts, which recorded an annual per capita GDP growth rate of 0.5 percent and 0.7 percent in each period, respectively. The higher growth rates of per capita household consumption observed during both periods are generally encouraging, as GDP per capita is a relatively crude measure of real income. However, from a pro-poor policy standpoint the central issue is how the growth of per capita consumption affects poverty rates. This depends on the elasticity of poverty in relation to per capita consumption, also known as the welfare indicator. 4 Table 2: Trends in Poverty Indicators, 2005-2011(*) 2005 2007/08 2011 Urban Rural Total Urban Rural Total Urban Rural Total P0 29.6 58.6 53.7 21.8 58.3 52.6 17.9 54.6 48.2 (3.07) (1.78) (1.53) (2.25) (1.98) (1.73) (2.18) (2.24) (1.97) P1 6.9 16.6 15.0 5.0 16.6 14.8 3.6 15.0 13.1 (0.87) (0.60) (0.52) (0.65) (0.87) (0.73) (0.49) (0.93) (0.79) P2 2.4 6.5 5.8 1.7 6.5 5.8 1.1 5.7 4.9 (0.38) (0.31) (0.27) (0.29) (0.45) (0.38) (0.18) (0.46) (0.39) % Population 16.8 83.3 100 17.6 82.1 100 17.3 83 100 % of Poor 9.3 90.7 100.0 6.4 93.6 100.0 6.4 93.6 100.0 # of Poor 624513 6123432 6747945 452724 6591040 7043764 511016 7452615 7963631 Source: Authors’ calculations using data from ECVMA/INS, Niger, 2011 (*) The Standard errors of indicators are in parenthesis 14. Determining the extent to which economic growth translates into poverty reduction is fundamental in designing effective pro-poor development policies. As noted above, Niger’s poverty rate is closely linked to its volatile agricultural sector. Due to the extreme concentration of poverty among the rural population, output growth in agriculture has a direct impact on poverty. However, this connection runs in both directions—strong harvests can generate a rapid reduction in poverty rates, but weak harvests and unpredictable exogenous shocks (adverse weather conditions, invasive pests, etc.) can have a devastating impact on the poor. 15. In Niger, poverty tends to be relatively sensitive to growth, which suggests that the country’s overall economic environment and public policy framework are generally pro-poor. This is especially true for the agricultural sector and the urban informal sector, which together account for the overwhelming majority of the country’s poor. In 2005, the growth elasticity of poverty was estimated at - 1.03 for the overall period. With the poverty rate at 53.2 percent, a 1 percent increase in per capita household consumption produced a drop in poverty of roughly 0.5 percentage points. Furthermore, growth not only had a highly positive impact on poverty incidence, but generated an even more significant improvement in indicators of the depth and severity of poverty. 5 Figure 1: Poverty/Growth elasticity 0 Poverty headcount Poverty gap Squared Poverty gap -0.5 -1 -1.5 -2 -2.5 -3 -3.5 2005-2007/08 2007/08-2011 Source: Authors’ calculations based on CWIQ-2005, ENBC-2007/08 and ECVMA-2011 16. The agricultural sector exhibited strong overall growth during the 2005-2011 period; however, severe drought-related shocks in 2004 and 2008 had a deeply negative impact on the poor. Nevertheless, broad indicators of sectoral growth were highly positive. Agricultural production expanded an average annual rate of 6.7 percent, and due to its overall economic importance agriculture made the largest contribution to GDP growth of any sector during the period. Moreover, the observed decline in rural poverty of 4 percentage points is directly attributable to the solid performance of agriculture. While the sector as a whole performed well over the entire period (with the exception of the two drought years), the livestock sector has lagged behind. Livestock production grew at an average annual rate of just 3 percent between 2005 and 2011, while the sector accounts for as much as 12.5 percent of GDP. The potential of the livestock sector remains largely unexploited. Nigeria and other countries in the region have large and growing consumer markets, but a combination of trade restrictions, high transaction costs and weak transportation infrastructure greatly limit livestock exports. 6 Table 3: GDP Growth by Sector, 2005-2011 Over the period Annual average 2005-2007/08 2007/08-2011 2005-2011 2005-2007/08 2007/08-2011 2005-2011 Primary sector 24.7 9.3 36.3 9.2 2.6 5.3 Agriculture 36.9 8.1 48.0 13.4 2.2 6.7 Livestock 12.1 6.3 19.2 4.7 1.8 3.0 Forestry and fishing -1.8 25.7 23.5 -0.7 6.8 3.6 Secondary sector 3.3 33.0 37.3 1.3 8.5 5.4 Extractive activities -1.0 77.4 75.5 -0.4 17.8 9.8 Manufacturing activities 4.7 19.3 24.8 1.8 5.2 3.8 Electricity, gas and water -10.7 32.7 18.5 -4.4 8.4 2.9 Construction 11.5 28.6 43.5 4.5 7.5 6.2 Tertiary sector 3.7 18.7 23.1 1.5 5.0 3.5 Trading 5.4 17.8 24.1 2.1 4.8 3.7 Transportation and communications 5.0 27.7 34.1 2.0 7.2 5.0 Other services -6.3 8.5 1.6 -2.6 2.4 0.3 Public administration 12.2 23.7 38.8 4.7 6.3 5.6 GDP to factor costs 13.0 15.7 30.7 5.0 4.3 4.6 Source: Authors’ calculations based on INS data 17. The extractive industries also experienced solid growth over the period, with especially high growth rates observed after 2009. Although the extractive industries tend to be capital intensive, import dependent and export oriented, urban households continue to benefit both directly and indirectly from the growth of the sector. The extractive industries are spurring growth in the non-tradable service sector and generating substantial public revenues, which have enabled the government to increase its investment spending, raise public-sector salaries and expand the provision of social services. In all cases these benefits have accrued primarily to urban residents. 