DIREC TIONS IN DE VELOPMENT Human Development Stepping Up Skills in Urban Ghana Snapshot of the STEP Skills Measurement Survey Peter Darvas, Marta Favara, and Tamara Arnold Stepping Up Skills in Urban Ghana Direc tions in De velopment Human Development Stepping Up Skills in Urban Ghana Snapshot of the STEP Skills Measurement Survey Peter Darvas, Marta Favara, and Tamara Arnold © 2017 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 20 19 18 17 This work is a product of the staff of The World Bank with external contributions. The findings, interpreta- tions, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other informa- tion shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Darvas, Peter, Marta Favara, and Tamara Arnold. 2017. Stepping Up Skills in Urban Ghana: Snapshot of the STEP Skills Measurement Survey. Directions in Development. Washington, DC: World Bank. doi:10.1596/978-1-4648-1012-1. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third- ­ party–owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-1012-1 ISBN (electronic): 978-1-4648-1013-8 DOI: 10.1596/978-1-4648-1012-1 Cover photo: © Peter Darvas/World Bank. Used with permission. Further permission required for reuse. Cover design: Debra Naylor, Naylor Design, Inc. Library of Congress Cataloging-in-Publication Data has been requested. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Contents Foreword xi Acknowledgments xiii Overview xv Abbreviations xxv Chapter 1 Country Context 1 Economic, Social, and Demographic Trends 1 Education and Skills 3 Persistent Challenges in Education 4 Notes 6 References 6 Chapter 2 Conceptual Framework: Why Is It Important to Focus on Skills? 7 Introduction 7 Which Skills Are Relevant? 9 How Are Cognitive, Behavioral, and Technical Skills Formed? 10 Understanding the Focus on Adults and Urban Areas 11 Notes 12 References 12 Chapter 3 Skills Toward Employment and Productivity Data 13 Introduction 13 Data and Sample Characteristics 13 Types of Skills Measured 14 Methodology 18 Note 19 References 19 Chapter 4 Education Issues in Ghana 21 Foundational Skills: Early Childhood Education 21 Socioeconomic Disparities 25 Constraints for Education: Late Entry, Completion Delay, and Dropouts 27 Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   v   vi Contents The Gender Gap in Education 29 Reference 31 Chapter 5 Labor Market Participation 33 Introduction 33 Wage Employment, Self-Employment, and Formality 35 Occupation Type 37 Economic Sector of Occupation 37 Gender Disparities 40 Notes 42 Reference 42 Chapter 6 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 43 Introduction 43 The Use of Cognitive Skills: Overall Use and Intensity of Use 43 Job-Relevant (or Task-Related) Skills 47 Socioemotional Skills 50 Reading Literacy Assessment 54 Developing Skills beyond Education: Training and Apprenticeships 65 Notes 67 References 68 Chapter 7 The Returns to Education and Skills: Building the Job-Relevant Skills That Employers Demand 69 Introduction 69 Returns to Education and Skills 70 Returns to Education: Is There a Gender Premium? 74 Notes 76 References 76 Chapter 8 Mismatch of Skills: A Measurement Issue and Unexploited Potential at Work 77 Mismatch between Self-Reported Skills and Core Literacy Test Results: Does Language Matter? 77 Are There Unexploited Skills in the Workforce? 78 Conclusions and Looking Forward 79 Appendix A Summary of Statistics 81 Appendix B Skills Definitions, Survey Questions, and Aggregation Strategy 85 Appendix C Definitions of Variables Used in the Analysis 89 Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Contents vii Appendix D Differences in Mean 91 Appendix E Returns to Education and Skills 103 Appendix F Mismatch of Skills and Unexploited Potential Tables 131 Appendix G Effect of Socioemotional Skills on Education and Labor Outcomes 145 Boxes 1.1 The Ghanaian Education System 3 2.1 Skills Toward Employment and Productivity 8 4.1 Definitions of Education Variables 25 5.1 Definitions of Labor Market Terms Used in the Skills Toward Employment and Productivity Survey 34 6.1 What Does It Mean to Pass the Core Literacy Test? 55 Figures 1.1 Net Migration Rate, by Region 2 B2.1.1 The Skills Toward Employment and Productivity Framework 9 2.1 Skills Classification 10 3.1 The STEP Household Survey Instrument 14 3.2 Workflow for the STEP Skills Measurement Survey 16 4.1 Participation in Early Childhood Education in Ghana 22 4.2 Regular Use of Reading, Writing, Numeracy, and Computer Skills in Ghana 22 4.3 Intensity of Skill Use in Ghana 23 4.4 Educational Attainment in Ghana 23 4.5 Educational Level in Ghana, by Age Groups 24 4.6 Regional Disparities in Changes in the Primary Completion Rate 26 4.7 Education Level Attained in Ghana, by Socioeconomic Status 26 4.8 Evolution of Educational Profile in Ghana, by Socioeconomic Status 27 4.9 Delay in the Official Age of School Entry and Graduation 28 4.10 Dropped Out of Highest Educational Level Started, by Maximum Level of Education Completed and Socioeconomic Status 28 4.11 Main Reason for Dropping Out 29 4.12 Educational Attainment, by Gender 30 4.13 Educational Composition: Is the Gender Gap Narrowing? 30 4.14 The Gender Gap in Educational Attainment (Male–Female), by Age Group 31 5.1 Labor Force Participation and Nonparticipation in Ghana 34 Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 viii Contents 5.2 Labor Status, by Educational Level 35 5.3 The (Slow) Transition from School to Work: Labor Status, by Age Group 36 5.4 Employment Status, by Age Group 36 5.5 Employment Status, by Education Level 37 5.6 Type of Occupation, by Employment Status 38 5.7 Type of Occupation, by Education Level 38 5.8 Type of Occupation, by Economic Sector 39 5.9 Characterizing Each Economic Sector by the Education Level of Its Labor Force 39 5.10 How People with Different Levels of Education Are Distributed across Economic Sectors 40 5.11 Employment Status, by Marital Status 41 5.12 Labor Status, by Gender and Age Group 41 6.1 The Use of Cognitive Skills, by Level of Education Completed 44 6.2 The Use of Cognitive Skills, by Level of Education Completed and Age 44 6.3 The Intensity of Use of Reading Skills, by Education Level 46 6.4 The Intensity of Use of Writing Skills, by Education Level 47 6.5 The Use of Cognitive Skills at Work, by Employment Status 48 6.6 The Use of Cognitive Skills at Work, by Economic Sector 48 6.7 Job-Relevant Skills, by Employment Status 49 6.8 Average Score of Personality Traits and Grit 50 6.9 Average Scores for Hostile Bias and Time and Risk Preferences 51 6.10 Coefficients of Socioemotional Skills on Years of Education, Controlling for Sociodemographic Characteristics 52 6.11 Coefficients of Socioemotional Skills on the Probability of Attaining at Least SHS Education, Controlling for Sociodemographic Characteristics 52 6.12 Coefficients of Socioemotional Skills on the Probability of Selected Labor Market Outcomes 53 B6.1.1 Main Language Spoken at Home and Work 55 B6.1.2 Self-Reported Ability to Speak and Read/Write in English at Work 56 B6.1.3 Performance in the Core Literacy Test According to Self-​ Reported Ability to Speak and Read/Write in English 56 B6.1.4 Did the Lack of English Keep You from Getting a Job? 57 6.13 Distribution of Reading Component and Core Literacy Test Respondents 58 6.14 Performance on the Reading Component 59 6.15 Performance According to the Self-Reported Use of Reading and Writing Skills 59 6.16 Performance According to Age Group and Education Level 60 6.17 Core Literacy Test Performance 61 6.18 Performance on Literacy Exercise Booklets: Proficiency Levels 63 Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Contents ix 6.19 Distribution of Core Literacy Test Results, by Proficiency Levels Achieved on Literacy Exercise Booklets 63 6.20 Literacy Level According to Gender, Education Level, Age Group, Reading Intensity, Labor Status, and Employment Status 64 6.21 Differences in Proficiency Levels, by Self-Reported Reading Intensity 64 6.22 Certificate, Training, and Apprenticeship, by Education Level and Age Group 65 6.23 Are Students Choosing Apprenticeship Instead of Formal Education? 66 6.24 Certificate, Training, and Apprenticeship, by Employment Status 66 6.25 Certificate, Training, and Apprenticeship, by Occupation and Economic Sector 67 7.1 Returns to Education 70 7.2 Returns to Education, by Type of Employment 71 7.3 Level of Education, by Employment Status 73 7.4 Linear Probability Model of Being Self-Employed 74 7.5 The Returns to Education 75 Tables 3.1 Definitions of Skill Types 15 3.2 Levels of Reading Proficiency 17 A.1 Summary of Statistics from the STEP Survey 81 A.2 Sampling Procedure Comparison of STEP, GLSS 5, and GLSS 6 83 A.3 Regional PSU Sample Size Comparison of STEP, GLSS 5, and GLSS 6 83 D.1 Difference in Mean of Those Passing the Core Literacy Test Threshold and Those Failing the Core 91 D.2 Difference in Mean of Those Passing the Core Literacy Test Threshold and Those English Illiterate 93 D.3 Difference in Mean of Those Answering Core Literacy Test and Those Who Did Not Answer It 95 D.4 Difference in Mean of Those Missing and Those English Illiterate 98 D.5 Difference in Mean of Those Answering the Socioemotional Section and Those Who Did Not 100 E.1 Returns to Years of Education (Mincer Equation), Controlling for Skills 103 E.2 Returns to Education Level (Mincer Equation), Controlling for Skills 105 Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 x Contents E.3 Returns to Years of Education (Mincer Equation) for Informal Wageworkers, Controlling for Skills 108 E.4 Returns to Education Level (Mincer Equation) for Informal Wageworkers, Controlling for Skills 110 E.5 Returns to Years of Education (Mincer Equation) for Formal Wageworkers, Controlling for Skills 112 E.6 Returns to Education Level (Mincer Equation) for Formal Wageworkers, Controlling for Skills 115 E.7 Returns to Years of Education (Mincer Equation) for Male Workers, Controlling for Skills 117 E.8 Returns to Education Level (Mincer Equation) for Male Workers, Controlling for Skills 119 E.9 Returns to Years of Education (Mincer Equation) for Female Workers, Controlling for Skills 122 E.10 Returns to Education Level (Mincer Equation) for Female Workers, Controlling for Skills 124 E.11 Linear Probability Model of Self-Employment, Controlling for Skills 127 F.1 Difference in Mean of Self-Reported Readers Who Passed and Who Failed the Reading Assessment Core Literacy Test 131 F.2 Unexploited Potential: Reading Skill 134 F.3 Unexploited Potential: Writing Skill 136 F.4 Unexploited Potential: Numeracy Skill 139 F.5 Unexploited Potential: Computer Skill 142 G.1 Years of Education, Controlling for Socioemotional Skills 145 G.2 Linear Probability Model of Completing SHS or Tertiary Education, Controlling for Socioemotional Skills 146 G.3 Linear Probability Model of Being Employed, Controlling for Socioemotional Skills 147 G.4 Linear Probability Model of Being Self-Employed, Controlling for Socioemotional Skills 149 G.5 Linear Probability Model of Working in a Medium- to High- Skilled Occupation, Controlling for Socioemotional Skills 150 Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Foreword The past two decades in Ghana have been marked by steady economic progress, which has transformed the country into a lower-middle-income economy, accom- panied by a decline in poverty, increases in incomes for families, improvements in health, and expanded educational opportunities. As Ghana looks forward to a future of economic growth, it needs to regain growth from the 2016 slowdown, including improving competitiveness and economic diversification and raising labor productivity. A well-equipped workforce will be key to obtaining these goals. The Government of Ghana and its development partners such as the World Bank have long recognized the importance of investments in human capital. Insufficient skills in young people will be an obstacle to improving competitiveness in all sec- tors across the economy, be they informal or formal, in traditional sectors or in the modern areas such as information and telecommunications technologies. An agenda for improving skills in the workforce relies on being able to ­ identify where the more practical and profitable investments in the current skills profile should be made. To that end, Skills Toward Employment and Productivity (STEP) is an innovative tool used across the world to assess the education, cogni- tive, work-related, and socioemotional skills stock in a population, as well as the impact of these traits on employment and earnings. Ghana is among the first two countries (along with Kenya) in Sub-Saharan Africa where this systematic assess- ment of skills has been carried out. The evidence collected through this assess- ment shows that the multidimensional nature of skills requires nurturing from early childhood education to school and university systems, as well as through school-based and on-the-job training. This broad concept of skills has a significant impact both on jobs and on earnings, and the relationship is also mutual: jobs attract and reward skills. The information from the Stepping Up Skills in Urban Ghana study provides detailed insights for policy makers. These insights cover areas including investments in early childhood education, the role of improve- ments in the quality of education, and the creation of incentives for economic actors to invest in on-the-job training to improve Ghana’s competitiveness and the well-being of its citizens. Henry Kerali Jaime Saavedra Chanduvi Country Director for Ghana Senior Director, Education Global Practice The World Bank The World Bank Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   xi   Acknowledgments The authors thank Henry Kerali, World Bank Country Director for Ghana; Jaime Saavedra Chanduvi, Senior Director; Amit Dar, Director; Luis Benveniste, Director; and Peter Nicolas Materu, Meskerem Mulatu, and Halil Dundar, Managers of the World Bank Education Global Practice, for their overall leader- ship and management guidance. We would also like to thank Alexandria Valerio, Kathleen Beegle, Maria Laure Sanchez Puerta, and Omar Arias for peer reviews and technical guidance. The team also received valuable advice from Deborah Mikesell and Eunice Ackwerh. The survey work was carried out by a team of the Institute for Social Statistical and Economic Research of the University of Ghana, led by Frank Ochere. The literacy assessment was completed with technical support by the Educational Testing Service (Princeton, New Jersey). Technical support to data management was provided by Tania M. Rajadel and Sebastian Monroy Taborda, both at the World Bank. Financial support was provided by the Multi-Donor Education Trust Fund and the Bank-Netherlands Partnership Support Program, both ­ managed by the World Bank. Editing and publishing support were provided by Jonathan Faull and Aziz Gökdemir. Janet Adebo provided invaluable administrative support throughout the process. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   xiii   Overview Introduction Ghana stands at the cusp of extraordinary opportunity. Since the country’s return to democratic rule and the advent of the Fourth Republic, per capita gross domestic product (GDP, current US$, PPP [purchasing power parity]) has increased almost fourfold, from US$375 in 1993 to US$1,442 in 2014. In 2011, Ghana was the only African economy to demonstrate double-digit economic growth, surpassing 14 percent that year. In the aftermath of the Chinese economic slowdown and the slump in global commodities markets, Ghana’s growth tempered to 4 percent in 2014 but is expected to recover above 8 percent in the medium term as the country begins exploiting signifi- cant oil and gas resources. Throughout the period of the Fourth Republic, fertility rates have remained relatively high, falling from 5.3 births per female in 1993 to 4.2 in 2014. Concurrently, the proportion of children under the age of 15 in the total popula- tion declined only marginally from 43 percent in 1993 to 39 percent in 2014. As a consequence, Ghana’s labor force has grown rapidly, from approximately 6.5 million in 1993 to 11.3 million in 2014, and is expected to continue to grow in the coming decades. Population growth has also contributed to an erosion of the effects of buoyant economic growth on poverty reduction. Although the country has made progress in reducing poverty—with the proportion of the population living below the government’s poverty line falling from 32 percent in 2005 to 24 percent in 2012—stubborn disparities persist with regard to access to economic, social, and political opportunities. Inequity is particularly evident in the differences between the populations of the poorer northern Savannah regions and the rest of the country. In the three northernmost administrative regions of the country, more than half, or Unless stated, all statistics cited in the Introduction are drawn from the World Bank Group’s Data Bank, accessible at data.worldbank.org (accessed March 23, 2016). Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   xv   xvi Overview 58 ­ percent, of the population falls below the poverty line, compared to 19 ­percent in the seven administrative districts of the south of the country. The spatial distribution of poverty and economic opportunity, in turn, has led to significant migration from north to south, and a swelling of the ranks of the urban labor force. Nevertheless, a significant economic dividend is implicit in Ghana’s increas- ingly abundant urban labor force. If current and future generations of workers can be empowered to realize their potential in a vibrant and increasingly competitive economy, the prospect of sustained poverty reduction, further economic development, and the reaping of benefits associated with a demo- graphic transition could profoundly reshape Ghana’s society and economy for future generations. Equipping current and future generations of workers with the skills they need to improve their livelihoods and to drive increases in national productivity and competitiveness requires that these workers have skills appropriately aligned with the needs of a growing economy. Some skills are innate, arbitrarily assigned through the accident of birth. Other skills are acquired through education, work, and life experience. The primary means through which a government can develop the skills of its labor force is through policies and strategies implemented through the education system. Jobs form the foundation of economic development, rising living standards, increases in productivity, and improved social cohesion. Equipping people with appropriate skills to access meaningful work constitutes the means for achieving these objectives. This report, premised on the Skills Toward Employment and Productivity (STEP) framework, considers education the instrument for learning and acquiring skills and the key for accessing employment. Although the Ghanaian government has made significant progress in expanding access to basic education, many challenges persist in the education sector. These include the low quality of learning in both basic and post-basic education, inequity of access, and the limited capacity of the education system to equip beneficiaries with skills aligned with increasing competitiveness and productivity in the economy. If education and training institutions are unable to provide the skills demanded by the market, economic inefficiencies could be compounded. Skills development strategies can be effective only if they are appropriately calibrated to the needs of the economy and only if they take into account the current skills endowment of the existing labor force. The most effective skills development strategies are informed by evidence, and effective implementation requires the ongoing collection of data to gauge shifting demand for skills and the changing character of the labor force and economy at large. To date, evidence regarding the stock of existing skills in Ghana’s labor force is relatively underdeveloped. The precise measurement of the prevalence of dif- ferent categories of skills within the population is required for the effective design of policy interventions that target improved training to reduce skill gaps Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Overview xvii aligned with the needs of specific sectors, improvement in productivity, and increased employability of workers. With the data it gathers, the first STEP household survey aims, in part, to address this deficiency through a rigorous analysis of the skills endowment of urban Ghanaian adults. The STEP survey was carried out between September 2011 and December 2013 in Ghana, as part of the first wave of surveys initiated under the STEP Skills Measurement Program. The Ghanaian sample consists of about 3,000 individuals between 15 and 64 years of age, living in urban areas across 71 ­ districts. In addition to standardized information captured at the household level, the STEP survey collects information regarding the level of skill, level of education completed, and work history. On skills, the STEP survey includes information about (i) self-reported cognitive skills (that is, a subjective assess- ment of an individual’s use of foundation skills—reading, writing, and numeracy—​ at work and in daily life); (ii) assessed cognitive skills (that is, an objective assessment of reading literacy based on the International Adult Literacy Survey); (iii) socioemotional skills (that is, personality traits, behavior, and risk and time preferences); and (iv) job-specific skills (that is, an indirect assessment of skills used at work). The STEP framework is structured according to five iterative steps and associated objectives: (i) getting children off to the right start; (ii) ensuring that all students learn; (iii) building job-relevant skills; (iv) encouraging entre- preneurship and innovation; and (v) facilitating labor mobility and job match- ing. Ultimately, if these five steps are achieved, Ghana will significantly advance the likelihood of realizing the potential of its labor force, with con- siderable positive implications for the livelihoods of its people, poverty reduc- tion, improved productivity, ­ economic development, and the betterment of society at large. Objectives This report is intended to complement and support the work of the Government of Ghana as it seeks to accelerate progress toward the achievement of the ­ education-related Millennium Development Goals (MDGs) and the work of the Ministry of Education (MOE) and the Ghana Education Service (GES) in advancing the reforms envisaged by the Education Strategic Plan (ESP) for 2010–20. This report is divided into eight chapters: • Chapter 1 of the report describes Ghana’s economic, social, and demographic trends and the current challenges faced by the education system. Chapter 2 outlines the report’s framework for enquiry and a detailed description of the STEP framework. Chapter 3 provides an overview of the STEP survey’s sam- pling methodology, a description of methodologies used to measure each subset of skills, and limitations to the analysis. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 xviii Overview • Drawing on data collected through the STEP survey, chapter 4 describes the education profile of the urban adult population; explores trends in preschool, primary, secondary, and tertiary completion rates, disaggregated by age group, gender, and region; explores the effect of socioeconomic and demographic factors on educational attainment, dropout rates, and age of entry into the education system; and looks at the relationship between education attainment and inequality. • Chapter 5 analyzes labor force participation, employment status, and underemployment; the correlation between labor market status and skills ­ and education; and regional disparities in education and labor market opportunity. • Chapter 6 focuses on the use of cognitive skills (both self-reported and assessed), socioemotional skills, and technical skills. The authors analyze the mismatch between self-reported and assessed cognitive skills and the mis- match between respondents who report high levels of education attainment but who performed poorly when these skills were tested and vice versa. This section of the report aims to answer the following research questions: How do socioemotional skills relate to cognitive skills? Does inequality with regard to years of schooling reflect inequality in cognitive skills? What are the char- acteristics of respondents who under- or over-report cognitive ability? Do incremental increases in educational achievement result in more developed skill sets? • Chapter 7 looks at the association between education and skills and labor market opportunity. It discusses the extent to which it is worthwhile for individuals to invest in (or for others to subsidize) education and/or skills development and to what extent these investments inform the develop- ment of skills demanded by the economy. In so doing, this chapter aims to answer the following research questions: Does education lead to better job market opportunities? Do skilled (educated) workers demonstrate higher earnings? Does education inculcate job-relevant skills? Which skills matter the most for employability? How much of the wage premium accruing to educated people is explained by skills? Is training relevant in producing good skills? • Section 8 quantifies the magnitude of unused skills and underexploited poten- tial within the labor force, examines the proportion of the employed popula- tion who do not use their skills at work, and weighs evidence of skills and education mismatching in the labor market. Key Findings The Educational Profile of Ghana’s Adult Urban Population Ghana has made significant progress in expanding access to basic education, having achieved near universal access to primary education. However, challenges persist with regard to improving the quality of basic education to enhance learning outcomes and to expanding access to post-basic levels of education. ­ Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Overview xix Moreover, a significant gender gap in educational achievement is evident for all levels of education in Ghana. The expansion of early childhood education (ECE) in Ghana’s urban areas has dramatically improved, and participation in ECE is positively correlated with household socioeconomic status and the intensity with which workers use basic skills. Only 30 percent of Ghanaians between ages 45 and 64 attended an ECE program compared to 87 percent of the youngest generation surveyed. Seventy-seven percent of adults living in households in the upper socioeco- nomic brackets had participated in an ECE program, compared to 49 percent of adults in the poorest households. Adults with ECE are more likely to read and write regularly and do so with greater intensity. Approximately two-thirds of students who complete primary school do so without demonstrating proficiency in core subject areas. In terms of the quality of senior high school (SHS) programs, a large disparity in further educational attainment is evident between students attending the highest-performing schools and the rest of the sector. The highest-performing 10 percent of high schools account for 90 percent of students entering university. There are significant regional, gender-based, and income-based disparities in access to and the returns accruing to post-basic edu- cation. Pass rates for the Basic Education Certificate Examination (BECE) vary greatly by region, with the poorest performance evident in the administrative regions of the north. The magnitude of difference in BECE pass rates for the north and south are upwards of 60 percent. Access to tertiary education has not changed substantially over time. Over three generations, the percentage of people with a tertiary qualification increased by only two percentage points, from 11 percent for adults ages 45 to 64 to 13 percent for respondents in the generation ages 25 to 34. A child’s age on entry to the school system is correlated with educational achievement. Officially, children should commence formal schooling at the age of six; however, in practice many children enter school later, with negative impli- cations for further education. Approximately 51 percent of those with incom- plete primary education enrolled in primary school at eight years of age, whereas 89 percent of those with postsecondary education enrolled at the official age of entry. Approximately one out of four members of the adult labor force dropped out of school prior to completing the highest grade of the level of education they had enrolled in, with the highest rates of dropout evident in the primary and junior high school (JHS) cycles of education. The most commonly cited reason for dropping out of school before the age of 16 is a lack of money for out-of-pocket expenses, regardless of the child’s socioeconomic status. Labor Force Participation and Employment Status Characteristics The overall urban labor force participation rate in Ghana is 67 percent: 62 per- cent of respondents ages 15 to 64 years reported being employed, 5 percent reported being unemployed but looking for work, 23 percent were deemed Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 xx Overview economically inactive (22 percent in education, 1 percent had retired), and 10 percent fell into the category “Not in Education, Employment, or Training” (NEET). A significant majority of the employed labor force works in the informal ­ sector, primarily in self-employed work (66 percent) or as informal wage workers (20 percent). Only 15 percent of the employed population is engaged in salaried employment in the formal sector. However, these trends are changing, with younger generations less likely to be self-employed and more likely to hold sala- ried employment in formal or informal firms. Although female workers are as likely to be employed as their male coun- terparts, a significant gender gap exists in the quality of employment enjoyed by male and female workers. Across all age groups, female workers are more likely than male workers to be self-employed. Approximately half of male workers are self-employed, with the remaining share split almost evenly between formal and informal wage employment. By contrast, 79 percent of female workers are self-employed, and only 7 percent are retained as employ- ees in the formal sector. A worker’s level of education is strongly associated with employment status. Approximately 37 percent of workers holding a tertiary qualification ­ work in the informal sector (20 percent as self-employed and 17 percent as informal employees), compared to 99 percent of workers with no formal edu- cation (86 percent as self-employed and 13 percent as informal employees). A key determinant of a worker’s likelihood of accessing a job in the formal wage sector is education. Approximately 44 percent of workers in the formal sector have a tertiary education, 32 percent graduated from SHS, and 24 per- cent report JHS education or less. The level of educational attainment is also strongly associated with employ- ment in higher-skilled occupations. The majority of the employed population works in low-skilled occupations in the informal sector, and just 13 percent of the working population is employed in high-skilled occupations, concentrated in the formal sector. SHS and tertiary graduates are more likely to be employed in mid- or high-skilled occupations than are workers with lower levels of educa- tional attainment. Approximately 29 percent of SHS graduates are employed in the high-value-added services sector, compared to just 3 and 8 percent of workers in this sector having terminated their education following the primary and JHS cycles, respectively. The Skills Profile of Ghana’s Adult Urban Population As mentioned previously, the STEP survey collects information about cognitive skills by asking people about their use of numeracy, reading, and writing skills and through an objective assessment of their reading proficiency. According to self-reporting by those surveyed, working-age Ghanaians use their numeracy skills on a regular basis, regardless of their level of edu- cation. At the same time, the use of reading and writing skills is significantly Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Overview xxi lower and strongly correlates with the level of education and with employ- ment status. The intensity of the use of reading skills is higher among SHS graduates and highest for respondents with a tertiary qualification. All respondents who received tertiary education report using their reading and writing skills on a regular basis, and approximately 64 and 60 percent of the population with primary education report the regular use of reading and writing skills, respectively. Nevertheless, even among those who com- pleted tertiary education, almost half write only with low intensity. Similarly, 84 percent of formal employees reported using their reading and writing skills on a daily basis, compared with approximately 41 percent of informal wageworkers and 21 percent of self-employed workers. Overall, the use of reading and writing skills has intensified over time, with the youngest cohort surveyed reporting regular use of their writing (89 percent of respondents) and reading (90 percent) skills, compared to the oldest age cohort of workers who reported a lower regular use of reading (41 percent of respon- dents) and writing (41 percent). The results of the reading proficiency tests mir- ror the respondents’ self-reported levels of skill use. Educational attainment was highly predictive of relative success on the Core Literacy Test component of the survey. Ninety-three percent of those who failed the test reported educational attainment of JHS or less (42 percent report no formal education, and 33 percent attended JHS), whereas about 57 percent of those who passed the test have at least an SHS education and approximately 21 percent report having tertiary education. Nevertheless, average performance across all three measures of reading is low. Respondents who reported using their reading and writing skills more fre- quently performed better in sentence processing and passage comprehension tests than those who did not. However, poor performance on the print vocabulary subcomponent of the reading assessment was relatively uniform across all ­ subgroups. Respondents with higher reported levels of educational attainment performed better on all aspects of the reading component. Younger respon- dents were more likely to perform better in sentence and passage comprehen- sion tests than older respondents. Respondents with lower socioeconomic status, and for whom English was not their main language at home or at work, were less likely to pass the Core Literacy Test. Of those who passed the test, 73 percent reported using English as their primary workplace language, compared with 17 percent of those who failed the test. However, in light of the fact that the reading test was administered in English, it is not surprising that a higher percentage of those who passed the test report speaking English as their primary language at home and work than those who failed. Of the respondents who were employed, 81 percent of those who failed the reading test were self-employed and only 3 percent were formal wage- workers. By contrast, 38 percent of those who passed the reading assessment were self-employed and 34 percent were formal wageworkers, with one-third retained in high-skilled occupations. Respondents who passed the reading Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 xxii Overview assessment earned on average 274 GHS (approximately US$80) per month more than those who failed the test. Average performance on the Literacy Exercise Booklets section of the test, which allows for an in-depth analysis of reading ability, was very poor. Sixty-one percent of the subsample to whom the Literacy Exercise Booklets were administered scored on the lowest rung of proficiency. Respondents who attained higher scores were more likely to be younger, to have reported higher levels of educational attainment (SHS or tertiary), to work in the for- mal sector, to have reported the use of English as their primary home or workplace language, and to have reported using their reading and writing skills more intensively. Finally, it is worth mentioning that the data show a substantial mismatch between self-reported use of skills and tested literacy: of the sample as a whole, approximately 69 percent of adults reported reading regularly, but only 42 ­percent passed the Core Literacy Test component of the literacy assessment. However, the poor results in the Core test may reflect poor English literacy rather than present a true measurement of general reading skills. Those who reported being able to read but who then failed the Core test are more likely to be women, more likely to work in low-skilled occupations and to be less edu- cated, and less likely to work in the high-value-added sector. Respondents who reported being able to read but then failed the Core test also earn approxi- mately 41 percent less than those who reported being able to read and who passed the Core test. The analysis suggests that socioemotional skills are closely associated with educational attainment and labor market outcomes. Individuals with higher self-reported scores relative to the “Big Five” personality traits (stability, agree- ableness, extraversion, conscientiousness, and openness) also had higher educa- tional attainment. Similarly, individuals scoring higher on the Big Five personality traits, and who demonstrated lower levels of hostile bias, are less likely to be self-employed. The association between socioemotional skills and labor market outcomes is reversed nevertheless when controlling for education. This finding confirms that education and socioemotional skills are (positively) correlated and also suggests that the development of socioemotional skills might be one of the channels through which education can contribute to better labor market outcomes. Job-Specific Skills Employment in the formal sector requires that a worker be capable of regularly learning new skills at work, more frequently involves the supervision of others’ work, and more often requires workers to make presentations. Survey data dem- onstrate that work in the informal sector is less cognitively demanding, more repetitive, and more physically demanding. Self-employed workers reported greater autonomy at work and a greater incidence of repetitive tasks than did wage workers. