Report No. 40328-GH Ghana Job Creation and Skills Development (In Two Volumes) Volume I: Main Report May 29, 2009 Human Development, AFTH2 Country Department AFCW1 Africa Region Document of the World Bank Acknowledgements This report was prepared by a team consisting of Setareh Razmara (AFTH2, task team leader), Ami1 Van Adams (Consultant), Quentin Wodon (AFTPM), Pieter Serneels (PREM), Harold Coulombe (Consultant), and Moukim Temourov (AFTH2). Moses Anwonoor, Alonso Sanchez and Kalpana Mehra (Consultants) helped with early data analysis. Montsemat Pallares-Miralles (HDNSP) also provided valuable inputs on pension reforms in Ghana. The report draws from two background papers that provided substantial inputs: (i)The Role of Employment and Earnings for Shared Growth: the Case of Ghana (Pieter Serneels, PREM); and (ii)Education, Skills and Labor Market Outcomes in Ghana: Planning reforms for skills development (Arvil V. Adams, Consultant). In addition, as part of this work, Francis Teal (Center for the Study o f African Economies, University o f Oxford), prepared two background papers on: (i)Apprenticeship in Ghana; and (ii)Formal and Informal Employment in Ghana: Job Creation and Skills. Significant suggestions and comments were provided by Ishac Diwan, Zeljko Bogetic, Pierella Paci, Harold Alderman, Peter Darvas, Katherine Bain, and Daniel Kwabena Boakye. Peer reviewers were Louise Fox, Gordon Betcherman and Zafiris Tzannatos. A major part of this report on the labor market outcomes and education and skills was incorporated in the background papers of the Country Economic Memorandum (CEM, report#40934-GH7November 2007). We would like to thank Mats Karlsson who was the Country Director at the time the study was launched. Kathryn Bach provided valuable editorial inputs. Administrative support was provided by Ruth Mulahi, Marietou Toure and Nicole Hamon (AFTH2). The preparation of this study also benefited from the cooperation of the PREM and Human DevelopmentBocial Protection Anchors based on the BNPP Project on Labor Market Indicators. The Volume 2 comprises seven background papers authored bythe following team members:Arvil Van Adams, Harold Coulombe and Quentin Wodon authors of "Education, Employment and Earnings in Ghana" (Annex 1); Courtney Monk, Justin Sandefur and Francis Teal authors of "Apprenticeship in Ghana" (Annex 2); Quentin Wodon and Harold Coulombe authors of: "Domestic or productive work? Changes in patterns of time use in Ghana" (Annex 4), "Estimating the Potential Cost for Firmsof Enforcingthe MinimumWage" (Annex 5) and, with contribution of Moukim Temourov, "Ghana's National Youth Employment Program and Poverty Reduction" (Annex 7); and Harold Coulombe, George Joseph, Caglar Ozden, Yoko Niimi, and Quentin Wodon authors of "International Migration from Ghana: Patterns, Brain Drain and Policy Implications" (Annex 6). Valuable support from the counterpart team in the Ghanaian ministries of Manpower, Youth and Employment (or new Ministries of Employment and Social Welfare and Youth and Sports), Education, Science and Sports (or new Ministry of Education and Sciences), Economic Development and Finance, and Ghana Statistical Services are gratefully acknowledged. The team also benefited from discussions with, the Development Partners (particularly DFID, JICA, ILO), Ghana Labor Unions, Employers Associations and representatives o f the private sector. To engage the new government, the findings of the report were presented in the context of the Country Management Unit's "development dialogue" series on competitiveness on April 15 and 17, 2009, and were endorsedby the authorities. GHANA Job Creationand Skills Development Table of Contents EXECUTIVE SUMMARY ............................................................................................................ 1 CHAPTER 1:INTRODUCTION ................................................................................................. 1 BACKGROUND .............................................................................................................................. 1 FRAMEWORK AND STRUCTUREOFTHE REPORT........................................................................... 2 CHAPTER2: THE CONTEXT FORJOB CREATION ........................................................... MACROECONOMIC ENVIRONMENT. WAGE GAINS AND POVERTYREDUCTION........................... 33 CHARACTERISTICSOF FIRMSAND INVESTMENTCLIMATE INGHANA...................................... 10 SUMMARYOF FINDINGS ............................................................................................................. 12 CHAPTER 3: LABORMARKET OUTCOMES ..................................................................... 15 DEMOGRAPHIC TRENDSANDDEPENDENCY RATIOS ................................................................. 16 LABOR OUTCOMESAMONG WORKERS AGED 25-64 ................................................................... 19 INCOMEFROMWORK INAGEGROUP25-64 .............................................................................. 29 HOURS WORKEDAND HOURLY EARNINGSINAGEGROUP25-64 ............................................. 32 SUMMARYOF FINDINGS ............................................................................................................. 34 CHAPTER 4: LABORMARKET POLICIES. INSTITUTIONS. AND PROGRAMS .......35 LABOR Labor Market Regulation....................................................................................................... 35 POLICIES ......................................................................................................................... 35 40 Non-wage labor costs............................................................................................................. Wage policies......................................................................................................................... 46 LABOR MARKET INSTITUTIONS 49 PROGRAMS......................................................................................................... .................................................................................................. EMPLOYMENT 51 ANOVERVIEWOFFINDINGSAND POLICYIMPLICATIONS......................................................... Labor regulations and institutions.......................................................................................... 55 55 Employment Programs........................................................................................................... 56 CHAPTER 5: EDUCATION,SKILLS, AND LABORMARKET OUTCOMES .................58 THELANDSCAPEFOR SKILLSDEVELOPMENT............................................................................. 59 Basic education ...................................................................................................................... 59 63 Technical and vocational education and training (TVET) ..................................................... Senior Secondary Schools...................................................................................................... 64 Formal and Traditional Apprenticeships ................................................................................ 67 ECONOMICRETURNSTO EDUCATION ................................................................ 70 Education and the Type of Job Held...................................................................................... AND TRAINING 72 Education and Earnings.......................................................................................................... 74 STRATEGIESFORREF0RM .......................................................................................................... 77 77 TVET reform: an internationalperspective ........................................................................... Government's Reform Strategy ............................................................................................. 78 An Overview of Findingsand Policy Implications................................................................ 82 REFERENCES ............................................................................................................................. 87 BOXES Box 1: Accounting for the increase inwelfare levels in Ghana ....................................................... 9 Box 2: Youth employment .................................................................................................................. 21 11 GHANA Job Creation and Skills Development Box 3: Time Use Patterns.................................................................................................................... 22 Box 4: Labor Law in Ghana................................................................................................................ 36 Box 5: Ghana hiringand firing procedures and costs..................................................................... 39 Box 6: Role of NationalLabor Commission................................................................................... -40 Box 7: PensionSystem inGhana........................................................................................................ 47 Box 8: NYEP financing sources......................................................................................................... 53 Box 9: Lessons from the Youth EmploymentInventory ............................................................... 54 Box 10:Migration,brain drain andlosses inpublic education spendmg ..................................... 62 Box 11:Summary of recommendations on education and s k d s developmentandnext steps.83 FIGURES Figure 1.a: Sectoral Reallocation ofLabor Force ............................................................................... 8 Figure 1.b:Growth Rate of Output and Labor Force by Sector (Yo) ............................................. 8 Figure2: Labor productivity is particularly low among small enterprises inGhana.................11 Figure3: Managers' responses on biggest investment climate constraints. MLEs and microenterprises .................................................................................................................................... 12 Figure 4: PopulationPyramids 2000, 2025, 2050............................................................................. 17 Figure 5.a: Populationby age groupinGLSS surveys, 1991-2006 ................................................ 18 Figure 5.b: Dependency Ratios in GLSS surveys, 1991-2006 ........................................................ 18 Figure 6: Income fromwork for 2005 inAge group 25-64............................................................ 30 Figure 7: Share of salaried workers with written contracts by sector (%) .................................... 38 Figure 8: The number of lay-offs registered at the Department of Labor................................... 39 Figure9: Dynamics of minimumwage andinflation, 1994-2006.................................................. 41 Figure loa: Density Function for Wages, Age 18-30, 1991/92...................................................... 42 Figure lob: Density Function for Wages, Age 18-30, 2005/06 ..................................................... 42 Figure 1Oc: Density Functionfor Wages, Age 25-64, 1991/92 ..................................................... 42 Figure 10d:Density Functionfor Wages, Age 25-64, 1991/92 ..................................................... 42 Figure11:Annual real earnings andminimumwage, 2005/06 (`000 cedis) ................................ 44 Figure 12:Monthly Reservation wages by gender, location, and education (2005/06) ..............44 Figure 13: Share OfWorkers With SocialBenefits, 1991/92 And 2005/06 (9`0)........................ 49 Figure 14: PublicExpenditure on PrimaryEducationas a Percent of Per Capita GDP:2005 60 Figure 15: Governmentof Ghana Per Capita Expenditures on Primary Education:2004 - 2006......................................................................................................................................................... 61 Figure 16: The Capacity of PublicWET Institutionsto Absorb the Pipehne ofJunior Secondary School Enrollment by Region: 2006/07 ....................................................................... 64 Figure 17: UnitCost of EducationinGhana: 2006......................................................................... 67 Figure 18: Impact of Education on the Likehhood of Wage Employment (vs no Education), . Urban...................................................................................................................................................... 73 Figure 19: Impact of Education on the Likelihoodof Non-AgricultureSelf-Employment (vs. no Education),Urban Figure 20: Impact of Education on the Likehhoodof Wage Employment(vs. no Education), ........................................................................................................................... 73 Rural........................................................................................................................................................ 74 Figure21: Impacty of Education on the Likelihoodof Son-AgricultureSelf-Employment (vs. no Education),Rural ............................................................................................................................ 74 Figure 22: Waghe Repressions with Sample Selection, Ghana 1991-2006, Urban..................... 75 Figure 23: Non-AgricultureSelf-Employed Regressions with Sample Selection, Ghana 1991- 2006, Urban........................................................................................................................................... 75 iii GHANA Job Creation and Skills Development TABLES Table 1: Growth rates and sources o f growth from a Solow-Denison Model PA).................... 4 Table 2: Ghana: Sectoral sources of economic growth. 1990-2006 PA) ................................... 6 Table 3: Labor Force Participation inage group 25.64. 1991-2005.............................................. 19 Table 4: Comparisonof employment rates by age group. 1991-2005 .......................................... 20 Table 5: Composition of the Labor Market. 1991-2005 ................................................................. 20 Table 6: EmploymentDistributionby Sector (percentage. for age group 25-64) ....................... 23 Table 7: EmploymentDistributionb y Sector (absolute numbers. for age group 25-64) ...........23 Table 8: Employment Status (percentage. for age group 25-64) 1991-2005................................ 24 Table 9: EmploymentStatus (Absolutenumbers for age group 25.64). 1991-2005 ..................24 Table 10: Percentage of Workers Residingina Poor Household. for Different Categories of Labor Market and EmploymentStatus for age group 25.64. 1991-2005..................................... 25 Table 11:Percentage of Workers Residingina Poor Household. for Different Categories of Labor Market and EmploymentStatus for Age Group 25.64. inRuralandUrbanAreas ........25 Table 12:Percentage ofWorkers residing ina Poor Household by Sector of Employmentfor Age Group 25.64. 1991-2005.............................................................................................................. 25 Table 13: Percentage of People Residmgina Poor Householdby Level of Education for Age Group25.64. 1991-2005...................................................................................................................... 26 Table 14:UnemploymentRates for Age Group 25.64. 1991-2005 .............................................. 27 Table 15: Poverty (head count index) among the Unemployedfor Age Group 25.64.199 1. 2005......................................................................................................................................................... 28 Table 16: Distributionof the Populationby Education LevelinAge Group 25.64.199 1.20055 ................................................................................................................................................................. 29 Table 17: Annual Real Earnings across Employment Status inAge group 25-64 (in`000 cedis) ................................................................................................................................................................. 31 Table 18: Earnings Ratio's to private formal sector wage inage group 25.64. 1991-2005 ........31 Table 19: Mean Number of Hours Workedper Year in age group 25-64 (1991-2005) ............32 Table 20: Earnings per Actual Hour Worked by Economic Activity. Employment Status and Poverty inage group 25.64. 1991-2005 (incedi) .............................................................................. 33 Table 21: Percentage increase inwages from belongingto the formal sector. beingin an organization with a trade unionandbeinginthe public sector on wages.................................... 45 Table 22: Mandatory non-wage labor costs to employers in selected countries inWest Africa ................................................................................................................................................................. 46 Table 23: Share of salaried workers with social protection ("/o) ..................................................... 48 Table 24: Share of salaried workers with benefits by firm size (%)............................................... 48 Table 25: The NYEP Youth EmploymentRegistry Data .............................................................. 51 Table 26: The NYEP Beneficiary Data. 2006-07 ............................................................................. 52 Table 27: Enrolmentby Level of Schooling for 2004/05 and 2006/07 ....................................... 59 Table 28: Governmentof Ghana Education Expendtures by Sub-sector: 2004....................... 61 Table 29: Types of TVET Institutions ............................................................................................. 65 Table 30: Share of youth (15-30 age group) gone throughTVET programs.............................. 66 Table 31: Share of youth (15-30 age group) gone throughApprenticeships............................... 69 Table A1 Employment. Unemploymentand Schooltng rates. age group 15-24....................... .. 21 Table A2: EmploymentDistributionby Labor Force Status (percentage) .................................. 21 iv Executive Summary Thanks to a steady pace of structural reforms and sound macroeconomic management, Ghana has experiencedfast and sustained growth. Since the 1990s, the annual GDP growth rate has averaged 5 percent, reaching about 6 percent in 2005-2006. This growth has outpaced that experienced by most Sub-Saharan Africa low-income countries during this period. Despite major gains, as discussedinthe recent Ghana CEM (see Bogetic et al, 2007), significant policy challenges remain in terms of consolidating gains and further accelerating economic growth. High economic growth has contributed to a sharp reduction in poverty. The core poor represented28 percent of the population in 2005, down from about 52 percent in 1991 (see Coulombe, Wodon, 2007). This poverty reduction came mainly from growth, despite an increase ininequality. However, the pace of poverty reduction has been weaker innorthern regions, and if poverty i s to be further reduced, spatial inequities need to be addressed through an effective regional policy. Population growth in Ghana is low by Sub-Sahara African standards and is decreasing, but the urban population is increasing. Ghana's population growth rate has fallen to about 2.5 percent, and this has translated into a higher ratio o f workers to dependents.Yet capitalizing on this opportunity will require a significant acceleration of the pace of job creation inthe economy. Moreover, the rapid urbanization (from 32 percent in 1983 to 48 percent in 2005) has brought with it an increase in the number o f new job seekers, especially incities and among youths. The Government is placing priority on employment, both in terms of fighting poverty and as an underpinning for economic growth. Especially given recent demographic trends and increasing urbanization, creating more and better jobs is an explicit priority for the Government, as reflected in the Government's second Poverty Reduction Strategy (GPRS 11). In late 2005, the Government launched the National Employment Summit to formulate the national employment policy and strategy. In addition, to promote the development of human capital and strengthen links between skills development and employment, inJuly 2006 the Parliament passed a Bill to establish a Council for Technical and Vocational Education and Training (COTVET) to formulate national policies for skills development across the broad spectrumof formal, informal, and non-formal education. I n this context, and complementingthe recent CEM and earlier Bank's work on youth employment, this report addresses labor market challenges that Ghana is facing, particularly as it strivesfor middle incomestatus by 2015. This report: (i) reviews briefly the determinants o f labor demand; (ii) analyzes labor market outcomes based on recent survey data; (iii) reviews the role of labor policies, institutions and programs; and (iv) examines education and skills development policies and their links to labor market outcomes. Themainfindings of the report can be summarized asfollows: Ghana's growth created newjobs during 1991-2005 and as a result of the structural changes in composition of output, labor has shifted from agriculture to better paid I jobs (self-employment and informal wage employment), leading to the growth of informal urban employment, This has created better income-earning opportunities and has contributed to poverty reduction. Nevertheless there are several challenges to sustaining these gains inthe future. 0 The prerequisite for preserving and strengthening recent growth and poverty reduction, as well as promoting job creation, i s macro policy (including investment climate) conducive to the ongoing shift towards higher productivity sectors. Since wage employment i s and will remain a small share o f the labor force for some time, the overall job strategy must continue to broaden the source o f agriculture growth, while stimulating growth o f other formal sector jobs (i.e., manufacturing, services). 0 In addition, other key elements for creating sustainable jobs, both in formal and informal sectors, are: (i)an adequate industrial relations framework (including realistic firing procedures, minimumwages, and non-wage costs); (ii) cost-effective employment programs; and (iii)a good quality education and training system. More specifically: P Although labor market functioning seems flexible, (i) effect of rigidities the (e.g., union mark-up on wages) on the economy may be masked by the dominant effect o f sustainable macro performance; and (ii)in the future, the growth of modern sectors may be slowed down by rigidities at the micro level; P Given the challenges of youth employment, the existing program could be an effective tool to increase the employability o f youth, but it needs (i)clear criteria and objectives, and (ii) close monitoring to ensure that it has a positive > impact on the beneficiaries; Ghana has been successful inexpanding basic education, but challenges remain in(i) addressing quality and equity issues, (ii) meetingthe increasingdemand for post-basic education; (iii) ensuring that the education andtraining system allows youth and older workers to acquire skills and return, when needed, for further education; and (iv) providing for those who leave educationearly, a traditional apprenticeship system that offers good quality skills recognized by employers and opens opportunities for further education. DETERMINANTSLABOR OF DEMAND Demand for labor cannot be viewed outside o f the overall development inproduct markets. The review of growth patterns and characteristics o f firms in Ghana during the last fifteen years provides the context in which to consider challenges and prospects for job creation over the medium term. Overall growth created new jobs, while structural changes in output composition led to shifts in labor shares creating better income-earning opportunities and contributing to poverty reduction. During the last 15 years, most sectors with strong growth in output experienced growth in employment. Growth has generated employment and wage gains which have inturn led to poverty reduction. The highest growth was in services, followed by industry, while agriculture share o f GDP declined. As a result, the inter-sectoral labor reallocationinGhana between 1991 and 2005 saw a sectoral shift inthe composition o fthe labor force, from the agricultural sector toward, among others, the financial services, transport and communication, and construction sectors. This confirms that people move to I1 better paying jobs. At the same time, wage employment in the formal sector is not growing fast enoughto absorb the new influx o f workers especially inurban areas, so that many workers are moving outside agriculture into self-employment and informal wage employment, leading to growth ininformal urbanemployment. Nevertheless, past trends in employment generation and wage gains need to be treated with caution. Based on the characteristics o f the growth and employment, as well as some possible overestimation o f wage gains inthe past, there are several challenges to sustaining these gains in the future First, much o f the growth in employment has taken place in the informal sector, which does not provide the same security and benefits as the formal sector and may not be attractive to skilled labor. Second, while there have certainly substantial gains in earnings, consumption, and poverty reduction, due to measurement issues, some o f the past gains may have been overestimated (low increase in the Consumer Price Index as compared to the GDP deflator, especially in urban areas outside o f Accra). Third, part o f the gains in earnings has been obtained through an increase in working hours, which may not be sustained in the future. Fourth, the pressure on the labor market to absorb a large number o f new entrants will remain in the future due to the demographic transition phased by the country. Fifth, at the macroeconomic level, the contribution to growth o f the improvement in factor productivity has been limited, and probably due over the last few years in large part to the boom of the cocoa sector, a success that may not be easily reproduced in future years. Finally, although there have been sectoral shifts in the labor force to better paying jobs, more than half of the labor force remains employed in agriculture, in most cases in low productivity jobs, so that the transformation o f the economy remains limited. It is expected that the movement away from the agriculture sector will continue in the future and therefore policies (among others in terms o f the investment climate) need to help increase productivity in the informal sector (agriculture and non-agriculture) as well as shift Ghana's economy towards its higher productivity sectors. Reflecting the Government's strategy, Ghana's growth has been driven by domestic absorption, and exports have not played an important role as an engine of growth, with the exception of cocoa in recent years. Experience in other countries has shown that the model o f growth that focuses on domestic demandand investmenti s not likely to provide a sustainable source o f employment. Moreover, due to lack o f diversification - 70 percent o f its exports are traditional products (cocoa, gold and timber) -Ghana remains vulnerable to commodity price shocks. Although the increase inthe price o f cocoa inrecent years had a direct impact on trade, its impact on the labor market was limited: the share o f workers in cocoa production remained stable during 1991-2005 (6 percent), while the share o f non- cocoa agriculture workers decreased (from 54 percent to 47 percent). The increase cocoa prices ledto an increase inrural earnings among cocoa producers, many o f whom are small farmers. Thus, poverty incidence among these producers decreased significantly (from 62 percent in 1991 to 21 in 2005) as compared to non-cocoa workers. Second round effects from the increase in rural spending o f these producers also contributed to rural poverty reduction. The characteristics of Ghanaian firms, which are less productive than firms in comparator countries, explains the concentration of job creation among self-employed. As discussed inthe CEM, Ghanaian firms are less productive mainly because they are less I11 capital and skill intensive andrelatively small insize, therefore, they do not account for the bulk of the economy's job creation. Ghanaian firms are mainly concentrated in the traditional sectors, such as manufacturing and trade, and given the country's growth model, their production is primarily oriented toward the domestic market. As such, they create temporary low-skilledjobs at low wages. As demonstrated by Teal, the small size o f most firms likely explains the concentration o f job creation among the self-employed. In this context, a business climate that i s favorable for investment will also promote employment by allowing large and labor-intensive firms to grow and create more jobs in the urban formal sector. LABOR MARKET OUTCOMES The economic growth has been associated with substantialjob creation, and mostpeople continue to be employed in agriculture. The employment rate o f the working age population (25-64 years o f age) has remained around 89 percent over time. Between 1991 and 2005, employment increased by about 2.8 million jobs (from 4.1 to 6.9 million for people from the 25-64 age group), which i s comparable to the labor force growth (from 4.8 to 7.9 million), as well as the population growth among that age group. As discussed earlier, although economic activity i s slowly shifting away from agriculture, the agriculture sector i s still by far the main sector o f employment, accounting for 54 percent o f the labor force (age group 25-64), followed by services (3 1percent), and industry (15 percent). Most of the labor force remains employed in the informal sector, which has accounted for most newjob creation in thepastfifteen years. Interms o f composition, employment i s dominated by self-employment (82 percent o f workers), o f which over two thirds are working inthe agriculture sector. A minority (about 18 percent o f the labor force) work in wage-earning jobs. Indeed, during the last 15 years, the percentage o f formal jobs (both public and private) as a share o f total employment has decreased (from 16 percent to 13 percent). This i s due to the fact that the decline inthe share o f public sector jobs has been only partially compensated by an increase in the share o f private formal jobs. Since wage employment in the formal sector is not growing fast enough to absorb the new influx o f workers especially in urban areas, many workers are moving outside agriculture into self- employment and informal wage employment, leading to growth in informal urban employment. Thus, the labor force has become more informal over time. The decrease in the formal sector as a share o f total employment i s also confirmed by a series o f other indicators, such as decrease in the share o f workers with contracts, and decrease in the share o f unionized workers (see below). Earnings have increased substantially during the last 15yearsfor all groups and the gap between the public sector and the private formal sector has continued to increase. Between 1991 and 2005, average annual earnings increased about 77 percent inreal terms. This rise in earnings has been an important driver of poverty reduction in Ghana. The largest gain was among the self-employed in the agriculture sector (92 percent increase) followed by wage workers in the public sector (89 percent). Despite these increases in earnings, the wage gap between different employment categories remains high, andthe gap between public sector workers and other workers i s increasing. Overall public sector workers earn 47 percent more than private formal sector employees. Wage workers in the private informal sector are the third highest-earning category, earning about 63 percent o f IV what private formal sector workers earn, followed by the non-agriculture self-employed. Finally, despite their earnings gains, agriculture self-employed workers - accounting for the majority of the employed - earn almost half of what non-agriculture self-employed workers are paid. Unemployment remains low and is primarily an urban phenomenon. Overall, the proportionof the unemployed remains very small (2.3 percent in2005), and as inother low income countries, is heavily concentratedinthe capital (Great Accra 4.9 percent). It is also concentrated among young persons (between 15-24 years old) and those who have an upper secondary education (6.1 percent unemployment rate). Although unemployment in low income countries has often been described as a luxury, there is heterogeneity among the unemployed in Ghana, with about one out of five unemployed workers residing in a poor household in 2005. Moreover, poverty decreased more slowly among the unemployed over the past years. This suggest that there may be two types of unemployed: (i) queuingforagoodjob(withhighearningsandotherattributes,particularlyinthe those public sector), and (ii)those not finding a job, probably due to urban migration by relatively young unskilled workers and lack of sufficient wage employment opportunities inthe urbancenters which are not expanding fast enoughto absorb the increasing number ofjob seekers. Returns to education are positive, suggesting that youngsters have an incentive to continue their education. As discussed below, education influence the type of job held and earnings received, confirming that the higher the level of education, the more likely one i s to become a wage and salary worker. Overall, education (including TVET) lowers the likelihood of becoming a self-employed worker in the agriculture sector and increases the chances of becoming either non-agriculture self-employed or a wage worker. Basic education has become increasingly important to the self-employed, improving their earnings. While TVET and tertiary education affect the likelihood of wage employment, education also influences earnings directly, with gains in earnings with TVET matching those of general secondary education. This is an important finding when it comes to choices for further investment inpost-basic education. LABOR MARKET POLICIESAND INSTITUTIONS Although the share o f the formal economy inGhana is below 20 percent and mostjobs are in the informal sector, increase injob creation in the private formal sector confirms that some formalization is taking place: notably, total wage employment in the formal private sector more than doubled during 1991-2005 (from 167,000 in 1991 to 498,000 in 2005). Therefore, it is important to examine the functioning of the labor market andto review the role of labor regulations and institutions for job creation inthe growing economy. Ensuring that labor market policies create a stimulating environment for a dynamic private sector that would createjobs and increaseproductivity while providing adequate social protection to workers is a challenging task for Ghana as it strives to reach middle-income economy status. Overall, Ghana labor regulations are relatively flexible compared to those of its neighbors, and the 2003 Labor Act has introduced moreflexibility to the labor market. Among the more important changes are: (i) consolidation of a large number of old labor laws to facilitate coherence and to improve application and enforcement o f the labor code; (ii)creation o f the National Labor Commission to tackle labor disputes; (iii) introduction of private employment agencies to facilitate job searches by the unemployed and to help employers find workers with desired skills; and (iv) flexibility in the hiring o f temporary workers. Although the changes introduced by the new law are a step in the right direction, dismissal procedures are still complex and costly. Ghana hiring and firing rules are overall flexible, but dismissal costs can be high and the procedures are cumbersome. According to employers' surveys, labor regulations in Ghana do not seem to be a constraint to job creation, but further investigation i s still needed to better understandthe effects o f existing labor regulations onjob creation inthe private formal sector. This will allow Government to take necessary steps to ensure that modern sector growth i s not impeded. According to 2008 Doing Business, firing costs inGhana are equal to 178 weeks o f wages, which is over twice the average cost in Sub-Saharan countries and about five times the cost in OECD countries. While Ghana's high dismissal costs could partially be explained by the methodology employed by Doing Business (based on cost for workers with over 20 years o f experience), they can also be due to the lack o f safety nets for those who may lose their jobs. Overall, based on international experience, the existing set-up raises several concerns for the labor market. First, flexible hiring procedures combined with costly and cumbersome termination regulations could lead to a fast increase in temporary jobs without security. Second, costly firing regulations impede enterprise restructuring and efficient resource - including human resource - allocation, resultingin a lower level o f formal job creation. Third, the impact could be substantially more negative for youth, women, andthose with low skills. The minimum wage, despite continues increase, does not currently seem to constrainjob creation, but if enforced, it could substantially increase the cost of unskilled labor. According to the 2003 Labor Act, the National Tripartite Committee annually sets the daily minimum wage, which is currently equal to $19,000 (US$1.7). The GLSS data shows that the minimum wage is high incomparison to prevailing wages, inthe sense that an increasing share o f workers now has wages below the minimum wage, especially among youth and inthe informal market. While this may not have yet had negative impact due to weak enforcement, the continuous increase o f the minimumwage inthe last fifteen years at a rate much faster than the inflation rate may cause problems in the future. Therefore, a close follow up o f the increase of the minimumwage and a reassessment o f its level are warranted, especially interms o f its potential impact on the cost o f unskilledlabor and access to formaljobs for youth. Mandatory social security charges are relatively low in Ghana, though firms also pay additional non-wage benefits agreed through