Report No. 59124-LR Liberia Employment and Pro-Poor Growth November 29, 2010 Poverty Reduction and Economic Management 4, Africa Region Development Dialogue on Values and Ethics, Human Development Network Country Department W1, Africa Region FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS (Exchange Rate Effective October 31, 2010 Currency Unit = Liberian Dollars (LR$) US$1 = LR$72.5 FISCAL YEAR July 1 ­ June 30 ABBREVIATIONS AND ACRONYMS ACPA Accra Comprehensive Peace Agreement BWI Booker Washington Institute CDD Community-Driven Development CWIQ Core Welfare Indicators Questionnaire CFSNS Comprehensive Food Security and Nutrition Survey CfWTEP Cash-for-Works Temporary Employment Project DAW Division for Advancement of Women DDP Development Data Platform DTIS Diagnostic Trade Integration Study ECOWAS Economic Community of West African States EITI Extractive Industries Transparency Initiative FIAS Foreign Investment Advisory Service FAO Food and Agricultural Organization GDP Gross Domestic Product GIS Global Information System GNI Gross National Income HCPI Harmonized Consumer Price Index HHS Household Survey IADB Inter-American Development Bank ICG International Crisis Group IFPRI International Food Policy Research Institute ILO International Labour Organization IMF International Monetary Fund KILM Key Indicators of the Labor Market LAC Liberia Agriculture Company LACE Liberia Agency for Community Empowerment LEITI Liberia Extractive Industries Transparency Initiative LISGIS Liberia Institute of Statistics and Geo-Information Services MEGS Maharashtra Employment Guarantee Scheme MoE Ministry of Education ii MoU Memorandum of Understanding MoYS Ministry of Youth and Sports NaCSA National Commission for Social Action NSAP National Social Action Project (Sierra Leone) NEEP National Emergency Employment Program NGO Non-Governmental Organization NSAP National Social Action Project OECD Organization for Economic Co-operation and Development PPP Public-Private Partnership PRSP Poverty Reduction Strategy Paper SFD Social Fund for Development (Yemen) SME Small and Medium Enterprise SOE State-Owned Enterprise SOL Substitution of Labor and Equipment in Civil Construction SPWP Special Public Works Program (Burkina Faso) TVET Technical and Vocational Education and Training UN United Nations UNMIL United Nations Mission in Liberia UNOPS United Nations Office for Project Services USAID United States Agency for International Development WDI World Development Indicators WDR World Development Report Vice President: Obiageli K. Ezekwesili Country Director: Ishac Diwan Country Manager Ohene Nyanin Acting Sector Director /Sector Manager : Jan Walliser/Miria A. Pigato Task Team Leaders: Errol Graham/Quentin Wodon Acknowledgments This report was prepared by a team led by Errol Graham (AFTP4) and Quentin Wodon (HDNDE) and comprising Rebecca Simson (AFTP4), Chris Jackson (AFTAR), Peter Darvas (AFTED), Emmanuel Fiadzo (AFTP4), Barbry Keller (AFCGH), and Clarence Tsimpo (HDNDE). Glaucia Ferreira (AFTP4) helped prepare the document. Overall guidance was provided by Ishac Diwan (Country Director, AFCW1) and Antonella Bassani (Sector Manager, AFTP4). iii Table of Contents EXECUTIVE SUMARY ................................................................................................ 1 1. EMPLOYMENT IS KEY FOR POVERTY REDUCTION IN LIBERIA ...................................... 3 2. ONE IN FIVE WORKERS IS UNEMPLOYED OR UNDEREMPLOYED ...................................... 7 3. THE CURRENT ECONOMIC STRUCTURE LIMITS PROSPECTS FOR EMPLOYMENT ........... 15 4. TRANSFORM AGRICULTURE TO DELIVER HIGHER-QUALITY JOBS.............................. 22 5. MAJOR CONSTRAINTS TO INVESTMENT AND JOB GROWTH IN THE FORMAL SECTOR ..... 30 6. PUBLIC WORKS ARE NECESSARY FOR THE VERY POOR ............................................. 40 7. EDUCATION AND SKILLS MUST ALSO BE IMPROVED..................................................... 55 List of Tables Table 1.1. Share of the Population in Poverty by Employment Characteristics ................ 4 Table 1.2. Most Important Measures for Government To Improve Living Standards, 2007 (%) .............................................................................................................. 6 Table 2.1. Labor Force Participation and Unemployment Rates, 2007 ............................. 9 Table 2.2. Underemployment Rate, 2007 ........................................................................ 10 Table 2.3. Employment Status in the Main Occupation, 2007 ........................................ 11 Table 2.4. Unemployment Rate (15-64) by Gender, Area, and Welfare Quintile, 2007 . 12 Table 2.5. Distribution of Household Incomes across Taxable Income Groups ............. 12 Table 2.6. Main Activity in the Main Occupation, 2007 .................................................. 13 Table 2.7. Hours Worked per Week by Gender, Area, and Welfare Quintile, 2007 ....... 14 Table 3.1. Projected Employment Growth in Liberia ...................................................... 19 Table 4.1. Rice Production and Yields for Selected African Countries ........................... 23 Table 4.2: Shocks with Negative Impacts on Households in Monrovia ........................... 25 Table 4.3. Agricultural Constraints by Type of Agricultural Household (HHS).............. 27 Table 4.4: Existing Rubber and Oil Palm Concessions .................................................... 28 Table 5.1. Commodity Price Forecast for Liberia's Primary Exports and Imports .......... 33 Table 5.2. Price Volatility of Liberia's Primary Exports, 1990-2008 ............................. 34 Table 5.3. Doing Business Indicators ­ ECOWAS Country Rankings in Sub-Saharan Africa, 2009-2010 .............................................................................................. 35 Table 5.4: Doing Business Indicators--Employing Workers........................................... 37 Table 6.1: Comparison of Project Daily Wages with Minimum/Agricultural Wages (US$) ........................................................................................................................... 45 Table 6.2. Estimates of Project Cost (Wages and Administrative Costs) ........................ 51 Table 6.3 Leakage Rate for Public Works by Wage Rate and Region, 2007 (%) ........... 53 Table 6.4. Potential Impact of Public Works on the Reduction of Poverty and Extreme Poverty (%) ........................................................................................................ 54 Table 7.1. School Enrollment for Selected Pre-War Years .............................................. 55 Table 7.2. Net and Gross Enrollment Rates in Primary and Secondary Schools, 2007 .. 56 Table 7.3. Distribution of the Labor Force by Gender and Literacy Status...................... 58 Table 7.4. TVET Enrollment by Course .......................................................................... 59 iv List of Figures Figure 2.1. Trend in Liberia's Population, 1960­2008 ...................................................... 7 Figure 3.1. Sectoral Contribution to GDP ........................................................................ 16 Figure 3.2. Structure of Liberia's Exports ....................................................................... 17 Figure 3.3. Reasons for Failed Formalization Attempts in Liberia ................................. 21 Figure 5.1. Sectoral Share of Employment ...................................................................... 31 Figure 5.2. Purchasing Power of Minimum Wage .......................................................... 37 Figure 6.1. Distribution of Potential Beneficiaries of Public Works (2007) ................... 53 Figure 7.1. Liberia Distribution of Labor Force by Literacy Status ................................ 57 List of Boxes Box 1.1. CWIQ Survey: Sampling and Quality of Employment Data .............................. 3 Box 2.1. Definitions of Employment, Unemployment, and Underemployment ............... 8 Box 4.1: China's Reduction in Rural and Aggregate Poverty .......................................... 24 Box 5.1. Issues Affecting Liberia's Investment Climate ................................................. 35 Box 6.1. World Bank Experience with Social Funds Involving Labor-Intensive Public Works ................................................................................................................ 43 Box 6.2. Liberia Cash-for-Work Temporary Employment Program ............................... 50 List of Annexes Annex 1. Employment Growth ........................................................................................ 62 Annex 2. Commodity Export Prices ................................................................................ 64 List of Annex Tables v EXECUTIVE SUMARY 1. Fourteen years of civil conflict (1989 ­ 2003) have destroyed Liberia's social and economic infrastructure and brought the economy nearly to a halt. Workers who came of age during the conflict are largely unskilled, and the supply of workers exceeds demand by a substantial margin. The negative effects of unemployment, underemployment, and low productivity on economic growth have made employment the most urgent demand of the population and the top priority for Government action. 2. This report offers guidance to the Government of Liberia in its development of a more strategic approach toward increasing productivity and employment, in order to achieve its pro-poor growth objectives. The main findings and policy recommendations are summarized below. 3. Section 1. Employment is key for poverty reduction. The poor tend to be underemployed, unemployed, or have low-quality jobs. While unemployment and underemployment clearly lead to poverty, the prevalence of low-productivity jobs is an even larger determinant of low standards of living. 4. Section 2. One in five workers is unemployed or underemployed. Liberia's official unemployment rate in its small formal sector is about 5.7 percent. However, there is also a high degree of underemployment in the large low-productivity, low-wage informal sector. Overall, about 20 percent of the labor force cannot find enough work to get out of poverty. 5. Section 3. The structure of Liberia's economy limits prospects for formal sector employment. A few primary export commodities--iron ore, rubber, and timber--account for almost all formal sector employment. Given the limited growth potential of these industries, informality will continue to dominate Liberia's labor market over the medium term. Forty percent of the labor force is currently sustained by subsistence agriculture. 6. Section 4. Transformation of the agriculture sector is essential for pro-poor growth. Higher productivity in domestic agriculture could increase employment and incomes of the rural labor force, improve food security, reduce the food import bill, and ease the burden on women. The plantation sector could expand opportunities for wage labor and help increase the productivity of small farmers by linking them to input and output markets and technical support. 7. Section 5. Investment and job growth in the formal sector are constrained by three main factors. Formal sector employment has not risen above about 20 percent of the labor force since the 1970s. One reason is that the modern sectors are more capital intensive, so growth has not translated to higher employment. Second, growth and employment in the traditional sectors have been negatively affected by commodity price shocks, and Liberia lacks the kind of stabilization fund used by many countries to cushion such shocks. Third, growth in Liberia's manufacturing sector has been constrained by lack of investor protections, poor quality infrastructure, and a low-skilled labor force. 8. Sections 6. Labor-intensive public works programs are necessary for the very poor. Such programs could buy crucial time to generate employment until the private sector expands. However, they must be effectively managed and evaluated to ensure proper targeting and maximum impact. 9. Section 7. Education and training must be improved to enhance employability. The significant de-skilling of the labor force during the civil war has left the economy in a low productivity trap. The gender inequities in education must be addressed, and the technical and vocational education and training (TVET) system needs adequate mechanisms to measure training outcomes. Specific Policy Recommendations Near-term Broad Policy Measures Near-term Actions Medium-term Broad Policy Measures Maintain prudent fiscal and monetary Maintain prudent fiscal and monetary policies to stabilize the macroeconomic policies to sustain strong economic environment. growth and low inflation and spur employment. Undertake labor-intensive public works Evaluate the targeting Strengthen administrative capacity to to provide employment and build performance and impact of existing continue to use labor-intensive public critical social infrastructure. cash-for-work program. works as a rapid response mechanism in economic downturns. Unlock the potential of the agriculture Strengthen land rights through the Improve farm roads to facilitate better sector to provide both short-term and provision of titles or form leases as an access to inputs and to market. long-term employment. interim arrangement. Establish national innovation systems to Formulate the framework for a national Help reduce/mitigate risks faced identify low-risk technologies for outgrower scheme, to enable small by poor farmers through the provision transfer to farmers. farmers to benefit from technical of: support and marketing arrangements · Improved extension services. for cash crops (rubber, oil palm, · High-quality seed and planting timber). material. · Crop loss insurance/automatic social safety nets. · Assistance with mechanical land preparation. Diversify the formal economy to Improve the investment climate Modernize the labor code to create a enhance job opportunities and reduce through: better balance between social protection the economy's vulnerability to external · Improvements in roads, ports, requirements and the need to be shocks. energy, and telecommunications competitive in job creation. infrastructure to support private sector activity · Modernization of the legal framework to facilitate contract enforcement. Increase investments in education and Establish a key role for the private Introduce certification of high-demand training to improve skills and sector in the governance of public skill categories, based on national and employability of the work force. education and training institutions, to regional standards. improve alignment between their curricula and labor market demands. Improve the quantity and quality of Establish the CWIQ or other data available for policy analysis and instrument as a regular survey planning. mechanism, including modules with higher-quality labor market data. -2- 1. EMPLOYMENT IS KEY FOR POVERTY REDUCTION IN LIBERIA 1.1. Liberia is one of the poorest countries in the world, with a per capita gross national income of US$150, about two thirds of the population living in poverty, and half in extreme poverty. At the national level, according to data from the nationally representative 2007 Core Welfare Indicators Questionnaire (CWIQ) survey (Box 1.1), 63.8 percent of the population is poor (Backiny-Yetna et al., 2009). This means that there are 1.7 million individuals in poverty in the country. The share of the population in extreme poverty is 47.9 percent, or 1.3 million people. Poverty is higher in rural (67.7 percent) than in urban areas (55.1 percent). Given that close to 70 percent of the population lives in urban areas, rural areas account for almost three quarters (73.4 percent) of the poor. 1.2. The poor tend to be underemployed, unemployed, or have low-quality jobs. In terms of the socioeconomic group of the head of household, those in the public sector or with a wage in the private formal sector have the lowest poverty rates (Table 1.1). The highest levels of poverty are observed among households with a head who is self-employed in agriculture, followed by those who are inactive (not working). Poverty rates by industry are lowest in the banking/financial sector, followed by utilities. Poverty rates are highest for those involved in fishing, crop farming, and mining/quarrying, and again for those who are unemployed or inactive. Household heads who have a second occupation tend to have a lower probability of being poor. The data also indicate that the probability of being poor is lower when there are one or two workers in the household, as opposed to none or more than two (in the latter case, because this denotes large households, which need to have many members working). Box 1.1. CWIQ Sur vey: Sampling and Quality of Employment Data Much of the empirical analysis presented in this report is based on data from the nationally representative Core Welfare Indicator Questionnaire (CWIQ) survey implemented by Liberia Institute of Statistics and Geo-Information Services (LISGIS) in 2007. This is the only such survey available for Liberia. Although the CWIQ is comprehensive and of relatively good quality, it was based in part on a 1984 sample frame and therefore may not precisely represent the current structure of the population and the labor market. Further, some of the weights used for the survey analysis were derived before the new census data became available). For example, the population size as measured using the survey weights is lower than the population size obtained from the 2008 census. Another issue relates to the share of the population living in urban as opposed to rural areas, with "major" urban areas defined as having a population of at least five thousand. The 2007 CWIQ shows that 31 percent of the population lives in urban areas, whereas the 2008 census shows an urban population of 39 percent. In terms of data quality, there do not appear to be any significant problems with the employment characteristics of the households, the estimation of wages required substantial corrections to the data to ensure reliability of the results. Source: World Bank staff. -3- Table 1.1. Shar e of the Population in Pover ty by Employment Char acter istics Share of the Population Poverty Headcount Urban Rural National Urban Rural National Socioeconomic group of household head Public 24.3 9.2 13.9 40.7 59.0 49.1 Private formal 5.6 5.2 5.3 37.5 63.0 54.6 Private informal 6.5 3.8 4.6 52.4 52.1 52.2 Self-agriculture 3.2 46.7 33.3 79.4 71.8 72.0 Self-other 27.4 16.4 19.8 54.7 62.2 59.0 Unemployed 12.1 2.5 5.4 67.6 62.9 66.1 Inactive, other 20.9 16.2 17.7 66.8 72.2 70.3 Industry of household head Crop farming 3.5 53.0 37.7 80.1 71.3 71.6 Forestry/logging 0.5 0.2 0.3 23.0 91.8 56.3 Fishing 0.7 0.1 0.3 77.4 67.3 74.3 Mining/quarrying 0.4 0.6 0.5 78.9 69.0 71.2 Manufacturing/processing 0.5 0.3 0.3 70.0 64.7 67.2 Electricity/gas/water supply 1.6 0.1 0.6 31.8 14.6 30.2 Construction 3.1 0.7 1.5 60.1 52.7 57.5 Wholesale/retail trades 10.4 3.2 5.4 49.6 38.0 44.8 Transport, storage, communications 2.8 0.3 1.1 36.9 46.4 38.6 Banking/financial services 1.0 0.2 0.4 24.7 34.6 27.6 Community services 13.7 7.4 9.3 42.0 57.1 50.3 Other 31.2 18.9 22.7 50.7 65.4 59.2 Unemployed, inactive 30.6 15.0 19.8 67.2 71.1 69.3 Source: Backiny et al. (2009). 1.3. Regression analysis suggests a strong correlation between employment and the level of consumption of households. Drawing a profile of poverty is a necessary step in identifying the characteristics of the population groups that are poor--however, it is not sufficient to measure the impact of various household characteristics on poverty or the correlation between these characteristics and poverty. The inherent problem with a poverty profile is that it provides information on who are the poor, or of which household characteristics make a household likely to be poor but it cannot be used to assess the correlates of poverty. That is, the variation of poverty rates across regions is sometimes better accounted for by the differences in households' characteristics than by the specificities of each region. To sort out the correlates or determinants of consumption and poverty, regression analysis is needed. Regressions on the correlates of the consumption level of households were run separately for Monrovia, other urban areas, and rural areas. Explanatory variables include: (a) geographic location; (b) demographic characteristics (number of infants, children, adults, and seniors, and their squared value), whether the household head is a woman, the age of the head, and the marital status of the head; (c) characteristics of the household head, including level of education, socioeconomic group, and whether the head has a second job; (d) the education level of the spouse of the household head, where there is one; and (e) other variables such as land cultivated, migration related to the war, and access to infrastructure. Key findings are as follows (Backiny-Yetna et al., 2009): · Demographic characteristics. An additional person in the household reduces consumption per equivalent adult, with the impact ranging from no loss to a loss of 25 percent of consumption per adult, depending on the case. Yet the impact on the probability of being poor is less statistically significant in urban areas (except for the number of male adults), and the impact is not present for subjective poverty, as has been observed in other countries. There are few statistically significant differences between male-headed and female-headed households. In terms of marital structure, -4- most of the coefficients are also not statistically significant, so no generalizations can be drawn. Finally, the age of the head does not seem to make a major difference in consumption levels. Thus, in terms of demographics, the main finding is that larger households have a lower consumption per equivalent adult, even after controlling for the differences in needs among different persons through the use of the adult equivalence scale. · Education level of the head and spouse. As expected, consumption levels increase and the probability of being poor decreases the higher the education level of the household head, but the effects are statistically significant only for secondary schooling. The impact of the spouse's education is, in most cases, of an order of magnitude similar to that of the head. Still, overall the impacts are not very large, which suggests that there are limited opportunities through good employment to benefit from the full returns that an education can provide. Employment of the head. After controlling for other variables, the type of employment does not seem to have much effect on the level of consumption of households, or on their probability of being poor. This is surprising to the extent that in many other countries, when the household head belongs to the public sector or the private formal sector, the household is typically better off than when the head is self-employed, especially in agriculture. By contrast, if the head is unemployed or inactive, the negative impact on consumption and poverty is rather large in most instances, and indeed larger than what has been observed in other West and Central African countries. Controlling for other characteristics, the unemployment of a household head reduces a household's consumption level by 37.5 percent in Monrovia (this estimate is only marginally statistically significant), 21 percent in other urban areas, and 17 percent in rural areas versus having a household head employed. Having an inactive head (i.e., not in the labor force) reduces consumption by 25 percent in other urban areas and 32 percent in rural areas (the effect is also large in Monrovia, at 30 percent, but not statistically significant). · Other variables. After controlling for other variables, if the household has a larger land size available for cultivation, consumption is higher, and the probability of being poor is lower, as expected. Displaced households that have returned to their place of origin actually seem to be better off, after controlling for other variables, than non-displaced persons, perhaps because those who were displaced had more liquid assets to enable them to leave their place of origin. Isolated households, as measured by the time it takes to reach the closest food market, tend to have lower consumption levels and a higher probability of being poor. Finally, there is some evidence that households in the South Central region of Liberia, and to some extent in the North Western region, are poorer after controlling for other observable characteristics. 