18. The extractive industries are expected to make an increasingly important contribution to Niger’s growth over the medium-term. Oil production began by 2011 and is expected to accelerate, while the opening of two new uranium mines during 2013-2014 is projected to make Niger the world’s second-largest producer of uranium. Although the extractive industries are driving positive income and employment dynamics, the sector is not nearly as structurally pro-poor as agriculture. Instead, public policy is the primary mechanism through which growth in the extractive industries translates into poverty reduction. 19. The pattern of economic growth in Niger from 2005-2011 largely explains the observed poverty trends. Robust agricultural production boosted rural incomes and decreased monetary poverty in rural areas, though the limited reach of public services in rural areas left nonmonetary poverty essentially unchanged. Meanwhile, a combination of urban service-sector growth and an expanding public administration (public-sector jobs grew by 5.6 percent over the period) generated improvements in urban living standards, which led to a decline in nonmonetary urban poverty. 7 Chapter 2: Inequality and the Structure of Household Consumption 20. The key element missing from this assessment is inequality. When analyzing trends in inequality it is important to bear in mind that different indicators each have their own advantages and drawbacks. The most popular inequality indicator, the Gini index, measures changes in the overall distribution of incomes. However, other statistical indices focus on specific ranges of the income distribution. The generalized entropy class-inequality indicators vary based on a parameter α, and the properties of the indicator depend on the value of this parameter. When α=0, the indicator is equal to the average logarithm of the variance and is sensitive to variations amongst low incomes. For α=1, the indicator is equal to the Theil index and is sensitive to variations throughout the distribution of incomes. Where α=2, the indicator is sensitive to variations in high incomes (Ferreira and Litchfield, 1999). Considering the complementarity in the properties of the inequality indicators, all four are used in this analysis in order to better understand the evolution of inequality in Niger. 21. Analyzing the distribution of consumption by deciles reveals important features of inequality trends in Niger. As shown in Error! Reference source not found. below, during the 2005- 2011 period household per capita consumption for the lowest three deciles declined while household per capita consumption for the highest three deciles increased sharply. This trend supports previous assessments indicating that the depth and severity of poverty were not improving even as the overall poverty incidence declined4. The situation which is stable during the 2005-2007/08 period even worsened during the 2007/08-2011 period, where the average per capita consumption levels for poorest households is decreasing while a huge increase is registered among the richest. Interestingly, the top 10 percent of households registered the largest increase in household consumption. 4 Ibid . 8 Box 1: Inequality Indicators Considering a distribution (consumption, expenses or revenues) on households or individuals , we assume that the size of the population is , and call ( ) the average of this distribution. The Gini index is the most widely-used inequality index and is based on the following formula: ∑ | | ( ) The Gini index has no essential property and cannot be broken down among different subgroups of a population. The generalized entropy class indices, however, do have this property and are calculated according to the following general formula: ( ) [ ∑( ) ] ( ) For values of the parameter equal to 0 and 1, the Hospital rule is used to obtain the formula. For values higher than or equal to 2, the formula is direct. Consequently, for α=0, we have: ( ) ( ) ∑ ( ) For α=1, we have the Theil index with the following formula: ( ) ∑ ( ) ( ) ( ) And finally, for α=2, we have: ( ) ∑[ ( )] ( ) 22. By the end of the entire 2005-2011 period per capita consumption had increased moderately for households living in extreme poverty (the poorest 30 percent), while consumption levels had increased faster among all other income groups. The highest decile again recorded the largest increase in consumption, and the overall structure of consumption shifted in favor of top-decile households at the expense of those on the lower half of the consumption distribution. In 2005, the poorest 30 percent of the population shared nearly one sixth of the total consumption; six years later this share declined to one seventh. As for 30 percent of the highest income group, the share increased from 49 percent to more than 51 percent. 