collective agreements with workers ' unions. Mandatory social security charges equal to 17.5 percent o f the gross wage bill are shared betweenthe employer (12.5 percent) and the employee (5 percent). Inaddition to social security contributions, employers, especially in large firms, also provide additional non-wage benefits agreed through collective bargaining with workers' unions. The Government i s currently considering several pension reform options, and it will be important to assess the impact o f the proposed reforms on the labor market to ensure that VI the new pension schemes do not distort workers' labor force participation decisions and do not increase inappropriately wage costs for employers. The share of unionized labor hasfallen in the last 15years, but there are still barriers that separatepublic and unionizedemploymentfrom private non-unionized employment. Currently about 10 percent o f the total labor force inGhana is unionized, which i s less than a half o f the 1990 rate o f 26 percent. This decrease was due mainly to a decline in the number o f public sector jobs and a fast increase in informal jobs, with no job-security. Nevertheless, the majority o f formal sector workers in Ghana, particularly in the public sector, are unionized (70 percent o f the public sector wage employees and about 30 percent of the private sector wage employees), and controlling for human capital, there is a wage premiumassociated with unions. However, despite the active and influential role played by the unions, the GLSS data show that this wage premium may well have decreased in recent years. Further work i s needed to better understand the factors affecting this decline in both union membership and its wage premium, including barriers that separate public unionized employment from private non-unionizedemployment. EMPLOYMENT PROGRAMS As Ghana aims to reach middle income status, an effective employment program and safety net will be essential to support workers, including youth. The government has recently developed a few employment-related interventions, but the National Youth EmploymentProgram (NYEP), launched in2006 by the Ministryo f Manpower, Youth and Employment, i s the major program that directly addresses job creation. Though it i s too early to evaluate the impact o f the NYEP on employability o f its beneficiaries, the design o fthe program raises some concerns. The NYEP has a high cost and mainly targets educated urban youth. Emphasis o f the program on youth (of 18-35 years o f age) seems justified, given that this group's unemployment rate is about twice as high as the national unemployment rate (6 percent compared to about 3). However, since 2006, the NYEP has consumed significant resources (about six times larger than the total budget o f the Ministry o f Manpower) to support a relatively small portion of youth (about 90,000 people). This scale o f spending could be justified only if programs improved labor market functioning by helping the youth to find employment. The design of the program raises several issues vis-a-vis labor marketfunctioning: (i) the wage paid by the program (ranging between 500,000- 2,000,000 Cedis per month) are highcompared to the national minimumwage (360,000 Cedis), creating disincentives for youth to look for jobs outside o f the program and building unrealistic expectations for future earning prospects; (ii) the targeted groups are primarily urbanyouth with secondary and higher education, thus from relatively better-off households and not really aimed to poverty alleviation; and (iii) the program mainly createstemporaryjobs inthe public sector (e.g. education and health services), and given the government's limited fiscal space, it raises issues related to jobs' sustainability. Theprogram's monitoring and evaluation system is weak. Currently, some information on the characteristics o f the beneficiaries is collected at the regional level. However, there i s no information on the results of the program and there i s no sound monitoring and VI1 evaluation system in place to ensure that resources are being effectively spent or to adjust program's design as needed. EDUCATION, SKILLSAND LABOR MARKET OUTCOMES For Ghana improving skills development i s key to the attainment of middle-income status. Intensified economic competition in countries worldwide has increased the focus on education and training system's ability to prepare youth for employment and support more seasoned workers in adjusting to structural changes taking place in labor markets. Inthis context, concerns exist as to whether skill deficits have or may become a constraint to Ghana's further growth and capacity for reducing poverty. The review of the education and training system in Ghana confirms that it is expanding, butfaces a number of challenges, such as: (i) regional disparities in access; (ii) gender inequalities that increase with the level of education; (iii) low rates of completion for basic education; (iv) infrastructure in need of repairs; and (v) teachers needing further qualifications. At the same time, demand is growing for post-basic education, adding pressure to Ghana's efforts to finance a good quality basic education for all. To acquire the vocational andtechnical skills neededby employers or necessary for starting one's own business, people must rely on a small public TVET sector with its own quality issues, the willingness of enterprises to provide workers with training, and a large traditional apprenticeship system. Education and skills training influence wages and earnings directly by raising the productivity of the worker and indirectly by promoting entry into more lucrativeforms of employment. Overall, education has a positive impact on the type of employment and earnings and GLSS data confirms that demand exists for education and skills at all levels, from primary to tertiary, including TVET and apprenticeship. Investing in education i s generally associated with movement out of the agriculture sector to wage or self- employment inthe non-agriculture sector, and the likelihood of becoming a wage work in urban areas increases with the level of education, controlling for other personal characteristics. In general, having less than a secondary education raises the chances o f self-employment in non-agriculture, and to a lesser extent it raises the chance of self- employment in agriculture. TVET exerts a positive influence on wage and self- employment, particularly in the non-agriculture sector. Apprenticeship also offers a pathway to self-employment, but is less effective than TVET in opening access to wage employment. Post-basic education improves the likelihood of wage employment,while those with only primary education largely work as self-employed in agriculture, and junior secondary education translates into self-employment in non-agriculture. To promote wage employment, primary education i s not sufficient and there is need to plan for more post- basic education. The analysis shows that primary education or the junior secondary education alone do not yield significant earning gains compared to those who have no education. However, the additional completion of a senior secondary education or TVET provides substantial earnings gains in both rural and urban areas and for virtually all categories o f employment. VI11 Ghana's success in expanding access to education, beginning with basic education, has built afoundation for the expansion of post-basic education. Economic growth, now in the 6 percent range, will increasingly demand more skilled workers, and the economic returns to education, evident in the results above, will justify this expansion. The investment is important, not only for its direct impact on the productivity of the workforce and their earnings, but for its indirect influence on the access to wage employment and the higher earnings found in these jobs. At the same time, evidence from the review of the education and training system shows that investment in quality at all levels has to accompany investments in expanding access, and that these investments need to take into consideration balancing opportunities among and withinregions and by gender. Against this backdrop, although enrolment in education is rising at all levels, 9 out of 10 youth aged 15-17 are not enrolled in senior secondary education either because they have dropped out of school or because they are behind in their curriculum due to repetition. Many of them have altogether left formal schooling tofind work or acquire skills through traditional apprenticeship or by learning on the job. Public initiatives to provide skills training reach nearly 48,000 young people, but suffer from low quality and limited connection to the market. Meanwhile, a 2003 Ghana Statistical Service survey indicates that there are over 207,000 active apprentices, mainly inthe informal economy. Traditional apprenticeshipsoffer the primary source of skills development today in Ghana, but they suffer from weak pedagogy, limited education of master craftsmen and apprentices, instruction in out-dated technologies, and a lack o f standards and quality assurance. A large private training sector exists, but it too i s of variable quality. There is ample room for a country with middle-income aspirations to invest in broadening access to good quality secondary education while also strengthening the foundation for basic education. The main policy findings confirm that: (i)Ghana has been successful in expanding basic education, but quality and equity issues remain; (ii) there are challenges inmeeting the demand for post-basic education for those who progress through basic education; (iii)there i s a need for an integrated education and training systemwith numerouspathways allowing youth and older workers to acquire skills needed by the market and return, when needed, for further education; and (iv) for those who leave education early, there is need for a traditional apprenticeship system that provides good quality skills, creates credentials accepted by employers and in a National Qualifications Framework, recognizesthese skills and opens opportunities for further education. POLICY AGENDAFOR THE LABOR MARKET The strategy for addressing Ghana's employment challenges must be based on multi- sectoral reforms that create a favorable environment for job creation in the private sector. Jobs are created when new firms are created and existing firms expand. This requires a multi-sectoral strategy that is based on macroeconomic stability, encourages investment,allows for the mobility of workers from declining to growing sectors, creates a highly skilled and flexible work force, and provides support for workers while creating incentives to find employment. Macroeconomic stability and a dynamic private sector are prerequisitesfor job creation. The recent CEM provided recommendations for sustaining high growth. Although IX recommendations in the area o f economic policy are beyond the scope o f this report, it i s important to emphasize that sustained growth i s essential for promoting job creation in Ghana. Though the 2008 World Bank's Doing Business report selected Ghana as one o f the five best reformers o f Sub-Saharan Africa for facilitating business practices, further efforts are needed to address procedures for formal business start ups, which are still cumbersome and so push firms to informality. Moreover, there are serious concerns regarding small and medium-size enterprises' access to electricity and finance, which need to be addressed. As part of the job creation agenda, thefunctioning of the labor market needs to be improved. Although the most important constraints to job creation lie outside the labor market (i.e., macro uncertainty, business climate), labor market policy reforms are also needed. Ghana has a relatively flexible labor market arrangement. However, rigidities at the micro level, such as labor regulations (particularly dismissal procedures) add additional cost to labor, particularly in larger firms, and may impede both firms' ability to undertake structural changes and job creation. According to lessons learned in other countries, it would be advisable to have fewer protective measures inside the firm (aimed at preserving jobs), and more effective schemes outside the firm, through a social safety net, including active labor market programs. This could be achieved through effective re-employment programs andperhaps income support programs. Growth in labor productivity will also require a highly qualified laborforce and thus a strengthened education and skills development system. For a country that aspires to middle-income status, interventions to improve education and training need to be balanced with interventions in economic and labor market policies to promote a favorable climate for private investment and job creation. Overall, Ghana has taken important steps toward opening access to education and introducing TVET reforms. Against this background, further investments inpromoting access to good quality education and training for all are essential and it i s important to invest in skills training beyond general education. Increasingly in today's global environment, the design o f education and training systems need to promote learning throughout all stages o f the people's lifecycle, particularly for workers who will change jobs many times during the course o f their working lives as the economy restructures in response to global economic pressures. In this context, priority actions could focus on: (i) improving the quality o f both formal education and TVET and strengthening the articulation within the two systems to provide choices to youth in acquiring skills and accommodating different learning needs; (ii)strengthening and improving the quality o f traditional apprenticeships; and (iii) providing incentives for using resources more effectively through introduction o f performance accountability and performance-based budgeting. Sequencingpolicy responses will be an important consideration; some reforms ought to take place over the next two years, while others should be part of a medium-term agenda. The most immediate actions that are required in the labor market are: (i) to continue investment inbasic and post basic education to improve access and quality andto improve the performance o f the skills development systems (both formal and informal); (ii)ensurethattheongoingpensionreformsdonotincreaselaborcostexcessively; (iii) to to review the design o f the NYEP; and (iv) to introduce a monitoring and evaluation system for the youth employment program. Over the mediumterm, Ghana could consider X (i) partofitspovertyalleviation,introducingself-targetedpublicworkprogramstooffer as temporary employment opportunities for youth; (ii) linking funding o f training systems to outcomes and performance; and (iii) reviewing dismissal procedures. In addition further work i s needed to (i)assess the impact of the minimum wage on formal job creation; (ii) review the barriers that separate public unionized employment from private non-unionized employment; and (iii) review industrial relations. This policy agenda is summarized inthe matrix below. XI SUMMARY OF MAINFINDINGS AND POLICYAGENDAINGHANALABOR MARKET Main Findings Desirable Outcomes Policv ODtionsto Achieve Outcomes Short term Labor market oolicies do not oose a maior constrainttojob creation in the-formal sector enable firms -to adjust on the fkms and propose necessaryreforms in the short run, but merit consideration in quickly to rapidly changing reforms to reduce costs for the medium term business conditions while employers as well as Hiring regulations are relatively flexible, ensuring the protection of considering social protection for those who may lose their but firing proceduresare complex and workers canbe costly jobs Monitor closely Minimum wage policy does not seem to Ensure minimum wage do Asses impact of the minimum minimum wage constrainjob creationbut it has morethan not increase faster than wage on access to formal jobs trends doubled in real terms inthe last 15 years. inflation for youth Non-wage costs are relatively low, but Cany out sound the proposedsocial security reforms Pension reforms do not Assess the impact of various pensionreforms couldleadto an increase inlabor cost increase non-wage labor pension reform options on costs labor market and labor costs Large firms pay additionalnon-wage benefits that increasethe cost of labor Non-wage labor costs do not Evaluate impact o f non-wage affect negatively job costsjob creation creation There are barriers that separate public and Labor unions do not impede Review barriers that separate unionized employment from private non- job creation in the formal public unionized employment unionized employment and state labor market sector or in large firms from private non-unionized institutionsare weak employment According to the GLSS data, share of unionizedlabor has fallen inthe last 15 years and union wage premiumhas decreased Given the role of unions inwage negotiations,unionizedworkers have about 20 percentwage premiumover the non-unionizedlabor force Develop a modernized New labor law has shiftedresponsibilities institutional framework Review the Industrial relations betweenunions and state institutions to refine the responsibilities and roles of various State labor institutionsare weak, with institutions and identify the little capacityto enforce regulationand areas that need further monitor the labor market strengthening and institutional capacity building. The National Youth Employment Program Revise the design of the Consider targets primarily educated youth while other employment programs that program to ensire that it introduction of governmentprogram concentrate mainly on provides short-term support reaches vulnerable youth self-targeted income generationandpoverty reduction to youth employment and beyond urban areas; the level public work The programwas launchedin 2006 to does not disturb the of wages do not distort the schemes to offer help the youth (betweenage of 18 and functioning of the labor labor market; involve more temporary 35) to find ajob market closely the private sector in employment the design and opportunities for The designof the programraises implementation; keep the youth. concerns:the level of wages paid is high durationof the programshort; comparedto the minimum wage; the and review the eligibility beneficiaries are mainly educatedurban criteria. youth from better-off families; and the program mainly creates temporaryjobs in Introduce a solid monitoring the public sector. and evaluation system. The program is costly, with weak M&E system. XI1 A strengthened foundation Continue investment in basic Strengthen secondary education force many youth to for basic education as well as post-basic education capacity for leave education after junior secondary school to improve both access and planning and and to search elsewherefor skills. quality at all levels and to take policy Broader access to good into ,account regional and development Enrolmentinbasic andsecondary educationis expanding, butthere are quality secondary education gender disparities substantialdisparities across regionsand the system is facing many issues interms of quality of teachers and infrastructure Net enrolmentrate in secondary educationis low (10.6%): only one out of ten youth (15 to 17 years of age) is enrolledin secondary education Public providers of skills training offer only a Reformed technical and Strengthen the capacity and Link funding of small number of places for youth and the vocational education at improve the quality of public the training quality is low secondary and tertiary levels and private technical system to Only 10%ofyouth (15-17) are enrolledin vocationaltraining outcomes and Improved quality and better performance public training centers policy development through Improve the performance of (performance- The resourcesallocated to TVET are strengthening links between public and private training based budgeting) insufficient to provide goodquality schools, training centers and institutionswith performance- instructors and adequatelyequipped employers based financing and Evaluate the workshops strengthening employers' impact of training engagement and introduce tracer studies National Training Authorities (coordination bodies) engage employers, worker representatives, and Gov. in coordination and governance of training system and more efficient use of public resources Traditional apprenticeships are a major Reformed non-formal Strengthen capacity and source of skills trainingfor youths, but quality training system through improve quality of traditional varies widely, and it is doubfful this system facilitating the improvement apprenticeships will meet the skill needs of an aspiring process (rather than the middle-incomecountry formalization of traditional Bring literacy to master Over 80% of youth pursueskills apprenticeships) craftsmen and apprentices, development through traditional provide access to new apprenticeships technologies for master craftsmen upgradepedagogical Most apprenticesand masterslack formal skills and technical knowledge vocational or technical training, of master craftsmen Some apprenticeslack sufficient basic Provide certification to education apprenticeshipskills The pedagogicalskills and knowledge of new technology is low amongmaster craftsmen There are no instruments for assuring the quality of traditional apprenticeships XI11 The many youth who miss first-chance Expanded choices for youth Improve the articulation Introduce market opportunitiesfor education have few second- and opportunities for them betweeneducation and training institutions that chance options to returnto formal education and open vertical and provide quality horizontal pathways to assurance o f The lack of basic educationreducesthe once they have left chances of youth beingselected for improvechoice training programs training by employers Provide better information to market participants Develop Those who miss early education information opportunities offer aweak foundation Children at risk to dropping instruments, such for the developmentof askilled out are able to find support Provide programs for at-risk as national workforce helping them remain in children that offer tutoring, qualifications school raise the demand for further frameworks, education, and lower income standards based Missededucationopportunities can Children who have left constraintsto keepingyouth in training, testing increaselater socialcosts school and do not wish to school and certification return to formal education process. can find equivalent non- Introduce "second-chance" formal educationprograms education programs that use non-formal education methods to deliver education that is accepted as equivalent to basic and secondary education programs XIV CHAPTER 1. INTRODUCTION BACKGROUND 1. According to Government, Ghana isfacing many challenges in the area of skill development and job creation. The Government is particularly concerned with (i) unemployment among the youth (6.1 percent in 2005 for 15-24 years old), which has been attributable to the rapid population growth rate o f the youth, the rapid urbanizationrate, the quality o f labor supply, and low labor absorption rate o f the economy; and (ii) external the efficiency o f the educatiodtraining system and its ability to supply the skills demanded by a diversified and competitive economy. Moreover, other related concerns include: channeling economic growth toward creating jobs, including `good jobs' '; better understandingthe functioning o f the informal sector; explaining the mismatch between skills development andjobs; improvinglabor market indicators; monitoring and evaluating employment programs' outcomes; reviewing the role o f labor market regulations in job creation; and reforming technical vocational education and training (TVET) systems. 2. The key objective of Ghana's development policy is to accelerate economic growth and put the creation of new and better jobs at the center of the Government agenda. Ghana aims to achieve middle income status by 2015, and the Government Second Growth and Poverty Reduction Strategy (GPRS I1for 2006-2009) indicates clearly the determination o f the authorities to mainstream employment policies and programs in the overall development agenda of the country. To achieve this objective, the Government launched the National Employment Summit with a High-Level Technical Workshop o f major stakeholders in employment in November 2005. Moreover, to promote human development capital and strengthen links between skills development and employment, a Council for Technical and Vocational Education and Training (COTVET) has been established to formulate national policies for skills development across the broad spectrum o f formal, informal and non-formal education. 3. To reach middle income status by 2015, an annual GDPgrowth of 6 to 8percent is needed. This implies that Ghana needs to accelerate its growth rate, which i s already at its highest point since 1960. To assist the Government in operationalizing its agenda, the World Bank has focused the recent Country Economic Memorandum (CEM, report # 40934-GHYNovember 2007) on accelerated and shared growth. To complement this work, as well as the earlier Bank's work on youth employment,* and assist the Government in its thinking on constraints and opportunities for job creation and skills development, the preparation o f this report on "Job Creation and Skills Development" was launched. 4. I n this context, understanding thefunctioning of labor markets, key constraints and implicationsfor policy makers is a priority. Based on updated information, the scope o f this report i s to focus on labor market outcomes, key barriers to job creation within labor markets, and issues related to the skills development. During the discussion of the 1 Good job is defined by a job that provides sufficient earning to leave a person out of poverty as well as ensure basic social protection. See: The Youth and Prospectsfor Quality Jobs inGhana; WE%,December2005. preliminary results o f the study, the authorities (Ministry o f Manpower, Youth and Employment; Ministry o f Finance and Economic Planning; and Ministry o f Education and Sports), confirmed that this report: (i)i s closely linked to the Government's own policy agenda for addressing job creation and skills development; (ii) will bring new knowledge regarding the functioning and outcomes o f labor markets in Ghana (both supply and demand sides), and policy options for consideration; and (iii) lay the groundwork for will Ghana to reach a Middle-Income Country (MIC) status. Inresponse to the Government's request, the preparation o f the study has been carried out on a consultative basis, and interimsteps have beenusedto share and discuss early analytical findings. Framework and Structure of the Report 5. This report follows a multi-sector approach to analyze and interpret the employmentsituation. The report recognizes that the challenge o fjob creation can only be addressed through a comprehensive strategy that has both economic and social elements and that addresses both labor supply and labor demand. The framework for the analysis considers the implications for employment o f Ghana's macroeconomic performance, the investment climate, labor market policies and institutions, safety net for workers, and education and skills development opportunities. By reviewing all the factors, the report aims to identify policy priorities that can promote creation o f more and betterjobs. 6. The analysis involves an up-to-date empirical assessment of Ghana's labor market trends and apolicy discussion thatfocuses specifically on labor market and skills development. The report is organized as follows. Chapter 2 sets the stage by focusing on major determinants o f labor demand (macroeconomic environment and the investment climate) and relying heavily on the findings o f the recent C E M that has been prepared in tandem with this report. Chapter 3 looks at the structure and trends o f labor markets. Chapter 4 reviews factors within the labor market that affect employment - labor market policies, institutions and programs. Finally, Chapter 5 reviews education and skills development opportunities andthe link to job creation. 2 CHAPTER 2. THE CONTEXT FORJOB CREATION Creating afavorable environmentfor growth and private sector development is key to promotingjob creation. This chapter looks at the contextsurrounding the labor market, focusing on major determinantsof labor demand-the macroeconomicenvironmentand the investmentclimate. Since the economic environment is the most critical determinant of labor market outcomes, this chapter: (0 highlights the relationship between output and employment; and (ii) reviews the extent the investment climate remains a constraint on investment,growth andjob creation. Overall, it shows that economicgrowth has been strong and has prompted substantialjob creation, which has been accompanied by a decrease in poverty. The shift in the structure of output increased demandfor labor, creating new income-earning opportunities. Most sectors with strong growth in output had a strong growth in employment, and composition of the labor force shifted from agriculture toward other sectors (e.g., financial services, construction and transportkommunication). This confirms, as discussedin the next chapter, thatpeople are moving outside agriculture into self-employment and informal wage employment, leading to the growth of informal urban employment. At the same time, some of the gains in wages andpoverty reduction may be slightly lower than the data suggest due to measurement issues, and it is notfully clearfrom the data that strong micro-economic foundations are there to sustain continuing high rates of growth in thefutures. Looking forward, further changes in the structure of employment are needed in order topromote continued a pro-poor growth. This will require among others strengthening infrastructure, increasing firm productivity, and diversifying outputs. Moreover, diversified exports may need to play an increasing role as an engine of growth andjob creation. The investment climate, while good, leaves room for improvement. Although firm surveys show that labor market regulations and workers skills do not currently impede firm development, reflecting the fact that most of the firms are in low productivity activities and are small in size, these are issues that Ghana will have to address ifit wishes to become a middle income country by 2015. MACROECONOMIC ENVIRONMENT, WAGEGAINSANDPOVERTY REDUCTION 7. During the last 15 years, Ghana has had one of the strongest growth rates amongst Sub-Saharan countries. The Government of Ghana's (GOG) substantial efforts to achieve macro economic stability have clearly paid off. Ghana is among the strongest policy performers among low-income African countries (see Bogetic et al., 2007, first volume ofthis CEM, for an analysis of Ghana's growthperformance) along with Tanzania, Uganda, and some other West African countries. Growth averaged 4 percent per year in the 1990s,increasedto 5 percent between2001and 2005, and surpassed6 percent in2006- 07. With a population growth rate of slightly less than 2.5%, this translates into a GDP per capita growth of up to 4%. 8. Apartfrom gains in export sectors in theperiodfrom 2001 to 2005, growth was driven largely by public investmentand domestic consumptionfunded inpart byforeign aid and remittances. Table 1 from the CEMprovides data on growth and a decomposition into contributions from capital, labor force, education, and total factor productivity (TFP). 3 Inthe 1990s, there was no increase inTFP. Growth came mainly from population growth (people must make a living) and capital accumulation financed through the budget thanks inpartto higher foreign aid. There were also some limitedgains through abetter educated workforce. In other words, with good macroeconomic and fiscal management, Ghana attracted foreign aid, which helped boost public investment and generated growth in per capita incomes apart from the growth that would naturally follow from population growth. For the period 2001-2005 however, there was in addition an increase in TFP, which contributed to almost athird ofthe growth. Table 1: Growth rates and sources of growth from a Solow-Denison Model (%) Source: See CEM 2007 Table 1.4.WDI, GhanaLive DataBase (LDB), authors' calculation. Notes: (a) Sensitivity analysis is based on various scenarios, including the initial capital output ratio at 1.3, estimated using the perpetual inventory method as in Nehruand Dhareshwar (1993), depreciation rate at 0.06, and a at 0.35 and 0.45. The results are basically the same: the factor accumulation accounted for most of the growth over the 15 years while TFP contribution to growth was more pronounced in the last 5 years. b a is the parameter that measures the importance of physical capital inCobb-Douglas production function. 9. A large part of the recent increase in TFP is likely to be related to gains in agricultural productivity, especially in the cocoa sector. Agriculture accounts for about 40 percent of GDP and the sector as a whole had a growth rate o f 5.7 percent over the period 2001-2005. About 30 percent of that growth was due to cocoa alone, even though cocoa represents a much smaller share of total agricultural output. It can therefore be inferredthat the cocoaboom addedapproximately 0.7 percentagepoint to the overall GDP growth. These gains are mostly captured through TFP in Table 1, since they cannot be accounted for by capital accumulation, increases in the overall labor force, or better schooling. Inaddition to the impact of the cocoa boom, growth also increasedinpart due to a construction boom linked to public investment as well as remittances, which are important in Ghana and increasing. Finally growth increased further to 6.3-6.4 percent in 2006-2007. In 2007, high growth sectors included mining (up 30 percent), finance, insurance, and real estate (up 15 percent) and construction (up 11 percent). Output however declined inmanufacturing (down 2.3 percent) and utilities (down 22 percent). A fiscal policy stimulus led to Government services growing as well, yielding an additional one percent of real GDP growth. Thus over the last 15 years apart from recent export- oriented mining and cocoa booms, long-term growth has been driven mostly by public investment and domestic consumption themselves financed in part externally through foreign aid andremittances. 4 10. Economic growth has led to massivepoverty reduction (see Volume I11- Chapter 2 of the CEM by Coulombe and Wodon, 2007). The share of the population inpoverty fell from 51.7 percent in 1991-02 to 39.5 percent in 1998-09, and 28.5 percent in2005-06. It i s expected that Ghana will meet the Millennium Development Goal of reducing poverty by half, with 1990 set as the baseline, well ahead of the target date of 2015. However, the reduction in poverty has been accompanied by an increase in inequality, and it is notable that the pace of poverty reduction has been weaker in the northern regions, which were already poorer inthe 1990s. Inaddition, as noted by Coulombe and Wodon (2007), there is a possibility that the decrease in poverty and the increase in consumption and wages observed in the GLSS5 data may have been overestimated due to a weaker increase than expected in the official consumer price data for urban areas outside of Accra which tends to result in relatively low poverty lines for these areas, and thereby lower poverty as well and higher wages in real terms. On the basis of the information available from the Consumer Price Index, consumption increased in real terms by about 35.5 percent according to the data available inthe household surveys, while the cumulative growth rate inper capita GDP was only at 18.4 percent over the same period. The difference between these two estimates can be explained through the GDP deflator, which grew more rapidly than the CPI, among others because the prices of some imports (such as cocoa) increased. Still, even if we assume that the CPI could have been underestimated in some areas, it i s clear that substantialpoverty reductiontook place inGhanabetween 1998/99 and 2005/06. 11. The labor market was able to absorb new entrants, and gains in wages were important, but asfor the case of poverty, may be overestimated due to data limitations. Labor force participation has remained stable over time at 89 percent (population aged 25- 64) over time, and unemployment has decreased slightly from 2.7 to 2.3 percent. Gains in standards of living, including wages, have come from three main sources: (i)lower dependency ratios within households which does not affect wages, but do enable households to spread their income on a smaller number of household members, which contributes to reducing poverty; (ii) shifts away from lower payingjobs in agriculture to better payingjobs in services, which are also related to migrationfrom rural to urban areas; and (iii)higher earnings inmostjob categories thanks to an overall increase inwage rates. The gain in average earnings between 1991/92 and 2005/06 were at 77 percent, which is roughly inline with household gains inconsumption per equivalent adult (64 percent) over the 15-year period (see Box 1). Part of the gains in earnings were related to an increase in average hours worked per person was at about 14 percent. The average gain in earnings per hour was at 35 percent (since we are dealing with ratios, we neednot have the average gains inwages per hour equal to the average gains inearnings less the average increase in hours worked). However, it should benoted that the divergence betweenthe GDP deflator and the CPI mentioned above also affects the link between growth as observed in the national Accounts and earning trends since the CPI is used to deflate earnings while the GDP deflator is used for growth analysis. Thus gains in earnings may have been overestimated somewhat (or alternatively the pace of growth inthe country may have been underestimated). 5 Table 2: Ghana: Sectoral sources of economic growth, 1990-2006 PA) Share of GDP Growth rate 1991 1998 2005 1991-2005 GDP 100.0% 100.0% 100.0% Agriculture 38.6% 35.2% 34.5% -0.8% Mining/Quarrying 4.9% 5.5% 4.9% 0.0% Manufacturing 9.5% 8.8% 8.6% -0.8% Utilities 2.3% 2.4% 2.5% 0.8% Construction 6.8% 7.4% 8.1% 1.3% Trading 4.1% 4.4% 4.8% 1.2% TransporKommunication 5.3% 6.4% 6.9% 2.0% Financial Services. 