1.4. Households themselves would like the creation of employment to be a top priority for government. Close to half (47 percent) of the population believes that employment creation should be the main priority of the government (Table 1.2). The importance of jobs appears to be stronger in urban than in rural areas (where self-employment may be easier through itinerant mining or subsistence agriculture, even though it has low productivity) and among the poorer segments of the population, as expressed by the quintile of household consumption (Table 1.2). Improving access to education is the top priority for 16 percent of households--a distant second; and providing paved roads is the top priority for 15 percent of households. -5- Table 1.2. Most Impor tant Measur es for Gover nment To Impr ove Living Standar ds, 2007 (% ) Residence Area Quintile Urban Rural Q1 Q2 Q3 Q4 Q5 Total Create employment 61.9 40.8 53.0 46.8 42.6 48.0 47.3 47.4 Improve access to education 11.5 17.4 12.2 15.4 14.1 16.9 18.0 15.6 Improve access to health care 1.1 4.0 2.5 3.8 5.2 3.0 1.6 3.1 Provide paved roads 3.5 20.1 15.8 14.9 21.0 12.2 11.7 14.9 Improve access to housing 0.9 3.8 2.9 4.3 2.9 2.5 2.1 2.9 Improve access to credit 3.2 2.4 1.5 2.0 2.7 3.9 2.8 2.6 Improve access to water or electricity 1.7 0.4 0.7 0.6 0.3 0.7 1.6 0.8 Increase salaries 3.2 2.5 1.6 1.7 1.4 4.3 4.0 2.7 Regulate basic prices 9.7 6.5 8.4 8.5 6.5 5.3 8.7 7.5 Fight against corruption 3.1 1.5 0.8 1.5 2.4 3.1 2.1 2.0 Other measures 0.3 0.5 0.5 0.4 0.8 0.4 0.1 0.4 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: World Bank staff calculations based on 2007 CWIQ survey data. 1.5. While unemployment and underemployment are clearly factors leading to poverty, low productivity jobs are an even larger determinant. As discussed in the next section, one in five Liberian workers is either unemployed or underemployed; therefore, improving employment opportunities would lead to poverty reduction. In the medium term, however, it is higher quality jobs that will have the greatest impact on poverty. Since workers already work a substantial number of hours per week (an average of 46 hours for the fully employed and underemployed), the scope for additional hours of work is limited. By contrast, increasing the productivity of work through investments in agriculture and infrastructure, as well as the creation of better paying jobs, could help to progressively raise the value added per worker and thereby the standard of living of poor households. -6- 2. ONE IN FIVE WORKERS IS UNEMPLOYED OR UNDEREMPLOYED 2.1. Liberia's population is rising, and so is the labor supply, especially in urban areas. Liberia's total population was estimated at 2.1 million in 1990 and 3.5 million in 2008 (Figure 2.1), 1 and is projected to grow at an average annual rate of 2.5 percent to about 4.2 million by 2015 (2008 National Census). The population is young. Approximately 47 percent is 0-14 years old, 51 percent is 15-64 years old, and only 2 percent is more than 65 years old. The population has been rapidly increasing in recent years, due in part to the return of refugees after the end of the conflict. Population growth has resulted in large increases in the working age population and has contributed to high rates of unemployment and underemployment, particularly in urban areas due to migration. The process of urbanization began in earnest in the 1970s after the first oil shock and the downturn in the economy but accelerated during the war, creating additional challenges for government in terms of service provision in urban centers, particularly in the two cities--Monrovia and Buchanan. Rapid urbanization has made heavy demands on the environment in Monrovia, which lacks the resources to deal with chronic solid waste disposal challenges. Figur e 2.1. Tr end in Liber ia's Population, 1960­2008 4.00 3.50 3.00 2.50 Millions 2.00 1.50 1.00 0.50 0.00 1965 1970 1975 1980 1985 1990 1995 2000 2005 2007 2008 Population (excl. Monrovia) Population of Monrovia Source: World Bank, Development Data Platform (DDP). 1 World Development Indicators (WDI), 2007, World Bank; Census results, 2008. -7- 2.2. The CWIQ survey provides useful data on employment. It has been said that between 75 and 80 percent of Liberia's population might be unemployed, but this statistic is not supported by the available data. However, it is clear that the share of the population employed in "gainful jobs" is limited because of a lack of available good jobs. According to the survey data, out of Liberia's population of 3.5 million, approximately 2 million are of working age (above the age of 15). This labor force is likely to grow at a rate of about 2.5 percent per annum in the near future. Both the labor force participation rate of 73 percent and the employment-to-population ratio of 68 percent are close to the average for Sub-Saharan Africa (74 and 67 percent, respectively, in 2006). Liberia's unemployment rate, at 5.7 percent, is also in line with unemployment rates in neighboring countries--although, as discussed below, this measure underestimates the extent of unemployment and underemployment. There are differences in labor force participation rates by age, gender, and level of household welfare (as measured by the quintile of household per capita consumption to which an individual household head belongs). Labor force participation rates are lower among youths (aged 15 to 25) than among other adults, especially in urban areas, where a larger share of youth is still in school. Participation is lower among women, although it is relatively high for Africa--females constitute about 40 percent of the labor force, and their participation rate is 69.3 percent compared to 77.0 percent for males (2007). Participation is also lower among the very poor (first quintile) than among the rest of the population. It is higher in rural than in urban areas--not because there are more employment opportunities in rural areas, but more likely because of the greater pressure to work to survive. Box 2.1. Definitions of Employment, Unemployment, and Under employment As defined by the International Labour Organization (ILO), a person is employed if he/she did any type of paid work during the past seven days. The unemployed are those who are (a) not working, (b) available for work, and (c) looking for work. The unemployment rate is computed as {U / (E+U)}*100, where U is the number of unemployed and E is the number of employed individuals. The inactive population is not included in this calculation. It is often useful to relax the strict definition of unemployment to include some individuals not actively looking for work. One such group consists of those who are not seeking work because they think no work is available. Another consists of those who are seasonally inactive. The first group is included under a partial extension of the definition; the second is included under the extended definition. For underemployment, part-time work is defined as working less than 40 hours per week and full-time employment as working at least 40 hours per week, and in both cases not wanting or being available to work additional hours. Visible underemployment is defined as working less than 40 hours per week and wanting and being available to work additional hours. Invisible underemployment is defined as working at least 40 hours per week and wanting and being available to work additional hours. Sources: ILO and World Bank staff. 2.3. Liberia's unemployment rate, using the standard ILO approach, is at 5.7 percent. This represents the share of the labor force not working, willing to work, and actively seeking work (Box 2.1). This definition, however, does not account for substantial disguised unemployment. If we broaden the definition of the unemployed to include those who would like to work but did not seek work either because there was no work available or because they were in seasonal activity, the unemployment rate doubles to 11.1 percent. The fact that unemployment is well below some estimates is not surprising, given that most adults in the labor force simply must work in order to survive--but this does not mean that these individuals are gainfully employed. In terms of comparisons between groups, the differences in -8- unemployment rates between poorer and richer individuals are reversed when one considers the extended definition of unemployment. Under this definition, which is more appropriate for Liberia, the poorest quintiles now have substantially higher unemployment rates than the richer quintiles (Table 2.1). For example, under the complete relaxation assumption, 16.5 percent of the individuals in the poorest quintile are unemployed, compared with about 11 percent in the rest of the population.2 Table 2.1. Labor For ce Par ticipation and Unemployment Rates, 2007 Residence area Quintile Urban Rural Q1 Q2 Q3 Q4 Q5 Liberia Labor force participation rate 15-24 36.4 69.5 48.8 65.3 62.7 55.6 58.1 58.1 25-64 74.6 84.9 69.2 85.0 85.2 85.2 82.7 81.6 Total 60.2 79.4 61.4 77.7 76.6 75.1 74.5 73.1 Unemployment rate, standard definition 15-24 14.0 2.0 2.1 4.2 4.3 4.4 7.9 4.6 25-64 13.8 2.9 7.4 5.7 6.3 5.7 6.0 6.2 Total 13.9 2.6 5.8 5.2 5.7 5.4 6.4 5.7 Unemployment rate, extended definition 15-24 24.5 7.3 14.5 9.6 10.4 9.8 12.3 11.2 25-64 20.2 7.0 17.4 8.9 10.0 9.5 10.1 11.0 Total 21.2 7.1 16.5 9.1 10.2 9.6 10.7 11.1 Source: World Bank staff, based on 2007 CWIQ survey. 2.4. In addition to substantial unemployment, Liberia also has considerable underemployment. Table 2.2 provides data on rates of part-time and full-time work as well as visible and invisible underemployment. Individuals identified as part-time workers not underemployed are not necessarily interested in working more hours; and individuals identified as full-time workers not underemployed are also not interested in working more hours. Individuals classified as visibly underemployed are part- timers who would like to work more. Those classified as invisible underemployed are full-timers who would like to work longer hours. Overall, 5.7 percent of the employed population can be considered in invisible underemployment, and 2.6 percent in visible underemployment. The total rate of underemployment is thus at about 8.3 percent, which is of a similar order of magnitude as the unemployment rate. Together with the unemployment rate, about 20 percent of the labor force therefore does not find enough work to satisfy the number of hours that they would be willing to work. 2 Among the 15-24 age group, unemployment rates are extremely high even among the well educated. Under complete expansion of the standard definition, unemployment among those with a post-secondary education reaches 70.4 percent for young women and 49.3 percent for young men. The issue of the unemployment of well-educated youth, however, is beyond the scope of this policy note, which focuses, rather, on the lack of good jobs for the poor, many of whom have no education. -9- Table 2.2. Under employment Rate, 2007 Residence area Quintile Urban Rural Q1 Q2 Q3 Q4 Q5 Liberia Age 15-24 Part-time, not underemployed 51.5 48.0 46.5 45.6 50.5 50.8 50.1 48.7 Full-time, not underemployed 39.4 47.8 47.7 50.7 43.2 43.7 45.1 46.1 Visible underemployment 4.3 3.0 5.0 2.1 4.3 2.4 2.4 3.3 Invisible underemployment 4.9 1.2 0.8 1.6 2.0 3.0 2.5 2.0 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Age 25-64 Part-time, not underemployed 18.1 18.5 25.4 18.3 17.7 16.1 16.1 18.4 Full-time, not underemployed 68.7 73.3 68.6 76.1 71.3 71.7 71.8 72.0 Visible underemployment 1.6 2.7 2.6 1.2 3.6 2.3 2.1 2.4 Invisible underemployment 11.7 5.6 3.3 4.4 7.4 9.9 10.0 7.2 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 All ages Part-time, not underemployed 25.7 27.7 32.1 26.9 28.1 24.9 24.8 27.2 Full-time, not underemployed 62.0 65.3 62.0 68.1 62.4 64.6 64.9 64.5 Visible underemployment 2.2 2.8 3.4 1.5 3.8 2.3 2.2 2.6 Invisible underemployment 10.1 4.2 2.5 3.5 5.7 8.2 8.0 5.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: World Bank staff, based on 2007 CWIQ survey. 2.5. Liberia's severe underemployment is masked by the fact that the labor market is characterized by a narrow formal sector and a large, low-productivity informal sector. Only an estimated 17 percent of the employed population consists of "paid employees," while unpaid family workers comprise 48 percent and self-employed without employees comprise 32 percent of total employment (Table 2.3). The agriculture sector is by far the largest employer, engaging almost 50 percent of the working population, followed by the services sector. Agricultural employment however, is lower than in Africa as a whole (66 percent in 2006), which may reflect the high level of displacement during the civil conflict, which drove the rural population into cities and towns that afforded greater protection. Significant numbers of Liberians returned to rural communities after the war, particularly those returning from neighboring countries, but large numbers remain in urban and peri-urban areas. - 10 - Table 2.3. Employment Status in the Main Occupation, 2007 Residence area Quintile Urban Rural Q1 Q2 Q3 Q4 Q5 Liberia Age 15-24 Paid employee 7.7 5.8 5.2 9.8 5.3 4.5 5.6 6.2 Self-employed with employees 1.2 0.9 2.2 0.5 0.6 1.0 0.7 0.9 Self-employed no employees 21.6 18.0 22.2 19.4 15.7 18.7 18.4 18.7 Unpaid family worker 66.4 74.7 67.8 69.6 77.6 74.9 74.4 73.0 Domestic employee 1.9 0.6 2.3 0.7 0.5 0.5 0.5 0.9 Apprentice 1.3 0.0 0.3 0.1 0.3 0.4 0.4 0.3 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Age 25-64 Paid employee 41.7 13.6 12.6 14.0 19.2 24.3 33.2 21.3 Self-employed with employees 4.1 1.9 4.7 1.9 1.2 2.6 2.7 2.5 Self-employed no employees 34.0 38.9 44.0 39.9 36.1 37.3 32.4 37.6 Unpaid family worker 19.0 44.7 37.1 43.9 42.8 34.6 30.6 37.7 Domestic employee 0.8 0.8 1.4 0.3 0.4 1.1 0.7 0.8 Apprentice 0.4 0.0 0.2 0.3 0.3 0.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 All ages Paid employee 33.8 11.2 10.3 12.7 14.8 19.4 26.2 17.0 Self-employed with employees 3.4 1.6 4.0 1.4 1.0 2.2 2.2 2.1 Self-employed no employees 31.1 32.5 37.3 33.5 29.6 32.6 28.9 32.1 Unpaid family worker 29.9 53.9 46.6 51.9 53.9 44.8 41.7 47.8 Domestic employee 1.1 0.7 1.7 0.4 0.4 1.0 0.7 0.8 Apprentice 0.6 0.0 0.2 0.0 0.3 0.1 0.3 0.2 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: World Bank staff, based on 2007 CWIQ survey. 2.6. Under the standard ILO definition, unemployment is slightly higher among youth, but it is also high among older groups, where it affects the poor more (Table 2.4). Among men, unemployment is higher among youth than among older individuals, while differences are small for women. For older individuals, however, unemployment is much more concentrated in the poorest quintiles of the population, while among youths it is concentrated among the higher quintiles. This would suggest that for poverty reduction, addressing high unemployment among older individuals might be a more important priority, at least in the short run, than among youths. When using a broader definition of employment, unemployment remains higher among youth than older groups (including for women), but it is less concentrated among better off quintiles, while for older workers it remains higher among the poorest quintiles. - 11 - Table 2.4. Unemployment Rate (15-64) by Gender , Ar ea, and Welfar e Quintile, 2007 Residence area Quintile Urban Rural Q1 Q2 Q3 Q4 Q5 Liberia Both sexes 15-24 14.0 2.0 2.1 4.2 4.3 4.4 7.9 4.6 25-64 13.8 2.9 7.4 5.7 6.3 5.7 6.0 6.2 Total 13.9 2.6 5.8 5.2 5.7 5.4 6.4 5.7 Males 15-24 18.0 1.9 2.7 4.5 4.4 7.3 9.0 5.3 25-64 15.5 4.2 11.1 7.5 7.4 7.6 6.1 7.7 Male total 16.0 3.5 8.3 6.5 6.5 7.6 6.8 7.1 Females 15-24 10.1 2.1 1.2 3.9 4.2 2.4 6.8 3.8 25-64 11.8 1.7 3.6 3.9 5.3 3.5 5.8 4.4 Females total 11.4 1.8 2.9 3.9 4.9 3.2 6.1 4.3 Source: World Bank staff, based on 2007 CWIQ survey. 2.7. Beyond unemployment and underemployment, a third issue is that many individuals are engaged in low-productivity work without pay. The CWIQ data show (Table 2.3) that nearly half of all workers (47.8 percent) are involved in unpaid family work; another 32.1 percent are self-employed without employees; and 17.0 percent are paid employees. An additional 2.1 percent are self-employed with employees. The last two categories (domestic employees and apprentices) together represent one percent of total employment. The very large share of unpaid family workers suggests that many workers are likely to be involved in subsistence and low-productivity work, mainly in agriculture but also in other sectors. The share of unpaid family workers is largest among youths, as expected, but it is also very large among older adults, especially women. As a result, most workers earn fairly low wages and overall household incomes are low as shown in Table 2.5, which gives the distribution of household taxable income in the various brackets of the tax code. Among the poor, most workers are still employed in agriculture and in rural areas, as expected (Table 2.6). Table 2.5. Distr ibution of Household Incomes acr oss Taxable Income Gr oups Sample Weighted Income group Freq. Percent Cum. Freq. Percent Cum. (Liberian dollars) Full sample 0 2,495 69.40 69.40 357,319 71.87 71.87 1 ­ 12000 422 11.74 81.14 57,560 11.58 83.45 12001-50000 414 11.52 92.66 51,102 10.28 93.73 50001-100000 147 4.09 96.75 16,621 3.34 97.07 100001-200000 58 1.61 98.36 6,414 1.29 98.36 200001-400000 42 1.17 99.53 5,546 1.12 99.48 400001-800000 14 0.39 99.92 1,824 0.37 99.84 800001-1200000 2 0.06 99.97 442 0.09 99.93 >1200000 1 0.03 100.00 331 0.07 100.00 Total 3,595 100.00 497,159 100.00 Source: World Bank staff based on 2007 CWIQ survey. Note: 1 Liberian dollar equals 0.014 US dollar. - 12 - Table 2.6. Main Activity in the Main Occupation, 2007 (shar e of labor for ce) Residence Area Quintile Total Urban Rural Q1 Q2 Q3 Q4 Q5 Age 15-24 Crop farming 4.6 54.6 53.5 48.7 42.3 42.5 37.2 44.9 Livestock/poultry - - - - - - - - Forestry/logging 0.0 0.2 0.9 0.0 0.0 0.0 0.1 0.2 Fishing 0.5 0.2 0.0 0.2 0.2 0.9 0.2 0.3 Mining/quarrying 0.0 0.4 0.1 1.1 0.1 0.0 0.1 0.3 Manufacturing/processing 0.2 0.2 0.4 0.0 0.4 0.2 0.2 0.2 Electricity/gas/water supply 0.5 0.0 0.0 0.0 0.0 0.5 0.0 0.1 Construction 0.3 0.6 0.9 0.2 0.5 0.1 1.3 0.6 Wholesale/retail trades 10.4 3.4 3.8 2.7 2.3 8.1 8.2 4.8 Transport, storage, communications 0.4 0.1 0.1 0.1 0.0 0.0 0.4 0.1 Banking/financial services 0.2 0.0 0.0 0.0 0.0 0.0 0.2 0.0 Community services 3.6 1.4 1.5 1.2 1.6 2.7 2.4 1.8 Other 79.4 38.8 38.9 45.8 52.7 45.0 49.8 46.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Age 25-64 Crop farming 4.7 63.1 55.4 57.4 51.1 45.6 30.1 47.3 Livestock/poultry 0.2 0.1 0.1 0.0 0.0 0.0 0.3 0.1 Forestry/logging 0.4 0.2 0.4 0.2 0.3 0.2 0.2 0.2 Fishing 0.9 0.1 0.4 0.6 0.2 0.2 0.2 0.3 Mining/quarrying 0.7 0.6 0.8 0.5 0.7 0.4 0.9 0.6 Manufacturing/processing 0.4 0.4 0.7 0.3 0.3 0.4 0.3 0.4 Electricity/gas/water supply 1.4 0.2 0.1 0.1 0.4 0.8 0.9 0.5 Construction 3.7 0.7 1.1 1.5 2.0 0.9 1.9 1.5 Wholesale/retail trades 17.3 4.3 4.5 5.5 6.4 10.2 11.4 7.8 Transport, storage, communications 3.4 0.4 1.0 0.3 0.7 1.4 2.6 1.2 Banking/financial services 1.5 0.1 0.1 0.3 0.5 0.4 1.0 0.5 Community services 16.8 6.0 5.0 6.4 8.2 10.0 13.9 8.9 Other 48.6 24.0 30.5 26.9 29.3 29.7 36.5 30.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Total 15-64 Crop farming 4.6 60.5 54.8 54.6 48.3 44.8 31.9 46.6 Livestock/poultry 0.1 0.0 0.1 0.0 0.0 0.0 0.2 0.1 Forestry/logging 0.3 0.2 0.6 0.1 0.2 0.1 0.2 0.2 Fishing 0.8 0.1 0.3 0.5 0.2 0.3 0.2 0.3 Mining/quarrying 0.5 0.5 0.5 0.7 0.5 0.3 0.7 0.5 Manufacturing/processing 0.4 0.3 0.6 0.2 0.3 0.3 0.3 0.3 Electricity/gas/water supply 1.2 0.1 0.1 0.1 0.2 0.7 0.7 0.4 Construction 2.9 0.7 1.0 1.1 1.5 0.7 1.7 1.2 Wholesale/retail trades 15.7 4.0 4.3 4.6 5.1 9.7 10.5 6.9 Transport, storage, communications 2.7 0.3 0.7 0.2 0.5 1.0 2.0 0.9 Banking/financial services 1.2 0.1 0.1 0.2 0.4 0.3 0.8 0.4 Community services 13.8 4.6 3.9 4.7 6.1 8.2 10.9 6.9 Other 55.7 28.6 33.2 32.9 36.7 33.6 39.9 35.3 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: World Bank staff, based on 2007 CWIQ survey. Note: The "Other" category captures a wide range of service-oriented activities among fairly heterogeneous groups. - 13 - 2.8. Many workers who are fully employed in terms of working hours remain in poverty. Table 2.7 shows that the average number of hours worked among employed workers is relatively high (given underemployment), at 44.8 hours per week in the workforce as a whole. This suggests that for most employed workers (except those underemployed) the potential for increasing work hours in order to reduce poverty is limited; indeed, poverty reduction is more likely to come in the medium term from a progressive improvement in the quality of the jobs held by a larger share of the population, than from an increase in work effort. While there is a question on seasonality in the CWIQ survey as one option for not working, the data suggest that at the time of the survey, seasonality was not a major factor in unemployment. Table 2.7. Hour s Wor ked per Week by Gender , Ar ea, and Welfar e Quintile, 2007 Residence area Quintile Urban Rural Q1 Q2 Q3 Q4 Q5 Liberia Both sexes 15-24 37.0 37.1 36.1 39.4 36.4 36.2 37.2 37.1 25-64 48.5 47.8 44.0 49.4 47.6 48.9 49.0 48.0 Total 45.9 44.5 41.5 46.3 44.0 45.7 45.9 44.8 Males 15-24 38.9 36.0 37.7 36.5 34.8 37.7 36.6 36.6 25-64 49.7 48.8 45.4 50.5 48.3 49.9 50.2 49.1 Males total 47.5 44.8 42.6 45.9 44.2 47.4 47.1 45.5 Females 15-24 35.4 38.2 33.9 42.5 37.7 35.2 37.7 37.6 25-64 47.0 46.8 42.7 48.4 46.9 47.8 47.5 46.9 Females total 44.0 44.1 40.3 46.7 43.9 44.0 44.7 44.1 Source: World Bank staff, based on 2007 CWIQ survey. - 14 - 3. THE CURRENT ECONOMIC STRUCTURE LIMITS PROSPECTS FOR EMPLOYMENT 3.1. The economy is structurally untransformed with exports, GDP growth, and employment being driven by a few traditional sectors, largely based on foreign investment. The increasing share of the primary, low-productivity sectors and the stagnation or decreasing share of the secondary and tertiary sectors of the economy suggest that the Liberian economy has regressed in its capacity to provide improved labor opportunities. Liberia's economy has remained heavily agrarian since the country was founded, with the majority of the labor force engaged in low-productivity farming. From its original base as a subsistence economy with some cash crop production (primarily cocoa, coffee, and oil palm), the Liberian economy received its first major stimulus in the 1930s, when the Firestone Rubber Company began producing rubber in the country. The rapid growth of the rubber industry created a demand for rural agricultural labor in the formal sector and increased government revenues in Monrovia, spurring an increase in government and services sector employment. In the mid-1950s, Liberia's first iron ore mine entered into production, and by the mid-1960s four large iron ore concessions were in production, providing a second massive boost to GDP, exports, and government revenues. The major structural shift in the economy over the last 20 years has been the decreasing contribution of the services (tertiary) sector and the increasing share of the agricultural (primary) sector (Figure 3.1). In 1987, services accounted for 44.8 percent of GDP, while the agriculture's share was 32.5 percent. By 1995, however, the services sector share of the economy had fallen to 16.3 percent while agriculture's share had risen to 60 percent. Since the end of the war in 2003, the services sector has been recovering, with its share of GDP rising to 24.7 percent in 2007 from 19.9 percent in 2003. At the same time, agriculture's share of GDP has fallen from 55.5 percent in 2003 to 51.2 percent in 2007. 