9 Table 4: Distribution of Household Consumption by Decile Consumption Average per capita consumption (CFAF 2011) Consumption structure Decile 2005 2007/08 2011 2005 2007/08 2011 1 83988 83205 88123 4.1 3.9 3.9 2 111785 111989 118396 5.5 5.1 5.3 3 129776 130976 137856 6.4 6.1 6.1 4 148669 149894 155042 7.3 7.0 6.9 5 166339 168327 175708 8.2 7.8 7.8 6 185532 188085 198686 9.1 8.8 8.8 7 209191 212728 227515 10.3 9.9 10.1 8 240000 248467 262668 11.7 11.5 11.6 9 287425 307749 323767 14.1 14.3 14.4 10 471764 550666 564339 23.1 25.6 24.9 Total 203375 215138 225103 100.0 100.0 100.0 Source: Authors calculations based on CWIQ-2005, ENBC-2007/08 and ECVMA-2011 23. This analysis of consumption trends confirms other measures showing that inequality in Niger increased between 2005 and 2011. All inequality indicators rose during the period; inequality deepened across the entire consumption distribution, and changes were especially pronounced among the highest and lowest income groups. The key question therefore is not whether inequality in Niger is growing, but rather what is driving it. 24. Short- and medium-term inequality trends often reflect the impact of public policies or exogenous shocks that affect the incomes of certain households or influence the prices for specific types of goods. Meanwhile, human capital formation and national economic integration tend to have longer-term effects. Understanding the links between growth and inequality can help policymakers establish a development strategy and policy framework that supports a more equitable distribution of the returns to growth. 25. A number of factors have affected the evolution of consumption and inequality dynamics since 2005. In 2008 a rapid increase in global food and fuel prices caused a spike in inflation. Between 2007 and 2008 food prices in Niger grew by more than 20 percent, while average consumer goods prices rose by more than 10 percent. The urban areas bore the brunt of the price hikes as the rural economy is less monetized; rural households are more likely to consume their own produce, and barter or quid-pro- quo arrangements often substitute for cash-based commerce. Meanwhile, oil prices increased in both 2010 and 2011, before falling in late-2011 when Niger began producing and refining its own oil.5 These successive increases in food and fuel prices tended to widen inequality, especially in urban areas. 26. Changes in tax policy also appear to have contributed to the rise in inequality over the period. In 2010 the government lowered the marginal rate for the single payroll tax (Impôt Unique sur les Traitements et Salaires – IUTS) from 45 percent to 35 percent. Reducing the IUTS benefitted salaried 5 The start of oil production is too recent to meaningfully impact the period examined in this analysis. 10 employees, who tend to be amongst the country’s wealthiest workers. The regressive nature of this tax cut is a likely factor in the widening inequalities observed at the upper end of the consumption distribution. 27. Finally, the periodic shocks to Niger’s agricultural sector continue to disproportionately impact the country’s poorest citizens. The vulnerability of smallholder farmers to recurrent droughts— as well as sporadic flooding, blights, pests and other threats to agricultural production —are a likely source of widening inequalities at the lower end of the consumption distribution. Humanitarian aid and relief efforts on the part of the government and its development partners can mitigate the most immediate effects of environmental shocks on household consumption, the longer-term loss of physical and human capital can prove much more difficult to address. While the rise in consumption among the wealthiest households appears to make a greater contribution to rising inequality than declining consumption among the poorest, from a development perspective the latter phenomenon is far more troubling. Table 5: Inequality Indicators Based on Annual Average Consumption Per Capita 2005 2007/08 2011 Urban Rural Total Urban Rural Total Urban Rural Total Average 298689 184146 203375 361957 188153 215138 377095 193390 225103 Median 247022 167260 176247 287154 168665 177829 304059 173492 186016 Gini 33.2 24.6 28.6 34.1 26.0 31.3 25.1 34.0 31.4 GE(0) 17.9 9.9 13.4 19.0 10.9 16.0 10.2 19.0 16.0 GE(1) 19.0 10.1 14.7 19.6 11.4 17.8 10.3 19.5 17.9 GE(2) 24.4 11.5 19.8 24.2 13.4 24.7 11.6 24.4 25.1 Source: Authors’ calculations based on CWIQ-2005, ENBC-2007/08, and ECVMA-2011 11 CONCLUSION 28. Although Niger’s overall poverty rate has declined in recent years the country’s progress in this area is fragile; poverty remains extremely widespread, and many Nigerien households are highly vulnerable to unpredictable and devastating shocks. In addition, despite the decline in poverty incidence observed since 2005, the absolute number of poor people in Niger continues to increase due to the country’s rapid population growth rate, which is especially high among the poorest households. In 2005 6.7 million Nigeriens lived below the poverty line; by 2011 that number had risen to nearly 8 million. While the modest improvements in both income poverty and some social indicators are encouraging, these trends will not be sufficient for Niger to achieve its poverty-related Millennium Development Goals by 2015. 29. There are many causes behind Niger’s persistently high levels of poverty. The limited diversification of its economy and reliance on rain-fed agriculture is a major contributing factor. The agricultural sector is highly volatile and although productivity growth tends to be strongly pro-poor, the vulnerability of the sector to exogenous shocks poses a constant threat to the sustainability of rural incomes. Low levels of diversification also constrain the economic opportunities available to individuals, and many rural workers are subject to chronic seasonal underemployment. Weaknesses in basic infrastructure further contribute to the narrow focus of economic activity. Finally, Niger’s high fertility rates, especially among poor rural households, tend to exacerbate both poverty and inequality. 