4.0% 4.1% 4.2% 0.4% Community& Other Services 24.8% 25.9% 25.5% 0.2% Share ofLabor Force Labor Force 100.0% 100.0% 100.0% Agriculture 60.3% 5 1.4% 53.2% -0.9% Mining/Quarrying 0.5% 0.9% 0.8% 3.2% Manufacturing 8.8% 12.3% 11.8% 2.1% Utilities 0.1% 0.2% 0.2% 2.7% Construction 1.2% 1.7% 1.9% 3.4% Trading 16.9% 20.0% 18.6% 0.7% TranspodCommunication 1.9% 2.6% 3.1% 3.5% FinancialServices 0.5% 0.9% 1.2% 6.1% Community& Other Services 9.7% 10.1% 9.2% -0.4% Average Productivity(GDP based) All 9.4 12.2 11.6 1.5% Agriculture 6.0 8.3 7.5 1.6% Mining/Quarrying 90.7 76.9 72.9 -1.5% Manufacturing 10.2 8.7 8.5 -1.3% Utilities 145.4 139.5 138.0 -0.4% Construction 52.6 54.2 48.6 -0.6% Trading 2.3 2.7 3.O 2.0% TranspodCommunication 25.6 30.2 26.0 0.1% Financial Services 70.1 55.4 40.0 -3.9% Community& Other Services 24.0 31.3 32.3 2.2% Average Productivity(Consumptionbased) All 6.0 7.8 10.1 3.8% Agriculture 4.7 5.6 7.4 3.3% Mining/Quarrying 9.1 13.1 12.2 2.1% Manufacturing 7.0 9.1 11.6 3.7% Utilities 13.7 11.2 20.1 2.8% Construction 8.0 8.5 12.1 3.0% Trading 7.4 9.7 12.7 4.0% TranspodCommunication 10.5 13.1 13.9 2.0% FinancialServices 11.3 17.4 19.5 4.0% Community& Other Services 9.2 10.6 15.3 3.7% Poverty Headcount A11 49.5% 35.2% 25.8% -4.5% Agriculture 61.5% 49.4% 38.4% -3.3% Mining/Quarrying 34.2% 7.6% 4.7% -13.2% Manufacturing 35.5% 25.9% 17.5% -4.9% Utilities 18.2% 11.6% 0.0% NA Construction 29.7% 20.9% 10.2% -7.3% Trading 33.8% 21.2% 11.O% -7.7% Transport'Communication 20.7% 8.3% 11.7% -4.0% FinancialServices 12.5% 5.8% 5.3% -5.9% Community& Other Services 26.3% 17.0% 6.7% -9.3% Average productivity=real GDPLF. Source: Staffestimates. 6 12. Structural changes in the composition of output led to shifts in labor shares creating better income-earning opportunities and contributing to poverty reduction. Table 2 provides data on the share o f GDP accounted for by different sectors, the share o f the labor force working inthese sectors, the average productivity o f workers computed as the total GDP or household consumption related to each sector (in the case o f consumption, this i s done on the basis o f the main occupation o f the household head), and finally the share o f the population living in poverty in each o f the sectors. Overall the degree o f "structural transformation" among the large sectors o f the economy has been limited in Ghana. The service sector has maintained an average growth rate o f 6 percent, compared to a total economic growth rate o f 4.7 percent (Table 2). More specifically, the following findings are observed: 0 The share o f agriculture in GDP i s decreasing over time, to the benefit o f the other sectors, and especially construction, trading, and transport and communications. Similarly, the share o f the workforce in agriculture is falling, while it i s increasing inall the other sectors, with the exception of community and other services which include public administration where employment in relative terms has been reduced. Within agriculture, although the increase inthe price o f cocoa had a direct impact on trade, its impact on the labor market was limited: the share o f workers in cocoa production remainedstable during 1991-2005 (6 percent), while the share of non-cocoa agriculture workers decreased (from 54 percent to 47 percent). 0 Figure l a and lb provide data on the growth in labor force by sector, the GDP output growth by sector, and the level o f worker productivity by sector. The size o f the bubbles on the Figures i s proportional to the GDP shares o f the various sectors. Two findings stand out. As expected, the growth in the labor force i s highest in those sectors where the growth inoutput i s highest as well (Figure la). Inaddition, there is a clear indication that workers are shifting to those sectors where productivity, as measures according to GDP or consumption per worker is higher (Figure 1b). Said differently, the inter-sectoral labor reallocation inGhana between 1991 and 2005 saw a sectoral shift in composition o f labor force from the agricultural sector toward among others the financial services, transport and communication, and construction sectors (Chapter 3, Table 4), which has helped increase the earnings o f the workers who have made this transition. 0 While the data confirms that people move to better paying jobs, at the same time, wage employment in the formal sector i s not growing fast enough to absorb the new influx of workers especially inurban areas, so that many workers are moving outside agriculture into self-employment and informal wage employment, leading to growth in informal urban employment. This transition i s expected to continue. Indeed, agriculture still accounts for about a quarter o f Ghana's GDP and almost halfo fthe labor force, with these shares expected to decrease inthe future. 0 Whenmeasured interms o f changes inconsumption levels, productivity per worker has increased substantially over time in all sectors o f the economy. However, as mentioned earlier the increase in consumption in real terms has been significantly higher than the increase in GDP, due to the fact that the CPI has increased less 7 rapidly than the GDP deflator. When looking at productivity in terms of GDP per worker, the data suggest that for some sectors there has beena gain inproductivity, while for others, there has been a loss. It is likely that the "true" changes in productivity are somewhere betweenthe changes obtained with the two indicators, GDP and consumption. 0 According to the consumption-baseddata that is usedto measure trends inpoverty, poverty was reduced substantially in all of the sectors, although in comparative terms, the reduction in poverty was slower in agriculture than for the rest of the sectors. As shown by Coulombe and Wodon (2007), the price and productivity gains in cocoa led to a decrease inpoverty among these producers from 62 percent in 1991to 21 in2005, while the decreasewas muchsmaller for other farmers. As a result, some northern regions where agriculture remains by far the main occupation are lagging behind. 8% - 7% - 1 6% - Financialservices I 5 % - Construction TransporWCom. 4 1 * Utilities 2% Manufacturing 0% -1 0 . 0 0 0% i -2% - 3 % J Productivity of labor force (consumption-based) I Figure 1.b: Growth Rate of Output and Labor Force by Sector (%) 10% FinancialServices 9% ~ 8% Construction Transport & Comnications 0 Manufacturing Utilities :;I Trading Cornunity & other services Agriculture ~ 3.0% 3.5% 4.0% 4.5% 5.0% 5.5% 6.0% 6.5% 7.0% 7.5% Annual output growth rate (consumptlon based) 8 Box 1: Accounting for the increase in welfare levels in Ghana The average consumptionper equivalent adult increasedin Ghana by about 64 percent inreal terms between 1991192 and 2005106, which i s very large. Various decomposition methods can be used to provide an account of what were the drivers for this change. Such decompositions typically include the following factors: a) sectoral shifts observed inthe labor force towards sectors with higher earnings; b) higher number of hours worked; c) demographic transition factors such that earnings are shared by a smaller number of householdmembers; d) potential over- or under-estimationof income or consumption gains (due inGhanato a potential underestimationof inflation in parts of the country); and e) productivity gains per worker within sectors. Ratherthan providing detailedresults from such a decomposition(these results are complex because ofthe interactionsbetweenthe different factors involved) we provide below an intuitive discussionofthe key factors that are likely to have had a large impact on householdwelfare. Sectoral shijts between sectors (gain of about 10percent in welfare). As shown intable 2, the share of the labor force inagriculture decreasedby 7 percentagepoints between 1991192and 2005/06,from 60 percentto 5 3 percent. In addition, the average productivity (consumption-based) of a worker in agriculture was only 7.4 million cedis per year in 200512006, versus an average of about twice that (13 million cedis) per year in the other sectors. In previous years as well, the productivity per worker is about twice as high outside of agriculture than within agriculture. If7 percent ofthe workers changejob and earn twice as much as they did previously in doing so, this generates an overall gain in welfare of about 7 percent. As workers try to move to the highestpaying sectors, the gains are likely to be slightly larger. Changein number of hours worked (gain of about 15percent in welfare). The average worker increased hidher working hours by about 14percentbetweenthe two years. On average, the increase inworking hours was larger outside of agriculture than in agriculture (where the increase was at about 10 percent). Thus, the impact of higher working hours on total earnings must have been at least 14 percent. Change in demographicstructure, laborforce participation and unemployment (gain of about I O percent in welfare). Consumption and poverty measures are based on total household consumption divided by the number of equivalent adults inthe household. InGhana, the number of equivalent adults decreasedbetween 1991192 and 2005106 from 4.17 to 3.90, a drop of about 6.5 percent. Similarly, the number of individuals in the average householddecreasedfrom 5.59 to 5.13 (a drop of 8.5 percent). Another useful statistic is that the dependency ratio based on the age of various members of the households decreased from 2.72 to 2.38 (a decrease of 12.5 percent). Overall, on the basis of these statistics, changes in demographic structure may have led to an increase of about 10 percent in welfare. By contrast labor force participation rates did not change much, and unemploymentrates decreasedmarginally from a low base. Potential over-estimation in gains of welfare (gain of 0 to 18 percent in welfare or about 9 percent in average). As discussed previously, there has been a divergence between the GDP deflator and the CPI in Ghana over the last few years. The increase in GDP per capita between 1998199 and 2005106 was at about 18 percent, while the increase in consumptionwas twice as large. It may be that part of the difference is due to an underestimation in the CPI in some areas of the country (non-Accra urban areas specifically). Thus, some of the divergence inthe GDP deflator and the CPI may reflect a measurement error that leads to higher "observed" welfare thanthe actual changes inwell-being for the population. Gains in productivity within sectors (gain of about dopercent in welfare). As shown intable 2, even when using productivity as measured by GDP per worker, there was an increase inproductivity in the two largest sectors of the economy interms of their labor force share (agriculture, with a 25 percent increase in GDP per worker, andtrading with a 32 percent increase). Despitethe loss inproductivity accordingto this measure in some of the other sectors, overall productivity gains may have generatedgains inwelfare. 13. Part of the gains in income orproductivity of workers is due to an increase in the number of working hours of those workers. There has been an increase in working hours per worker o f about 14 percent between 1991-92 and 2005-06. This increase is observed for all sectors, although there are some differences between sectors in terms o f the 9 magnitude of the increase. Once again, this increase in working hours suggests that the growth of the economy has generated substantial employment, especially in the trading sector where the increase in working hours per workers was one o f the highest. But this also means that interms o f wage earnings per hour worked, the gains have been lower than one might be led to believe if one were not observing the increase in working hours over the period. 14. Since exports have not diversified, the economy remains vulnerable to commodity price shocks. Reflecting the Government growth strategy, Ghana's export contribution to total GDP has been lower than that in rapidly growing comparator economies (Bostwana, Mauritius, Mozambique, Malaysia): between 1990-2005, exports in Ghana contributed to only 6 percent o f GDP growth (compared to over 40 percent in comparator economies), government spending and investment accounted for 40 percent o f growth, and the remainder was due to private consumption. Moreover, as shown in the recent CEM, Ghana's exports by commodity have not diversified significantly, and cocoa, gold, and timber still account for about 70 percent o f total commodity exports. As a result, Ghana's economy remains vulnerable to commodity price shocks. The growth success stories in Africa (Mauritius and Bostwana) testify that for sustained, long-term growth, the strength of economic policies rest on fiscal soundness, transparency, and export orientation o f the economy. 15. Microeconomic evidence suggests that Ghanaian firms lag behind the most productive African firms, indicating scope for improvements. According to Teal and others (2006), based on microeconomic firm-level data, productivity (values for value- added per worker) for the median manufacturing firm in Ghana was approximately one- tenth o f the productivity for the median firm in South Africa. Firms in Ghana compared more favorably with productivity in other countries; in particular, the median Ghanaian firm was slightly more productive than the median Tanzanian firm, approximately half as productive as the median Kenyan firm, and approximately one-third as productive as the median Nigerianfirm. According to the 2007 Investment Climate Assessment (see Clarke 2007), productivity appears to be relatively low in Ghana. The median firm in Ghana produces far less output per worker than firms in more successful manufacturing economies elsewhere in Sub-Saharan Africa or inEast Asia. CHARACTERISTICSOF FIRMSAND INVESTMENT CLIMATE INGHANA 16. Low Productivity offirms reflects the characteristics of the enterprises in Ghana. As discussed in the 2007 CEM, firms in Ghana are less productive than firms in the comparator countries or other low-income countries in Sub-Saharan Africa. This i s mainly because firms in Ghana are (i)less capital and skill intensive (less machinery and equipment per worker); (ii) relatively small in size: the median manufacturing firm has only about 10 employees in Ghana, compared to 60 employees in Swaziland, and 50 employees in Kenya; and (iii) likely to export: only one-quarter o f manufacturing less Small, Medium and Large Enterprises (MLEs) in Ghana exports, compared to more than half in most o f the comparator countries (see Figure 2). Labor productivity appears to be particularly low among small firms-small firms from Ghana are less productive and less capital intensive than small firms even in other low income countries in Sub-Saharan 10 Africa, while large firms are relatively productive and relatively capital intensive. The small size o f firms has been noted in other studies, along with the substantial increase in their number, which potentially explains the continuation o f the job creation in the informal sector (Teal and others, 2006). Figure 2: Labor productivity is particularly low among small enterprises in Ghana. Small Large Senegal Senegal Kenya Mali Mali Tanzania Zambia Mauritania Malawi Kenya Mauritania Nigeria Benin Eritrea Eritrea Ghana Nigeria Zambia Mozambique Uganda Tanzania Burundi Congo, DR Benin Rwanda Malawi Uganda Congo, DR Burundi Madagascar Madagascar Rwanda Ghana Ethiopia Ethiopia Mozambique $0 $5,000 $10,000 $15,000 $0 $5,000 $10,000 $15,000 Value-added per worker Value-added per worker (US$) Source:Enterprise Surveys; See CEM 2007. Note: Comparisons only include manufacturingenterprises. Countries where there were fewer than 10 large enterprises and middle incomecountries are excludedfrom the graph. Definitions are providedin footnote to previousfigure. 17. According to Enterprise Surveys, impediments tofirm development exist outside of the labor market. The overall investment climate in Ghana, as measured by the World Bank's enterprise investmentclimate survey (2007), i s relatively good.3 This is confirmed by the 2008 Doing Business Report, which lists Ghana among the top ten reformers. Inthe report, Ghana scores relatively well on the basic indicators such as law and order, taxes, trading across border, and enforcement o f contracts, but less well on issues affecting starting a business, getting a license and hiring workers. Based on the 2007 Investment Climate Survey, the two main areas o f investment climate that deserve particular attention and stand out as particular concerns for small, medium and large enterprise (SMLE) are power (electricity) and access to finance, especially for small and medium size firms (Figure 3): close to nine out o f ten SMLE managers and over half o f the microenterprise managers rated electricity as a serious problem (Le., a major or very constraint) and about seven out of ten SMLE and microenterprise managers rated access to finance as a serious problem. Far fewer enterprise managers were concerned about other constraints, such as 3 An Investment Climate Firm Survey was conducted in Ghana in May through July 2007. The survey covered firm performance and productivity, manager's perceptions about the main obstacles to doing business in Ghana, and objective indicators of the investment climate. Results from the survey will be discussedingreater detail inthe forthcomingInvestment Climate Assessment, which is ubderpreparation. 11 access to land, tax rates, transportation, macroeconomic instability, skills, tax administration, labor regulation, courts, and political instability. Figure 3: Managers' responseson biggest investment climate constraints, MLEsand microenterprises Source: World Bank EnterpriseSurveys. 18. Characteristicsofflrms, explain to a large extent the results of the ICA. The low level o f firms' concern about some issues-such as labor regulations and worker skills- might reflect the small size, low capital intensity and low productivity observed among firms in Ghana, rather than indicate that labor regulations and skills are not a problem for the country overall. Iffirms are mostly engaged inlow productivity activities, they are less likely to demand skilled workers. In addition, since small firms are less unionized, they are less affected by labor regulations. These issues will also be discussed in greater detail inthe following Chapters. SUMMARYOF FINDINGS 19. Growth has generated employment and wage gains which have in turn led to poverty reduction, but there are challenges to sustaining these gains in the future. Several challenges can be highlighted. First, much o f the growth inemployment has taken place in the informal sector, which can be helpful for reducing poverty, but it may be less attractive to more skilled labor, particularly since it will not provide the same security and benefits as the formal sector. Second there have been substantial gains in earnings, consumption, and poverty reduction, but it could be that some o f these gains have been overestimated due to a comparatively low increase in the Consumer Price Index as compared to the GDP deflator, especially in urban areas outside o f Accra. Third, part o f the gains in earnings have been obtained through an increase in working hours, which could potentially be increased more inthe future, but only up to point since there is a limit to the number o f hours that workers can work in a day. Fourth, the pressure on the labor 12 market to absorb a large number o f new entrants will remain in the future due to the demographic transition phased by the country. Fifth, at the macroeconomic level, the contribution to growth o f the improvement in factor productivity has been limited, and probably due over the last few years inlarge part to the boom o fthe cocoa sector, a success that may not be easily reproduced in future years. Finally, although there have been sectoral shifts in the labor force to better paying jobs, more than half o f the labor force remains employed in agriculture, in most cases in low productivity jobs, so that the transformation o f the economy remains limited. 20. These results suggest that efforts need to be made to upgrade the skills of the labor force for Ghana's economy to acquire the features observed in middle-income countries. The macroeconomic analysis o f the sources o f growth in Ghana (see table 1) suggest that: (i) improvements inthe education o f the labor force have contributed only for about 11-12 percent o f the growth observed over the last fifteen years; (ii)total factor productivity has also not risen much, with the exception perhaps of the cocoa boom between 2001 and 2005; (iii) although real wages have increased, it i s likely that part o f those gains will be reversed due to the increase inthe cost o f living (due to higher food and oil prices); (iv) some o f the gains in earnings have been obtained by working more rather than by obtaining higher hourly wages. While the increase inworking hours is positive since it suggests better employment opportunities, it cannot be repeated indefinitely; and (v) a large share o f the population remains engaged in low productivity unskilled work. Therefore, for a country that aspires to middle-income status, a key priority should be to increase the skills o f the labor force, and ensure that interventions to improve education and training are balanced with interventions in economic and labor market policies to promote a favorable climate for private investment andjob creation. Inthis context, labor market policies, institutions and programs are discussed in Chapter 4, and the issue o f ensuring that the education and training systems prepare and support workers to respond to the skills needs o f the economy i s discussed inChapter 5 of the report. 21. A business climate that is favorable for investments will also be favorable for employment, including employment in the formal sector. Cross-country comparisons demonstrate that reforms that make business development easier allow large and labor- intensive firms to grow faster and create more jobs in urban areas (World Bank WDR 2005). Therefore a sustained employment growth requires a dynamic private sector and a friendly investment climate. Ghana has made clear progress in improving the laws and regulations that define the business climate. However, as discussed in the CEM, there are serious concerns in respect to electricity and access to finance by small and medium-size enterprises, which need to be addressed. Moreover, certain business regulations need to be improved, such as procedures for formal business start ups, which are still cumbersome and push firms to informality. Currently, according to the recent investment climate, the main obstacles are not labor market policies or lack o f skills, but as other constraints are addressed, labor market regulations and inadequate skilled labor may become important (see Chapters 4 and 5). 22. Exports need toplay an increasing role as an engine of growth andjob creation. The model o f economic growth that has focused primarily on domestic demand and investmentis not likely to provide a sustainable source o f employment growth. At present, 13 the increases in the demand for labor come mainly from the domestic market, where self- employment activities have expanded far more rapidly than wage employment (see Chapter 3). As discussed in the CEM, exports need to play a larger role as an engine o f growth, and export diversification need to be promoted. Currently, agriculture exports are the primary source of foreign exchange accounting for over half o f export receipts. Government GPRS I1i s focused on an agriculture-led strategy which seeks to diversify the economy's structure from dependence on cocoa to cereals and other cash crops for export markets. In this context, the review o f the sector in the recent C E M suggests substantial scope to increase physical yields, opportunities to substitute low-value production for higher value crops, including those for export. However, to promote job creation as well as productivity gains, future macroeconomic growth policies need to improve the competitiveness o f the manufacturing sector and enable firms to enter and grow in the export market. 14 CHAPTER3. LABORMARKET OUTCOMES: BASIC STATICTICS As discussedin theprevious chapter, during the last 15years, Ghana has had one of the strongestgrowth rates amongstSub-Saharan countries and an outstandingperformance in terms of poverty reduction. Sincethe vast majority of people dependon their labor as theprimary source of income, the quantity and quality of employmentplay a central role in the translation of growth into poverty reduction. This chapter reviews labor market trends in Ghana using simplestatistical analysis (regression analysis is used in the next chapter). The analysis focuses on data spanning the last 15 years to answer the following questions: How has growth been translated in terms of job creation?How has the distribution of employment across various sectors changed? Who works and what type of skills do workers possess? What is the quality of the jobs created in both the public and private sectors? What share of workers in various sectors belongs to poor households? Which groups among the workingpopulation are relatively disadvantaged, as measured by labor market status, earnings, education, skills and otherjob attributes? In answering these questions, the chapter shows that jobs are being created, although more so in the informal than in other sectors. Although wages are increasing, the labor market remains somewhat segmented between high wage, high benefit public sector jobs, and lower wage, lower benefit jobs in the informal sectors and among the self- employed. Unemployment is primarily found among the better-educated urban youth. Looking forward, job generation needs to be promoted in higher productivity sectors, public sector pay policy may need to be reviewed, and given the positive impact of education on earnings, investmentsmust be made in the education level and skills of the workingpopulation (this is discussedin chapter 5). 23. In order to present basic trends on Ghana's labor market outcomes, we rely in this chapter on simple statistics, leavingfor later the analysis of the determinants of labor outcomes. This chapter provides basic statistics on jobs and earnings, as well as some data on time use patterns. We start with some information on dependency ratios (section l), and follow with sections on labor outcomes (section 2), with a focus on the adult population aged 25 to 64. We then present data on total individual earnings (Section 3), and finally on hours worked and hourly earnings (Section 4). The analysis is based on the last three rounds o f the GLSS surveys for 1991/92, 1998/99 and 2005/06. The objective is to provide a simple picture of the labor markets at three points in time, and assess how the labor market has evolved. An analysis o f the determinants o f the occupational choices o f workers as well as o f their wages i s provided inchapter 5. 24. There is an important caveat related to the comparability of the surveys that needs to be emphasized- we do not consider here the issue of youth employment. Most of the statistics presented inthe next section o f this chapter are based on GLSS surveys for 1991/92, 1988/99, and 2005/06. Though the surveys are comparable for most questions, there i s one major difference in the design o f the 1998/99 questionnaire that needs to be taken into account. In the 1998/99 questionnaire, once a child or youth was enrolled in 15 school, no questions were asked about his labor activities; however, this was not the case inthe 1991/92 and 2005/06 surveys. This means that when considering the whole adult population aged 15 to 64, only the data from the 1991/92 and 2005/2006 surveys are comparable, as estimates based on the 1998/99 surveys do not account properly for the labor activities of youths aged 15 to 24 (or children below the age of 15). By contrast, when considering the population aged 25 to 64, the three surveys are to a large extent comparable, although in some cases, we still find some divergence in the results obtained with the 1998199 survey as compared to the other two surveys. Inthis chapter, we report separately on the trends for the 25-64 age group using all three surveys, and the issue of youth employment for the 15-24 age group using the first and third surveys (1991 and 2005) only. DEMOGRAPHICTRENDS DEPENDENCY AND RATIOS 25. Ghana's population is experiencing a decline in fertility which is resulting in changing demographicstructure and improving falling) dependency ratios. As a result of the declining fertility the age structure is changing. This will affect dependency ratios which will fall, and will create a window of opportunity, as there will be more working people relative to dependents. Annual population growth has fallen from about 4% in the early eighties to slightly less than 2.5% in2005. With this population growth rate Ghana is below the Sub-Sahara African average, but the country is still above the average for low income countries as a whole. In addition to a reduction in its population growth rate, Ghana is experiencing a demographic transition characterized by a declining fertility and improving (falling) dependencyratio (this i s the ratio of the total population to the working age population). The working age population is indeed increasing as a share of the total population (from 52 percent in 1983 to 57 percent in 2005), while the share of children (0- 14 age cohort) has decreased from 45% to 39%. The proportion of people above 65 i s increasing rapidly, but from a low base, and therefore remains low at about 4% in 2005. According to projected population pyramids (see Figure 3), this trend will continue. O f course, for households to benefit from the falling dependency ratio the economy must create sufficient jobs so that both the existing active population and the new cohorts who are entering the labor force are able to find work. Yet as will be documented in this chapter, this has largely been the experience in Ghana over the last 15 years, with the growth inemployment closely matching the growth inthe supply of labor among the adult population, especially inthe 25-64 age group. 26. Thefall in dependency ratios is already observed in the various rounds of the GLSS surveys. As shown inFigures 4 and 5, while the population grew for all age groups between 1991 and 2006, the largest increase was observed among the working age population, and especially among youths. This has resulted ina substantial decrease inthe dependency ratio, whether calculated by comparing the total population to the number o f people aged 15 and 64, or to the number of people aged 25-64. For example, when using the population between the ages of 25 and 64 as the denominator, the dependency ratio decreased from 2.72 in 1991/92 to 2.38 in2005/06. This is likely to have been a key factor inthe improvement of the consumptionpercapita indicators usedto measurepoverty. 16 27. However, while Ghana is benefiting from a demographic transition, the urban population is increasing rapidly, which implies a rapid rise in the urban labor supply, especially amongyouth. While the youngest age cohort (0-14) is growing at the slowest rate (about 1% per year), the working age population (15-64 years) is growing at a much higher of 3% per year, and the rate is higher for youth (15-24 years). Another important trend is the rapid rate of urbanization, with the share of the urban population increasing from 32% in 1983 to 48% in 2005. Together, the high rate of population growth and the rapidurbanization have yielded a large increaseinnewjob seekers, especially incities, and especially among youth. We will document in this chapter the fact that poverty has decreased much less among the unemployed than among other groups over time. This suggests that interventions to help the unemployed, and especially young workers with limited experience that may have difficulty finding good jobs, will be important if the aspirations of the new cohorts entering the labor force are to be addressed. Figure 4: Population Pyramids 2000,2025,2050 *r#-%+g.@,E Ghana: 2888 1.0 1:5 1:0 0:5 0:O 0:O 0:5 1:0 1:5 2.0 Population (in m i l l i o n s ) ource: U.S.Census Bureau, International Data Base. Ghana: 2825 1 E I"[ 2.0 1.5 1.0 0.5 0.0 0.0 0.5 1.0 1.5 2.0 Population (in m i l l i o n s ) ;ource: U.S. Census Bureau, I n t e r n a t i o n a l Data Base. IFEMALE Source 1:5 1:0 0 : 5 0 : O 0 : O 0:5 1:O 115 2.0 Population (in m i l l i o n s ) U.S. Census Bureau, I n t e r n a t i o n a l Data Base. 17 Figure 5.a: Populationby age group in GLSS surveys, 1991-2006 Population by age group 10.0 9.0 8.7 8.0 8.0 7.4 7.2 7.0 6.0 5.9 5.0 4.6 4.0 3.0 2.0 1.o 0.0 2006 1998/99 1991/92 0 0 to 14 w 15 to 24 25 to 59 60 and over Source: Basedon Coulombe and Wodon (2007), usingGLSS surveys. Figure 5.b: DependencyRatios in GLSS surveys, 1991-2006 iatio of population to lumber of working age adults 2.90 2.12 2.70 2.56 2.50 2.38 2.30 2.10 1.90 1.70 1.50 2006 1998/99 1991/92 Dependencyratio(adults 15-59) Dependencyratio (adults20-59) Source: Basedon Coulombe andWodon (2007), usingGLSS surveys. 18 LABOR OUTCOMES AMONG WORKERS AGED 25-64 28. Although labor force participation rates are high in Ghana, the data need to be interpreted with caution. There is substantial debate whether labor force participation is being measured accurately in a Sub-Sahara African context, especially for work in rural areas and particularly for women. InGhana, the data suggeststhat labor force participation i s very high among the adult population aged 25-64 (more than 90% for men, and more than 85% for women) and has remained stable over time. Stylized facts suggest that (i) participation is higher inrural areas; (ii) educated adults are less likely to participate better in the labor force, both in urban and rural areas - after controlling for age and family situation; and (iii)at least in rural areas, being married increases the probability of participating in the labor force for both men and women (Table 3). However, in order to better appreciate the extent to which different individuals work, an analysis of hours worked (given below) is required. Table 3: Labor Force Participation in age group 25-64,1991-2005 I991/92 I998/99 2005/06 change Sex Male 92.2 87.4 92.2 0.0 Female 86.4 80.5 86.2 -0.2 Age 24-54 89.5 84.0 89.7 0.2 55-64 85.1 80.4 83.6 -1.5 EducationLevel N o education 88.5 81.7 88.5 0.0 Primary 88.8 86.5 90.0 1.2 Secondary lower 90.3 84.5 90.1 -0.2 Secondary Higher 86.3 82.3 84.5 -1.8 Post Secondary 87.8 87.2 88.9 1.1 UrbadRural Urban 86.5 81.3 85.0 -1.5 Rural 90.2 84.8 91.7 1.5 Region Western 93.4 86.3 89.1 -4.3 Central 92.9 93.1 94.8 1.9 Greater Accra 85.4 78.6 81.6 -3.8 Volta 90.5 89.1 91.1 0.6 Eastern 87.6 82.7 91.6 4.0 Ashanti 88.0 86.2 92.0 4.0 Brong Ahafo 94.4 88.6 93.9 -0.5 Northern 89.8 73.2 89.7 -0.1 UpperEast 90.0 58.6 76.8 -13.2 Upper West 70.6 81.7 79.8 9.2 All 88.9 83.5 88.9 0.0 Source: Based on Coulombe andWodon (2007), usingGLSS surveys. 29. The employment rate has remained high among the adult population over 25 years of age, and has decreased among children and youths in part due to better school enrollment rates. According to the GLSS surveys, the employment rate of the working age population (25-64) has remained at about 89 percent over time (with a slight drop in 1998/99). The unemployment rate has also remained stable. This implies that the labor 19 force participation rate has remained stable as well. By contrast, labor force participation has dropped among younger individuals, in part due to higher enrollment in schools (see Coulombe and Wodon, 2007). Finally, the employment rate among the elderly has also fallen consistently, which probably suggests that at least some elderly persons may not need to work anymore to maintain a decent standard of living (see Table 4). Table 4: Comparisonof employmentrates by age group, 1991-2005 I991/92 1998/99 2005/06 Employmentrate (15-64) 75.9 70.2 Employmentrate (25-65) 86.5 82.0 85.9 Child laborrate (7-14) 27.9 15.9 Elderly Employmentrate (65+) 63.0 55.0 52.4 1Jnemnlovmentrate (25-64) 2.7 2.4 2.3 Source:Basedon Coulombe and Wodon (2007), usingGLSS surveys. 30. Growth has been associated with substantialjob creation, with the increase in the number of workers matching the increase in the adult population. Between 1991 and 2005, as shown in table 5, the labor force among the population aged 25 to 64 grew by about two thirds (from 4.2 to 7.1 million people), which is comparableto the growth rate in the population itself for that age group: 2.8 million jobs were created for the population aged 25-64 over the last fifteen years (see Box 2). Table 5: Compositionof the Labor Market, 1991-2005 I991/92 I998/99 2005106 Working age population(25-64) 4,764,710 6,038,507 7,941,919 Inactive(last 7 days) 527,060 994,123 878,437 Active (last 7 days) 4,237,650 5,044,3 84 7,063,482 Unemployed 114,610 120,715 163,668 Employed 4,123,040 4,923,669 6,899,8 14 Child Population(7-14) 3,506,190 4,074,978 4,552,153 ChildLaborers 979,660 592,308 Elderly Population(65+) 580,350 806,838 1,015,964 EmployedElderly 365,730 420,228 547,062 Source: Basedon Coulombe and Wodon(2007), usingGLSS surveys. 31. Most of the inactive youth are students aged below 27 years old, with older inactives more likely to befemales who are busy at home. Based on 2005106 data, it can be shown that the large majority of inactive individuals in the labor market are younger people still at school. However, ifwe concentrate on the group aged between 25 and 64, more than two thirds are women staying at home, undertaking domestic chores. Excluding child rearing, these women spend an average of close to 25 hours a week on housekeeping (see Box 3). From a statistical point of view, we had to remove child rearing as an at-home activity, as many women considered it to be a 24 hour a dayjob, and reported it as such inthe survey. 