3.2. While the mining sector was dramatically weakened by the war, the forestry sector's share of GDP has grown (Figure 3.1). The mining sector's share of GDP fell from 10.5 percent in 1989 to less than one percent in 1993 and has remained below one percent since then--due mainly to the cessation of iron ore mining with the onset of the war in 1989. In contrast, the contribution of the forestry sector to GDP grew from 4.8 percent in 1987 to peak at nearly 21 percent in 1994--largely driven by the charcoal and wood sector. Since then, the sector's share of the economy has gradually fallen to about 17 percent in 2007. Over the same 20-year period, the performance of the manufacturing sector has been unremarkable. The manufacturing sector's share of GDP increased from 7.3 percent in 1987 to 11.4 percent in 1991, but since then has remained below 10 percent--well below the world's average and lower than most sub- Saharan African countries.3 Although the manufacturing and services sectors' share of the economy has been improving since the end of the war in 2003, the long-term trend suggests that these sectors, on average, have been growing more slowly than GDP. 3 African Development Indicators (2007) and World Development Indicator (2007). - 15 - Figur e 3.1. Sector al Contr ibution to GDP (1987- 2007) 80.0 70.0 60.0 50.0 % 40.0 30.0 20.0 10.0 0.0 Agriculture and fishries Forestry Mining and panning Manufacturing Services Sources: LISGIS and IMF. 3.3. The relatively undiversified nature of the Liberian economy and its dependence on a few primary commodities are also seen in the structure of its exports (Figure 3.2). Before the war, Liberia's exports were dominated by iron ore, which accounted for more than 50 percent of total exports in US dollar terms. Rubber and timber exports accounted for about 24 and 15 percent. At the onset of the civil war, the iron ore industry virtually disappeared and rubber assumed the dominant export position; its share peaked at 76.6 percent of total exports in 1997. That same year, the other major export, timber represented about 19 percent. With rubber and timber accounting for approximately 95 percent of total exports between 1997 and 2002, "other" exports, including cocoa and coffee, made up the other 5 percent. In 2003, the United Nations banned the export of timber from Liberia pending the passage of a new Forestry Law. Consequently, since 2004 rubber has been the dominant export, accounting for an average of 91 percent of total exports between 2004 and 2007. The ban on timber exports was lifted in 2006, but no timber exports were reported during 2007 and 2008. - 16 - Figur e 3.2. Str uctur e of Liber ia's Expor ts (1987-1988 and 1997-2007) 100.0 90.0 80.0 70.0 60.0 % 50.0 40.0 30.0 20.0 10.0 0.0 1987 1988 Civil 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 War Rubber Timber Cocoa Coffee Iron Ore Other Sources: IMF and Central Bank of Liberia. 3.4. In the wake of the destruction of key infrastructure and the weak (though improving) security situation in Liberia, diversifying the economy will take some time. Thus a key challenge is to maximize the economic and social benefits from the economy given its current structure. In large part, this is a question of how to better manage natural resources and use them more effectively to increase their contribution to the economy, while preventing unwanted effects, including Dutch disease (appreciating of the exchange rate and decline in competitiveness across most sectors) and negative environmental impacts. To maximize the benefits of the natural resources to the economy, the government would need to act on three important fronts: (a) improve governance of the natural resources sector; (b) develop strategic interventions to take full advantage of the value chain spin-offs from the concessions; and (c) monitor the dynamics of the concessions and garner evidence to inform future evidence-based policies. - 17 - 3.5. The Government has made remarkable progress in improving governance of the natural resources sector, but additional actions are required. In June 2007, following extensive consultations, the Government, in partnership with companies and persons in the mining, oil, and logging sectors, agreed to establish the Liberia Extractive Industries Transparency Initiative (LEITI). A Memorandum of Understanding (MoU), formally committing the stakeholders to full implementation of the LEITI, was signed in April 2008. In September 2008, President Ellen Johnson Sirleaf released a proclamation adopting LEITI as a significant policy of the Government, and requiring all government agencies, as well as all mining, logging, and oil companies to comply with all its requirements. Liberia's first EITI report, issued in February 2009, was the first in any country to include forestry. LEITI was signed into law in July 2009, and that October, Liberia became the first country in Africa and the second country in the world to be recognized as EITI compliant. Going forward, Liberia needs to work towards implementing a broader governance framework for the natural resources sectors. This framework, known as "EITI++," provides countries with a mechanism to address governance issues across the entire supply chain, from the negotiation and award of contracts; through regulations, taxes and royalties, investment plans, and enforcement of best practice standards; to anticipating, monitoring, and measuring an activity's social and environmental impacts. 3.6. The experiences of Australia, Sweden, and Finland have shown that natural resources can provide the launching pad for successful rapid and sustained pro-poor growth (Lederman and Maloney, 2007).To harness more economic benefits from the natural resources sectors, as well as to localize and maximize the benefits of value additions, the Government needs to encourage strategic private sector interventions at all points along the sector's value chains. Unfortunately, the value chains for most of Liberia's natural resources sectors are short, since most raw materials are exported before being processed. In the long term, policies are needed to incentivize more value addition within Liberia. In the short term, Government should take steps to maximize the spillovers from the large foreign concessions, such as opportunities for food services, security, training, and courier services. There are also opportunities to involve small farmers, artisanal miners, or itinerant producers in production processes. 3.7. Monitoring operation of the concessions, including their positive and negative spin-offs and the number of jobs they generate, is critical for facilitating learning by doing and developing evidence-based policies going forward. Information to facilitate such monitoring should come from, inter alia, the regular publication of all material payments by companies to government and material revenues received by government; regular consultations between government and companies operating along the entire value chain; and regular consultations with service enterprises. Employment Prospects (2009-2013) 3.8. The supply of workers in the labor market exceeds demand by a substantial margin. Liberia's labor force grew at an average rate of 2.5 percent between 1991 and 1997. Due to 14 years of conflict, however, most of the labor is unskilled. The Government expects the private sector to be the major engine of growth and employment 4 and the pace of private sector investment has accelerated as the security situation has improved. Foreign direct investment, in particular, has increased twenty-fold since 4 See Poverty Reduction Strategy , Republic of Liberia; (2008) - 18 - the Accra Comprehensive Peace Agreement (ACPA) was signed in 2003 and UN peacekeepers have been in the country. Most of this investment has been in the traditional sectors of mining, rubber production, and forestry. Even under an optimistic short- to medium-run scenario, however, the private sector-led demand for labor in these sectors is unlikely to absorb a substantial part of the labor force. The exception is the telecommunication sector, primarily the rapidly expanding cell phone market, where the penetration rate of the country's four providers is still only about 15 percent. 3.9. To develop a strategy to address unemployment and underemployment over the short to medium term, policymakers need to understand the structure of the current labor market and likely employment growth patterns. Table 3.1 below provides projected employment growth relative to the growth of the labor force over the next five-year period, 2009-2013. The projections are based on projected sectoral growth and estimated employment elasticity with respect to growth, using results from the 2007 CWIQ survey as the baseline (see Annex 1 for methodology). These projections provide a rough outline of likely labor market developments. As more information becomes available, more rigorous estimates could be developed and more analysis done to further inform and refine the policy actions. Table 3.1. Pr ojected Employment Gr owth in Liber ia (2009-2013) 2008 (Est.) 2009 2010 2011 2012 2013 Total population 3,489,000 3,576,225 3,665,631 3,757,271 3,851,203 3,947,483 Population 15-64 2,029,770 2,080,514 2,132,527 2,185,840 2,240,486 2,296,498 Total labor force (active labor force) 1,461,800 1,498,345 1,535,804 1,574,199 1,613,554 1,653,893 TOTAL EMPLOYED 1,378,477 1,411,928 1,454,151 1,495,699 1,534,934 1,574,500 FORMAL SECTOR (PAID) 231,584 238,071 249,996 262,149 270,335 277,659 Growth rate of formal employment 2.8% 5.0% 4.9% 3.1% 2.7% Share of total labor force 15.8% 15.9% 16.3% 16.7% 16.8% 16.8% Agriculture (including forestry) 37,643 39,382 42,858 46,311 49,005 51,768 Agriculture 34,949 36,382 38,858 41,311 43,005 44,768 Forestry 2,694 3,000 4,000 5,000 6,000 7,000 Industry 9,026 8,991 12,117 15,329 15,653 15,951 Mining 3,099 3,099 6,099 9,099 9,099 9,099 Manufacturing (incl. arts and crafts) 3,233 3,021 2,953 2,964 3,097 3,236 Electricity, gas, and water 2,694 2,871 3,065 3,267 3,458 3,616 Services 141,639 144,048 146,666 149,364 151,906 154,002 Government 61,758 61,758 61,758 61,758 61,758 61,758 Transport 7,140 7,552 7,999 8,461 8,895 9,254 Construction 10,508 11,114 11,773 12,452 13,091 13,618 Commerce, hotels, restaurants 4,715 4,987 5,283 5,587 5,874 6,111 Other services (incl. com services, 57,517 58,636 59,852 61,106 62,287 63,261 NGOs and intl organizations) 38,111 38,111 38,111 38,111 38,111 38,111 Other 43,162 45,651 48,356 51,144 53,771 55,937 INFORMAL SECTOR 1,146,893 1,175,566 1,204,955 1,235,079 1,265,955 1,297,604 Share of total labor force 78.5% 78.5% 78.5% 78.5% 78.5% 78.5% Agriculture (including forestry and 642,260 658,317 674,775 691,644 708,935 726,658 Industry 11,469 11,756 12,050 12,351 12,660 12,976 Services 114,689 117,557 120,495 123,508 126,596 129,760 Other 367,006 376,181 385,585 395,225 405,106 415,233 UNEMPLOYED 83,323 84,708 80,853 76,971 77,263 78,629 Share of total labor force 5.7% 5.7% 5.3% 4.9% 4.8% 4.8% Sources: Projections developed based on: 2007 CWIQ data; population estimates: 2008 preliminary census results; sectoral growth estimates: IMF, Article IV, 2009; employment elasticities: ILO, KILM, 2007. - 19 - 3.10. The analysis distinguishes between paid employment, used as a proxy for the formal sector, and unpaid informal sector employment. Out of the total labor force, the formal sector employs only 17 percent, while the informal sector employs 78.5 percent. Liberia's formal sector remains dominated by low or semi-skilled employment and relatively high poverty rates, but it still affords a significantly steadier source of income and greater safety net than the informal sector. For example, formal sector employees make up 10.3 percent of the poorest quintile, compared with 37.3 percent for informal sector employees. Formal employment growth is therefore an important step towards enhancing the quality of work in Liberia. Informal employment, often characterized by significant underemployment, low productivity, and consequently low wages can be regarded as a buffer that absorbs excess labor without a corresponding increase in output, thus masking the true employment gap. 3.11. Over the five-year period 2009-2013, assuming that the informal sector stays constant as a proportion of the labor force, the unemployment rate (ILO definition) is expected to decrease from 5.7 percent to 4.8 percent. Approximately 45,000 new jobs are expected to be created in the formal sector, expanding it only marginally, from 15.8 percent of the labor force to 16.7 percent. The baseline data show that the formal sector is likely to remain heavily dominated by the services sector, particularly public sector employment. The government, state-owned enterprises (SOEs), NGOs, and international organizations combined account for almost 50 percent of paid employment. Public sector and NGO employment are expected to remain relatively constant over the coming years as the Government moves to "right-size" the civil service and aid flows through NGOs and international organizations gradually decline. This will depress formal sector employment growth, which will increase by only about 4 percent a year, driven mainly by growth in other services, industry, and commercial agriculture. 3.12. Liberia's culture of corruption and non-compliance, in the face of weak legal and regulatory structures and high transactions cost for formalization, has given rise to a high level of informality. However, there is no clear evidence to suggest that informality is a drag on growth. A 2007 report by the World Bank's Foreign Investment Advisory Service (FIAS) that examined, among other things, the causes of informality in Africa, argues on the basis of modeling and survey evidence that small and medium enterprise (SME) taxation regimes in Africa have been key contributors to firms' decision to operate informally. The report found that in all 11 African countries analyzed, the measured weight of the tax system faced by SMEs (outside of the value-added tax net) was higher than for every other sector of the economy subject to the general tax regime. In addition, a survey of the compliance costs for SMEs (FIAS, World Bank 2007) shows that the time and financial burden of complying with tax reporting is higher for small firms, and decreases dramatically as firm size increases. Based on the survey evidence, the primary causes of informality are the cost of formality (24.9 percent) and complicated procedures (22.7 percent). Figure 3.3 below highlights the primary reasons for the high level of informality. To its credit, Liberia has been moving to address some of the disincentives for formalization. Specifically, it has lowered the corporate tax rate from 35 to 25 percent and has reduced the time, cost, and number of procedures involved in starting a formal business in Liberia. - 20 - Figur e 3.3. Reasons for Failed For malization Attempts in Liber ia Attempts to become formal and reasons for failure 100 90 Never tried to formalize 80 55 Tried to formalize 70 60 50 Reasons for failure in formalization 40 30 24.9 22.7 20.4 45 16 20 12.2 4.4 10 0.6 0 Steps too Steps too Couldn't Could not The process Other DK//NA expensive complicated comply with find all got stuck requirements necessary information Source: Small Business Taxation: Is This the Key to Formalization? Evidence from Africa and Possible Solutions. FIAS, World Bank (2007). 3.13. Given the limitations to growth in the formal sector and the dominance of subsistence agriculture, which sustains about 40 percent of the labor force, informality will remain the central feature of Liberia's labor market over the medium term. The vast majority of the working poor and a disproportionate number of women and youth are in the informal sector. For example, based on CWIQ data, 23.5 percent of females are engaged in informal agriculture, compared with 19 percent for males. This has considerable implications for labor market policy and tax policy. Tax policy is beyond the scope of this report. However, in terms of labor market policy, while formalization--appropriately defined-- may be the ultimate policy objective to improve wages and working conditions, enforcing formality where informal businesses are operating at the margin could be detrimental to workers. Rather, economic strategies that help to improve the productivity of informal sector activities and thereby raise wages will be necessary to improve the quality of work and life in Liberia. 3.14. Liberia's farming sector is characterized by low investment, low productivity, low wages, and high poverty incidence. The poverty rate among self-employed agricultural workers is 72 percent-- highest of all employment sectors. In the short to medium term, poverty-reducing measures need to focus on transforming agriculture to improve its productivity and increase the incomes of those employed in the sector. This is the issue that is addressed in the next section of the report. - 21 - 4. TRANSFORM AGRICULTURE TO DELIVER HIGHER-QUALITY JOBS 4.1. The recovery and transformation of Liberia's agriculture to unlock its tremendous potential is not only critical to overall growth and development, but also lies at the heart of any strategy to ensure sustained job creation and shared growth for the country. As Timmer (2003) has pointed out, few countries have solved their problem of poverty without creating a dynamic agriculture sector. Similarly, it is well documented that agriculture has the strongest forward and backward linkages to other sectors of the economy and has the greatest growth elasticity of poverty (Byerlee, Diao and Jackson, 2005). Agricultural growth is unlikely to have any unfavorable distributional effects--although there can be important intra-household impacts on, for instance, the relative roles of men and women. 4.2. Liberia's considerable agricultural potential is largely unrealized. The World Bank's recent Diagnostic Trade Integration Study (DTIS) highlights Liberia's comparative advantage in a number of tradable sectors, including rubber, cocoa, and oil palm. 5 However, the country's potential in domestic agriculture, including food production, remains largely unexploited, even in the face of its substantial food import bill. Between 2006 and 2008, Liberia's food import bill in 2008 averaged 25.6 percent of total imports, with rice accounting for more than 60 percent. Liberia's agricultural potential lies in its vast acreage of arable, fertile land; availability of water for irrigation; and substantial unemployed labor. The total land area in Liberia is estimated at 9.8 million hectares (USAID 1998). 6 Of this total, it is estimated that forest occupies 4.9 million hectares and that arable land occupies 4.6 million hectares. Potential pasture land is estimated at some 0.2 million hectares. Liberia's tropical climate is ideally suitable for a number of crops. The rainy season--April to November--records an annual average rainfall of 2,400 millimeters, with spatial variation from 2,000 to 5,000 millimeters. This level of rainfall is more than adequate for most crop growth. In addition, Liberia's irrigation potential is approximately 600,000 hectares (FAO). Even at current low yields, this amount of arable land is more than adequate to provide for national food self-sufficiency in the key staples; with improved yields and subject to cost competitiveness, Liberia could become a net exporter of rice. 4.3. Agricultural yields are low even by African standards, with scant evidence of the application of modern varieties or contemporary practices. Over two decades through 1980, average rice yields were similar to those observed in Ghana, at just over 1mt/ha. Over the next two decades, Liberia's yields increased to an average of 1.18mt/ha while those in Ghana increased to 1.49mt/ha. More recently (2001-2007), local yields actually fell below their 1960-1980 average, to less than 1mt/ha, while Ghana's more than doubled, to 2.14mt/ha.7 Liberia also has substantial yield gaps compared to other countries in West Africa, such as Mali and Senegal. 8 5 Liberia Tapping Nature's Bounty for the Benefit of All: Diagnostic Trade Integration Study, World Bank, December, 2008. 6 USAID (1998). Agricultural Sector Assessment for Liberia and Draft Agricultural Strategy, Liberia. 7 Staff calculations, based on FAOSTAT Database, 2008. 8 Yield data are notoriously unreliable. For instance, while USDA data show yields of 1.33mt/ha in 2006, FAO data show yields of 0.55mt/ha for the same year. - 22 - Table 4.1. Rice Pr oduction and Yields for Selected Afr ican Countr ies (metr ic ton, 2008) Gross Production Yield (per ha) Ghana 250,000 2.08 Guinea 900,000 1.71 Liberia 160,000 1.33 Madagascar 3,600,000 2.72 Mali 909,000 2.33 Senegal 262,000 2.62 Sierra Leone 350,000 1.75 Source: Data from USDA, reported at www.irri.org. 4.4. With an annual population growth estimated at about 2.5 percent, an annual average real GDP growth rate of 10 percent would be required to double per capita incomes over the next decade. Given the importance of the agriculture sector--it accounts for 40 percent of the country's employment and 42 percent of its value added--agriculture will have to grow at or very close to this aggregate rate. Such high rates of agricultural growth are impressive by any measure--since 1980, only two low-income countries have reported four consecutive years with agricultural GDP growth rates in excess of 8 percent. 9 Sustaining high rates over a longer period, even allowing for annual fluctuations, is also the exception rather than the rule. Since 1999, only three-low income countries have exhibited growth rates in excess of 8 percent for 6 years within a 10-year period (Rwanda 1995-2004; Malawi 1991-2000; and Tajikistan 2000-2008). Rwanda is an example of a post-conflict scenario, where higher sector growth rates are possible as the economy returns to its production frontier. As the economy transitions from post-conflict recovery to a normal development trajectory, one would expect to see Liberia exhibit the typical growth dynamics of agriculture growth, but at a lower rate than non-agriculture sectors, given the low productivity of the sector. This transformation facilitates labor exit from agriculture into higher-paying employment in manufacturing and services, thereby increasing labor productivity-- and hence incomes--for those remaining in the sector. 4.5. The domestic non-tradable agriculture subsector has received relatively much less attention than the tradable sector, which has attracted foreign direct investment. However, the domestic agriculture subsector could make substantial contributions to the Government's pro-poor agenda. Cross- country experiences show that given the concentration of the poor in rural areas, improvements in agriculture can play a major role in poverty reduction. As pointed out by Besley and Cord (2007), taking into account the indirect effects on non-agricultural households, agriculture (principally food crops) accounted for 44 and 77 percent of poverty reduction in the 1990s in Ghana and Uganda, respectively; and up to three quarters of the poverty reduction from 1984 to 1996 in Indonesia. In Vietnam, 71 percent of the workers who moved out of poverty between 1993 and 1997 either remained employed in informal agriculture or moved into formal agricultural employment. The 2008 World Development Report (WDR) also highlights the fact that declining rural poverty has been a key factor in aggregate poverty reduction (see case of China, Box 4.1). 