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Overview xxiii Further Education and Training Approximately 32 percent of the population surveyed had actively engaged in activities to further develop their skills in the year prior to taking the survey. More than one in four respondents surveyed by STEP reported participating in an apprenticeship, and less than 7 percent had participated in a training course of 30 hours or more. Apprenticeship-based training is more common among self-employed work- ers, whereas formal wageworkers are more likely to enroll in skills training courses. Approximately 26 percent of formal workers reported attending a work-related or personal skills training course in the year prior to the survey, compared to 3 and 6 percent of self-employed and informal wageworkers, respectively. On the other hand, approximately 38 percent of self-employed and 25 percent of wage- workers had participated in an apprenticeship in the 12 months prior to the survey. Returns to Education and Skills The analysis demonstrates that an additional year of education is associated with an increase of 6–10 percent in monthly earnings. After controlling for socioeconomic characteristics, primary education is associated with no distin- guishable earnings premium for people with no formal education or incomplete primary education. On the contrary, the wage premium accruing to SHS and tertiary graduates is 29–79 percent (SHS) and 112–172 percent (tertiary) higher than the average wage of workers with no formal education or incom- plete primary education. The premium for workers with JHS over workers with no formal education or incomplete primary education is much lower (35–59 percent). Returns to education and skills vary by type of employment and with respect to the type of skill. An additional year of education increases the monthly earnings of informal wageworkers by 4–6 percent, compared to 7–10 percent for formal wageworkers. However, taking into account the individual’s skills (cognitive, socioemotional, and job-related skills), the earnings premium associated with any additional year of education decreases ­ by 2 percent. The premium accruing to a female worker through an additional year of education is higher than that accruing to male workers (between 9 and 13 per- cent for female workers, compared to between 6 and 9 percent for men). For male workers, a pronounced premium associated with educational attainment becomes evident only at the tertiary level, whereas the data suggest that sub- stantial payoffs accrue to female workers from JHS and above. Furthermore, with respect to skills, the pro-women gender gap is persistent at tertiary educa- tion (female workers with tertiary education make 164 percent more than those without compared to a 118 percent advantage for men). However, this apparent gender gap may be misleading because it does not take into account the fact that average labor market participation and educational attainments are lower for female workers. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 xxiv Overview Evidence of an Unrealized Potential in the Labor Force An apparent mismatch between skills and their use at work could reflect unex- ploited human capital in the labor force. Within the overall population of employed adults, the evidence suggests that there is a subset of individuals who use their cognitive and computer skills at home but not at work. Moreover, there is a disproportionately large number of young workers who report being able to use a computer but who are not required to do so at work. The evidence suggests that the greatest residue of unexploited potential is located in low- skilled occupations and low-value-added sectors. Approximately 34 percent of self-employed workers report reading at home but never using this skill at work. On the other hand, only 6 percent of formal sector workers report this skills mismatch. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Abbreviations ALL Adult Literacy and Life Skills Survey BECE Basic Education Certificate Examination ECE early childhood education ESP Education Strategic Plan GDP gross domestic product GES Ghana Education Service GLSS V Ghana Living Standards Survey (Fifth Round) GSS Ghana Statistical Service IALS International Adult Literacy Survey JHS junior high school MDG Millennium Development Goals MOE Ministry of Education NEET not in education, employment, or training OECD Organisation for Economic Co-operation and Development PIAAC Programme for the International Assessment of Adult Competencies PPP purchasing power parity SES socioeconomic status SHS senior high school SSA Sub-Saharan Africa STEP Skills Toward Employment and Productivity TVET technical and vocational education and training WAEC West African Examinations Council WASSCE West African SHS Certificate Examinations Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   xxv   Chapter 1 Country Context Economic, Social, and Demographic Trends In the contemporary period, Ghana has achieved sustained economic growth and significant poverty reduction. The country has experienced 20 years of positive economic growth (in the range of 4–5 percent) and in 2011, with year- on-year growth of 14.4 percent, was one of only seven countries in the world, and the only country in Sub-Saharan Africa (SSA), to achieve double-digit economic growth (IMF 2012).1 Improved economic performance has been accompanied by significant pov- erty reduction, underpinned by rural development and increasing urbanization. In rural areas, small-scale agriculture has benefitted from improved agricultural productivity (notably in cocoa), rising incomes, and rising domestic demand. Concurrently, Ghana’s rapidly growing urban centers have led to a significant expansion of the service sector and a growing labor force, inclusive of migrants from rural areas who have been absorbed into better-paying jobs in both the formal and the informal sectors of the economy. Despite this impressive progress, deep inequity continues to characterize Ghanaian society and is reflected in significant disparities in access to economic, social, and political opportunities. This is especially evident when observing dif- ferences in access to opportunity between the populations of the poorer northern “Savannah regions” and the rest of the country. The bulk of Ghana’s poverty is concentrated in the three northernmost administrative regions of the country: the Upper East, Upper West, and Northern regions. These Savannah regions (home to approximately one quarter of the country’s population) are the locus of an average poverty rate of 58 ­percent, compared to 19 percent in the seven administrative districts in the south of the country. Between 1995 and 2005, the number of people living in poverty fell by 2.5 million in the south. Over the same period the number of people living in poverty in the north increased by 0.9 million, although the poverty rate for this area declined because of an absolute increase in the population.2 Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   1   2 Country Context Unemployment and underemployment in Ghana are structural in nature, which is not atypical in SSA. However, when demographic trends are taken into account, the structural nature of the challenges becomes more pro- nounced, with most economies on the continent being unable to generate suf- ficient rates of growth to create the jobs required to absorb the ever-increasing number of workers entering the labor force. The scale of the challenge is evi- dent in the fact that formal private sector employment accounts for only a small proportion of available manpower. To sustain and accelerate economic growth and poverty reduction, Ghana’s education strategy is focused on building the skills profile of the youth demo- graphic (IMF 2012). The urgent need to intervene in this regard is underlined by a national median age of approximately 20 years (UNDP 2011). Although the country’s youth demonstrate the highest levels of educational attainment relative to older generational cohorts, younger workers also depend most on salaried employment because of the inability of rural agriculture to sustain workers and their families and the contingent movement of migrants from the countryside to Ghana’s urban centers. From a developmental perspective, the most important demographic trend is the pace and character of migration from the northern regions of the country to the south (figure 1.1). Migrants to cities in search of jobs, education, skills, and business opportunities disproportionately include large family groups with children of school-going age and unemployed youth with poor educational attainment. Figure 1.1 Net Migration Rate, by Region Upper West Upper East Northern Brong-Ahafo Ashanti Eastern Volta Greater Accra Central Western –400 –300 –200 –100 0 100 200 300 400 Number of migrants, in thousands Source: GSS 2013. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Country Context 3 Education and Skills Ghana has made significant progress in expanding access to basic education over the course of the past 15 years. Having achieved near universal access to primary education, the country now faces two new urgent challenges: (i) improving the quality of basic education to enhance learning outcomes and (ii) expanding access beyond basic levels of education. The majority of students receiving upper secondary education in Ghana are enrolled in three-year senior high schools (SHS), following the completion of 11 years of basic education (comprising preprimary school, primary school, and junior high school [JHS]) (see box 1.1 for more information on Ghana’s education system). The government intends to universalize access to upper sec- ondary education by enabling students who are unable to afford fees to attend secondary school at no charge. Informal apprenticeships currently equip many more workers with skills than do formal public technical and vocational education and training (TVET) programs.3 Fewer than 10 percent of workers with technical and vocational skills acquired their skills through public TVET institutions. Moreover, there are approximately one and a half times as many trainees enrolled in private TVET institutions as in public institutions, and the number of students engaged in informal apprenticeships is 10 times that of students enrolled in the formal TVET sector. Box 1.1 The Ghanaian Education System In theory, the age range for children and young people engaged in the Ghanaian education system spans from 3 to 21 years. The educational system in Ghana consists of the following cycles or levels of education: • Preschool, equivalent to U.S. kindergarten (ages 4–5) • Primary school (ages 6–11) • Junior high school (JHS), equivalent to U.S. middle school (ages 12–14) • Senior high school (SHS), equivalent to U.S. high school (ages 15–17) • Tertiary education/institution, equivalent to U.S. college/university (ages 18–21) A full cycle of basic education is optimally provided over the course of 11 years, comprising two years of kindergarten, six years of primary school, and three years of JHS. Ghanaian children enter class one (first grade) during the calendar year in which they reach their sixth birthdays. During the first three years, the medium of instruction is either English or a combination of English and local languages. JHS education comprises forms 1 through 3 (U.S. grades seven through nine). Admission to JHS is open to any student who has completed primary class six. There is no entrance exam, and JHS education is considered integral to the ­ country’s nine-year cycle of basic education, to which all Ghanaian children are entitled. At the end of JHS form 3 (ninth grade), students sit box continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 4 Country Context Box 1.1  The Ghanaian Education System (continued) for the  Basic Education Certificate Examination (BECE) in nine or ten subjects. The BECE is administered and graded by the West African Examinations Council (WAEC). Admission to SHS is based exclusively on a student’s BECE results. After JHS, students may choose to go into different streams within the SHS system or to pursue further skills acquisition through an apprenticeship scheme with some support from the government. All SHS courses prepare students for university education, but most TVET students are likely to join the labor market once they graduate. SHS consists of forms 4 through 6 (equivalent to U.S. grades 10 through 12). The core SHS curriculum comprises six subjects—English, science, mathematics, social studies, physical education, and religious and moral education—which are studied throughout the three-year SHS cycle. Students undergo examinations only in the first four of these subjects. To complete a full course of SHS education, each student must choose—in addition to the core ­curriculum— one program (general arts, general science, agriculture, economics, business, or technical and vocational) and three or four elective subjects from within that chosen program. At the end of the three-year SHS cycle, all students are required to sit the West African Senior School Certificate Examinations (WASSCE). Usually, the student’s overall score is deter- mined by aggregating the student’s grades in his or her elective subjects and then adding this score to the aggregate score of the student’s best “core” subjects, with scores in English and mathematics considered first. The Ministry of Education considers any SHS graduate with an aggregate score of 24 (a D average) or better to be a successful school-leaver, equivalent to a graduate of a U.S. high school. Entrance to universities is by examination following the com- pletion of SHS. Students obtaining aggregate scores of 36 or above (six subjects) on the WASSCE can enter university. After completing SHS, students also have the option to pursue further education through a polytechnic. Persistent Challenges in Education Despite Ghana’s steady economic growth and improved access to basic educa- tion, many challenges persistent in the education sector. Among other things, the primary challenges concern • The poor quality of learning outcomes, in both basic and post-basic education; • Inequity of access, especially in secondary education; and • The limited capacity of the education system to create relevant skills aligned with increasing competitiveness and productivity within the economy. The National Education Assessment carried out biannually since 2005 shows a persistent trend: approximately two-thirds of students who complete primary school do so without demonstrating proficiency in core subject areas. In terms of the quality of SHS programs, the annual WASSCE demonstrates a large disparity in further educational attainment between students attending the 100 highest- performing 10 percent performing schools and the rest of the sector. The highest-­ Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Country Context 5 of high schools account for 90 percent of students entering university. Approximately 65–70 percent of SHS graduates do not continue education at the tertiary level. Students exiting the SHS cycle with low WASSCE scores are unable to access tertiary education, with many of these graduates going on to enroll in some form of training or apprenticeship program. There are significant regional, gender-based, and income-based disparities in access to, and the relevance of, post-basic education. The 2010 Population and Housing Census and the 2013 Ghana Living Standards Survey (GLSS) demon- strate very low pass rates for the BECE in the Upper East (11 percent), Volta (17 percent), and Northern and Upper West (both 22 percent) administrative regions, compared to high pass rates in the Greater Accra (90 percent), Western (88 percent), Brong-Ahafo (86 percent), and Ashanti (81.5 percent) regions (GSS 2013, 2014). Survey data also demonstrate that the limited capacity of Ghana’s education and training systems inhibits their ability to produce human capital of sufficient quality to meet the needs of the labor market and to drive a more competitive and diversified economy. Many JHS and SHS graduates are unable to find jobs in the formal sector, or are unable to pursue further education and training because of poor performance in exit examinations, a lack of information, and/or the suboptimal supply of training providers. Consequently, many JHS and SHS graduates are limited to self-employment or finding employment in the informal sector. According to data from the most recent National Census (2010), approxi- mately two-thirds of the adult population is self-employed, and the proportion of employment accounted for by the formal private sector has declined since 2000 (GSS 2013). Formal (public or private) employment accounts for only 17 percent of total employment and is disproportionately concentrated in urban areas, although a relatively large minority of public employees—primarily teachers—are present in rural areas. ­ Skills development in Ghana encompasses the inculcation of foundational, or basic, skills (literacy, numeracy); transferable and soft skills; and technical and vocational skills. These skills are acquired over the course of a lifetime through formal education, training, and higher education; through on-the-job and profes- sional training; and through the family, community, and media. The majority of young Ghanaians acquire technical and vocational skills through informal on- the-job apprenticeships (Darvas and Palmer 2014). Although the scale and scope of Ghana’s TVET systems are difficult to measure, clear opportunities exist for further skills development at the intersection of education, youth, and the needs of the labor market (Darvas and Palmer 2014). Enterprise surveys demonstrate mixed perceptions on the part of firms with respect to the quality of workers’ skills and the extent to which poor skills act as a constraint to improved economic performance (Darvas and Palmer 2014). However, in light of the demonstrable inadequacy of the supply of relevant and quality skills to the economy, a more deliberate analysis of enterprise survey data is required to more effectively understand the underlying causes of the apparent low demand for skills. If strategies for the further development of Ghana’s Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 6 Country Context human capital and economic development are to be successful, it is imperative that skills development strategies are premised on a thorough understanding of the existing supply of and demand for skills, as well as of the use of skills in the economy, in order to more effectively match the supply of skills to demand (Campbell 2012; Darvas and Palmer 2014). Despite the central role skills play in shaping employment outcomes, there is very little information about the distribution of different types of skills in the Ghanaian labor force or of their contribution to labor market outcomes. Ultimately, the paucity of relevant information undermines the design of more effective skills development policies and programs (World Bank 2014). The Skills Toward Employment and Productivity (STEP) project—through its use of a framework enabling the analysis of Ghana’s labor market conditions based on internationally comparable skills data—will help to address this knowledge gap and assist in the development of policies aligned to the needs of the economy. Notes 1. Real gross domestic product (GDP) growth was 8 percent in 2010, 4 percent in 2009, 8.4 percent in 2008, 6.5 percent in 2007, 6.1 percent in 2006, 6 percent in 2005, 5.3 percent in 2004, 5.1 percent in 2003, and 4.5 percent between 1993 and 2002 (IMF 2012b, 196; IMF 2011, 185). 2. Calculations based on the Ghana Living Standards Surveys GLSS3, GLSS4, GLSS5 available at http://documents.worldbank.org/curated/en/2011/03/14238095/tackling​ -poverty-northern-ghana. 3. This report uses the terms “technical and vocational education and training (TVET)” and “technical and vocational skills development” interchangeably. References Campbell, M. 2012. “Skills for Prosperity? A Review of OECD and Partner Country Skill Strategies.” LLAKES Research Paper 39, University of London, United Kingdom. Darvas, P., and R. Palmer. 2014. Demand and Supply of Skills in Ghana: How Can Training Programs Improve Employment? Washington, DC: World Bank. GSS (Ghana Statistical Service). 2013. 2010 Population & Housing Census: National Analytical Report. Accra: GSS. ———. 2014. Ghana Living Standards Survey Round 6 (GLSS 6): Main Report. Accra: Ghana. IMF (International Monetary Fund). 2012. “Ghana: Poverty Reduction Strategy Paper.” IMF Country Report 12/203, IMF, Washington, DC. UNDP (United Nations Development Programme). 2011. Human Development Report 2011. Sustainability and Equity: A Better Future for All. New York: UNDP. World Bank. 2014. “STEP Skills Measurement Program.” World Bank, Washington, DC. http://microdata.worldbank.org/index.php/catalog/step/about. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Chapter 2 Conceptual Framework: Why Is It Important to Focus on Skills? Introduction It is generally recognized that there is a strong positive correlation between the educational and skills profile of the workforce and a country’s per capita gross domestic product (GDP)—which is intended here as a proxy of productivity. Thus, education and training are the foundations for a skilled workforce and would benefit not only individuals but also the economy as a whole. If the education and the training system are of low quality, however, workers will be ill-prepared for the labor market. This could lead to several types of mismatches. For example, the education level of an individual may not match his or her skills, or individuals may overrate their own skills, thus leading to a mismatch between self-reported skills and tested skills. In another type of mismatch, workers may lack the skills required for available jobs or lack access to high-quality training programs that would increase their skill level and enable them to apply for higher-productivity jobs. As a result, many employees are either under- or overqualified for their occupa- tions. These mismatches lead to a loss of human capital and show that having post-basic or higher education does not automatically translate into getting a good job. In all of these cases of mismatch, investments in education may not pay off. Education policy makers and other education stakeholders in Ghana are pay- ing more attention to learning outcomes by testing all students from early grades through basic and post-basic education. Various research and diagnostic work has highlighted the fact that the reading and mathematics proficiency of Ghana’s students continues to be below acceptable levels. This means that most of them are unable to reach important milestones in literacy, to access post-basic educa- tion, or to develop a foundation for lifelong learning. As a result, their employ- ment prospects are limited. Policy makers have identified the low quality of inputs and the limited relevance of science and technology education as being Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   7   8 Conceptual Framework: Why Is It Important to Focus on Skills? among the main causes of these limited prospects. Policy makers also highlight Ghana’s large disparities in learning outcomes and related inequalities in the delivery of education services.1 Furthermore, policy makers and researchers have also homed in on several persistent challenges on the demand side. Not only do services and educational performance vary greatly by students’ social and economic status and by geogra- phy, but demand for skills, job opportunities, productivity, and expectations also vary widely and are affected by similar disparities. Despite the central role played by skills in improving employment outcomes and increasing productivity and growth in Ghana, information about the supply and demand of skills is sparse. Assessments of supply have largely focused on the outputs of mostly school and tertiary-level education and training institutions, whereas assessments of demand rely only on the responses to a few questions in enterprise surveys and rate-of-return analyses based on the last three Ghana Living Standard Surveys (GLSS) carried out in 1995, 1999, and 2005.2 Recently, more systemic analyses, surveys, and impact evaluations have been initiated, and we hope their results will inform future policy making. The lack of data and information on the skills endowment has made it difficult to design skills development policies and programs in Ghana, as in many other Sub-Saharan African (SSA) countries. Precise measurement of the prevalence of different types of skills among the population is needed to inform the design of public policies to reduce skill gaps in specific sectors, to increase the employ- ability of the population, to improve training, and to enhance productivity. The Skills Toward Employment and Productivity (STEP) framework consid- ers education as the instrument that enables individuals to learn and acquire skills, and subsequently access employment (box 2.1). Jobs are the foundation of economic development, improved standards of living, improved productivity, Box 2.1 Skills Toward Employment and Productivity (STEP) The STEP framework (see figure B2.1.1) is informed by the following five interlinked steps: • Step 1. Getting children off to the right start—by developing the technical, cognitive, and behavioral skills conducive to high productivity and flexibility in the work environment through early child development (ECD), emphasizing nutrition, stimulation, and basic cog- nitive skills. Research shows that the handicaps built early in life are difficult if not impossible to remedy later in life and that effective ECD programs can have a very high payoff. • Step 2. Ensuring that all students learn—by building stronger systems with clear learning standards, good teachers, adequate resources, and a proper regulatory environment. Lessons from research and ground experience indicate that key decisions about education systems involve how much autonomy to allow and to whom, accountability from whom and for what, and how to assess performance and results. box continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Conceptual Framework: Why Is It Important to Focus on Skills? 9 Box 2.1  Skills Toward Employment and Productivity (STEP) (continued) Figure B2.1.1 The Skills Toward Employment and Productivity (STEP) Framework Productivity and growth Getting children Ensuring that Building Encouraging Facilitating labor o to the right all students learn job-relevant entrepreneurship mobility and start skills and innovation job matching Source: Banerji et al. 2010. • Step 3. Building job-relevant skills that employers demand—by developing the right incentive framework for both pre-employment and on-the-job training programs and insti- tutions (including higher education). There is accumulating experience showing how public and private efforts can be combined to achieve more relevant and responsive training systems. • Step 4. Encouraging entrepreneurship and innovation—by creating an environment that encourages investments in knowledge and creativity. • Step 5. Matching the supply of skills with the demand—by moving toward more flexible, efficient, and secure labor markets. Avoiding rigid job protection regulations while strength- ening income protection systems, complemented by efforts to provide information and intermediation services to workers and firms, is the final complementary step transforming skills into actual employment and productivity. Source: Banerji et al. 2010. and social cohesion. As a consequence, equipping people with the right mix of skills is the means for achieving these objectives. Which Skills Are Relevant? A worker’s skill set comprises different categories of skills, including cognitive skills, social and behavioral skills, and technical skills (see figure 2.1). These cat- egories of skills relate to job skills relevant to specific occupations as well as to Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 10 Conceptual Framework: Why Is It Important to Focus on Skills? Figure 2.1 Skills Classification Cognitive Social and Behavioral Technical Involving manual dexterity and Involving the use of logical, Soft skills, social skills, life-skills, the use of methods, materials, intuitive and creative thinking personality traits tools and instruments Openness to experience, Technical skills developed Raw problem solving ability vs. conscientiousness, extraversion, through vocational schooling knowledge to solve problems agreeability, emotional stability or acquired on the job Verbal ability, numeracy, problem Self-regulation, perseverance, Skills related to a specific solving, memory (working and decision making, interpersonal occupation (e.g. engineer, long-term) and mental speed skills economist, IT specialist, etc) Source: Pierre et al. 2014. cognitive ability and the various personality traits that inform relative success in the labor market (Pierre et al. 2014). Cognitive skills include the use of logic, intuition, and critical thinking, as well as problem-solving skills deployed using acquired knowledge. They also include literacy, numeracy, and the ability to understand complex ideas, to apply lessons accrued through experience, and to analyze problems using logical thought processes. Social and behavioral skills relate to personality traits that are linked to labor market success, such as an individual’s relative openness to new experiences, conscientiousness, extraver- sion, agreeability, and emotional stability. Technical skills range from manual dexterity in the use of complex tools and instruments to occupation-specific knowledge and skills for use in professional and technical occupations such as engineering or medicine. How Are Cognitive, Behavioral, and Technical Skills Formed? The process of skills formation should be considered as a continuum spanning an individual’s life, with critical stages for the development of both cognitive and socioemotional skills. Four features of skills formation are particularly relevant to the development of a skills strategy: 1. Foundational cognitive and behavioral skills are formed early in an individual’s life and serve as a platform upon which other skills are developed. Children who fall behind early face significant disadvantages in catching up with their peers. Non-cognitive skills appear to be more malleable throughout life, but early interventions appear to have significant positive effects for their further development in the longer term. 2. Aggregate skills formation and development benefit from previous invest- ments in skills development and are cumulative over an individual’s life. For example, a child who has learned to read fluently by second grade will be able Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Conceptual Framework: Why Is It Important to Focus on Skills? 11 to absorb more in third grade than a child of equivalent age who cannot read at the commencement of grade three. This implies that early investments in skills development are likely to have a greater longer-term impact on aggregate skills formation because it is easier and less costly to develop these skills when children are young and most receptive to learning. 3. Social and behavioral skills are particularly valuable early in a child’s life because they support, and benefit from, the development of cognitive skills. For example, children who are open to new experiences are more likely to be imaginative and creative and to apply themselves at school. 4. The acquisition of technical and job-specific skills is facilitated by strong cog- nitive and behavioral skills acquired earlier in the education system. These skills are often acquired last, through technical and vocational education and training (TVET), higher education, and on-the-job learning. The skills learned in formal education help workers to continuously update their technical skills during their working lives. Understanding the Focus on Adults and Urban Areas Early investments are smart and cost-effective and prepare future generations of workers for the labor market. However, skills development strategies should also focus on maximizing the potential of current generations of workers. In recent decades Ghana has experienced large-scale migration from more rural and comparatively impoverished areas to Ghana’s more prosperous and urban areas. Migrants are disproportionately young adults who are motivated by educational opportunities and jobs in cities. Urbanization represents both a chal- lenge and an opportunity for development because it has been associated with rising per capita income in Europe, Latin America, and, more recently, Asia. However, Africa has been an exception in this regard, because of the fact that in many African countries, including Ghana, industrialization has generally not accompanied urbanization. The majority of migrants to Ghana’s urban areas find employment in the com- paratively poorly productive informal sector. However, even informal employ- ment is, in many instances, much more than migrants could aspire to in their home regions. Equipping migrants with skills aligned with the needs of the urban labor market would give them a further opportunity for self-improvement. Accurately measuring the skills of the adult population is critical for informing policies aimed at increasing the productivity of those already in the labor force. Understanding the demographic structure and the socioeconomic status of the adult population in Ghana, as well as the associations between levels of educa- tion and an individual’s skills endowment, will help to inform the development of effective nonformal education and skills development programs aligned with the specific contextual characteristics of Ghanaian society and the labor force. It is for this reason that the STEP Skills Measurement Survey (discussed in the next chapter) is premised on gathering information about the skills endowment of the adult urban population. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 12 Conceptual Framework: Why Is It Important to Focus on Skills? Notes 1. References on equity, demand and supply work, and other analyses include Balwanz and Darvas (2014) and Darvas and Palmer (2014). 2. The GLSS 6 survey results have just been released, and a later version of this report will also analyze these results (GSS 2014). References Balwanz, David, and Peter Darvas. 2014. Basic Education beyond the Millenium Development Goals in Ghana: How Equity in Service Delivery Affects Educational and Learning Outcomes. Washington, DC: World Bank. Banerji, Arup, Wendy Cunningham, Ariel Fiszbein, Elizabeth King, Harry Patrinos, David Robalino, and Jee-Peng Tan. 2010. Stepping Up Skills for More Jobs and Higher Productivity. Washington, DC: World Bank. Darvas, Peter, and Robert Palmer. 2014. Demand and Supply of Skills in Ghana: How Can Training Programs Improve Employment? Washington, DC: World Bank. GSS (Ghana Statistical Service). 2014. Ghana Living Standards Survey Round 6 (GLSS 6): Main Report. Accra: Ghana. Pierre, Gaëlle, Maria Laura Sanchez Puerta, Alexandria Valerio, and Tania Rajadel. 2014. STEP Skills Measurement Surveys: Innovative Tools for Assessing Skills. Social Protection and Labor Discussion Paper 1421. Washington, DC: World Bank. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Chapter 3 Skills Toward Employment and Productivity Data Introduction The Skills Toward Employment and Productivity (STEP) survey provides a unique opportunity to collect information on the use of skills and skills profi- ciency in Ghana. The objective of the study is to yield a clearer understanding of the complex relationship between skills, employment, and productivity. This chapter discusses the characteristics of the sample used in the Ghana STEP sur- vey, the data arising from the survey, and the definitions of skill and other vari- ables used in this analysis. Data and Sample Characteristics This household survey was carried out in Ghana as part of the first wave of sur- veys under the STEP Skills Measurement Program between September 2011 and December 2013. The Ghanaian sample was gathered through a two-stage random sampling of households and individuals. It consists of 2,987 individuals between 15 and 64 years of age. Table A.1 in appendix A provides a complete description of the sample in terms of demographic characteristics, education, language, labor and employment status, economic sector and occupation, and geographic region. It is important to note that the weighted sample represents only the urban population. Therefore, the findings of this report cannot be extended to the national level because the urban population is likely to differ from the rural population along many dimensions. For example, workers in the urban areas are likely to have higher education levels and to be relatively more concentrated in the services sector than workers in the rural areas. Along with the standard information captured at the household level, the STEP survey collects extensive information on the skills level, education, and work history of one individual between ages 15 and 64 randomly selected from Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   13   14 Skills Toward Employment and Productivity Data Figure 3.1 The STEP Household Survey Instrument Household Random selection of Urban individual selected information one HH respondent aged 15–64 PART A Cognitive Job-specific Socioemotional Background Household roster skills skills skills PART B Education Dwelling pathways characteristics Self-reported PART C Health Direct assessment Identification of (Basic or Extended respondent Test) Employment history Family background Source: Banerji et al. 2010. Note: HH = household; STEP = Skills Toward Employment and Productivity. each sampled household (figure 3.1). The survey includes three innovative modules on skills: (i) a specially designed assessment of reading literacy and competence to access, identify, integrate, interpret, and evaluate information, scored on the same scale as the test in the Program for the International Assessment of Adult Competencies (PIAAC) of the Organisation for Economic Co-operation and Development (OECD); (ii) a battery of questions capturing self-reported information on personality traits and behaviors; and (iii) a series of questions on job-specific skills that the respondent possesses or uses in his or her jobs. The skills of the entire sampled population are captured, irrespec- tive of their labor force status (employed, unemployed, or inactive) or sector of employment. Types of Skills Measured The STEP survey evaluates three categories of skill: (i) cognitive (self-reported and direct assessment); (ii) socioemotional skills; and (iii) job-specific skills. Table 3.1 and appendix B summarize the dimensions of skills captured in each category, and the corresponding survey questions used to create the relevant variables for analysis. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Skills Toward Employment and Productivity Data 15 Table 3.1  Definitions of Skill Types Skills type Definition Measure Skills Cognitive The “ability to understand Indirect assessment (self- Reading complex ideas, to adapt reported) of individual’s use of Writing effectively to the foundation skills at work and Numeracy environment, to learn from in daily life experience, to engage in Direct assessment of reading Reading proficiency various forms of reasoning, literacy based on the [and] to overcome obstacles International Adult Literacy by taking thought”a Survey (IALS) using the same scale as the Programme for the International Assessment of Adult Competencies (PIAAC) Socioemotional Related to traits covering Personality traits Openness skills multiple domains (social, Conscientiousness emotional, personality, Extraversion behaviors, attitudes, Agreeableness and so on) Emotional stability Persistence Behavior Decision making Hostility bias Risk and time preferences Risk-taking preferences Time preferences Job-specific skills Task related and build on a Qualifications required for the job Qualification requirement for combination of cognitive and and job learning times current job noncognitive skills Learning times Indirect assessment of skills used Autonomy and repetitiveness at job Computer use Contact with clients Solving and learning Supervision Physical tasks Note: For further details, see appendix B and Pierre et al. 2014. a. Neisser et al. 1996. Cognitive Skills The STEP survey measures cognitive skills in two ways: First, respondents are asked to report whether and with what intensity they use their reading, writing, and numeracy skills in daily life and at work (if they work). These measures are likely to capture a combination of an individual’s actual ability to conduct tasks involving these skills and their motivation or opportunity to do so. Second, the survey makes use of an objective literacy assessment to directly measure cognitive skills. The self-reported questions capturing the use of reading and writing skills may substantively differ from an individual’s actual ability to read or write. The STEP literacy test has been aligned with a series of large-scale international surveys, such as the International Adult Literacy Survey (IALS), the Adult Literacy and Life Skills Survey (ALL), and the PIAAC. The assessment consists of three parts: Reading Components, the Core Literacy Test, and the Literacy Exercise Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 16 Skills Toward Employment and Productivity Data Booklets, as shown in figure 3.2. The Reading Components focus on foundational reading skills, including a respondent’s ability to assess word meaning, process sentences, and comprehend passages. The Core Literacy Test consists of eight lit- eracy questions, which function as a screen to sort the least literate respondents from those with the highest literacy skills. Only respondents who were capable of answering two or more questions of the Core Literacy Test correctly are asked to complete the Literacy Exercise Booklets. The Booklets are broken down into four parts with a total of 18 questions, which allow for a more granular evaluation of reading skills among the relatively more literate respondents. All tests are con- ducted in English (see box 6.1 for further discussion). Cumulatively, the data gathered by the STEP survey consist of: (i) the Reading Components score(s),1 (ii) pass/fail information for the Core Literacy Test, and (iii) the score obtained in the Exercise Booklets grouped into five levels of com- petency (table 3.2), inclusive of descriptions of basic literacy skills from 1, the lowest level, to 5, the most advanced level. Socioemotional Skills Socioemotional skills refer to skills and traits that are not directly related to intelligence, such as social, emotional, personality, behavioral, and attitudinal skills. The measures used to capture behavioral attributes are less established Figure 3.2  Workflow for the STEP Skills Measurement Survey General booklet Section A. Reading components (10 min) (Part 1. Print vocabulary; Part 2. Sentence processing; and Part 3. Passage Comprehension Section B. 8 Core literacy items (7 min) Fail core Pass core Exercise booklets 1, 2, 3 or 4 [18 Literacy items (28 min), random assignment] End of Module Source: Pierre et al. 2014. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Skills Toward Employment and Productivity Data 17 Table 3.2 Levels of Reading Proficiency Literacy Below Level 1 (0–175) The tasks at this level require the respondent to read brief texts on familiar topics in order to locate a single piece of specific information. Only a basic knowledge of vocabulary is required to complete these tasks, and the reader is not required to understand the structure of sentences or paragraphs or make use of other text features. There is seldom any competing information in the text, and the requested information is identical in form to information in the question or directive. Although the texts can be continuous, the information can be located as if the text were noncontinuous. Tasks below level 1 do not make use of any features specific to digital texts. Literacy Level 1 (176–225) Respondents are required to read relatively short digital or continuous print, noncontinuous print, or mixed texts to locate a single piece of information that is identical to, or synonymous with, the information given in the question or directive. Some tasks may require the respondent to enter personal information into a document, in the case of some noncontinuous texts. Little, if any, competing information is present. Some tasks may require simple processing of more than one piece of information. Tasks at level 1 require knowledge and skill in recognizing basic vocabulary, evaluating the meaning of sentences, and reading paragraph text. Literacy Level 2 (226–275) At this level the relative complexity of the text increases. Texts may be digital or printed, and may comprise continuous, noncontinuous, or mixed types of text. Competing pieces of information may be present. The completion of level 2 tasks requires respondents to match the text to information and may require the ability to paraphrase or make low-level inferences. Tasks require the respondent to • Cycle through or integrate two or more pieces of information against given criteria; • Compare, contrast, and/or reason about information requested in the question; or • Navigate within digital texts to access and identify information from various parts of a document. Literacy Level 3 (276–325) Texts at this level are comparatively dense and lengthy, and include continuous, noncontinuous, mixed, and/or multiple pages. Respondents must demonstrate an understanding of textual and rhetorical structures to successfully complete tasks, especially in the navigation of complex digital texts. Tasks require the respondent to identify, interpret, or evaluate one or more pieces of information, and often require varying levels of inference. Many tasks require the respondent to construct meaning across larger chunks of text, or perform multistep operations in order to identify and formulate responses. Tasks may also require that the respondent disregard irrelevant or inappropriate text to answer the question accurately. Competing information is present, but it is not more prominent than correct information. Literacy Level 4 (326–375) Tasks at this level require respondents to perform multiple-step operations to integrate, interpret, and/or synthesize information from complex or lengthy continuous, noncontinuous, mixed, or multiple-page texts. Complex inferences and the application of background knowledge may be needed to perform successfully. Many tasks require the identification and understanding of one or more specific, noncentral ideas in the text to interpret or evaluate subtle evidentiary claims or persuasive discursive relationships. Conditional information is frequently present in tasks at this level and must be taken into consideration by the respondent. Competing information is present and may be as prominent as correct information. Literacy Level 5 (376–500) At this level, tasks may require the respondent to search for and integrate information across multiple, dense texts; construct syntheses of similar and contrasting ideas or points of view; and/or evaluate evidence-based arguments. Application and evaluation of logical and conceptual models of thinking may be required to accomplish tasks. Evaluating the reliability of evidentiary sources and selecting key information is frequently a requirement to successfully complete tasks. Level 5 tasks require respondents to be aware of subtle, rhetorical cues, to make high-level inferences, and/or to use specialized background knowledge. Source: Adapted from Valerio et al. 2014. Note: Scores are out of a possible 500. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 18 Skills Toward Employment and Productivity Data than those used to measure cognitive skills, in part because of the absence of a consensus on the structure and evolution of personality. STEP builds on the “Big Five” personality traits (openness, conscientiousness, extraversion, agree- ableness, and stability), a widely accepted taxonomy for capturing personality traits. This taxonomy has been found to be replicable across cultures and can be used to capture the evolution of a personality over an individual’s lifetime (John and Srivastava 1999). Measures of grit and hostility bias were also included. The survey of socioemotional skills includes a module aimed at assessing respondents’ time and risk preferences. Job-Specific Skills Job-specific skills are task related and build on a combination of cognitive and noncognitive skills. The STEP survey asks individuals about specific tasks they perform and the skills that they use in their current job. Questions on job-specific skills are intended to capture an individual’s technical skills, reflecting his or her acquired knowledge in particular areas. Because technical skills are often specific to a certain discipline, they are difficult to capture using a survey instrument aimed at the general population. As a consequence, the STEP survey focuses on a range of skills relevant to performance in multiple jobs. Methodology This report uses a combination of descriptive statistics, simple probability models (to estimate what determines the probability of acquiring education and skills and of finding employment), and the Mincer equation (to estimate the returns to education and skills). The descriptive analysis includes simple distributions of the use and intensity of use of skills across genders, age cohorts, different socio- economic statuses, and regions. The estimation is done using the sample weights, and the results are representative for the urban population of Ghana. The returns to education are estimated using a Mincerian wage regression. The total log monthly earnings of the main and second occupation are estimated as a function of education and a number of control variables (as detailed in appendix C). The sample is restricted to those individuals with monthly earnings different from zero, and models are estimated using ordinary least squares (OLS) with robust standard errors. For all skills, dummies of nonresponse (missing variables) have been included in the regression. These models describe associations between earnings and education or skills but do not claim any causal relation. If a person with a college degree earns more than a person with lower education it does not necessarily mean that the college education is the cause of the difference in pay. Rather, the person who went to college might have some characteristics that make him or her more productive in the labor market, thus resulting in higher earnings. It is possible, for example, that high-ability people self-select themselves into college. Thus, higher productivity might be the result of this selection rather than the consequence of the education level achieved. Therefore, it is uncertain whether the education premium reflects Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Skills Toward Employment and Productivity Data 19 the returns to education or a higher market value of unobserved skills. In an attempt to alleviate this selection problem, the regression model included a broad set of controls. Nevertheless, there are many other observable and unobservable factors that might affect both the probability of being employed and of working in specific sectors and occupations and the returns to education. Note 1. The scores for Reading Components will depend on whether there are sufficient data to report three subscales or just a single reading components score. References Banerji, Arup, Wendy Cunningham, Ariel Fiszbein, Elizabeth King, Harry Patrinos, David Robalino, and Jee-Peng Tan. 2010. Stepping Up Skills for More Jobs and Higher Productivity. Washington, DC: World Bank. Carroll, J. B. 1993. Human Cognitive Abilities: A Survey of Factor-Analytic Studies. New York: Cambridge University Press. Cattell, R. B.1971. Abilities: Their Structure, Growth, and Action. Boston, MA: Houghton, Mifflin. Horn, J. L., and R. B. Cattell. 1967. “Age Differences in Fluid and Crystallized Intelligence.” Acta Psychologica 26: 107–29. John, O. P., and S. Srivastava. 1999. “The Big Five Trait Taxonomy: History, Measurement and Theoretical Perspectives.” In Handbook of Personality: Theory and Research, edited by L. A. Pervin and O. P. John. New York: Guilford Press. Neisser, U., G. Boodoo, T. J. Bouchard, Jr., A. W. Boykin, N. Brody, S. J. Ceci, D. F. Halpern, J. C. Lochlin, R. Perloff, R. J. Sternberg, and S. Urbina. 1996. “Intelligence: Knowns and Unknowns.” American Psychologist 51 (2): 77–101. Pierre, Gaëlle, Maria Laura Sanchez Puerta, Alexandria Valerio, and Tania Rajadel. 2014. STEP Skills Measurement Surveys: Innovative Tools for Assessing Skills. Social Protection and Labor Discussion Paper 1421. Washington, DC: World Bank. Valerio, Alexandria, Maria Laura Sanchez Puerta, Gaelle Pierre, Tania Rajadel, and Sebastian Monroy Taborda. 2014. STEP Skills Measurement Program: Snapshot 2014. Washington, DC: World Bank. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Chapter 4 Education Issues in Ghana Foundational Skills: Early Childhood Education Early childhood is a critical stage for long-lasting skills development. Investment in children between the time of birth and the beginning of primary education is a major determinant of their future school performance and labor mar- ket success. Various studies find a strong association between cognitive and socioemotional skills gained at a young age and school achievement, graduation rates, and employment outcomes. Provision of early childhood education (ECE) in Ghana has dramatically increased since 2000. Only 30 percent of Ghanaians between the ages of 45 and 64 attended an ECE program (kindergarten, crèche, day care, and/or nursery school) against 87 percent of the youngest generation ages 15–19 years old. Interestingly, there are no significant differences in ECE participation between males and females. Participation in preschool is positively correlated with household socioeco- nomic status. Seventy-seven percent of adults living in households in the upper socioeconomic brackets participated in an ECE program as compared to 49 ­percent of individuals from the poorest households (figure 4.1). This is con- sistent across all age cohorts. Adults who received ECE are more likely to read and write regularly. Adults who did not attend preschool programs report lower levels of reading and writing regularly (47 and 45 percent, respectively) compared with 82 and 77 percent of those who benefitted from ECE (figure 4.2). Almost all of the population reported using their numeracy skills on a daily basis. Figure 4.3 demonstrates that those who participated in ECE also use their reading, writing, and numeracy with more intensity. Only 27 percent of the adult population of Ghana uses a computer. A third of those who attended ECE programs use one, whereas only one in every ten adults who did not attend ECE uses one. Over time, access to primary and secondary education has substantially increased. In fact, most of the population has completed at least primary edu- cation (80 percent) and junior high school (JHS) (69 percent) (figure 4.4). Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   21   22 Education Issues in Ghana Figure 4.1 Participation in Early Childhood Education in Ghana 100 90 80 Adults who received ECE (%) 70 60 50 40 30 20 10 0 15–19 20–24 25–34 35–44 45–64 Male Female Low Middle High SES SES SES Note: The figure includes people currently attending school. ECE = early childhood e ­ ducation; SES = socioeconomic status. Figure 4.2 Regular Use of Reading, Writing, Numeracy, and Computer Skills in Ghana 100 80 Adults using skill regularly (%) 60 40 20 0 Did not attend ECE Attended ECE Reading Writing Numeracy Computer Note: The regular use of skills is defined on the basis of general questions about daily activities that involve the use of skills at work and outside work. The figure includes people currently attending school. ECE = early childhood education. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Education Issues in Ghana 23 Figure 4.3 Intensity of Skill Use in Ghana a. Reading b. Writing 100 100 80 80 Percent Percent 60 60 40 40 20 20 0 0 No ECE ECE No ECE ECE c. Numeracy d. Computer use 100 100 80 80 Percent 60 Percent 60 40 40 20 20 0 0 No ECE ECE No ECE ECE High Mid Low Not used Note: Intensity of skill use is defined as follows: High = more than 25 pages a week; Medium = 6–25 pages; Low = 1–5 pages requiring the use of skills at work and outside work. The figure includes people currently attending school. ECE = early childhood education. Figure 4.4 Educational Attainment in Ghana 100 Adults with education level completed (%) 80 60 40 20 0 None ECE Primary JHS SHS Tertiary Note: Attainment is the ratio of the population who complete a specific level of education to the population who could have finished this level of education (see box 4.1). The figure includes people currently attending school. ECE = early childhood education; JHS = junior high school; SHS = senior high school. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 24 Education Issues in Ghana Figure 4.5 Educational Level in Ghana, by Age Groups 50 Maximum level of education (%) 40 30 20 10 0 45–64 35–44 25–34 20–24 None JHS SHS Primary Tertiary Note: The figure excludes those who are currently attending school. JHS = junior high school; SHS = senior high school. In general, younger generations have more years of schooling than their elders. Whereas adults in the 45–64 age cohort have an average of 7.4 years of educa- tion, adults ages 20–24 have 2.6 more years of schooling. About 14 percent of adults ages 20–24 report not having completed primary education, but the equivalent figure was 31 ­ percent among the oldest generation (ages 45–65) ­(figure 4.5). Also the proportion of adults who have completed senior high school (SHS) has increased substantially over time. About 37 percent of adults ages 20–24 have completed SHS against only 13 percent of the oldest generation. Tertiary education is also increasing but at a slower rate. In three generations the percentage of people with a tertiary education increased by only two ­ percentage points, from 11 percent for the 45–64 age group to 13 percent among the 25–34 age group (figure 4.5). About 29 percent of the 20–24 age cohort is still attending tertiary education, and 5 percent have already graduated (see box 4.1 for a discussion of educational attainment and education levels). Regional Disparities in Access to Primary Education Although primary education attainment among younger adults has increased in all regions, the increase, as measured in the Skills Toward Employment and Productivity (STEP) survey, was more significant in the Northern, Brong-Ahafo, and Volta regions. Nevertheless, it is worth remembering that, although similar trends are likely to have occurred in the Upper West and Upper East, the STEP Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Education Issues in Ghana 25 Box 4.1  Definitions of Education Variables The education variables include five subgroups of respondents: (i) those with no formal educa- tion or incomplete primary (henceforward called the “No education” or “None” subgroup); (ii) those with complete primary education; (iii) those who completed junior high school (JHS); (iv) those who completed senior high school (SHS); and (v) those who completed tertiary education. We define two variables for education: educational attainment and educational level. Educational attainment identifies the percentage of the population that has completed a spe- cific education level whereas education level identifies the maximum level of education achieved. For example, someone whose education level is tertiary has attained primary, sec- ondary, high school, and tertiary education. Thus, tertiary attainment is the ratio of those with tertiary-level education to all those who achieved (or are currently attending) a lower level of education. Notably, educational attainment is age independent. Indeed, at the denominator, it takes into account only those who, given their age, might have achieved the same level of education. Thus, for example, tertiary education attainment is the ratio of individuals at least 21 years of age having completed tertiary education to the same age-specific group with lower educational attainment. The age references for each educational level are defined as following: 11 years old and above for primary level, 14 years old and above for JHS, 17 years old and above for SHS, and 21 years old and above for tertiary level. Educational attainment is therefore useful for understanding the coverage of a specific education level within a population. For instance, if 80 percent of the population in a country has attained primary education, this includes those who have also achieved higher levels of education. In contrast, when analyzing educational level, the focus is on the highest level com- pleted by the population. Educational level is useful for analyzing how the overall years of education completed by individuals might be correlated with their labor force participation, health, and life outcomes. survey’s small sample of urban households made it impossible to measure them. In the Northern region the proportion of people who attained (or currently attend) primary education increased from 22 percent among adults ages 45–65 to 100 percent among the youngest population. Similarly, in Brong-Ahafo only 42 percent of the 45–64 cohort attained primary education compared to 96 ­percent of the youngest cohort. The Central region is the only region that is lagging behind, with only 83 percent of those 15–19 years old having attended primary education (figure 4.6). Socioeconomic Disparities As might be expected, educational attainment increases with household wealth. More than 36 percent of individuals from the lowest quintile have not com- pleted formal education compared with only about 19 percent of individuals from the highest income quintile (figure 4.7). Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 26 Education Issues in Ghana Figure 4.6 Regional Disparities in Changes in the Primary Completion Rate Population completing primary education (%) 100 90 80 70 60 50 40 30 20 10 0 Northern Brong- Volta Central Eastern Ashanti Western Accra Ahafo 15–19 year-olds 45–64 year-olds Note: The figure includes those who are currently attending school. Upper West and Upper East not reported because of their small sample size. Figure 4.7 Education Level Attained in Ghana, by Socioeconomic Status 100 80 Urban adults (%) 60 40 20 0 Low SES Middle SES High SES Tertiary SHS JHS Primary None Note: The figure excludes those who are currently attending school and uses a subjective measure of socioeconomic status (SES) at age 15. JHS = junior high school; SHS = senior high school. Nevertheless, Ghana is far from universal attainment in basic education, even among the urban population, and the education disparity is increasing. The rich- est segment of the population has been driving most of the improvement in education achievement—especially at the secondary level. Educational level has increased more quickly among the wealthiest, thus increasing the education gap between the poor and the wealthy (figure 4.8). Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Education Issues in Ghana 27 Figure 4.8 Evolution of Educational Profile in Ghana, by Socioeconomic Status 120 100 80 Urban adults (%) 60 40 20 0 4 4 4 4 4 4 4 4 –2 –3 –4 –6 –2 –3 –4 –6 20 25 35 45 20 25 35 45 Bottom 20 percent Top 20 percent Tertiary SHS JHS Primary None Note: The figure excludes those who are currently attending school and uses a subjective measure of socioeconomic status at age 15. JHS = junior high school; SHS = senior high school. Constraints for Education: Late Entry, Completion Delay, and Dropouts Delaying school entry correlates to educational achievement. The average age for starting school for all Ghanaians is 7 years even though the official entry age is 6 years old. On average, those who completed a higher level of education began school around the official age of entry. About 51 percent of those who never complete primary education enroll in school at age 8, whereas 89 percent of those with a post-secondary education enrolled at the official entry age. Figure 4.9 shows the average age of school entry and completion for each level of education achieved compared with the official ages for entry and graduation. Although education attainment is increasing, delays in completion remain prevalent. Primary school should be completed by age 11, JHS by age 14, and SHS by age 17. However, those whose maximum level of education is primary graduate at an average age of 16.4. Considering their late entry, it takes them almost nine years to complete this level, four years more than expected figure 4.9). For JHS and SHS, the difference between the actual and expected (­ ages of graduation decreases, with the difference being about three years for both JHS and SHS. However, students are graduating from SHS at the age at which they should be finishing tertiary education instead. Tertiary education is the level with the highest delay in completion: although the expected age of completion is 21, the graduation age is on average almost six years later. Almost one out of four adults dropped out of school. The dropout rate is high- est during primary education and JHS: among dropouts, 44 percent left school before completing primary education, 37 percent while attending JHS, 16 ­ percent while attending SHS, and 4 percent during tertiary education. Among those starting primary education, 82 continue with a higher level of education; for 8 percent, primary education is their highest level of education, and 10 percent dropped out of primary education before completing it (figure 4.10, panel a). Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 28 Education Issues in Ghana Figure 4.9  Delay in the Official Age of School Entry and Graduation Tertiary Tertiary SHS SHS JHS JHS Age of entry Primary Primary None 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 Age of entry Age of completion Note: The figure excludes those who are currently attending school. The dash lines indicate the expected age of completion for each education level. JHS = junior high school; SHS = senior high school. Figure 4.10  Dropped Out of Highest Educational Level Started, by Maximum Level of Education Completed and Socioeconomic Status a. Dropouts by maximum level of b. Dropouts by socioeconomic status education completed 100 40 80 30 Dropouts (%) Adults (%) 60 20 40 10 20 0 0 Primary JHS SHS Tertiary Low SES Middle SES High SES Dropout Continued with education Highest level completed Note: The figure excludes those who are currently attending school. JHS = junior high school; SES = socioeconomic status; SHS = senior high school. Among those starting JHS, 10 percent dropped out before completing this level; 9 percent of SHS students dropped out before completing SHS. Most of the dropouts belonged to households with low socioeconomic status when they were 15 years old. Whereas the dropout rate for individuals from the richest households is about 26 percent, this rate increases to 38 percent among the poor- est households (figure 4.10, panel b). Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Education Issues in Ghana 29 Figure 4.11 Main Reason for Dropping Out Money High opportunity cost Low SES Others Poor grades Unsafe Money Middle SES High opportunity cost Others Poor grades Unsafe Money High opportunity cost High SES Others Poor grades Unsafe 0 20 40 60 80 Percent citing this reason for dropping out Note: The figure excludes those who are currently attending school and uses a subjective measure of socioeconomic status (SES) at age 15. The lack of money for out-of-pocket expenses is the top reason for dropping out at age 15 for all households, regardless of their socioeconomic status. Among the poorest households, which suffered the highest dropout rate, 66 percent mentioned lack of money for fees, uniforms, or school materials as a reason for dropping out (figure 4.11). High opportunity costs were the second most fre- quently mentioned reason. These high opportunity costs stemmed from the fact that students with lower qualifications might (relatively) easily find self-­ employment in the informal sector. The Gender Gap in Education The gender gap in education attendance is significant for all levels of education in Ghana (figure 4.12). For example, at 26 percent, the proportion of female students completing SHS education is significantly lower than the completion rate of male SHS students, at 47 percent. Moreover, in primary education, 74 percent of female students enrolling in primary education complete a full course of primary education, compared to 88 percent for their male counterparts. The gender gap in the primary completion rate is still significant. Females make up a larger proportion of adults with no formal education. About 13 ­ percent of men and 26 percent of women have no or an incomplete primary education. Nevertheless, women’s educational attainment at both the primary and second- ary level is increasing over time (figures 4.13 and 4.14). Within the oldest generation, Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 30 Education Issues in Ghana Figure 4.12 Educational Attainment, by Gender 100 Adults with education level completed (%) 80 60 40 20 0 Male Female None Primary JHS SHS Tertiary Note: Includes people currently attending school. For the definitions of the variables, see box 4.1. JHS = junior high school; SHS = senior high school. Figure 4.13 Educational Composition: Is the Gender Gap Narrowing? 100 80 Urban adults (%) 60 40 20 0 Male Female Male Female Male Female Male Female 20–24 25–34 35–44 45–64 Tertiary SHS JHS Primary None Notes: The figure excludes those who are currently attending school and uses a subjective measure of socioeconomic status at age 15. JHS = junior high school; SHS = senior high school. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Education Issues in Ghana 31 Figure 4.14 The Gender Gap in Educational Attainment (Male–Female), by Age Group 30 20 Difference in percentage points 10 0 –10 –20 –30 15–19 20–24 25–34 35–44 45–64 None Primary JHS SHS Tertiary Note: The figure excludes those who are currently attending school. 41 percent of women have no formal education (compared with 18 percent of men), but this proportion is 23 percentage points lower among the youngest cohort. Similarly, only 8 percent of women ages 45–64 have SHS ­ education against an SHS completion rate of 35 percent among the youngest. In sum, the STEP survey data show a clear trend of increases in education attainment across generations. Although driven mostly by deprived regions catching up with the rest of the country, these increases were also led by the highest-income groups. Despite improvements, inequality persists because lower- income groups continue to have high opportunity costs. Reference World Bank. 2014. “STEP Skills Measurement Program.” World Bank, Washington, DC. http://microdata.worldbank.org/index.php/catalog/step/about. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Chapter 5 Labor Market Participation Introduction Following the analysis of education attainment based on the Skills Toward Employment and Productivity (STEP) survey, we analyze the survey’s data on labor force participation and employment. The labor force participation rate in Ghana is 67 percent, with 62 percent reported to be working and only 5 percent to be unemployed (figure 5.1, panel a). About 23 percent of the population are inactive either because they are enrolled in school1 (22 percent) or because they are retired (1 percent). The remaining 10 percent are “Not in Education, Employment, or Training” (NEET) (see box 5.1 for definitions of various labor market terms). The most common reasons that respondents gave for being NEET were being a housewife, being unfit to work, or having been discouraged from working (figure 5.1, panel b). Unemployment rates are generally low, and inactivity varies very little across levels of education achieved. Unemployment is up to 6 percent for those with tertiary education, and the rate of NEET varies between 11 percent for people with no formal education and those with senior high school (SHS) education and 6 percent among tertiary graduates (figure 5.2). The transition from school to work is very slow. Individuals ages 35–44 have the highest employed rate. Whereas unemployment and NEET rates vary little across age groups, inactivity declines significantly (by about 28 percent- age points) between the age 20–24 and 25–34 cohorts and falls by 7 additional percentage points for those ages 35–44 (figure 5.3). Although this trend might be explained by a change over time of the labor market structure, it also shows (together with the prevalence of late school entry and late completion) a slow transition from school to work. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   33   34 Labor Market Participation Figure 5.1 Labor Force Participation and Nonparticipation in Ghana percent a. Labor force b. Reason for being NEET Other Inactive, 23 Looking for work In transition Pregnancy/ nursing Discouraged NEET, Unfit 10 Housewife Employed, 62 0 10 20 30 40 50 Unemployed, 5 Reasons for NEET (%) Note: NEET = Not in Employment, Education, or Training. Box 5.1  Definitions of Labor Market Terms Used in the Skills Toward Employment and Productivity (STEP) Survey Ghana’s labor force includes people ages 15–64, both employed and unemployed (and those seeking a job). The employed labor force includes those working for pay, those who have found a job but are not working because they are sick or waiting to start their new job, those working for profit, and those in apprenticeships. The unemployed labor force includes those who are looking for job and who have been laid off for 30 days or less. The employment/unemploy- ment rate can be computed as share of employed/unemployed people over the whole popu- lation aged 15–64 or over the labor force (which excludes the inactive population). The inactive population includes people enrolled in school and those who are “Not in Education, Employment, or Training,” or NEET. The NEET rate is defined as a share of NEET over the 15- to 64-year-old population, including both the active and inactive population. The inactive subgroup is ultimately broken down to distinguish between married and unmarried people. ­ The underemployed are those working less than 40 hours a week (considering the primary and secondary occupations together) and willing to work more hours. Underemployment is defined both as a share of the active labor force and as a share of the employed labor force only. A further distinction is made between those employed in the formal sector and those work- ing in the informal sector. The STEP survey defines informal sector workers as unpaid workers (mainly employed in family businesses), the self-employed and wageworkers who report not having any social security or benefits, or informal wageworkers. The rest of the employed popu- lation constitutes the formal sector, made up predominantly of wageworkers. box continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Labor Market Participation 35 Box 5.1  Definitions of Labor Market Terms Used in the Skills Toward Employment and Productivity (STEP) Survey (continued) Occupations are classified into three categories on the basis of the level of skills required. Low-skilled occupations include agricultural, forestry, and fishery workers; craft and related trades workers; plant and machine operators and assemblers; and elementary occupations. Mid-skilled occupations include technicians and associate professionals; clerical support workers; and service and sales workers. High-skilled occupations include managers and professionals. Finally, the economic sectors are classified in four groups: (i) agriculture, fisheries, and mining; (ii) manufacturing; (iii) low-to-mid-value-added services; and (iv) high-value-added services. The low-to-mid-value-added services sector includes electricity, gas, steam, and air condition- ing supply; water supply and sewage; waste management; construction; the wholesale and retail trade; transportation and storage; accommodation and food services; information and communication; arts, entertainment, and recreation; other services activities; and the activities of households and extraterritorial organizations and bodies. The high-value-added services sector includes financial and insurance activities; real estate activities; professional, scientific, and technical activities; administrative and support service activities; public administration and social security; and education, human health, and social work activities. Figure 5.2 Labor Status, by Educational Level 100 80 Urban adults (%) 60 40 20 0 None Primary JHS SHS Tertiary Inactive at school NEET Unemployed Employed Note: The figure includes people currently studying but not inactive retired people. JHS = junior high school; NEET = Not in Education, Employment, or Training; SHS = senior high school. Wage Employment, Self-Employment, and Formality Most of the employed labor force works in the informal sector, either as self- employed (66 percent) or as informal wageworkers (20 percent).2 In fact, only 15 percent of the employed population works as salaried workers in the formal sector. However, this trend is changing: younger generations are less likely to be self-employed and more likely to be salaried workers, with either formal or Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 36 Labor Market Participation Figure 5.3 The (Slow) Transition from School to Work: Labor Status, by Age Group 100 90 80 70 60 Percent 50 40 30 20 10 0 20–24 25–34 35–44 45–64 Inactive NEET Unemployed Employed Note: The figure includes people who are currently studying. NEET = Not in Education, Employment, or Training. Figure 5.4 Employment Status, by Age Group 100 Maximum level of education (%) 80 60 40 20 0 45–64 35–44 25–34 20–24 Age group Self-employed Formal employee Informal employee Note: The figure excludes people currently studying. informal firms. Nevertheless, still more than 95 percent of the youngest age cohort (20- to 24-year-olds) work in the informal sector, most of them as infor- mal wageworkers rather than self-employed (55 percent) (figure 5.4). Education is strongly associated with employment status, as shown in figure 5.5. About 37 percent of those with a tertiary education work in the informal sector (20 percent as self-employed and 17 percent as informal employees), and this Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Labor Market Participation 37 Figure 5.5 Employment Status, by Education Level 100 80 Urban adults (%) 60 40 20 0 None Primary JHS SHS Tertiary Self-employed Informal employee Formal employee Note: The figure excludes people who are currently studying. JHS = junior high school; SHS = senior high school. share increases to 99 percent for those who do not have any formal education (86 percent as self-employed and 13 percent as informal employees). Occupation Type Informality is synonymous with low-skilled occupations. Only 13 percent of the employed population works in high-skilled occupations, mainly in the formal sector (figure 5.6). Conversely, the majority is employed in low-skilled (39 percent) and mid-skilled occupations (48 percent). The prevalence of workers employed in low- or mid-skilled occupations is relatively higher among the self-employed than among informal wageworkers. Increasing education leads to better occupations. Nevertheless, although this is true for SHS and/or tertiary graduates, those with primary education and those with no formal education are equally distributed among low-skilled or mid-skilled jobs (figure 5.7). Junior high school (JHS) graduates do slightly better, with about 2 percent more working in mid-skilled occupations and about 4 percentage points fewer in low-skilled occupations. The scenario changes for those with an SHS education, with 18 percent working in high-skilled occupations and only 32 percent in low-skilled occupations. The contrast in terms of high-skilled jobs is even more pronounced for those with a tertiary education, for whom the figure reaches 70 percent, with only 6 percent working in low-skilled occupations. Economic Sector of Occupation The largest share of the employed population works in services (61 percent in low-to-mid-value-added services and 17 percent in high-value-added services). The smallest share works in manufacturing (about 10 percent), followed by Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 38 Labor Market Participation Figure 5.6 Type of Occupation, by Employment Status 100 80 Urban adults (%) 60 40 20 0 Formal employee Informal employee Self-employed High-skilled Mid-skilled Low-skilled Note: The figure excludes people currently studying. Figure 5.7 Type of Occupation, by Education Level 120 100 Urban adults (%) 80 60 40 20 0 None Primary JHS SHS Tertiary High-skilled Mid-skilled Low-skilled Note: The figure excludes people currently studying. JHS = junior high school; SHS = senior high school. 11 percent in agriculture, fishery, and mining combined. Almost all of the occu- pations in the manufacturing sector and the agriculture, fishery, and mining sector are low-skilled occupations (figure 5.8). Education attainment increases in line with employment in higher-value eco- nomic sectors. Almost half of the workers in the agriculture, fishery, and mining sector have no formal education (53 percent) compared to 21 percent in the manufacturing sector and 28 percent in the low-to-mid-value-added services sector (figure 5.9). The education level of employees in the manufacturing sector is very similar to that of workers in the low-to-mid-value-added services sector. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Labor Market Participation 39 Figure 5.8 Type of Occupation, by Economic Sector 100 80 Urban adults (%) 60 40 20 0 Agriculture Manufacturing Low-to-mid- High-value- fishing, and value added added mining High-skilled Mid-skilled Low-skilled Note: The figure excludes people currently studying. Figure 5.9 Characterizing Each Economic Sector by the Education Level of Its Labor Force 120 100 80 Urban adults (%) 60 40 20 0 Agriculture Manufacturing Low-to-mid- High-value- fishing, and value added added mining Tertiary SHS JHS Primary None Note: The figure excludes people currently studying. JHS = junior high school; SHS = senior high school. The pool of people with tertiary education is concentrated in the high-value- added services sector (43 percent). Although completing primary education reduces the probability of work- ing in agriculture, the picture for those with a primary education and those with a JHS education is again very similar (figure 5.10). Conversely, Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 40 Labor Market Participation Figure 5.10 How People with Different Levels of Education Are Distributed across Economic Sectors 100 Urban adults (%) 80 60 40 20 0 None Primary JHS SHS Tertiary High-value-added Manufacturing Low-to-mid-value-added Agriculture, fishery, and mining Note: The figure excludes people currently studying. JHS = junior high school; SHS = senior high school. completing SHS makes a difference, given that about 29 percent of SHS graduates work in the high-value-added services sector compared with 3 and 8 percent for those with primary and JHS educations, respectively (­figure 5.10). Nevertheless, 8 percent of SHS graduates still work in the low-to-mid-value-added services sector. ­ Gender Disparities On average, labor force participation is very similar for men and women. However, when distinguishing between married and unmarried men and women, we find a significant gender gap among those who are married. About 89 percent of married men are employed compared with about 77 percent of married women. Married women are more likely to be out of the labor force than men and more likely to be NEET in their capacity as housewives (see figure 5.11). The difference between the labor market status of males and females across age cohorts is not as significant as might be expected (see figure 5.12). Notably, the percentage of females who are NEET is highest for the 20–34 age group, which likely correlates to age-specific life events such as childbearing rather than a change over time in the women’s labor market status. Increasing access to childcare and flexible work options might help to expand labor market participa- tion and opportunities for women. Although women are as likely as men to be employed, the gender gap is very pronounced in the quality of employment. At all ages, females are more likely to be self-employed than men. About half of the male working population is Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Labor Market Participation 41 Figure 5.11 Employment Status, by Marital Status 120 100 80 Percent 60 40 20 0 Male Female Male Female Male Female Single Married All Inactive NEET Unemployed Employed Note: NEET = Not in Education, Employment, or Training. Figure 5.12 Labor Status, by Gender and Age Group 100 80 Urban adults (%) 60 40 20 0 9 4 4 4 4 9 4 4 4 4 –1 –2 –3 –4 –6 –1 –2 –3 –4 –6 15 20 25 35 45 15 20 25 35 45 Male Female Inactive NEET Unemployed Employed Note: The figure includes people currently studying. NEET = Not in Education, Employment, or Training. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 42 Labor Market Participation self-employed whereas the other half is almost evenly distributed between formal and informal employment. In contrast, 79 percent of females are self- ­ employed and only 7 percent are formal sector employees. Notes 1. The inactive category also included a few part-time students (6 percent of the total population) who reported that they were studying and working at the same time. The NEET group also included those who reported being employed as unpaid family workers (3.5 percent of the total population). 2. Unpaid family work is not very common in Ghana, making up only 3.5 percent of the population. In this report, unpaid family workers are classified as NEET. Reference World Bank. 2014. “STEP Skills Measurement Program.” World Bank, Washington, DC. http://microdata.worldbank.org/index.php/catalog/step/about. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Chapter 6 The Use of Cognitive Skills, Job-Specific Skills, and Literacy Introduction To the extent that workers’ productivity depends on their skills and is reflected by their wages, individuals with more skills should expect higher returns from their labor market participation. Most studies investigate the returns to invest- ments in human capital using the education qualification attained. Only a few recent studies examine the returns to skills (Leuven et al. 2004; Tyler 2004). The data arising from the Skills Toward Employment and Productivity (STEP) survey provide a unique opportunity to analyze skills across a range of dimensions and measure their implications for an individual’s well-being and relative opportu- nity in the labor market. As mentioned in chapter 3, the STEP survey focuses on three types of skills: (i) cognitive skills (self-reported and direct assessment); (ii) socioemotional skills; and (iii) job-specific skills. The following section will assess the distribution of each category of skill across population subgroups and thereafter proceed to an analysis of the returns to skills. The Use of Cognitive Skills: Overall Use and Intensity of Use The use of cognitive skills refers to the overall use of reading, writing, and numeracy skills either in daily life or at work.1 Almost everybody uses numeracy skills regularly, at all levels of education. Conversely, the regular use of reading and writing skills is positively correlated with education. The large majority of survey respondents who attended tertiary education read and write regularly. This can be compared with the population with a primary education who reported using their reading (64 percent) and writing skills (60 percent) on a regular basis (figure 6.1). Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   43   44 The Use of Cognitive Skills, Job-Specific Skills, and Literacy The use of reading and writing skills has increased greatly over time, especially among people with low educational attainment. In fact, the highest disparity in the use of skills by age groups is among people who completed only primary education. In the youngest cohort with only primary education, 89 percent use their writing skills and 90 percent use their reading skills. However, less than half of the oldest age cohort uses them, with only 40 percent using their reading skills and 41 percent using their writing skills (figure 6.2). The intensity of the use of various skills is linked to the level of profi- ciency and increase with the level of education attained. Overall, 31 percent Figure 6.1 The Use of Cognitive Skills, by Level of Education Completed 100 90 Adults using skill regulary (%) 80 70 60 50 40 30 20 10 0 None Primary JHS SHS Tertiary Reading Writing Numeracy Computer Note: The figure includes those who are currently in school. JHS = junior high school; SHS = senior high school. Figure 6.2 The Use of Cognitive Skills, by Level of Education Completed and Age a. No education b. Primary 100 100 Those using skill (%) Those using skill (%) 80 80 60 60 40 40 20 20 0 0 45–64 35–44 25–34 20–24 15–19 45–64 35–44 25–34 20–24 15–19 figure continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 45 Figure 6.2  The Use of Cognitive Skills, by Level of Education Completed and Age (continued) c. Junior high school d. Senior high school 100 100 Those using skill (%) Those using skill (%) 80 80 60 60 40 40 20 20 0 0 45–64 35–44 25–34 20–24 15–19 45–64 35–44 25–34 20–24 15–19 e. Tertiary 100 Those using skill (%) 80 60 40 20 0 45–64 35–44 25–34 20–24 15–19 Numeracy Reading Writing Computer Note: The figure excludes those who are currently in school. of the population reported not using their reading skills at all. Among the 69 percent using them regularly, half of the population read with low inten- sity, 24 percent read with medium intensity, and 29 percent read with high intensity. Among those with no formal education, 87 percent do not read at all, and 11 percent read only very short documents, whereas only 1 to 2 percent read intensively (figure 6.3). Reading intensity is higher among graduates of senior high school (SHS) and tertiary education. Nevertheless, there are still 40 percent of SHS and 19 percent of tertiary graduates who do not read or who do so rarely. With regard to writing skills, about 35 percent of the adult population reported not writing regularly. Furthermore, most people use their writing skills less regularly than their reading skills. As in the case of reading skills, the use of writing skills increases with education. Writing skills significantly improve for those completing primary education and for those completing SHS and tertiary education. Nevertheless, even among those who completed tertiary education, almost half write only with low intensity (figure 6.4). The use of skills is strongly related to employment status. Formal employment requires an intensive and frequent use of reading skills. More than 84 percent of Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 46 The Use of Cognitive Skills, Job-Specific Skills, and Literacy Figure 6.3 The Intensity of Use of Reading Skills, by Education Level percent a. No education b. Primary c. Junior high school Medium, 1 Skill Skill High, 16 High, 1 not used, High, 17 not used, Low, 11 37 20 Medium, 16 Medium, 16 Skill not used, Low, 87 Low, 31 46 d. Senior high school e. Tertiary Skill High, 33 not used, High, Low, 5 52 19 Medium, 30 Low, 35 Medium, 26 Note: The figure includes those currently in school. Percentages in some panels do not add up to 100 percent because figures were rounded up. formal employees use their reading and writing skills on a daily basis compared with about 41 percent of informal employees and about 21 percent of the self- employed (figure 6.5). Although the use of numeracy skills is almost universal within all employment groups, computer use is low across the entire workforce. More than one-third of formal wageworkers reported using their computer skills at work, whereas only 11 percent of informal wageworkers and less than 3 percent of the self-employed reported using them (figure 6.5). Finally, the percentage of workers who use reading and writing skills and com- puters is significantly higher among those who work in the high-value-added services sector than in any other sectors (figure 6.6). Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 47 Figure 6.4 The Intensity of Use of Writing Skills, by Education Level percent a. No education b. Primary c. Junior high school High, Skill not Medium, High, High, 8 used, Low, 1 0 10 Medium, Medium, 26 11 8 9 Skill not used, 41 Skill not Low, used, Low, 58 88 41 d. Senior high school e. Tertiary High, Skill not Skill not 10 used, High, used, Medium, 9 20 1 16 Medium, 28 Low, 51 Low, 64 Note: The figure includes those currently in school. Percentages in some panels do not add up to 100 percent because figures were rounded up. Job-Relevant (or Task-Related) Skills Job-relevant skills are task related and consist of a combination of cognitive and socioemotional skills. The STEP survey includes information on whether employed people perform any of the following tasks at work: repairing and maintaining elec- tronic equipment, operating heavy machinery, making presentations, and supervis- ing others. The survey also asked about the intensity of their computer use, solving and learning from problems, physical tasks, and autonomy and repetitiveness. For each skill, a score ranging from 0 to 3 was computed, with 0 being equivalent to not using the skill, 1 for low use, 2 for medium use, and 3 for high use. The data also include the respondents’ self-reported information of the educational attain- ment required to do their job and the time required to learn how to do it. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 48 The Use of Cognitive Skills, Job-Specific Skills, and Literacy Figure 6.5 The Use of Cognitive Skills at Work, by Employment Status 100 90 Adults using skill regularly at work (%) 80 70 60 50 40 30 20 10 0 Formal employee Informal employee Self-employed Reading Writing Numeracy Computer Note: The figure excludes those currently in school. Figure 6.6 The Use of Cognitive Skills at Work, by Economic Sector 100 Adults using skill regularly at work (%) 80 60 40 20 0 Agriculture, fishing, Manufacturing Low-to-mid- High-value-added and mining value-added Reading Writing Numeracy Computer Note: The figure excludes those currently in school. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 49 Working in the formal sector implies regularly learning new things at work and more frequently involves supervising others’ work and making pre- sentations. Conversely, work in the informal sector is less cognitively but more physically demanding and is more repetitive. Self-employed workers reported having more autonomy at work and performing more repetitive tasks than wageworkers ­(figure 6.7). Figure 6.7  Job-Relevant Skills, by Employment Status a. Intensity of activities by employment status 3.0 2.5 2.0 Average score 1.5 1.0 0.5 0 Self-employed Informal employee Formal employee Autonomy and repetitiveness Physical task Solving and learning b. Frequency of people engaging in various types at work by employment status 100 80 Workers using skill (%) 60 40 20 0 Self-employed Informal employee Formal employee Operates heavy machines Supervises others Repairs electronic equipment Makes presentations Drives a car, truck, or three-wheeler Note: Excludes those currently in school. In panel a, the intensity of activities by employment status measured as a score ranging from 0 to 3 was computed, with 0 being equivalent to not using the skill, 1 for low use, 2 for medium use, and 3 for high use. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 50 The Use of Cognitive Skills, Job-Specific Skills, and Literacy Socioemotional Skills The survey includes information on socioemotional skills, more specifically on the “Big Five” personality traits (openness, conscientiousness, extraversion, agree- ableness, and stability), grit, and hostile bias. Hostile bias is a behavioral charac- teristic that measures how often the individual believes that others are hostile to him or her. In addition, the survey also asks about risk and time preferences. Risk lovers have a high score, whereas risk-averse people score low. For time prefer- ence, those with a high score are more future oriented and those with a low score are more present oriented (more detail about the socioemotional skills can be found in appendix B). This self-administered section was presented in English to the full survey sample. Because of low literacy, about 37 percent of respondents were not able to answer. Therefore, the results are not representative of the whole urban popu- lation because those who did not complete this section were not randomly selected and have different characteristics than those able to answer. For exam- ple, nonrespondents are older, less educated, and more likely to work in low- skilled occupations than those able to answer (see table D.5 in appendix D for further details). All questions relating to the measurement of socioemotional skills were answered on the following scale: 1 corresponding to “Almost never,” 2 corre- sponding to “Some of the time,” 3 corresponding to “Most of the time,” and 4 ­corresponding to “Almost always.” As reported in figure 6.8, the respondents Figure 6.8 Average Score of Personality Traits and Grit 4 3 Averge score 2 1 0 Openness Extraversion Girt Conscientiousness Agreeableness Mean of grit Note: The data include only those answering the socioemotional section. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 51 scored around 3 in nearly all the personality traits, which means that most of the time they identify themselves positively with the Big Five and grit. Among them, extraversion is the trait with the lowest average score (2.5) and conscientiousness the one with the highest average score (3.2). On average, respondents located on the middle to low level in the hostile bias scale. With respect to their prefer- ences, they are on average more likely to be present than future oriented and more likely to be risk averse than risk loving (figure 6.9). There is a growing literature underscoring the role of socioemotional skills in education and labor market achievement. Figure 6.10 illustrates the estimated coefficients for socioemotional skills as factors associated with the number of school years completed, controlling for sociodemographic characteristics (age, gender, maternal education, socioeconomic status at age 15, and the region of residency). Similarly, in figure 6.11 the same associations are reported for the estimated probability of attending at least SHS. Finally, figure 6.12 illustrates the association between socioemotional skills and the probability of being employed, being self-employed (versus wageworkers), and employment in mid- to high-skilled occupations (versus low-skilled occupations). Overall, the results suggest that socioemotional skills are closely associated with educational attain- ment (figures 6.10 and 6.11). More specifically, a higher self-reported score relative to the Big Five personality traits is associated with higher educational attainment. Among the personality traits, conscientiousness has the highest positive correlation, followed by extraversion, openness, and agreeableness. Grit and emotional stability are also positively associated but in smaller magnitude. Figure 6.9 Average Scores for Hostile Bias and Time and Risk Preferences 4 3 Averge score 2 1 0 Hostile bias Risk aversion Time preference Note: The data include only those answering the socioemotional section. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 52 The Use of Cognitive Skills, Job-Specific Skills, and Literacy Figure 6.10 Coefficients of Socioemotional Skills on Years of Education, Controlling for Sociodemographic Characteristics Stability 0.3* Agreeableness 0.5*** Extraversion 0.7*** Conscientiousness 1.0*** Openness 0.6*** Grit 0.3* Hostile bias –0.9*** Time preference –0.4*** Risk 0.2** –1.0 –0.9 –0.8 –0.7 –0.6 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Note: The “Big Five” (stability, agreeableness, extraversion, conscientiousness, and openness) are included simultaneously in the regression model. The rest of the skills are included on their own, with controls. Including each of the Big Five personality traits on their own with controls does not influence the results. Each coefficient should be interpreted as an indication of how a change in one unit of the skills score is related to years of education. *** p<0.01, ** p<0.05, * p<0.1. All models are estimated using ordinary least squares (OLS) controlling for maternal education, region, socioeconomic level at age 15, and gender. See table G.1 in appendix G for full results. *** p<0.01, ** p<0.05, * p<0.1. Figure 6.11 Coefficients of Socioemotional Skills on the Probability of Attaining at Least SHS Education, Controlling for Sociodemographic Characteristics Stability 5.3%** Agreeableness 5.2%** Extraversion 8.5%*** Conscientiousness 14.1%*** Openness 7.8%*** Grit 2.9% Hostile bias –11.5%*** Time preference –5.2%*** Risk 3.6%*** –16 –14 –12 –10 –8 –6 –4 –2 0 2 4 6 8 10 12 14 16 Note: The dependent variable is a binary variable that takes the value of one if the highest educational level is SHS or higher, and zero for all other levels of educational attainment. The Big Five (stability, agreeableness, extraversion, conscientiousness, and openness) are all included simultaneously in the same regression model. The rest of the skills are included on their own with controls. Including each of the Big Five personality traits on their own with controls does not influence the results. Each coefficient should be interpreted as an indication of how a change in one unit of the skills score is related to the probability of attaining SHS education or higher. All models are estimated using linear probability model controlling for maternal education, region, socioeconomic level at age 15 and gender. See table G.2 in appendix G for full results. *** p<0.01, ** p<0.05, * p<0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 53 The results are in line with the literature that relates better educational out- comes to those with higher levels of the Big Five personality traits and grit. Furthermore, those who are more trustful, more present oriented, and less risk averse have a higher probability of studying one additional year (figure 6.10). The results are substantially confirmed when analyzing the probability of com- pleting at least SHS education (figure 6.11). In this case, a change in one score unit of conscientiousness is associated with 14 percentage points on the probability of completing SHS or tertiary attainment. As with years of education, higher agree- ableness and stability levels, perceiving others as less hostile, being more present oriented, and being a risk lover increase the probability of completing higher levels of education Finally, we present the results of the association of socioemotional skills and the probability of being employed, being self-employed (versus wageworkers), and working in mid- to high-skill occupations (versus low-skill occupations). Figure 6.12 shows that in most cases the correlation is not statistically significant Figure 6.12 Coefficients of Socioemotional Skills on the Probability of Selected Labor Market Outcomes Stability –5.2%* 7.2%** –4.0%* Agreeableness Extraversion –7.1%** Conscientiousness –8.8%*** Openness Grit Hostile bias 8.8%*** Time preference Risk –4.1%*** –15 –10 –5 0 5 10 15 Prob. employed Prob. self-employed Prob. high-to-medium-skilled occupation Note: The Big Five (stability, agreeableness, extraversion, conscientiousness, and openness) are all included simultaneously in the same regression model. The rest of the skills are included on their own with controls. Including each of the Big Five personality traits on their own with controls does not influence the results. Each coefficient should be interpreted as a measure of how a change in one unit of the score on the skill is related to the probability of being associated with the following labor market outcome: The probability of being employed is relative to those unemployed or inactive. The probability of being self-employed is relative to being a wage earner. The probability of being in high-to-medium-skilled occupations is relative to low-skilled ones. All models are estimated using linear probability model controlling for maternal education, region, socioeconomic level at age 15, gender, and level of education. See tables G.3 through G.5 in appendix G for full results. *** p<0.01, ** p<0.05, * p<0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 54 The Use of Cognitive Skills, Job-Specific Skills, and Literacy (once we controlled for education, one of the control variables included in the model). However, noncognitive skills seem to play a role in determining the qual- ity of the occupation. Individuals with higher levels of the Big Five personality traits and lower levels of hostile bias are less likely to be self-employed (and therefore more likely to be wageworkers). Interestingly, the only socioemotional skill that is significantly associated with having a mid- to high-skill occupation is emotional stability, which relates to the ability to cope with stress. Included in mid- to high-skill occupations are managers, professionals, clerical support workers, and service and sales workers. Reading Literacy Assessment The reading literacy assessment consists of three parts, all administered in English. The first part (Reading Components) evaluates foundation reading skills, including word meaning, sentence processing, and passage comprehension. The second part (Core Literacy Test) consists of a core literacy assessment that is intended to sort the least literate adults from those with higher reading skill levels. The Core Literacy Test has a total of eight items, and respondents with three or more correct responses are regarded as having met a minimum reading literacy threshold (see box 6.1 for more information on the Core Literacy Test). The third part (the Literacy Exercise Booklets) is administered only to those respondents who have passed the Core Literacy Test (see box 6.1). The Literacy Exercise Booklets use a variety of questions drafted around daily life situations. They require respondents to carry out different types of tasks such as accessing and identifying information (in both text-based and nonprose materials such as tables, graphs, and forms), integrating and interpreting information, and evaluat- ing information by assessing the relevance, credibility, or appropriateness of the material for a particular task. Overall reading proficiency scores are reported on a scale ranging between 0 and 500, which is divided into five levels, with level 1 characterized by the least demanding tasks and level 5 the most demanding. For each respondent, 10 plausible values were generated. All survey respondents were asked to answer the Core Literacy Test. Nevertheless, the nonresponse rate is quite high—about 38 percent (figure 6.13). There are two types of nonrespondents: those who fail to complete the section because they are unable to read or understand English (“No English”) and those who did not complete the tests for multiple reasons (“Missing”).2 Those with no English literacy represent 21 percent of the adults, whereas the missing group represents 17 percent of the sample. The three groups (those who answered the Reading Components and the Core Literacy Test, the group with no English literacy, and the missing group) are significantly different; therefore, the findings summarized may not be applicable to the entire urban sample. For example, those who answered are more likely to speak English at work (57 percent versus 18 percent of the missing and those with no English together) and to read and write with higher intensity than those who did not answer the test. Compared to the nonrespondents, they have higher Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 55 Box 6.1  What Does It Mean to Pass the Core Literacy Test? A sizable proportion of adults reported that they read regularly, even though they did not pass the minimum literacy threshold of the Core Literacy Test. One reason could be that the tests are in English whereas many individuals mainly speak a local language. Only 1 percent and 40 percent speak English as their main language at home and at work, respectively (figure B6.1.1). In contrast, Akan is widely used in daily life. Thus, although the literacy assessment captures reading proficiency in English, the self-reported measure could be a better proxy of respon- dents’ capacity to read and write in their daily life in their local language. Even though only 40 percent of adults speak English at work, 58 percent of the urban adults report being able to speak and to read and write English well enough to work in a job that requires them to do so (figure B6.1.2). As expected, the vast majority (90 percent) of those with no English and a third of those who failed the Core Literacy Test are unable to write, read, or speak in English at work. In con- trast, among those who passed the Core Literacy Test, about 96 percent report being able to write, read, and speak English at work. (figure B6.1.3). Notably, about half of those who failed the Core Literacy Test report being able to use English at work—that is, there is a discrepancy between the self-reported and measured skills. Poor English language literacy does not seem to prevent people from getting a job (figure B6.1.4). Only 12 percent report that the lack of English kept them from getting a job. However, those who speak English at work tend to have better-quality jobs. They are more likely to be formal employees in high-value-added sectors and in high-skill occupations. They also earn on average 248 GHS per month (or 60 percent) more than those who do not speak English at work. Figure B6.1.1 Main Language Spoken at Home and Work a. At home b. At work Akan Akan Ewe Ewe Ga-Adangme Ga-Adangme Mole-Dagbani Mole-Dagbani English English Other Other 0 20 40 60 80 100 0 20 40 60 80 100 Percent Percent Note: This figure includes all individuals regardless of whether or not they are attending school as well as those who worked at any point during the previous 12 months. box continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 56 The Use of Cognitive Skills, Job-Specific Skills, and Literacy Box 6.1  What Does It Mean to Pass the Core Literacy Test? (continued) Figure B6.1.2 Self-Reported Ability to Speak and Read/Write in English at Work percent Speak, write/read, 58 Don’t speak or write/read, 32 Don’t speak but do write/read, 1 Speak but don’t write/read, 8 Note: This figure includes individuals regardless of whether or not they are attending school. Proportions may not add up to 100 percent because figures are rounded. Figure B6.1.3 Performance in the Core Literacy Test According to Self-Reported Ability to Speak and Read/Write in English 100 80 60 Percent 40 20 0 Missing No English Failed Passed Don’t speak or read/write Only speak Speak and read/write Note: This figure includes all individuals regardless of whether or not they are attending school. box continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 57 Box 6.1  What Does It Mean to Pass the Core Literacy Test? (continued) Figure B6.1.4  Did the Lack of English Keep You from Getting a Job? percent Don’t know/ haven’t applied, Yes, 11 12 No, 76 Note: The figure includes all individuals regardless of whether or not they are attending school and those who worked at any point during the previous 12 months. Proportions may not add up to 100 percent because figures are rounded. education levels, are on average nine years younger, and are more likely to be students, to be wageworkers rather than self-employed, and to work in high- value occupations. Although one-third of the missing group use English at work, 6 percent of those with no English literacy do. Furthermore, almost three-quarters (72 percent) of those with no English did not complete any formal education, whereas only 18 percent of the missing group has no education and 17 percent completed at least SHS. Also a lower percentage of those with no English reported using their reading and writing skills compared to those in the missing group. Reading Components The Reading Components section evaluates the extent to which participants can recognize the printed forms of common objects (print vocabulary), comprehend sentences of varying levels of complexity (sentence processing), and comprehend the literal meaning of connected text (basic passage comprehension). Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 58 The Use of Cognitive Skills, Job-Specific Skills, and Literacy Figure 6.13  Distribution of Reading Component and Core Literacy Test Respondents percent No English, 21 Answered test, 62 Missing, 17 Note: This figure includes only those who are currently in school. The overall performance is low in all three sections (figure 6.14). Sentence processing has the highest percentage of correct answers (41 percent) at all edu- cation levels, age groups, and levels of self-reported use of skills. However, it is also the section with the highest percentage of nonresponse (19 percent) com- pared to 11 percent in print vocabulary and 9 percent in passage comprehension. In passage comprehension only 38 percent of the items are answered correctly. The lowest performance is in print vocabulary with only 21 percent correct answers. Respondents reporting to be medium- to high-intensity readers or writers answer a higher percentage of correct answers in both the sentence processing and passage comprehension tests. The only exception is for the print vocabulary com- ponent, which does not vary much across groups (figure 6.15, panels a and b). The percentage of correct answers increased with the education level achieved (figure 6.16, panel a), and younger cohorts performed better than older age groups (figure 6.16, panel b). Those with no formal education answered fewer than 24 percent of the questions in each section correctly. Even though the results improve across all levels of education, there is a more noticeable difference between junior high school (JHS) and SHS graduates. Those with a primary or a JHS education answered less than one-third of the test questions correctly, whereas those with an SHS or a tertiary education answered about 44 percent of the pas- sage questions and 46 percent of the sentence questions correctly. However, even among those with tertiary education, the scores on the print vocabulary section are low, with an average of only 23 percent correct answers—only 4 percentage points Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 59 Figure 6.14 Performance on the Reading Component 100 80 60 Answer (%) 40 20 0 Passage Sentence Vocabulary Nonreponse Incorrect Correct Note: This figure excludes the 38 percent of the sample that did not take these tests and includes people currently attending school. Proportions may not add up to 100 percent because figures are rounded. Figure 6.15 Performance According to the Self-Reported Use of Reading and Writing Skills a. Reading b. Writing 80 80 60 60 Correct answers (%) Correct answers (%) 40 40 20 20 0 0 Skill not Low Medium High Skill not Low Medium High used used Passage Sentence Vocabulary Note: This figure excludes the 17 percent of the sample that did not take these tests. Includes people currently attending school. higher than those in primary education. Younger generations are more likely to perform better in the sentence and passage comprehension tests than older age groups (figure 6.16, panel b). This might suggest an improvement of the literacy level over time or might simply reflect a greater familiarity of the younger genera- tions with taking assessment tests. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 60 The Use of Cognitive Skills, Job-Specific Skills, and Literacy Figure 6.16 Performance According to Age Group and Education Level a. Age group (years) b. Education level 80 80 60 60 Correct answers (%) Correct answers (%) 40 40 20 20 0 0 15–19 20–24 25–34 35–44 45–64 None Primary JHS SHS Tertiary Passage Sentence Vocabulary Note: This figure excludes the 17 percent of the sample that did not take these tests and includes people currently attending school. JHS = junior high school; SHS = senior high school. Core Literacy Test The Core Literacy Test contains eight basic literacy questions and is designed to identify those with a minimum level of literacy, namely those who can answer at least three of the eight questions correctly. Only 42 percent of the respondents pass this threshold; 17 percent do not answer because they are unable to read the questions, and 22 percent are missing for other reasons. When taking out the missing group and adding those with no English to the group that did not pass the Core Literacy Test, the data show that only half of the population passed the Core threshold (figure 6.17, panel a). The distribution of correct answers has a U shape: about 15 percent of the sample answered all questions correctly, and 14 percent had no correct answer. The median value is about three to four correct answers (figure 6.17, panel b). This section provides a profile of those who pass and those who fail the test (considering together those who failed and the “No English” subgroup) in terms of gender, age, education level, language, labor and employment status, self-reported use of skills, socioemotional characteristics, and performance on the Reading Components and Core Literacy Tests. Full results are reported in appendix D. Young men with at least an SHS education are most likely to pass the test. More than half of those who pass the exam were male and seven years younger than the average. Ninety-three percent of those who failed the test had a JHS education or lower (42 percent have no formal education and 33 percent attended JHS), whereas about 57 percent of those who passed the test had at least an SHS educa- tion and about 21 percent have tertiary education. Notably, about 81 percent of those who passed the exam had received early childhood education compared with 47 percent of those who failed. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 61 Figure 6.17 Core Literacy Test Performance percent a. Sample distribution b. Correct answer distribution No English, 40 25 Passed, 30 50 Adults (%) 20 10 Failed, 24 0 No 0 1 2 3 4 5 6 7 8 English Total correct answers in core Note: This figure excludes the 17 percent of the sample that did not take these tests and includes people currently attending school. Proportions may not add up to 100 percent because figures are rounded. The variables for socioeconomic status (SES) and whether English was the respondent’s main language are both highly correlated with the probability of pass- ing the test. About 30 percent of those who failed the test have been poor during their adolescence, and only 15 percent of them passed the Core Literacy Test. Although English is Ghana’s official language, only 1 percent of all respon- dents reported speaking English at home and 40 percent at work. Because the reading test was in English, it is not surprising that a higher percentage of those who passed the test than of those who failed reported English as the main lan- guage at home and at work. Among those who passed the test, 73 percent use English at work compared with 17 percent of those failing the test. Furthermore, the probability of succeeding is higher for those working in the formal sector and in high-skilled and better-paid occupations. Within the employed group, 81 percent of those who failed the test are self-employed and only 3 percent are formal wageworkers. More than half work in mid-skilled occu- pations and only 3 percent are in high-skilled jobs. In contrast, 38 percent of those who passed the exam were self-employed and 34 percent were formal wage- workers. One-third worked in high-skilled occupations and earn on average 274 GHS (approximately US$80) more per month than those who failed the test. Self-reported measures of the use and intensity of use of reading and writing skills is highly correlated with the results of the Core Literacy Test. Nevertheless, a significant number of people who reported using their skills regularly failed the test. Indeed, of those who did not pass the test, about 42 and 38 percent, respectively, reported using their reading and writing skills. The (self-reported) intensity of use of reading and writing skills is a better predictor of the test results. Indeed, 41 per- cent of those who pass the test are regular (high-intensive) readers compared to Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 62 The Use of Cognitive Skills, Job-Specific Skills, and Literacy only 12 percent of those who fail the test. Similarly, 83 percent of those who fail the test self-report being low-intensity writers, and the percentage of low-intensity writers is lower (63 percent) among those who did not pass the test. The differences among respondents are more significant when we compare those who successfully pass the Core Literacy Test with those who do not answer any questions correctly. On average the latter have lower education (half of them have not completed any formal education); only 5 percent speak English at work and none of them use it at home. Eighty percent of these respondents are employed and most of them (79 percent) are self-employed. Fewer than one-third of them report using their reading skills (only 9 percent of them are high-intensive readers) or their writing skills (only 2 percent of them are high-intensive writers). Literacy Exercise Booklets Those who succeed in the Core Literacy Test are asked to take the Literacy Exercise Booklets, which provide a more in-depth evaluation of reading skills. The Literacy Booklets are scored on a scale that ranges from 0 to 500 and includes six literacy levels from “below level 1” to level 5. The description of the literacy competencies for each one of the levels is provided in table 3.2. The “below level 1” includes those who failed the Core test (half of the original sample) and an imputed score for those who did not answer (about 17 percent of the eligible sample).3 The average performance on the Literacy Exercise Booklets is very poor. Most of the population (61 percent) is below level 1, 18 percent reach level 1, 17 percent reach level 2, only 4 percent of the respondents reach level 3 or level 4 (reported together in figure 6.18), and no one is in level 5. The level achieved is positively correlated with the number of correct answers in the Core Literacy Test. Those with a reading proficiency below level 1 are those who did not reach the minimum threshold set for the Core Literacy Test (three correct answers) (about 33 percent) and the group with no English literacy (34 per- cent). However, 11 percent of those below level 1 passed the Core test but per- formed poorly in the subsequent assessment. On the other hand, those in level 3 and 4 answered correctly at least six of the Core test questions. Indeed, 93 percent of them correctly answered seven or more questions in the Core test (figure 6.19). Like those passing the Core test, those scoring high in the literacy assessment are younger, have higher education (SHS or tertiary), and are more likely to work in the formal sector, to use their reading and writing skills more intensively, and to use English more frequently at work and at home (figure 6.20). Performance on the Literacy Exercise Booklets is correlated with the self- reported reading intensity as well. For example, among those who passed the Core Literacy Test with three to six correct answers, more of those who self-reported using their reading skills with high intensity reached level 1 than did those who read with low intensity (52 percent versus 43 percent) and level 2 (25 percent versus 14 percent) (figure 6.21). Similarly, among those who completed the whole Core Literacy Test correctly, 27 percent of the high-intensity readers reached levels 3 and 4 on the Literacy Exercise Booklets versus only 7 percent of those who self- reported reading with low intensity. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 63 Figure 6.18 Performance on Literacy Exercise Booklets: Proficiency Levels percent Level 3/4, 4 Level 2, 17 Below level 1, 61 Level 1, 18 Note: The “below level 1” portion includes those who failed the core test (half of the original sample) and an imputed score for those who did not answer (about 17 percent of the eligible sample). Figure 6.19  Distribution of Core Literacy Test Results, by Proficiency Levels Achieved on Literacy Exercise Booklets 100 80 Urban adults (%) 60 40 20 0 Below level 1 Level 1 Level 2 Level 3/4 Illiterate Missing Score 0, 1, 2 Score 3, 4, 5, 6 Score 7, 8 Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 64 The Use of Cognitive Skills, Job-Specific Skills, and Literacy Figure 6.20 Literacy Level According to Gender, Education Level, Age Group, Reading Intensity, Labor Status, and Employment Status a. Gender b. Education c. Age group (years) 300 300 300 Level 3: 276 to 325 points 275 275 275 250 250 250 Reading proficiency score Reading proficiency score Reading proficiency score Level 2: 226 to 275 points Level 2: 226 to 275 points 225 225 225 200 200 Level 1: 176 to 200 Level 1: 176 to 225 points 225 points Level 1: 176 to 225 points 175 175 175 150 150 150 125 125 125 100 100 100 75 75 75 50 50 50 25 25 25 0 0 0 Female Male None Primary JHS SHS Tertiary 15–19 20–24 25–34 35–44 45–64 d. Reading intensity e. Labor status f. Employment status 300 300 300 275 275 275 Reading proficiency score Reading proficiency score Reading proficiency score 250 250 250 Level 2: 226 to 275 points Level 2: 226 to 275 points Level 2: 226 to 275 points 225 225 225 200 Level 1: 176 to 225 points 200 200 Level 1: 176 to 225 points Level 1: 176 to 225 points 175 175 175 150 150 150 125 125 125 100 100 100 75 75 75 50 50 50 25 25 25 0 0 0 Skill Low Medium High Emp- Unemp- NEET Inactive Formal Informal Informal not used loyed loyed at school employees self- employees Note: JHS = junior high school; NEET = Not in Education, Employment, or Training; SHS = senior high school. Figure 6.21  Differences in Proficiency Levels, by Self-Reported Reading Intensity 100 80 Urban adults (%) 60 40 20 0 Low High Low High intensity intensity intensity intensity ETS Score 3–6 ETS Score 7–8 Below level 1 Level 1 Level 2 Level 3/4 Note: ETS = Educational Testing Service. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 65 Developing Skills beyond Education: Training and Apprenticeships In Ghana 32 percent of the adult population living in urban areas continue their skills development by participating in skills training courses, by achiev- ing industry- or government-recognized certificates, and through informal ­ apprenticeships. Apprenticeship is the most common: in the 12 months prior the survey, 25 percent of adults participated in an apprenticeship, whereas only 6 percent received a recognized training certificate and less than 7 percent par- ticipated in a formal training course of at least 30 hours. It is worth noting that, although apprenticeship is more common for older age cohorts (figure 6.22, panel b) and among those with JHS or lower education, training is more com- mon for those with tertiary education (figure 6.22, panel a). After completing each level of formal education, students might continue with the next level of formal education or leave the education system and take an apprenticeship instead. As shown in figure 6.23, apprenticeship does not replace formal education. Indeed, most of people who complete an education level either decide to drop out of school or continue in formal education. Among those completing primary education, only 4 percent left the education system for an apprenticeship, 65 percent start JHS, and 10 percent drop out. The percent of those opting for an apprenticeship is the highest among those completing JHS: 13 percent of them choose an apprenticeship while 30 percent continue to SHS. Formal workers generally have more training outside the formal education system, whereas apprenticeship is more common among the self-employed. Twenty-six percent of formal workers participated in work-related or personal- related skill training courses, 25 percent participated in apprenticeships, and 14 percent achieved an industry- or government-recognized certificate. About 38 percent of the self-employed participate in an apprenticeship (figure 6.24). Apprenticeships are also more common among those working in low-skilled occupations and in the manufacturing sector (figure 6.25). The most common Figure 6.22 Certificate, Training, and Apprenticeship, by Education Level and Age Group a. Education level b. Age group (years) 60 60 Urban adults (%) Urban adults (%) 40 40 20 20 0 0 None Primary Junior Senior Tertiary 15–19 20–24 25–34 35–44 45–64 education high school high school education Certificate Training courses Apprenticeship Note: The figure excludes those who are currently at school. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 66 The Use of Cognitive Skills, Job-Specific Skills, and Literacy Figure 6.23 Are Students Choosing Apprenticeship Instead of Formal Education? 100 80 60 Percent 40 20 0 None Primary Junior Senior Tertiary education high school high school + Higher education level No more + Apprenticeship Note: The figure excludes those who are currently at school. The subgroup “+ Higher education level” includes those who, after completing a certain education level, continue with the next level of education. The subgroup “+ Apprenticeship” includes those who complete a certain education level, drop out, and opt for an apprenticeship. The subgroup “No more” includes those who, after completing a certain education level, left school without completing another level of education. Figure 6.24 Certificate, Training, and Apprenticeship, by Employment Status 80 60 Urban adults (%) 40 20 0 Formal employee Informal employee Self-employed Certi cate Training courses Apprenticeship Note: The figure excludes those who are currently at school. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Use of Cognitive Skills, Job-Specific Skills, and Literacy 67 Figure 6.25 Certificate, Training, and Apprenticeship, by Occupation and Economic Sector a. Occupation b. Economic sector 80 80 60 60 Urban adults (%) Urban adults (%) 40 40 20 20 0 0 Low-skilled Mid-skilled High-skilled Agriculture, Manufac- Low-to- High- fishing, and turing mid-value- value- mining added added Certificate Training courses Apprenticeship Note: The figure excludes those who are currently at school. field of apprenticeship is personal service, retail, and related (65 percent), where the category barber, hairdresser, hair stylist, and cosmetologist accounted for close to 40 percent of the apprenticeships. Notes 1. See appendix B for a definition of skills use and intensity of use variables. 2. This group includes for example those refusing to begin the test booklet (because of time constraints, not wanting to bother, or other general refusal), those who began the booklet but refused to continue, or those who were unable to continue because of an unusual circumstance. 3. The literacy proficiency is modeled using the item response theory in a scale that ranges from 0 to 500. Each point in the 500-point scale has a probability of completing the tested item. The higher the score for a given individual, the more proficient the individual is. The design divides the test into partially linked booklets so that only a portion of the entire battery of items is administered to a single individual. Otherwise, the assessment would be too long if every individual had to take the entire battery of items. In order to estimate the proficiency level, multiple literacy proficiency scores (plausible values) are estimated for each individual on the basis of response to the items as well as some background information. Only those who passed the Core had to respond to this booklet, providing a finer evaluation of reading skills for the most literate respondents. However, imputation allows estimating the reading proficiency level for all respondents. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 68 The Use of Cognitive Skills, Job-Specific Skills, and Literacy References Leuven, E., H. Oosterbeek, and H. van Ophem. 2004. “Explaining International Differences in Male Skill Wage Differentials by Differences in Demand and Supply of Skill.” The Economic Journal 114 (495): 466–86. Tyler, John H. 2004. “Basic Skills and the Earnings of Dropouts.” Economic of Education Review 23 (3): 221–35. World Bank. 2014. “STEP Skills Measurement Program.” World Bank, Washington, DC. http://microdata.worldbank.org/index.php/catalog/step/about. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Chapter 7 The Returns to Education and Skills: Building the Job-Relevant Skills That Employers Demand Introduction This section discusses the payoffs of education and skills in term of earnings and job opportunities. This section answers the following questions in detail: To what extent does education explain better labor market opportunities and higher earn- ings? (And at which educational level does further investment in education not generate a premium?) Do skilled workers have higher earnings? Which skills yield the best labor market opportunities and higher earnings? The analysis carried out in this section involves descriptive statistics and two multivariate regression models. The first model is a Mincer equation for earnings. The second is an employment participation model. These models look at the correlation between education and/or skills and monthly earnings and employment status, respectively. For the two models we tried different specifications. The first specification reported in the figures and tables below does not include any controls and reports only the “pure” correlation between monthly earnings (or self-employment status) and the education variables (either years of education or education levels where those with no formal edu- cation or incomplete primary are used as a reference category). In the second model, we control for a core set of sociodemographic variables including age, gender, work experience, economic sectors, and regional dummies. Then we include a set of controls one at time to see how education returns vary when taking into account (i) the results of the proficiency assessment test (the num- ber of correct answers in the Core Literacy Test and the percentage of correct answers on the Reading Components); (ii) individual personality traits; (iii) the intensity of the use of cognitive skills; (iv) the intensity of the use of job-specific skills (cognitive challenges, physical work, autonomy and repetitiveness, Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   69   70 The Returns to Education and Skills: Building the Job-Relevant Skills That Employers Demand presentation skills, supervision skills, and the use of computer at work); and (v) all of the controls together. All control variables are defined in appendix C and full results are reported in appendixes D and E.1 Returns to Education and Skills Our data show some returns for continued education. An additional year of edu- cation increases monthly earnings by 3–7 percent (figure 7.1). Notably, primary education has no distinguishable premium in terms of earnings, considering those with no formal education or incomplete primary education as a base category. We also found that market rewards for primary education and junior high school (JHS) are lower than those for senior high school (SHS) and higher levels of education.2 For higher education, the premium is between 16 and 63 percent for SHS graduates and between 77 and 156 percent for tertiary graduates compared to workers with no formal education or incomplete primary education. The pre- mium for JHS is much lower (between 10 and 40 percent). Faced with uncertain and binding constraints, people are more likely to drop out of school before SHS. In fact, as mentioned above, the majority of the population (66 percent) leaves the education system before SHS. Figure 7.1 Returns to Education 160 156*** 156*** 8 7*** 7*** 140 141*** 7 6*** 131*** Return of years of education (%) Return of level of education (%) 120 6*** 116*** 6 111*** 100 5*** 5 5*** 80 77*** 4 63*** 63*** 3*** 60 3 50*** 43*** 43*** 45*** 40 39*** 40*** 36*** 38*** 2 28** 26*** 17 17 21**16 20 17 15 1 12** 8 15 0 0 ls ills lls ic t y ls en lit ro tro om ki sk na m nt cs on on of ss o co ifi rs lc se ec e ec No Pe Us Al as cio sp y b- So ac Jo er Lit Primary Junior high school Senior high school Tertiary Years of education Note: All models are estimated using ordinary least squares. The estimated coefficients for each education level (bars) and for the years of education (dotted line) are reported in the figure. ***p < .01, **p < .05, *p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Returns to Education and Skills: Building the Job-Relevant Skills That Employers Demand 71 Overall skills are an important determinant of earnings beyond education. In fact the returns of any additional year of education decrease by 2 percentage points after all skills have been controlled for. The best-paid jobs are those that require intensive writing and job-specific skills such as the use of the computer and also involve cognitive challenges and supervision responsibilities. Out of the “Big Five” personality traits (openness, conscientiousness, extraversion, agreeable- ness, and stability), only the estimated coefficient for conscientiousness is statisti- cally significant (see appendix E). Notably, even after accounting for all skills (cognitive, socioemotional, and job-specific skills), SHS and tertiary education have a significantly higher earnings premium. The returns to education and skills vary a great deal according to types of employment, with the lowest returns among self-employed and the highest among formal wageworkers. One additional year of education increases the monthly earnings of the self-employed by 1–5 percent, those of informal wage- workers by 4–6 percent, and those of formal wageworkers by 7–10 percent (see figure 7.2 and appendix E). Earnings differentials are a typical feature of labor market segmentation. More specifically, an earnings gap between informal sector workers and equally quali- fied formal wage and salaried employees can be interpreted as a measure of the degree of labor market segmentation (Schultz 1961; Becker 1962; and Mincer 1962). In countries such as Ghana, the formal economy is not capable of Figure 7.2 Returns to Education, by Type of Employment a. The self-employed 160 154*** 154*** 10 142*** 9 140 Return of years of education (%) 127*** 8 Return of level of education (%) 120 113*** 7 100 6 5*** 5*** 80 5*** 5 68** 60*** 4*** 4 60 60*** 3*** 53*** 3*** 45 3 40 38*** 38*** 37*** 38*** 34** 33** 25** 30** 1 2 23* 20 13 13 15 17 8 11 11 1 4 7 0 0 ls ills lls ic t ity ls en ro tro om ki l sk na sm nt cs on on of so co ifi es lc ec e r ec ss No Pe Us Al cio sp ya b- So ac Jo er Lit figure continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 72 The Returns to Education and Skills: Building the Job-Relevant Skills That Employers Demand Figure 7.2  Returns to Education (continued) b. Informal wageworkers 160 10 140 9 Return of years of education (%) 132*** Return of level of education (%) 8 120 119*** 119*** 112*** 7*** 106*** 7 100 6*** 6*** 5*** 90*** 90*** 6 80 5*** 5 75*** 4*** 64*** 64*** 4** 60 58*** 4 53** 53*** 43 48** 39*** 40*** 3 40 39*** 32* 34** 33 29 29 29 27* 2 24 21 20 12 1 0 0 ls ills lls ic t y ls en lit ro tro om ki sk na sm nt cs on on of so co ifi es lc ec e r ec ss No Pe Us Al cio sp ya b- So ac Jo er Lit c. Formal sector wageworkers 140 125*** 12 120 Return of years of education (%) 114*** 114*** 10*** 10 Return of level of education (%) 9*** 9*** 9*** 9*** 96** 9 100 104*** 9*** 8 7** 82* 80 7 71 6 60 5 50 45 45 45 4 40 29 28 29 24 29 24 26 30 3 22 20 17 15 16 2 10 4 1 0 0 ls ills lls ic t y ls en lit ro tro om ki sk na sm nt cs on on of o co ifi es rs lc ec e ec ss No Pe Us Al cio sp ya b- So ac Jo er Lit Primary Senior high school Years of education Junior high school Tertiary Note: All models are estimated using ordinary least squares. The estimated coefficients for each education level (bars) and for the years of education (dotted line) are reported in the figure. See appendix E for full results. ***p < .01, **p < .05, *p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Returns to Education and Skills: Building the Job-Relevant Skills That Employers Demand 73 providing enough good, high-wage jobs. The majority of self-employment con- sists of informal income-earning activities and small businesses pursued by indi- viduals who are unable to secure (formal) wage employment. The key to accessing the formal wage sector is once again education (figure 7.3). About 44 percent of workers employed in the formal sector have a tertiary education and 32 percent have graduated from SHS, whereas only 24 percent have a JHS education or less. In Ghana the formal sector mainly consists of high-value-added service companies, which require a highly qualified labor force. Only tertiary education has a significant premium in the formal sector given the selection bias. This means, for example, that individuals with higher ability and motivation are more likely to choose and complete tertiary education and receive higher earnings. Graduates of SHS and JHS and those who have com- pleted primary education all have the same probability as workers with no for- mal education of being self-employed (figure 7.4). Furthermore, self-employed workers use their numeracy skills more intensively than all other employees and perform fewer repetitive tasks and use the computer less. With respect to socioemotional skills, only grit is positively correlated with being self-employed (see appendix E). Figure 7.3 Level of Education, by Employment Status 100 80 Urban adults (%) 60 40 20 0 Formal employee Informal employee Self-employed Tertiary Senior high school Junior high school Primary None Note: The figure excludes people currently attending school. Proportions may not add up to 100 percent because figures have been rounded. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 74 The Returns to Education and Skills: Building the Job-Relevant Skills That Employers Demand Figure 7.4 Linear Probability Model of Being Self-Employed 20 10 8 8.2* 3 3 5 4 3 2 1 3 2 –8** 1 1 0 –1 –2 –5 –4 –10 –8 –9** Percent –12*** –20 –30 –40 –41*** –50 ls ic ity ills ls ls t en ro kil tro om al sk sm nt cs on on on of co ifi es rs lc ec e ec ss Pe No Us Al cio sp ya b- So ac Jo er Lit Primary Junior high school Senior high school Tertiary Note: All models are estimated using linear probability model. The estimated coefficients for each education level (bars) are reported in the figure. See appendix E for full results. ***p < .01, **p < .05, *p < 0.1. Returns to Education: Is There a Gender Premium? The returns to an additional year of education are very similar—between 3 and 5 percent for men and between 3 and 6 percent for women (figure 7.5). The returns to primary education are not significantly different from zero for both men and women. For males only tertiary education guarantees a premium, whereas for women education starts to pay off starting from JHS. The market reward is always higher for women than for men. In fact, after controlling for all skills, the returns to tertiary education are 98 percent for females and 73 percent for men. This gender gap might be misleading because it does not take into account that both labor market participation and educational attainments are lower for women. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Returns to Education and Skills: Building the Job-Relevant Skills That Employers Demand 75 Figure 7.5 The Returns to Education a. Male workers 160 10 Return of level of education (%) 140 9 Return of years of education (%) 120 117*** 8 109*** 109*** 100 97*** 95*** 7 6*** 80 73*** 6 5*** 5*** 66*** 60 4*** 5 40 5** 4 20 2* 3* 3 0 2 –20 1 –40 0 ls ills lls ic t y ls en lit ro tro om ki sk na sm nt cs on on of co so ifi es lc ec e r ec ss No Pe Us Al cio sp ya b- So ac Jo er Lit b. Female workers 160 155*** 155*** 156*** 10 142*** 9 140 Return of years of education (%) 128*** 123*** 8 Return of level of education (%) 120 7 6*** 6*** 6*** 98*** 100 6 5*** 80 5*** 5 5*** 60 56*** 56*** 58*** 4 52*** 38*** 3** 3 40 35*** 35*** 33* 38** 33*** 24** 27** 2 20 1 0 0 ls ills lls ic t ity ls en ro tro m ki al sk m nt no cs on on of ss co ifi o rs lc se ec e ec No Pe Us Al as cio -sp cy So b a Jo er Lit Primary Senior high school Years of education Junior high school Tertiary Note: All models are estimated using ordinary least squares. The estimated coefficients for each education level (bars) and for the years of education (dotted line) are reported in the figure. See appendix E for full results. ***p < .01, **p < .05, *p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 76 The Returns to Education and Skills: Building the Job-Relevant Skills That Employers Demand Notes 1. We recognize the possible limitations of this estimation procedure in terms of possible multi-collinearity. Furthermore, a causal interpretation of the finding shown in the following sections is not either legitimate or claimed. 2. This is consistent with recent studies for many low-income countries showing that the returns to primary education are lower than those for the secondary level—see Mwabu and Schultz (2000) for South Africa, Kingdon (1997) for India, Siphambe (2000) for Botswana, among others. For example, Moll (1996) argues that a decline in school quality and changes in demand and supply factors largely account for the low returns to primary schooling in South Africa. References Becker, Gary S. 1962. “Investment in Human Capital: A Theoretical Analysis.” Journal of Political Economy 70 (5, Part 2): 9–49. Kingdon, Geeta Gandhi. 1997. “Does the Labour Market Explain Lower Female Schooling in India?” LSE Research Online Documents of Economics 6715, London School of Economics and Political Science. Mincer, Jacob. 1962. “On-the-Job Training: Costs, Returns, and Some Implications.” Journal of Political Economy 70 (5, Part 2): 50–79. Moll, Peter G. 1996. “The Collapse of Primary Schooling Returns in South Africa 1960–90.” Oxford Bulletin of Economics and Statistics 58 (1): 185–209. Mwabu, Germano, and T. Paul Schultz. 2000. “Wage Premiums for Education and location of South African Workers, by Gender and Race.” Economic Development and Cultural Change 48 (2) 307–34. Schultz, Theodore W. 1961. “Investment in Human Capital.” American Economic Review 51 (1): 1–17. Siphambe, Happy Kufigwa. 2000. “Rates of Return to Education in Botswana.” Economics of Education Review 19 (3): 291–300. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Chapter 8 Mismatch of Skills: A Measurement Issue and Unexploited Potential at Work Mismatch between Self-Reported Skills and Core Literacy Test Results: Does Language Matter? Several of the cognitive and job-specific skills captured in the Skills Toward Employment and Productivity (STEP) survey are self-reported. These self- reported measures relate to the intensity of use of these skills at work or outside work. The self-reported reading skill measure can capture a subjective assessment of the mastery and usefulness of the skill, but a key question is how well these assessments fit in with some objective measures such as the Core Literacy Test. If subjective and objective measures reflected a real ability, those reporting that they read frequently should also be able to pass the Core test. However, although 69 percent of adults reported reading regularly, only 42 percent of the population passed the Core test. This considerable mismatch is due to both the high nonresponse rate for the objective literacy assessment (in contrast to low missing values for the subjective measurements) and the difficulties encountered in the Core test because of the language used. As a consequence, the Core test is more likely to be a measure of the individual’s English literacy than of his or her general reading skills. Notably, the self-reported assessment of the use of skills does not specify a particular lan- guage, and respondents might indeed take it to mean any of the local languages widely used not only in the rural Ghana but also in urban areas. There is a considerable mismatch between the self-reported and the assessed reading skills (that is, between those reporting that they read and those passing the Core test). Indeed, about 26 percent of those who reported reading did not pass the Core test. Still, those not passing the Core test and reporting reading are 41 percentage points more likely to speak English at work (74 percent use English at work whereas 33 percent do not) and slightly more likely to speak English at home (close to 2 percentage points). Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   77   78 Mismatch of Skills: A Measurement Issue and Unexploited Potential at Work The mismatch between self-reported and measured literacy is not randomly distributed (see appendix F). Those who failed the Core test and reported being able to read are more likely to be women, to be less educated, to work in low-skilled occupations and less in high-value-added sectors, and to earn about 41 percent less than those who report being able to read and also pass the Core test. The difference in the intensity of use of reading skills might definitively explain the skills mismatch between self-reported and objective measures of skills. The intensity of use of the skill is a more closely related measure to the Core test. Among those who self-reported reading with low intensity, 66 percent failed the text and 30 percent passed it. Only 14 percent of those who failed indicated reading frequently. Are There Unexploited Skills in the Workforce? Among the employed population, a group of individuals reported using their cognitive skills and computer skills in their daily life (that is, using skills out- side the realm of a paid job, including a family business) but not at work. This group of people might represent unexploited human capital that the labor market is not able to absorb. In this section, we analyze the profile of those who report using their skills only outside work and those who use their skills at work. Women—more frequently than men—are likely to be literate but employed in occupations not requiring the use of reading and writing skills. Similarly, there is an unexploited potential among the youngest generations who are able to use computers but work in jobs where this skill is not demanded. Being able to use their computer skills at work would guarantee them a monthly income 38 percent higher than they receive in their current occupations. Those using a skill only at home are less likely to work at a high-skilled occupation and in a high-value-added sector and are more likely to work at a mid-skilled occupation and a low- to mid-value-added sector. The greatest unexploited potential, as expected, is among those in low-skilled occupa- tions and low-value-added sectors. For example, about 34 percent of the self-employed read at home but never use their reading skills at work. Conversely, only 6 percent of employees in the formal sector have a skills mismatch. The level of education seems to be closely related to the use of certain skills at work. The level of education of those who read, write, or use computers only at home is lower than for those using their skills at work. The latter are predomi- nantly senior high school (SHS) graduates. The fact that the mismatch of skills occurs more at lower levels of education than SHS might suggest that SHS pro- vides students with the right skills demanded by the labor market and that SHS education provides a clear signal to employers of the cognitive skills potential employees have. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Mismatch of Skills: A Measurement Issue and Unexploited Potential at Work 79 To a certain extent the mismatch of skills used at home but remaining unused at work does reflect the quality of those skills rather than an unexploited potential. Indeed, on average those using their (reading, writing, and computer) skills only at home performed worse on both the Reading Component and the Core test than those who also use them at work. To sum up, there is a clear connection between the availability of skills and the use of skills at work, whether the skills are cognitive or work related. Moreover, the use of the skills at work appears to explain the difference between objectively measured and self-reported skills. Those who work in low-skilled occupations are more likely to report higher skills than are measured objectively through the literacy assessment. Conclusions and Looking Forward In conclusion, the study has provided evidence in a number of areas that are relevant for policy makers, for setting government investment priorities and also for continued diagnostic work. The study demonstrated the positive impact of early childhood education on schooling, on jobs, and on earnings. The country continues to demonstrate progress in education attainment at all levels, but there is also evidence of persistent disparities in attainment by income, gender, and geographical location. There are also continued difficulties with delayed educa- tion, dropout, and low completion rate—mostly for the poor, given the high opportunity costs. Importantly, improved education attainment did not bring about comparable improvements in cognitive skills as measured by the adult literacy assessment. The labor market also shows persistent segregation between employment in higher-value sectors, where employees have higher attainment and education pays off, and employment in sectors such as agriculture, fishing, and mining, where there appears to be an equilibrium of low education and low income. In the first group, employees demonstrate and report higher skills. There is a clear connection between the availability of skills and the use of skills at work, whether the skills are cognitive or work related. Moreover, the use of the skills at work appears to explain the difference between objectively measured and self-reported skills. Those who work in low-skilled occupations are more likely to report higher skills than are measured objectively through the literacy assessment. These findings offer some priorities and options for policy and investment. For the country, investment and scaled-up access and quality in early childhood edu- cation would lead to sustained long-term growth and shared prosperity. For education policy, continued investment into improved literacy and also improved socioemotional and work skills helps build the country’s human capital (increas- ing demand for postbasic, postsecondary education) and improved labor force. Sustainable forms of skills development should incentivize demand and supply of skills through both formal and on-the-job training. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 80 Mismatch of Skills: A Measurement Issue and Unexploited Potential at Work The study also offers options for continued diagnostic work. The household survey data could help carry out more in depth analysis on the return to educa- tion and skills (and the role of training), on skills formation (and how cognitive and noncognitive skills correlate), and cross-countries comparison with other similar datasets. Finally, this research could be complemented with a STEP employer survey that focuses on the demand for education and both cognitive and noncognitive skills by both public and private employers in Ghana’s key economic sectors. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Appendix A Summary of Statistics Table A.1 Summary of Statistics from the STEP Survey Sample observations Mean SD Min Max Female (%) 2,987 58 49 0 100 Age range (years) (%) 2,987 33 13 15 64   15–19 2,987 14 35 0 100   20–24 2,987 18 38 0 100   25–34 2,987 28 45 0 100   35–44 2,987 19 39 0 100   45–64 2,987 21 40 0 100 Education (%)   No education 2,987 21 40 0 100   Primary education 2,987 14 35 0 100   Junior high school 2,987 36 48 0 100   Senior high school 2,987 20 40 0 100   Tertiary education 2,987 10 29 0 100   Received early childhood education 2,948 64 48 0 100 Socioeconomic status (%)  Low 2,960 23 42 0 100  Middle 2,960 56 50 0 100   High 2,960 21 41 0 100 Region (%)  Western 2,987 9 29 0 100  Central 2,987 11 31 0 100   Greater Accra 2,987 26 44 0 100  Volta 2,987 5 22 0 100  Eastern 2,987 6 24 0 100  Ashanti 2,987 22 41 0 100   Brong-Ahafo 2,987 10 31 0 100  Northern 2,987 7 26 0 100   Upper East 2,987 2 15 0 100 table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   81   82 Summary of Statistics Table A.1  Summary of Statistics from the STEP Survey (continued) Sample observations Mean SD Min Max Labor status (%)  Employed 2,981 62 48 0 100  Unemployed 2,981 5 22 0 100  NEET 2,981 10 30 0 100  Inactive 2,981 23 42 0 100 Employment status (%)   Formal employee 1,973 15 35 0 100   Informal employee 1,973 20 40 0 100   Self-employed 1,973 66 48 0 100 Economic sector (%)   Agriculture, fishing, and mining 1,973 12 33 0 100  Manufacturing 1,973 10 30 0 100   Low- to mid-value-added 1,973 61 49 0 100   High-value-added 1,973 17 37 0 100 Occupation (%)   Low-skilled occupation 1,973 39 49 0 100   Mid-skilled occupation 1,973 48 50 0 100   High-skilled occupation 1,973 13 34 0 100   Monthly earnings (in GHS) 1,993 516 1,580 0 34,800 Language spoken at work (%)  Akan 2,344 80 49 0 100  Ewe 2,344 8 27 0 100   Ga-Adangme 2,344 13 33 0 100   Mole-Dagbani 2,344 8 27 0 100  English 2,344 40 49 0 100  Others 2,344 12 32 0 100 Language spoken at home (%)  Akan 2,972 63 48 0 100  Ewe 2,972 8 27 0 100   Ga-Adangme 2,972 7 26 0 100   Mole-Dagbani 2,972 8 28 0 100  English 2,972 1 10 0 100  Others 2,972 13 33 0 100 Note: Variables are described in appendix C. NEET = not in employment, education, or training; STEP = Skills Toward Employment and Productivity. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Summary of Statistics 83 Table A.2 Sampling Procedure Comparison of STEP, GLSS 5, and GLSS 6 GLSS 5, 2005–06 GLSS 6, 2012–13 STEP, 2011–13 Sampling frame 2000 PHC 2010 PHC 2010 PHC Target population All noninstitutionalized All noninstitutionalized All noninstitutionalized persons living in persons living in persons living in private private dwellings private dwellings dwellings in urban areas of the country at the time of data collection Sample design Sample design Two-stage stratified Two-stage stratified Four-stage (implicit stratified) Sample unit PSU: EA PSU: EA PSU: EA SSU: household SSU: household Second stage sample unit: PSU partition Third stage sample unit: household Fourth stage sample unit: individual Sample size Number of 580 1,200 201 PSUs Number of 8,700 18,000 3,015 household Observations of 4,074 11,538 2,987 urban population Note: EA = Census Enumeration Area; GLSS = Ghana Living Standards Survey; PHC = Population and Housing Census; PSU = primary sample unit; SSU = secondary sample unit; STEP = Skills Toward Employment and Productivity. Table A.3 Regional PSU Sample Size Comparison of STEP, GLSS 5, and GLSS 6 GLSS 5 GLSS 6 STEP Region n % n % n % Western 20 8 51 9 16 8 Central 17 7 55 10 17 8 Greater Accra 73 30 130 24 57 28 Volta 13 5 39 7 12 6 Eastern 22 9 56 10 18 9 Ashanti 53 22 90 17 47 23 Brong-Ahafo 20 8 52 10 16 8 Northern 14 6 35 6 12 6 Upper East 5 2 21 4 4 2 Upper West 3 1 16 3 2 1 Total 240 545 201 Note: GLSS = Ghana Living Standards Survey; PSU = primary sample unit; STEP = Skills Toward Employment and Productivity. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Appendix B Skills Definitions, Survey Questions, and Aggregation Strategy Self-Reported Reading, Writing, and Numeracy Survey Questions Self-reported reading “Do you read anything [in daily life/at this work], including very short notes or instructions that are only a few sentences long?” “Among the things that you normally read [in daily life/at this work], what is the size of the longest document that you read?” Self-reported writing “Do you ever have to write anything (else) [in daily life/at work], including very short notes, lists, or instructions that are only a few sentences long?” “Thinking about all the things you normally write (wrote) [in daily life/at work], what is the longest document that you write (wrote)?” Aggregation: Intensity of use Level Does not do read/write = Does not use 0 Reads/writes documents of 5 pages or less = Low 1 Reads/writes documents of 6 to 25 pages = Medium 2 Reads/writes documents of more than 25 pages = High 3 Self-reported numeracy “[As a normal part of this work /in daily life], do you do any of the following...?” Aggregation: Complexity of use Level Does no math = Does not use 0 Measures or estimates sizes, weights, distances Calculates prices or costs = Low 1 Performs any other multiplication or division Uses or calculates fractions, decimals or percentages = Medium 2 Uses more advanced math such as algebra, geometry, trigonometry = High 3 Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   85   86 Skills Definitions, Survey Questions, and Aggregation Strategy Specific On-the-Job Skills Survey Questions Computer use “As a part of your work do you (did you) use a computer?” “As a part of your life [outside of work as (OCCUPATION)] have you used a computer in the past 3 months?” Aggregation Intensity of use Level Does not use a computer/use a computer almost never = Does not use 0 Uses computer less than three times per week = Low 1 Uses computer three times or more per week = Medium 2 Uses computer every day = High 3 External interpersonal skills “As part of this work, do you (did you) have any contact with people other than co-workers, for example with customers, clients, students, or the public?” Using any number from 1 to 10, where 1 is little involvement (…) and 10 means much of the work involves meeting or interacting (…) what number would you use to rate this work? Aggregation Intensity of use Level Does not have any contacts with clients = Does not use 0 Involvement scale ranges from 1 to 4 = Low 1 Involvement scale ranges from 5 to 7 = Medium 2 Involvement scale ranges from 8 to 10 = High 3 Physical tasks “Using any number from 1 to 10 where 1 is not at all physically demanding (such as sitting at a desk answering a telephone) and 10 is extremely physically demanding (such as carrying heavy loads, construction worker, etc), what num- ber would you use to rate how physically demanding your work is?” Aggregation Intensity of use Level Not at all physically demanding = Does not use 0 Physical demand scale ranges from 2 to 4 = Low 1 Physical demand scale ranges from 5 to 6 = Medium 2 Physical demand scale ranges from 7 to 10 = High 3 Cognitive challenge: average of two indicators “Some tasks are pretty easy and can be done right away or after getting a little help from others. Other tasks require more thinking to figure out how they should be done. As part of this work as [OCCUPATION], how often do you have to undertake tasks that require at least 30 minutes of thinking?” Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Skills Definitions, Survey Questions, and Aggregation Strategy 87 Aggregation Intensity of use Level Never = Does not use 0 Less than once per month = Low 1 Less than once a week but at least once a month OR at least once a week but not every day = Medium 2 Every day = High 3 “How often does (did) this work involve learning new things?” Aggregation Intensity of use Level Rarely = Does not use 0 At least 2–3 months or at least once a month = Low 1 At least once a week = Medium 2 Every day = High 3 Autonomy and repetitiveness: average of two indicators “Still thinking of your work as [OCCUPATION] how much freedom do you (did you) have to decide how to do your work in your own way, rather than fol- lowing a fixed procedure or a supervisor’s instructions? Use any number from 1 to 10 where 1 is no freedom and 10 is complete freedom.” Aggregation Intensity Level Decision freedom scale from 1 to 2 = Close to none 0 Decision freedom scale from 3 to 6 = Low 1 Decision freedom scale from 7 to 9 = Medium 2 Decision freedom scale 10 = High 3 “How often does (did) this work involve carrying out short, repetitive tasks?” Aggregation Intensity Level Almost all the time = Close to none 0 More than half the time = Low 1 Less than half the time = Medium 2 Almost never = High 3 Job learning time: based on three questions “What minimum level of formal education do you think would be required before someone would be able to carry out this work? How many years of work experience in other related work do you think would be required before someone with [FILL Q22] would be able to carry out this work? About how long would it take someone to learn to do this work well if they had [FILL Q22] education and [FILL Q23] years of related work experience?” Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 88 Skills Definitions, Survey Questions, and Aggregation Strategy Socioemotional Skills Survey Questions Scale Almost never 1 Some of the time 2 Most of the time 3 Almost always 4 Openness - Do you come up with ideas other people haven’t thought of before? - Are you very interested in learning new things? - Do you enjoy beautiful things, like nature, art, and music? Conscientiousness - When doing a task, are you very careful? - Do you prefer relaxation more than hard work? - Do you work very well and quickly? Extraversion - Are you talkative? - Do you like to keep your opinions to yourself? - Are you outgoing and sociable? For example, do you make friends very easily? Agreeableness - Do you forgive other people easily? - Are you very polite to other people? - Are you generous to other people with your time or money? Emotional stability (neuroticism) - Are you relaxed during stressful situations? - Do you tend to worry? - Do you get nervous easily? Grit - Do you finish whatever you begin? - Do you work very hard? For example, do you keep working when others stop to take a break? - Do you enjoy working on things that take a very long time (at least several months) to complete? Hostile bias - Do people take advantage of you? - Are people mean/not nice to you? Decision making - Do you think about how the things you do will affect you in the future? - Do you think carefully before you make an important decision? - Do you ask for help when you don’t understand something? Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Appendix C Definitions of Variables Used in the Analysis Definition Female A dummy variable equal to 1 if respondent is female, and 0 otherwise. Age Age of respondents: a dummy variable for each one of the following age groups: 15–19 years; 20–24 years; 25–34 years; 35–44 years; 45–64 years Education level A dummy variable for each one of the following education levels completed: None (equal to 1 if respondent has not completed any formal education level. Takes value 0 if respondent completed at least primary education); primary education (1 for those whose highest level of education completed is primary education); junior high school (1 for those whose highest level of education completed is junior high school); senior high school (1 for those whose highest level of education completed is senior high school); tertiary education (1 for those whose highest level of education completed is tertiary education) Early childhood A dummy variable equal to 1 if the respondent attended early childhood education (%) education, and 0 otherwise. Low socioeconomic A dummy variable equal to 1 if the respondent classified him or herself as having status (%) belonged to the three first income deciles at age 15 years, and 0 otherwise. Middle socioeconomic A dummy variable equal to 1 if the respondent classified themselves as having status (%) belonged to the 4th to 6th income deciles at age 15 years, and 0 otherwise. High socioeconomic A dummy variable equal to 1 if the respondent classified themselves as having status (%) belonged to the 7th to 10th income deciles at age 15 years, and 0 otherwise. Region A dummy variable equal to 1 if the respondent lives in this region, and 0 otherwise. Labor status Employed (%) A dummy variable equal to 1 if the respondent is employed and not in education, and 0 otherwise. Unemployed (%) A dummy variable equal to 1 if the respondent is unemployed and not in education, and 0 otherwise. NEET (%) A dummy variable equal to 1 if the respondent is inactive and not in education, or training or is retired, and 0 otherwise. Inactive (%) A dummy variable equal to 1 if the respondent is inactive and not NEET, and 0 otherwise. table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   89   90 Definitions of Variables Used in the Analysis Definition Employment status only defined for employed population Formal employee (%) A dummy variable equal to 1 if the respondent is a wageworker and does have social security, and 0 otherwise. Informal A dummy variable equal to 1 if the respondent is a wageworker and does not employee (%) have social security, and 0 otherwise. Self-employed (%) A dummy variable equal to 1 if the respondent is self-employed or an unpaid family worker, and 0 otherwise. Economic sector defined only for employed population following Economic sector coded to 2 digits, ISIC rev.4 Agriculture, fishing, A dummy variable equal to 1 if economic sector is agriculture, fishing, and and mining (%) mining and 0 otherwise. Manufacturing (%) A dummy variable equal to 1 if economic sector is manufacturing Low- to mid-value- A dummy variable equal to 1 if economic sector is electricity, gas, steam and air added (%) conditioning supply; water supply; sewerage, waste management, and remediation activities; construction; wholesale and retail trade; repair of motor vehicles and motorcycles; transportation and storage; accommodation and food service activities; information and communication; arts, entertainment and recreation; other service activities; activities of households; extraterritorial organizations and bodies, and 0 otherwise. High-value-added (%) A dummy variable equal to 1 if economic sector is 64–66 financial and insurance activities; real estate activities; professional, scientific and technical activities; administrative and support service activities; public administration and social security; education; human health and social work activities, and 0 otherwise. Occupation defined only for employed population Low-skilled A dummy variable equal to 1 if occupation is skilled agricultural, forestry, and occupation (%) fishery worker; craft and related trades worker; plant and machine operator and assembler; elementary occupations and 0 otherwise. Mid-skilled A dummy variable equal to 1 if occupation is technician and associate occupation (%) professional; clerical support worker; service and sales worker, and 0 otherwise. High-skilled A dummy variable equal to 1 if occupation is manager or professional and 0 occupation (%) otherwise. Monthly earnings Considering the main and second occupation (in GHS) Language Language at work A dummy variable equal to 1 if the respondent speaks this language at work, and 0 otherwise. Language at home A dummy variable equal to 1 if the respondent speaks mainly this language at home, and 0 otherwise. Note: NEET = not in employment, education, or training. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Appendix D Differences in Mean Table D.1  Difference in Mean of Those Passing the Core Literacy Test Threshold and Those Failing the Core Y1: Failed Core test Y2: Passed Core test Obs Mean SD Obs Mean SD Difference: Y2 – Y1 Female (%) 1,274 70.06 45.82 1,193 43.16 49.55 −26.897*** Socioeconomic status (%)  Low 1,265 29.92 45.81 1,192 14.69 35.42 −15.233***  Middle 1,265 51.23 50.00 1,192 61.81 48.61 10.576***   High 1,265 18.84 39.12 1,192 23.50 42.42 4.657** Age range (years) (%) 1,274 35.55 12.73 1,193 28.85 12.27 −6.695***   15–19 1,274 9.15 28.85 1,193 23.23 42.25 14.080***   20–24 1,274 12.51 33.10 1,193 25.69 43.71 13.171***   25–34 1,274 29.25 45.51 1,193 26.25 44.02 −3.002   35–44 1,274 25.28 43.48 1,193 10.79 31.03 −14.491***   45–64 1,274 23.80 42.61 1,193 14.05 34.76 −9.759*** Education (%)   No education 1,274 42.14 49.40 1,193 0.60 7.75 −41.540***   Primary education 1,274 18.56 38.89 1,193 9.44 29.25 −9.116***   Junior high school 1,274 32.75 46.95 1,193 32.42 46.83 −0.324   Senior high school 1,274 6.25 24.22 1,193 36.81 48.25 30.558***   Tertiary education 1,274 0.30 5.45 1,193 20.72 40.55 20.422***   Received early childhood education 1,260 47.09 49.94 1,184 81.25 39.04 34.162*** Language spoken at work (%)  Akan 1,077 79.04 40.72 837 77.88 41.53 −1.161  Ewe 1,077 6.60 24.84 837 7.10 25.69 0.495   Ga-Adangme 1,077 9.51 29.35 837 14.81 35.54 5.299***   Mole-Dagbani 1,077 12.10 32.62 837 3.09 17.31 −9.006***  English 1,077 16.72 37.33 837 73.28 44.28 56.559***  Others 1,077 16.30 36.96 837 7.82 26.87 −8.480*** Language spoken at home (%)  Akan 1,274 58.43 49.30 1,193 65.35 47.60 6.925***  Ewe 1,274 5.98 23.72 1,193 9.03 28.67 3.048**   Ga-Adangme 1,274 3.55 18.50 1,193 9.75 29.68 6.204*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   91   92 Differences in Mean Table D.1  Difference in Mean of Those Passing the Core Literacy Test Threshold and Those Failing the Core (continued) Y1: Failed Core test Y2: Passed Core test Obs Mean SD Obs Mean SD Difference: Y2 – Y1   Mole-Dagbani 1,274 12.86 33.48 1,193 5.42 22.65 −7.438***  English 1,274 0.23 4.79 1,193 1.98 13.92 1.745***  Others 1,274 18.96 39.21 1,193 8.47 27.86 −10.483*** Labor status (%)  Employed 1,274 73.31 44.25 1,193 45.84 49.85 −27.472***  Unemployed 1,274 4.05 19.72 1,193 6.39 24.46 2.337*  NEET 1,274 11.97 32.47 1,193 7.74 26.74 −4.226***  Inactive 1,274 10.67 30.89 1,193 40.03 49.02 29.362*** Employment status (%)   Formal employee 964 3.29 17.85 611 33.81 47.34 30.517***   Informal employee 964 15.62 36.32 611 28.03 44.95 12.404***   Self-employed 964 81.09 39.18 611 38.17 48.62 −42.922*** Occupation (%)   Low-skilled occupation 964 44.74 49.75 611 25.04 43.36 −19.695***   Mid-skilled occupation 964 52.75 49.95 611 41.42 49.30 −11.326***   High-skilled occupation 964 2.52 15.67 611 33.54 47.25 31.021*** Economic sector (%)   Agriculture, fishing, and mining 964 17.20 37.76 611 5.22 22.26 −11.981***  Manufacturing 964 10.04 30.07 611 8.82 28.39 −1.220   Low-to mid-value-added 964 68.61 46.43 611 47.40 49.97 −21.208***   High-value-added 964 4.15 19.94 611 38.55 48.71 34.409*** Earnings   Monthly earnings 928 388.33 1,213.30 674 662.31 2,046.88 273.980** Socioemotional skills   Extraversion (score) 437 2.43 0.60 1,175 2.58 0.60 0.149***   Missing extraversion 1,274 67.27 46.94 1,193 1.53 12.26 −65.742***   Conscientiousness (score) 431 2.93 0.64 1,175 3.31 0.53 0.375***   Missing conscientiousness 1,274 67.37 46.90 1,193 1.53 12.26 −65.844***   Openness (score) 432 2.82 0.63 1,175 3.22 0.52 0.395***   Missing openness 1,274 67.45 46.88 1,193 1.53 12.26 −65.920***   Emotional stability (score) 426 2.65 0.59 1,173 2.75 0.55 0.097**   Missing stability 1,274 67.93 46.69 1,193 1.87 13.56 −66.052***   Agreeableness (score) 430 2.79 0.69 1,175 3.15 0.58 0.363***   Missing agreeableness 1,274 67.47 46.87 1,193 1.53 12.26 −65.944*** Self-reported skills Reading   Use reading skill 1,274 41.71 49.33 1,193 96.66 17.97 54.957***   Read with low intensity 546 68.68 46.42 1,131 30.46 46.04 −38.220***   Read with medium intensity 546 19.12 39.36 1,131 28.23 45.03 9.111***   Read with high intensity 546 12.20 32.76 1,131 41.31 49.26 29.109*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Differences in Mean 93 Table D.1  Difference in Mean of Those Passing the Core Literacy Test Threshold and Those Failing the Core (continued) Y1: Failed Core test Y2: Passed Core test Obs Mean SD Obs Mean SD Difference: Y2 – Y1 Writing   Use writing skill 1,274 38.02 48.56 1,193 93.31 25.00 55.294***   Write with low intensity 506 82.98 37.62 1,096 63.43 48.18 −19.545***   Write with medium intensity 506 9.95 29.96 1,096 19.82 39.89 9.873***   Write with high intensity 506 7.07 25.66 1,096 16.74 37.35 9.672*** Numeracy   Use numeracy skill 1,274 93.71 24.28 1,193 95.82 20.01 2.110*   Numeracy with low intensity 1,201 44.55 49.72 1,134 11.94 32.44 −32.616***   Numeracy with medium intensity 1,201 53.01 49.93 1,134 62.01 48.56 9.007***   Numeracy with high intensity 1,201 2.44 15.43 1,134 26.05 43.91 23.610*** Computer   Use computer skill 1,274 6.63 24.90 1,193 56.37 49.61 49.736***   Computer with low intensity 70 45.36 50.14 643 32.73 46.96 −12.623   Computer with medium intensity 70 12.67 33.50 643 18.12 38.55 5.456   Computer with high intensity 70 41.97 49.71 643 49.14 50.03 7.168 Literacy assessment Reading proficiency 1,274 77.11 1,193 221.68 144.571*** Note: NEET = not in employment, education, or training; Obs = number of observations. ***p < .01, **p < .05, *p < 0.1. Table D.2  Difference in Mean of Those Passing the Core Literacy Test Threshold and Those English Illiterate Y1: English illiterate Y2: Passed Core test Obs Mean SD Obs Mean SD Difference: Y2 – Y1 Female (%) 659 75.12 43.27 1,193 43.16 49.55 −31.956*** Socioeconomic status (%)  Low 656 36.07 48.06 1,192 14.69 35.42 −21.379***  Middle 656 48.08 50.00 1,192 61.81 48.61 13.723***   High 656 15.84 36.54 1,192 23.50 42.42 7.656*** Age range (years) (%) 659 39.33 12.18 1,193 28.85 12.27 −10.474***   15–19 659 2.43 15.41 1,193 23.23 42.25 20.804***   20–24 659 7.48 26.33 1,193 25.69 43.71 18.201***   25–34 659 28.61 45.23 1,193 26.25 44.02 −2.358   35–44 659 30.33 46.00 1,193 10.79 31.03 −19.541***   45–64 659 31.15 46.35 1,193 14.05 34.76 −17.106*** Education (%)   No education 659 72.14 44.87 1,193 0.60 7.75 −71.534***   Primary education 659 13.41 34.10 1,193 9.44 29.25 −3.967**   Junior high school 659 13.85 34.57 1,193 32.42 46.83 18.572***   Senior high school 659 0.60 7.74 1,193 36.81 48.25 36.209***   Tertiary education 659 0.00 0.00 1,193 20.72 40.55 20.720***   Received early childhood education 651 30.96 46.27 1,184 81.25 39.04 50.288*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 94 Differences in Mean Table D.2  Difference in Mean of Those Passing the Core Literacy Test Threshold and Those English Illiterate (continued) Y1: English illiterate Y2: Passed Core test Obs Mean SD Obs Mean SD Difference: Y2 – Y1 Language spoken at work (%) Akan 570 70.99 45.42 837 77.88 41.53 6.890** Ewe 570 6.82 25.22 837 7.10 25.69 0.279 Ga-Adangme 570 10.90 31.19 837 14.81 35.54 3.916* Mole-Dagbani 570 17.46 38.00 837 3.09 17.31 −14.373*** English 570 5.77 23.34 837 73.28 44.28 67.508*** Others 570 21.33 41.00 837 7.82 26.87 −13.508*** Language spoken at home (%) Akan 659 45.85 49.87 1,193 65.35 47.60 19.500*** Ewe 659 6.68 24.98 1,193 9.03 28.67 2.355* Ga-Adangme 659 3.70 18.89 1,193 9.75 29.68 6.049*** Mole-Dagbani 659 18.43 38.80 1,193 5.42 22.65 −13.013*** English 659 0.07 2.68 1,193 1.98 13.92 1.903*** Others 659 25.27 43.49 1,193 8.47 27.86 −16.795*** Labor status (%) Employed 659 81.53 38.84 1,193 45.84 49.85 −35.690*** Unemployed 659 3.67 18.81 1,193 6.39 24.46 2.716* NEET 659 12.64 33.25 1,193 7.74 26.74 −4.896** Inactive 659 2.17 14.57 1,193 40.03 49.02 37.869*** Employment status (%) Formal employee 539 0.76 8.68 611 33.81 47.34 33.049*** Informal employee 539 13.93 34.66 611 28.03 44.95 14.095*** Self-employed 539 85.31 35.43 611 38.17 48.62 −47.143*** Occupation (%) Low-skilled occupation 539 45.81 49.87 611 25.04 43.36 −20.771*** Mid-skilled occupation 539 51.91 50.01 611 41.42 49.30 −10.490*** High-skilled occupation 539 2.28 14.93 611 33.54 47.25 31.261*** Economic sector (%) Agriculture, fishing, and mining 539 20.89 40.69 611 5.22 22.26 −15.666*** Manufacturing 539 9.43 29.26 611 8.82 28.39 −0.609 Low- to mid-value-added 539 67.81 46.76 611 47.40 49.97 −20.410*** High-value-added 539 1.87 13.56 611 38.55 48.71 36.685*** Earnings   Monthly earnings 519 409.79 1,442.84 674 662.31 2,046.88 252.526* Socioemotional skills   Extraversion (score) 28 2.38 0.51 1,175 2.58 0.60 0.203**   Missing extraversion 659 96.43 18.57 1,193 1.53 12.26 −94.905***   Conscientiousness (score) 27 3.12 0.56 1,175 3.31 0.53 0.187*   Missing conscientiousness 659 96.53 18.31 1,193 1.53 12.26 −95.008***   Openness (score) 26 2.55 0.70 1,175 3.22 0.52 0.664***   Missing openness 659 96.59 18.16 1,193 1.53 12.26 −95.065*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Differences in Mean 95 Table D.2  Difference in Mean of Those Passing the Core Literacy Test Threshold and Those English Illiterate (continued) Y1: English illiterate Y2: Passed Core test Obs Mean SD Obs Mean SD Difference: Y2 – Y1   Emotional stability (score) 25 2.85 0.72 1,173 2.75 0.55 −0.100   Missing stability 659 96.83 17.54 1,193 1.87 13.56 −94.956***   Agreeableness (score) 26 2.55 0.69 1,175 3.15 0.58 0.605***   Missing agreeableness 659 96.62 18.10 1,193 1.53 12.26 −95.089*** Self-reported skills Reading   Use reading skill 659 13.19 33.87 1,193 96.66 17.97 83.471***   Read with low intensity 98 81.36 39.15 1,131 30.46 46.04 −50.899***   Read with medium intensity 98 13.57 34.43 1,131 28.23 45.03 14.663***   Read with high intensity 98 5.07 22.06 1,131 41.31 49.26 36.236*** Writing   Use writing skill 659 11.83 32.32 1,193 93.31 25.00 81.481***   Write with low intensity 90 89.48 30.85 1,096 63.43 48.18 −26.051***   Write with medium intensity 90 10.52 30.85 1,096 19.82 39.89 9.307**   Write with high intensity 90 0.00 0.00 1,096 16.74 37.35 16.744*** Numeracy   Use numeracy skill 659 91.56 27.82 1,193 95.82 20.01 4.263**   Numeracy with low intensity 611 53.60 49.91 1,134 11.94 32.44 −41.665***   Numeracy with medium intensity 611 46.16 49.89 1,134 62.01 48.56 15.853***   Numeracy with high intensity 611 0.24 4.85 1,134 26.05 43.91 25.813*** Computer   Use computer skill 659 0.59 7.65 1,193 56.37 49.61 55.783***   Computer with low intensity 3 50.79 61.23 643 32.73 46.96 −18.053   Computer with medium intensity 3 0.00 0.00 643 18.12 38.55 18.124***   Computer with high intensity 3 49.21 61.23 643 49.14 50.03 −0.071 Literacy assessment Reading proficiency 659 104.82 1,193 221.68 116.864*** Note: NEET = not in employment, education, or training; Obs = number of observations. *** p < 0.01, ** p < 0.05, * p < 0.1. Table D.3  Difference in Mean of Those Answering Core Literacy Test and Those Who Did Not Answer It Y1: Did not answer Core Y2: Answered Core Obs Mean SD Obs Mean SD Difference: Y2 – Y1 Female (%) 1,179 70.37 45.68 1,808 50.26 50.01 −20.110*** Socioeconomic status (%)  Low 1,159 30.85 46.21 1,801 17.60 38.09 −13.248***  Middle 1,159 50.72 50.02 1,801 59.42 49.12 8.693***   High SES 1,159 18.43 38.79 1,801 22.99 42.09 4.555** Age range (years) (%) 1,179 38.27 12.32 1,808 29.78 12.29 −8.493*** 15–19 1,179 3.76 19.04 1,808 20.88 40.66 17.117*** 20–24 1,179 9.13 28.81 1,808 23.06 42.14 13.938*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 96 Differences in Mean Table D.3  Difference in Mean of Those Answering Core Literacy Test and Those Who Did Not Answer It (continued) Y1: Did not answer Core Y2: Answered Core Obs Mean SD Obs Mean SD Difference: Y2 – Y1 25–34 1,179 29.40 45.58 1,808 27.45 44.64 −1.954 35–44 1,179 27.77 44.81 1,808 13.83 34.53 −13.942*** 45–64 1,179 29.94 45.82 1,808 14.78 35.50 −15.159*** Education (%) No education 1,179 47.55 49.96 1,808 4.14 19.94 −43.401*** Primary education 1,179 13.58 34.28 1,808 14.14 34.86 0.559 JHS education 1,179 30.74 46.16 1,808 38.85 48.76 8.111*** SHS education 1,179 5.93 23.63 1,808 28.72 45.26 22.784*** Tertiary education 1,179 2.20 14.67 1,808 14.14 34.86 11.947*** Received ECE 1,155 44.52 49.72 1,793 75.50 43.02 30.983*** Language spoken at work (%) Akan 1,000 77.15 42.01 1,344 81.67 38.70 4.527** Ewe 1,000 9.20 28.92 1,344 6.82 25.22 −2.382 Ga-Adangme 1,000 13.43 34.11 1,344 12.24 32.79 −1.188 Mole-Dagbani 1,000 12.19 32.73 1,344 4.25 20.17 −7.945*** English 1,000 17.70 38.19 1,344 56.54 49.59 38.836*** Others 1,000 15.38 36.09 1,344 8.92 28.52 −6.457*** Language spoken at home (%) Akan 1,164 55.04 49.77 1,808 67.30 46.93 12.255*** Ewe 1,164 8.74 28.25 1,808 7.80 26.83 −0.935 Ga-Adangme 1,164 5.88 23.53 1,808 7.67 26.62 1.794 Mole-Dagbani 1,164 12.37 32.93 1,808 5.98 23.72 −6.382*** English 1,164 0.48 6.92 1,808 1.46 11.99 0.977*** Others 1,164 17.50 38.01 1,808 9.79 29.72 −7.709*** Labor status (%) Employed 1,173 79.48 40.40 1,808 52.07 49.97 −27.410*** Unemployed 1,173 3.72 18.93 1,808 5.75 23.28 2.029* NEET 1,173 11.32 31.70 1,808 8.90 28.48 −2.423 Inactive 1,173 5.48 22.77 1,808 33.28 47.14 27.803*** Employment status (%) Formal employee 937 5.80 23.39 1,036 22.70 41.91 16.901*** Informal employee 937 15.41 36.13 1,036 23.86 42.64 8.443*** Self-employed 937 78.79 40.90 1,036 53.44 49.91 −25.344*** Occupation (%) Low-skilled occupation 937 46.89 49.93 1,036 32.50 46.86 −14.391*** Mid-skilled occupation 937 48.98 50.02 1,036 46.47 49.90 −2.507 High-skilled occupation 937 4.13 19.90 1,036 21.03 40.77 16.898*** Economic sector (%) Agriculture, fishing, and mining 937 16.26 36.92 1,036 8.17 27.41 −8.085*** Manufacturing 937 10.83 31.09 1,036 9.64 29.53 −1.185 Low- to mid-value-added 937 66.05 47.38 1,036 56.46 49.60 −9.591*** High-value-added 937 6.87 25.31 1,036 25.73 43.74 18.861*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Differences in Mean 97 Table D.3  Difference in Mean of Those Answering Core Literacy Test and Those Who Did Not Answer It (continued) Y1: Did not answer Core Y2: Answered Core Obs Mean SD Obs Mean SD Difference: Y2 – Y1 Earnings   Monthly earnings 910 477.44 1,421.51 1,083 549.53 1,702.99 72.094 Socioemotional skills   Extraversion (score) 336 2.50 0.65 1,584 2.55 0.60 0.044   Missing extraversion 1,179 72.99 44.42 1,808 13.26 33.92 −59.736***   Conscientiousness (score) 333 3.11 0.59 1,579 3.22 0.58 0.103**   Missing conscientiousness 1,179 73.33 44.24 1,808 13.29 33.95 −60.045***   Openness (score) 327 2.90 0.67 1,581 3.13 0.57 0.228***   Missing openness 1,179 73.58 44.11 1,808 13.32 33.99 −60.264***   Emotional stability (score) 320 2.71 0.58 1,574 2.72 0.56 0.010   Missing stability 1,179 74.03 43.87 1,808 13.79 34.49 −60.237***   Agreeableness (score) 323 2.90 0.67 1,579 3.07 0.62 0.168***   Missing agreeableness 1,179 73.85 43.96 1,808 13.33 34.00 −60.526*** Self-reported skills Reading   Use reading skill 1,172 37.08 48.32 1,808 88.24 32.22 51.163***   Read with low intensity 444 73.13 44.38 1,579 39.89 48.98 −33.245***   Read with medium intensity 444 15.66 36.38 1,579 26.11 43.94 10.452***   Read with high intensity 444 11.21 31.58 1,579 34.00 47.39 22.792*** Writing   Use writing skill 1,172 34.27 47.48 1,808 84.00 36.67 49.732***   Write with low intensity 415 86.94 33.74 1,512 68.08 46.63 −18.858***   Write with medium intensity 415 8.39 27.76 1,512 17.30 37.84 8.906***   Write with high intensity 415 4.67 21.13 1,512 14.62 35.35 9.952*** Numeracy   Use numeracy skill 1,172 92.27 26.73 1,808 95.85 19.94 3.588***   Numeracy with low intensity 1,093 44.80 49.75 1,724 19.71 39.79 −25.084***   Numeracy with medium