20 Box 2: Youth employment Most of the analysis found in this report concentrates on the 25 to 64 age group, rather than the usual 15-64 group found inmost economic literature. The mainreason for this departure from convention is a problem of comparability between the different rounds of the Ghana Living Standards Survey. Inboth the 1991/92 and 2005/06 rounds, the questions related to time use and employment were asked of all participants, student or not. However, the 1998/99 questionnaire instructed interviewers to skip the employment section for individuals who were attending schools. Since a significant proportion of the students also participate inthe labor market, any 1998/99 figures for an age-group that is likely to include students would be biased. Therefore, the analysisbelow refers to the 1991/92and 2005/06 surveys only. Employment declined over the past 15 years, while rates of schooling increased. Table A1 shows that between the early 90s and 2006, the youth employment rate dropped from close to 56% to less than 40%, a decline that was much more significant inrural areas (though employmentrates inrural areas remainhigher). That decline in employment is partly explained by a significant increase in school attendance around 10 percentage points - - in both urban and rural areas. The main difference between urban and rural areas concerns unemployment. Although the youth unemployment rate increasedslightly over the 15 year period andremainsabove the rate for the population as awhole, the urbadrural gap i s huge. In 1991/92, close to 19 percent of the youth labor force in urban areas was considered unemployed, with only 1.3 percent unemployed in rural areas. Even though the rural unemployment rate rose more quickly between 1991192 and 2005/6 than the urban unemployment rate, urban areas still experienced a much higher unemployment rate (13.5 vs 3.4). Table Al.: Employment,Unemploymentand Schooling rates, age group 15-24 1991/92 2005/06 Ghana Employment rate 55.8 39.1 Unemployment rate 5.3 6.6 Schoolingrate 37.1 46.5 Urban Employment rate 29.4 26.7 Unemployment rate 18.6 13.5 Schoolingrate 43.4 50.0 Rural Employment rate 71.8 48.5 Unemployment rate 1.3 3.4 Schoolinrrrate 32.2 43.8 Around half of working youth are still doing unpaid work onfamily farms. Table A2 shows that even if unpaid work on family farm decline since 1991, half the 15-24 year old individuals remain inthis category. Comparedto the older age group (see table 8 inthe maintext), a muchhigher proportion of younger workers are self-employed. However the change in status pattern is similar between the two age-groups: a significant decline inpublic employment but a more than compensatingincrease inthe private wage earningjobs. Table A2: EmploymentDistribution by Labor Force Status (percentage) 1991/92 2005/06 Wage Public 3.0 1.1 Wage Private Formal 2.1 4.6 Wage Private Informal 2.8 5.5 Self-empl Agro -Paid 15.5 11.2 Self-empl Agro -Unpaid 55.0 49.6 Self-empl NonAgro 21.5 28.1 All 100.0 100.0 21 Box 3: Time Use Patterns Standard labor statistics do not take into account time worked at home. Work includesboth productive work (for the productionof goods and servicesfor the market) and domestic or reproductivework (productionof goods and services for household consumption). In most societies gender is a key means o f determiningthe division of labor and responsibilities among household members. Women's domestic work -particularly that relatedto biological reproduction- increasesdisproportionatelyin relationto other adults in the householdas the number of infants and children grows. Mothers of large families spend more time pregnant, breast-feeding, and caring and cooking for children and other family members. While beingmale or female is a biological fact, becominga man or a woman i s a cultural and socialprocesswhich affects labor outcomesand leadsto gender disparities in labormarkets. Detailed data are available in the GLSS on time use, and these data have been used by Coulombe and Wodon (200a) to provide an analysis of patterns of time by age and gender, and of changes over time in those patterns. The individual-level indicator used is the total amount of time spent by individuals working, whether inthe formal and informal labor market, in domestic chores or in collecting water and wood. Only a small share of the adult populationdoes not work at all (between2 and 3%). The meanworking time at the national level is 47.0 hours for the adult populationbetween25 and 64 years of age in 1991192, 46.2 hours in 1998199 and 48.4 hours in 2005/06. There are however substantial differences between adult men and women in all three years. In 1991/92, the average workload reaches 51.9 hours for women, and 40.8 hours for men. This gender gap of around 11 working hours is also observed in 1998199 (workload of 51.1 hours per week for women versus 40.4 hours for men). However,this gender gap had shrunk by 2005106 to 5.4 hours. The diminishing gender gap was due to male being more occupied by about 5 hours a week. However, a closer look at the results shows a rather more complex and interesting stories. If we compare 1998199 with 2005/06, the additional hours by male workers were essentially spent on the labor market, beingawage earner or self-employed. Inthe case of female workers, their time spent on domestic choreswas reducedby 4.8 hours (from 25.9 hours in 1998199to 21.1 hours in2005/06)but that time free of domestic choreswas compensatedby an almost equivalent number of additionalhours (4.6) on the labormarket. This labor market activity seems to be due to increasing earnings and job opportunities in both urban and rural areas. While men spend a few more hours on average in the labor market than women, the amount of time spent by women in domestic chores is much higher than for men, andthis explainsthe large differences in total working time. Among the 7-24 age group, girls also work longerhoursthan boys due to a higher burdenfrom domestic work, but the total amount of work remains reasonable (about 24.3 hours per week on average in 2005/06). There is a reductionin working hours observed for children and youth aged 7 to 24 over time between the surveys that may be related in part to an increase in school enrollment rates. As for comparisons across urban and rural areas, the results suggest differences in total working time, with on average rural individuals working five hours more per week than their urban counterparts. The main difference by location is the higher burden of domestic chores for women as well as children (boys and girls) in rural areas, especially due to the need for rural householdsto fetch water and wood. However it is worth noting that fetching water and wood has taken half as muchtime in 2005106 than in early 90s, which may be due in partto better access to infrastructureservices. It is possible to estimate the share of individuals in various groups that can be considered as "time poor", i.e. working extra long hours. In the absence of a well-establishedprocedure to compute time poverty lines, we use arbitrary multiplesof mediannumbersof hours spend on work. After having computedthe mediantime use of the 7-24,25-64 and over 64 age groups, a first line was set at 1.5 times the median. A secondtime poverty line was at twice those medians. The lower poverty line was set at 25.5 hoursfor the 7 to 24 year old, 72 hours for the adults (25-64) and at 37.5 for the elderly. The higher time poverty lines were set at 34, 96 and 50 hours respectively. Some 17 percentof adults aged25-64 were time poor in 2005106. This is an increaseof almost 4 percentagepoints over 1991/92, and a lower increase o f 0.6 point over 1998/99. We also find that a much higher percentage of women (21.1 percent) are time poor than men (12.3 percent). Although the gender gap in time poverty seems to have declined slightly for this population, it remains large, with still almost twice as many women who were time poor in 2005/06 than men. The data also suggest that time poverty is higher for adults aged 25-64 in rural than in urban areas. Similarly, for young persons aged 7 to 24, time poverty is higher inrural than inurbanareas, although the thresholdof relativetime poverty is muchlower for this age group, so it is less likely to be an issue. Who works the most in Ghana? To a large extent, the results providedinthis paper confirm conventionalwisdom. Women are found to work much more than men, especially on domestic tasks and in rural areas. Thus, while it is often argued on the basis of labor statistics that women participate less than men in the labor force of developing countries such as Ghana, especially in urban areas, this reflects more the impact of traditional definition of what constitutes work in household survey questionnaires than any real differencesbetween genders, as women tend to work at least as hardas men. Source: Coulombeand Wodon (2008a). 22 32. Agriculture remains by far the most important sector of employment, but its importance has slightly decreased over time. As shown intables 6 and 7, agriculture still accounted for more than half o f total employment in2005/06. The services sector (trading, transport and communication, financial services, and other services) comes in second place, employing roughly one third o f the labor force, while the industry sector employed about 15% o f workers. Over the past fifteen years, the percentage o f employed people working inthe agriculture sector has fallen, along with that inthe services sector (due to a drop inthe shares o f workers employed in community and other services), while the share o f workers employed in industrial sectors has risen. Of the 2.8 million new jobs created between 1991-2005 for the 25-64 age group, approximately 44.5% were in agriculture (1.24 million), with close to one million newjobs in services (including Government), and more than half a million inindustry (see Tables 6 and 7). Table 6: EmploymentDistribution by Sector (percentage, for age group 25-64) 1991/92 1998/99 2005/06 changes Agriculture 55.6 52.9 54.1 -1.5 MininglQuanying 0.6 0.8 0.7 0.1 Manufacturing 9.1 11.9 11.6 2.5 Utilities 0.2 0.2 0.3 0.1 Construction 1.4 1.4 1.8 0.4 Trading 18.4 19.3 18.8 0.4 TransportICommunication 2.4 2.3 2.7 0.3 Financial Services 0.7 0.9 1.3 0.6 Community& Other Services 11.6 10.3 8.8 -2.8 All 100.0 100.0 100.0 0.0 Source: Basedon CoulombeandWodon (2007), usingGLSS surveys. Table7: EmploymentDistributionby Sector (absolute numbers, for age group 25-64) 1991/92 I998/99 2005/06 change distribution Agriculture 2,294,390 2,419,690 3,533,997 1,239,607 44.6 MininglQuanying 23,360 49,794 54,436 31,076 1.1 Manufacturing 375,950 619,075 835,085 459,135 16.5 Utilities 8,030 9,188 17,475 9,445 0.3 Construction 59,130 85,022 134,635 75,505 2.7 Trading 757,010 1,027,445 1,387,163 630,153 22.7 TransporWornmunication 99,280 134,920 206,843 107,563 3.9 FinancialServices 27,010 47,304 92,098 65,088 2.3 Community& Other Services 478,880 531,231 638,082 159,202 5.7 All 4,123,040 4,923,670 6,899,813 2,776,773 100.0 Source: Basedon CoulombeandWodon(2007), usingGLSS surveys. 33. A large share of laborforce remains employedin the informal sector. Intables 8 and 9, which group workers by category, we see that more than 80 percent o f total employment remains inthe informal sector, and this share has increased slightly over time. This increase is due mostly to a rise in the share o f unpaid self-employed agriculture workers. However, since the employment status o f these workers i s not entirely clear (many are women who combine domestic and other work), it i s quite possible that while some may declare themselves to be working, others in similar situations may not. As such, statistics on these workers can appear volatile. If all or some o f these workers had been 23 excluded from the analysis, the rate o f formalization would probably appear to have increased over time. 34. There has been a shijlfrom public toprivate formal sector employment. Though Aryeetey et a1 (2004) argued that the decline in public sector employment over time has been compensated for, in terms o f total employment, by growth in informal sector employment, tables 8 and 9 demonstrate that in this case, total employment figures have been sustained by an increase inprivate sector formal employment. It is also worth noting than inurban areas over 50 percent of employment is salaried, while inrural areas, most workers are self-employed. Note also that the distinction between paid and unpaid agriculture workers is an imperfect way to try to identify those workers who belong to larger farms or small enterprises, versus subsistence farmers working on their own; unfortunately, differences indefinitions between surveys make it difficult to measure over time more precisely the share of informal workers working infirms or farms with multiple employees as opposed to by themselves. Table 8: Employment Status (percentage,for age group 25-64) 1991-2005 1991/92 1998/99 2005/06 change Wage Publicsector 12.3 8.1 6.6 -5.7 Wage Private sector Formal 4.1 4.5 6.7 2.6 Wage Private sector Informal 2.2 2.2 4.3 2.1 Self-employedAgro -Paid 36.9 38.0 30.7 -6.2 Self-employedAgro -Unpaid 16.4 13.6 21.8 5.4 Self-employed NonAgro 28.0 33.6 29.9 1.9 All 100.0 100.0 100.0 0.0 Source: Basedon Coulombe and Wodon (2007), usingGLSS surveys; Changes are expressed in percentage points. Table 9: Employment Status (Absolute numbersfor age group 25-64), 1991-2005 1991/92 1998/99 2005/06 change distribution Wage Publicsector 508,080 417,258 487,019 21,061 -0.8 Wage Private sector Formal 167,900 246,77 1 497,876 329,976 11.9 Wage Private sector Informal 92,7 10 128,162 335,948 243,238 8.8 Self-employedAgro Paid 1,s 19,860 1,734,566 2,214,7 10 694,850 25.0 Self-employedAgo Unpaid -- 678,170 622,588 1,193,344 515,174 18.6 Self-employed NonAgro 1,156,320 1,774,324 2,170,917 1,014,597 36.5 All 4,123,040 4,923,670 6,899,813 2,776,773 100.0 Source: Basedon CoulombeandWodon (2007), usingGLSS surveys; Changesare expressedinpercentagepoints. 35. The share of workers living in poverty has decreased for all groups, but the decrease has been largestfor workers with better jobs. As the tables below show, there has been a decrease over time in poverty for all categories o f individuals, but the decrease has been largest inproportional terms for the categories o f workers with the smallest rates o f poverty in the initial year. For example, poverty among formal sector workers has virtually been eliminated, while it remained substantial among the self-employed in 2005/06. This suggests that improvements in wages and standards o f living have been larger for better off individuals, which confirms that inequality in consumption per capita has increased over time in Ghana (See Table 12). Also, in 2005/06, the probability o f being poor remained much higher for agricultural workers than for any other group o f workers. 24 Table 10: Percentage of Workers Residing in a Poor Household, for Different Categories of Labor Market and Employment Status for age group 25-64,1991-2005 1991/92 1998/99 2005'06 ~~~~~~ change Wage Public sector 26.4 15.9 5.9 -20.5 Wage Private sector Formal 22.6 9.0 5.7 -16.9 Wage Private sector Informal 29.1 15.3 13.1 -16.0 Self-employedAgro -Paid 59.5 47.7 32.4 -27.1 Self-employedAgro -Unpaid 67.8 54.4 45.3 -22.5 Self-employed NonAgro 35.3 22.6 13.1 -22.2 Unemployed 21.7 8.0 18.9 -2.8 Inactive 42.7 40.1 29.3 -13.4 All 46.6 34.5 24.3 -22.3 Source: Based on Coulombe and Wodon (2007), usingGLSS surveys; Changesare expressedinpercentagepoints. Table 11: Percentage of Workers Residing in a Poor Household, for Different Categories of Labor Market and Employment Status for Age Group 25-64, in Rural and Urban Areas 1991/92 1998/99 2005/06 Change v Urban Rural Urban Rural Urban Rural Urban Rural Wage Public sector 15.5 43.0 10.1 23.7 3.5 13.5 -12.0 -29.5 Wage Privatesector Formal 20.4 27.4 5.0 16.7 3.4 12.8 -17.0 -14.6 Wage Private sector Informal 27.4 32.6 2.9 26.7 11.9 15.1 -15.5 -17.5 Self-employedAgro - Paid 37.4 61.7 40.0 48.7 17.1 34.6 -20.3 -27.1 Self-employedAgro - Unpaid 51.7 70.1 41.4 56.6 30.6 46.9 -21.1 -23.2 Self-employed Non Agro 21.3 52.5 12.6 32.9 5.3 24.4 -16.0 -28.1 Unemployed 16.7 41.9 3.9 20.8 12.4 42.5 -4.3 0.6 Inactive 24.5 56.0 17.6 55.2 14.0 48.6 -10.5 -7.4 All 23.5 59.0 15.7 44.8 8.8 35.1 -14.7 -23.9 Source:Based on CoulombeandWodon (2007), usingGLSS surveys; Changesare expressedinpercentagepoints. Table 12: Percentage of Workers residing in a Poor Household by Sector of Employment for Age Group 25-64,1991-2005 1991/92 1998/99 2005/06 Agriculture 61.0 48.9 36.1 Mining1Quan-ying 28.1 7.7 3.4 Manufacturing 37.9 25.3 16.6 Utilities 18.2 10.0 0.0 Construction 27.2 21.8 10.5 Trading 34.5 21.2 10.6 TranspordCommunication 18.4 5.4 10.0 FinancialServices 10.8 6.3 5.3 Community & Other Services 25.8 16.0 6.2 All 47.8 34.0 23.8 Source: Basedon Coulombe andWodon (2007), usingGLSS surveys; Changesare expressedinpercentagepoints. 25 Table 13: Percentageof PeopleResidingin a Poor Householdby Levelof Educationfor Age Group 25- 64, 1991-2005 1991/92 I998/99 2005/06 No education 58.0 48.5 39.3 Primary 46.6 32.7 20.1 Secondary lower 33.5 23.2 12.8 Secondary Higher 14.0 12.3 6.9 Post Secondary 17.4 12.9 3.8 All 46.6 34.5 24.3 Source: Based on Coulombe and Wodon (2007), usingGLSS surveys; Changes are expressedinpercentagepoints. 36. Who are the unemployed? Unemployment, as in other Low Income Countries (LICs), i s predominantly an urban phenomenon, and heavily concentrated in the capital, with an unemployment rate in Greater Accra more than twice that of the rest o f the country. It i s also concentrated among the young and relatively well-educated - especially those with secondary education (Table 14). This phenomenon i s often explained by the description o f unemployment in LICs as being a `luxury' : only the non-poor can afford not to be working in some capacity. However, the reality is more nuanced for Ghana. In2005 about one out o f five unemployed workers resided in a poor household. Furthermore, poverty decreased much more slowly among the unemployed than for the population as a whole, with unemployed people in Accra, the Volta, the Northern, and the Upper West regions actually getting poorer. Clearly, unemployment i s progressively becoming a serious issue for at least part o f Accra's population, probably in part due to urban migration by relatively young workers, some o f whom have a hard time finding work. However, it most be noted that sample sizes for the unemployed are often small, which may lead to misleading large variations over time in poverty rates. Also, the proportion o f the unemployed remains very small (at least when using a "narrow" definition o f unemployment), and has actually decreased over time at the national level. As such, underemployment is probably a more serious issue for many Ghanaians than unemployment per se. 26 Table 14: Unemployment Rates for Age Group 25-64,1991-2005 1991/92 1998/99 2005/06 change Sex Male 2.8 3.0 2.2 -0.6 Female 2.6 1.8 2.4 -0.2 Age 24-54 2.9 2.5 2.5 -0.4 55-64 1.1 1.5 1.2 0.1 Education Level No education 1.3 0.9 1.4 0.1 Primary 2.1 2.5 2.3 0.2 Secondarylower 4.6 4.0 2.7 -1.9 SecondaryHigher 7.9 4.2 6.1 -1.8 Post Secondary 4.0 2.3 3.4 -0.6 UrbadRural Urban 6.4 5.3 4.6 -1.8 Rural 0.8 0.9 0.8 0.0 Region Western 1.2 2.5 2.0 0.8 Central 1.2 0.6 3.0 1.8 GreaterAccra 7.5 8.1 4.9 -2.6 Volta 3.2 0.6 1.4 -1.8 Eastern 0.8 1.4 1.2 0.4 Ashanti 4.7 3.1 2.8 -1.9 BrongAhafo 0.8 0.7 1.1 0.3 Northern 1.6 1.o 0.7 -0.9 Upper East 1.o 3.3 2.4 1.4 Upper West 0.7 0.4 3.6 2.9 All 2.7 2.4 2.3 -0.4 Source: Based on Coulombe and Wodon (2007), using GLSS surveys; Changes are expressedin percentage points. 27 Table 15: Poverty (head count index) amongthe Unemployedfor Age Group 25-64,1991-2005 1991/92 1998/99 2005/06 change Sex Male 23.3 7.8 22.1 -1.2 Female 20.2 8.2 16.2 -4.0 Age 24-54 21.5 5.9 19.3 -2.2 55-64 25.0 34.0 11.4 -13.6 Education Level No education 30.8 13.3 35.0 4.2 Primary 33.3 19.3 24.5 -8.8 Secondary lower 17.5 4.8 13.5 -4.0 Secondary Higher 20.0 0.0 9.7 10.3 Post Secondary 0.0 11.5 3.5 3.5 UrbadRural Urban 16.7 3.9 12.4 -4.3 Rural 41.9 20.8 42.5 0.6 Region Western 37.5 20.5 0.0 -37.5 Central 42.9 17.5 3.0 -39.9 Greater Accra 16.4 0.3 23.9 7.5 Volta 8.0 25.9 14.4 6.4 Eastern 75.0 13.7 10.2 -64.8 Ashanti 25.0 5.2 5.6 -19.4 Brong Ahafo 20.0 0.0 0.0 -20.0 Northern 11.1 51.2 34.6 23.5 Upper East 100.0 6.4 71.8 -28.2 Upper West 0.0 0.0 92.5 92.5 All 21.7 8.0 18.9 -2.8 Source: Basedon Coulombe andWodon (2007), usingGLSS surveys; Changes are expressedinpercentagepoints. Note: The drop in the poverty statistics for then unemployedin 1998199 is probably due in part to a sampling issue in the greater Accra area where poverty was observed to fall much more than expected that year (see Coulombe and Wodon, 2007). 37. The population is becoming more educated and is increasing its skills. The percentage of people (not workers) with no formal education has dropped by 4.4 percentage points over the last 15 years, while the share o f population with tertiary education increased from 2.6% in 1991 to 7.6% in 2005 (Table 16). As suggested by others (e.g., Aryeetey et a1 2004; Dabalen et a1 2003), it i s likely that part of this improvement is due to a number of factors, including the fact that : (i)workers are upgrading their skills inorder to be better able to "take advantage of growing opportunities arising from increased globalization"; (ii) inthe manufacturing sector, which is growing, 28 training i s more easily available andpaid for andprovided by the employer; (iii) or foreign large firms also provide more training than domestic and small one; and (iv) already educated workers receive more training. A more detailed discussion of education as it relatesto the labor market is provided inchapter 5. Table 16: Distribution of the Populationby EducationLevel in Age Group 25-64,1991-20055 1991/92 1998/99 2005/06 Change No education 53.6 45.1 49.2 -4.4 Primary 9.9 11.8 10.7 0.8 Secondary lower 29.4 32.9 27.0 -2.4 Secondary Higher 4.5 5.3 5.4 0.9 Post Secondary 100.o 2.6 4.8 7.6 0.o 5.0 All 100.0 100.0 Source: Based on Coulombe and Wodon (2007), using GLSS surveys; Changes are expressedin percentagepoints. Note: The drop in the share of the population with no education in 1998199is probably overestimated. INCOME FROMWORK INAGEGROUP 25-64 38. I t has long been a source of debate as to whether earnings data are of good quality in Low Income Countries (LICs) with large rural population shares. Many LICS collect wage data only from salaried workers. This approach, though straightforward, assesses only a small section of the labor force - often as small 10% - which also represents better off workers. It i s therefore difficult to carry out a credible analysis of labor markets when restricting oneself to wage and salary workers. The Ghana Living Standard Surveys (GLSS) collect wage data on a much larger group of individuals, and includes a separate module for agricultural and non-agricultural business, and the Ghana Statistical Services provide data on business profits (see GSS (2000)). Based on this information, income for the self-employed can be calculated as the profit of the business, divided by the number of household members working in the business. For agricultural businesses, a similar approachcan be followed. The GSS computes household agricultural income by deducting total input costs (seeds, fertilizer, etc), inclusive of wages paid to non-household membersand an allowance for depreciation from the value of gross output, (either sold to the market or consumed), from total revenues. This, divided by the number of working household members, can be usedto calculate income per worker. Income from rentandremittances should not be incorporated. 39. I n Ghana, the labor market is characterized by large differences in earnings between different types of workers, with public sector employment paying the highest wages. Figure 6 shows that there i s a clear hierarchy inwages, with salariedworkers inthe public sector faring best, followed by salaried workers inthe private formal sector, salaried workers in the private informal sector and self-employed workers not in agriculture, and, finally, self-employed workers in agriculture. In addition, the groups of wage and salary workers with the highest earnings also tend to have a distribution of earnings with a limited spread (less variance in earnings between individuals). Finally, formal sector workers not only have better paid jobs, but in addition they usually bring home other valuable job attributes (such asjob security, worker protection, access to social security, etc.). 29 Figure 6: Income from work for 2005 in Age group 25-64 Source: Basedon Coulombe and Wodon (2007), using GLSS surveys. 40. Earnings increased substantially from 1991-2005, but the gap between formal public and private sector workers increased. While table 17 shows that the largest earnings gain i s observed for the self-employed inagriculture who are paid, this increase i s from a low base, and subject to some extent to measurement error (the ratio o f household farm business profits to the number o f family workers is very sensitive to movements in and out o f the household, and probably leads to an underestimation o f true earnings). The second highest increase was for wage workers in the public sector (89%). Other groups have seen earning gains between 40% and 50% inreal terms over the 15 year period under review. Overall, the wage gaps between different employment categories remain high, as expected. Yet what i s striking is that in2005, public sector workers earned 47% more than private formal sector workers, with an increasing differential between both groups over time. Private informal sector wage workers earn about 63% o f what private formal sector workers earn, while individual self-employed earn less than a third o f what private sector workers are paid. This increase in the gap between public sector workers and other workers may be due to multiple reasons and it does not necessarily mean that there is a systematic premiumfor public sector workers. For example, the gap may be related to the fact that public sector workers tend to work in Accra and are better educated. We will come back in Chapters 4 and 5 to the analysis of the determinants o f earnings, and the marginal impact o f beinginthe public sector (as well as having a formal job and benefiting from a trade union) on earnings controlling for other individual characteristics and geographic effects on wages. Also of interest is the fact that the self employed (paid) agricultural workers gained substantially in terms o f earnings, which could perhaps be related to the fact that hired workers in cocoa production probably have been able to improve their earnings due to the boom inthe sector. 30 Table 17: Annual Real Earnings across Employment Statusin Age group 25-64 (in `000 cedis) 1991/92 I998/99 2005/06 Change Wage Public sector 7,470 8,992 14,120 89.0 Wage Private sector Formal 6,480 8,054 9,574 47.7 Wage Private sector Informal 4,205 3,915 6,025 43.3 Self-employedAgro-Paid 1,597 1,385 3,064 91.9 Self-employedAgro-Unpaid Self-employedNonAgro 3,455 3,733 5,180 49.9 All 1,442 1,293 2,551 76.9 Source: Basedon Coulombe andWodon (2007), usingGLSS surveys; Changes are expressedin percentagepoints.Note: Median annual earnings expressed in local currency in 1991 prices. Some divergences in trends observed with the 1998199 datamaybe due to samplingissues and/or lower level o f comparability. Table 18: Earnings Ratio's to private formal sector wage in age group 25-64'1991-2005 1991/92 1998/99 2005/06 Wage Public sector 1.15 1.12 1.oo 1.47 Wage Private sector Formal 1.oo 1.oo Wage Private sector Informal 0.65 0.49 0.63 Self-employedAgro-Paid 0.25 0.17 0.32 Self-employedAgro-Unpaid Self-employedNon Agro 0.53 0.46 0.54 All 0.22 0.16 0.27 Source: Based on Coulombe and Wodon (2007), using GLSS surveys; Some divergences in trends observed with the 1998/99datamay be due to samplingissuesand/or lower level o f comparability. 41. The relatively high wages in the public sector and their increase over time is of potential concernfor the rest of the economy. Analyses in other countries (including in the Africa region) have shown that highpublic sector wages prompt rising formal private wages, due to competition for human resources. As shown inEthiopia, it is also likely that high public sector earnings will encourage individuals to queue for public sector jobs instead o f pursuing less attractive private sector opportunities. Based on earlier works on Ghana, there i s evidence that both the size o f firms and unionization are important explanatory factors for this wage premium (Soderbom and Teal; Kingdon, Sandefur and Teal and others, 2005). Although, the size effect is likely to reflect a difference in productivity (it may also be due to a higher cost to monitor), the union effect indicates the importance o f institutions and labor regulation (see chapter 4). 42. There are also disparities in earnings by region, socio-economic groups, sector of activities, gender, age and skills. Findings include: women earn consistently less than men, regardless of their employment status; the earnings-age profile follows an inverse U- shape pattern; ingeneral, earnings increase with education; earnings inthe urbansector are higher than those in the rural sector; agriculture workers have consistently the lowest earnings; and earnings in the manufacturing and trading sectors are at the tail end o f the earnings distribution. Overall, according to GLSS data the share o f low earners has consistently decreased for virtually all groups, but the pace o f progress has been uneven between groups. 31 HOURS WORKEDAND HOURLYEARNINGS AGEGROUP IN 25-64 43. Labor supply expressed in hours worked per year is lowest in the agriculture sector and highestfor wage workers in the private sector. Based on average work load reported in hours per year, the hours worked are the highest in transport and communication, and lowest in agriculture where seasonality plays a larger role (see Table 19). The number of hours worked has increased for all categories of workers over time, which has contributed to increasing living standards. However, these increases have been more dramatic for certain categories, and there are notable differences betweencategories in terms of hours worked. For example, public sector workers, who earn more, are working fewer hours than private sector workers. For every eight hours worked by a wage worker inthe private formal sector, the public sector workers work 7 hours. Over time, the gap between public sector and formal private sector workers has widened, despite the faster increase in the former's wages documented earlier. It i s also worth noting that the poor supply less labor than the non-poor (they work about 10% fewer hours), so that underemployment, defined as working limited number of hours, i s probably one of the contributing factors to poverty (inrural areas mainly). Table 19: Mean Number of Hours Worked per Year in age group 25-64 (1991-2005) 1991/92 1998/99 2005/06 change Sector Agriculture 1,400 1,356 1,54 1 10.1 MininglQuarrying 2,067 2,254 2,662 28.8 Manufacturing 1,653 1,658 1,953 18.1 Utilities 2,332 2,202 2,380 2.1 Construction 1,868 1,403 1,894 1.4 Trading 1,893 1,943 2,230 17.8 TransportlCommunication 2,707 2,687 2,885 6.6 Financial Services 2,160 2,3 12 2,452 13.5 Community & Other Services 1,904 2,086 2,064 8.4 Employment Status Wage Public sector 1,974 2,09 1 2,074 5.1 Wage Private sector Formal 2,237 2,187 2,516 12.5 Wage Private sector Informal 1,902 2,230 2,2 17 16.6 Self-employed Agro-Paid 1,451 1,389 1,579 8.8 Self-employed Agro-Unpaid 1,211 1,210 1,429 18.0 Self-employed Non Agro 1,817 1,846 2,091 15.1 Poverty Status Very poor 1,418 1,286 1,520 7.2 Poor 1,558 1,359 1,727 10.8 Nonpoor 1,765 1,827 1,924 9.0 All 1,620 1,651 1,846 14.0 Source:Basedon Coulombe andWodon (2007), usingGLSS surveys; Changes are expressed inpercentagepoints. 44. There are large disparities between hourly earnings for different categories of employment. Overall, the hourly earnings in industry and services are broadly similar (with differences according to selected sub-sectors), while the agriculture sector has 32 significantly lower hourly earnings (although it i s likely that these are underestimated). Among employment status categories, public sector workers have benefited from the largest increase in hourly earnings, at about twice the rate of the national average. The increase in hourly earnings of poor workers has been much lower (and in some cases negative) than for non-poor workers (Table 20). Table 20: Earningsper Actual Hour Worked by Economic Activity, EmploymentStatus and Poverty in age group 25-64,1991-2005 (in cedi) I991/92 1998/99 2005/06 change Sector Agriculture 757 724 937 23.8 MiningiQuanying 5,34 1 6,685 5,297 -0.8 Manufacturing 2,255 2,384 3,021 34.0 Utilities 3,437 4,473 6,988 103.3 Construction 3,539 3,925 4,992 41.1 Trading 2,269 2,271 2,598 14.5 TransportKommunication 3,004 3,843 3,078 2.5 FinancialServices 4,779 7,829 5,793 21.2 Community & Other Services 4,123 3,563 5,183 25.7 Employment Status Wage Public sector 3,954 4,869 6,925 75.1 Wage Private sector Formal 3,232 3,806 3,930 21.6 Wage Private sector Informal 2,306 2,189 2,782 20.6 Self-employedAgro-Paid 1,277 1,211 2,033 59.2 Self-employedAgro-Unpaid Self-employedNonAgro 2,388 2,309 2,713 13.6 Poverty status Very poor 912 829 552 -39.5 Poor 1,270 1,26 1 1,346 6.0 Nonpoor 2,157 2,117 2,483 15.1 All 1,521 1,669 2,060 35.4 Source:Basedon CoulombeandWodon (2007), usingGLSS surveys; Changesare expressed inpercentagepoints. 45. Differences in earnings across different types of employment raise several issues. Wage and salary workers' earnings are up to three times higher (inthe case of public sector workers) than the hourly earnings of the individual self-employed. Further preliminary analysis based on regression techniques also suggests that men are more likely to be in wage and salary work than women. Young people are by contrast less likely to be wage and salary workers (see the following section for more details related to education). Overall, workers (menand women) with secondary education, TVET or Tertiary education have the highest probability of being wage and salary workers, while those with lower education are more likely to be self-employed, and many agriculture workers have no education (see Chapter 5). 33 SUMMARY OFFINDINGS 46. From this descriptive analysis o f labor market trends and characteristics, a few conclusions stand out: Earningsfrom work have risenfor all groups, although less sofor thepoor. This rise in earnings has been an important driver o f poverty reduction in Ghana. However, a majority o f the labor force remains in low productivity, mostly agricultural activities, which yield low earnings. The percentage o f formal jobs (both public and private), as a share o f total employment, has decreased, as the decline inthe share o f public sector jobs has been only partially compensated by an increase inthe share o f private formal jobs. Thus, the labor force has become more informal over time, especially as many workers have shifted out o f agriculture to join the informal service sector inurban areas. The decrease inthe formal sector as a share o f total employment i s confirmed by a series o f statistics to be discussed in the next chapter on contracts, trade unions, and other aspects. The public sector's attractive pay policies are likely to affect labor supply decisions. As discussed inthis chapter the increase inthe gap between public sector workers and other workers can be due to multiple reasons. Also the attractive pay inthe public sector encourages queuing for higher paying public sector jobs rather than less attractive opportunities in the private sector. These effects can limit the potential dynamism o f the private sector in attracting the needed labor with the appropriate skills. The public sector can also affect wage setting in the private sector. These are important issues that needfurther investigation. A large amount of the working population remains low skilled, lacking even primary education. Based on the preliminary analysis, it seems the returns to education are positive, but low at low levels of schooling. The analysis also confirms that education plays a role in the attainment o f high paying jobs, which suggests that the poor require more education to get access to such jobs. Apprenticeships may offer one solution. Chapter 5 provides more details analysis of links between education and labor market outcomes 34 CHAPTER 4. LABORMARKETPOLICIES,INSTITUTIONS, AND PROGRAMS As mentioned in theprevious chapter, less than 20 percent of the labor force is engaged in the formal sector, but this sector is growing, with private wage employment doubling in absolute terms during 1991-2006. For a country like Ghana that aims to reach a middle-income country status, the challenge is to ensure that labor market policies are flexible enough to encourage growth of jobs and productivity. I n this context this chapter examines the functioning of the labor market itserf,looking at the implicationsof labor regulations, institutions andprograms on job creation andformalization of the economy. Thefirst section reviews labor regulations; the second section examines the institutional settings of the labor market; and the third section reviews employment programs. The analysis shows that: (i) overall, Ghana has relatively flexible labor regulations that do not create a major obstaclefor private sector development, but further reforms might be necessary to address dismissal costs and procedures that seem to be costly and complex; (ii) the increase in informal employment and reduction inpublic jobs in the last 15 years, leading to decline in the share of workers with contracts and social protection benefits, raises concerns; (iii) non-wage labor costs are moderate comparedto other countries in the region, but the Government needs to be careful to ensure that the ongoingpension reforms do not distort the labor market and do not raise labor costs; (iv) the workers' unionsplay an important role in labor relations and influence the wage-setting process, especially in large firms, though recent institutional changes introduced by the Labor Act have changed the dynamics of industrial relationsand requirefurther investigation;and (v) the Governmentyouth employment program can be an effective tool to increase the employability of youth but the design needs to avoid distorting the labor market functioning and ensure efficient use of the publicfund, and the monitoring and evaluation system needs to be strengthened. LABOR MARKET POLICIES Labor Regulations 47. Although labor market regulations affect mostly formal sector employment, they also influence workers in the informal sector and the unemployed. As discussed in the previous Chapter, workers engaged in formal employment - including wage workers in public sector and wage workers in private formal sector - account for about 13 percent o f employment in Ghana. One o f the main objectives o f the government's poverty reduction strategy i s to create favorable conditions for private sector development, which i s seen to be essential for long-term economic growth and poverty reduction. A strong and sizable formal private sector would create better-quality and higher-paying jobs and would contribute to increased public revenues, which, inturn, could be channeled to help the poor through various social programs. Inthis context and given the fact that most o f the labor force i s outside o f the.forma1 sector, it is important to consider whether the regulatory and institutional framework in Ghana i s sufficiently flexible to encourage overall job creation andto provide right incentives for informal businesses to enter the formal sector. 