9 Data from the World Bank. These cases are: Ethiopia (2004-2007) and Tajikistan (2000-2004). - 23 - Box 4.1: China's Reduction in Rur al and Aggr egate Pover ty China's poverty reduction in the past 25 years is unprecedented. Estimates by Ravallion and Chen (2007) indicate that poverty fell from 53 percent in 1981 to 8 percent in 2001, pulling about 500 million people out of poverty. Rural poverty fell from 76 percent in 1980 to 12 percent in 2001, accounting for three-quarters of the total. The evolution of poverty has been very uneven over time, however. The sharpest reduction was in the early 1980s, with some reversal in the late 1980s and early 1990s. The sharp decline in poverty from 1981 to 1985 was spurred by agricultural reforms that started in 1978. The household responsibility system, which assigned strong user rights for individual plots of land to rural households, along with the increase in government procurement prices and a partial price liberalization, all created strong incentives for individual farmers. In the initial years of the reforms, agricultural production and productivity increased dramatically, in part through farmers' adoption of high-yield hybrid rice varieties (Lin 1992). Rural incomes rose by 15 percent a year between 1978 and 1984 (Von Braun, Gulati, and Fan 2005), and the bulk of national poverty reduction between 1981 and 1985 can be attributed to this set of agrarian reforms. The role of agricultural growth in China's poverty reduction remained important in subsequent years, as the reforms created the rural nonfarm sector, which provided employment and income to millions of people whose work was no longer needed on farms. The share of the rural nonfarm sector in GDP went from close to zero in 1952 to more than one-third in 2004 (Von Braun, Gulati, and Fan 2005). Considering the entire period, Ravallion and Chen (2007) concluded that growth in agriculture did more to reduce poverty than did either industry or services. Source: World Development Report 2008. 4.6. A transformation of Liberia's domestic agriculture could increase employment and contribute to pro-poor growth in a number of direct and indirect ways. First, domestic agriculture could be a source of considerable growth in employment, particularly for the rural labor force. Prior to the war, the agricultural sector provided employment for more than three-quarters of the labor force, while in 2008 this ratio had dropped to less than half. The bulk was employment was in the informal sector, including subsistence farming. Cross-country experience has shown that the agricultural sector generally has a higher elasticity of employment with respect to growth than other sectors. Current data available for Liberia do not permit an estimation of the employment elasticity in agriculture. However, recent estimates by the ILO (Key Indicators of the Labor Market--KILM), based on data from 1991 to 2003, suggest an employment elasticity of growth of 0.82 for sub-Saharan Africa. This means that a 10 percentage point growth in the agricultural sector would lead to about an 8 point growth in employment in the sector. Using the 2008 employment in the sector as a baseline, this would mean an increase of more than 50,000 new workers (assuming no increases in labor productivity). The elasticity for the agriculture sector is higher than that for the services sector (0.79), but lower than for the manufacturing sector (0.90). Given the weakness in infrastructure--including roads, electricity, and telecommunications--the higher public investment cost per job in manufacturing means that agricultural employment is currently a much cheaper source of job creation for Liberia. 4.7. Second, a vibrant domestic agriculture sector could help reduce the food import bill, thereby contributing to increased national- and household-level food security and improvement in the balance of payments. As noted above, food imports account for 25.6 percent of total imports. The recent food price crisis has highlighted Liberia's vulnerability to global food price increases, due to its heavy dependence on imported food. Food insecurity is a major issue for Liberia. The comprehensive - 24 - Food Security and Nutrition Survey (CFSNS) conducted in 2006 revealed serious levels of national food insecurity. It found that 11 percent of surveyed households are food insecure and 40 percent are highly vulnerable. The survey also found that 39 percent of children under the age of 5 are stunted, 29 percent are underweight, and 7 percent are wasted. In addition, over-reliance on a few foods prevents nutritional balance, particularly in rural areas. A recent survey of households in Monrovia suggests that food shocks are major concerns for those in the urban areas as well. Table 4.2 below shows the response from households in Monrovia to a request to name up to three shocks that had negative impacts. Based on the analysis of the responses, 79 percent of household reported that shock from high food prices was of most concern, followed by the concern of not having enough money to buy food or cover other basic needs. It is likely that Liberia's urban population will continue to expand faster than the rural population in the short to medium term, driven by perceived economic opportunities and security concerns. Consequently, urban and peri-urban agriculture could help to improve food security in urban areas, as the experience in Asia has shown (Tsubota 2007 and Razak and Roff 2006). Table 4.2: Shocks with Negative Impacts on Households in Monrovia Shocks % of respondents High food prices 79% Not enough money to buy food or cover other basic needs 44% High fuel/transportation prices 36% Sickness/health expenditures 33% Death household member/funerals 14% House damaged 9% Loss employment/reduced salary 8% Debt to reimburse 8% Insecurity/theft 8% Other shock 6% Irregular/unsafe drinking water 5% Payment house rental 4% Bad weather/heavy rains 4% Source: The Impact of High Prices on Food Security in Liberia, United Nations, July 2008. 4.8. Third, increasing domestic food production, particularly through increased productivity, could help reduce inflation, which is a tax on the poor. Food has the single highest weight (45.20 percent) in the Liberia Harmonized Consumer Price Index (HCPI). The possible direct and indirect negative effects of this were seen in 2008 during the global food price crisis. The rate of inflation for 2008 was 17.5 percent, up from 11.4 percent in 2007. However, core inflation, which excludes food and transport (with transport a relatively smaller portion), was only 6 percent--indicating the impact of the price of food on the general price level. Since the supply of labor far exceeds demand in both the formal and informal labor markets and productivity is very low, nominal wages are expected to increase only very slowly in the short to-medium term. Sharp increases in the general price level, and food prices in particular, could therefore be a severe tax on the poor including the working poor, eroding the real value of their only asset--their labor. As suggested above, it is critical that the rate of productivity increases must exceed the rate of price reduction if farm incomes are to be protected. - 25 - 4.9. Fourth, increasing productivity and production in domestic agriculture could help improve incomes in the rural sector, with other positive developmental effects. Increased rural incomes can be a driver of demand for both agricultural and non-agricultural products; while, conversely, low productivity and low earnings in the agriculture sector may in part explain the underdevelopment of the manufacturing sector. There are important intra-household dynamics that imply differing consumption effects of increased incomes for men compared to women. The existence of defined gender roles in agriculture is well documented, and augmenting on-farm opportunities for women can have important differences in terms of their consumption impacts. It is estimated that in African countries, women do at least 70 percent of the agricultural work (OECD, 2006). Comparable data are not available for Liberia, but women are disproportionately represented as unpaid employees (mainly in agriculture)--55 percent compared with 45 percent for their male counterparts. Additional incomes earned by women typically have a greater impact on children's nutrition and education than additional income accruing to men. On the other hand, research by Glick and Sahn (1998) suggests that when African women work outside the home, their families reap more income but often with potentially "deleterious consequences on the health of their young children." Increasing the productivity of domestic agriculture in a manner that allows women to get the same or greater output per unit of time spent working on the family farm appears to be a good option for both increasing the income of women and also allowing them to spend more time in child care at home. 4.10. Why has the considerable potential in domestic agriculture remained unrealized? Agriculture became an increasingly important part of the Liberian economy starting the 1930s, with the entry of Firestone into rubber production. By 1960, agriculture accounted for more than 40 percent of Liberia's GDP. Over the years, however, Government policies towards the sector have to a large extent focused on attracting investment into plantations, particularly rubber and oil palm, with relatively less attention paid to commercial crops such as cocoa and coffee, and even less to noncommercial farming. With the country's ability to finance imports (including fuel and rice, the main food staple) being driven by the earnings from the few key traditional export sectors, there has been little or no incentive to develop the domestic agricultural sector. The low investment in domestic agriculture is reflected in its expenditure budget. In the 2008/09 budget, agriculture accounted for less than 3 percent of the total expenditure. This places Liberia well below the 2003 average of 5.6 percent for 31 African countries, as well as the 2005 average of 6.6 percent for 24 African countries. 4.11. One of the primary issues affecting domestic agriculture is secure access to land. In the Comprehensive Food Security and Nutrition Survey (CFSNS) conducted in 2006, 66 percent of those surveyed reported having access to land, although 41 percent of households reported that farm sizes were smaller than what they had prior to the war. In terms of the security of tenure, the majority of households (67 percent) did not have deeds for the land to which they currently have access. - 26 - Table 4.3. Agricultur al Constr aints by Type of Agr icultur al Household (HHS) Farming HHS with land but HHS without Total HHS (49%) not farming (18%) land (34%) Lack of seeds 50% 56% 46% 50% Lack of tools 47% 52% 54% 50% Lack of financial capital 29% 39% 30% 31% Lack of household labor 27% 37% 23% 28% Groundhog attack 30% 10% 7% 19% Bird attacks 17% 5% 5% 19% HH engaged in other activity 10% 12% 18% 13% Lack of arable land 3% 3% 34% 13% Returned late for planting season 2% 25% 3% 6% Source: Comprehensive Food Security and Nutrition Survey (CFSNS), 2006. 4.12. As reported in the CFSNS survey, among both farming and non-farming households, the three most important constraints to agriculture are lack of seeds, lack of tools and lack of financial capital. The fourth major constraint, reported as lack of household labor, is rather surprising given the average household size as well as the high level of unemployment and underemployment in Liberia. There are two possible explanations: (a) poor households have no savings or other assets to fall back on during the planting and growing period and therefore cannot afford to forego paid employment to prepare land, plant, and tend the crops to harvest; (b) while underemployment is the norm at specific points in the agricultural cycle, such as rain-dependent land preparation, labor availability does become a binding constraint. In either case, it is worth exploring whether support from the government for mechanical land preparation would allow the farmer to undertake paid employment while the land is being prepared and also have a larger plot of land to plant than he could otherwise prepare with his own labor. An even stronger case can be made for the government assisting with mechanization of post-harvest processes, which would not only help reduce post-harvest losses, but also result in higher quality and more marketable produce. Given the high level of unemployment and underemployment, however, it will be important to strike the correct balance between the use of labor and machines. 4.13. Poor-quality seeds and planting material, loss from animal pests, low labor productivity due to lack of tools, the absence of savings or other assets to provide a safety net, and uncertain and variable prices for produce, all combine to make domestic agriculture an extremely risky venture for the poor in Liberia. There is considerable anecdotal evidence to suggest that the perceived risks attached to agriculture, compared with the probable high pay-off from itinerant mining for gold, make domestic agriculture relatively less attractive. Households with little or no income-generating assets are likely to be more risk averse. 4.14. Investors' reengagement with the plantation sector since the war ended has the potential to provide substantial employment opportunities, either directly as wage labor or through outgrower schemes for surrounding small farmers. Liberia's rubber and oil palm production was badly affected by the civil conflict, with many (typically foreign) concessionaires withdrawing their capital, extensive - 27 - destruction of physical assets (in particular processing facilities), and the occupation of a number of plantations by ex-combatants, often leading to the "slaughter-tapping" of trees. However, existing concession holders are now re-engaging and new entrants are taking over old plantations and expanding into new areas, spurred by recent commodity price booms. 4.15. Complete employment data for all the concessions do not exist, but available estimates illustrate the direct employment potential from a vibrant plantation sector. Existing employment is substantial, wages are comparatively high, and employees benefit from additional health- and education- related benefits provided by concessionaires. According to the 2006 joint UNMIL ­ Government of Liberia Rubber Task Force Report, Firestone employs 6,500 company workers plus an additional 3,800 contractors (although it employed 20,000 in the 1960s). The minimum wage for employees is $2.65 per day. Similarly, Liberia Agriculture Company (LAC) employs 1,300 company workers and 1,500 contractors. Cavalla provides a further 1,350 jobs and Guthrie--whose plantations were once home to 3,700 ex-combatants--now employs about 1,500 workers (many drawn from ranks of ex-combatants). Importantly, there is substantial scope for increasing employment as plantations are rehabilitated and as underutilized areas within concession agreements are brought into productive use. Short-run impacts are likely to be even greater since expanding the planted area (augmented by the need to replant existing planted areas) is more labor-intensive that tapping. Table 4.4: Existing Rubber and Oil Palm Concessions a Plantation Location Concession Area Planted Area Cavalla Maryland 20,000 acres 15,000 acres Cocopa Nimba 25,000 acres 8,500 acres Sinoe Sinoe 600,000 acres 50,000 acres Guthrie Bomi (80%) & GCMb 300,000 acres 22,000 acres LAC Grand Bassa 300,000 acres 32,000 acres Firestone Margibi 118,000 acres 55,000 acres Salala Margibi 21,000 acres 7,000 acres Total 1,384,000 acres 189,500 acres Source: IMF (2008). Notes: (a) These are colloquial names by which plantations are generally known. (b) Grand Cape Mount. A second important source of improved livelihoods from the plantation sector is through the expansion of arrangements that link large-scale nucleus plantations with surrounding farms, including small-holder family farms. Globally, tree crop plantations typically involve both large-scale plantations and associated outgrowers, many of whom are small-holders with a few hectares. Such arrangements benefit the nucleus by internalizing economies of scale from increased production volumes beyond what is feasible for a single vertically integrated plantation; as well as benefiting the smallholders by linking them to input and output markets and providing essential extension advice for improved productivity. Indeed, the Government of Liberia recognizes this potential and has mandated that all future concession agreements include outgrower schemes. 10 Technical assistance from the World Bank Group is supporting development of a framework for a national outgrower scheme that will set the parameters for such schemes. Specific metrics of what constitutes an appropriate outgrower framework in the Liberia 10 See Liberia's Poverty Reduction Strategy Paper (Government of Liberia, 2008), para. 14. - 28 - context still need to be defined. However, discussions with the industry suggest that 20-25 percent of developed land in new oil palm plantations should be for cultivation by outgrowers, and that each household could be expected to cultivate three to five hectares. On this basis, every 15,000 hectares rehabilitated or replanted--a realistic figure given the development plans of existing concessions--would imply between 600 and 1,250 additional beneficiary households. The availability of candidate outgrowers is not likely to be a problem, given that the population residing within any single concession is generally much larger than the number of existing employees. (For instance, while an estimated 100,000 people reside within the Firestone plantation, 30,000 to 35,000 within the LAC, and 20,000 within the Cavalla concession areas, less than 10 percent of them are employed at the plantations). 4.16. Along with its focus on transforming domestic agriculture, the Government must also give attention to improving the business environment for the formal sector. The formal sector is not only key to maintaining the growth momentum, but is also an important source of higher-quality jobs and fiscal revenues. The next section examines issues surrounding the limited contribution of the formal sector to employment, its vulnerability to exogenous shocks, and some of the constraints to its further development and diversification. - 29 - 5. MAJ OR CONSTRAINTS TO INVESTMENT AND J OB GROWTH IN THE FORMAL SECTOR 5.1. Between 1960 and 2007, despite a significant rise and subsequent decline in national income, Liberia's labor market has not undergone any significant structural transformation. The informal labor market has remained steady at approximately 80 percent of the labor force, the formal labor market is still dominated by the public sector and concessions agriculture, and educational attainment has remained low. Real wages appear to have peaked in the 1960s and 1970s, and have since been sharply eroded. The most significant shift in employment pattern has been a move from rural informal agricultural employment to urban informal employment (largely petty trading). As a result, Liberia's urban population has swelled from 25 percent of the population in 1970 to an estimated 40 percent of the population in 2007 (Figure 2.1). 5.2. While the modern sector grew rapidly between the 1950s and the early 1970s, the share of the labor force employed in the modern sector (wage earners and salaried employees) has remained relatively constant at around 20 percent of the total labor force, or 10 percent of the population. Over this period, the bulk of formal sector employment was accounted for by the public sector, the services sector, and commercial agriculture (Figure 5.1). The lack of growth in employment in the modern sector may in part reflect the dynamics of the increasing capital intensity of the modern sector. Growth in the modern sector has changed little since the civil war and the rebound of the traditional sector. In 2007, formal employment accounted for an estimated 22.8 percent of the employed labor force, the bulk of which was accounted for by the services sector, including government services. 5.3. The civil service, the core of public sector employment and the main machinery of government, is dysfunctional and poorly paid. One of the legacies of the war is a civil service which, though relatively large in size, does not have an adequate cadre of personnel with the requisite skills, competencies, and motivation to deliver efficient services to the population. As of 2008/09, the civil service consisted of an estimated 38,200 persons, or about one percent of Liberia's population. The education sector accounts for approximately one third of the civil service, while the security services account for about 14 percent. In the 2008/09 budget, personnel cost accounted for almost 30 percent of the overall budget. From the early 1980s to 1995, Liberia's civil servants enjoyed relatively high wages by West African standards. For example, in 1995, the highest grade (grade 16) and lowest grade (grade 1) earned US$16,717 and US$2,400, respectively. However, salaries collapsed after 1995, with the highest and lowest grades now earning US$600 and US$360, respectively. 5.4. The Government's current strategy is to "right size" the civil service and improve the pay structure, to motivate civil servants to deliver more efficient services. Over the next five years, the Government plans to almost double the size of health sector employment, as government takes up most of the health delivery gaps that have been left by the departure of many health NGOs as the country stabilized. The security service is expected to increase by 49 percent, while the education sector will see a modest 10 percent increase. At the same time, civilian central government employment in the other ministries (mainly the economic ministries) is expected to see a reduction from 18,173 in 2008/09 to about 15,584 in 2013/14. On balance, therefore, employment in the civil service is expected to increase by about 9 percent over the next five years. - 30 - Figur e 5.1. Sector al Shar e of Employment (1962, 1970, 1979, and 2007) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1962 1970 1979 2007* Monetary agriculture Industry (excl mining) Mining Services (excluding Govt) Government Sources: World Bank, Growth with Development, 1975; World Bank, Recent Economic Developments and Medium-Term Prospects, 1982.* Estimates for 2007 developed from CWIQ data, but exclude observations where the sector of employment could not be categorized. 5.5. Foreign investments in the traditional export sectors (rubber, oil palm, forestry, and mining) may raise and even sustain exports and high levels of growth, but are unlikely to create significant employment opportunities. The economy posted steady growth, averaging 4 to 7 percent in the 1960s, driven by investments and the rapid growth in the rubber, iron ore, and forestry sectors. Growth also rebounded sharply after the civil war to 5.3 percent in 2005, 7.8 percent in 2006, and 9.5 percent in 2007, before falling to 7.1 percent in 2008 as a result of the food and fuel price shocks. However, even with the strong growth in the 1960s, the total number of wage workers was estimated at only 80,000, or approximately 20 percent of the labor force; while the remaining 80 percent were employed in the informal/subsistence agriculture sector. High levels of growth driven by the traditional sectors do not translate into substantial growth in employment, because these sectors represent capital- intensive enclaves with very little direct or even indirect links to the rest of the economy. Some investors in these enclaves have even been forced to self-provide in non-core areas such as the wooden sleepers for the railway tracks as there is no domestic capacity to supply these inputs at competitive cost. This represents a clear opportunity for local private entrepreneurs and government to act to remove the "walls" of the enclaves and promote a greater integration of these sectors into the overall economy. One approach could be to provide creative financing for start-ups that includes forward contracts with upfront payments (with the appropriate discounts). For this to be feasible, however, low-cost legal remedies must be in place to enforce these contracts if necessary. - 31 - 5.6. Traditional commercial agriculture currently employs about 35,000 low-skilled and unskilled workers. Strong employment growth in the medium to long term can be expected from the planned development from the several new concessions, as well as the expansion of current holdings. The Diagnostic Trade Integration Study (2008) estimates that the areas under rubber and oil palm cultivation could increase by 25 and 30 percent, respectively, over the next 10 years. A recently awarded 220,000- hectare oil palm and rubber concession is expected to create more than 20,000 jobs within the next 10 years. In the forestry sector, based on concessions recently been negotiated as well as those currently under negotiation, the sector is expected to grow by 8-10 percent in 2009-11 and by 19 percent in 2012, before slowing to 15 percent in 2013. However, commercial logging is fairly capital intensive and employment estimates are generally unreliable. Based on past experience, employment in the forestry sector is unlikely to exceed 7,000 at full production, which is expected to be reached by 2012/13. The impact of the financial crisis on the forestry sector remains uncertain, but may slow down the rate of investment if prospective logging companies find it more difficult to raise the necessary investment capital. 