intensity 1,093 53.33 49.91 1,724 61.26 48.73 7.923***   Numeracy with high intensity 1,093 1.87 13.55 1,724 19.03 39.27 17.161*** Computer   Use computer skill 1,168 5.20 22.22 1,808 42.14 49.39 36.936***   Computer with low intensity 67 39.31 49.21 710 33.90 47.37 −5.418   Computer with medium intensity 67 17.97 38.69 710 17.65 38.15 −0.326   Computer with high intensity 67 42.71 49.84 710 48.46 50.01 5.744 Literacy assessment Reading Proficiency 1,179 90.66 1,808 165.16 74.498*** Note: NEET = not in employment, education, or training; Obs = number of observations. *** p < 0.01, ** p < 0.05, * p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 98 Differences in Mean Table D.4  Difference in Mean of Those Missing and Those English Illiterate Y1: Missing Y2: English illiterate Obs Mean SD Obs Mean SD Difference: Y2 – Y1 Female (%) 520 64.61 47.86 659 75.12 43.27 10.505*** Socioeconomic status (%) Low 503 24.30 42.93 656 36.07 48.06 11.774*** Middle 503 54.03 49.89 656 48.08 50.00 −5.946 High 503 21.67 41.24 656 15.84 36.54 −5.828** Age range (years) (%) 520 36.99 12.37 659 39.33 12.18 2.340** 15–19 520 5.38 22.59 659 2.43 15.41 −2.953** 20–24 520 11.12 31.47 659 7.48 26.33 −3.634* 25–34 520 30.36 46.03 659 28.61 45.23 −1.752 35–44 520 24.68 43.16 659 30.33 46.00 5.648* 45–64 520 28.46 45.17 659 31.15 46.35 2.691 Education (%) No education 520 17.72 38.22 659 72.14 44.87 54.422*** Primary education 520 13.80 34.52 659 13.41 34.10 −0.390 Junior high school 520 51.23 50.03 659 13.85 34.57 −37.375*** Senior high school 520 12.40 32.98 659 0.60 7.74 −11.793*** Tertiary education 520 4.86 21.53 659 0.00 0.00 −4.863*** Received early childhood education 504 61.33 48.75 651 30.96 46.27 −30.369*** Language spoken at work (%) Akan 430 85.02 35.73 570 70.99 45.42 −14.033*** Ewe 430 12.25 32.83 570 6.82 25.22 −5.437** Ga-Adangme 430 16.67 37.31 570 10.90 31.19 −5.771** Mole-Dagbani 430 5.44 22.71 570 17.46 38.00 12.020*** English 430 32.98 47.07 570 5.77 23.34 −27.205*** Others 430 7.76 26.79 570 21.33 41.00 13.571*** Language spoken at home (%) Akan 505 66.54 47.23 659 45.85 49.87 −20.685*** Ewe 505 11.32 31.71 659 6.68 24.98 −4.640** Ga-Adangme 505 8.60 28.06 659 3.70 18.89 −4.898*** Mole-Dagbani 505 4.78 21.35 659 18.43 38.80 13.652*** English 505 0.99 9.92 659 0.07 2.68 −0.921* Others 505 7.78 26.81 659 25.27 43.49 17.492*** Labor status (%) Employed 514 76.97 42.15 659 81.53 38.84 4.561 Unemployed 514 3.78 19.09 659 3.67 18.81 −0.111 NEET 514 9.71 29.64 659 12.64 33.25 2.925 Inactive 514 9.54 29.41 659 2.17 14.57 −7.375*** Employment status (%) Formal employee 398 12.34 32.93 539 0.76 8.68 −11.580*** Informal employee 398 17.34 37.91 539 13.93 34.66 −3.407 Self-employed 398 70.32 45.74 539 85.31 35.43 14.987*** Occupation (%) Low-skilled occupation 398 48.29 50.03 539 45.81 49.87 −2.480 table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Differences in Mean 99 Table D.4  Difference in Mean of Those Missing and Those English Illiterate (continued) Y1: Missing Y2: English illiterate Obs Mean SD Obs Mean SD Difference: Y2 – Y1 Mid-skilled occupation 398 45.18 49.83 539 51.91 50.01 6.732 High-skilled occupation 398 6.53 24.73 539 2.28 14.93 −4.252** Economic sector (%) Agriculture, fishing, and mining 398 10.25 30.37 539 20.89 40.69 10.634*** Manufacturing 398 12.63 33.26 539 9.43 29.26 −3.197 Low- to mid-value-added 398 63.77 48.13 539 67.81 46.76 4.044 High-value-added 398 13.35 34.06 539 1.87 13.56 −11.482*** Earnings Monthly earnings 391 564.52 1,390.53 519 409.79 1,442.84 −154.735 Socioemotional skills Extraversion (score) 308 2.51 0.66 28 2.38 0.51 −0.134 Missing extraversion 520 44.56 49.75 659 96.43 18.57 51.874*** Conscientiousness (score) 306 3.11 0.59 27 3.12 0.56 0.008 Missing conscientiousness 520 45.19 49.82 659 96.53 18.31 51.343*** Openness (score) 301 2.93 0.66 26 2.55 0.70 −0.374** Missing openness 520 45.67 49.86 659 96.59 18.16 50.919*** Emotional stability (score) 295 2.70 0.57 25 2.85 0.72 0.145 Missing stability 520 46.37 49.92 659 96.83 17.54 50.462*** Agreeableness (score) 297 2.93 0.66 26 2.55 0.69 −0.382*** Missing agreeableness 520 46.24 49.91 659 96.62 18.10 50.374*** Self-reported skills Reading Use reading skill 513 66.38 47.29 659 13.19 33.87 −53.189*** Read with low intensity 346 71.12 45.39 98 81.36 39.15 10.234* Read with medium intensity 346 16.17 36.87 98 13.57 34.43 −2.599 Read with high intensity 346 12.71 33.35 98 5.07 22.06 −7.635* Writing Use writing skill 513 61.79 48.64 659 11.83 32.32 −49.965*** Write with low intensity 325 86.35 34.39 90 89.48 30.85 3.138 Write with medium intensity 325 7.90 27.01 90 10.52 30.85 2.618 Write with high intensity 325 5.76 23.33 90 0.00 0.00 −5.756*** Numeracy Use numeracy skill 513 93.13 25.32 659 91.56 27.82 −1.569 Numeracy with low intensity 482 34.18 47.48 611 53.60 49.91 19.426*** Numeracy with medium intensity 482 61.98 48.59 611 46.16 49.89 −15.821*** Numeracy with high intensity 482 3.84 19.24 611 0.24 4.85 −3.605*** Computer Use computer skill 509 10.92 31.22 659 0.59 7.65 −10.332*** Computer with low intensity 64 38.87 49.13 3 50.79 61.23 11.913 Computer with medium intensity 64 18.66 39.27 3 0.00 0.00 −18.664*** Computer with high intensity 64 42.46 49.82 3 49.21 61.23 6.750 Literacy assessment Reading Proficiency 520 73.49 659 104.82 31.3294*** Note: NEET = not in employment, education, or training; Obs = number of observations. *** p < 0.01, ** p < 0.05, * p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 100 Differences in Mean Table D.5  Difference in Mean of Those Answering the Socioemotional Section and Those Who Did Not Y1: Did not answer Y2: Answered socioemotional socioemotional section section Obs Mean SD Obs Mean SD Difference: Y2 – Y1 Female (%) 1,045 73.60 44.10 1,930 49.32 50.01 −24.281*** Socioeconomic status (%) Low 1,038 35.69 47.93 1,922 15.44 36.15 −20.245*** Middle 1,038 48.58 50.00 1,922 60.27 48.95 11.692*** High 1,038 15.73 36.43 1,922 24.29 42.89 8.553*** Age range (years) (%) 1,045 37.79 12.02 1,930 30.41 12.75 −7.378*** 15–19 1,045 3.91 19.40 1,930 20.12 40.10 16.206*** 20–24 1,045 9.21 28.93 1,930 22.45 41.74 13.245*** 25–34 1,045 30.09 45.89 1,930 27.05 44.43 −3.046 35–44 1,045 28.49 45.16 1,930 13.98 34.68 −14.514*** 45–64 1,045 28.30 45.07 1,930 16.41 37.04 −11.891*** Education (%) No education 1,045 54.16 49.85 1,930 2.33 15.07 −51.833*** Primary education 1,045 16.41 37.06 1,930 12.66 33.26 −3.755** JHS education 1,045 27.35 44.60 1,930 40.34 49.07 12.986*** SHS education 1,045 2.05 14.18 1,930 29.87 45.78 27.819*** Tertiary education 1,045 0.03 1.63 1,930 14.81 35.53 14.783*** Received ECE 1,029 41.24 49.25 1,907 76.03 42.70 34.783*** Language spoken at work (%) Akan 907 80.06 39.98 1,436 79.52 40.37 −0.544 Ewe 907 6.17 24.08 1,436 8.97 28.59 2.799** Ga-Adangme 907 9.16 28.86 1,436 15.20 35.91 6.037*** Mole-Dagbani 907 11.22 31.57 1,436 5.18 22.18 −6.032*** English 907 10.17 30.25 1,436 60.39 48.93 50.213*** Others 907 14.96 35.68 1,436 9.44 29.24 −5.518*** Language spoken at home (%) Akan 1,045 60.78 48.85 1,927 63.73 48.09 2.952 Ewe 1,045 5.61 23.02 1,927 9.54 29.39 3.936*** Ga-Adangme 1,045 3.92 19.41 1,927 8.68 28.16 4.765*** Mole-Dagbani 1,045 11.84 32.32 1,927 6.49 24.65 −5.347*** English 1,045 0.04 2.06 1,927 1.66 12.79 1.621*** Others 1,045 17.81 38.28 1,927 9.89 29.85 −7.928*** Labor status (%) Employed 1,045 79.95 40.05 1,930 52.85 49.93 −27.108*** Unemployed 1,045 3.64 18.73 1,930 5.73 23.25 2.092** NEET 1,045 11.72 32.18 1,930 8.76 28.28 −2.960** Inactive 1,045 4.69 21.15 1,930 32.67 46.91 27.976*** Employment status (%) Formal employee 838 1.88 13.60 1,131 25.00 43.32 23.119*** Informal employee 838 14.64 35.37 1,131 24.00 42.73 9.362*** Self-employed 838 83.48 37.16 1,131 51.00 50.01 −32.481*** Occupation (%) Low-skilled occupation 838 48.34 50.00 1,131 32.17 46.73 −16.174*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Differences in Mean 101 Table D.5  Difference in Mean of Those Answering the Socioemotional Section and Those Who Did Not (continued) Y1: Did not answer Y2: Answered socioemotional socioemotional section section Obs Mean SD Obs Mean SD Difference: Y2 – Y1 Mid-skilled occupation 838 49.56 50.03 1,131 46.11 49.87 −3.458 High-skilled occupation 838 2.09 14.32 1,131 21.72 41.26 19.633*** Economic sector (%) Agriculture, fishing, and mining 838 19.71 39.80 1,131 5.72 23.23 −13.985*** Manufacturing 838 9.70 29.61 1,131 10.68 30.90 0.984 Low- to mid-value-added 838 67.58 46.83 1,131 55.75 49.69 −11.831*** High-value-added 838 3.01 17.10 1,131 27.84 44.84 24.832*** Earnings Monthly earnings 804 402.45 1,217.70 1,184 602.31 1,803.02 199.860** Self-reported skills Reading Don’t use read skill 1,040 72.62 44.61 1,901 8.85 28.41 −63.768*** Read with low intensity 1,040 21.96 41.42 1,901 37.50 48.43 15.542*** Read with medium intensity 1,040 3.59 18.60 1,901 23.44 42.37 19.854*** Read with high intensity 1,040 1.84 13.44 1,901 30.21 45.93 28.372*** Writing Don’t use writing skill 1,043 75.34 43.13 1,911 12.94 33.57 −62.401*** Write with low intensity 1,043 22.34 41.67 1,911 60.03 49.00 37.688*** Write with medium intensity 1,043 1.71 12.99 1,911 14.65 35.37 12.938*** Write with high intensity 1,043 0.61 7.77 1,911 12.38 32.95 11.776*** Numeracy Don’t use numeracy skill 1,045 6.93 25.41 1,930 4.74 21.25 −2.195* Numeracy with low intensity 1,045 43.97 49.66 1,930 18.28 38.66 −25.699*** Numeracy with medium intensity 1,045 48.68 50.01 1,930 58.73 49.24 10.055*** Numeracy with high intensity 1,045 0.42 6.44 1,930 18.26 38.64 17.840*** Computer Don’t use computer skill 1,044 98.23 13.18 1,897 58.63 49.26 −39.600*** Computer with low intensity 1,044 0.60 7.75 1,897 14.17 34.88 13.562*** Computer with medium intensity 1,044 0.00 0.00 1,897 7.48 26.32 7.482*** Computer with high intensity 1,044 1.16 10.73 1,897 19.72 39.80 18.555*** Literacy assessment Reading Component (average) Sentence missing (%) 1,045 77.44 41.82 1,930 15.99 36.66 −61.450*** Sentence correct answers (%) 218 19.77 18.41 1,590 43.60 12.33 23.827*** Passage correct answers (%) 218 14.50 17.11 1,590 41.20 12.38 26.704*** Vocabulary correct answers (%) 218 13.67 8.46 1,590 21.90 3.81 8.230*** Core Score Core Test 218 0.59 1.31 1,590 4.95 2.71 4.357*** Passed Core Test 847 2.15 14.50 1,620 75.00 43.31 72.856*** Literacy assessment Reading Proficiency 1,045 77.54 1,930 169.10 91.556*** Note: NEET = not in employment, education, or training; Obs = number of observations. *** p < 0.01, ** p < 0.05, * p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Appendix E Returns to Education and Skills Table E.1 Returns to Years of Education (Mincer Equation), Controlling for Skills No Literacy Job-specific controls Socioeconomic assessment Personality Use of skills skills All controls Years of education 0.1019*** 0.1019*** 0.0898*** 0.0956*** 0.0802*** 0.0843*** 0.0632*** (0.0111) (0.0111) (0.0180) (0.0135) (0.0135) (0.0126) (0.0168) Socioemotional skills (relative to score 2 or lower) Extraversion>2 0.0017 0.0228 (0.1460) (0.1387) Conscientiousness>2 0.4564** 0.2193 (0.1915) (0.2118) Openness>2 0.1216 0.0110 (0.1955) (0.1820) Stability>2 0.0430 0.0589 (0.2930) (0.2870) Agreeableness>2 −0.0304 0.0138 (0.1988) (0.1958) Grit>2 0.0479 0.0085 (0.1503) (0.1711) Cognitive skills (relative to skill not used) Reading Read with low intensity −0.2853 −0.2766* (0.1742) (0.1655) Read with medium −0.3427* −0.4154** intensity (0.2007) (0.2037) Read with high intensity −0.4598** −0.5218** (0.2241) (0.2185) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   103   104 Returns to Education and Skills Table E.1  Returns to Years of Education (Mincer Equation), Controlling for Skills (continued) No Literacy Job-specific controls Socioeconomic assessment Personality Use of skills skills All controls Writing Write with low intensity 0.4973*** 0.4204** (0.1578) (0.1746) Write with medium 0.8394*** 0.8296*** intensity (0.1913) (0.2183) Write with high 0.8989*** 0.8828*** intensity (0.2746) (0.3125) Numeracy Numeracy with low 0.0310 −0.0623 intensity (0.1996) (0.1991) Numeracy with 0.3047** 0.1953 medium intensity (0.1510) (0.1636) Numeracy with high 0.3670 0.2565 intensity (0.2665) (0.2658) Job-specific skills Cognitive challenge Low think and learn −0.0004 −0.0649 (0.1358) (0.1252) Medium think and learn 0.1657 0.1059 (0.1380) (0.1359) High think and learn 0.0314 0.0003 (0.1851) (0.1615) Physical Low physical demand −0.1321 −0.0895 (0.1782) (0.1969) Medium physical −0.0698 −0.0717 demand (0.1842) (0.1907) High physical demand 0.0529 0.1037 (0.1730) (0.1801) Autonomy and repetitiveness Low autonomy and 0.0838 0.0747 repetitive (0.1580) (0.1274) Medium autonomy 0.1405 0.1079 and repetitive (0.2023) (0.1703) High autonomy and −0.0986 −0.0269 repetitive (0.2574) (0.2419) Make presentations −0.0138 −0.1471 (0.2261) (0.2226) Supervise others 0.2121 0.2407** (0.1297) (0.1131) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 105 Table E.1  Returns to Years of Education (Mincer Equation), Controlling for Skills (continued) No Literacy Job-specific controls Socioeconomic assessment Personality Use of skills skills All controls Computer use at work (relative to no use) Computer work with 0.0272 −0.0526 low intensity (0.2801) (0.2717) Computer work with 0.4043 0.3359 medium intensity (0.2760) (0.2886) Computer work with 0.3084 0.2992 high intensity (0.2124) (0.1873) Reading proficiency Sentence correct −0.0040 −0.0063 answers (%) (0.0058) (0.0054) Passage correct 0.0116* 0.0096 answers (%) (0.0067) (0.0063) Vocabulary correct −0.0044 0.0072 answers (%) (0.0141) (0.0137) Passed Core test −0.0220 −0.0915 (0.2915) (0.2664) Failed Core test −0.0846 −0.1091 (0.2127) (0.1990) Number of observations 675 675 675 675 675 675 675 R2 0.228 0.228 0.239 0.235 0.283 0.265 0.329 Note: All models estimated using ordinary least squares. Excluding self-employed. Robust standard errors are in parentheses. Controls include experience, experience squared, gender, economic sector and dummies for region. For all skills dummies of nonresponse (missing) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. *** p < 0.01, ** p < 0.05, * p < 0.1. Table E.2 Returns to Education Level (Mincer Equation), Controlling for Skills No Literacy Use of Job-specific controls Socioeconomic assessment Personality skills skills All controls Education level (relative to none) Primary education (%) 0.3186* 0.3186* 0.2609 0.2408 0.1857 0.3176* 0.1388 (0.1834) (0.1834) (0.1989) (0.1937) (0.1683) (0.1829) (0.1872) Junior high school (%) 0.5921*** 0.5921*** 0.4815*** 0.4413*** 0.4197*** 0.5858*** 0.3525** (0.1548) (0.1548) (0.1640) (0.1537) (0.1487) (0.1558) (0.1523) Senior high school (%) 0.7882*** 0.7882*** 0.5513* 0.5791** 0.5097** 0.6962*** 0.2917 (0.2160) (0.2160) (0.2978) (0.2353) (0.2057) (0.2132) (0.2465) Tertiary education (%) 1.7174*** 1.7174*** 1.4883*** 1.4939*** 1.3415*** 1.5899*** 1.1183*** (0.1681) (0.1681) (0.2584) (0.2040) (0.2050) (0.2082) (0.2618) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 106 Returns to Education and Skills Table E.2  Returns to Education Level (Mincer Equation), Controlling for Skills (continued) No Literacy Use of Job-specific controls Socioeconomic assessment Personality skills skills All controls Socioemotional skills (relative to score 2 or lower) Extraversion>2 −0.0158 0.0338 (0.1370) (0.1327) Conscientiousness>2 0.4631** 0.2422 (0.1987) (0.2141) Openness>2 0.0709 −0.0060 (0.1968) (0.1836) Stability>2 0.0593 0.0872 (0.3046) (0.2903) Agreeableness>2 −0.0460 0.0049 (0.1975) (0.2018) Grit>2 0.0008 −0.0430 (0.1480) (0.1547) Cognitive skills (relative to skill not used) Reading Read with low intensity −0.2164 −0.2617* (0.1581) (0.1561) Read with medium −0.2754 −0.3666* intensity (0.1824) (0.1906) Read with high intensity −0.3958** −0.4742** (0.2013) (0.2002) Writing Write with low intensity 0.5616*** 0.4663*** (0.1515) (0.1691) Write with medium 0.8029*** 0.7854*** intensity (0.1921) (0.2151) Write with high 0.9096*** 0.8858*** intensity (0.2677) (0.2894) Numeracy Numeracy with low 0.0125 −0.0881 intensity (0.1921) (0.1912) Numeracy with 0.2467 0.1257 medium intensity (0.1598) (0.1702) Numeracy with high 0.1919 0.1478 intensity (0.2668) (0.2586) Job-specific skills (relative to score 0) Cognitive challenge Low think and learn 0.0172 −0.0554 (0.1349) (0.1217) Medium think and learn 0.1504 0.0822 (0.1377) (0.1345) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 107 Table E.2  Returns to Education Level (Mincer Equation), Controlling for Skills (continued) No Literacy Use of Job-specific controls Socioeconomic assessment Personality skills skills All controls High think and learn −0.0431 −0.0686 (0.1782) (0.1563) Physical Low physical demand −0.0731 −0.0208 (0.1816) (0.1943) Medium physical 0.0414 0.0441 demand (0.1878) (0.1879) High physical demand 0.1540 0.2231 (0.1677) (0.1737) Autonomy and repetitiveness Low autonomy and 0.0327 0.0274 repetitive (0.1379) (0.1149) Medium autonomy 0.0177 −0.0071 and repetitive (0.1740) (0.1538) High autonomy and −0.1011 −0.0259 repetitive (0.2393) (0.2214) Make presentations −0.0298 −0.1428 (0.2309) (0.2146) Supervise others 0.2123* 0.2320** (0.1259) (0.1084) Computer use at work (relative to no use) Computer use with low 0.0029 −0.0624 intensity (0.2832) (0.2744) Computer use with 0.2490 0.1699 medium intensity (0.2935) (0.3049) Computer use with 0.1793 0.1872 high intensity (0.1855) (0.1708) Reading proficiency Sentence correct −0.0072 −0.0088 answers (%) (0.0057) (0.0055) Passage correct 0.0137** 0.0117* answers (%) (0.0064) (0.0061) Vocabulary correct 0.0012 0.0099 answers (%) (0.0136) (0.0133) Passed Core test −0.0134 −0.0607 (0.2965) (0.2764) Failed Core test −0.0098 −0.0526 (0.2288) (0.2163) Number of observations 675 675 675 675 675 675 675 R2 0.270 0.270 0.286 0.281 0.320 0.298 0.364 Note: All models estimated using ordinary least squares. Excluding self-employed. Robust standard errors are in parentheses. Controls include experience, experience squared, gender, economic sector and dummies for region. For all skills dummies of nonresponse (missing) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. *** p < 0.01, ** p < 0.05, * p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 108 Returns to Education and Skills Table E.3 Returns to Years of Education (Mincer Equation) for Informal Wageworkers, Controlling for Skills Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Years of education 0.0591*** 0.0667*** 0.0513*** 0.0431*** 0.0508*** 0.0392*** (0.0132) (0.0163) (0.0157) (0.0142) (0.0132) (0.0153) Socioemotional skills (relative to score 2 or lower) Extraversion>2 0.2324* 0.2113** (0.1401) (0.1384) Conscientiousness>2 0.3148 0.1219 (0.2440) (0.2126) Openness>2 0.0927 0.1104 (0.2083) (0.2005) Stability>2 0.1037 0.0595 (0.2398) (0.2302) Agreeableness>2 0.0063 0.0704 (0.2609) (0.2514) Grit>2 (−0.1268) (−0.2334) (0.1511) (0.1461) Cognitive skills (relative to skill not used) Reading Read with low intensity −0.3320** −0.2426 (0.1696) (0.1685) Read with medium intensity −0.3557** −0.2863 (0.2048) (0.2059) Read with high intensity −0.3564* −0.3573 (0.2127) (0.2050) Writing Write with low intensity 0.5207*** 0.4269*** (0.1568) (0.1701) Write with medium intensity 0.6020*** 0.5452*** (0.2254) (0.2230) Write with high intensity 1.2538*** 1.2017*** (0.3060) (0.3337) Numeracy Numeracy with low intensity −0.0310 −0.2354* (0.1867) (0.1883) Numeracy with medium intensity 0.0421 −0.1644 (0.1753) (0.1915) Numeracy with high intensity −0.1045 −0.1206 (0.2778) (0.2866) Job-specific skills (relative to score 0) Cognitive challenge Low think and learn −0.0847 −0.1587 (0.1224) (0.1164) Medium think and learn 0.0201 −0.0215 (0.1491) (0.1535) High think and learn 0.1683 0.1503 (0.1781) (0.1705) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 109 Table E.3  Returns to Years of Education (Mincer Equation) for Informal Wageworkers, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Physical Low physical demand 0.0114 0.0691 (0.1843) (0.2108) Medium physical demand 0.1611 0.1819 (0.2098) (0.2256) High physical demand 0.4199 0.4646 (0.1748) (0.2112) Autonomy and repetitiveness Low autonomy and repetitive −0.0737 −0.0447 (0.1043) (0.1072) Medium autonomy and −0.1788 −0.1817 repetitive (0.1860) (0.1923) High autonomy and repetitive −0.3558 −0.1763 (0.1866) (0.2018) Make presentations −0.1404 −0.1978* (0.2063) (0.1868) Supervise others 0.3356* 0.3211** (0.1327) (0.1298) Computer use at work (relative to no use) Computer use with low intensity 0.4667 0.3215 (0.3951) (0.3338) Computer use with medium 0.2478 −0.1337 intensity (0.3652) (0.5208) Computer use with high intensity 0.1323** 0.0676* (0.2700) (0.2438) Reading proficiency Sentence correct answers (%) (−0.0037) (−0.0100) (0.0068) (0.0068) Passage correct answers (%) (0.0101) (0.0112) (0.0073) (0.0078) Vocabulary correct answers (%) (−0.0090) (0.0029) (0.0161) (0.0157) Passed Core test (−0.2260) (−0.2427) (0.3442) (0.3458) Failed Core test −0.0133 0.0380 (0.2371) (0.2401) Number of observations 387 387 387 387 387 387 R2 0.098 0.108 0.125 0.186 0.192 0.300 Note: All models estimated using ordinary least squares. Robust standard errors are in parentheses. Controls include experience, experience squared, gender, economic sector, and dummies for region. For all skills dummies of nonresponse (missing) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. *** p < 0.01, ** p < 0.05, * p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 110 Returns to Education and Skills Table E.4 Returns to Education Level (Mincer Equation) for Informal Wageworkers, Controlling for Skills Job-specific No controls Socioeconomic Literacy assessment Use of skills skills All controls Education level (relative to none) Primary education (%) 0.2856 0.2883 0.2390 0.2093 0.3256* 0.1215 (0.1899) (0.1945) (0.2090) (0.1734) (0.1895) (0.2034) Junior high school (%) 0.3940*** 0.4268*** 0.3244* 0.3360** 0.3970*** 0.2673* (0.1469) (0.1499) (0.1716) (0.1590) (0.1436) (0.1585) Senior high school (%) 0.6408*** 0.7543*** 0.5273** 0.5258*** 0.5829*** 0.4845** (0.1699) (0.2173) (0.2092) (0.1920) (0.1789) (0.2154) Tertiary education (%) 1.1907*** 1.3209*** 1.0638*** 0.8967*** 1.1197*** 0.9016*** (0.3063) (0.3383) (0.3128) (0.2973) (0.2869) (0.2699) Socioemotional skills (relative to score 2 or lower) Extraversion>2 0.2066 0.1925 (0.1347) (0.1365) Conscientiousness>2 0.3094 0.1052 (0.2498) (0.2156) Openness>2 0.0463 0.0944 (0.2094) (0.2016) Stability>2 0.1368 0.0888 (0.2640) (0.2385) Agreeableness>2 −0.0205 0.0504 (0.2606) (0.2586) Grit>2 −0.1343 −0.2493* (0.1531) (0.1469) Cognitive skills (relative to skill not used) Reading Read with low intensity −0.3387** −0.2445 (0.1611) (0.1637) Read with medium −0.3869* −0.2993 intensity (0.2032) (0.2064) Read with high intensity −0.4016* −0.3925* (0.2137) (0.2040) Writing Write with low intensity 0.5317*** 0.4376*** (0.1556) (0.1669) Write with medium 0.6135*** 0.5278** intensity (0.2270) (0.2191) Write with high intensity 1.1775*** 1.1263*** (0.3292) (0.3317) Numeracy Numeracy with low −0.0352 −0.2713 intensity (0.1904) (0.1874) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 111 Table E.4  Returns to Education Level (Mincer Equation) for Informal Wageworkers, Controlling for Skills (continued) Job-specific No controls Socioeconomic Literacy assessment Use of skills skills All controls Numeracy with medium 0.0155 −0.2245 intensity (0.1785) (0.1911) Numeracy with high −0.1288 −0.1731 intensity (0.2807) (0.2806) Job-specific skills (relative to score 0) Cognitive challenge Low think and learn −0.0825 −0.1535 (0.1193) (0.1146) Medium think and learn 0.0005 −0.0370 (0.1493) (0.1548) High think and learn 0.1528 0.1491 (0.1736) (0.1663) Physical Low physical demand 0.0552 0.1066 (0.1880) (0.2030) Medium physical demand 0.2426 0.2305 (0.2072) (0.2161) High physical demand 0.4886*** 0.5131** (0.1771) (0.2024) Autonomy and repetitiveness Low autonomy and −0.0889 −0.0595 repetitive (0.1035) (0.1039) Medium autonomy and −0.1960 −0.1992 repetitive (0.1909) (0.1969) High autonomy and −0.3438* −0.1597 repetitive (0.1905) (0.2027) Make presentations −0.1499 −0.2023 (0.2114) (0.1891) Supervise others 0.3106** 0.3119** (0.1278) (0.1301) Computer use at work (relative to no use) Computer use with low 0.3657 0.2876 intensity (0.3604) (0.3321) Computer use with 0.1611 −0.2094 medium intensity (0.4063) (0.5478) Computer use with high 0.0307 −0.0090 intensity (0.2671) (0.2430) Reading proficiency Sentence correct −0.0038 −0.0104 answers (%) (0.0068) (0.0069) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 112 Returns to Education and Skills Table E.4  Returns to Education Level (Mincer Equation) for Informal Wageworkers, Controlling for Skills (continued) Job-specific No controls Socioeconomic Literacy assessment Use of skills skills All controls Passage correct 0.0086 0.0110 answers (%) (0.0072) (0.0079) Vocabulary correct −0.0046 0.0051 answers (%) (0.0159) (0.0158) Passed Core test −0.2775 −0.2738 (0.3481) (0.3502) Failed Core test 0.0203 0.0681 (0.2392) (0.2489) Number of observations 387 387 387 387 387 387 R2 0.116 0.126 0.142 0.194 0.207 0.310 Note: All models estimated using ordinary least squares. Robust standard errors are in parentheses. Controls include experience, experience squared, gender, economic sector and dummies for region. For all skills dummies of non-response (missing) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. JHS = junior high school; SHS = senior high school. *** p < 0.01, ** p < 0.05, * p < 0.1. Table E.5 Returns to Years of Education (Mincer Equation) for Formal Wageworkers, Controlling for Skills Job-specific No controls Socioeconomic Literacy assessment Use of skills skills All controls Years of education 0.0944*** 0.0749** 0.1025*** 0.0918*** 0.0946*** 0.0914*** (0.0273) (0.0359) (0.0268) (0.0319) (0.0353) (0.0320) Socioemotional skills (relative to score 2 or lower) Extraversion>2 −0.3074* −0.3692* (0.1859) (0.2155) Conscientiousness>2 0.6214* 0.3945 (0.3329) (0.4364) Openness>2 −0.1851 −0.1046 (0.2369) (0.2385) Stability>2 −0.1854 −0.0490 (0.4990) (0.4707) Agreeableness>2 −0.2409 −0.3834 (0.2402) (0.2700) Grit>2 0.3177 0.3255 (0.2913) (0.2965) Cognitive skills (relative to skill not used) Reading Read with low intensity −0.0587 −0.2557 (0.3558) (0.3472) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 113 Table E.5  Returns to Years of Education (Mincer Equation) for Formal Wageworkers, Controlling for Skills (continued) Job-specific No controls Socioeconomic Literacy assessment Use of skills skills All controls Read with medium −0.1809 −0.4497 intensity (0.3793) (0.4015) Read with high intensity −0.3465 −0.5582 (0.3646) (0.3905) Writing Write with low intensity 0.2104 −0.0757 (0.2699) (0.3083) Write with medium 0.4048 0.1047 intensity (0.2860) (0.3381) Write with high intensity 0.3263 0.1177 (0.3529) (0.3707) Numeracy Numeracy with low −0.0341 0.0043 intensity (0.3980) (0.3314) Numeracy with medium 0.5189** 0.3861 intensity (0.2194) (0.2554) Numeracy with high 0.3131 0.2708 intensity (0.3604) (0.3435) Job-specific skills (relative to score 0) Cognitive challenge Low think and learn 0.1443 0.2795 (0.2725) (0.2631) Medium think and learn 0.2753 0.4169* (0.2501) (0.2508) High think and learn −0.2385 0.0778 (0.3188) (0.2652) Physical Low physical demand 0.0282 −0.1511 (0.2477) (0.2695) Medium physical −0.0172 −0.2942 demand (0.2335) (0.2392) High physical demand −0.1174 −0.2737 (0.2795) (0.2728) Autonomy and repetitiveness Low autonomy and 0.2523 0.2650 repetitive (0.2804) (0.2098) Medium autonomy and 0.2755 0.2476 repetitive (0.3134) (0.2647) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 114 Returns to Education and Skills Table E.5  Returns to Years of Education (Mincer Equation) for Formal Wageworkers, Controlling for Skills (continued) Job-specific No controls Socioeconomic Literacy assessment Use of skills skills All controls High autonomy and 0.7046 0.9453** repetitive (0.4544) (0.4556) Make presentations −0.0052 −0.0519 (0.3045) (0.2646) Supervise others 0.0980 0.1014 (0.1961) (0.1496) Computer use at work (relative to no use) Computer use with low −0.2233 −0.1649 intensity (0.3276) (0.3313) Computer use with 0.4662** 0.5208** medium intensity (0.1962) (0.2286) Computer use with high 0.1477 0.0986 intensity (0.2220) (0.2147) Reading proficiency Sentence correct −0.0115 −0.0108 answers (%) (0.0079) (0.0084) Passage correct answers 0.0107 0.0048 (%) (0.0089) (0.0099) Vocabulary correct 0.0043 0.0181 answers (%) (0.0175) (0.0193) Passed Core test 0.4454 0.2707 (0.4625) (0.4119) Failed Core test 0.0791 −0.1837 (0.4179) (0.3463) Number of observations 288 288 288 288 288 288 R2 0.093 0.134 0.125 0.158 0.170 0.277 Note: All models estimated using ordinary least squares. Robust standard errors are in parentheses. Controls include experience, experience squared, gender, economic sector, and dummies for region. For all skills dummies of nonresponse (missing) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. *** p < 0.01, ** p < 0.05, * p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 115 Table E.6 Returns to Education Level (Mincer Equation) for Formal Wageworkers, Controlling for Skills Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Education level (relative to none) Primary education (%) 0.2928 0.0374 0.2616 −0.0484 0.4489 0.1584 (0.4344) (0.6760) (0.4813) (0.4963) (0.5645) (0.6628) JHS education (%) 0.4450 0.1683 0.2855 0.2182 0.4958 0.3042 (0.3439) (0.3831) (0.3214) (0.3345) (0.3275) (0.3958) SHS education (%) 0.2360 −0.1607 0.1456 −0.0278 0.2834 0.1042 (0.4336) (0.5481) (0.4459) (0.4512) (0.3710) (0.4470) Tertiary education (%) 1.1384*** 0.7088 1.0394*** 0.8243* 1.2545*** 0.9646** (0.3398) (0.4805) (0.3549) (0.4251) (0.3818) (0.4473) Socioemotional skills (relative to score 2 or lower) Extraversion>2 −0.2490 −0.2514 (0.1744) (0.1946) Conscientiousness>2 0.4825* 0.2224 (0.2737) (0.3698) Openness>2 −0.3270 −0.2816 (0.2264) (0.2647) Stability>2 −0.1719 −0.0512 (0.4698) (0.4299) Agreeableness>2 −0.1443 −0.2504 (0.2345) (0.2619) Grit>2 0.2738 0.2488 (0.2522) (0.2557) Cognitive Skills (relative to skill not used) Reading Read with low intensity 0.1837 −0.1505 (0.3971) (0.3778) Read with medium 0.1256 −0.2512 intensity (0.4144) (0.4191) Read with high −0.0101 −0.3388 intensity (0.3943) (0.4239) Writing Write with low 0.2481 0.0166 intensity (0.2696) (0.2880) Write with medium 0.3944 0.1636 intensity (0.2878) (0.3235) Write with high 0.4449 0.2797 intensity (0.3438) (0.3437) Numeracy Numeracy with low 0.0016 0.0034 intensity (0.3639) (0.3262) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 116 Returns to Education and Skills Table E.6  Returns to Education Level (Mincer Equation) for Formal Wageworkers, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Numeracy with 0.5156** 0.3465 medium intensity (0.2395) (0.2675) Numeracy with high 0.2631 0.2520 intensity (0.3622) (0.3446) Job-specific skills (relative to score 0) Cognitive challenge Low think and learn 0.2136 0.2547 (0.2573) (0.2456) Medium think and 0.2371 0.3076 learn (0.2395) (0.2403) High think and learn −0.3254 −0.0825 (0.2974) (0.2671) Physical Low physical demand 0.1116 −0.0010 (0.2456) (0.2631) Medium physical 0.1269 −0.1099 demand (0.2557) (0.2496) High physical demand 0.0359 −0.0434 (0.2729) (0.2704) Autonomy and repetitiveness Low autonomy and 0.1599 0.1615 repetitive (0.2377) (0.1891) Medium autonomy 0.0943 0.0690 and repetitive (0.2618) (0.2359) High autonomy and 0.5873 0.7399* repetitive (0.3939) (0.3807) Make presentations 0.0141 −0.0667 (0.2859) (0.2522) Supervise others 0.1955 0.1946 (0.1844) (0.1543) Computer use at work (relative to no use) Computer use with low −0.1210 −0.1003 intensity (0.3586) (0.3482) Computer use with 0.2826 0.3083 medium intensity (0.1787) (0.1962) Computer use with 0.0900 0.0659 high intensity (0.1963) (0.1986) Reading proficiency Sentence correct −0.0129 −0.0117 answers (%) (0.0086) (0.0081) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 117 Table E.6  Returns to Education Level (Mincer Equation) for Formal Wageworkers, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Passage correct 0.0183* 0.0103 answers (%) (0.0101) (0.0096) Vocabulary correct −0.0015 0.0138 answers (%) (0.0165) (0.0178) Passed Core test 0.3477 0.2030 (0.4756) (0.4121) Failed Core test 0.0700 −0.1231 (0.4205) (0.3474) Number of observations 288 288 288 288 288 288 R2 0.157 0.208 0.182 0.216 0.236 0.322 Note: All models estimated using ordinary least squares. Robust standard errors are in parentheses. Controls include experience, experience squared, gender, economic sector, and dummies for region. For all skills dummies of nonresponse (missing) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. *** p < 0.01, ** p < 0.05, * p < 0.1. Table E.7 Returns to Years of Education (Mincer Equation) for Male Workers, Controlling for Skills Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Years of education 0.0821*** 0.0686*** 0.0944*** 0.0806*** 0.0662*** 0.0624*** (0.0156) (0.0258) (0.0174) (0.0173) (0.0172) (0.0220) Socioemotional skills (relative to score 2 or lower) Extraversion>2 −0.0059 0.0422 (0.1703) (0.1637) Conscientiousness>2 0.4383* −0.0224 (0.2357) (0.3114) Openness>2 −0.0065 −0.0437 (0.1801) (0.1831) Stability>2 −0.1271 −0.2001 (0.3880) (0.3785) Agreeableness>2 −0.2579 −0.2465 (0.1941) (0.2005) Grit>2 0.1604 0.1718 (0.2199) (0.2288) Cognitive skills (relative to skill not used) Reading Read with low intensity −0.3859* −0.3590 (0.2260) (0.2206) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 118 Returns to Education and Skills Table E.7  Returns to Years of Education (Mincer Equation) for Male Workers, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Read with medium −0.5224** −0.5639** intensity (0.2427) (0.2752) Read with high intensity −0.6218** −0.5933** (0.2819) (0.2924) Writing Write with low intensity 0.4664** 0.5645** (0.2058) (0.2307) Write with medium 0.6412** 0.7750** intensity (0.2591) (0.3042) Write with high intensity 0.8619** 0.9770** (0.3523) (0.3977) Numeracy Numeracy with low 0.2629 0.2190 intensity (0.2687) (0.2606) Numeracy with medium 0.4744** 0.3842* intensity (0.1892) (0.2289) Numeracy with high 0.4698 0.2906 intensity (0.3668) (0.3631) Job-specific skills (relative to score 0) Cognitive challenge Low think and learn −0.0341 −0.0369 (0.1921) (0.1693) Medium think and learn 0.1034 0.1151 (0.1757) (0.1699) High think and learn −0.1847 −0.1343 (0.2354) (0.2000) Physical Low physical demand −0.1036 −0.1291 (0.2332) (0.2360) Medium physical −0.0518 −0.1612 demand (0.2431) (0.2422) High physical demand −0.0943 −0.1206 (0.2201) (0.2239) Autonomy and repetitiveness Low autonomy and 0.1612 0.1707 repetitive (0.2188) (0.1815) Medium autonomy and 0.2790 0.2713 repetitive (0.2492) (0.2115) High autonomy and 0.0510 0.1784 repetitive (0.3501) (0.3418) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 119 Table E.7  Returns to Years of Education (Mincer Equation) for Male Workers, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Make presentations −0.0657 −0.1311 (0.2940) (0.2691) Supervise others 0.2440 0.2224* (0.1547) (0.1307) Computer use at work (relative to no use) Computer use with low 0.1596 0.1160 intensity (0.3874) (0.3717) Computer use with 0.5847** 0.4686 medium intensity (0.2976) (0.3126) Computer use with high 0.3244 0.3399 intensity (0.2750) (0.2454) Reading proficiency Sentence correct −0.0068 −0.0074 answers (%) (0.0064) (0.0064) Passage correct answers 0.0091 0.0070 (%) (0.0075) (0.0074) Vocabulary correct 0.0064 0.0197 answers (%) (0.0168) (0.0156) Passed Core test 0.0079 −0.0787 (0.3671) (0.3641) Failed Core test −0.0972 −0.2103 (0.2759) (0.2484) Number of observations 429 429 429 429 429 429 R2 0.142 0.154 0.157 0.195 0.194 0.271 Note: All models estimated using ordinary least squares. Robust standard errors are in parentheses. Controls include experience, experience squared, gender, economic sector, and dummies for region. For all skills dummies of nonresponse (missing ) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. *** p < 0.01, ** p < 0.05, * p < 0.1. Table E.8 Returns to Education Level (Mincer Equation) for Male Workers, Controlling for Skills Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Education level (relative to none) Primary education (%) 0.0169 −0.0538 0.0343 −0.0514 0.0684 0.0198 (0.2009) (0.2217) (0.2200) (0.2000) (0.2066) (0.2194) Junior high school (%) 0.3425** 0.1980 0.3531** 0.3072* 0.3531** 0.2752 (0.1645) (0.1802) (0.1716) (0.1788) (0.1687) (0.1900) Senior high school (%) 0.3803 0.1119 0.4190 0.3132 0.3598 0.1930 (0.2696) (0.3837) (0.2924) (0.2594) (0.2601) (0.3096) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 120 Returns to Education and Skills Table E.8  Returns to Education Level (Mincer Equation) for Male Workers, Controlling for Skills (continued) Literacy Job−specific No controls Socioeconomic assessment Use of skills skills All controls Tertiary education (%) 1.3544*** 1.1057*** 1.3990*** 1.3153*** 1.3461*** 1.1786*** (0.1928) (0.3225) (0.2354) (0.2526) (0.2648) (0.3298) Socioemotional skills (relative to score 2 or lower) Extraversion>2 0.0021 0.0786 (0.1588) (0.1556) Conscientiousness>2 0.3845* −0.0435 (0.2291) (0.2742) Openness>2 −0.0416 −0.0800 (0.1734) (0.1782) Stability>2 −0.1689 −0.2126 (0.3825) (0.3580) Agreeableness>2 −0.2124 −0.2296 (0.1873) (0.1970) Grit>2 0.0783 0.0887 (0.2118) (0.2013) Cognitive skills (relative to skill not used) Reading Read with low intensity −0.2334 −0.2992 (0.2209) (0.2153) Read with medium −0.3417 −0.4566* intensity (0.2295) (0.2540) Read with high −0.4847* −0.5189* intensity (0.2579) (0.2686) Writing Write with low 0.4947** 0.5589** intensity (0.2058) (0.2312) Write with medium 0.5499** 0.6607** intensity (0.2516) (0.2985) Write with high 0.8401** 0.9446** intensity (0.3374) (0.3696) Numeracy Numeracy with low 0.2834 0.2168 intensity (0.2392) (0.2483) Numeracy with 0.4737** 0.3465 medium intensity (0.2014) (0.2385) Numeracy with high 0.2603 