48. Government has taken major stepsforward in consolidating and harmonizing a large number of different labor laws, bringing more flexibility to labor relations. Adopted in 2003, the Ghana Labor Act serves as the main legal document on labor 35 relations and i s more detailed and comprehensive than previous laws. The new Labor Act has: (i)unified and modified a large number of old labor laws, bringing them into conformity with the core principles o f the International Labor Convention, to which Ghana i s a signatory; (ii) introduced more flexible hiring conditions through short term contracts and temporary workers; (iii) clarified rights and obligations o f employees and employers, particularly in terms o f disputes, dismissals, and severance pay; (iv) defined the functioning o f trade unions and employers' organizations, collective bargaining, and collective agreements; and (v) allowed the establishment o f private employment agencies (see Box 4). Compared to the previous labor laws in Ghana, and to existing labor laws in neighboring West African countries, the new Ghana labor law has a relatively flexible framework with a limited number o f direct state interventions in the labor market. Discussions carried out in the country during the preparation of this study showed that there i s a general consensus among the employers' groups, workers' unions, and the government that the recent reforms are a right step towards simplification and flexibility o f labor regulations inGhana. Box 4: Labor Law in Ghana Prior to 2003, the main sources of regulations relating to termination of employment in Ghana were the Industrial RelationsAct of 1965, the Labor Decree o f 1967, as well as the Civil Service Act and the Civil Service Disciplinary Code Regulations, 1971, which regulated only State employees. Based on extensive tripartite consultations the Labor Act of 2003 (LA) was prepared and is now the main source of legislation relating to labor law and termination of employment. The 2003 L A applies to all workers and employers, except the Armed Forces, the Police, and Security Agencies. However,the L A does not apply to the workers under acontractof employment for specifiedperiod of time or specified work and workers engaged on a casual/seasonalbasis. According to the 2003 LA, the employment of a worker by an employer for aperiodof six months or more must be secured by awritten contractof employment. Termination of employment and dismissal. The L A allows the termination of contract (i)by mutual agreement between the employer and the worker; or (ii) by the worker on grounds of inappropriatetreatment. The employer can also dismiss a worker (i) if the worker is unable to carry out his or her work due to sickness or accident; (ii) if the worker is incompetent and lacks qualification inrelationto the work; (iii) for provenmisconductofthe worker; and (iv) for economic reasons (redundancy when employer contemplates the introduction of major changes in production or technology).The terminationof worker is considered unfair if it is linked (i) to the worker membership in trade union; (ii) to the worker's gender, race, color, religion, social, political or economic status; (iii) to temporarily illness/ injury; (iv) to the current level of qualification requiredinrelationto the work for which the worker was employed. Notice and prior procedural safeguards. A contract of employment may be terminated at anytime by either party giving to the other party: (i)one month's notice, or one month's pay in lieu of notice, in the case of a contract of three years or more; (ii) notice of two weeks or payment in lieu inthe case of a contract of less than three years; (iii)seven days' notice or payment in lieu in the case of a contract from week to week; or (iv) without notice if the amount of remunerationfor the period of the notice is paid. However,the legalprovisionsare not applicablewhere in a collective agreement the terms and conditions for termination of employment are defined and which are more beneficial to the worker. In respect to collective dismissalsfor economic reasons, the employer has to provide in writing to the Chief Labor Officer and the trade union concerned, not later than three months before the contemplated changes, all relevant information including the reasons for any termination, the number and categories of workers likely to be affected, and the periodwithin which any terminationis to be carried out. The employer must also consult the trade unionconcerned on measures to be taken to avert or minimize the termination, as well as measures to mitigatethe adverse effectsof any terminationsonthe workers concerned, suchas finding alternativeemployment, Severance pay. In case of dismissals for economic reasons (severance), where the worker becomes unemployed or where the worker suffers any diminution in the terms and conditions of employment (measured by past services and accumulatedbenefits), such worker is entitledto compensation("redundancy pay"). According to the LA, the terms and conditions of redundancy pay are subject to negotiation between the employer and the worker or the trade union concerned. 36 49. The new Labor Act has introducedjlexibility in hiring conditions. Temporary and fixed-term contracts and regulations o f private agency work are two labor market indicators commonly used to compare employment conditions across countries. According to the 2003 Labor Act, workers can be hiredon a casual or temporary basis, and there i s no need to have a written contract for casual employment. On the other hand, all temporary workers employed by the same employer for a continuous six months or longer period are entitled to the same employment conditions (concerning the minimum wage requirements, hours o f work, rest period, paid public holidays, night work and sick leave) as permanent workers. As a result o f this flexibility inhiring, one can expect that a large number o fjobs created since early 2000s involve employment on a temporary basis. Further work using the 2007 I C A survey i s neededto provide additional information on contract terms and to shed light on whether this flexibility has ledto increase intemporary jobs. 50. Tofacilitate job placement, the 2003 Labor Act has allowed thefunctioning of private employment agencies. Currently there are about 30 private employment agencies. Since they target mainly a limited number o f high-skill white-collar workers for large formal enterprises, these agencies have a very marginal impact on job placements. While the employment agencies could be a useful resource for firms looking for both permanent andtemporary labor, the highcost of obtaining a license andthe cumbersome renewal and reporting procedures hinder their efficiency. According to the 2003 law, the placement agencies need to obtain a license and renew it annually with the labor ministry, and they are also requiredto report to the ministry every three months on their activities. The law stipulates that private agencies could charge fees for finding a job placement for job seekers. However, in practice, it i s the employing firm that covers recruitment expenses related to posting job announcements and pre-screening o f job candidates through preliminaryinterviews. 51. The share of wage workers without written contracts has increased substantially in the last 15years confirming the rise of informal wage employment and jobs without protection. Employment in Ghana i s mostly informal and it i s estimated that only about 50 percent o f the wage earners (or about 6.5 percent o f total employment) have written contracts. According to the GLSS data, between 1991/92 and 2005/06 the share o f wage earners with written contracts decreased from 64 percent to 51 percent, mainly due to a substantial decrease o f written contracts in the private sector (from 39 percent in 1991/92 to 29 percent in 2005/06). On the other hand, the share o f public sector jobs with written contracts increased from 77 percent to 87 percent during the same period (Figure 7)4. The drop in the share o f workers with written contracts confirms the earlier discussed trend towards informality o f the Ghanaian economy and its labor market (see Chapter 3), and it can also be a result o f an increasing share o f casual workers who do not need written contracts. In addition to more informality, less written contracts also means less social protection for workers since informal sector jobs are low-paid and often do not provide non-wage benefits (see below). 4 The apparent inconsistency in the figure is due to the huge increase inthe number o f private sector wage employees (with a much lower rate of written contracts) comparedto public ones. According to the survey data, the private/public ratio increased significantly during 1991-2005 so the weighted average was skewed toward the public sectorjobs with written contacts in 1991/92, but toward the private ones in2005/06. 37 Figure 7: Share of salaried workers with written contracts by sector (YO) 87.4 38.6 I i private public total Note: The share of salariedworkers is equal to about 18 percent of the active working age population Source: GLSS3 (1 991/92)& GLSSS (2005/06) 52. Ghana'sfiring rules are rather moderate but the procedures are complex and firing could be costly, especially for seasoned workers. According to 2008 Doing Business indicators, dismissal costs in Ghana are equal to 178 weeks o f wages, which i s quite high compared to the OECD countries and other countries inthe region (See Box 5). This non-favorable high cost of firing could be explained by the fact that the Doing Business methodology takes into account the severance pay for dismissal of workers with at least 20 years o f experience, who might have a previously-negotiated very protective and costly employment contracts. Moreover, high dismissal costs are more common in well-established large formal enterprises with strong trade unions. 53. The number of lays-off is small but has almost doubled between 2004 and 2006, reflecting the fact that the needfor restructuring is sufficiently strong to outweigh the high costs of laying off workers. According to the Department o f Labor, there were 70 cases o f labor redundancies in 2006 that resulted in laying-off 5,719 workers (see Figure 8). These numbers were similar to the number of redundancies (54 cases) and laid-off workers (5,635) in 2005, but were almost twice the number o f lay-offs in 2004 (2,967). Even at the current level o f close to 6,000 lay-offs per year, the numbero f workers affected by redundancies represents less than one percent of the formal sector employment. Financial constraints and falling workload were recorded as two major reasons for the redundancies. 38 Figure 8: The number of lay-offs registered at the Department of Labor 5635 5719 2004 2005 2006 Source: Ministry ofManpower, Youth andEmployment Box 5: Ghana hiringand firing proceduresand costs Procedures: The 2003 LaborActs defineshiring andfiring regulationsfor all types of contracts. According to the law, an employer can only lawfi~llyterminate employees on the grounds of redundancy (the employer is contemplating a close-down or the introduction of major changes in production, program, organization, structure or technology o f its business. on the basis of which the employment of the employees will no longer be necessary). The employer is requiredto give anotice in writing to the Chief Labor Officer not later than three months before the plannedchanges or close-downsetting out all relevantinformation, including the reasonsfor the termination, the number and categories of employees to be affected, and the period within which the termination is to be carried out. The employer is also required to notify the relevant union, if employees are unionized. The notification must be in writing and must be delivered three months before the contemplated changes or close down. The employer is further required to consult with the workers union on measures to mitigate the adverse effects of the terminations on the employees such as finding alternativeemployment. Costs: According to 2008 Doing Business, an annual publication by the World Bank and IFC, the rigidity of employment indcx in Ghana is 37. The rigidity of employment index is the average of three sub-indices: dificulty o f hiring index, rigidity of hours index, and difficulty of firing index. Each of these indices assigns values betwecn 0 and 100; with higher values representing more rigid regulations. Ghana scores well compared to the Africa region's average, but its employment rules are more rigid,than the OECD average. Ghana scores particularly poorly onfiring costs, which measure the cost of advance notice requirements, severance payments and penalties, as expressed in weekly wages. At 178, Ghana's firing costs are high comparedto the region and the OECD average, which could be explained by the fact that the DoingBusinessmethodologyis based on the firing costs of workers with at least20 years of experience, Historically such contractshavehigh severancepaymentsnegotiatedinthe past, The difficulty of firing index (workerslegalprotectionsagainst dismissal, includingthe grounds for dismissal, andprocedures for dismissal) is 50. Ghana's hiring index, which measures flexibility of contractsandthe ratio of minimum wage to the value-addedper worker, stands at a favorable 22, and the rigidity of hours index , which measures restrictions on weekend and night work, workweekrequirements, and mandateddays of annual leave with pay, is 40. DoingBusinessrankedGhanaa low 138 among 178 countries interms o f ease of employingworkers. 11 Difficult4 Doing Business Indicators Indicator I Ghana 1 Region I OECD of HiringIndex I 22 I 41.7 1 25.2 I1 Rigidity ofHoursIndex 40 43.9 39.2 Difficulty of FiringIndex 50 42.2 27.4 Rigidity ofEmploymentIndex 37 42.6 30.8 Non-wagelabor cost (% of salary) 13 12.3 20.7 Firingcosts (weekso f wages) 178 68.3 25.7 39 54. I n Ghana the state does not intervene directly in defining the amount of a redundancypayfor laid-off workers, leaving the matterfor negotiation between workers or their trade unions and the employer. Individual employment contracts or collective agreements negotiated between trade unions and employers include conditions for a redundancy payment and other compensations for the laid-off. However, there are also cases in which workers do not have written contracts or employment contracts do not contain a redundancy clause. In such cases, a redundancy package is decided via negotiations right before such redundancy takes place. Disputes relating to the redundancy pay and the terms and conditions are handled by the Labor Commission for settlement (see Box 6). Box 6: Role of NationalLabor Commission According to the Labor Act, a worker who claims that hisiher employer has unfairly terminated hisiher employment may present a complaint to the National Labor Commission. If the Commission finds that the termination of the employment is unfair, it may order the employer to: (i) re-instate the worker from the date of the termination of employment; (ii) re-employ the worker, either in the same work or other suitable work; or (iii) order the employer to pay compensationto the worker. The main functions of the National Labor Commission are (i)to facilitate and settle industrial disputes; (ii)to investigatelabor relatedcomplaints (specifically unfair labor practices) and take steps such as mediation necessary to prevent labor disputes; (iii) to maintaina database of qualified personsto serve as mediatorsand arbitrators;and(iv) to promote effective labor co-operation between labor and management. In order to maintain its credibility and independence in Ghana, the exercise o f the Commission's adjudicating and dispute settlement function cannot be subject to the control or directionof any personor authority. The Commissionexercisesthe following powers:(i) receivescomplaintsfrom workers, trade unions, andemployers, or employers' organizationon industrialdisagreementand allegationsof infringementof any requirements of the LA; (ii) requires an employer to furnish information and statistics concerningthe employmentof its workers and the terms and conditions of their employment in a form and manner the Commissionconsiders necessary; (iii)requires a trade union or any workers' organizationto provide such information as the Commissionconsiders necessary; and (iv) can notify employers and employers' organizationsor workers and trade unions in cases of contravention of the LA and direct them to rectify any default or irregularities. Any dispute that concerns redundancy pay may be referred to the Commission by the aggrieved party for settlement. The decision of the Commission in such matters is final and in settling industrialdisputes, the Commissionhas the powers of the High Court inrespect of enforcingthe attendance of witness and examiningthe cases. Wage Policies 55. The minimum wage in Ghana is set through tripartite negotiations between workers unions, employers, and the government. The current daily minimum wage in Ghana i s $19,000 (US$1.7) or about #6,150,000 (US$380) per year. According to the Labor Act, the National Tripartite Committee composed o f representatives o f workers unions, employers associations, and the Government is responsible for determining the daily minimum wage. When the daily minimum wage i s determined by the National Tripartite Committee, the Minister for Youth and Manpower Development publishes a notice o f the national daily minimum wage in the media. The main goal o f setting a minimum wage is to promote decent jobs and reduce poverty among workers, but its effectiveness i s questionable in developing countries where a minimum wage i s often set unreasonably high. 40 56. There has been a large increase in real terms in the minimum wage over time. In 2005/06, the year of the latest Ghana Living Standard Measurement survey, the minimum wage was at 13500 cedis per day (US$ lS), versus the equivalent in real terms (adjusted for inflation) o f 8,242 cedis per day in 1998/99, and 5,985 cedis per day in 1991/92. Thus, in real terms, the minimum wage has more than doubled in the last 15 years. Figure 9 provides data on the trend inthe CPI andthe minimumwage over time. Figure 9: Dynamics of minimum wage and inflation, 1994-2006 1500.0 1300.0 t-MW-D--- Inflation 1100.0 900.0 700.0 500.0 300.0 100.0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Source:Government of Ghana, IMF 57. The minimum wage does not seem to be a binding constraint in the sense that many wage workers today earn less than the minimum wage. Density functions for actual wages in 2005/06, as well as the reservation wages of the unemployed and their expected wage given their characteristics are provided in Figures 1Oa to 1Od. The sample excludes workers without pay as well as the self-employed. Data are provided separately for workers aged 18 to 30, and for workers aged 25 to 64. There is a bit of overlap between the two groups, but we definedyoung workers as being between 18 and 30 years of age to increase sample size for that group. The vertical lines on the figures represent the values of the minimumwage. The messages are clear. The minimumwage has increasedrapidly, and while in 1991/92, very few wage workers were earning less than the minimumwage, this was not the case anymore in2005/06. The fact that so manywage workers are earning less than the minimumwage in2005/06 shows that enforcement i s weak, but it also means that if enforcement were stronger, the costs for firms could be high. There is an interesting discrepancy between the imputed wages for the unemployed and their reservation wage, with many workers reporting a reservation wages higher than what they could expect given their characteristics. This suggests that part of the unemployment, especially among youth, may be due to high expectations. 41 N e 3 s Q\ s- 3 N sc"UF W W d W L E ez 0 .I c) h c) .I B.. v) B 3 W L .-a M 3 I I I I I 1 9' 9' P' 2' 0 &!==alaOa)( PI Y 58. At the same time, the increase in the minimum wage implies that the cost of enforcing it could be highforjirm. Annex 4 provides estimates of the potential cost for firms of enforcing the minimum wage. The main parameter of interest i s the wage gap, which computes the extra cost of hiring workers at the minimumwage as a proportion of the minimumwage (the wage gap takes into account the distance separating workers from the minimum wage; it takes a positive value only for those workers below the minimum wage, but the statistic i s computed on all workers). For workers employed and between the ages of 18 and 30, the wage gap has increasedat the national level from 3.7 percent in 1991/92 to 15.1 percent in 2005/06. The increase is even larger for the unemployed according to their imputed wages, from 0.3 percent in 1991/92 to 20.4 percent in2005106. For workers betweenthe ages of 25 and 64, the wage gap has increased as well, although the level i s smaller than for younger workers. The wage gaps are larger in rural than in urbanareas, but they are still substantial even inurban areas. In2005/06 for example, for current wage earners betweenthe age of 18 and 30, the wage gap inurban areas i s at 13.9 percent, while it i s at 6.6 percent for the older age group. The increase in the wage gap over time is due essentially to an increase in the share o f workers below the minimum wage. For example, in urban areas, 32.1 percent of actual wage earners today are below the minimumwage among the 18 to 30 years age group, and the proportion i s 17.4 percent for older workers. For the unemployed, the proportion i s a bit lower on the basis of the worker's declared reservation wage, but it i s significantly higher on the basis of the imputedwage. Evenwhen one restricts the analysis to formal sector workers, the share of current workers below the minimum wage i s 25 percent for the age group 18-30 in 2005/06, and 11percent for the age group 25-64. 59. Another way to look at the level of the minimum wage is to compare it to actual earningsfor various groups of workers. The GLSS data suggest that the minimum wage i s comparable with average earnings for workers inthe informal sector, and lower than the earnings in the public and formal private sectors, but this does not mean as discussed below that a substantial share of workers are not below the minimum wage. The comparison of the 2005 monthly minimumwage with the reservationwage declaredby the unemployed indicates that the minimum wage is below the average reservation wages for most workers, but as discussed previously, reservation wages seem to be unrealistically high in comparison to what unemployed workers may expect to earn on the basis of predictions made using standardwage regressions (see Figures 11& 12)5. The reservation wage i s calculated based on the 2005 survey, where the unemployed were asked "what is the lowest wage that you are ready to accept?" 43 Figure 11: Annual real earnings and minimum wage, 2005/06 ('000 cedis) I 15,376 I public sector formal private sector informalprivate sector minimumw age Source: GLSS5 and Departmentof Labor Figure 12: Monthly Reservation wages by gender, location, and education (2005/06) '' ; 12,0004 10,000 ' 8,000 i 6,000 ~ ~ 4,000 2,000 0 Note: The share of salaried workers is equal to about 18 percent ofthe active working age population Source: GLSSS (2005106). 60. While the minimum wage is high in comparison to prevailing wages, it is not enough to emergefroinpoverty. The minimum wage is adjusted annually to compensate for inflation, and in fact it has been rising faster than the CPI (Figure 11). It is not sufficient, however to pull out of poverty an average family of four with one working adult. This then poses a difficult dilemma. On one hand, there i s a desire to make sure that the minimum wage is high enough to enable workers not to be poor, but on the other hand a high minimum wage provides incentives for firms to stay informal, and if the minimum wage becomes too high, it may become irrelevant for the poor while also implying costs for the formal sector. Given that the minimum wage is now rather high, more detailed assessment o f its potential impact i s probably called for. 44 61. There are large wage discrepancies within thepublic sector with the government taking measures to address these imbalances. Since early 2000s, there has been a steady increase in the wage bill, partly reflecting employment increases and partly correcting historically low levels o f pay6. In 2006 the overall average remuneration for all public sector employees increased 44% compared to 2005, with large variations among different sectors: the average remuneration o f staff o f the Ghana Education Services (GES), the largest component o f the wage bill, grew by 23%, remuneration o f health service staff, the second largest, increased by 181%. These variations reflected the separate approaches taken by different sectors to wage negotiations with the government (e.g., for the health sector, the remuneration increase aimed to prevent the further loss o f skilled staff to the private sector and abroad). With the wage bill now equal to about one half o f tax revenues, and pay anomalies prevalent between services, the Government recognizes that continuation o f existing pay setting arrangements i s unsustainable. Currently, the Fair Wages Commission i s looking at equity issues across services, and the Government has undertaken a comprehensive payroll review as part o f its public sector reform program. The latter envisages shedding surplus staff through rightsizing and linking pay more closely to performance. These initiatives are critical steps towards developing a sustainable policy for pay and employment in the public sector and reducing the costs o f public service reform (Le., separation payments and pensions) and their implications on the labor markets will need to be closely followed.' In the GLSS data, in comparison with other types of workers, public sector workers are found to be paid much better (as discussed in chapter 3). This remains the case after controlling for a range o f characteristics including education and experience. Table 21below provides the marginal impact of being inthe formal sector, being inan organization with a trade union, and being inthe public sector on wages. While the coefficients for public sector affiliation are not statistically significant, they are positive, and public sector workers also benefit from the benefit o fbeing inthe formal sector and inorganizations with unions. Table 21: Percentageincrease in wages from belongingto the formal sector, being in an organization with a trade unionand being inthe public sector on wages Urban Rural Urban Rural Urban Rural 1991/92 1991192 1998199 1991/92 1005/06 2005/06 Formaljob 0.420** -0.063 0.217* 0.125 0.180*** -0.141 (2.36) (0.33) (I.75) (0.87) (2.81) (1.18) Union -0.003 0.169 0.317*** 0.294** 0.144*** 0.179* (0.03) (I.56) (4.51) (2.20) (2.86) (I.68) Public 0.097 0.077 0.078 -0.167 0.095 0.158 (0.85) (0.66) (1.12) (I.38) (I.68) (1.40) Source: See Annex 1(Adams et al. - 2008). T-statistics in parenthesis. Ghanaianpublic servants also had salary increases in 1994 (20%), 1996 (38%), I997 (10-1 8%), 1999 (20- 42%) and 2000 (21% inMarchand 20% inNovember). 'See "External Review of Public Financial Management", September 2007, WB, Marcel0 R.Andrade 45 Non-wage Labor Costs 62. The mandatory non-wage costs of labor in Ghana are comparable with the regional rates. According to Social Security Law o f 1991,workers and employers in the formal sector (public and private) have to contribute the equivalent o f 17.5 percent o f the gross wage bill to the Social Security and National Insurance Trust fund. The contribution by employers amounts to 12.5 percent of gross wages, with the rest paid by workers. Self- employed persons are required to contribute the same amount o f their monthly income to social security. The social security funds are used to pay old-age and disability pensions and family allowances to dependants. To qualify for a pension, a person has to contribute to social security for at least 240 months by age 60 (age 55 for underground mine workers). The law also outlines conditions for early retirement and monthly pensions versus a lump-sumpayment. The non-wage labor costs in Ghana are comparable to the regional average, and even lower than insome neighboring countries (see Table 22). Table 22: Mandatory non-wage labor costs to employers in selected countries in West Africa ~~~ Non-wage CountryIRegion labor costs, % Benin 29.00 Burkina Faso 20.00 Cameroun 16.20 Ghana 12.50 Mali 26.90 Nigeria 8.50 Senegal 21.40 Sub-Saharan Afi-ica I 12.70 Source: Doing Business 2006. 63. Upcomingpension refopms need to consider effects on the labor market. Ghana i s about to start a transition towards a new three pillar pension structure and i s considering several proposals for pension reforms (see Box 7). To ensure that such reforms would lead to wider social security coverage, and to avoid distorting the labor market, several important considerations should be taken into account: (i) further increases o f non-wage costs could contribute to evasion and expansion o f informal work; (ii) parametric reforms are needed to improve financial sustainability o f the pension system in the medium and long term; and (iii) links between contributions and pension benefits, as well as a closer more transparent system with good governance, would increase incentives to contribute. 46 Box 7: PensionSystem in Ghana There are currently two mandatory pension schemes in Ghana: the Social Security and National Insurance Trust (SSNIT), and "CAP 30". . On the aggregate, both systems cover less than 10 percent of the labor force inGhana and cost around2.2 percentof GDP (2005). 0 SSNIT is the main system and covers employees inthe private sector, and some civil andpublic servants. It also covers on voluntary basis the self-employed. SSNIT was establishedby the Social Security Law in 1991, which converted a Provident Fund(from 1972) into a pension scheme (partially-funded PAYG). The institution provides old-age pensions, invalidity, and death benefits.Itis financedby a contribution rate of 17.5 percentofthe employee's salary (5 percent from the employee and 12.5 percent from the employer). Workers become eligible for full benefits after contributing for 20 years, and the pensionbenefit is basedon the average of the workers' 3 highest earning years. The normal retirement age is 60, but workers inhazardousjobs may retire with a f i l l benefit at age 55. Early retirement is availablewith reducedbenefitsfrom the age of 55. In2005 SSNIT hadnearly 900,000 activemembers (contributors), and around65,000 pensioners. CAP 30 is a small scheme, currently phasing-out, and only covers military, police, and a few civil servants. However, despite its limited coverage, CAP 30 expendituresaccount for 1.3 percent of GDP (2004), with its annual cost rising exponentially since 2000. CAP 30 is an unfundeddefined benefit system. Although its contribution rates are the same as those of the SSNIT, it provides muchmore generous pensions. Employeesare eligible for monthly pensionwith only 10years of service, with benefits basedon the final year salary. Although the compulsory retirementage is 60, employees may retire as early as at the age of45. Police and other security agency workers as well as members ofthejudicial and legal services are exempt from contributing. In2004 CAP 30 system coveredmore than 3 1,000 public workers and paid benefitsto 97,000 pensioners. According to early work in Ghana and the white paper on pension reform, the main weaknesses of the current system can be summarized as follows: Benefits inCAP 30 lack any actuarialbasis; there is inequity betweenworkers that contributeand those who do not contribute; there are arrears incollecting contributions and paying benefits; and there is lack oftransparency. 0 InSSNIT the mainissues are: (i) inadequateinvestmentreturns; (ii) coverage with low marginalization of informal sector (public service employees are basically all covered; more than 20 percent ofprivate sector employees are covered; while only less than 2 percent of non- agricultural self-employed, and farmers are covered); (iii) high administrative costs and low efficiency; and (iv) substantial slippage inreal value of the pensions. To reform the system, a Presidential Commission on Pension (established in 2004) made the following recommendation for a contributory three-pillar pension structures, comprising two mandatory and one voluntary scheme: First pillar. A mandatory basic state social security scheme to be administeredby a restructured SSNIT, which will pay only periodic monthly (no lump-sums), and other pension benefits (such as survivors and invalidity benefits). Contributions would be 13.5 percentofpayroll. Second pillar. A mandatory, privately-managed occupational pension scheme, based on a defined contribution pension scheme, paying mainly lump-sum benefits with flexibility that allows the contributor to purchase additional annuities to enhance monthly pensionbenefits. The minimum contribution rate will be 5 percent (4 percent contributed by the employee and the remaining 1 percent contributed by the employer). Third pillar. A voluntary private pension scheme, offering attractivetax incentives.This would be operated inline with the provisions ofthe Long-Term Savings Act (LTSA), 2004. While some of the proposals for such structure are still under revision, the implementation of the new three pillar pensionis planedto start in2008. Under the new pensionsystem all workers, currentlyon the SSNIT scheme andbelow 55 years, would automaticallyjoin the new scheme. 47 64. Formal sector workers, and particularly public sector employees, also receive other non-wage benefits. According to the Labor Act, every employee in the formal sector i s entitled to not less than 15 working days leave with full pay in any calendar year of continuous service. Workers also have a right to paid or unpaid leave in a number o f personal circumstances. The GLSS5 data analysis shows that non-wage benefits more common inthe public sector than inthe private sector. According to the data, for instance, 90 percent o f public sector employees have paid holidays which are significantly higher than among workers inthe private sector (34 percent) (see Table 23). Private Public sector sector Total Paidholidays 34.2 90.2 54.8 Paid sick leave 31.7 65.5 44.1 Pensions 20.9 87.0 45.2 Medical care 26.6 65.1 40.8 Social security 20.8 61.0 35.6 Note: The Source: GLSS5 (2005106) 65. Larger firms provide more non-wage benefits than smaller firms as part of the benefits package negotiated between employees and employers via collective bargaining. According to the GLSS data, for instance, in large firms (with 500 or more employees) 64 percent o f workers received transportation subsidies and 81 percent were also given free medical care. In general, as a result o f negotiations between the workers' unions and the employer, large firms tend to pay higher wages or provide additional non-wage benefits to raise labor productivity and efficiency (such as group insurance, safety provisions, etc,). These benefits, which are not mandated by the state (except for pension and social security contributions), can nevertheless increase total social charges on labor for employers to nearly 20 percent. While a complete assessment o f these charges i s beyond the scope o f this study, it appears that the many small charges add up to a substantial supplementary indirect cost on labor (Table 24). t Table 24: Share of salariedworkers with benefitsby firm size (YO) Non-wage benefits 51-100 101-500 501+ Total Housing subsidies* 20.6 20.2 19.9 12.3 Transport subsidies* 17.3 24.6 63.5 14.1 Paid holidays 33.1 63.7 77.5 71.1 69.0 54.3 Paid sick leave 27.1 45.4 61.9 78.1 69.9 43.4 Pensions 21.3 51.6 75.8 81.3 85.5 44.2 Free medical care 21.5 42.7 59.6 84.4 80.7 39.4 Social security 1 17.6 I 44.7 I 46.8 I 42.4 I 41.8 I 72.9 I 76.3 34.9 :e: The share of salaried workers is equal to about 18 percent of the active working age population. Source: *GLSS4 (1998199) & GLSS.5(2005106) 48 66. There was a sharp decrease in the share of workers with social protection benefits during the last 15years due mainly to the reduction of public sectorjobs and an increase in informal jobs. The data show, for instance, that during this period the share o f workers with paid holidays decreased from 73 percent to 55 percent, while the share o f workers with paid sick leave fell from 74 percent to 44 percent. Also, in 2005/06, only 41 percent of the wage-earners had free or subsidized medical care (compared to 55 percent in 1991/92) and only 36 percent o f them (compared to 65 percent in 1991/92) had social security benefits (See Figure 13). The analysis shows that the fall inthe share of workers with social protection was mainly due to the reduction of well-protected public sector jobs during this period, which declined from 12.3 percent in 1991 to 6.6 percent in 2005 (see Chapter 3). On the other hand, the private sector jobs created during this time were mostly informal and did not offer social protection benefits (including social security, pension, etc.). I Figure 13: Share OfWorkers With Social Benefits, 1991/92 And 2005/06 (YO) 73.2 74.1 65.3 I paid holidays paid sick leave pensions free medical care social security benefits ~ ~~~ ~~ Source: GLSS3 and GLSS.5.Socialsecurity benefitsincludefamily and work injury allowances LABOR MARKET INSTITUTIONS 67. Social dialogue is operating through a tripartite mechanism, involving employer, employee and government. Labor market institutions inGhana are represented by workers unions, employers associations, and governmental labor ministries and agencies.s The Labor Act provides all workers with the right o f association and the freedom to form or join unions. The law requires that all trade unions register with the office o f the Chief Labor Officer. While workers are represented by the labor unions, the employers have formed a number o f employers a~sociations.~ The Ministry o f Manpower, Youth and Employment represents the government at the collective bargaining negotiations between workers unions and employers. In addition, the ministry's Department o f Labor oversees 8Ghana has more than 15 workers unions, including Trade Unions Congress, Public Services Workers, Teachers andEducational Workers Union, and HealthServices Workers, amongthe largesttrade unions. 