5.7. The large-scale mining sector is expected to grow rapidly over the medium term, owing to the anticipated development of three large iron ore mines. It is expected that mining will increase as a share of GDP from about 1.2 percent in 2008 to about 28 percent by 2013. However, commercial mining operations are highly capital intensive and are expected to employ 10,000 workers at most by the time all three iron ore mines are in full operation. Labor needs will include unskilled, semi-skilled (machine operators, etc.) and skilled managers/professionals. In the short term, mining company activities are expected to generate employment in the construction and service sectors, as infrastructure is rehabilitated and the demand for services increases. The artisanal mining sector in Liberia is poorly mapped, and employment estimates vary widely. The CWIQ reported employment in artisanal mining at less than 5,000. The diamond and artisanal gold industries are small, and are not expected to grow significantly (total combined export value of just above US$20 million in 2008). Poverty among mine workers is high (71.2 percent). Improving productivity, increasing the safety of mining practices, and gradually formalizing the sector and thereby reducing illegal smuggling and exploitative practices, could help to reduce poverty among artisanal mining communities. 5.8. While the traditional sectors are key to growth, they are vulnerable to external shocks that can have immediate direct and indirect impact on employment. As Collier (2002) points out, the world prices of primary commodities are highly volatile, and since most African countries, Liberia included, are dependent on a narrow range of commodities, this exposes them to severe macroeconomic shocks. Thus the global economic crisis, which began in the latter half of 2008, has constrained their ability to create jobs in the short to medium term. The crisis has resulted in a sharp markdown of projected growth in the developing world for 2009, from 4.4 percent to 2.1 percent; and global GDP growth is expected to fall from 2.5 percent in 2008 to less than 1 percent in 2009. Although some recovery of the global situation is expected in 2010, the pace and timing of the recovery is uncertain, and the negative effects of the crisis may continue for several more years as global growth remains below potential. More importantly, recent commodity price forecasts (Table 5.1) suggest that external conditions may worsen for Liberia before they get better. Liberia's traditional exports are expected to suffer further price declines in 2010, while its two main imports--rice and fuel--are expected to increase in price in 2010. These anticipated developments in part explain the expected deterioration in the balance of - 32 - payments, from a deficit of 32.6 percent of GDP in 2009 to a much larger deficit of 47.5 percent of GDP in 2010. Table 5.1. Commodity Pr ice For ecast for Liber ia's Pr imar y Expor ts and Impor ts Commodity 2007 2008 2009 2010 Primary Commodity Exports Cocoa (c/kg) 195 260 200 190 Palm oil $/mt 780 950 700 650 Logs, Cameroon ($/cum) 381 535 440 400 Rubber (c/kg) 229 270 210 205 Gold ($/toz) 697 860 750 725 Iron ore($/dmt) define 85 141 120 105 Primary Commodity Imports Rice, Thailand, 5%, $/mt 326 660 446 457 Crude oil, average $/bbl 71 101 74 76 Memo: External Current Account -35.2 -52.5 -32.6 -47.5 Balance, incl. grants (% of GDP) Note bbl=barrel; mt=metric ton; cum=cubic meter; toz=troy oz; dmt=dry metric ton; kg=kilogram. Sources: World Bank and IMF. 5.9. Given the price volatility of Liberia's primary export commodities, could a stabilization fund help to mitigate shocks from downturns? Commodity price data from 1990 to 2008 show that the prices of more than half of Liberia's key exports have a coefficient of variation higher than 25 percent, and that rubber, which has accounted for more than 90 percent of exports since 2004, has a coefficient of variation of 49 percent (Table 5.2). Furthermore, as the correlation matrix in Annex 2 shows, the prices of Liberia's key exports are all positively correlated, and in the case of the major exports (rubber, gold, and iron ore) they are highly correlated. This means that the cycle of boom and bust is likely to be very pronounced for Liberia, with a relatively large impact on fiscal revenues. A number of countries with concentrated export structures--including Norway, Chile, Venezuela, Kuwait, Colombia, Nigeria, and Canada--when faced with the issue of cushioning the domestic economy from volatility in commodity prices--have turned to stabilization funds. In some countries, in addition to the income smoothing function, these funds also serve the purposes of mitigating the Dutch Disease effect and ensuring intergenerational asset transfer from natural resources wealth. 5.10. There is broad theoretical and empirical support for the use of these funds as a mechanism to self-insure, particularly in cases where the country has limited access to the international capital market and where export diversification is of necessity a longer-term prospect. This is currently the case for Liberia. In a review of the experience with stabilization and savings funds in Norway, Chile, Venezuela, USA (State of Alaska Permanent Oil Fund), Kuwait, and Oman, Fasano (2000) concluded that: (a) the savings funds in Kuwait, Norway, and Alaska have helped build sizable assets to meet future needs; (b) in most cases, stabilization funds have contributed to enhancing the effectiveness of fiscal policy; (c) in Venezuela and Oman, the experience has been less successful because of frequent changes - 33 - in the fund's rules and the deviation from its intended purposes; and (d) stabilization schemes have been more successful in countries with a strong commitment to fiscal discipline and sound macroeconomic management. Table 5.2. Pr ice Volatility of Liber ia's Pr imar y Expor ts, 1990-2008 Cocoa Coffee Palm Logs Rubber Gold Iron Timber oil ore Average price 169.42 214.95 391.0 390.23 126.43 613.10 86.55 122.07 Standard 40.37 87.67 187.32 62.31 61.83 153.02 28.44 22.65 deviation of prices Coefficient of 24% 41% 48% 16% 49% 25% 33% 19% variation of prices Source: Staff calculations, based on monthly price data from World Bank database. 5.11. Liberia's weak enabling environment constrains investment in the manufacturing sector, limiting its ability to provide more and better-quality jobs and help diversify the economy to reduce vulnerability to external shocks. Liberia's nascent manufacturing sector currently employs an estimated 3,000 skilled and unskilled workers. But as experiences in other countries have shown, there is potential for a larger manufacturing sector with greater employment. Labor-intensive manufacturing has been successfully used by China, India, Indonesia, and Morocco to increase employment opportunities and reduce poverty. This has depended on those countries improving the investment climate to enable the development of manufacturing enterprises. The Government of Liberia has also taken some actions to improve the investment climate, including revision of the Investment Code. In 2007, Liberia ranked 170 out of 178 countries in the World Bank's Doing Business Survey. After some of the reforms took effect, the ranking in 2009 improved to 157 out of 181 countries. The ranking further improved to 149 out of 183 countries in 2010. In 2010, Liberia was cited as one of the top ten reformers for key indicators, including starting a business, processing construction permits, and trading across borders. Overall, however, the enabling environment remains weak, and further structural reforms are needed to make Liberia a more competitive location for business. Liberia is currently ranked at the mid-point of Economic Community of West African States (ECOWAS) countries on the Doing Business indicators for 2009 and 2010 (Table 5.3). Relative to ECOWAS better performers, Liberia is weak in a number of areas, including licensing, registering property, protecting investors, enforcing contracts, and closing businesses. A FIAS mini- diagnostic analysis of Liberia's investment climate conducted in 2006 identified a number of structural issues (Box 5.1). In addition to the issues raised in the FIAS diagnostic report, poor infrastructure (roads, ports, water and telecommunications) and the low levels of skills in the labor force are major constraints to increasing private sector investment. - 34 - Table 5.3. Doing Business Indicator s ­ ECOWAS Countr y Rankings in Sub-Sahar an Afr ica, 2009-2010 Dealing with Paying taxes Registering Employing Protecting Starting a Enforcing contracts investors Closing a business business property workers licenses borders ranking Trading Getting Overall across credit 2009 2010 --------------------------------------------------2010----------------------------------------------- Ghana 6 7 24 32 27 1 13 5 14 5 5 16 Nigeria 12 13 13 36 4 46 8 8 26 23 17 13 The Gambia 15 19 14 12 12 18 20 44 44 4 9 22 Cape Verde 19 20 25 14 38 21 28 21 20 3 2 34 Burkina Faso 22 21 15 13 11 17 28 27 30 44 19 18 Sierra Leone 23 22 5 42 37 44 17 3 36 18 30 29 Liberia 27 23 4 26 21 43 20 27 15 11 38 30 Mali 33 25 26 18 17 14 28 27 34 29 26 20 Senegal 23 26 12 24 40 39 28 43 42 2 33 11 Togo 31 30 39 31 34 34 28 27 32 6 35 14 Côte d'Ivoire 30 32 40 40 24 29 28 34 31 33 22 5 Benin 36 36 32 25 28 21 28 34 39 13 44 24 Guinea 37 37 43 41 10 38 40 44 41 14 25 17 Niger 38 38 33 39 41 10 28 34 29 42 27 27 Guinea Bissau 44 44 46 20 43 45 28 21 24 12 29 34 Source: World Bank Doing Business database. Box 5.1. Issues Affecting Liber ia's Investment Climate 1. The size and breadth of the informal economy impedes competition and has a strong negative impact on growth; likely causes of the large informal economy include years of conflict, weak rule of law, administrative barriers, rent seeking, and lack of information; 2. By law, the economy is not open to all residents of Liberia, which impedes growth and prevents job creation; for those accorded partial rights, the cost of doing business is made artificially more expensive and creates a disincentive to invest; 3. Compliance with the law is weak due to a culture of corruption and lack of training and information; 4. Low capacity in both the public and private sectors threatens the creation and implementation of a new policy framework to encourage private investment. Considering the low level of government capacity to encourage and support investment, it appears that existing government resources are being misallocated; instead of creating unnecessary barriers to doing business (licensing, inspections), often driven by rent seeking, resources would be better allocated toward legitimate revenue collection and implementation of investment-friendly laws and procedures. Source: Liberia: FIAS Mini-Diagnostic Analysis of the Investment Climate, March 2006. 5.12. Wage data are scarce in Liberia, as there is no systematic data collection by either the public or private sector. The only data available are for the basic wage of rubber tappers in 1960 and 1975, which can be assumed to be a proxy for semi-skilled agriculture labor. However, this data excludes bonuses, benefits, and food subsidies, which together account for a substantial portion of total earnings. This reflects a common feature of the Liberian labor market, where allowances are significantly larger - 35 - than basis pay. (It is not clear whether this practice is intended to avoid taxes.) Wage data suggest a decline in real earnings between 1960 and 1975. In 2007 US dollars (deflated by CPI), tappers earned an estimated US$6.75/day in 1960 and US$5.30/day in 1975. The fall in base earnings in the 1970s may, however, have been more than compensated by the introduction of a production-based bonus program in the late 1960s. Liberia has statutory minimum wages, which are set by the Minimum Wage Board in the Ministry of Labor. The Minimum Wage Law provides that an unskilled laborer must be paid not less than twenty-five cents (US currency) an hour if he is an industrial laborer, and not less than one dollar and fifty cents per eight-hour day if he is an agricultural laborer, exclusive of fringe benefits. Individual and cross- country studies in Africa have found no significant effect of minimum wage on employment. Although the lack of data precludes such an analysis for Liberia, the likelihood is that employment in most sectors--with the possible exception of agriculture--is invariant of the current minimum wage, since in most cases wages (cash and kind) are already higher than the statutory minimum. 5.13. The current industrial minimum wage of US$2 per 8-hour day equivalent has been maintained since 1972. Figure 5.2 uses data from the Food and Agriculture Organization (FAO) to plot the purchasing power of the minimum wage over 1972-2008, expressed in terms of the kilograms of rice it can buy at international prices.11 Overall, the purchasing power of the minimum wage increased slightly after dropping sharply in 1972. Since 2002, however, its purchasing power has been pushed below the trend line as a result of price increases, which were particularly sharp in 2004 (20 percent), 2005 (20 percent), and 2008 (99 percent). 5.14. An important policy question for the Government is whether to raise the minimum wage to attempt to capture a larger portion of the rent from the natural resource-based sectors for the workers. Raising the minimum wage could be beneficial for those in the formal wage sector. It could also have a positive impact on other workers through increased demand for goods and services. On the negative side, there are two primary concerns: (a) the possible negative impact of a higher wage rate on the demand for labor within the targeted sectors; and (b) the possible negative effect on the demand for labor in the agriculture and services sector, where productivity is already low. These questions can only be adequately answered through the construction of an economy-wide model, for which there are not sufficient data in Liberia. 11 Rice is a staple in Liberia. According to Tsimpo and Wodon (2008), rice (imported and locally produced) is by far the largest food consumption item, accounting for more than one third of the value of total food consumption. - 36 - Figur e 5.2. Pur chasing Power of Minimum Wage (1972-2008) 14.00 12.00 10.00 KG of Rice 8.00 6.00 4.00 2.00 0.00 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 Years Series1 Linear (Series1) Source: Staff calculations, based on FAO rice price data. Note: Series 1 is the amount of kilograms of rice that can be purchased with the daily minimum wage. Linear (Series 1) is a linear time trend of the kilograms of rice that can be purchased with the daily minimum wage. Labor Market Regulations 5.15. In the 2009 World Bank Doing Business Survey, Liberia ranked 105th out of 181 countries on the "Employing Workers" indicator. This was the same ranking as the previous year. On the "Rigidity of Employment Index," however, Liberia's score was 31, suggesting that labor market regulations governing the hiring and firing of workers are not very rigid. In fact, Liberia ranked 18th out of 46 sub-Saharan countries on the Doing Business indicator for "Employing Workers". Liberia's scoring on key indicators related to the hiring and firing of workers in comparison with other countries in the region, as well as with the average for OECD countries, is summarized below (Table 5.4). Table 5.4: Doing Business Indicators--Employing Workers Indicator Liberia Region OECD Difficulty of Hiring Index 33 39.0 25.7 Rigidity of Hours Index 20 43.5 42.2 Difficulty of Firing Index 40 41.5 26.3 Rigidity of Employment Index 31 41.3 31.4 Firing Costs (weeks of salary) 84 68.3 25.8 Source: Doing Business 2009, World Bank. - 37 - 5.16. Liberia scores above the average for the region on most indicators, with the exception of "Firing Costs." The survey estimates that it would cost a firm approximately 84 weeks of salary to fire a worker in Liberia, compared with an average of 68.3 weeks for the region and just 25.8 weeks for the OECD countries. The Firing Cost indicator measures the cost of advance notice requirements, severance payments, and penalties due when terminating a worker. 5.17. The Liberia Labor Code provides relatively generous termination benefits for workers. Section 1508 (1) of the Code provides that "No employer shall dismiss any employee with whom he is bound by a contract for a definite period before the end of that period unless it is shown that the employee has been guilty of a gross breach of duty or a total lack of capability to perform. Where this has not been proven, the dismissed employee shall be entitled to claim full remuneration for the unexpired portion of the contractual period." Section 1508 (2) defines "gross breach of duty" as follows: "The following acts and violations shall be deemed to be gross breaches of duty (without limiting the generality of the term) within the meaning of Section 1 of this Chapter and shall dispense the employer from payment of compensation for dismissal under the provisions of that Section: (a) any unprovoked assault by an employee upon the employer or his agents in the course of or arising out of employment; (b) persistent disregard by any employee of the technical measures for safety of the staff of the undertakings; provided that the said measures have been in rules posted as required by law and the employer or his agenda has ordered the employee in writing to comply with the said rules; (c) disclosure by an employee of the working secrets of the employer's undertaking; (d) absence of an employee for more than ten consecutive days (or more than 20 days over a period of six months) without good cause, in which case the employee shall be deemed to have terminated his contract. Save in the case of vis major, an employee shall be required to notify the employer or his agent of the reason for his absence." 5.18. The specific provision for termination benefits, as contained in section 1508(3), are as follows: "Where the contract is concluded between the employer and the employee for an indefinite period, the employer shall have the right to dismiss the employee on condition that he gives him/her two weeks written notice in the case of non-salaried employee and four weeks written notice in the case of salaried employee or payment in lieu of such notice, provided however that the employer shall also pay to the employee as follows: (a) Non-salaried employee: In the case of a non-salaried employee, the employer shall pay six weeks for each year of service, including any accrued wages and all unpaid benefits. (b) Salaried employee: In the case of a salaried employee, the employer shall pay one and a half month[s] salary for each year of service, including any accrued wages and all unpaid benefits. - 38 - (c) An employee who has worked not less than ten (10) years under the same employer, shall not be dismissed without cause as defined in Subsection (2) of Section 1508 of the Labor Practices Law of Liberia." 5.19. A 2007 review of Liberia's Labor Law compiled by the America Bar Association and the Africa Rule of Law Institute found that section 1508 (3) of the Labor Code provides for compensation that is heavily in favor of the employee, and that this could discourage formal employment in Liberia. The reviewers also noted that Liberia's Labor Law provides for the highest level of compensation in the world in the event of an employee being dismissed without cause. - 39 - 6. PUBLIC WORKS ARE NECESSARY FOR THE VERY POOR 6.1. Labor-intensive employment programs, including public works, can help to rebuild social and economic assets quickly, and also buy crucial time until the private sector expands and the diversification strategy takes root, to allow the economy to absorb a larger proportion of the labor force at reasonable wages. Labor-intensive programs should not be seen as a single "silver bullet" but as one of the elements of a comprehensive strategy which has short-term, medium-term and long-term elements to address the issue of the lack of gainful employment in Liberia. At the same time, labor- intensive public works have advantages over other strategies for infrastructure building and employment creation in Liberia. This section provides a brief review of some of the literature on the topic, building on the experience of other countries that have implemented such programs. Comparing labor-intensive versus equipment-intensive public works 6.2. The ILO defines employment-intensive or labor-intensive projects as those projects where labor is the dominant resource. When considering labor-intensive public works to build infrastructure, policymakers must first look at whether this type of program would have a negative impact on the quality of the infrastructure built, as opposed to a program using equipment­intensive techniques. Over time, the provision of infrastructure in many countries has shifted from being predominantly labor based to equipment based. This shift has been particularly dramatic for developed countries where wage rates have been increasing, but it is also taking place in some developing countries. Nevertheless, work done by the World Bank and the ILO has shown that for countries facing strong demand for infrastructure in the face of significant unemployment, labor-based provision of infrastructure remains a viable alternative. 6.3. World Bank research12 in the first half of the 1970s found that under certain circumstances, labor-intensive construction could compete with machines on both technical and financial terms. This finding was based particularly on the experiences in India and Indonesia. Beginning in 1975, the Bank's research expanded to Honduras, Kenya, Chad, Benin, Lesotho, and the Dominican Republic. Importantly, the study also found that for countries with no traditional experience with labor-intensive technologies, a minimum of three years of preparation was required for mobilization, staff training, and the introduction of specialized institutional arrangements before the large-scale use of unskilled labor could commence. This important finding points to a possible trade-off between the speed of implementation of public works versus their labor intensity. 6.4. A comparative analysis for Lesotho and Zimbabwe in 1995 13 concluded that the labor-based approach was technically and financially competitive with the equipment-based approach in both countries. The analysis was based on the construction of gravel roads. In a similar comparative analysis of rural road work in Cambodia in 2003, 14 the ILO found that: (a) the overall weighted average cost of labor-based works was 17 percent lower than equipment-based works; (b) that irrespective of the implementation modality (i.e., force account or contracted works), the weighted average cost of labor- 12 World Bank. 1986 The Study of the Substitution of Labor and Equipment in Civil Construction (SOL). 13 International Labour Organization, "Technology Choice, Man or Machine: Including Case Studies from Lesotho and Zimbabwe," Maria Lennartsson and David Stiedl, 1995. 14 International Labour Organization. "Jobs or Machines: Comparative Analysis of Rural Road Work in Cambodia," Paul Munters, 2003. - 40 - based project was consistently lower than that of equipment-based projects. When works were carried out using force account operations, the cost savings was 9 percent compared with using equipment-based operations. For the contracted labor-based operations, the cost savings increased to 37 percent compared with equipment-based operations; and (c) there was a very large employment potential in rural road works. The average unskilled labor content of equipment-based work was as low as 1 percent of total costs, compared to 37 percent for labor-based works. 