30. The persistence of high poverty rates is not inevitable, and over the past several years the authorities have launched a series of initiatives designed to transform the national economy. Among its most ambitious initiatives is the Kandadji dam project, which is expected to provide both electrification and irrigation to a large section of the countryside. Meanwhile, the “3N� program (“Nigeriens Nourish Nigeriens�) is advancing a multi-pronged rural development strategy, although it has yet to be fully implemented. However, the country’s high fertility rate not only limits per capita income growth, but equally exerts constant pressure on public infrastructure and services, and the government’s development efforts should be reinforced by a comprehensive plan to manage population growth over both the medium and long term. In this context, policymakers must continue working to improve the quality of education and health care and extend access to these services, especially in remote rural communities. 31. Measures to stabilize the agricultural sector and boost its production capacity will be crucial to sustainable poverty reduction. The overwhelming reliance of poor households—as well as those just above the poverty line—on a structurally volatile agricultural sector poses a serious threat to Niger’s modest but important progress in reducing poverty. Investment in rural infrastructure, particularly irrigation, and extension services can help to mitigate shocks to agricultural production, but Niger’s unpredictable climate and limited water resources will continue to constrain productivity. In addition, population growth in the rural sector also presents a serious long-term concern, as the country’s most productive farmland is already under cultivation and the continuing expansion of the rural population is forcing younger farmers onto increasingly marginal lands. Left unchecked, this trend will have deeply troubling implications for poverty, food security, and the future of Niger’s fragile Sahelian ecosystems. 12 32. Ultimately, Niger’s extractive industries may present its most valuable opportunity for lasting growth. As the country’s new uranium mines begin production and its oil deposits continue to be exploited, this sector may come to serve as an important source of employment and provide the government with the resources needed to finance the expansion of essential infrastructure and services. However, these positive effects cannot be taken for granted, and Niger’s government must continue to strengthen the institutional and policy framework for natural-resource management if it is to harness the sector’s power to promote diversified, inclusive growth and sustainable, broad-based poverty reduction. 13 Annex A: Public Health and Poverty in Niger In this exercise health variables were not included in the calculation of the welfare indicator, but they can provide an important measure of household wellbeing and reveal a key dimension of poverty. In 2005 about 60 percent of respondents reported that they had recently been ill but did not seek health care; this share declined to around 50 percent in 2011. Inadequate healthcare access may be due to supply-side constraints, i.e. limited availability exacerbated by a high population growth rate and deficiencies in complementary infrastructure. However, this can also reflect demand-side limitations, as even when services are available households may lack the means to pay, or long travel times, low-quality services and other factors may dissuade people from using health facilities. Meanwhile, access to health care by pregnant women has increased substantially, with the number of pregnant women reporting that they had received a prenatal exam rising from 60 percent in 2005 to roughly 82 percent in 2011. However, the percentage of births that take place in health centers or that are assisted by formally trained healthcare personnel remains very low at less than 1 in 3, exposing women and infants to heightened risks of maternal and child mortality. In 2006 the maternal mortality rate averaged 648 deaths per 100,000 live births over the 9 years period prior to the survey. More recent figures are not yet available. The Nigerien health system is extremely limited in terms of the scope and availability of services. According to the World Bank (2005) a pervasive lack of health personnel, especially doctors and midwives, as well as lack of access to healthcare services, especially in rural areas, were amongst the major obstacles to improving the performance of the health system. The study revealed that medical staff was frequently unavailable; only 20 percent of patients met with a doctor during their visit; while 17 percent of pregnant women saw a midwife and just 53 percent saw a nurse. In addition, Niger’s relatively small group of trained medical professional strongly prefer to work in cities, especially Niamey, while remote rural areas have a very difficult time attracting quality healthcare workers. In 2006/07 Niger’s public health budget allocated 25 dollars per person per year, yet only 9.4 dollars per person per year were spent during the period. The health budget stood at 121.8 million dollars in 2008 and remained at virtually the same level in 2010. 14