9Ghana's employers at collectivebargainingnegotiations are representedby: Ghana Employers' Association (GEA), Ghana Industries, andthe Chamber o f Mines. These serve as a professionaland lobbying network to foster enterprise development andto promote free enterprise. 49 working conditions, work disputes, the social security system, and labor force developments. The law provides unions with the right to organize and to bargain collectively to set wages, benefits and work conditions. Negotiations betweenworkers and employers usually take place at the national level, with the Government playing the role of arbitrator betweenworkers unions and employers association. 68. As a result of decliningpublic sector jobs, and slow increase in private formal jobs, the share of unionized labor has beenfalling. It is estimated that currently about 10 percent of the total labor force in Ghana is unionized, which is less than `ahalf of the 1990 rate of 26 percent". Nevertheless, the majority of formal sector workers in Ghana are unionized: according to GLSS data, in 2005 about 70 percent of public sector wage employees and 27 percent of private sector wage employees belonged to trade unions. 69. Although the share of unionized labor has beenfalling in the last 15 years, the unionizedlabor still has a sizeable wagepremium. It i s estimated that currently about 10 percent of the total labor force in Ghana i s unionized, which is less than a half of the 1990 rate of 26 percent". This decrease was due mainly to a decline in the number of public sectorjobs and slow increase in private formal jobs. Nevertheless, the majority of formal sector workers in Ghana are unionized: according to GLSS data, in2005 about 70 percent of public sector wage employees and 27 percent of private sector wage employees belonged to trade unions. A study by the University of Oxford using manufacturing firms survey data concluded that without controlling for differences in human capital there is a very large difference in wages between unionized and non-unionized firms with the latter paying wages three times higher (US$75 as against US$25)12. This study showed that, controlling for human capital, the union wage differential was estimated to be about 60 percent. On the other hand, the GLSS data indicates that the union wage premium for wage workers has declined from about 30 percent in 1998/99 to about 15 percent in 2005 (the premiumwas not statistically significant in 1991/92 however). While the GLSS data i s not comparable to the data from the manufacturing sector that was used to estimate the 60 percent premiumassociated with unions, there i s still evidence that there is a premium associated with unions but that this premiummay well have decreased in recent years, as has union membership. Further work is needed to better understand the factors affecting this decline inbothunionmembershipandthe wage premiumassociatedwithunions. 70. The labor relations have evolved in the country with the government institutions taking more responsibilityfor dealing with industrial issues. Under the new labor law, the Chief Labor Officer, in the Ministry of Youth and Employment, rather than the Trade Union Congress (TUC), issues collective bargaining agreements. This effectively limits the TUC's monopoly, since the old collective bargaining provisions implicitly compelled all unions to be part o f the TUC. Also the National Labor Commission has been established and tasked with resolving labor and industrial issues, removing that responsibility from the Labor Court. However, the capacity of the government institutions lo J.P. Azam et al., "Labor Institutions, Labor-Management Relations, and Social Dialogue inAfrica", 2005 11 J.P. Azam et al., "Labor Institutions, Labor-Management Relations, and Social Dialogue inAfiica", 2005 12F.Teal "Labor Demand inGhana for Skilled andUnskilledWorkers", 2007 50 i s overall weak and they lack resources (financial and technical) to enforce labor regulations andmonitor the labor market. EMPLOYMENT PROGRAMS 71. The National Youth Employment Program (NYEP) is the main employment program. Ghana has a number of employment-related programs, such as the Special Presidential Initiatives (SPI), the Rural Enterprise Development Program (REDP), the National Board of Small Scale Enterprises (NBSSE), and others run by a number of sectoral ministries and agencies. However, the National Youth Employment Program (NYEP) i s the only major program that focuses primarily on job creation, with the rest targeting overall poverty reduction and private sector development rather thanjob creation per se. The analysis inthis section, therefore, will focus primarily on the NYEP. 72. TheNYEP addressesjob creationfor theyouth, dejined asyoungpeople between the ages of 18 to 35. Launched inOctober 2006, the program aims to "empower the youth to be able to contribute more productively towards the socio-economic and sustainable development of the nati~n"'~.The NYEP program is built on the experience of the Skills Training and Employment Placement Program (STEP), which focused mainly on vocational training, including apprenticeship for JSS/SSS graduates, agricultural training for rural areas, and the teaching of entrepreneurship skills to college-graduates. According to the available data, between2002 and 2004 the government spent about 19 million cedis from the HIPC funds to train some 18,000 STEP beneficiaries. No. of youth ActualNo. of Share of youth Region registered youthemployed employed(%) Ashanti 24,322 7,537 31.0 BrongAhafo 19,868 7,932 39.9 Central 13,016 7,697 59.1 ' Eastern 19,100 8,600 45.0 Greater Accra 22,363 7,922 35.4 Northern 21,959 16,528 75.3 Upper East 13,271 9,530 71.8 UpperWest 12,590 9,688 76.9 Volta 18,094 8,674 47.9 Western 10,087 7,967 79.0 Total 174,670 92,075 52.7 73. The NYEP is more ambitious and aims to employ 500,000 young people between 2006 and 2009. The government introduced the program to help unemployed youth find jobs. According to the 2005 survey, the unemployment rate among youth ages 15-24 is about twice as high as the national unemployment rate (6 percent compared to 3 percent). The NYEP aims to train and provide temporary employment for about 170,000 youth per year, with the overall target of 500,000 for 2006-2009. This target number i s basedon the l3 Youth EmploymentImplementationGuidelines, Government of Ghana, 2006 51 national youth employment surveyhegistry carried out prior to the program, which identified and registered about 175,000 young people, with only about 50 percent o f them employed at the time o f the survey. The survey confirmed large regional disparities in youth unemployment, with the highest unemployment rates in large urban areas, particularly Ashanti and Greater Accra regions (Table 25). 74. The NYEP is a broad-basedprogram, involving a number of national ministries and agencies, district assemblies, community-level groups, as well as NGOs and the private sector. According to the design of the program, youth with all types o f educational background are eligible, though a large share o f the employment modules requires at least basic education. The youth employment program targets a wide range of activities in different sectors, such as education, health, water and sanitation, agriculture, and others, and operates inall 10 regions o f the country. To reach its objectives, the program interacts on regular basis with a number o f governmental structures at the national and regional levels and it also contracts out some o f its activities to NGOs and the private sector. Most of the NYEP beneficiaries are employed in thepublic sector toprovide basic social services. Since October 2006, the NYEP has provided employment to 92,075 young people, with about 42,000 o f them working as teaching assistants and health and sanitation workers. Agrobusiness (16,383) i s another important employment module that promotes farm and non-farm income-generating activities in rural areas. The internships module (5,041) targets mainly the educated youth in urban areas seeking employment with the private and public sectors. The employment modules for trades and vocation and ICT are still being developed (see Table 26). Overall the largest share o f NYEP employment i s in the public sector, with teaching and nurse assistants financed by the NYEP filling in for absent teachers and medical staff, especially in remote areas o f the country. While the program is providing important and necessarily social services, the youth hired by the program often lack proper training and do not have all necessary qualifications to carry out their tasks. Table 26: The NYEP Beneficiary Data, 2006-07 Employment modules Beneficiaries Community education teaching assistants 23,021 Agro-business 16,383 Health extension workers 14,000 Internship 5,200 Waste and sanitation 5,041 Community protection 1,300 Trades and vocation I Others I 26.76014 I Program staff 370 Total 92,075 l4No information on the break-down by types o f employment was available for this category 52 75. The NYEP is costly, with an annual .budgetseveral times larger than that of the Ministry oflabor itseg Originally the Government of Ghana has planned an earmarked annual allocation of 1,300 million cedis (US$120 million) to finance NYEP activities. Based on the available data, the NYEP budget allocations for 2007 are estimated at 677,000 million cedis or about 0.7 percent of GDP, which corresponds to about six times the total budget of the Ministry of Manpower, Youth, and Employment (estimated at 103,000 million cedis for 2007). Since the program's launch in October 2006, the government has spent about 445,000 millions cedis (US$42 million) to benefit 92,075 young people, or about US$450 per beneficiary (see Box 8)15. Box 8: NYEP financingsources The program is financed from four main sources: (i)specialized funds and national programs, such as Poverty Alleviation Fund, HIPC, Road Fund, Ghana Education Trust (GET) Fund, National Health Insurance System (NHIS), Women Development Fund, Food and Agriculture Budget Support Funds; (ii) cost-sharing schemes and collaborative funding by district assemblies common funds (DACF), government agencies, civil society organizations, etc.; (iii)h d s recovered from the program participants; and (iv) other state sources. An earmarkedamount from each specializedfund i s usedannually to financing various NYEP employment modules. However, the budget execution is well below the planned allocation: in 2006, the NYEP received 270 billion cedis (US$25 million) out of 552 billions cedis planned for the program; and inthe first half of 2007, about 22 percent of the budget allocatedto the program was actually executed. Source: Ministry of Manpower, Youth and Employment, May 22, 2007. 76. The design of the program raises several concerns that can affect the labor marketfunctioning: & The level ofwages paidby the program (ranging between500,000- 2,000,000 Cedis per month16), varying by level of education of beneficiaries, is very high (between2 to ten times the national minimumwage). This can create (i) disincentives for youth to look for jobs outside of the program; and (ii) create unrealistic expectations among youth andtheir future earning prospects. P The targetedgroups are mainly youthwith secondaryandhigher education and frombetter-off urbanareas (See Annex 7). l6Wages paid by the program depend on the educational level of participants: 2,000,000 cedis (graduate degree); 1,500,000 cedis (bachelor degree); 1,000,000 cedis (higher nationaldiploma -HND); 800,000 cedis (SSS); 500,000 cedis (JSS). However the majority of programbeneficiaries are HND and SS graduates. 53 P The programmainly creates temporaryjobs inthe public sector (e.g. education and health services) and there are concerns on their sustainability. 77. The impact of the NYEP on poverty seems to be minimal. An analysis using GLSS data (see Annex 7) shows that the poverty impact o f the program is likely to be very limited, especially in comparison to its cost, This is because most o f the potential beneficiaries belong to the better-off quintiles o f the population, given the fact that the program targets youth with at least ajunior secondary education completed. Inrural areas, on the other hand, other types o f programs also providing employment such as public works could be considered to target individuals belonging to households who tend to be poor. At the current wage o f the NYEP one can conjecture that with its current size, the program i s reducing poverty at the national level by slightly more than half a percentage point, which is small incomparison to cost. A self-targeted public works program inrural areas with the same budget, lower wages, and no selection criteria in terms o f education, could probably have an impact five times larger at the same cost as the NYEP. 78. International experience provides useful lessons to improve the relevance of youth programs. Overall international experience suggests that better longer-term results could be achieved if employment programs are combined with other education-related interventions. The inventory o f youth employment interventions carried out by the World Bank provides some insights on how other countries deal with the issue of youth unemployment (see Box 9). Among the most important lessons learned is the fact that youth interventions should involve the private sector from the beginning to ensure an appropriate match betweenthe skills provided by youth programs and employers' needs. Box 9: Lessons from the Youth Employment Inventory The Youth Employment Inventory is a World Bank inventory of 289 programs and interventions from 84 countries designed to integrate young people into the labor market. Based on these premises and the programs' primary purpose, the inventoryidentifiedthe following categories of interventionsapplyingto Sub-SaharanAfrica: 1. Public works program could be more suitable than formal sector wage subsidy programs for developing countries. The review shows that wage subsidies do not go far in developing countries with a small size of the wage sector. Creatingjobs through public works offer alternativepathways to employment with investments in infrastructure continuingto be important in the near future. Public work programs provide good opportunitiesto acquire initial work experience, especially for low-skilled andrural youth. 2. Targeted youth entrepreneurship schemes improvechances for young entrepreneursin low-incomecountries where job growth inthe formal economy tends to be slow. The evidence, however, indicatesthat not all youth will be suited to self-employmentand failures rates can be high. Careful targeting and screening for these programs will be importantto success and cost-effectiveness. 3. Training programs are most successful when they involve the private sector in providing practical work experience and in identifyingthe kind of skills required. Engagement of the private sector in training is an effective tool to mitigate the risk of high-cost training disconnected from market demand and to increase on-the-job training. Targetingof training programsvaries from disadvantagedyouth to the longterm-unemployed, underemployed, or simply job seekers in urban and rural areas. Training programs are among the most costly of interventions for youth, and therefore attentionto connecting trainingwithjobs is importantto success. 4. Traditional apprenticeships can prove to be a cost-effective alternative for especially lower skilled youth. Throughout West Africa, traditional apprenticeships between a master craftsman and apprentice are a popular source of skills. The main strengths of traditional apprenticeship are its practical orientation, se(f-regulation,and self-fiancing. These apprenticeshipsalso cater to individualswho lackthe educationalrequirementsfor formaltraining (see Chapter 3). 5. Comprehensive interventions including education equivalency programs, training, counseling, job search assistance and other services lead to better outcomes. Such programs are among the more successful, though expensive, approaches for improving the employability of disadvantagedyouth. Multi-service programs are common in 54 LatinAmerica and the U.S. In Latin America, the J6venes Programs have been widely analyzed using a welfare framework and cited as a successful story in assistingyoung workers in developing countries. They use a demand-driven model that targets economically disadvantaged youth, fosters private sector participation, and promotes competition amongtraining providers.Although successfulin improvingjob placementand earnings,the programsbecame expensive for some countries and have been replaced by smaller and more focused interventions.Comprehensive programs have also worked in sub-SaharanAfrica. Source:Global Inventoryof Interventionsto Support Young Workers; WB, 2007. 79. The NYEP's monitoring and evaluation system is weak and there is no information on the project's impact on youth employment. Given the multi-sectoral nature of the program and its huge costs, it i s crucial to put in place a strong monitoring and evaluation system to monitor results on the ground and ensure efficient use of public resources. Currently the information collected on the beneficiaries is not sufficient to evaluate the outcomes of the program. ANOVERVIEW OFFINDINGS POLICYIMPLICATIONS AND Labor regulations and institutions 80. The changes introduced by the new law are steps in the right direction, but further revisions of the labor code are needed to ease dismissalprocedures that are still complex and costly. Ghana hiring and firing rules are overall flexible, but firing costs can be high and the procedures are cumbersome. There is still need to better understand the dismissal procedures and costs in order to suggest further legislation reforms, if necessary. Based on lessons learned from other countries, hiring and firing regulations need to find a balance between providing security to workers and reducing business costs. High dismissal costs can hinder job creation and slow the flow of FDI into the country. Although regulations on hiring and firing are usually justified, as they protect workers from arbitrary actions and provide stability of employment, onerous employment protection legislation can discourage formal job creation because firms will be reluctant to hireworkers ifthey face significant costs inadjusting the workplace to changes indemand. 81. The minimum wage policy does not seem to constrain job creation because it is not strictly enforced, but if it were enforced, costs could be high due to its more than doubling in real terms in the last 15 years. According to the 2003 Labor Act, the National Tripartite Committee annually sets the daily minimum wage, which is currently equal to $19,000 (US$1.7). The GLSS data show that the minimum wage i s high in comparison to prevailing wages, in the sense that an increasing share of workers now has wages below the minimum wage, especially among youth and in the informal market. - While this may not have had negative impact due to weak enforcement, the continuous increase in the minimum wage at a rate much faster than inflation may cause problems in the future, so that a reassessment of the level of the minimum wage is warranted, especially interms of its potential impact on access to formaljobs for youth. 82. The impact of the ongoingpension reforms on the labor market, including non- wage labor costs, needs to be carefully evaluated. Overall, the current non-wage labor costs inGhana are moderate comparedto the region, butthe Government has to ensure that 55 the new pension reforms do not increase non-wage labor costs since it would decrease employer incentives to hire new workers as well as reduce incentives for workers to participate in the labor market. It i s advisable that during the preparation o f the pension reforms, the authorities assess the impact o f various pension reform options on the labor market overall and on labor costs inparticular. 83. A better understanding of the Industrial relations is needed. Recent changes in the labor law and shifting responsibilities between unions and state institutions require a better understanding o f the nature and dynamics o f industrial relations in Ghana. This task, which i s a responsibility o f the Government, employers, and workers, would help in developing a modernized institutional framework that would match Ghana's ambitions for becoming a middle-income economy. It will also help the Government refine the responsibilities and roles o f various institutions and identify the areas that need further strengthening and institutional capacity building. Employment Programs 84. Addressing youth employment requires both short and long term solutions. While cost-effective employment programs could provide short-term support to youth employment, long term solutions would require investmentsand policy changes that could contribute to economic growth as well as investments in education and training to better prepare future youth for more productive jobs in the growth sectors o f the economy. International experience shows that training alone i s not sufficient to ensure the transition o f young people from school to work and it i s important to link training to the real needs o f the economy, particularly inthe private sector (see Chapter 5). 85. To improve the effectiveness o f the existing NYEP, the following measures could be considered: k Reviewing the design of the program to better achieve its objectives. The attention given to youth employment is important and based on international experience focus on this group i s justified. However, the Government needs to reviewthe design o fthe NYEP to make sure that: (i) programtargets vulnerable the youth well beyond urban areas; (ii) the level o f wages paid by the program are below the market wages in order to avoid distorting the labor market; and (iii) temporary employment offered by the program i s not de facto disguised public sector employment and involves private sector. As part o f this review, it would be helpfulalso to revise the concept o f youth since the current definition of youth (18 to 35 years old) i s too broad and includes diverse groups: adolescents (18-19), youth (20-24) and young adults (25-35). These different groups are in different stages inthe transitions from child to adulthood and as such require different policy andprograms. k Strengthening the program's monitoring and evaluation system to ensure that thepublicfunds are spent efficiently and that theprogram delivers long term and sustainable results. The existing youth program in Ghana i s relatively new and 56 there i s not detailed information on its beneficiaries and their socio-economic characteristics and there i s no information on its impact. However, given the program's high costs, the Government needs to: (i)improve the monitoring indicators; (ii)undertake sound evaluations to ensure that funds are being effectively spent; (iii)monitor closely high expenditure modules o f the program in order to see whether the priority given to them i sjustifiable; and (iv) make sure that the beneficiariesmeet the selection criteria for the program. P As part of the poverty alleviation strategu, considering introduction of sev- targeted public work schemes to offer temporary employment opportunitiesfor youth. Given infrastructure needs inthe country, the Government will continue to invest significantly inpublic works innear future. This provides an opportunity for low-skilled youth in both urban and rural areas to acquire initial work experience and a temporary income, while at the same time constructing/maintaining needed infrastructure. Developing innovative mechanism for road maintenance and other public services through SME also hold the potential to absorb both unskilled and skilled youth. Based on lessons learned from other countries, the public work schemes should include: (i) self targeted mechanism to ensure participation o f both low-skilled and high-skilled youth (i.e., level o f wage below market); and (ii) training for basic skill development o f low skilledyouth. Based on the design o f the program, this safety net scheme can play an important role in poverty alleviation strategy. 57 CHAPTER 5. EDUCATION, SKILLS,AND LABORMARKET OUTCOMES Trade, rapid advances in science and technology, and intensiped economic competition are shaping the demandfor skills in countries worldwide. Theseforces have increased the attention given to the ability of education and training systems toprepareyouth for entry to the world of work and to support more seasoned workers in adjusting to structural changes takingplace in labor markets." Concern exists whether skill deficits have or may become a constraint to Ghana'sfurther growth and capacityfor reducing poverty. Noting the limited opportunitiesfor skills development beyond basic education, a WhitePaper wasprepared in 2004, building on theEducation StrategicPlan (ESP)for 2003-2015 prepared by the Ministry of Education Science and Sports (MOESS) and callingfor increased emphasis on technical, vocational, and agricultural education and apprenticeship.18 I n a sector review, MOESS refers to evidence of a widespread disparity between what education institutions produce and what the labor market wants.19 This chapter begins with an overview of the landscapefor skills development, starting with basic education and moving on to consider optionsfor further education and skills development, comparing school-based and post-school options for training and highlighting issues of access, quality, efficiency andfinancing in variousprograms. Thesecond section estimates returns to education and training, lookingfor evidence of growing skills gaps, and the third section reviews the ESP recommendationsfor sector reforms involving skills, and benchmarks recommendations against regional and international experience,emphasizingpriorities. Overall, the analysis shows that though Ghana has made substantial investments in basic education, improvements in quality are necessary. Furthermore, education and skills training are influencing wages and earnings directly by raising the productivity of the worker and indirectly by promoting entry into more lucrativeforms of employment. Therefore improving skills development is key to the attainment of middle-income status. To this end, Government needs to continue improving the quality of basic education while opening access to post-basic education and improving articulation betweenformal education and trainingprograms, including traditional apprenticeships. Government can play an important role in promoting the efficient operation of training markets by engaging employers in setting skills standards,focusing on developmentof a quality assuranceframework, and linking publicfinancing of training to outcomes. 17 A review of global experience and the role played by skills development in helping youth make the transition from school to work is available inAdams (2007). 18Ministry of Education, Science and Sports (2004). 19Ministry ofEducation, Science and Sports (2006). 58 THELANDSCAPE FORSKILLSDEVELOPMENT Basic Education 86. Ghana has made progress in increasing young people's access to primary education aspart of its efforts to meet the Millennium Development Goalsfor education. Capitation grants were first piloted in 2004/05 and mainstreamed in 2005/06.20 As a consequence, net enrolment rates of youth 6-11years of age increased nationally from 59 percent in2004/05 to nearly 79 percent in2006/07, with net admission rates for 6 year olds more than doubling to 69.3 percent from the 2004/05 figure of 26.2 percent. Gross enrolment rates, including over-age and under-age youth, rose to nearly 91 percent in 2006/07 (see Table 27). The increase in the net enrolment rate for primary education to 78.6 percent pushedGhanaabove the Sub-SaharanAfrica average o f 66.3 percent.21 Table 27: Enrolment by Level of Schooling for 2004/05 and 2006/07 IN e t Enrolment I59.1 I78.6 I31.6 I 50.7 n.a. I10.6 87. National primary enrolment rates, however, hide disparities across and within regions and between sexes, and efficiency remains low. Within the country's 10 regions, the Northern and Upper West regions show the lowest net enrolment rates, with 68 and 70 percent respectively in2006/07. The net enrolment rate of young girls inprimary education trailed the rate for young boys in2006/07, 77.3 percent to 79.8. Girls' share o f enrolment ranged from 46.3 percent inthe NorthernRegion to 51.2 percent in Greater Accra. As an indicator of the system's inefficiency, for every 100 students enrolled in the first grade, approximately 56 complete the sixth grade. The MOESS estimates it will not meet its target for universal primary completion by 2012, preventing achievement of universal basic completion by 2015. 88. The growth in enrolment in public primary schools has put pressure on teacher training and assignments, as well as on school infrastructure. The expansion in enrolment in public schools led to new teachers being hired. Many, however, lacked appropriate training, and the percentage of trained teachers fell from 67 percent in2005/06 to 62 percent in2006/07. Inprivate schools, less than 20 percent of teacherswere properly trained. Pupil-teacher ratios inpublic primary schools rangedwidely in2005/6, from 29.4 pupilsper teacher inthe Volta Regionto 42.8 inthe Upper East Region, with the range by District even wider. Furthermore, students in public primary schools face crowded conditions, and infrastructure that needs repairs. Nationally, seats available per 100 20For every child enrolled in basic education, a school receives 30,000 Cedis, about US$3.10 to offset a reduction inschool levies. 21This section is based on data drawn from Ghana's Education Management Information System (EMIS) and the Preliminary Education Sector Performance Report 2006, using EMIS data preparedby the Ministry of Education, Science and Sports. 59 students fell from 90 to 80 in 2006/007, and twenty-seven percent of public primary classrooms were inneed of major repairs. 89. Junior secondary schools (JSS), which have also seen a rise in enrolments,face many of the same difficulties. Net admission rates for 12 year old first-year JSS students in 2004/05 surged from 12.2 to 44.4 percent in 2006/07. Much of the expansion came from the increased number of students completing primary and moving on to JSS (93% of primarycompleters). As with primary education, the accessibility and condition of schools varies across regions and districts. Gender parity continues to be an issue. Teachers lack appropriate credentials, especially in private schools (which account for 15.9 percent of JSS level schools). In public schools, nearly eight of every ten teachers have completed Teacher Training College, but only 27 percent meet this standard among private JSS. Facilities are also in need of repairs and the growth of JSS enrollments has produced crowding. For every 100 studentsentering JSS, 65 are expectedto complete JSS. 90. These conditions are reflected in low pass rate for those taking the Basic Education Certificate Examination (BECE), which determines whether or not a student is able to continue to the second cycle of education. The BECE i s currently the only indicator measuring achievement and, thus, the quality of basic education. Sixty-two percent passedthe BECE in2005/06, a figure that has been relatively stable since 2002/03. The variation inpass rates across regions rangedfrom 47 percent inthe NorthernRegionto 70 percent in the Greater Accra Region. Boys performed better than girls. Those failing the BECE join others who drop out or simply never enroll and who now must enter the labor market with limited education and skills. A 2003 survey by the Ghana Statistical Service found a significant number o f children on the street engaged inlabor who were not attending school. The proportion increased from 8 percent among 5 to 9 year olds to 54 percent for those ages 15 to 17. Figure 14: Public Expenditure on Primary Educationas a Percent of Per Capita GDP: 2005 20 u 15 C g 0, 10 n 5 0 60 91. Despite keeping pace with Table 28: Governmentof GhanaEducationExpenditures by Sub. international trends in education Sector: 2004 spending, Ghana faces a continued needfor investment in basic education, and also faces Sub-sector Recurrent Investment growing demands for investment YO % in post-basic education. Ghana Pre-School 5.1 has matched international norms in Primary 35.5 Junior Secondary 22.3 spendingon education, with actual Senior Secondary 13.1 47.3 expenditures in 2006/07 reaching Non-formal Education 1.O 6 percent o f GDP, and spending Special Education 0.6 3.4 on primary education accounting Teacher Education 4.7 2.6 for 12.6% o f its per capita GDP TVET 1.o 2.7 (roughly average for the middle- Tertiary 16.7 7.8 income countries listed in Figure Subvented Agencies 3.0 14). Despite rising enrolments, Mgmt 33.2 increased spending on primary HIV/AIDS education has led to an increase in Total ____-_________________________ 100.0 100.0 spending per student (Figure 15). Source: Government o f Ghana However, the primary education sector still has need for more, and at the same time, Government has to respond to the growing numbers o f youth expecting to enter a senior secondary school. Furthermore, the evidence above shows that less thantwo of every three students successfully complete primary education and JSS. This emphasizes the importance o f improving the efficiency o f resource allocation. Moreover, based on some preliminary analysis, it seems that about 8 percent o f total investment in education is lost to migration as a result o fthe brain drain(See Box 10 and Annex 6). Although some o f this investment may be recuperated through the returns from the migrants, further work i s neededto assess the net effect o fthe brain drain inGhana. Figure 15: Governmentof Ghana Per Capita Expenditureson PrimaryEducation:2004 2006 - 2004 2005 2006 Source: Government of Ghana 61 Box 10: Migration, brain drain and losses in public education spending Migration operates as an important equilibrating mechanism by allocating labor to regions and sectors where it is in most in demand. As such, migration, whether internal and international, can generate significant welfare gains for the migrants, their families and both the destination and the source economies. Since the majority o f the gains accrue to the migrants through higher wages, and to their families through the remittances sent back, migration may become an important tool for economic development and poverty reduction. However, international migration data in countries such as Ghana reveal that a significant share o f the migrants, especially to the OECD countries, has tertiary education. Given the importance of human capital in economic development and growth, brain drain i s likely to have important development and labor market implications. Who are the international migrants? According to preliminary data from GLSSS, a total o f 422,436 individuals or 1.91% o fthe population i s comprised o f international migrants. The most favored destinations for external migrants are the US, the UK and other OECD countries followed by other African countries. The most popular destinations within Africa are Cote d'Ivoire, Nigeria, Burkina Faso and Guinea, which are all neighboring countries. A total o f more than 288,000 households (5.3% of total households) have an international migrant. The vast majority o f households with migrants to the OECD countries (including the US and the UK) are from urban areas and belong to the fourth and fifth consumption quintiles.Families with migrants to African countries tend to be poorer and rural. Migrants to OECD countries are predominantly from Ashanti and Greater Accra whereas there i s significant number o f migrants to other African countries in the Volta, Eastern and Brong Ahafo regions. Profile of international migrants based on GLSSS. On average, migrants constitute a better educated and more skilled section o f the Ghanaian labor force, and education affects the place o f destination. Some 14 percent o f OECD migrants have tertiary education and another 75 percent have secondary education. This i s rather different than the averages for migrants to other African countries where 57 percent have secondary education and only 1 percent has post-secondary education. Another important indicator o f the loss o f educated workforce is the share o f educated workers at home and abroad. Only 22 percent and 2.15 percent o f the total population have secondary, and post secondary education within Ghana while around 70 percent o f migrants have secondary education and 10 percent have post secondary education. This suggests fixther that the present migration trends in Ghana are leading to brain drain and have the potential o f hrther aggravating the problem to low human capital at home. The data also suggests that a vast majority o f international migrants (82.8 percent) have completed their education inGhana. This means that a significant proportion o f the expenditure on education inGhana has gone to those who later migrated to other countries. Loss in education spending due to brain drain. Annex 5 combines household survey data with detailed information on international migrants and their education level with administrative data on public spending for education in order to compute the share o f total education investment that is lost due to migration. The migration module o f GLSS5 provides information on the highest level o f education achieved by each migrant in Ghana before they migrated. Thus the cumulative cost of education is calculated for each migrant by summing over the educational cost of each level o f education the person have obtained. Unit costs from education are derived from data available in the Preliminary Education Sector Performance Report o f the Ministry of Education and Sports inGhana. Due to lack of data intrends inunit costs over long periods of time, we estimate the investmentmade ineducation for all individuals, even older ones, usingthe 2004 unit costs. The total cumulative expenditure on education for the country as a whole i s the sum o f the education expenditure for all Ghanaians who have remained intheir country as well as the expenditure for those who have migrated (estimated at 49,116 billion cedis or $ 5 billion, Table 4 Annex 5). The total investment in education allocated to individuals who have migrated comes up to 4,3 12 billion cedis. Thus about 8 percent o f the total investment in education is lost to migration. Some o f this may be recuperated if some migrants come back in their country, but the evidence from the survey i s that the number o f return migrants is very small. Source: Joseph et al. (2008) See Annex 6. - 62 Senior Secondary Schools 92. Secondary school enrolments are expanding, but conditions are similar to basic education. Though enrolment in senior secondary schools (SSS) doubled between 1999 and 2006, net enrolment remains low at 10.6 percent. This means that nine out of every ten youth 15-17 years of age are not enrolled insenior secondary school (although some of these youth may still be enrolled in lower levels of education if they have been repeating one or more grades).. Many of these youth are out of school either employed, unemployed or out of the labor force, and for those who have left school, access to skills i s taking place mainly through apprenticeships, informal learning on the job, or non-formal training programs. Similar to basic education, opportunities for access are uneven across Ghana's 10regions, with net enrolment rates ranging from 5.3 percent inthe Upper West Region to 18.5 percent in the Central Region. Gender parity remains a problem: nationally, young women accounted for 44 percent of SSS enrolments in 2006/07, lower than that of basic education, with regional enrollment rates varying between 32.1 percent in the Northern Region and 47.5 percent inthe Central Region (Figure 16). As observed, the percentage of girls enrolled declines as they progressthroughschool. 