6.5. The employment-based approach has been shown to have a much wider positive general macroeconomic impact. An ILO study (1999) on Madagascar estimated the differential effects of employment- versus equipment-intensive approaches to investment projects, with regard to key economic variables including production, consumption, employment, public finance, and foreign trade. The analysis suggested the superiority of the employment-based approach, which was 30 to 80 percent less costly, created 2.5 times more jobs, increased national income and household consumption by 2.5 times, and required only 30 percent of the foreign currency used by the equipment-based investment projects. Comparing labor-intensive public works to other employment programs 6.6. Although the cross-country review suggests that direct job creation, including through labor-intensive public works, is by far the most popular, governments have also made use of indirect job creation programs (Adato and Haddad 2001, ILO 1995 and 1999,Tesfaye 1995, Subbarao 1997, von Braum et al 1991). Indirect job creation programs include wages subsidy schemes, entrepreneurship and apprenticeship programs for youth, and job search assistance. The use of wage subsidies is based on the theory that the position of low-skilled workers can be improved by reducing their cost to the employers. The lower wage cost is expected to increase the demand for low-skilled labor and consequently increase the level of employment. Wage subsidy programs have been found to be administratively difficult even for developed countries. Targeted schemes are complicated and may cause serious labor market distortions. General subsidies are relatively easier to administer but more costly, since they also subsidize those workers who would still have a job, absent the scheme. Wage subsidy schemes have been used mostly in developed countries and in middle-income countries such as Argentina. Experience with labor-intensive public works 6.7. Cross-country experience with labor-intensive initiatives shows that the welfare impacts have generally been positive, but that results have been mixed in terms of the quality and sustainability of the assets produced. Public employment programs were pioneered in South Asia to deal with huge open unemployment. The Maharashtra Employment Guarantee Scheme (MEGS) which dates back to 1971 has served as an effective safety net for the poor in India. Bangladesh has used public employment programs since 1962, largely financed by external donors. In Sri Lanka, labor-intensive public works were used to cushion the adverse effects of structural adjustment in the 1980s. In Latin America, the outcome has been equally positive. According to Subbarao (2003), nearly 100 percent of the participants in Chile's public works program belonged to poor households. In Argentina's Trabajar program, 60 to 70 percent of the participating households were poor. In Africa, labor-intensive public works programs have been implemented both as freestanding programs and as components of Social Fund programs. In general, stakeholders have had a positive impression of the programs' income-generation - 41 - and capacity-building impacts, and were particularly pleased at the speed with which the temporary jobs were created. The public works program in South Africa, which is considered one of the most innovative, has multiple objectives, including job creation, poverty reduction, infrastructure development, job training, and community capacity building. 15 6.8. In general, public works or other job creation programs tend to have multiple, non- exclusive objectives. Cross-country experiences show that projects or programs in support of labor- intensive work focus on a range of assets, including roads (mostly rural but also urban), markets, schools, health centers, urban drainage systems, water supply systems, irrigation systems, reforestation, anti- erosion structures, land reclamation, housing, and solid waste management. Roads tend to be the most popular choice across the world. Whichever works are chosen, cross-country experience clearly shows that works that are demand driven and reflect the choice of the communities will likely be better implemented and more sustainable that those that are supply driven, even when the communities benefit from the jobs created by the latter. Social funds have been particularly successful at encouraging community participation (Box 6.1), using different participatory tools to get communities involved in selecting and designing priority projects, in supervising implementation, and in maintaining the public works. Some have also successfully experimented with community contracting (e.g., Jamaica Social Investment Fund, Malawi Social Action Fund, Bolivia Social Fund). Beneficiary participation not only builds ownership of the project, but is also an essential component of good governance. 15 Adato and Haddad, Targeting Poverty through Community-Based Public Works Programs: A Cross-Disciplinary Assessment of Recent Experience in South Africa." IFPRI Discussion Paper 121. - 42 - Box 6.1. Wor ld Bank Exper ience with Social Funds Involving Labor -Intensive Public Wor ks The Second Social Fund for Development (SFD) Project in Yemen (2000-2006) aimed at improving the range of services and options available to the poor through a combination of community development, capacity building, and microfinance programs. The project promoted (a) community consultation and participation in the operation and maintenance of community infrastructure and services, with SFD providing capacity building and support; (b) increased access of the poor to financial services, through the development of sustainable microfinance intermediaries; (c) increased income-generating opportunities, with SFD providing technical and financial assistance, contingent on time-based performance benchmarks; and (d) strengthening of community capacity to effectively identify, implement, and operate development projects. Lessons learned. Strengthening targeting mechanisms based on periodic reviews of implementation experience allows for more effective channeling of project development resources to the poorest and most vulnerable. Well- designed monitoring and evaluation systems allow for meaningful tracking of project performance and achievements, responsiveness to priority needs of target beneficiaries, and informed decision making. Aligning capital investments with capacity development can maximize the impact of project resources. Commitments, action, a strong drive for results, and flexibility to respond to evolving needs are important elements in project effectiveness. Maximizing development impact requires that interventions be designed in an integrated fashion and that the necessary inputs, including software, are in place. The National Social Action Project (NSAP) in Sierra Leone (2003-2009) aimed at assisting war-affected communities to restore infrastructure and basic services and build local capacity for collective action. The project supported: (a) demand-driven, community-based social and economic infrastructure projects, social mobilization, capacity building for community groups, and project monitoring and evaluation; (b) rehabilitation of rural infrastructure and provision of basic shelter to needy households, especially female-headed households; and (c) capacity building activities for communities as well as ministries. The National Commission for Social Action (NaCSA) was set up by an Act of Parliament to implement the project. Lessons learned. Community-driven development (CDD) approaches that are introduced in a gradual and politically sensitive way are likely to be more successful. Project processes that are driven by community demand build greater project ownership and greater community cohesiveness. One important lesson is that infrastructure constructed under direct supervision of the beneficiary community can result in lower costs even as technical soundness is maintained. Community mobilization should be carried out not only when identifying community needs and priorities, but throughout the implementation process. As they are mobilized, beneficiaries come to understand the importance of staying involved throughout the project cycle and even beyond, to maintain the infrastructure and services put in place. Source: Sum and Hackett, 2007. 6.9. Reports of the experience of labor-intensive public works have generally been positive, particularly on the welfare objective. In a review of programs implemented in selected countries to protect the poor during transition or adjustment, Subbarao, Braithwaite, and Jalan (1995) note that the welfare benefits of a public employment program in India were reasonably well-targeted to the poor. According to Rao, Subbarao, and Ray (1988), public employment programs in India have effectively reduced the adverse consequences of droughts. Public works program also accounted for a substantial - 43 - portion of national employment in Botswana in 1985-86 (21 percent) and Chile in 1983 (13 percent).16 The assessment of the welfare impact of the Bangladesh program was also positive. In other cases, the welfare impacts were not as significant as they could have been, largely due to ineffective targeting. Selected challenges faced by labor-intensive public works 6.10. Labor-intensive programs face many challenges. At the political/institutional level, a major challenge is to reconcile the short-term need for job creation with longer-term strategic goals. There is also tension between the need to complete projects quickly and for time-consuming consultations and other community-based activities to ensure greater ownership of the assets. The following section elaborates on the challenges that countries face in implementing labor-intensive public works. 6.11. Financing. To meet the overall social objective of providing short-term employment and building social and economic assets, financing for labor-intensive public works should not only cover the implementation of the project but also, where relevant, make provision for operation and maintenance. This will help prevent the creation of "white elephants." Financing for labor-intensive initiatives has come from a mix of sources, including government budgets and bilateral and multilateral donors. In very poor countries with severe fiscal constraints, financial support for such programs has also come as grants from several bilateral donors, including the Netherlands, Germany, and Denmark. Financing has also come indirectly through social funds financed by the World Bank (e.g., Egypt Social Development Fund) and regional development banks. 6.12. Fiscal space. One of the key challenges for many developing countries is limited fiscal space, and consequently the lack of resources to support a large enough labor-intensive public works program to have a meaningful impact on unemployment. In such cases, targeting a specific unemployed population could help narrow the scope of the project to better fit the resource envelope. Poverty maps and global information system (GIS) techniques could be extremely useful in tying the poor to the works. Cross- country studies (Tesfaye 1995) that highlight government financing of labor-intensive programs suggest that many of these programs have been part of the capital or development budget. In Botswana between 1984 and 1987, for example, such programs accounted for between 14 and 18 percent of total development expenditure. In Tanzania, the Government contributed about 11 percent of the total capital fund to such a program between 1980 and 1990. 6.13. Wage rate. Setting an effective wage rate for labor-intensive public works is another critical challenge, since the wage rate explicitly or implicitly determines (a) the program's effectiveness in targeting the poor; (b) its implications for the formal labor market (through its influence on the reservation wage); (c) the productivity of labor; (d) the overall scope of the program (the number of poor who can be reached); and (e) the sustainability of the program (the length of time that fiscal or grant resources can sustain the level of employment at the current wage rate). 6.14. Most labor-intensive programs use the wage rate as the primary targeting mechanism to ensure that the benefits accrue to the poor and the unemployed. The higher the project wage rate, the higher the welfare impact (holding other factors constant). However, the higher the project wage rate, the more 16 Coady David. P. "Designing and Evaluating Social Safety Nets: Theory, Evidence and Policy Conclusions" International Food Policy Research Institute, Washington D.C. January 2004. - 44 - attractive employment by the project becomes for the non-poor. Coady (2004) has also argued that higher wages may have a beneficial second-round effect if they help reduce employer power in monopolistic labor markets. Coady suggested that one way of reducing the participation of the non-poor at the higher wage rates is to combine higher wages with an administrative targeting method to ration employment. However, a higher wage rate could also mean a smaller project scope, given the ever-present budget constraints. 6.15. Wage setting practices vary across countries. In Botswana, for example, wages were set at about 70 percent of the lowest minimum statutory wage rate for unskilled labor in the formal wage sector. This wage rate was considered low enough to make the program self-targeting. On the other hand, in Tanzania, the established wage was close to the minimum wage, and it was applied uniformly across all projects regardless of location. There was also no gender differentiation in the wages paid. In some cases, to bypass the official minimum wage, which was considered too low to attract labor, some projects have shifted to a task rate or have provided workers with food in addition to the official wage. 17 In Ethiopia, wages varied across income groups and projects. For example, on some project sites, the poorest households received payments that exceeded standard wages (i.e., the wages received by less poor households in the same community). 18 In a survey of ILO-supported labor-intensive investments across 11 countries, Miller (1992) found that the average wage was about US$0.83 per day, or about 79 percent of the statutory minimum wage of US$1.05 per day and about 90 percent of agricultural wages (Table 6.1). Table 6.1: Compar ison of Pr oject Daily Wages with Minimum/Agr icultur al Wages (US$) Country Statutory Agricultural Wages in Project wages minimum wages wages projects as % of : surveyed Minimum wage Agricultural wage Burkina Faso 3.37 1.13 0.97 28.8 85.8 Burundi 0.47 0.47 0.59 125.5 125.5 Cape Verde 1.54 1.79 0.77 50.0 43.0 India 0.92 1.15 0.92 100.0 80.0 Nepal 0.77 0.93 0.77 100.0 82.8 Rwanda 1.27 0.93 1.27 100.0 136.6 Sierra Leone 0.25 0.57 0.51 204.0 89.5 Sudan 0.72 0.90 0.60 83.3 66.7 Tanzania 0.16 0.33 0.18 112.5 112.5 Togo 1.95 1.15 2.16 110.8 187.8 Uganda 0.12 0.76 0.42 350.0 55.3 Average 1.05 0.92 0.83 79.1 90.2 Sources: Based on table in Miller 2002; and on Bank staff calculations. Note: Wages include cash and non-cash payments. 17 This practice is also observed in the private sector. For example, some private sector companies in Liberia give workers bags of rice in addition to their official wage. 18 See Webb and Shubh (1995). - 45 - 6.16. Distortions. In a review of the cross-country experience of the payment system and level of wages in the labor market, the Inter-American Development Bank (IADB) 19 found that public works programs generally do not create major distortions in the labor market to the extent that they offer wages below the relevant market and can provide a source of income for unemployed workers. On the other hand, the study argued that wage subsidy programs tend to generate large labor market distortions because they change the relative price of different types of workers in favor of a target group. 6.17. Gender. There is an increasing body of research which suggests that when women are empowered with resources, they have a larger favorable impact on household food security and investments in children's health, nutrition, and education. Labor-intensive programs have the potential to provide resources to poor women with appropriate gender targeting mechanisms. In a review of the gender dimensions of public works programs, Swamy (2001) poses the issues of women's access to the direct and indirect benefits from public employment schemes in terms of three key questions: (a) do women have equal access to the direct wage employment benefits offered by public works? (b) what factors of design and implementation determine women's participation? and (c) do women benefit equally from the assets created by public works? 6.18. Drawing on the cross-country experiences of 13 completed and ongoing ILO-supported projects in Africa and Asia, (Dejardin 1996) highlighted some of the issues related to these gender questions. On the issue of access to employment, the experience varied. On the Ruvuma Roads Improvement Project in Tanzania, women' share of employment ranged from zero in the Tunduru district to a high of 25 percent in the Songea district. In the first two years of the Burkina Faso Special Public Works Program (SPWP), women's share of unskilled employment ranged from zero in one village to 43 percent in Woro. Women comprised 30 percent of the unskilled in the Zambia Roads Project in 1987-1989. At the high end, the Rural Roads Maintenance Project in Tanzania registered female participation ranging from 16 to 59 percent. 6.19. On the question of factors (both project specific and non-project specific) that determine women's participation in public works, Dejardin (1996) suggested that in African countries, much of the information points to women's domestic work (including house work and farm work) as a major hindrance to their participation. In many countries, including Tanzania, Rwanda, Burundi, and Burkina Faso, most female construction workers were young (less than 30 years old) and unmarried. Of the project-specific factors, Dejardin noted the importance of the recruitment strategy. Recruitment procedures tend to assume that men and women respond in the same way. However, women tend to be less mobile, tend not to frequent administrative centers, and have less access to information about the availability of work. 6.20. Youth. Lack of employment for youth (age 15-24) is a problem for countries across the world. In several countries, where their anti-social behavior is reaching crisis proportions (including a disproportionately high level of serious crimes), their high rate of unemployment is receiving special attention. An inventory of youth employment based on some 289 programs and interventions in 84 countries (Betcherman et al 2007) provides the basis for some key policy lessons that are relevant for youth employment in Africa. One lesson is that in developing countries, public works programs might be 19 IADB, 2004. Good Jobs Wanted: Labor Markets in Latin America. - 46 - more suitable than formal sector wage subsidy programs, since wage subsidies do not have much effect in the absence of a large wage sector. Creating jobs through public works offers an alternative pathway to employment while at the same time investing in critical infrastructure needs. Public works programs also provide youth, especially low-skilled rural youth, with good opportunities to acquire initial work experience. 6.21. Overall, as discussed by Subbarao (1997), Ravallion (1998), and Hicks and Wodon (2001), key features of good public works programs include the following: · The projects should be targeted to poor areas and try to ensure that the assets created are of maximum value to poor people in those areas. Where possible, GIS technology should be used to overlay poverty maps with priority assets, to enable simultaneous targeting of poverty, assets, and employment creation. · Public works should be synchronized to the timing of the agricultural slack season. · The appropriate form of wage is important to encourage female participation. For example, women can benefit from piece rates or task-based wages; sometimes wages in the form of food have attracted more women to work sites. The provision of childcare or preschool services can also improve participation by women. · The wage rate should be no higher than the prevailing market wage rate for unskilled manual labor in the setting in which the scheme is introduced. · Restrictions on eligibility should be avoided; the fact that one wants work at this wage rate should ideally be the only requirement for eligibility. · If rationing is required (because demand for work exceeds the budget available at the wage rate set), then the program should be targeted to poorer areas, as indicated by a credible poverty map. However, future budget allocations should flexible to reflect differences across areas in demand for the scheme. · The program's labor intensity (share of the wage bill in total cost) should be as high as possible. The level of labor intensity will depend on the relative importance attached to immediate income gains versus other, longer-term gains to the poor from the assets created. This will vary from setting to setting. · Transaction costs to the poor should be kept low by locating project sites close to villages. · The program should provide for appropriate mediation of NGOs to protect the rights of the poor vis-ŕ-vis program managers. · The program should include an asset maintenance component. The Liberian context 6.22. Job creation is an imperative for the Government of Liberia. Fourteen years of civil conflict destroyed the country's social and economic infrastructure base and brought economic activity to standstill. This has resulted in large-scale unemployment and significant poverty. Growth has picked up since the signing of the Accra Comprehensive Peace Agreement in 2003, spurred in part by increasing foreign direct investment in traditional sectors. However, the current rate of job creation in these sectors, even under the most optimistic scenario, is unlikely to absorb a significant portion of the - 47 - unemployed and underemployed labor force in the short term. The lack of gainful employment, including for many unskilled youth (whose education was terminated by the conflict) and many ex-combatants, poses a significant risk to maintaining peace. One of the Government's major challenges is to address the immediate employment situation in the context of establishing a strategic framework for more sustainable, long-term job creation by the private sector. The strategy that is ultimately adopted should reflect not only the country's social, economic, and political needs, but also the Government's administrative capacity to manage implementation. The strategy should be financially and politically sustainable given the current and evolving fiscal space and political situation. 6.23. In terms of strategy options, labor-intensive public works appear to be the most natural fit to the current situation in Liberia. Labor-intensive public works can respond to the country's dual needs to (a) create social and economic assets to improve welfare and help create an environment for private sector-led growth; and (b) provide employment for a large number of unskilled workers, including women and youth. Wage subsidy schemes may be less appropriate for the current situation in Liberia, given the public and private sectors' limited capacity to absorb additional labor in the short term. Furthermore, targeted wage subsidy schemes are administratively more difficult to administer and generalized schemes would be more expensive. 