93. For those enrolled in secondary education, quality is an issue. Inpublic schools (accounting for 91 percent of SSS enrolments) in2006/07, 15 percent o f teachers were not fully trained, and this figure rises to over 40 percent in private schools. One out of five public secondary schools required major repairs in 2006/07 compared with only three percent of private schools. Crowding exists in public secondary schools with 0.78 seats per student available in 2005/06, and 0.63 writing places available per student. The number of textbooks for each student ranges from one for every two students of science and social studies to just over one textbook for each student of mathematics and the Englishlanguage. 94. Most of the secondary students pursue general secondary education and the curriculum choices vary by gender. Slightly over nine out of ten secondary students pursue a general secondary education with less thanone out of ten following atechnical or vocational curriculum (see Figure 16 for data by region). O f those enrolled in general secondary education, more than one-third are enrolled in a business curriculum, and about the same share ingeneral arts. Elevenpercent pursuesa science education, andjust over 9 percent enroll inagriculture. Overall, girls are more likely than boys to enroll in secretarial studies, general arts, and home economics, while boys are found more frequently in agriculture, accounting, technical, and visual arts programs. 95. Like basic education, pass ratesfor secondary students are low. Pass rates on the Senior Secondary School Certificate Examination (SSSCE), the main instrument for assessment of learning and access to further formal education at the tertiary level, stood at 59 percent in 2005/06. Boys uniformly score higher than girls on mathematics, science, social studies, and Englishlanguage exams. Adding the students who fail to the numbers who withdraw from education or who never enter the system creates a weak foundation for building a skilledworkforce. 63 Figure 16: The Capacity of Public TVET Institutionsto Absorb the Pipeline of Junior Secondary School Enrollment by Region :2006/07 Total Western volt0 Upper West Upper East Northern h a t e r Accra Eastern Central Brow Ahafo Ashanti 0 2 4 Percent 6 8 10 12 Technicaland vocationaleducation and training (TVET) 96. Governmentprovides skills training through technical and vocational education and training (TWT) institutions. These institutions serve the need for intermediate, advanced, and technical skills with entry requirements varying from none to passing o f the BECE and even the SSSCE. There are two main public systems: (i) one i s handled by the Ghana Education Service (GES) in Technical Training Institutes under the MOESS; and (ii) otheristheNationalVocationalTrainingInstitutes(NTVI)runbytheMinistryof the Manpower, Youth and Employment (MMYE) (see Table 29). Inaddition, other technical ministries offer sector-specific training programs as do community (for-profit and non- profit) institutions. Public capacity accounted for 47,935 training places in 2006/07.22 Formal and traditional apprenticeships are also an important source o f training on-the-job (see below). 97. Public TWT institutions provide places for a small share of the potential demandfor training. The enrolment injunior secondary schools in 2006/07 i s a measure of the potential pipeline o f students who could advance to a senior secondary education and attempt to enroll in a TVET program. Figure 15 expresses public enrolment inTVET as a percent o f this pipeline for each o f the 10 Regions in Ghana. Nationally, 5 percent o f '*Ministry o f Education, Science and Sports, E M I S 2006, Table 3.2 64 junior secondary students can expect to find a place for study ina public TVET institution, with regional rates varying from 1.7 percent inBrong Ahafo to 10.6 percent inthe Upper East. If private TVET capacity i s included, the percentage rises to 7.2 percent. O f those students who actually advance beyond ajunior secondary education, 12.7 percent enroll in a TVET institution. 98. TWT reaches only a small share of the 1 Table 29: Types of TVET Institutions population, but the share has increased over the Public 188 past 15years. Using data from the Ghana Living GES Technical Training Inst. 24 Standards Survey from 1991/92 and 2005.06, just NationalVocational Training Inst. 38 over 2 percent o f youth 20 to 30 years o f age have Integrated Community Centers for 1 i1 been through a TVET program, but this Employable skills (ICCES) I 61 I percentage has doubled since 1991/92 (see Table Social Welfare Centers 30). Much o f this growth has come from the LeadershipTraining Inst. increased number o f young men enrolling in Opportunities Industrialization Centers TVET as their share in the 2005/06 survey Community Development Centers 24 surpassed that o f young women. As will be Auiculture Training Inst. confirmed later in this chapter when looking at IRoad& Transport Training Centers I 1 I the determinants o f different types o f Gratis Foundation 9 employment, TVET i s an important source o f Private I1252 skills for those who hold wage employment, ITotal I440 I especially those working in the private formal sector. It also i s connected with those creating their own jobs outside agriculture. Access to TVET, like that for secondary and higher education, favors those who come from better off families. The share o f those who have been through a TVET program rises uniformly with the level o f consumption for families. For example, the percentage o f those in the highest standard o f living quintile with a TVET background, 4.3 percent, is seven times the percentage o f those inthe lowest standard o f living quintile. 99. The upgrading of instructors is important to public and private TWT institutions, as is the improvement of workshops. Only about half o f TVET instructors hold the Teacher's Certificate A, with a slightly higher percentage in public than private centers, 57 versus 43 percent. About 84 percent o f teachers in public and private centers have obtained the minimumtechnical qualification o f a Technician I1Certificate or above. Of the public centers responding, 46 percent o f TVET instructors report they hardly ever have in-service training, and another 23 percent say they have this training only once a year. The figures for private TVET institutions are 43 and 23 percent respectively. Furthermore, in 2005/06, only 12 percent o f public training centers and 29 percent o f private training centers described themselves as "well-equipped". In contrast, thirty-seven percent o f public training institutions and 12 percent o f private institutions described their facilities as "poorly equipped" or with "no equipment." Public training sites in 2006/07 report that 18 percent o f their classrooms need major repairs compared with 37 percent o f private institutions. There are public and private institutions that offer high quality training, but these are exception^.^^ 23Suchas the private DonBosco training institutes inTema/Ashiamanand Sunyaniand some public training centers benefiting from Donors supports likethe Netherlands. See Palmer (2007a), Chapter 5. 65 Table 30: Share of youth (15-30 age group) gone through TVET programs 1991/92 2005106 Urban Rural Total Urban Rural Total Gender Male 1.9 0.3 0.6 5.1 0.7 2.0 . Female 2.4 0.4 0.9 3.6 0.7 1.6 Age group 15-19 0.0 0.1 0.1 0.0 0.0 0.0 20-24 1.3 0.4 0.7 3.5 1.2 2.0 25-30 3.4 0.4 1.4 5.9 0.8 2.5 Type of employment Wage Public Sector 6.8 2.5 4.8 5.6 2.2 4.4 Wage Private Sector Formal 4.2 0.0 2.7 9.6 4.6 8.3 Wage Privatesector Informal 3.1 0.0 1.9 2.6 1.7 2.3 Self-employedAgro, paid 0.0 0.3 0.3 1.7 0.4 0.4 Self-employedAgro, unpaid 0.0 0.1 0.1 2.3 0.3 0.4 Self-employedNon-Agro 1.6 0.7 1.2 3.5 1.4 2.6 Standard of living Lowestquintile 0.0 0.3 0.3 2.0 0.5 0.6 Second quintile 0.0 0.0 0.0 2.5 0.3 0.6 Third quintile 2.2 0.2 0.6 2.5 0.6 1.1 Fourthquintile 2.3 0.2 0.9 4.5 1.2 2.6 Highestquintile 2.6 1.1 1.9 5.6 1.7 4.3 All 2.2 0.3 0.8 4.3 0.7 1.8 Source:Authors usingGLSS data. 100. Programs availablefor helping youth make the transition to employment are not carefully evaluated. Rigorous evaluations using appropriate control and treatment groups are not available for training programs. Moreover, only a few programs have introduced tracer studies to track placement and earnings o f graduates, and those that have lack adequate coverage. What evidence i s available is largely anecdotal and i s generally unfavorable, as reported in the White Paper. Furthermore, public training capacity in Ghana, like that inmany other countries, is heavily focused on the skill needs o f the small formal sector and, with few exceptions, does not address the need for entrepreneurial skills inthe muchlarger informal economy.24 101. Financing good quality TWT is costly, and spending on TWT remains low. The recurrent unit cost o f TVET in 2006 was 5 times higher than that o f a primary education and 2.7 times that o f a senior secondary education. Internationally, the unit cost o f TVET averages 2 to 3 times that o f a secondary education due to the cost o f workshops and lower teacher-pupil ratios. While falling within international norms, the unit cost for TVET hides quality issues. In 2004, this unit cost was only a fraction higher than that of 24Haanand Serriere (2002). 66 senior secondary education, implying there were inadequate levels o f spending at that time. The higher unit cost ratio in 2006 is actually the result of enrollment in TVET declining from 21,424 in 2004 to 18,005 in 2006 creating underutilized capacity and higher unit costs. The MOESS allocates one percent o f its education budget to the 14 percent o f secondary students enrolled in TVET and the MMYE spends about 12 percent o f its budget on VTIs (see Table 28). Smaller sums are spent by other technical ministries, like Roads and Local Government. Only 7 percent o f donor spendin on services went to TVET in2006, andthere was no allocationfor investment inTVET. h Figure 17: Unit Cost of Educationin Ghana: 2006- ~ ~ ~ . - [-`~-5=OO r 4500000 4000000 3500000 .- 3000000 `0 U a 2500000 2000000 1500000 1000000 500000 1 0 Source: MOESS EMlS2007 Primary JSS sss TVET 102. With the diversity of TVETproviders, it is surprising tofind little coordination in the use of scarce public resources. The National Coordinating Committee for Technical and Vocational Education and Training (NACVET) under the MOESS i s charged with coordinating public and private providers, but there is little evidence o f success with many different bodies and vested interests involved.26 The legal framework surrounding the TVET sector is itself fragmented, and UNESCO reports a "dizzying array" o f examination^.^' The lack o f a comprehensive policy for TVET development i s apparent and is addressed inthe ESP with the introduction in2006 o fthe Council for Technical and Vocational Education and Training Act with a new coordinating body. Formaland TraditionalApprenticeships 103. Enterprises in Ghana, as in most countries throughout WestAfricd', are active trainers offering formal and traditional apprenticeships and training on-the-job. Formal ~ 25Ibid. Japan is the only major development partner currently engaged in TVET with its focus on helping develop the policy framework for TVET. JICA is currently upgrading five workshops, with plans to upgrade another five in2007. 26UNESCO (2003). 27Ibid. 28JohansonandAdams, 2004; Batra, 2000; Tan and Savchenko,2003; Tan 2005. 67 apprenticeships are found in the wage sector of the economy, which is small in Ghana, accounting for about 16 percent o f employment based on the 2006/07 Ghana Living Standards Survey.29 These apprenticeships are coordinated by the NVTI. Traditional apprenticeships, found inthe informal sector of the economy, consist of private contractual arrangements between a parent or apprentice and a master craftsperson who agrees to provide practical training in the workplace, ranging from several months to three or four years, and subsequently certify the training in return for a fee or reduced earnings while learning,30 104. Enterprise-basedtraining reaches morepersons thanpublic training institutions. While numbers are difficult to come by, Atchoarena and Delluc (2001) report 80 to 90 percent of all basic skills training comes from traditional or informal apprenticeships in Ghana, compared with 5 to 10 percent from public training institutions. A 2006 urban- based labor market survey in Ghana conducted by the Centre for the Study of African Economies (CSAE) in conjunction with the Ghana Statistical Office found one-third of those 16 to 65 years of age had some form of training.31 Apprenticeship was by far the most common form with 55% being either current or past apprentices. Enterprises in the formal sector also provide short-term training to workers beyond formal apprenticeships, helping upgrade skills and introducenewtechnologies. On-the-job training was the second most common form of training inthe 2006 survey with 25% having received such trainin The training offered by enterprises favors those with higher levels of general education. 5 Formal vocational training trailed in importance with 16% receiving their training from this source. Almost all training programs taken at VTIs can be mastered through traditional apprenti~eship.~~ 105. Apprenticeship has become more important over time as a means to acquire skills for employment, especiallyfor young women. According to the Ghana Living Standards Surveys of 1991/92 and 2005/06, the percentage of youth 20 to 30 years o f age participating in an apprenticeship has increased over the survey period (see Table 30). Currently, about one of every three youth in this age group has experience as an apprentice. The percentagefor young women overall has doubled during this period, a rate higher than that of young men.. Education influences the choice of an apprenticeship. A much smaller percentage of those with a senior secondary or post-secondary education are likely to choose an apprenticeship, while those with a primary or junior secondary education are more likely to participate in an apprenticeship. This is a pattern backed up by the CSAE 2006 urbanlabor survey.34While those with an apprenticeship are present in all types of employment, they are more frequently found in non-agricultural self- employment and wage employment inthe informal sector. Significant numbers also show 29Coulombe and Woden 2007. 30Haan and Serriere (2002) estimate that fees for traditional apprenticeships average about US$70 per year. ILO in2006 estimatedfees to average US$160, ranging from US$22 to US$616. In2005, Palmer estimated the averagefee inthe Ashanti Regionto be $42 with arange from $13 to $173. 31Monk, Sandefur and Teal, 2007 (see Annex 2). 32Rosholm, Nielsen, and Dabalen, 2007. 33Haan and Serriere (2002). 34Monk, Sandefur and Teal, 2007 (see Annex 2) 68 up inwage employment inthe private formal sector, Like TVET, apprentices do not come from households with the lowest levels o f consumption. Table 31: Share of youth (15-30 age group) gone throughApprenticeships 1991192 2005106 Urban Rural Total Urban Rural Total Gender Male 36.0 18.1 21.9 46.7 19.0 27.3 Female 21.1 11.1 13.9 40.7 20.1 26.3 Age group 15-19 10.6 5.2 6.1 30.9 6.9 11.1 20-24 34.5 16.4 20.8 44.3 24.4 31.0 25-30 28.8 21.4 23.7 46.2 24.3 31.9 Education level No Education 15.1 5.8 7.5 21.4 6.9 8.7 Some Primary 16.7 15.9 16.0 42.1 16.6 21.4 Primary 25.4 10.5 13.3 54.1 25.5 32.6 Secondary (lower) 38.5 25.9 29.9 60.4 43.2 50.9 TVET 22.2 12.5 19.2 20.0 19.2 19.8 Secondary (higher) 14.7 13.5 14.3 29.6 17.5 25.7 Post Secondary 13.3 11.1 12.1 7.6 3.4 6.5 Type of employment Wage Public Sector 9.1 12.7 10.8 12.4 10.9 11.9 Wage Private SectorFormal 50.7 56.4 52.7 28.0 42.5 31.6 Wage Privatesector Informal 53.8 57.5 55.2 36.0 49.6 41.2 Self-employedAgro, paid 12.2 16.3 16.0 32.2 18.8 19.7 Self-employedAgro, unpaid 16.2 7.6 8.4 22.1 8.4 9.2 Self-employedNon-Agro 27.0 24.2 25.7 56.8 44.1 51.2 Standard of living Lowestquintile 12.1 8.4 8.6 30.3 9.4 10.9 Secondquintile 6.9 10.3 9.9 36.2 15.2 18.3 Third quintile 25.9 14.0 16.4 35.9 25.1 28.0 Fourthquintile 27.0 17.1 20.3 43.4 31.5 36.7 Highestquintile 31.4 28.4 30.0 49.6 41.3 46.9 All 26.9 14.4 17.5 43.5 19.6 26.8 Source: Authors usingGLSS data. 106. Since traditional apprenticeships form a major source of skills for youth in Ghana, their quality is an issue.35 In many countries and business environments, traditional apprenticeships serve the informal sector well, but are proving too narrowly focused to cope with the increasing pace o f technological change, skills upgrading, and expanding markets.36 Although traditional apprenticeships have not been carefully evaluated in cost-benefit terms, their main strengths lie in their practical orientation, self- 35Hams and Serriere (2002). 36Ziderman, (2003). 69 regulation, and self-financing. They cater to individuals who lack credentials for formal training, serve important target groups (rural populations and urban poor) and are generally cost-effective. However, traditional apprenticeships have many weaknesses: (i)they are gender-biasedby occupation; (ii) exclude applicants from the very poorest households they unable to finance modest fees; (iii) they are based on traditional technologies as master craftsmen fail to keep up with technological change; and (iv) they depend on the master craftsman for their standards and quality assurance, whichvaries widely. 107. Almost all apprentices and masters lack formal vocational or technical training andfew have more than 9years of formal schooling. Labor force data for Ghana in2000 show that there were 207,047 economically active persons (15 years and older) in apprenticeship training.37 Over three-quarters of apprentices were 15 to 29 years of age, and 57 percent were males. Males were mainly inauto mechanics, carpentry, tailoring and driving, while females were primarily in dressmaking, hairdressing, and catering. Out of the total, only 5.3 percent had formal vocational or technical training.38 According to the GLSS data, most apprenticeships are paid by parents or by youth themselves (either incash or in kind) and the average length of the apprenticeships is about two to three years to complete. The 2006 CSAE urban-based labor market survey shows that the most common pattern for apprentices i s to enter an apprenticeship after the end of junior secondary school. O f those who had been apprentices in the past, 74 percent entered at the junior secondary level. While 10 percent of apprentices had completed primary school and 9 percent had no education, only 7% had a secondary education or higher.39This pattern i s consistent with the results of the GLSS data (see Table 31). EconomicReturnsto Educationand Training 108. Education influences wages and earnings directly by raising the productivity of the worker and indirectly by promoting entry into more lucrative categories of employment. This section estimates the economic returns to education and training using three Ghana household living standards surveys (GLSS) conducted in 1991/92, 1998/99, and 2005/06.40 As a guide to future investment in skills, the returns highlight variation across levels of education and for different modes of training (covering institution-based TVET and apprenticeship^.^^). We look for indications o f a tightening labor market for 37Ghana Statistical Service, (2005). 38Ibid. 39Monk, Sandefur and Teal, 2007. See Annex 2. 40The specification o f the earnings function expresses education in categorical form, including TVET, along with experience and other job and demographic characteristics. Measures o f apprenticeship by level o f education are also included in binomial form testing for differences in returns to apprenticeship by level o f education. Separate regressions are estimated for urban and rural areas to allow for interactions with education. Regression estimates are based on ordinary least squares methods controlling for selection bias. The regression coefficients for each education category reflect the percentage difference in log earnings from the omitted category o f no education. These regressions can be found inthe background paper prepared for this chapter by Adams, Coulombe, Wodon, and Razmara (2008) andattached as Annex 1. 41 For a historical review o f the economic returns to education in Ghana, see Palmer (2007b). Recent evidence points to private economic returns for post-basic education that now exceed those for primary education. 70 skills as reflected through returns to investments in education and training in the three survey periods. impact of education on wages and earnings and Table 32: Median Earnings per Hour estimating the likelihood of entry into each of three Workedin2005 employment categories: wage employed, self- employed agriculture and self-employed non- Employment Status agriculture. I s higher education, for example, a factor Wage workers 6,925 influencing entry to wage employment? What role do Wageworkerspublic sector 3,930 TVET and apprenticeship play inthis choice and self- Wageworkersprivate formal Wageworkersprivate informal 2,782 employment? Second, attention turns to estimating Self-employedAgriculture 2,033 the direct effects of education on wages and earnings, Self-EmployedNon-Agrjculture 2,713 controlling for other individual and family All 2,060 characteristic^.^^ The sample covers 25 to 64 year olds, excluding youth due to problems in coding Source: GLSS2005106 110. The three employment categoriesexhibit different mean earnings with additional differences observedfor wage workers in the public and private sectors (see Table 32). Wage employees in the public sector earn 2.6 times the earnings o f the self-employed in non-agriculture and 3.4 times the earnings o f the self-employed in the agriculture sector. Inturn, wage workers inthe public sector earn more than their counterparts inthe private formal sector and the private informal sector. These differences are unadjusted for the education and training profile o f workers inthe different employment categories. 111. Workers in the three employment categories each have different education and skill profiles. Workers inall three employment categories show rising levels o f education over the period with wage workers tending to have higher levels o f education than both categories o f the self-employed. See Table 1 in Annex 1. For a self-employed worker, a junior secondary education or less tends to be the educational norm, while for a wage worker the norm i s likely to be a senior secondary or even a post-secondary education. A smaller share o f wage workers and the self-employed in non-agriculture have a TVET background, but this share has also increased over the three surveys. Apprenticeship features prominently in the background o f wage workers and the self-employed in non- agriculture, more so thandoes TVET. 42The impact o f education on the choice o f employment category and on wages and earnings is done using a two-step procedure. We treat the employment choice as endogenous and first analyze this choice by using a multinomial logit procedure and then we analyze the determinants o f wages and earnings taking into account selection bias fiom the multinomial logit equation. For a more detailed explanation o f this methodology and the procedures for controlling for selection bias see Annex 1. 71 112. Important changes have taken place over the three surveys with employment becoming lessformal, less unionized, and less likely to be in thepublic sector. The share o f employment in each o f the three employment categories (see inAnnex 1, Table l), has remained fairly stable over the survey period. However, other indicators show that important changes have taken place in employment, especially among wage workers. For wage workers, the share o f employment in formal sector jobs has dropped. Consistent with this pattern, the percentage of unionizedjobs has declined along with the percentage o fjobs inthe public sector. The share o f wage employment inthe public sector alone has fallen from 64.5 percent in the 1991192 survey to 35.7 percent in the 2005/06 survey. Thus, while education and skill levels have risen in employment, there is evidence that much o f this employment has shifted to the private sector and i s no longer unionized. Employment inall three categories has become more urbanized over the survey period. Educationand the Type ofJob Held 113. Education influences earnings indirectly through the type of job held. Figures 17 through 20 show the influence o f education on the type o fjob held (See Tables 2 and 3 in Annex 1 for coefficients from the multinomial logit functions). As shown inFigure 17, the likelihood o f becoming a wage work in urban areas increases with the level o f education, controlling for other personal characteristic^.^^ This pattern, however, does not hold for non-agriculture self-employment inurbanareas as seen inFigure 18. Instead, those with a basic education are more likely to be self-employed inthe non-agriculture sector thanthose with higher levels o f education. The larger effect of basic education observed in 2005/06 could well reflect the growing importance o f this education to being able to establish one's own business, but it i s also consistent with the hypothesis o f expanding numbers o f youth completing a basic education who are unable to find wage employment and are left to create their own employment. The results do not allow us to choose between these competing hypotheses. 114. The relationship between education and the type of employment held is weaker in rural areas than in urban areas. The likelihood o f finding wage employment in rural Ghana, shown in Figure 19, rises with the level of education. The relationship o f education to wage employment in rural areas shows evidence o f strengthening in 2005/06 with the chances of finding wage employment with a post-secondary education increasing sharply. Though not reported here, the influence of education on the likelihood o f becoming self-employed in agriculture, primarily in farming, i s weaker than that for self- employment in non-agriculture. In general, having less than a secondary education raises the chances o f self-employment inagriculture, but by substantially less than the chances o f self-employment in non-agriculture. Investing in education is generally associated with movement out o f the agriculture sector to wage or self-employment in the non-agriculture sector. 43The logit function coefficients are relative odds ratios, andthus, a value of 1.321for a primary education in the wage earner regressions for 2005/06 means that workers with a primary education have a 32.1 percent higher likelihood of being a wage earner than a worker with no educationat all. 72 115. TVET exerts a positive influence on wage and self-employment. The likelihood of holding wage employment in Figure 17 is slightly higher for those with a TVET education than for those with a senior secondary education. Similarly, by providing a technical skill, TVET increases the chances of establishing one's own job in the non- agriculture sector, as reflected in Figure 18, although the effect appears to weaken in 2005/06. TVET's impact on the likelihood of obtaining wage employment inrural areas is generally lower than that in urban areas. However, it appears to have become a major factor influencing non-agriculture self-employment in rural areas, as observed in Figure 19. The previous section on labor supply identified the weak capacity of TVET in school settings, includingthat offered by the MOESS and MMYE, but the results here confirmthe importance of this training to finding wage employment or becoming self-employed. TVET exerts a positive influence on the probability of both types of employment. Its effect is particularly strong affecting the chances for obtaining wage employment. Figure 18: Impact of Educationon the Likelihood Figure 19: Impact of Educationon the Likelihoodof of Wage Employment(vs. no Education), Urban Non-Agriculture Self-Employment (vs. no Education), Urban 16 .6 I .4 1991/92 .2 1998199 0 .2 F4/q .6 .8 -1 73 Figure 20: Impact of Educationon the Likelihood Figure 21: Impacty of Educationon the Likelihood of Wage Employment(vs. no Education), Rural of Non-Agriculture Self-Employment (vs. no ~~-~~ Education), Rural -- 111 ~ 7n I l2 I I V 60 50 40 30 .....I" I1991/92 1991/92 1998/99 1998/99 2005/06 116. Apprenticeship offers a pathway to self-employment, but is less effective than TWT in opening doors to wage employment. As shown earlier in Table 1 o f Annex 1, approximately 30 percent o f the workers holding wage employment previously had been an apprentice, but when other factors are controlled inthe logit function, an apprenticeship does not appreciably increase the chances for wage employment. There i s evidence that it may have this impact for those with low levels o f education, but not for those with higher levels o f education. Formal education is a better investment for increasing the chances o f obtaining wage employment. On the other hand, an apprenticeship does enhance the probability o f becoming self-employed in the non-agriculture sector. The relationship has strengthened in the two later surveys for all levels o f education. When compared with TVET likelihoods, however, TVET remains a more effective mechanism for opening doors to wage employment. Educationand Earnings 117. Education influences earnings directly for wage and self-employment. Higher levels of education for wage workers in rural and urban settings generally translate into higher earnings in all three surveys with earnings differences statistically significant for secondary and post-secondary education (See Figures 21 and 22).44 The impact o f education i s thus two-fold affecting the likelihood o f obtaining wage employment and the wage level received. The relationship i s largely stable over the three surveys without any clear trends showing rising returns to education consistent with a tightening o f labor markets and demand for skills. The relationship for education also holds true for the self- employed in non-agriculture in urban areas, but i s weaker in rural areas where earnings differences tend to rise for junior secondary education and higher, but are less likely to be statistically significant (See Figure 22). Education's impact on the earnings o f the self- employed inagriculture favors higher earnings with education, but mainly inrural areas. 44 The regressions used infigures 21 22 are found inAnnex 1(see Tables 4 and 5). - 74 118. The private economic returns to primary education in Ghana are very low or even zero. Evidence in the past has favored primary education over alternative investments in higher levels of education. Continuing to invest in good quality primary education remains important, but more from the perspective of the options it opens for further education. Alone, primary education, and in most cases junior secondary education, do not yield statistically significant earnings gains compared to those who have no education. However, the additional completion of a senior secondary education or TVET provides substantial earnings gains that are statistically significant in rural and urban areas and for virtually all categories of employment. Further gains are found for those with post-secondary education, especially for wage employment, supporting the argument for investmentinpost-basic education. Figure 22: Waghe Repressionswith Sample Figure 23: Non-Agriculture Self-Employed Selection, Ghana 1991-2006,Urban Regressionswith Sample Selection, Ghana 1991- 2006, Urban I 2 L 1.5 1.5 1 1 I 1 9 9 1 0.5 I 1 9 9 1 0.5 I 1998 0 2005 0 2005 -0.5 i 119. Wage workers in rural and urban areas with TWT show earnings differences that comparefavorably with those of senior secondary education. School-based TVET fares well in comparison with senior secondary education for wage workers (see Figures 21 and 22). The quality issues notwithstanding, incremental gains inlog-earnings for those with a TVET education in 2005/06 largely match the gains in earnings for those with a senior secondary education. The earnings differences are statistically significant for wage workers inurbanareas. This is an important finding when it comes to making choices for further investment in post-basic edu~ation.~~However, the relationship of TVET to differences in the earnings of those self-employed in non-agriculture i s not generally statistically significant in rural or urban areas, and thus, TVET's impact on self- employment is largely indirect through its positive impact on the likelihood o f self- ~ 45While the private rates of return may place TVET and senior secondary on an even footing, the higher unit cost of TVET, shown above inFigure 16, would lower the social rate of return to TVET in comparisonwith senior secondary education. 75 employment. Although not shown here, these results also hold for TVET's impact on self- employmentinagriculture. 120. Participation in an apprenticeship is not statistically associated with higher earnings in rural or urban areas. While apprenticeships do influence the likelihood o f having a job by increasing the chances o f becoming self-employed, they do not translate into higher earnings when compared to those without an apprenticeship. The relationship i s examined for those with an apprenticeship and different levels o f education without identifying a positive impact o f apprenticeship on earnings. Instead, for the three categories o f employment, apprenticeship i s frequently associated with having a lower level o f earnings when compared to those without an apprenticeship. Monk, Sadefur and Teal (2007) (see Annex 2) offer an exception to this in their analysis o f the 2006 CSAE urban-based labor force survey in Ghana. They report that workers with the lowest levels o f education ('junior secondary or less) and an apprenticeship have a 42 percent higher income than those with the same level o f education and no apprenticeship. We find a similar result only when using regression analysis that does not control for self-selection into type of employment choice. 121. Beyond education and skills, other social and economic factors influence earnings. Men earn more than women with the differences being smaller in the case o f wage employment than self-employment in rural and urban areas. As observed in Table 1 o f Annex 1, men are a majority o f those employed inthe wage sector and agriculture self- employment and a minority o f those employed in non-agriculture self-employment. Women dominate this latter employment category. Only one out o f four persons in non- agriculture self-employment inurbanareas i s male, yet their earnings are higher than those o f women in this employment category and differences are statistically significant. The decline o f formal employment, unionization, and public sector employment in the wage sector, noted earlier, also has implications for earnings. The first two of these features o f wage employment enjoy an earnings premium while holding other factors like education constant. The declining share of workers in wage employment with these features translates into fewer workers who will realize the higher earnings associated with this employment. 122. Controlling for selection bias reduces the observed impact of education, including TVET,on wages and earnings. There is a debate inthe literature as to whether the two step procedure outlined earlier and the control for selection bias i s the best way to proceed. Puhani (2000) offers a good survey o f this literature showing that the procedure can be harmful where it i s difficult to obtain good identifying variables that are statistically significant in the multinomial logit equation. Since this problem i s observed in our regressions, we have estimated separately the same earnings equations using the recommended ordinary least squares method without adjusting for selection bias (see Annex 1 and Tables 6 and 7). There are differences in the results, but especially for education. We find there are positive returns to junior secondary education in urban and rural areas, which were not observed uniformly inthe regressions controlling for selection bias. And second, we find much stronger effects for TVET in urban and rural areas for 76 wage employment, but also for non-agriculture self-employment in urban areas. These results strengthen the conclusions about the importance o f education and skills. STRATEGIESFOR REFORM 123. Ghana launched the reform of its education and training systems early in the decade. The findings of the previous section point to the importance o f post-basic education and skills to obtaining employment, and depending on the type o f employment, to the wages and earnings received. TVET plays an important role, especially in influencing the type o f employment acquired. The White Paper, introduced in2004, refers to the mismatch o f skills and jobs with an education and training system that i s poorly connected with the skills that employers need. This section examines the strategy Ghana has adopted to address the skills problem, first in the ESP and then in the White Paper. This section places the strategy alongside other country reformers in the Region and globally and, in the context faced by Ghana, reflects on priorities for the reform program that would prepare youth with the skills needed for overcoming poverty and raising living standards. Government's ReformStrategy 124. The White Paper callsfor accelerating the ESP goalsfor education and placing increased emphasis on the expansion of TVET. The White Paper laid out plans to elevate TVET at the secondary level as a credible alternative to senior secondary education, and open access to apprenticeship for junior secondary school leavers who do not continue to senior secondary schools.46 The paper calls for a ten-fold increase by 2015 in technical and agricultural training in secondary schools from a gross enrolment rate o f 1.5 percent to 15 percent, and an increase in apprenticeships, beginning in 2008, from a gross enrolment rate of 2 percent to 20 percent. Inthis context, traditional apprenticeships are to be brought into the formal system. The TVET sub-committee identified the following strategies to implement the reform: (i) a structure for implementation and coordination o f the reforms; (ii) developmentofaNationalApprenticeship System; and(iii)sustainablefunding the a base for TVET. The objective was to improve the impact and image o f TVET, which was generally considered second-class education, and align TVET more closely with the needs of industryand trainees for skills. 