6.24. A critical policy trade-off for the Government to consider is the need to (a) deliver infrastructure quickly to restore basic services and facilitate private sector investment for long- term growth, versus (b) the need to provide a high number of short-term jobs through labor- intensive reconstruction. Even under the most optimistic scenario, a labor-intensive public works program requires time to plan, undertaking the required training, and establish an effective institutional framework. One way to reconcile the trade-off is to undertake the labor-intensive work related to road maintenance, including bush clearance and drainage repairs, while the institutional framework is being established and the training is being done to facilitate more skill-intensive work such as road building. A clear delineation of priority roads, and the modalities for their repair and reconstruction, would facilitate planning to optimize the labor content without compromising delivery time. 6.25. The high unemployment of women and youth suggests that the strategy should address the particular needs of these two groups. Poverty rates do not show large gender differences, but when non-income dimensions of poverty are taken into account, women's vulnerability becomes apparent. Several non-income poverty factors affect women disproportionately; these include high fertility rates, high maternal mortality, high illiteracy, and low access to education. In addition, inadequate quality of and access to infrastructure, as well as standard housing conditions, contribute to women's time poverty. Time poverty reinforces income poverty by limiting women's access to economic opportunities (Blackden and Wodon, 2006). In Liberia, almost half of non-working women reported household and family responsibilities as the main reason for their inactivity, compared to 12.6 percent of men. As discussed above, youth in Liberia have been severely disadvantaged by the war in terms of their formal schooling and the opportunities for training. 6.26. Cross-country experience (del Ninno, Subbarao and Milazzo, 2009) suggests that the needs of women in Liberia could be addressed through labor-intensive public works designed with sensitivity to their needs. One important design feature is the flexibility of work time, which allows women to participate in the project and still meet their domestic responsibilities. Another important - 48 - design feature could be the provision of daycare facilities close to work sites, which would allow women to bring their children to work. The strategy for youth employment needs to be more forward looking, since youth are likely to have few skills and will remain in the labor market for a longer period of time. The strategy, therefore, needs to have some transitional component that will allow youth to move from the temporary employment program to more sustainable longer-term employment. Experiences with youth employment programs (Betcherman et al 2007) suggest that initiatives that combine paid work with substantial training or education activities appear to have the greatest impact on participants' future employment and earnings, producing gains well beyond those achieved through traditional "workfare" programs. Based on lessons from cross-country experiences with other programs and interventions in addition to labor-intensive programs, two additional types of strategic interventions seem relevant for Liberia. These are (a) targeted youth entrepreneurship schemes; and (b) traditional apprenticeships programs. 6.27. Opportunities for labor-intensive public works in both urban and rural areas include rehabilitation or reconstruction of roads, schools, health centers, and other small civil works, as well as the management of solid waste. Low-volume community and feeder roads lend themselves to labor-intensive works for both construction and maintenance. Some aspects of highway maintenance are also suitable for labor-intensive operation, except where the reconstruction of critical high-volume roads requires a machine-finished surface. Such roads generally need to be delivered quickly to support private sector investment. The current road network is about 10,000 km, about half of which may be suitable for labor-intensive operations. 6.28. Liberia's initial attempt at executing a formal temporary employment project has yielded some positive impacts and useful lessons,(Box 6.2) although a comprehensive assessment of the project has yet to be done. In 2008, the Government launched a pilot Cash-for-Work Temporary Employment Project (CfWTEP) as part of its strategic response to the global food crisis. To maintain access to food among vulnerable households, the Government established directed assistance through school feeding and supplementary feeding for pregnant and lactating mothers, as well as a cash-for-work scheme to provide additional income-generating opportunities for underemployed households during the lean season. The cash-for-work scheme was implemented through an existing institutional framework-- the Liberia Agency for Community Empowerment (LACE)--which had a good track record for managing community empowerment projects (Box 6.2). The planned assessment of CfWTEP in the context of the broader LACE program should help to determine whether the Government will be able to build on this pilot to establish a more comprehensive public works program. - 49 - Box 6.2. Liber ia Cash-for -Wor k Tempor ar y Employment Pr ogr am 1. Introduction The Liberia Agency for Community Empowerment (LACE) is a not-for-profit autonomous agency of the Government of Liberia, established in 2004 with support from the World Bank. Its primary goals are to implement national initiatives and strategies aimed at alleviating poverty, and to assist Liberia in achieving Millennium Development Goals through community-driven development (CDD). LACE is currently executing a Cash-for-Works Temporary Employment Project (CfWTEP), financed by a US$3 million grant from the World Bank. 2. Objective The objective of CfWTEP is to finance activities that help mitigate the short- and medium-to-long-term impact of rising food prices precipitated by the global food price crisis, as well as to support the Government's national strategy to improve food security in Liberia. The project uses a demand-driven works approach to provide the most vulnerable with seasonal employment. It aims to provide temporary jobs to 17,000 persons for 40 days at US$3 per day. 3. Targeting LACE works with communities to inform potential beneficiaries of CfWTEP activities and to identify activities that would be most beneficial. Participants are recruited on a first-come, first-served basis, as stipulated in the project's Operations Manual. The wage of US$3 per day was selected to allow for effective self-targeting. At least 30 percent of participants must be female. Youths below 18 years of age are excluded from the recruitment process to help ensure that they remain in school. Pregnant women are also excluded from the program. 4. Impact The impact of the CfWTEP intervention has been twofold. First, through the transfer of more than US$1.3 million to the project's participants, it has improved the lives of these vulnerable persons. Second, the work completed under CfWTEP has benefited the target communities. In Lofa County, project participants cleared an area for a community airfield. In Bong County, the clearing of lowland areas allowed the community to plant rice. In Grand Bassa County, CfWTEP workers cleared roadways and cleaned the town hall, police station, and other public areas. 5. Findings and lessons learned To date, project implementation has produced several key findings and lessons learned: · Vulnerable beneficiaries are willing to work for US$3 per day, and would like the project to continue beyond this first phase. · Payment through the banking system (Ecobank) has been very smooth and effective. Beneficiaries are happy with the payment process, through which they receive cash payments from the local Ecobank branch, and proud of identification card that they use to receive their payments. For many of the 11,000 beneficiaries, this project has provided them with their first ID card and access to the banking system for the first time. · LACE has observed that transferring used project tools to established community organizations has a greater ongoing impact on the community than when tools are transferred to individual beneficiaries through a lottery. 6. Sustainability Local authorities and relevant stakeholders will work with community organizations, CfWTEP participants, and the community as a whole to help ensure the sustainability of CfWTEP activities. For example, to maintain feeder/farm access roads, community organizations will periodically mobilize community members to carry out maintenance activities, including cleaning roads, repairing drainage systems, and filling potholes. Former project participants will play a key role in these efforts. Source: Bank staff. - 50 - The cost and impact of public works in Liberia 6.29. The cost of a labor-intensive public works program in Liberia will depend on the number of beneficiaries and the wages paid to program participants. An analysis of the 2007 CWIQ data suggests that only a small share (less than 10 percent) of the population earns more than the minimum wage of US$2/day, or US$480/year. While this suggests that the targeting of public works programs would be better if wages were set lower than the minimum wage, this may not be socially or legally defensible. In our simulations, therefore, we consider two scenarios and three wage levels, annualized to US$240, US$480, and US$720 per year (Table 6.2). We assume that public works participants will benefit from the program for six months per year. Under the first scenario, the program would reach 25,000 beneficiaries, and its cost would range from US$3.75 million (0.4 percent of GDP) at wages of US$240 per year (about half the minimum wage), to US$11.25 million (1.2 percent of GDP) at wages of US$720 per year (about 150 percent of the minimum wage). Under the second scenario, the program would reach 50,000 beneficiaries at a cost ranging from US$7.5 million (0.8 percent of GDP) to US$22.5 million (2.5 percent of GDP). These costs include wage and administrative costs, but do not cover the other costs of public works such as construction materials. Table 6.2. Estimates of Pr oject Cost (Wages and Administr ative Costs) Parameters Scenario I Scenario II Beneficiaries 25,000 25,000 25,000 50,000 50,000 50,000 Minimum wage (US$) --------------480.00------------------- Paid wage (US$) 240 480 720 240 480 720 % minimum wage 50% 100% 150% 50% 100% 150% Employment duration (yrs) 0.50 0.50 0.50 0.50 0.50 0.50 Wage cost (US$m) 3.0 6.0 9.0 6.0 12.0 18.0 Adm. cost @25% (US$m) 0.75 1.5 2.25 1.5 3.0 4.5 Total cost (US$m) 3.75 7.5 11.25 7.5 15.0 22.5 Cost/GDP (%) 0.4% 0.8% 1.2% 0.8% 1.7% 2.5% GDP (US$m) 904.35 904.35 904.35 904.35 904.35 904.35 Source: World Bank staff, based on 2007 CWIQ survey. 6.30. To assess the potential impact of labor-intensive public works on poverty in Liberia, we rely on simulation techniques using data from the 2007 CWIQ. Following Coulombe et al (2008), the approach begins with an assessment of who may be potentially interested in participating in the program. This is done by identifying individuals working without pay, as well as--for every level of the proposed wage in the program--those individuals who work but now earn less than the program wage, since all these individuals may be interested in participating in the program to increase their earnings. The unemployed whose reservation wage is likely to be below the proposed program wage are also considered as potential beneficiaries. Next, we randomly select a number of participants from the pool of potential beneficiaries. Finally, we estimate a leakage rate, 20 which represents the share of program outlays that do 20 The leakage rate is computed as the product of one minus the poverty or extreme poverty status for the household to which the public works participant belongs (since when an individual is not poor or extreme poor, the public works wage does not help in reducing poverty or extreme poverty) times the additional wage for the public works - 51 - not directly contribute to poverty reduction. This leakage rate depends on two key parameters: (a) the targeting performance of the program; and (b) its substitution effect, since only part of the wages paid to beneficiaries generates additional income (they would probably have done other work if they had not participated in the program). Our simulations for the impact of public works on poverty are based on the assumption of a 50 percent substitution effect; i.e., program participants give up half their current earnings to participate in the public works program, which is assumed to take place in the lean season. 6.31. We focus here on the key results. Figure 6.1 summarizes the data on the potential number of participants by consumption quintiles of the households to which potential beneficiaries belong. This is done separately for urban and rural areas as well as for the three potential wage levels. Two findings stand out. First, the number of individuals who could potentially be interested in the program appears to be very large, especially because many are working with low pay (observed or imputed) and might therefore be interested in getting higher cash income through public works. Second, the targeting performance or likely benefit incidence of the program depends on whether it is implemented mostly in urban or rural areas. In urban areas, the program would probably be regressive, since most of the potential beneficiaries belong to the better-off quintiles (because urban households tend to have higher levels of consumption than rural households, so relatively few households in urban areas belong to the bottom quintiles). By contrast, the program could be well targeted to individuals belonging to poor households if the focus is placed on providing employment and reconstructing infrastructure in rural areas. There is also a clear relationship between the wage level of workers and the poverty status of households. participant divided by the reference wage of the program (since this is the proportion of the wage outlays that contributes to the reduction of poverty or extreme poverty). - 52 - Figur e 6.1. Distr ibution of Potential Beneficiar ies of Public Wor ks 80 (2007) 100 Number of Beneficiaries, in '000 Number of Beneficiaries, in '000 80 60 60 40 40 20 20 0 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 $240 $480 $720 $240 $480 $720 Urban areas Rural areas Source: World Bank staff, based on 2007 CWIQ survey. 6.32. In order to assess the potential impact of public works on poverty, we start by discussing targeting performance. The leakage rate for the program is the share of program outlays that is not reaching the poor or the extreme poor If the public works program is implemented with an annual wage of US$720, the leakage rate for poverty in Monrovia is estimated at 58.0 percent, and for extreme poverty at 80.2 percent (Table 6.3). These leakage rates are relatively high, given that the share of program participants who are poor or extremely poor is lower in the capital than elsewhere. Overall, however, the variation in leakage rates across the various wage levels is not very high. This is because a higher wage level implies less targeting to the poor, but on the other hand it reduces the substitution effect through which part of the gains from the public works wage are lost due to the need to give up other work. In terms of results for the country as a whole, at a wage rate of US$240 per year, the overall leakage rate is 52.4 percent, and it remains between 45.4 percent and 52.4 percent when we change the wage rate. The leakage rates are lowest in the South Eastern area of the country, where poverty and extreme poverty are higher. Table 6.3 Leakage Rate for Public Wor ks by Wage Rate and Region, 2007 (% ) Wage rate US$240 US$480 US$720 Poverty Extreme Poverty Extreme Poverty Extreme leakage poverty leakage poverty leakage poverty Region rate leakage rate rate leakage rate rate leakage rate Greater Monrovia 60.8 81.1 57.9 80.1 58.0 80.2 North Central 51.5 59.9 45.5 55.5 42.3 53.1 North Western 53.1 62.2 44.4 55.4 39.4 51.6 South Central 52.0 65.7 49.4 64.5 48.4 63.9 South Eastern A 45.5 58.0 41.2 53.2 37.2 50.5 South Eastern B 52.8 63.8 46.1 58.5 42.9 56.2 Total 52.4 63.7 47.8 61.2 45.4 59.6 Source: World Bank staff, based on CWIQ survey. - 53 - 6.33. The estimated potential impact of the program on poverty is given in Table 6.4. With the provision of 50,000 jobs and assuming an annual public works wage of US$720 (each workers then gets US$360 over a six-month period), the headcount index of poverty is reduced by 15.6 percentage points among program beneficiaries. The reduction for the population as a whole is 1.61 percentage point. The impact on extreme poverty is similar. While a reduction of less than two percentage points may not appear to be very large compared to the existing share of the population in poverty (64 percent), it is not negligible. A large share of the population would still benefit from improved standards of living through the public works program. The impact is about half if only 25,000 jobs are created. Table 6.4. Potential Impact of Public Wor ks on the Reduction of Pover ty and Extr eme Pover ty (% ) Beneficiaries Whole population Wage rate (US$) Headcount Poverty gap Headcount Poverty gap Impact on poverty, 50,000 jobs 240 3.33 3.29 0.35 0.34 480 10.16 6.85 1.05 0.71 720 15.61 10.06 1.61 1.04 Impact on extreme poverty, 50,000 jobs 240 3.40 3.01 0.36 0.31 480 10.08 6.02 1.04 0.62 720 16.07 8.62 1.66 0.89 Source: World Bank staff, based on 2007 CWIQ survey. Note: The impact on poverty is expressed in percentage points. - 54 - 7. EDUCATION AND SKILLS MUST ALSO BE IMPROVED 7.1. In addition to the challenges of creating employment posed by the global economy, the Government also faces significant challenges on the supply side of the labor market. The de-skilling of the labor force during the 14 years of civil war has left the economy in a low productivity trap, unable to compete in global markets except in the natural resources-based sectors, which are largely controlled by concessionaires. The skills gap exists across all levels and sectors. For example, an external contractor on a large road project, when asked about the ease of finding skilled workers, said he had difficulty finding skilled carpenters and masons. Unless the skills shortage is addressed, Liberia may find itself with a significant number of unemployable workers once the global economy rebounds. To achieve its pro- poor growth policy objectives, the Government will need to take a two-approach. First, it needs to rapidly develop technical and vocation training programs to improve the employability of Liberian workers, and ensure that such training is better aligned to market demand. This will not only increase the incentives for the private sector to share the cost of such training, but also improve the social returns on such investments. And second, the Government will need to provide at least three years of short-term temporary employment to bridge the employment gap until the private sector rebounds. 7.2. Liberia's education system has been shaped by its history. Until the Second World War, schools were run primarily by religious missions and other charitable organizations. Liberia's school system was highly underdeveloped compared to that of its neighbors. Beginning in the 1950s, public resources were increasingly devoted to education, and by the 1960s the majority of schools were publicly owned, although access to schooling in rural areas remained limited. Total enrollment in primary and secondary school was approximately 60,000 students in 1960, and the gross primary enrollment rate was estimated at 31 percent. Since the 1960s, a number of educational institutions have been providing vocational training, aimed at preparing Liberians over the age of 15 for productive employment. 7.3. By the mid-1970s, access to education had increased significantly. The primary gross enrollment rate was greater than 50 percent in 1975, and secondary gross enrollment was at 12 percent. Literacy rates among youth (age 15-19) had improved from 16 to 47 percent between 1962 and 1970. Education indicators continued to improve through the 1970s, with primary gross enrollment rates increasing to 67 percent by the late 1970s (Table 7.1). Table 7.1. School Enr ollment for Selected Pr e-War Year s School enrollment (%) 1960 1970 1979/80 Primary enrollment (gross) 31 53 67 Secondary enrollment (gross) 2 9 22 Adult literacy 9 15 25 Sources: 1960, 1970 and 1979/1980 data: World Bank, Liberia: recent economic developments and medium-term prospects, 1982. 2007 data: World Bank, Education in Liberia: Basic diagnostics using the 2007 CWIQ survey. 7.4. The damage to the educational system during the war has resulted in a generation of youth who grew up with little or no access to formal education. According to the 2008 Population and Housing Census, more than half of the population aged six and older has no formal education, less than - 55 - 25 percent have primary education, and less than 25 percent have an education above the primary level. This reflects, in part, Liberia's severe "brain drain" as a result of the conflict, when many skilled and professional people emigrated. Using medical doctors as a proxy for the professional class highlights the considerable brain drain which began in the 1980s. The International Crisis Group (ICG) reported that in the 1980s there were 400 doctors in the country, but by 2002 there were only about 30. 21 According to the 2007 CWIQ, about 45 percent of the working population has never been to primary school, and fewer than 15 percent has completed secondary education. Table 7.2. Net and Gr oss Enr ollment Rates in Pr imar y and Secondar y Schools, 2007 Residence Area Quintile Urban Rural Q1 Q2 Q3 Q4 Q5 Total Primary enrollment rates Net enrollment (6-11) Total 47.5 32.8 28.8 32.1 36.1 41.8 49.5 37.3 Male 48.0 33.2 32.8 33.9 33.3 40.0 50.9 37.5 Female 47.1 32.3 23.8 30.0 39.1 44.0 48.3 37.1 Gross enrollment Total 93.1 83.3 77.8 86.5 87.5 86.6 94.3 86.3 Male 88.7 87.7 86.9 93.8 80.9 81.4 99.3 88.0 Female 97.4 78.2 66.7 78.0 94.4 92.7 90.1 84.5 Secondary enrollment rates Net enrollment (12-17) Total 25.4 10.1 11.4 12.2 13.0 19.8 21.5 15.2 Male 27.7 11.2 12.6 13.3 15.4 19.1 22.8 16.0 Female 23.4 8.7 9.6 10.7 10.8 20.4 20.3 14.2 Gross enrollment Total 74.4 39.7 40.1 43.0 42.8 65.3 71.8 51.3 Male 86.8 44.9 43.4 51.0 54.5 72.5 75.6 57.2 Female 63.7 33.1 35.1 32.6 31.7 59.2 68.0 44.7 Source: World Bank staff, based on 2007 CWIQ survey. 7.5. Liberia's education system is recovering, however, and literacy among the younger population is rising. The net primary school enrollment rate in 2007 was 37.3 percent, compared with 34.7 percent in 2001 (Table 7.2). Moreover, the much higher gross enrollment rate (86.3 percent) suggests that older children who missed out on education during the war are returning to school. Data from the 2008 Population and Housing Census shows a significantly higher literacy rate among the younger group of the working-age population. Among the 15-29 age group, 73 percent are literate, compared to only 25.8 percent among the 60-64 age group. Figure 7.1 below shows the distribution of the working age population by literacy status. The dynamic of improving literacy is a positive for the labor market going forward. In addition, an estimated 20,000 students are currently enrolled in tertiary education, which means that the number of university-trained workers could double within the next five years. However, continued lack of access to education and training remains acute in the rural areas. 21 Liberia: The Key to Ending Regional Instability, ICG Africa Report No. 43, 24 April 2002. - 56 - Figur e 7.1. Liber ia Distr ibution of Labor For ce by Liter acy Status 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 % Share of Labor Force % Literate Source: World Bank staff, based on 2008 Population and Housing Census. 7.6. The gender bias in education is manifest, but the gap is closing rapidly. Women have lower levels of educational attainment than men across all age groups. Data from the 2008 Population and Housing census shows that for the working age population (15-64), the literacy rate for males is more than one and a half the rate for females, although the gap is closing in the younger generation. As Table 7.3 shows, while the ratio of literate males to females in the 60-64 age group is more than four to one, it falls dramatically to almost parity in the 15-19 age group. In rural areas, however, the illiteracy rate is as high as 74.1 percent for women compared with 40.1 percent for men. Although the gender gap is closing, the gender difference in gross enrollment, particularly in the secondary schools--44.7 percent enrollment for girls compared to 57.2 percent for boys in 2007--suggests the importance of gender equity in education for poverty reduction. In fact, the much lower rate of female participation in the economy, and women's disproportionate representation in unpaid and informal employment, may in part be explained by their lower educational attainment. An assessment by the United Nations showed that women held only 2 percent of the formal jobs in Liberia in 2006. 22 The situation may have improved somewhat, as data from the 2008 Population and Housing Census suggest that women now account for 21.5 percent of paid employment. 7.7. Demand for secondary education is sharply increasing. The recent progress in enrollment and completion rates for six-year primary education will generate demand for post-primary (secondary and tertiary) education. Even a modest increase (from 71 to 75 percent) in first year admissions to junior high schools and a somewhat more significant increase in tertiary admissions, from 26 to 39 percent , will require a five-fold increase in the recurrent education budget. Moreover, places and (consequently) enrollments are unevenly distributed across the country, and access is a major challenge in rural areas. This means that increases in access will require policy changes as well as significant investments in 22 United Nations (UN), 2006, United Nations Country Common Assessment for Liberia: Consolidating Peace and National Recovery for Sustainable Development. Monrovia. - 57 - school construction and in reforming the curriculum to accommodate a more diverse cohort of young people. These investments are essential if Liberia is to rely on more literate, competent, and better-skilled future generations in the medium to long term. Table 7.3. Distr ibution of the Labor For ce by Gender and Liter acy Status Age Male Female Literate group males to females # Literate Total % Literate # Literate Total % Literate Ratio 15-19 145,300 189,407 76.7 129,062 186,288 69.3 1.11 20-24 120,802 161,951 74.6 102,179 180,979 56.5 1.32 25-29 95,543 141,006 67.8 67,628 150852 44.8 1.51 30-34 69,936 107,326 65.2 43,911 112,306 39.1 1.67 35-39 64,062 99,136 64.6 35,547 104,400 34.0 1.90 40-44 53,273 81,670 65.2 22,218 74,065 30.0 2.17 45-49 39,107 63,827 61.3 13,489 54,980 24.5 2.50 50-54 26,589 44,870 59.3 7,350 38,070 19.3 3.07 55-59 16,737 30,975 54.0 3,695 25,485 14.5 3.73 60-64 10,901 25,473 42.8 2,730 27,357 10.0 4.29 Total 642,250 945,641 67.9 427,809 954,782 44.8 1.52 Source: World Bank staff, based 2008 Population and Housing Census. Technical and Vocational Education and Training 7.8. The institutional structure of TVET in Liberia is highly fragmented in terms of providers, governmental oversight, and funding. A World Bank Scoping Mission 23 in March 2009 found that TVET providers fall into two groups: one group comprises 10 to 15 public institutions with a potential capacity of 500 trainees; and the other comprises 100 to 150 non-public institutions with an average potential capacity of 20 to 40 trainees. 7.9. A study of vocational training institutes conducted in 2006 24 found that approximately 15 percent were government run, while the remaining 85 percent were managed by private individuals, religious missions, and NGOs. Among more than 500 teachers surveyed, almost 80 percent were untrained and held only trade certificates (no degrees). The institutions had a total enrollment of more than 18,000 for all subjects (Table 7.4). However, the report did not distinguish between short and long-term training opportunities. 23 The objective of the mission was to assess demand and supply factors in TVET ongoing operations. 24 UNESCO, Situational Analysis of the Technical, Vocational Education and Training System in Liberia" December, 2006. - 58 - Table 7.4. TVET Enr ollment by Cour se Subject Enrollment Agriculture 1,546 Carpentry and construction 1,007 Masonry 689 Plumbing 245 Electricity 443 Electronics 1,279 Auto Mech. 528 Computers 4,852 Secretariat science 904 Typing 1,177 Cosmetology. 559 Home arts 711 Pastry 391 Soap marking 596 Tailoring/sewing 2,042 Tie & dye 734 Other 329 Total 18,032 Source: World Bank staff, based on assessments of 113 TVET institutions (2006). 7.10. Governance and management of TVET are inadequate at both the institutional and central government levels. A few institutions have full budget control and show some relatively strong strategic and management capabilities, as well as some accountability. Most others have neither the capacity nor the mandate to define and implement their institutional strategies or manage resources. The small private providers are set up to respond to the mostly short-term demand for quick programs, without linkages to any type of qualification or articulation. Most of them are private, NGO, or church-owned operations without certification or accreditation. Governance at the central level is fragmented across several ministries (Education, Youth, Labor, Planning, Agriculture, and Finance), and intergovernmental coordination is lacking. A National Council for Technical and Vocational Education and Training is largely dormant. There are no accreditation or certification systems for the institutions and programs, and no qualification system for the trainees. 7.11. Outcomes, skills, and results of training are not measured in most cases, and there are generally no adequate mechanisms to align training services with economic demand. Both providers and employers acknowledge the mismatch between the limited scale of TVET programs (and thus the skills that the job seekers have) and short-term employment opportunities. It appears that the growth of certain industries is constrained by the lack of skilled manpower, while the expansion of training programs remains limited by the lack of jobs. 7.12. In the absence of quality assurance systems and a mature labor market that could provide reliable information about the outcomes of TVET, quality can best be assessed based on inputs (such as qualification of trainers and quality of equipment) and processes (such as curricula). The large majority of teachers are inadequately trained and equipment quality is uneven. Curriculum - 59 - development is almost non-existent, with the exception of some institutional initiatives to respond to a particular employer or economic demand. 7.13. The unit costs of most TVET programs are high, and externally financed activities are not sustainable. According to Ministry of Education (MoE) estimates, the cost per person of TVET is around US$750. World Bank estimates put the cost somewhat lower, at around US$650, mostly because of the lower unit costs at the Ministry of Youth and Sports (MoYS). Under the MoE, unit costs are about US$1000, while at the MoYS they are roughly US$400. TVET programs within the school system (providing three years of training) cost around US$1000 and others of significantly lower quality and shorter duration cost about a third to a fourth that amount. Using the unit cost of US$750 as a benchmark, TVET costs are more than 13 times the costs of senior high school unit costs, 35 times the costs of primary education, and 3 times the costs of tertiary education. The high costs of the programs appear to be linked to high operating and reconstruction costs. The TVET review (UNESCO, 2006) found that fuel, rent, utilities, and administrative staff costs accounted for almost 40 percent of expenditure, with building reconstruction and other capital investment costs accounting for an additional 30 percent. Some of the costs of the programs are borne by the students themselves, in the form of course fees. Most of the TVET institutions are in need of physical rehabilitation, new equipment, and training of teachers. 7.14. Donors have spent about US$40 million on various TVET activities over the last three years, targeting at-risk groups such as ex-combatants and internally displaced persons (IDP). The entire TVET budget for the MoE and the MoYS combined is about US$4 million a year (of which US$1.3 million is for Liberia's Booker Washington Institute alone). For IDP operations, unit costs often reach US$2000 to US$3000. Yet, owing to fragmentation and high overheads, not many of these operations appear to be effective or sustainable. Most donor-financed TVET activities target demobilized combatants, which contributes to maintaining peace and stability and limiting the risk of mass violence. Yet despite these significant government and donor inputs, there are no mechanisms in place to measure the sector's performance, in terms of its contribution to either employment or productivity. Rehabilitation of the sector will depend on investments being accompanied by institutional and governance reforms. With the present political momentum, there are opportunities for significant structural changes, as well as options for programs that bring rapid and sustainable results. 7.15. A focused effort on education and skills training is needed to prepare workers for the transformation of the economy from a primarily natural resource-based, labor-intensive economy to one that is natural resource based and skill intensive. According to a 2006 UNESCO report, "Poor access to and quality of education present a major impediment to economic development, livelihoods, prosperity, and ultimately peace and security." 25 Although temporary employment and increased agricultural productivity are partial solutions, but Liberia must begin to set the stage for its transition to a skill-intensive economy. Infrastructure modernization has already begun, with support from the World Bank and other donors, and it needs to continue, particularly for energy infrastructure and ports. At the same time, a focused effort on education and skills training is needed to incentivize investors in skills- intensive manufacturing to consider Liberia as a possible investment location. A vibrant agriculture sector could also provide the basis for agro-industry. 25 UNESCO, Situational Analysis of the Technical, Vocational Education and Training System in Liberia" December, 2006. - 60 - 7.16. These programs need to address three general objectives. They need to (a) provide rapid response to emerging employment issues; (b) while bringing skills services closer to industry needs; and (c) addressing these issues in ways that are sustainable and mainstreamed within public and private education. The programs also need to be cost-effective, supported by reliable public and private sector revenue sources, and aligned with coherent qualification and certification frameworks. Some innovative programs have already been developed and could be scaled up or replicated. These include the consistently high-quality training provided by the Booker Washington Institute, as well as donor programs such as the Liberia Community Infrastructure Project, which combines non-formal TVET education with business development support. Finally, large industrial concessions, such as Arcelor Mittal, combine large-scale skills development with employment generation. The common characteristics of these public, NGO-based, and industry-based programs are that they are innovative and results focused, and they exemplify a close, productive working relationship between economic actors and training providers. Most of them also have high unit costs (US$1,200 to US$2,400) which can only be sustained through public-private partnerships. - 61 - Annex 1. Employment Gr owth Note on methodology for determining employment growth, 2009-2013 1. The projections used 2008 as the base year with a population of 3.5 million. The 2007 CWIQ was used to determine the labor force participation rate and the percentage of the labor force engaged in each sector. These shares were then applied to the 2008 census population estimate. This was done because the 2007 CWIQ was based on an old sampling framework that yielded a population size well below the subsequent 2008 preliminary census results. 2. The labor force was divided into "formal" and "informal" employment. The formal category comprised "paid employees," while the informal category comprised the self-employed, unpaid family workers, domestic workers, and apprentices. Sectoral breakdowns were then applied to the formal and informal sectors. These were determined based on two questions from the CWIQ survey: question E9, "main activity," and question E8, "employer in main job." Employment growth in each sector was determined separately: 3. Agriculture: Employment growth in the formal agriculture sector was calculated based on expected sectoral growth rates and employment elasticity of sectoral growth (Annex Table 1). Employment growth in the forestry sector was based on the estimated number of new jobs created by new concessions coming on board and by the expansion of existing concessions. 4. Mining: Employment growth in the formal mining sector was based on the estimated number of jobs created by the two new iron ore concessions (Bong and Western Cluster), according to consultations with Ministry of Mines representatives. It was assumed that employment would take place at the front end, as the companies begin developing the mines, and be phased in over a two-year period. 5. Other industry: Employment in manufacturing and the electricity, gas, and water sectors was calculated based on expected sectoral growth rates and employment elasticity of sectoral growth (Annex Table 1). 6. Government services: Employment in government, calculated to be 62,000 based on the CWIQ, is much higher than the actual number of civil servants on payroll (35,000). This is due to large numbers of volunteer teachers, health workers, and office workers who consider themselves government employees despite lacking formal contracts. Government employment is assumed to remain constant for the purpose of this exercise. However, government employment may in fact decrease over the projection period due to the rationalization of public sector employment, as proposed in the civil service reform strategy 26. 7. Services: Employment in other service sectors was calculated based on expected sectoral growth rates and employment elasticity of sectoral growth (Annex Table 1). Employment by NGOs and international organizations, included under "other services," is assumed to remain constant over the 26 Smaller Government Better Service" Civil Service Reform Strategy (2008 ­ 2011), June 2008 - 62 - projection period due to declining humanitarian aid channeled through NGOs and other agencies. This number is subtracted from the "other services" total before applying the employment growth rate. 8. Informal sector: Employment in the informal sector, which can be regarded as a "buffer category" that absorbs surplus labor, is projected to grow at the rate of population growth, and thus stay constant as a proportion of the total population and labor force. Annex Table 1. Sector al Gr owth Rates and Employment Elasticities with Respect to Gr owth GROWTH RATES (%) ELASTICITIES 2009 2010 2011 2012 2013 Elasticity Agriculture 5.0 8.3 7.7 5.0 5.0 Agriculture 0.82 Forestry 9.9 5.0 6.7 6.3 12.7 Industry 0.9 Mining -7.4 84.9 150.1 134.8 61.9 Services 0.79 Manufacturing -7.3 -2.5 0.4 5.0 5.0 Services 7.3 7.5 7.3 6.5 5.0 Sources: IMF, June 2009; ILO, KILM 5th Edition. - 63 - Annex 2. Commodity Expor t Pr ices Cor r elation Matr ix of Commodity Expor t Pr ices (1990-2008) Commodity Cocoa Coffee Palm oil Logs Rubber Gold Iron ore Cocoa 1.00 Coffee 0.21 1.00 Palm oil 0.72 0.55 1.00 Logs 0.37 0.04 0.44 1.00 Rubber 0.56 0.27 0.64 0.58 1.00 Gold 0.61 0.10 0.60 0.77 0.92 1.00 Iron ore 0.61 0.10 0.58 0.79 0.84 0.95 1.00 Source: Staff calculations, based on World Bank commodity price database. - 64 - REFERENCES Adato, Michelle and Lawrence Haddad. 2001. Targeting Poverty through Community-Based Public Works Programs: A Cross-Disciplinary Assessment of Recent Experience in South Africa. International Food Policy Research Institute, Discussion Paper 121. Backiny-Yetna, P., Q. Wodon, R. Mungai, and C. Tsimpo, 2008. Poverty in Liberia: Level, Profile and Determinants, in Q. Wodon, Editor, Poverty and Policy in Liberia: Results from the 2007 CWIQ Survey. World Bank Working Paper, Washington DC. Bautista, Romeo M., Sherman Robinson, and El-Said Moataz. 1999. Alternative Industrial Development Paths for Indonesia: SAM and CGE Analyses. International Food Policy Research Institute (May). Besley, Timothy and Louise J. Cord. 2007. Delivering on the Promise of Pro-Poor Growth: Insights and Lessons from Country Experiences. World Bank. Beasley, Kenneth W. 2006. Job Creation in Post-conflict Societies. USAID. Betcherman, Gordon, Martin Godfry, Susana Puerto, Friederike Rother, Antoneta Stavreska. 2007. A Review of Interventions to Support Young Workers: Findings of the Youth Employment Inventory. Social Protection Discussion Paper No. 0715. World Bank. Blackden, C. M., and Q. Wodon, eds. 2006. Gender, Time Use and Poverty in Sub-Saharan Africa, World Bank Working Paper No. 73, Washington DC. Byerlee, D., X. Diao, and C. Jackson. 2005. "Agriculture, Rural Development, and Pro-poor Growth: Country Experiences in the Post-Reform Era." Agriculture and Rural Development Discussion Paper 21. World Bank. Coady, D. 2004. Designing and Evaluating Social Safety Nets: Theory, Evidence, and Policy Conclusions. Food Consumption and Nutrition Division, International Food Policy Research Institute, Discussion Paper 172. Collier, Paul. 2002. Primary Commodity Dependence and Africa's Future. World Bank, Washington DC. Coolidge, Jackie, Richard Stern, and Jan Loeprick. 2007. Business Taxation: Is This the Key to Formalization? Evidence from Africa and Possible Solutions, FIAS. Coulombe, H., M. Temourov, and Q. Wodon, 2008, Ghana's National Youth Employment Program and Poverty Reduction, in World Bank, Ghana: Job Creation and Skills Development, Report No. 40328­ GH, Volume 2, Washington DC. - 65 - Dejardin, Amelita King. 1996. Public Works Programs, a Strategy for Poverty Alleviation: The Gender Dimension. Issues in Development, Discussion Paper 10. del Ninno, Carlo, Subbarao Kalanidhi, and Annamaria Milazzo. 2009. How to Make Public Works Work: A Review of the Experiences. World Bank (May). Devereux, Stephen and Colette Solomon. 2006. Employment Creation Programs: The International Experience. International Labor Organization, Geneva. Glick, P. and D. Sahn. 1998. Maternal Labour and Child Nutrition in West Africa. Oxford Bulletin of Economics and Statistics 60(3):325-355. Fasano, Ugo. 2000. Review of the Experience with Oil Stabilization and Savings Funds in Selected Countries. IMF Working Paper WP/00/112 ( June). Foreign Investment Advisory Service, International Finance Corporation, and World Bank. 2006. Liberia: Mini-Diagnostic Analysis of the Investment Climate (March). Hicks, N. and Q. Wodon, 2001, Social Protection for the Poor in Latin America, CEPAL Review, 73: 93- 113. Inter-American Development Bank. 2004. Good Jobs Wanted: Labor Markets in Latin America. International Crisis Group. 2002. Liberia: The Key to Ending Regional Instability. ICG Africa Report No. 43 (April). International Labour Organization. 2003. A Global Program: Investing in Employment for Poverty Reduction and Local Economic Growth. International Labour Organization. Key Indicators of the Labor Market--KILM 5th Edition. International Labour Organization. 1999. Investing in Jobs for Development: The ILO's Employment ­ Intensive Programme (EIP). Geneva. International Monetary Fund. 2008. Liberia: Fiscal Terms for Large-Scale Agriculture and Forestry Concessions (December). International Monetary Fund. 2009. Liberia Country Report No. 09/177 (June). Lederman, Daniel and William F. Maloney. 2007. Natural Resources Neither Curse nor Destiny. World Bank. Lennartsson, Maria and David Stiedl. 1995. Technology Choice, Man or Machine: Including Case Studies from Lesotho and Zimbabwe. International Labour Organization. - 66 - Lin, Justin Yifu. 1992. "Rural Reforms and Agricultural Growth in China." American Economic Review 82(1):34­51. Miller, Stephen K. 1992. Remuneration Systems for Labor-Intensive Investments: Lessons for Equity and Growth. International Labor Review, Vol. 131, 192. No.1. Munters, Paul. 2003. "Jobs or Machines: Comparative Analysis of Rural Road Work in Cambodia." International Labor Organization. Organisation for Economic Co-operation. 2006. Promoting Pro-Poor Growth in Agriculture. Quisumbing, Agnes R. and Yisehac Yohannes. 2004. How Fair Is Workfare? Gender, Public Works, and Employment in Rural Ethiopia. International Food Policy Research Institute. Rao, C.H.H., K. Subbarao, and S. K. Ray. 1988, Unstable Agriculture and Droughts: Implications for Policy. Vikas: Delhi. Ravallion, Martin and Shaohua Chen. 2007. "China's (Uneven) Progress Against Poverty," Journal of Development Economics 82: 1-42. Ravallion, Martin. 1988. Appraising Workfare Programs. Policy Research Working Paper No. 1955, World Bank. Razak Abd S. and M. N. Mohamad. Roff. 2006. Status and Potential of Urban and Peri-Urban Agriculture in Malaysia. Smaller Government Better Service. 2008. Civil Service Reform Strategy 2008 ­ 2011 (June). Subbarao, Kalanidhi. 2003. Systemic Shocks and Social Protection: Role and Effectiveness of Public Works Programs. World Bank. Subbarao, K. Public Works As An Anti-Poverty Program: An Overview of Cross-Country Experience. 1997 Subbarao, K., Jeanine Braithwaite, and Jyotsna Jalan. 1995. Protecting the Poor During Adjustment and Transitions. Working Paper 21229. World Bank, May 1995. Sum, June-Wei and June Hackett. 2007. Sierra Leone: The Road to Recovery ­Results from a Beneficiary Impact Assessment. World Bank. Swamy, G. 2001. Gender Dimensions of Public Employment Schemes, World Bank, Washington D.C. Teklu, Tesfaye. 1995. Labor­Intensive Public Works: The Experience of Botswana and Tanzania. - 67 - Thurlow, James, Jane Kiringai, and Gautam. 2007. Rural Investments to Accelerate Growth and Poverty Reduction in Kenya. International Food Policy Research Institute. Timmer, Peter C. 2003. Agriculture and Pro-Poor Growth: Pro-Poor Economic Growth Research Studies. Tsimpo, Clarence and Questin Wodon. 2008. Rice Prices and Poverty in Liberia. Policy Research Working Paper Series 4742, World Bank. Tsubota, Kunio. 2007. Urban Agriculture in Asia: Lessons from Japanese Experience. UNESCO. 2006. Situational Analysis of the Technical, Vocational Education and Training System in Liberia (December). United Nations. 2008. The Impact of High Prices on Food Security in Liberia (July). United Nations. 2006. United Nations Country Common Assessment for Liberia: Consolidating Peace and National Recovery for Sustainable Development. Monrovia. United Nations Mission in Liberia. 2006. Government of Liberia Rubber Plantation Task Force Report (May). von Braun, Joachim, Ashok Gulati, and Shenggen Fan. 2005. Agricultural and Economic Development Strategies and the Transformation of China and India. International Food Policy Research Institute. von Braun, J., T. Teklu,and P. Webb, P. 1991. Labor-Intensive Public Works for Food Security: Experience in Africa. Working Papers on Food Subsidies No.6. International Food Policy Research Institute. Webb, Patrick and Shubh Kumar. 1995. Food and Cash for Work in Ethiopia: Experience during Famine and Macroeconomic Reform. International Food Policy Research Institute. Thwala Wellington Didibhuku. Employment Creation through the Use of Labor-Intensive Methods in South Africa: Experience and Problems. University of Johannesburg. World Bank. 1975. Liberia Growth with Development: A Basic Economic Report. Report No. 425a-LBR (March). World Bank. 1982. Liberia: Recent Economic Developments and Medium-Term Prospects. Report No. 4178-LBR (December). World Bank. 1986. The Study of the Substitution of Labor and Equipment in Civil Construction. World Bank. 2007. African Development Indicators. World Bank. 2007. World Development Indicators. - 68 - World Bank. 2008. World Development Report. World Bank. 2008. Liberia: Tapping Nature's Bounty for the Benefit of All. Diagnostic Trade Integration Study. - 69 -