125. The TVET strategy of the White Paper contains a holistic view. To achieve the goal o f a more market responsive training system, the strategy calls for reforms at school and market levels: (i)improving linkages of providers with industry; (ii)shifting to an outcomes focus with competency-based training; (iii)improving infrastructure and instructor training; and (iv) strengthening sector management and leadership. Plans were announced for development o f a national qualifications framework that would grant academic credit for past learning and open pathways to further education and training options as part o f a lifelong learning system. The framework would be extended to traditional apprenticeships. With TVET receiving only one percent o f the public education 46The White Paper identifies four curriculum streams for senior secondary students: general, vocational, technical, and agricultural. 77 budget, the strategy calls for an increase in public expenditure on TVET and the introduction o f additional sustainable sources o f financing. TVET Reform: a Regional Perspective 126. Lessons learned from experience. There i s a rich record o f reform experience in the Africa region. This record serves as a benchmark and guide for the reform o f TVET in Ghana:47 0 The importance o f building on a solid foundation o f basic education. Good training requires young people who have the requisite general education for acquiring skills for the workplace. Knowledge o f basic math, sciences, and communications i s essential to acquiring job-specific skills and encouraging employers to invest inworkers. 0 The importance o f maintaining sound macroeconomic policies and sustaining a favorable investment climate for iob creation.48 Training itself does not create jobs and i s dependent on investment to produce these jobs. Government and donors can only provide so much o f the investment and the remainder has to come from the private sector. Creating an investment-friendly climate i s important alongside policies that provide access to good quality skills training and encourage youth to invest in themselves. In Ghana, this includes improving the business climate inthe informal economy. 127. Ghana's awareness of the education and training system's weaknesses and the potential adverse consequencesfor growth and poverty reduction place it on a familiar path traveled by other countries. TVET reforms are observed throughout Sub-Saharan Africa, and also worldwide, with nations seeking to improve the performance o f training systems to smooth the transition o f young people from school to work and create a more competitive workforce for the global economy.49 While these reforms vary from country to country, they share many features and are divided into two categories: (i)reform o f technical and vocational education (TVE) at secondary and tertiary levels o f education, which i s applied mainly to ministries o f education and influences when TVE i s first offered and the design o f curricula for vocational and general education; and (ii) o f reform non-formal training, which includes a more diverse set o f providers and objectives outside ministries of education including technical ministries (e.g. agriculture, health, labor), for- profit and non-profit institutions, and employers. (QReforming TW:Issues to Conside?' 47Johansonand Adams (2004) 48World Bank (2005) 49Johanson and Adams (2004); Adams (2007); CINTERFOWILO (2000); Gill, Fluitman, Dar (2000); Atchoarenaand Delluc (2001); OECD (2000) 5oAdams (2007); Unevoc (2006) 78 128. The timing of the TVEintroduction is important. The pattern emerging worldwide i s to defer vocational specialization until senior secondary education, with advanced countries moving vocational content even later. Ghana's White Paper finds efforts to vocationalize JSS programs unsuccessful. Delaying vocational content provides room for a solid basic education foundation and avoids pressuring students to make career choices prematurely. Ghana i s part o f this trend with its focus on TTIs for junior secondary school leavers and the availability o f polytechnics for those seeking more advanced vocational content. However, low enrolment rates in secondary education means that there are large numbers of youth in Ghana (nearly 90 percent of the 15-17 year age group) who cannot follow this pattern o f skills development. The option for these youth is: (i) a VTI which provides a pathway back to the formal system in TTI's; or (ii) o f the other non-formal one sources o f TVET available from public and private providers, including traditional apprenticeship. 129. It is critical to expand choices for youth, enabling them to recover from situations in whichfurther education was not possible, or had to be abandoned. Within the current TVET strategy inGhana, VTIs appear to be the main instruments for providing this choice, but they are severely limited in capacity. Though youth benefit from other options, like traditional apprenticeships, which help them acquire skills, these options do not offer a second chance to return to formal education, and this places a ceiling on the future for these youth andthe humanresources capacity o fthe economy. 130. Governmentsare improving the articulation of education and training.51TVET i s often the option offered students where there is inadequate space for continuing in general secondary education. As such, TVET has gained the reputation o f being a dead- end, or a `second-class' education. Countries are taking steps to change this image allowing students to move to other vocational programs or even back to general education and vice-versa without losing time in studies. Curricula that give vocational students more academic content and general students more practical exposure provide for this fle~ibility.~~ Ghana's education system does not offer this flexibility. Leaving JSS at age 15, students face a choice of SSS or a TTI, or alternatively, to leave formal schooling and pursue skills through non-formal means o f training, including apprenticeship. Once out o f the education system, it is difficultto re-enter and continue one's formal education. 131. More attention is being given in developed countries to the combination of schooling and work experience, including apprenticeship. Sweden and Australia are examples o f countries that include work experience in the secondary curriculum. In the United States, career academies in secondary schools are providing schooling and work options. The long standing example o f this i s found in the "dual system" o f Germany where students divide time between school and work. These programs work well in advanced countries with large formal sectors that are creating jobs, but may not work as well in developing countries like Ghana where the formal sector is much smaller. Ghana is, however, moving forcefully with plans for a more active role for apprenticeship. With ''Adams (2007) 52 Bowers, Sonnet, Bardone(1999:27) 79 its large informal sector, the attention Ghana i s giving to traditional apprenticeships i s more consistent with the nature of its economy. 132. Few countries, however, have been successful in formalizing traditional apprenticeships as Ghanaproposes. The White Paper calls for Government to assume the cost of the first year of the apprenticeship program.53 Government will want to proceed carefully to avoid destroying what works well intraditional apprenticeships. A review in West Africa points to actions that could improve traditional apprenticeships, including: (i) providing literacy and basic education for master craftsmen and apprentices; (ii)improving access to technology for master craftsmen; and (iii)offering further technical and pedagogical training for master craftsmen.54 The ILO expects to release soon the proceedings of its May 2007 workshop on informal apprenticeship training inWest Africa, which contains more examples of how traditional apprenticeshipscan be improved. (io Reforming non-formal training, (the second "T"in TWT):Issues to consider 133. Non-formal training is offered by technical ministries (labor, agriculture, health); private for-profit and non-profit institutions; and employers. Non-formal training i s frequently of short duration, competency-based rather than time-based as found inschools, and delivered by skilledcraftsmen rather thaninstructors withteaching degrees. Non-formal training does not necessarily open pathways to further formal education, although NQFs are beginning to open this door. The goals of non-formal training go beyond preparation of youth for first employment to preparation of youth and adults for coping with changes found in a market economy. The objectives include skills upgrading for the employed as new technologies are introduced, equipping the unemployed with skills to search for employment, facilitating occupational mobility, and providing at-risk youth and young adults with "second chance" opportunities for skills missed in early education. 134. Faced with a diversepublic andprivate provider communityfor TWT, countries havefocused on governanceframeworks to integrate the system and improve links to the market and efficiency. The creation of COTVET as a coordination body is a familiar tool for this and i s found in countries like Cote d'Ivoire, Mozambique, South Africa, Tanzania, Zambia, andinother countries world-wide. These coordination bodies, either embeddedin ministries or establishedas autonomous bodies, bring employers and government together in decision-making and allocation of public expenditure on TVET. Both parties can be joined by representatives of workers and civil society. Different governance models are found with powers ranging from advisory to decision-making and responsibilities covering different activities supporting amarket-led training system. 135. Decentralization of delivery is placing training decisions closer to the market. Countries are moving away from centralized TVET systems and bringing decisions on instructor hiring and firing, curriculum and course offerings, choice of pedagogy, and management o f locally generated resources closer to the market at the training institution 53See White Paper (2004): 13 54Haan and Serriere (2002) 80 level. Managers of training institutions, rather than ministry officials far from local markets,are taking responsibility for these decisions, with the expectation they will be held accountable for performance. This decentralization, while an important tool for reform, requires building capacity of local managers, who must become skilled in reading local marketdemand for skills andorganizing resources to meet this demand. Ghana's plans for strengthening educational management are consistent with this, but could usefully be extended to private providers, improving their performance and reducing pressures on public provision. 136. Other than financing, the most important issue in TVET reforms is quality assurance. Quality can vary widely in training systems and efforts to promote the integration of public and private provision, placing them on the same level of performance, require introducing quality assurance frameworks. 0 Providing;certification of skills. Inadvanced countries and some developing countries, National Qualification Frameworks (NQFs), as proposed in Ghana, have been created. These frameworks give individuals credit for past learning from different modes of delivery and offer a widely-accepted credential. Whether the skill is acquired in the workplace or in a school setting, the learning can be measured and credit given, placing all forms of training on a level playing field. However, NQFs require considerable capacity to develop and manage and may not be suitable for all countrie~.~~ a Involving emulovers in setting; standards. This approach may work better than a NQF in improving quality and standards for traditional apprenticeships. Education and training institutions frequently set standards for skills development without engaging employers. Reforms that bring employers into this process can improve the relevance and quality o f the training offered. Independenttesting and certification of trainees can determine if the standard has been met. In Ghana, though employer engagement in standard setting is not emphasizedinthe ESP, it is expectedto be part of the reform. a Shifting attention to outcomes. Building a standards-based training system balances attention given to the input side with attention to the impact of training on employment, earnings, and further education and training. Ghana can proceed with development o f a NQF and should link this with other regional efforts to promote the portability o f skills.However, giving a higher priority to development of a standards-based training system and a focus on training outcomes i s likely to provide a faster payoffto quality improvements. a Building.a quality assurance framework. This is essential and a priority for Ghana with its large non-government provider base. Other instruments, such as licensing, can help improve quality by setting minimum standards for training 55Young (2005) 81 providers, while accreditation can establish even higher standards. One o f the most glaring omissions from the TVET strategy in Ghana is its monitoring and evaluation plan. The TVET sub-committee's logical framework identifies results, timetables, and agencies responsible for activities, but says little about the monitoring o f key performance indicators or the conduct o f impact evaluations. Effective TVET systems are developing "learning cultures," whereby performance i s monitored carefully for lessons to guide ongoing policy development and investment. 137. Diversification of financing is important. Ghana's public expenditure on TVET (slightly over one percent o f the education budget) is well below spending in similar country settings. Post-conflict Sierra Leone spends over 4 percent o f its education budget on TVET, with other low-income countries like Ethiopia spending up to 10 percent. The education budget for TVET will be increased to 7.5 percent, while the ministry o f labor budget will rise to 20 percent. In view o f the low level o f public expenditure, private financing from households and enterprises will continue to play an important role in spending on TVET. Other sources o f financing can be sought from sale o f contract training services, sale o f products produced by trainees, community contributions, and external donor assistance. Payroll and turnover taxes are used in some countries where there i s a stable, formal economy as well as tax expenditures where there i s an effective tax collection system. 138. Beyond the mobilization of resourcesfor the system, a more important issue is how these resources are spent and the incentivescreatedfor goodperformance. Ghana's TVET strategy focuses on resource mobilization, and says little about how resources will be spent or performance incentives. Countries like Chile have separated financing from provision. The national training agency, SENCE, buys training services on the open market frompublic and private providers (rather thanproviding it itself), usingcompetition to reduce costs and get the best training for the money spent. In Sub-Saharan Africa, Mauritius has recently adopted this model. Performance-based budgeting, competitive procurement o f training and vouchers can all improve training institutions' incentives for performance, yet none are part o f the Ghana TVET reform. Improving management capacity for TTIs and VTIs and holding managers accountable for results would be necessary to make a performance-based system work. An Overview of Findingsand PolicyImplications 139. Ghana has taken important steps toward opening access to education and introducing TVET reforms. While progress has been made in opening access to basic education, quality remains a challenge. As reflected in completion rates, a large majority o f youth leave the formal education system without a good educational foundation. The expansion o f access for basic education i s now creating a pipeline o f youth with demand for further education. The pressure for expansion i s backed by analysis o f household data in the previous section providing evidence of positive economic returns to post-basic education that open access to better paying jobs and higher earnings. In this analysis, TVET plays an important role influencing access to wage and self-employment. 82 Opportunities for further education, however, are limited. Of those successfully completing nine years o f basic education, only one out o f ten will find a place in senior secondary education and fewer than this will find their way into a TVET program. A much larger number will pursue skills through a traditional apprenticeship or by learning on the job. 140. For a country that aspires to middle-income status, tackling the problem of education and skills is a priority. Ghana plans large infrastructure investments in the future covering energy, water and sanitation, rural roads, and ICT as part o f its broader strategy for promoting growth and private inve~tment.~~ Assuring a skilled workforce for constructing, managing and maintaining this infrastructure i s important, but also for creating incentives to introduce new technologies in other sectors that can raise productivity and incomes. A skilled workforce is a necessary, though not a sufficient condition, to achieve the goal of becoming a middle-income country. Promoting better access to good quality education and skills has to be accompanied by policies that are friendly to private investment. Against this background, the education and TVET reforms Ghana has launched are an important step toward achieving the middle-income goal and, as such, merit support. Box 11 summarizes recommendations related to education and skills development and next steps. Box 11: Summary of recommendationson education and skills development and next steps In respect to skills development, Ghana has adopted a strategy for further development of technical and vocational education and training and created a coordination body known as the Council for Technical and Vocational Education and Training (COTVET). It now needs to translate the strategy into an expenditure program and action plan setting priorities for implementation of the strategy. In this context the fmdings of this report provide a fiamework for hrther development of the sector and setting of implementation priorities. The recommendationsof the report can be divided into policy reforms for school-based skills development and post-school development of skills. The former addresses the needs of youths now coming through the education system, preparing them for entry to the world of work, while the latter reaches out to a wider set of beneficiaries covering those who are entering the labor market and need skills, those with work needing to upgradetheir skills, and those who have lostjobs andneed skills to improve their employability. The recommendationsfor school-basedreforms in TVET set the priority on improving quality in basic and post-basic education with the latter including technical and vocational education. Improving quality is given a higher priority than expansion of coverage because the low demand that now exists for TVE in schools, private and public, is attributed to the low quality andrelevance ofthe programs available. Before expansion of any program, attention needs to be given to enhancingthe quality ofthe training offered. For those seeking skills after leaving school, establishing COTVET and building its capacity to coordinate skill development activities is recognized as a priority. This includes strengthening the quality assurance system that sets standards for training with industry input and provides for testing and certification of skills and for licensing and accreditation of providers. These steps are important to improve the use of public expenditure that is now fragmented among different providers and to provide the marketplace with information about the quality of skills training offered by a diverse set ofproviders. The majority of workers have used traditional apprenticeships to prepare themselves for employment in 56World Bank (2007) 83 Ghana. This report identifies quality as an issue for these apprenticeships and offers a number of recommendationsto improve the quality of traditional apprenticeships. Given the scope of apprenticeship programs, improving the quality of these programs should be given a high priority among post-school reforms, second to that of establishing COTVET. Ongoing donor initiatives to create a new Skills DevelopmentFundcan be usefulto promoting an expanded role for enterprises inthe delivery and financing of training, but given the small size of Ghana's current formal sector which is expectedto be the primary beneficiaryofthe Fund, its impact on skills developmentis likely to be small. Therefore, next steps call for preparationof an expenditureprogramimplementingGovernment's strategy for skills development. This report provides a useful framework of investmentsand activities for consideration as part of the expenditureprogramto help Ghanaestablish a flexible, market-drivenTVET system that offers equitable access to skills for all andprepares beneficiaries for productivewage andself-employment. 141. The challenge will be sustaining the gains made in basic education and improving quality, while responding to growing demand for post-basic education. Alongside this is the challenge o f promoting choice and strengthening the role played by TVET, both for pre-employment and in-service training needs. Current quality notwithstanding, a school-based TVET investment compares well in employment and earnings outcomes with the alternative o f a senior secondary education for those who expect to enter the labor market on completion. Currently, less than one percent of the education budget i s spent on the approximately 14 percent o f secondary students in a school-based TVET program. Given the expected higher unit cost o f these programs, reforms need to focus on strengthening their linkage with market demand, ensuring a good match o f skills andjobs inan economy that i s increasingly led by the private sector, and in providing incentives for the efficient use o f resources for skills development. 142. Benchmarked against regional and international experience, Ghana's TWT reforms adhere to good practices, but as shown above, there are also missed opportunities. This review has highlighted challenges that face skills development in Ghana and provided evidence from household surveys o f the economic benefits that can be realized from meeting these challenges. The findings support expanded attention to post- basic education, and therein, the reform o f TVET. Addressing school-based TVET's image as "second-class" education i s not featured as part o f current reforms and unless this is included, the proposed reforms are likely to fall short o f their goal. EnteringTVET after lower secondary school offers few pathways back into the general education system for further education, and as such, reinforces the image o f TVET as a dead end. This can be addressed in two ways. First, improving the quality and relevance o f the TVET offered can enhance its image and outcomes. Second, improving articulation within the education and training systems and opening vertical and horizontal pathways between these two systems can open choices to youth to move between general secondary and vocational programs and from TVET to higher levels of education. Curriculum reforms that blend general and vocational content can do this as can accelerating development o f the National Qualifications Framework (NQF). Development o f an NQF will take time, however, and intermediate steps can be taken by engaging employers in setting skill standards and building capacity to test and certify the training available against these standards, including skills intraditional apprenticeships. 84 143. While strengthening traditional apprenticeships can improve their quality and outcomes, caution needs to be exercised in trying to formalize this mode of skills development. Apprenticeships are by far the most frequent means used for skills development in Ghana. We have shown their importance in opening employment for youth outside the farm sector. Yet, while referring to the benefits o f apprenticeships, we have also highlighted their weaknesses. Traditional apprenticeships are probably not the future for skills development in Ghana with its aspirations for becoming a middle-income country, but they are an important means for improving the current welfare of workers. Government initiatives to intervene and finance the first year o f apprenticeship and create a regulatory framework open Ghana to the risk o f distorting the incentives that make the system work well. The proposed cost o f this i s probably not affordable and comes at a very high opportunity cost for other reforms in TVET and in education more generally. There are steps, however, that can be taken that would lower the risk o f destroying the conditions that make traditional apprenticeships work well. These include: (i)bringing literacy to master craftsmen and apprentices, (ii) upgrading the technical and pedagogical skills o f master craftsmen, (iii) opening access to new technologies for master craftsmen, and (iv) certifying apprenticeship skills as a guide to their quality. 144. The attention given to firtancing TVET is important, but incentivesfor using resources effectively are even more important. This is currently missingfrom the TKET reform program. Introducing greater accountability for performance i s needed especially within the public sector. Subjecting institutions to competition for public financing and placing public and private institutions on equal footing for receiving this financing can change the incentives for performance. Vouchers, for example, can subject training institutions to competition for resources directly from the beneficiary. Performance-based budgeting is another tool requiring training institutions to compete for public financing where performance may consider training outcomes such as reduced dropouts, higher skill certification rates, and job placement, along with training inputs such as the number o f classes and workshops, number o f instructors, and number o f programs offered. By moving to funding formulas that focus as much on the outcomes and performance o f the providers as on inputs, and by holding managers accountable for results, the incentive can be changed in a meaningful way to alter the outcomes. Moreover, encouraging public training institutions to earn more revenue through their design and delivery o f customized training for employers can further improve incentives for performance. 145. Building a "learning culture" around TVET reforms will improve policy development and the use of public spending on skills in a setting of limited resources. What works in one country setting may not work as well in another. Monitoring and evaluating outcomes i s necessary to determine if goals have been met and programs are performing as expected. Ghana's TVET reform program has a results framework, but does not offer a plan for monitoring and evaluation and without this, little can be said about the effectiveness o f the reforms and modifications that may be needed. This is a missed opportunity. Experience in other countries shows that TVET reform i s an extended journey o f 10 to 15 years. The program demands a strong champion ingovernment to lead the way and a good strategy. The ESP and the follow-up White Paper provide this strategy, but a "learning culture" needs to be built around the strategy to guide future 85 investments. This paper provides justification for support o f the TVET reforms proposed by Ghana. These reforms can be strengthened by addressing the missed opportunities above. 86 References Adams, Arvil V. (2007). The Role of Youth Skills Development in the Transition from School to Work: A Global Review, HDNCY DiscussionPaper No. 5, Washington, D.C.: World Bank. Adams, Arvil V., Harold Coulombe, Quentin Wodon, Setareh Razmara. (2008). "Education, Employment and Earnings inGhana." World Bank (processed). Anyemedu K. (2000). "Trade Union Responses to Globalization: Case Study of Ghana," University of Ghana. Aryeetey, E.andA. K.Fosu (2002). Economic GrowthinGhana: 1960-2000, mimeo. Atchoarena, David and Andre Marcel Delluc (2001). Revisiting Technical and Vocational Education in Sub-Saharan Africa: An Update on Trends, Innovations, and Challenges, IIEP/Prg.DA/1,320. Paris: International Institute for Educational Planning. Baah-Boateng W. (2005). "Employment Policies for Sustainable Development: the Experienceof Ghana," Department ofEconomics, University of Ghana, 2005. Batra, Geeta (2000). "Foreign Direct Investment and Skills Upgrading in Developing Countries," Foreign Investment Advisory Service. Washington, D.C.:World Bank reference available at http://www.oecd.org/dataoecd/l 1/12/2731690.pdf. Betcherman G., Luinstra A., and Ogawa M. (2001). "Labor Market Regulation: International Experience in Promoting Employment and Social Protection," Social Protection DiscussionPaperNo. 0128, World Bank. Betcherman G., Olivas K., andDar A. (2004). "Impacts of Active Labor Market Programs: New Evidence from Evaluations with Particular Attention to Developing and Transition Countries." Bogetic, Z., et al. (2007). Ghana's Growth Story. Ghana CEM: Meeting the Challenge of Accelerated and Shared Growth. Bowers, Norman, Anne Sonnet, and Laura Bardone (1999). "Giving Young People a Good Start: the Experience of OECD Countries," in Preparing Youth for the 2lSt Century: the TransitionfromEducation to the Labour Market. Paris: OECD: 7-86. Coulombe, H.and Q. Wodon (2007). "Poverty, Livelihoods, and Access to Basic Services in Ghana." Ghana CEM: Meeting The Challenge of Accelerated and Shared Growth. World Bank. Dabalen, H. S. Nielsen, et al. (2003). "Labor Markets and Enterprise Training in African Manufacturing." Fields, G. (2007). Employment in Low-Income Countries: Beyond Labor Market Segmentation. Employment and Shared Growth. Rethinking the Role of Labor Mobility for Development. P. Paci and P. Serneels, The World Bank. Ghana Statistical Service (2000). Ghana Living Standard Survey, Report of the Fourth Round(GLSS 4). Ghana Statistical Service (2000). Poverty Trends inGhana inthe 1990s. 87 Ghana Statistical Service (2005). Population data analysis report Vol. 1:Socio-economic and demographic trends analysis, Ghana Statistical Service, Accra. Ghana Statistical Service (2007). Pattern and Trends of Poverty inGhana 1991-2007. Gill, Indermit, Fred Fluitman, Amit Dar (2000). Vocational Education and Training Reform: Matching Skills to Markets and Budgets. World Bank and International Labour Organisation: Washington, D.C. and Geneva. Government of Ghana (2006). Youth Employment Implementation Guidelines. Government of Ghana, ILO and UNDP (2004). "An Employment Framework for Poverty reduction inGhana." Haan, Hans Christiaan and Nicolas Serriere (2002). "Training for Work in the Informal Sector: Fresh Evidence from West and Central Africa," Occasional Papers of the International Training Centre ofthe ILO, Turin: ILO. ISSER (Institute of Statistical, S. a. E. R. (2006). The State of the Ghanaian Economy in 2005. I. Ghana. Legon, Ghana. 0. Johanson, Richard and Arvil V. Adams (2004). Skills Development in Sub-Saharan Africa, Washington, D.C.: World Bank Kelley, A. C. and R. Schmidt (2001). Economic and Demographic Change: A Synthesis of Models, Findings, and Perspectives. Population Matters. N. Birdsall, A. C. Kelley and S. W. Sinding, Oxford University Press. Kingdon G., Sandefur J., Teal F. (2004). "Patterns of Labor Demand," mimeo, World Bank, Africa Region. Maloney, W. (2003). Informal Self-Employment: Poverty Trap or Decent Alternative? Pathways Out of Poverty. G. Fields and G. Pfeffermann. Boston, Kluwer. McKay, A. and E. Aryeetey (2004). Operationalizing Pro-Poor Growth. Mimeo. McKay, A. and E. Aryeetey (2007). Ghana: The Challenge of Translating Sustained Growth into Poverty Reduction. Delivering on the Promise o f Pro-Poor Growth. T. W. Bank.Washington / New York, The World Bank and Palgrave Macmillan. Ministry of Education, Youth and Sports (2003). "Education Strategic Plan Iand 11," Accra: (2004). white Paper on the Report of Education Reform Review Committee, Ministry of Education, Youth and Sports, Accra. (2006). "Preliminary Education Sector Performance Report 2006," Accra. and Ministry of Manpower, Youth and Employment (2004). "Executive Summary: TVET Policy for Ghana," Accra. Ministry of Manpower, Youth andEmployment(2006). TheNationalEmployment Policy, Draft. Ministry of Manpower, Youth and Employment (2006). The National Social Protection Strategy: Investing inPeople, Draft. Ministry of Manpower, Youth, and Employment (2006). The National Employment Policy - draft. 88 Monk, Courtney, Justin Sandefur, and Francis Teal. 2007. "Apprenticeship in Ghana," Centre for the Study of African Economies, Department of Economics, University of Oxford, (processed). National Education Reform Implementation Committee (2006). "Report of the TVET Sub- Committee," Accra. Rosholm, M., H. S. Nielsen and A. Dabalen 2007. "Evaluation of Training in African Enterprises," Journal of Development Economics, vol. 84(l), 310-329. pages Palmer, Robert (2007). Skills Development, the Enabling Environment and Informal Micro-Enterprise in Ghana, doctoral thesis (mimeo), Edinburgh: University of Edinburgh. Puhani, P.A. (2000). "The Heckman Correction for Sample Selection and its Critique," Journal of Economic Surveys, 14:53-68. Sandefur, J. (2006). Explaining the Trend toward Informal Employment in AFrica: Evidence from GhanaianManufacturing. mimeo, The World Bank. Sandefur, J., P. Serneels, et al. (2007). Poverty and Earnings Mobility in Three African Countries. Employment and Shared Growth. Rethinking the Role of Labor Mobility for Development. P. Paci andP. Serneels, The World Bank. Schultz, Paul (2003). "Evidence of Returns to Schooling in Africa from Household Surveys; Monitoring and Restructuring the Market for Education," Economic GrowthCenter, Discussion PaperNo. 875, New Haven: Yale University. Soderbom, M. and F. Teal (2001). "Firm Size and human capital as determinants of productivity and earnings." CSAE Working paper 2001-09. Soderbom, M., F. Teal, et al. (2001). "Does firm size really affect earnings?" CSAE Working paper 2002-08. Tan, Hong (2005). "In-service Skills Upgrading and Training Policy: Global and Regional Perspective," Paper presented at the MNA Job Creation and Skills Development Conference, December 2005, Cairo. Tan, Hong and Pei Savchenko (2003). "In-service Training and Productivity: Results from Investment Climate Surveys," Washington, D.C.: World Bank (processed). Teal F. (2005). "Ghana Investment Climate Assessment," Draft, University of Oxford. UNESCO (2003). "Synthesis of Main Findings from Two Case Studies Carried Out in Ghana and Zambia on Private Technical and Vocational Education and Training (TVET)," Paris: UNESCO. Unevoc (2006). "Participation inFormalTechnical and Vocational Education andTraining Programmes Worldwide," UNESCO International Centre for Technical and Vocational Education and Training: Bonn. World Bank (1995). "Workers inanIntegrating World", World Development Report. World Bank (2005). World Development Report 2005: A Better Investment Climatefor Everyone. Washington, D.C. World Bank (2007). Ghana: Meeting the Challenge of Accelerated and Shared Growth. Country Economic Memoradum, Vol. I. Young, Michael (2005). "National qualifications frameworks: Their feasibility for effective implementation indeveloping countries." Geneva: ILO (processed) 89 Ziderman, Adrian (2003). Financing Vocational Training in Sub-Saharan Africa, Africa RegionHumanDevelopmentSeries. Washington, D.C.:World Bank. 90 MAP SECTION IBRD 33411 2°W 0° 2°E To Tenkodogo BURKINA FASO To Hamale Navrongo U P P E R E A S T Bobo- Diolasso WalewaleTumu Bolgatanga GHANA Nakpanduri U P P E R W E S T Walewale To Dapaong 10°N Black Volta Kolpawn Wa Wa 10°N Gushiegu White Volta To Djougou N O R T H E R N Yendi To Tamale Ferkéssédougou BENIN Sawla Fufulsu Bole Daka To Djougou CÔTE Nakpayili Oti TOGO D'IVOIRE To BlackVolta Salaga Bouna Makongo Yeji Kintampo Dambai 8°N 8°N Jema V O L T A B R O N G - A H A F O Atebubu Kwadwokurom Tain Techiman Pru Berekum To Sokodé Sunyani Mount Afadjato (880 m) K w Bia a h Afram u Lake Kpandu nges AgbovilleoT P Volta To l Agogo Ra Abomey Goaso a t e a u Kumasi Bibiani E A S T E R N -Togo Ho Krokosue A S H A N T I Anum im To Porto- no Obuasi Diaso Birim p Novo a aT Kade Koforidua kw Volta Aflao 6°N Oda W E S T E R N A 6°N Dunkwa Enchi GREATER 2°E ACCRA AbidjanoT Tema Twifo Praso ACCRA Prestea C E N T R A L GHANA Ankobra Pra Tarkwa Winneba SELECTED CITIES AND TOWNS Cape Coast REGION CAPITALS Newtown NATIONAL CAPITAL This map was produced by Sekondi the Map Design Unit of The Axim Takoradi RIVERS World Bank. The boundaries, Gulf of G uinea colors, denominations and any other information shown MAIN ROADS on this map do not imply, on the part of The World Bank 0 20 40 60 80 Kilometers RAILROADS Group, any judgment on the legal status of any territory, REGION BOUNDARIES or any endorsement or 0 20 40 60 Miles a c c e p t a n c e o f s u c h boundaries. 2°W 0° INTERNATIONAL BOUNDARIES SEPTEMBER 2004