0.1549 intensity (0.3515) (0.3530) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 121 Table E.8  Returns to Education Level (Mincer Equation) for Male Workers, Controlling for Skills (continued) Literacy Job−specific No controls Socioeconomic assessment Use of skills skills All controls Job-specific skills (relative to score 0) Cognitive challenge Low think and learn −0.0205 −0.0157 (0.1968) (0.1684) Medium think and 0.1175 0.1261 learn (0.1746) (0.1655) High think and learn −0.2634 −0.1828 (0.2307) (0.1932) Physical Low physical demand 0.0281 0.0124 (0.2289) (0.2245) Medium physical 0.1739 0.0532 demand (0.2480) (0.2398) High physical demand 0.1060 0.0863 (0.2077) (0.2173) Autonomy and repetitiveness Low autonomy and 0.1165 0.1432 repetitive (0.1886) (0.1605) Medium autonomy 0.1580 0.1665 and repetitive (0.2116) (0.1890) High autonomy and 0.0369 0.1470 repetitive (0.3103) (0.2957) Make presentations −0.0655 −0.0975 (0.2923) (0.2533) Supervise others 0.2130 0.1847 (0.1505) (0.1231) Computer use at work (relative to no use) Computer use with low 0.0905 0.0623 intensity (0.3887) (0.3746) Computer use with 0.3364 0.2039 medium intensity (0.3220) (0.3327) Computer use with 0.1473 0.1806 high intensity (0.2261) (0.2148) Reading proficiency Sentence correct −0.0112 −0.0097 answers (%) (0.0069) (0.0067) Passage correct 0.0149* 0.0106 answers (%) (0.0078) (0.0074) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 122 Returns to Education and Skills Table E.8  Returns to Education Level (Mincer Equation) for Male Workers, Controlling for Skills (continued) Literacy Job−specific No controls Socioeconomic assessment Use of skills skills All controls Vocabulary correct 0.0092 0.0208 answers (%) (0.0155) (0.0148) Passed Core test −0.0174 −0.1081 (0.3760) (0.3545) Failed Core test −0.0572 −0.1977 (0.2890) (0.2551) Number of 429 429 429 429 429 429 observations R2 0.205 0.228 0.214 0.256 0.243 0.319 Note: All models estimated using ordinary least squares. Robust standard errors are in parentheses. Controls include experience, experience squared, gender, economic sector, and dummies for region. For all skills dummies of nonresponse (missing) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. JHS = junior high school; SHS = senior high school. *** p < 0.01, ** p < 0.05, * p < 0.1 Table E.9 Returns to Years of Education (Mincer Equation) for Female Workers, Controlling for Skills Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Years of education 0.1248*** 0.1260*** 0.1092*** 0.0939*** 0.1103*** 0.0886*** (0.0122) (0.0173) (0.0183) (0.0180) (0.0170) (0.0204) Socioemotional skills (relative to score 2 or lower) Extraversion>2 0.0796 0.0980 (0.1718) (0.1541) Conscientiousness>2 −0.0219 −0.2704 (0.3320) (0.3749) Openness>2 −0.0474 −0.4414 (0.3065) (0.3278) Stability>2 0.3484 0.5365* (0.3328) (0.3052) Agreeableness>2 0.6757* 0.7056* (0.3852) (0.3919) Grit>2 −0.1010 −0.3204* (0.1875) (0.1941) Cognitive skills (relative to skill not used) Reading Read with low −0.0896 −0.1866 intensity (0.2482) (0.2422) Read with medium −0.0275 −0.0439 intensity (0.3582) (0.3606) Read with high −0.1740 −0.2822 intensity (0.3050) (0.2868) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 123 Table E.9  Returns to Years of Education (Mincer Equation) for Female Workers, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Writing Write with low 0.2225 0.1639 intensity (0.2531) (0.2464) Write with medium 0.8702*** 0.8462*** intensity (0.2724) (0.2701) Write with high 0.5741 0.4050 intensity (0.3906) (0.3540) Numeracy Numeracy with low −0.2961 −0.3563* intensity (0.1958) (0.1955) Numeracy with −0.0280 −0.0250 medium intensity (0.1865) (0.1906) Numeracy with high 0.1788 0.2868 intensity (0.2565) (0.2959) Job-specific skills (relative to score 0) Cognitive challenge Low think and learn −0.0188 −0.0368 (0.1904) (0.1543) Medium think and 0.1362 0.0164 learn (0.1929) (0.1665) High think and learn 0.2075 0.1928 (0.2390) (0.2169) Physical Low physical demand −0.1409 −0.1024 (0.2323) (0.2868) Medium physical −0.2135 −0.1550 demand (0.2664) (0.2870) High physical 0.1024 0.1532 demand (0.2267) (0.2643) Autonomy and repetitiveness Low autonomy and −0.0378 0.0739 repetitive (0.1500) (0.1250) Medium autonomy −0.1028 −0.1493 and repetitive (0.2240) (0.2157) High autonomy and −0.3109 −0.1523 repetitive (0.2452) (0.3167) Make presentations 0.1021 −0.0914 (0.2217) (0.1969) Supervise others 0.0606 0.2123 (0.2088) (0.1690) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 124 Returns to Education and Skills Table E.9  Returns to Years of Education (Mincer Equation) for Female Workers, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Computer use at work (relative to no use) Computer use with −0.1790 −0.3520 low intensity (0.3608) (0.4408) Computer use with −0.5476 −0.6718 medium intensity (0.4123) (0.4293) Computer use with high intensity 0.2913 0.3859* (0.2810) (0.2111) Reading proficiency Sentence correct −0.0073 −0.0126 answers (%) (0.0098) (0.0084) Passage correct 0.0110 0.0120 answers (%) (0.0102) (0.0104) Vocabulary correct −0.0088 −0.0052 answers (%) (0.0202) (0.0190) Passed Core test −0.0018 0.0561 (0.4059) (0.3695) Failed Core test −0.0935 0.0858 (0.3014) (0.3175) Number of observations 246 246 246 246 246 246 R2 0.401 0.413 0.445 0.462 0.439 0.559 Note: All models estimated using ordinary least squares. Robust standard errors are in parentheses. Controls include experience, experience squared, gender, economic sector, and dummies for region. For all skills dummies of nonresponse (missing) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. *** p < 0.01, ** p < 0.05, * p < 0.1. Table E.10 Returns to Education Level (Mincer Equation) for Female Workers, Controlling for Skills Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Education Level (relative to none) Primary education (%) 0.2825 0.2438 0.1458 0.3042 0.2806 0.0419 (0.2219) (0.2306) (0.2219) (0.2485) (0.2328) (0.2410) Junior high school (%) 0.5477*** 0.5532** 0.4536** 0.4565** 0.5437*** 0.4290* (0.2013) (0.2158) (0.2198) (0.2105) (0.2003) (0.2272) Senior high school (%) 1.0766*** 1.1053*** 0.9449*** 0.8253*** 1.0569*** 0.7940*** (0.2110) (0.2679) (0.2714) (0.2605) (0.2487) (0.3078) Tertiary education (%) 1.9808*** 2.0489*** 1.8514*** 1.6017*** 1.9526*** 1.6394*** (0.2128) (0.2928) (0.2755) (0.2955) (0.2808) (0.3437) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 125 Table E.10  Returns to Education Level (Mincer Equation) for Female Workers, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Socioemotional skills (relative to score 2 or lower) Extraversion>2 0.0386 0.1002 (0.1643) (0.1574) Conscientiousness>2 −0.2087 −0.4056 (0.3219) (0.3695) Openness>2 −0.2421 −0.4383 (0.2903) (0.3224) Stability>2 0.4671 0.6062* (0.3625) (0.3330) Agreeableness>2 0.5900 0.6056 (0.3774) (0.3996) Grit>2 −0.0939 −0.3274* (0.1880) (0.1909) Cognitive skills (relative to skill not used) Reading Read with low intensity −0.1334 −0.2002 (0.1876) (0.2046) Read with medium −0.1886 −0.2094 intensity (0.3042) (0.3423) Read with high −0.2695 −0.3422 intensity (0.2808) (0.2672) Writing Write with low 0.3236 0.2774 intensity (0.2136) (0.2065) Write with medium 0.8419*** 0.8777*** intensity (0.2622) (0.2356) Write with high 0.5962 0.4322 intensity (0.4008) (0.3281) Numeracy Numeracy with low −0.3490* −0.4421** intensity (0.1895) (0.1827) Numeracy with −0.1930 −0.1827 medium intensity (0.1857) (0.1776) Numeracy with high −0.0451 0.1409 intensity (0.2632) (0.2830) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 126 Returns to Education and Skills Table E.10  Returns to Education Level (Mincer Equation) for Female Workers, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Job-specific skills (relative to score 0) Cognitive challenge Low think and learn −0.0642 −0.0479 (0.1707) (0.1434) Medium think and −0.0388 −0.0624 learn (0.1736) (0.1660) High think and learn 0.0030 0.0947 (0.2210) (0.2173) Physical Low physical demand −0.0583 −0.1043 (0.2353) (0.2800) Medium physical −0.1113 −0.1502 demand (0.2470) (0.2757) High physical demand 0.1334 0.1088 (0.2232) (0.2528) Autonomy and repetitiveness Low autonomy and −0.0761 0.0210 repetitive (0.1450) (0.1274) Medium autonomy −0.2425 −0.2699 and repetitive (0.2162) (0.2214) High autonomy and −0.2488 −0.1428 repetitive (0.2571) (0.3168) Make presentations 0.0234 −0.1881 (0.2496) (0.2210) Supervise others 0.0740 0.2096 (0.1935) (0.1548) Computer use at work (relative to no use) Computer use with low −0.2007 −0.3515 intensity (0.3747) (0.4651) Computer use with −0.6498 −0.9862* medium intensity (0.6101) (0.5453) Computer use with 0.1299 0.3077 high intensity (0.2855) (0.2136) Reading proficiency Sentence correct −0.0082 −0.0148* answers (%) (0.0091) (0.0082) Passage correct 0.0077 0.0119 answers (%) (0.0095) (0.0096) Vocabulary correct −0.0047 −0.0075 answers (%) (0.0211) (0.0191) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 127 Table E.10  Returns to Education Level (Mincer Equation) for Female Workers, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Passed Core test 0.0648 0.1867 (0.4226) (0.3835) Failed Core test 0.2063 0.3686 (0.3268) (0.3355) Number of 246 246 246 246 246 246 observations R2 0.457 0.462 0.500 0.496 0.481 0.590 Note: All models estimated using ordinary least squares. Robust standard errors are in parentheses. Controls include experience, experience squared, gender, economic sector, and dummies for region. For all skills dummies of nonresponse (missing) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. *** p < 0.01, ** p < 0.05, * p < 0.1. Table E.11 Linear Probability Model of Self-Employment, Controlling for Skills Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Education Level (relative to none) Primary education (%) −0.0803** 0.0051 0.0135 −0.0003 0.0064 0.0187 (0.0381) (0.0347) (0.0350) (0.0350) (0.0334) (0.0338) Junior high school (%) −0.1164*** 0.0073 0.0244 −0.0002 −0.0036 0.0232 (0.0264) (0.0266) (0.0302) (0.0310) (0.0243) (0.0298) Senior high school (%) −0.3106*** −0.0411 −0.0114 −0.0560 −0.0345 0.0074 (0.0333) (0.0345) (0.0406) (0.0398) (0.0318) (0.0393) Tertiary education (%) −0.6318*** −0.2107*** −0.1804*** −0.2207*** −0.2283*** −0.1812*** (0.0328) (0.0387) (0.0452) (0.0468) (0.0409) (0.0488) Socioemotional Skills (relative to score 2 or lower) Extraversion score >2 −0.0148 0.0003 (0.0277) (0.0254) Conscientiousness >2 −0.0233 0.0023 (0.0653) (0.0571) Openness>2 0.0070 0.0102 (0.0453) (0.0408) Stability>2 −0.0277 −0.0396 (0.0367) (0.0348) Agreeableness>2 0.0101 −0.0100 (0.0406) (0.0395) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 128 Returns to Education and Skills Table E.11  Linear Probability Model of Self-Employment, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls Cognitive skills (relative to skill not used) Reading Read with low intensity −0.0150 0.0078 (0.0316) (0.0296) Read with medium −0.0033 0.0119 intensity (0.0403) (0.0375) Read with high intensity −0.0039 0.0300 (0.0426) (0.0388) Writing Write with low intensity 0.0395 0.0521* (0.0290) (0.0268) Write with medium 0.0104 0.0139 intensity (0.0430) (0.0401) Write with high intensity −0.0137 −0.0030 (0.0561) (0.0528) Numeracy Numeracy with low 0.1551*** 0.1147*** intensity (0.0424) (0.0409) Numeracy with medium 0.1625*** 0.1134*** intensity (0.0412) (0.0398) Numeracy with high 0.1508** 0.1005* intensity (0.0625) (0.0591) Cognitive challenge Low think and learn 0.0471** 0.0483** (0.0220) (0.0222) Medium think and learn 0.0598** 0.0603** (0.0247) (0.0247) High think and learn 0.1071*** 0.1085*** (0.0304) (0.0307) Physical Low physical demand 0.0146 0.0156 (0.0388) (0.0380) Medium physical demand 0.0260 0.0236 (0.0394) (0.0386) High physical demand −0.0080 −0.0118 (0.0396) (0.0387) Autonomy and repetitiveness Low autonomy and 0.2061*** 0.2105*** repetitive (0.0314) (0.0307) Medium autonomy and 0.4185*** 0.4231*** repetitive (0.0323) (0.0314) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Returns to Education and Skills 129 Table E.11  Linear Probability Model of Self-Employment, Controlling for Skills (continued) Literacy Job-specific No controls Socioeconomic assessment Use of skills skills All controls High autonomy and 0.4370*** 0.4398*** repetitive (0.0391) (0.0383) Make presentations 0.0048 0.0131 (0.0318) (0.0320) Supervise others 0.0060 0.0043 (0.0218) (0.0221) Computer use at work (relative to no use) Computer use with low −0.1850** −0.1853** intensity (0.0741) (0.0738) Computer use with 0.0994 0.1025 medium intensity (0.1026) (0.1007) Computer use with high −0.0567 −0.0664* intensity (0.0401) (0.0383) Number of observations 1,963 1,963 1,963 1,963 1,963 1,963 R2 0.159 0.390 0.392 0.399 0.477 0.487 Note: All models estimated using ordinary least squares. Robust standard errors are in parentheses. Controls include gender, economic sector, and dummies for region. For all skills dummies of nonresponse (missing) were created and included in the regression, but they are not displayed in the table. Dependent variable is monthly earning considering main and second occupation. Y = 1 if self-employed and 0 if wage employed. *** p < 0.01, ** p < 0.05, * p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Appendix F Mismatch of Skills and Unexploited Potential Tables Table F.1  Difference in Mean of Self-Reported Readers Who Passed and Who Failed the Reading Assessment Core Literacy Test Y1: Read and passed Core Y2: Read and failed Core Obs Mean SD Obs Mean SD Difference: Y2–Y1 Female (%) 1,150 42.59 49.47 457 62.56 48.45 19.970*** Low SES (%) 1,149 14.16 34.88 452 19.62 39.76 5.463** Middle SES (%) 1,149 62.36 48.47 452 52.76 49.98 −9.605*** High SES (%) 1,149 23.48 42.40 452 27.62 44.76 4.142 Age range (years) (%) 1,150 28.86 12.32 457 30.98 12.16 2.119** 15–19 years 1,150 23.64 42.51 457 17.68 38.20 −5.957** 20–24 years 1,150 25.29 43.49 457 16.59 37.24 −8.701*** 25–34 years 1,150 26.08 43.93 457 32.11 46.74 6.022** 35–44 years 1,150 10.95 31.24 457 18.70 39.04 7.755*** 45–64 years 1,150 14.04 34.75 457 14.92 35.67 0.882 Education (%) No education 1,150 0.62 7.88 457 5.67 23.15 5.043*** Primary education 1,150 9.44 29.25 457 23.18 42.24 13.737*** JHS education 1,150 31.52 46.48 457 54.44 49.86 22.920*** SHS education 1,150 37.03 48.31 457 15.87 36.58 −21.159*** Tertiary education 1,150 21.39 41.02 457 0.85 9.19 −20.541*** Received ECE 1,141 81.54 38.81 453 70.84 45.50 −10.699*** Language spoken at work (%) Akan 806 78.92 40.81 370 88.20 32.30 9.278*** Ewe 806 7.34 26.10 370 7.83 26.90 0.486 Ga-Adangme 806 14.84 35.57 370 9.65 29.57 −5.191** Mole-Dagbani 806 2.83 16.60 370 3.60 18.65 0.767 English 806 74.40 43.67 370 33.14 47.14 −41.256*** Others 806 7.83 26.88 370 9.48 29.34 1.658 table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   131   132 Mismatch of Skills and Unexploited Potential Tables Table F.1  Difference in Mean of Self-Reported Readers Who Passed and Who Failed the Reading Assessment Core Literacy Test (continued) Y1: Read and passed Core Y2: Read and failed Core Obs Mean SD Obs Mean SD Difference: Y2–Y1 Language spoken at home (%) Akan 1,150 65.56 47.54 457 72.80 44.55 7.239** Ewe 1,150 9.04 28.69 457 6.54 24.75 −2.504 Ga-Adangme 1,150 9.67 29.57 457 3.34 17.98 −6.338*** Mole-Dagbani 1,150 5.22 22.25 457 5.00 21.82 −0.217 English 1,150 2.04 14.15 457 0.25 4.99 −1.794*** Others 1,150 8.47 27.85 457 12.08 32.63 3.615* Labor status (%) Employed 1,150 45.20 49.79 457 62.44 48.48 17.235*** Unemployed 1,150 6.40 24.49 457 3.38 18.08 −3.029** NEET 1,150 7.52 26.38 457 9.93 29.93 2.411 Inactive 1,150 40.88 49.18 457 24.26 42.91 −16.617*** Employment status (%) Formal employee 587 35.46 47.88 310 8.36 27.72 −27.103*** Informal employee 587 27.71 44.79 310 17.58 38.13 −10.125*** Self-employed 587 36.83 48.28 310 74.06 43.90 37.228*** Occupation (%) Low-skilled occupation 587 23.85 42.65 310 46.61 49.97 22.758*** Mid-skilled occupation 587 41.11 49.24 310 49.80 50.08 8.690** High-skilled occupation 587 35.04 47.75 310 3.59 18.64 −31.448*** Economic sector (%) Agriculture fishing, and mining 587 5.18 22.18 310 10.10 30.18 4.916* Manufacturing 587 8.82 28.39 310 13.76 34.50 4.935 Low- to mid-value-added 587 45.55 49.84 310 66.55 47.26 20.993*** High-value-added 587 40.44 49.12 310 9.60 29.51 −30.844*** Earnings Monthly earnings 648 678.52 2,091.11 304 401.69 970.30 −276.839** Socioemotional skills Extraversion (score) 1,135 2.59 0.59 344 2.43 0.59 −0.163*** Missing extraversion 1,150 1.11 10.46 457 27.99 44.95 26.887*** Conscientiousness (score) 1,135 3.32 0.52 340 2.96 0.63 −0.361*** Missing conscientiousness 1,150 1.11 10.46 457 28.45 45.17 27.343*** Openness (score) 1,135 3.22 0.52 343 2.86 0.62 −0.359*** Missing openness 1,150 1.11 10.46 457 28.11 45.00 27.000*** Emotional stability (score) 1,133 2.75 0.55 340 2.65 0.57 −0.099** Missing stability 1,150 1.47 12.02 457 28.51 45.20 27.049*** Agreeableness (score) 1,135 3.15 0.58 341 2.81 0.69 −0.342*** Missing agreeableness 1,150 1.11 10.46 457 28.40 45.14 27.293*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Mismatch of Skills and Unexploited Potential Tables 133 Table F.1  Difference in Mean of Self-Reported Readers Who Passed and Who Failed the Reading Assessment Core Literacy Test (continued) Y1: Read and passed Core Y2: Read and failed Core Obs Mean SD Obs Mean SD Difference: Y2–Y1 Self-reported skills (%) Reading Don’t use reading skill 1,131 0.00 0.00 448 0.00 0.00 n.a. Read with low intensity 1,131 30.46 46.04 448 66.28 47.33 35.822*** Read with medium intensity 1,131 28.23 45.03 448 20.17 40.17 −8.061*** Read with high intensity 1,131 41.31 49.26 448 13.55 34.26 −27.761*** Writing Don’t use writing skill 1,136 5.26 22.33 455 17.38 37.93 12.120*** Write with low intensity 1,136 60.25 48.96 455 67.05 47.05 6.800** Write with medium intensity 1,136 18.40 38.77 455 8.78 28.33 −9.626*** Write with high intensity 1,136 16.08 36.76 455 6.79 25.19 −9.294*** Numeracy Don’t use numeracy skill 1,150 3.96 19.52 457 4.25 20.21 0.292 Numeracy with low intensity 1,150 10.85 31.12 457 31.45 46.48 20.592*** Numeracy with medium intensity 1,150 59.51 49.11 457 58.94 49.25 −0.569 Numeracy with high intensity 1,150 25.68 43.70 457 5.36 22.55 −20.314*** Computer Don’t use computer skill 1,126 43.00 49.53 453 86.03 34.70 43.038*** Computer with low intensity 1,126 18.56 38.90 453 6.68 24.99 −11.882*** Computer with medium intensity 1,126 10.29 30.39 453 1.10 10.46 −9.183*** Computer with high intensity 1,126 28.16 45.00 453 6.19 24.12 −21.973*** Literacy assessment No answer on Reading Component or Core (%) 1,150 0.00 0.00 457 0.00 0.00 n.a. Reading Component (average) (%) Sentence incorrect answers 1,150 38.47 11.30 457 40.62 27.96 2.156 Sentence correct answers 1,150 46.64 9.50 457 31.70 16.63 −14.940*** Sentence no answers 1,150 14.90 11.70 457 27.68 18.61 12.785*** Passage incorrect answers 1,150 51.28 8.87 457 55.51 25.16 4.231*** Passage correct answers 1,150 44.67 7.96 457 27.59 17.80 −17.081*** Passage no answers 1,150 4.05 6.54 457 16.90 17.17 12.850*** Vocabulary incorrect answers 1,150 72.85 4.42 457 63.37 14.12 −9.482*** Vocabulary correct answers 1,150 22.84 2.39 457 18.24 6.36 −4.608*** Vocabulary no answers 1,150 4.31 5.21 457 18.39 18.53 14.089*** Core (average) (%) Score Core test 1,150 6.24 1.61 457 0.76 0.85 −5.478*** Passed Core test 1,150 100.00 0.00 457 0.00 0.00 −100.00*** Literacy assessment Reading proficiency 1,150 222.86 457 53.02 −169.841*** Note: n.a. = not applicable; NEET = Not in Employment, Education, or Training. *** p < 0.01, ** p < 0.05, * p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 134 Mismatch of Skills and Unexploited Potential Tables Table F.2  Unexploited Potential: Reading Skill Y1: Read at home and at work Y2: Read at home but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Female (%) 650 36.84 48.28 560 61.41 48.73 24.562*** Low SES (%) 646 16.24 36.91 555 21.61 41.20 5.371* Middle SES (%) 646 61.79 48.63 555 56.43 49.63 −5.355 High SES (%) 646 21.97 41.44 555 21.96 41.43 −0.015 Age range (years) (%) 650 36.47 11.87 560 37.22 11.36 0.750 15–19 years 650 1.17 10.77 560 2.69 16.19 1.517 20–24 years 650 14.81 35.55 560 10.70 30.93 −4.113 25–34 years 650 36.81 48.27 560 31.37 46.44 −5.446* 35–44 years 650 20.24 40.21 560 30.33 46.01 10.089*** 45–64 years 650 26.97 44.41 560 24.92 43.29 −2.046 Education (%) No education 650 1.92 13.72 560 6.35 24.41 4.438*** Primary education 650 4.19 20.05 560 12.22 32.78 8.031*** JHS education 650 27.30 44.59 560 57.41 49.49 30.106*** SHS education 650 37.13 48.35 560 20.80 40.62 −16.336*** Tertiary education 650 29.46 45.62 560 3.22 17.67 −26.240*** Received ECE 645 69.31 46.15 548 66.08 47.39 −3.237 Language spoken at work (%) Akan 649 80.75 39.45 559 85.78 34.96 5.026** Ewe 649 7.75 26.76 559 12.26 32.83 4.509* Ga-Adangme 649 16.14 36.82 559 12.44 33.03 −3.701 Mole-Dagbani 649 2.48 15.56 559 2.19 14.64 −0.293 English 649 76.57 42.39 559 31.36 46.44 −45.205*** Others 649 7.74 26.74 559 9.89 29.87 2.149 Language spoken at home (%) Akan 648 69.86 45.92 557 64.88 47.78 −4.977 Ewe 648 7.80 26.83 557 12.81 33.45 5.013** Ga-Adangme 648 8.83 28.40 557 7.33 26.09 −1.500 Mole-Dagbani 648 4.41 20.55 557 3.06 17.23 −1.355 English 648 2.00 14.03 557 0.29 5.40 −1.713*** Others 648 7.10 25.70 557 11.63 32.09 4.532** Labor status (%) Employed 650 100.00 0.00 560 100.00 0.00 n.a. Unemployed 650 0.00 0.00 560 0.00 0.00 n.a. NEET 650 0.00 0.00 560 0.00 0.00 n.a. Inactive 650 0.00 0.00 560 0.00 0.00 n.a. Employment status (%) Formal employee 650 36.91 48.29 560 5.96 23.69 −30.954*** Informal employee 650 24.76 43.20 560 21.64 41.22 −3.122 Self-employed 650 38.32 48.66 560 72.40 44.74 34.076*** Occupation (%) Low-skilled occupation 650 29.05 45.43 560 40.77 49.18 11.725*** Mid-skilled occupation 650 35.78 47.97 560 56.52 49.62 20.733*** High-skilled occupation 650 35.17 47.79 560 2.71 16.27 −32.458*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Mismatch of Skills and Unexploited Potential Tables 135 Table F.2  Unexploited Potential: Reading Skill (continued) Y1: Read at home and at work Y2: Read at home but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Economic sector (%) Agriculture, fishing, and mining 650 4.42 20.57 560 9.38 29.17 4.956** Manufacturing 650 10.99 31.31 560 11.22 31.59 0.228 Low- to mid-value-added 650 42.75 49.51 560 71.56 45.15 28.815*** High-value-added 650 41.84 49.37 560 7.84 26.90 −33.998*** Earnings Monthly earnings 626 771.72 2,391.23 540 383.97 669.19 −387.751*** Socioemotional skills Extraversion (score) 596 2.58 0.57 387 2.45 0.59 −0.131*** Missing extraversion 650 8.40 27.76 560 32.17 46.75 23.773*** Conscientiousness (score) 594 3.33 0.52 385 3.13 0.61 −0.198*** Missing conscientiousness 650 8.49 27.89 560 32.46 46.86 23.973*** Openness (score) 595 3.14 0.55 385 2.92 0.63 −0.219*** Missing openness 650 8.51 27.92 560 32.34 46.82 23.826*** Emotional stability (score) 593 2.81 0.55 382 2.73 0.56 −0.078* Missing stability 650 8.60 28.06 560 33.31 47.18 24.711*** Agreeableness (score) 594 3.09 0.60 383 2.94 0.69 −0.149*** Missing agreeableness 650 8.55 27.98 560 32.64 46.93 24.091*** Self-reported skills (%) Reading Don’t use reading skill 649 0.00 0.00 547 0.00 0.00 n.a. Read with low intensity 649 44.37 49.72 547 70.70 45.55 26.333*** Read with medium intensity 649 24.79 43.21 547 18.29 38.69 −6.501** Read with high intensity 649 30.84 46.22 547 11.01 31.33 −19.832*** Writing Don’t use writing skill 646 3.47 18.32 557 23.72 42.58 20.251*** Write with low intensity 646 71.84 45.01 557 67.40 46.92 −4.437 Write with medium intensity 646 14.24 34.97 557 6.88 25.34 −7.358*** Write with high intensity 646 10.45 30.62 557 1.99 13.99 −8.456*** Numeracy Don’t use numeracy skill 650 2.36 15.20 560 4.81 21.41 2.446* Numeracy with low intensity 650 17.32 37.87 560 33.84 47.36 16.523*** Numeracy with medium intensity 650 68.90 46.32 560 59.96 49.04 −8.940** Numeracy with high intensity 650 11.42 31.82 560 1.39 11.70 −10.029*** Computer Don’t use computer skill 641 55.66 49.72 556 90.23 29.72 34.571*** Computer with low intensity 641 9.30 29.07 556 4.08 19.81 −5.220*** Computer with medium intensity 641 6.50 24.66 556 2.39 15.30 −4.101*** Computer with high intensity 641 28.54 45.20 556 3.29 17.86 −25.251*** Literacy assessment No answer on Reading Component or Core (%) 650 18.54 38.89 560 36.87 48.29 18.335*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 136 Mismatch of Skills and Unexploited Potential Tables Table F.2  Unexploited Potential: Reading Skill (continued) Y1: Read at home and at work Y2: Read at home but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Reading Component (average) (%) Sentence incorrect answers 527 39.98 15.20 353 42.80 24.52 2.814 Sentence correct answers 527 44.30 12.34 353 37.26 17.06 −7.045*** Sentence no answers 527 15.72 13.39 353 19.95 16.23 4.230*** Passage incorrect answers 527 52.79 13.62 353 56.04 21.32 3.242** Passage correct answers 527 41.61 12.54 353 33.54 17.82 −8.074*** Passage no answers 527 5.59 9.71 353 10.42 13.56 4.832*** Vocabulary incorrect answers 527 71.40 8.17 353 68.35 12.03 −3.052*** Vocabulary correct answers 527 22.24 3.63 353 20.54 5.42 −1.698*** Vocabulary no answers 527 6.36 10.45 353 11.11 15.70 4.750*** Core (average) Score Core test 527 5.29 2.74 353 2.97 2.71 −2.320*** Passed Core test 527 78.76 40.94 353 49.42 50.07 −29.338*** Literacy assessment Reading proficiency 650 176 560 107 −69.405*** Note: n.a. = not applicable; NEET = Not in Employment, Education, or Training. *** p < 0.01, ** p < 0.05, * p < 0.1. Table F.3  Unexploited Potential: Writing Skill Y1: Write at home and at work Y2: Write at home but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Female (%) 812 42.82 49.51 248 55.20 49.83 12.383*** Low SES (%) 808 17.56 38.07 245 22.07 41.56 4.514 Middle SES (%) 808 62.38 48.47 245 51.68 50.07 −10.700** High SES (%) 808 20.06 40.07 245 26.24 44.08 6.186 Age range (years) (%) 812 37.36 11.81 248 35.29 10.86 −2.065** 15–19 years 812 1.21 10.95 248 3.54 18.51 2.325* 20–24 years 812 12.16 32.70 248 13.69 34.44 1.533 25–34 years 812 34.79 47.66 248 35.76 48.03 0.974 35–44 years 812 23.65 42.52 248 26.76 44.36 3.114 45–64 years 812 28.19 45.02 248 20.25 40.27 −7.946** Education (%) No education 812 1.93 13.75 248 7.56 26.49 5.633*** Primary education 812 4.70 21.19 248 12.95 33.64 8.243*** JHS education 812 34.91 47.70 248 53.19 50.00 18.276*** SHS education 812 33.94 47.38 248 23.25 42.33 −10.684*** Tertiary education 812 24.52 43.05 248 3.05 17.24 −21.467*** Received ECE 805 67.32 46.93 243 62.77 48.44 −4.555 table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Mismatch of Skills and Unexploited Potential Tables 137 Table F.3  Unexploited Potential: Writing Skill (continued) Y1: Write at home and at work Y2: Write at home but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Language spoken at work (%) Akan 810 80.21 39.86 248 84.73 36.04 4.516 Ewe 810 9.47 29.30 248 12.00 32.56 2.525 Ga-Adangme 810 17.33 37.88 248 11.61 32.10 −5.724** Mole-Dagbani 810 2.38 15.25 248 4.91 21.65 2.530 English 810 68.95 46.30 248 24.60 43.15 −44.346*** Others 810 8.02 27.18 248 10.41 30.60 2.389 Language spoken at home (%) Akan 808 68.53 46.47 247 63.61 48.21 −4.912 Ewe 808 9.66 29.56 247 10.97 31.32 1.316 Ga-Adangme 808 9.72 29.65 247 5.92 23.64 −3.807* Mole-Dagbani 808 3.49 18.37 247 5.72 23.26 2.222 English 808 1.64 12.69 247 0.76 8.68 −0.879 Others 808 6.96 25.47 247 13.02 33.73 6.060* Labor status (%) Employed 812 100.00 0.00 248 100.00 0.00 n.a. Unemployed 812 0.00 0.00 248 0.00 0.00 n.a. NEET 812 0.00 0.00 248 0.00 0.00 n.a. Inactive 812 0.00 0.00 248 0.00 0.00 n.a. Employment status (%) Formal employee 812 31.27 46.39 248 4.87 21.56 −26.403*** Informal employee 812 20.55 40.43 248 24.75 43.24 4.203 Self-employed 812 48.18 50.00 248 70.38 45.75 22.200*** Occupation (%) Low-skilled occupation 812 32.62 46.91 248 43.42 49.67 10.803** Mid-skilled occupation 812 39.66 48.95 248 54.58 49.89 14.914*** High-skilled occupation 812 27.72 44.79 248 2.00 14.02 −25.717*** Economic sector (%) Agriculture, fishing, and mining 812 4.60 20.95 248 15.44 36.20 10.842*** Manufacturing 812 14.54 35.28 248 4.85 21.53 −9.691*** Low- to mid-value-added 812 46.75 49.92 248 72.88 44.55 26.135*** High-value-added 812 34.11 47.44 248 6.83 25.27 −27.286*** Earnings Monthly earnings 783 704.36 2,110.72 236 315.79 457.89 −388.565*** Socioemotional skills Extraversion (score) 698 2.56 0.58 178 2.43 0.59 −0.134** Missing extraversion 812 13.81 34.52 248 30.64 46.19 16.837*** Conscientiousness (score) 695 3.32 0.51 176 3.04 0.67 −0.281*** Missing conscientiousness 812 14.10 34.82 248 31.30 46.46 17.198*** Openness (score) 695 3.06 0.60 177 3.03 0.57 −0.030 table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 138 Mismatch of Skills and Unexploited Potential Tables Table F.3  Unexploited Potential: Writing Skill (continued) Y1: Write at home and at work Y2: Write at home but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Missing openness 812 14.17 34.90 248 30.86 46.29 16.689*** Emotional stability (score) 693 2.81 0.55 175 2.77 0.54 −0.039 Missing stability 812 14.27 34.99 248 32.36 46.88 18.092*** Agreeableness (score) 692 3.09 0.61 176 2.93 0.69 −0.168** Missing agreeableness 812 14.49 35.22 248 31.34 46.48 16.849*** Self-reported skills (%) Reading Don’t use reading skill 806 4.40 20.52 245 12.69 33.35 8.285*** Read with low intensity 806 46.12 49.88 245 58.20 49.42 12.086*** Read with medium intensity 806 24.00 42.73 245 19.14 39.42 −4.864 Read with high intensity 806 25.48 43.60 245 9.98 30.03 −15.507*** Writing Don’t use writing skill 809 0.00 0.00 244 0.00 0.00 n.a. Write with low intensity 809 77.14 42.02 244 89.66 30.50 12.529*** Write with medium intensity 809 13.33 34.01 244 9.58 29.49 −3.752 Write with high intensity 809 9.53 29.38 244 0.76 8.68 −8.777*** Numeracy Don’t use numeracy skill 812 3.06 17.24 248 4.58 20.95 1.519 Numeracy with low intensity 812 18.59 38.92 248 35.98 48.09 17.393*** Numeracy with medium intensity 812 68.91 46.31 248 57.71 49.50 −11.208** Numeracy with high intensity 812 9.44 29.25 248 1.73 13.08 −7.704*** Computer Don’t use computer skill 800 64.05 48.01 247 90.57 29.28 26.519*** Computer with low intensity 800 8.11 27.31 247 2.50 15.63 −5.611*** Computer with medium intensity 800 6.19 24.11 247 1.62 12.64 −4.574*** Computer with high intensity 800 21.65 41.21 247 5.31 22.48 −16.334*** Literacy assessment No answer on Reading Component or Core (%) 812 24.30 42.92 248 32.46 46.92 8.157** Reading Component (average) (%) Sentence incorrect answers 611 40.06 15.84 165 42.09 24.57 2.034 Sentence correct answers 611 43.51 12.83 165 37.73 17.31 −5.782*** Sentence no answers 611 16.43 13.83 165 20.18 16.53 3.748** Passage incorrect answers 611 53.18 14.91 165 55.94 21.38 2.755 Passage correct answers 611 40.93 13.37 165 33.52 17.78 −7.404*** Passage no answers 611 5.89 9.85 165 10.54 14.23 4.649*** Vocabulary incorrect answers 611 71.57 7.80 165 68.44 10.10 −3.132*** Vocabulary correct answers 611 22.18 3.67 165 20.46 5.42 −1.718*** Vocabulary no answers 611 6.25 9.88 165 11.10 13.74 4.850*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Mismatch of Skills and Unexploited Potential Tables 139 Table F.3  Unexploited Potential: Writing Skill (continued) Y1: Write at home and at work Y2: Write at home but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Core (average) Score Core test 611 5.01 2.80 165 2.80 2.60 −2.207*** Passed Core test 611 75.95 42.77 165 47.20 50.07 −28.749*** Literacy assessment Reading proficiency 812 164 248 109 −54.538*** Note: n.a. = not applicable; NEET = Not in Employment, Education, or Training. *** p < 0.01, ** p < 0.05, * p < 0.1. Table F.4  Unexploited Potential: Numeracy Skill Y1: Use mathematics at home Y2: Use mathematics at home and at work but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Female (%) 1,603 58.30 49.32 88 39.40 49.14 −18.905*** Low SES (%) 1,591 25.27 43.47 88 31.19 46.59 5.916 Middle SES (%) 1,591 55.57 49.70 88 49.81 50.29 −5.764 High SES (%) 1,591 19.15 39.36 88 19.00 39.46 −0.152 Age range (years) (%) 1,603 37.41 11.61 88 36.92 12.18 −0.497 15–19 years 1,603 1.80 13.32 88 0.65 8.06 −1.158 20–24 years 1,603 11.78 32.25 88 13.00 33.82 1.217 25–34 years 1,603 32.05 46.68 88 41.38 49.53 9.329 35–44 years 1,603 27.47 44.65 88 18.62 39.15 −8.855* 45–64 years 1,603 26.89 44.35 88 26.36 44.31 −0.533 Education (%) No education 1,603 25.58 43.64 88 25.16 43.64 −0.417 Primary education 1,603 9.96 29.95 88 7.95 27.21 −2.003 JHS education 1,603 34.33 47.49 88 27.04 44.67 −7.289 SHS education 1,603 19.54 39.67 88 20.46 40.57 0.911 Tertiary education 1,603 10.60 30.79 88 19.40 39.77 8.799 Received ECE 1,585 55.63 49.70 87 56.09 49.92 0.459 Language spoken at work (%) Akan 1,599 80.59 39.56 88 70.73 45.76 −9.862 Ewe 1,599 6.15 24.03 88 17.59 38.29 11.435* Ga-Adangme 1,599 13.17 33.83 88 14.51 35.43 1.345 Mole-Dagbani 1,599 7.85 26.90 88 6.72 25.17 −1.131 English 1,599 38.72 48.73 88 48.77 50.27 10.049 Others 1,599 12.16 32.70 88 12.14 32.85 −0.023 Language spoken at home (%) Akan 1,598 64.12 47.98 88 61.81 48.86 −2.307 Ewe 1,598 6.71 25.03 88 12.13 32.84 5.420 table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 140 Mismatch of Skills and Unexploited Potential Tables Table F.4  Unexploited Potential: Numeracy Skill (continued) Y1: Use mathematics at home Y2: Use mathematics at home and at work but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Ga-Adangme 1,598 6.89 25.34 88 6.04 23.96 −0.847 Mole-Dagbani 1,598 8.00 27.13 88 11.02 31.49 3.021 English 1,598 0.84 9.11 88 0.00 0.00 −0.835*** Others 1,598 13.44 34.12 88 8.99 28.77 −4.451 Labor status (%) Employed 1,603 100.00 0.00 88 100.00 0.00 n.a. Unemployed 1,603 0.00 0.00 88 0.00 0.00 n.a. NEET 1,603 0.00 0.00 88 0.00 0.00 n.a. Inactive 1,603 0.00 0.00 88 0.00 0.00 n.a. Employment status (%) Formal employee 1,603 13.44 34.12 88 37.91 48.79 24.468*** Informal employee 1,603 17.85 38.31 88 45.14 50.05 27.282*** Self-employed 1,603 68.71 46.38 88 16.95 37.74 −51.750*** Occupation (%) Low-skilled occupation 1,603 38.82 48.75 88 40.95 49.46 2.123 Mid-skilled occupation 1,603 48.24 49.98 88 28.93 45.60 −19.314*** High-skilled occupation 1,603 12.93 33.57 88 30.12 46.14 17.191** Economic sector (%) Agriculture, fishing, and mining 1,603 11.12 31.45 88 11.06 31.55 −0.061 Manufacturing 1,603 11.10 31.42 88 3.29 17.94 −7.807*** Low- to mid-value-added 1,603 62.73 48.37 88 35.03 47.98 −27.707*** High-value-added 1,603 15.05 35.76 88 50.62 50.28 35.575*** Earnings Monthly earnings 1,560 566.46 1,770.94 82 315.20 490.52 −251.257*** Socioemotional skills Extraversion (score) 931 2.51 0.60 56 2.61 0.54 0.102 Missing extraversion 1,603 44.55 49.72 88 33.74 47.55 −10.818 Conscientiousness (score) 925 3.23 0.57 56 3.39 0.53 0.161 Missing conscientiousness 1,603 44.95 49.76 88 33.74 47.55 −11.218 Openness (score) 924 3.04 0.60 56 3.05 0.61 0.009 Missing openness 1,603 44.91 49.76 88 33.74 47.55 −11.180 Emotional stability (score) 918 2.77 0.56 56 2.74 0.57 −0.032 Missing stability 1,603 45.30 49.79 88 33.74 47.55 −11.563 Agreeableness (score) 919 3.02 0.64 56 3.12 0.55 0.099 Missing agreeableness 1,603 45.16 49.78 88 33.74 47.55 −11.422 Self-reported skills (%) Reading Don’t use reading skill 1,593 37.74 48.49 88 30.28 46.21 −7.464 Read with low intensity 1,593 35.92 47.99 88 39.41 49.15 3.489 table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Mismatch of Skills and Unexploited Potential Tables 141 Table F.4  Unexploited Potential: Numeracy Skill (continued) Y1: Use mathematics at home Y2: Use mathematics at home and at work but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Read with medium intensity 1,593 14.01 34.72 88 12.36 33.10 −1.650 Read with high intensity 1,593 12.32 32.88 88 17.95 38.60 5.625 Writing Don’t use writing skill 1,596 40.78 49.16 88 32.45 47.09 −8.325 Write with low intensity 1,596 48.11 49.98 88 52.27 50.23 4.160 Write with medium intensity 1,596 6.98 25.48 88 9.97 30.14 2.998 Write with high intensity 1,596 4.13 19.91 88 5.30 22.53 1.167 Numeracy Don’t use numeracy skill 1,603 0.00 0.00 88 0.00 0.00 n.a. Numeracy with low intensity 1,603 29.59 45.66 88 43.38 49.84 13.786* Numeracy with medium intensity 1,603 65.55 47.53 88 54.51 50.08 −11.044 Numeracy with high intensity 1,603 4.86 21.51 88 2.12 14.48 −2.741 Computer Don’t use computer skill 1,590 81.99 38.44 88 74.34 43.92 −7.650 Computer with low intensity 1,590 4.26 20.20 88 4.93 21.77 0.667 Computer with medium intensity 1,590 3.27 17.79 88 1.49 12.20 −1.775 Computer with high intensity 1,590 10.48 30.63 88 19.23 39.64 8.758 Literacy assessment No answer on Reading Component or Core (%) 1,603 46.91 49.92 88 45.95 50.12 −0.962 Reading Component (average) (%) Sentence incorrect answers 843 43.19 23.48 51 44.15 19.78 0.964 Sentence correct answers 843 38.34 17.12 51 42.38 15.53 4.038 Sentence no answers 843 18.47 16.27 51 13.47 11.85 −5.003** Passage incorrect answers 843 55.78 20.31 51 54.78 17.14 −1.007 Passage correct answers 843 35.11 17.65 51 40.02 15.18 4.908* Passage no answers 843 9.11 14.14 51 5.21 7.93 −3.901*** Vocabulary incorrect answers 843 67.41 14.25 51 69.97 11.44 2.559 Vocabulary correct answers 843 20.20 6.12 51 22.05 4.09 1.849*** Vocabulary no answers 843 12.38 18.97 51 7.97 14.92 −4.408* Core (average) Score Core test 843 3.86 3.09 51 5.15 2.52 1.286*** Passed Core test 843 58.99 49.21 51 83.36 37.62 24.364*** Literacy assessment Reading proficiency 1,603 120 88 144 24.144** Note: NEET = Not in Employment, Education, or Training. *** p < 0.01, ** p < 0.05, * p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 142 Mismatch of Skills and Unexploited Potential Tables Table F.5  Unexploited Potential: Computer Skill Y1: Use computer at home Y2: Use computer at home and at work but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Female (%) 172 27.59 44.83 190 28.29 45.16 0.706 Low SES (%) 172 8.74 28.32 190 18.68 39.08 9.943** Middle SES (%) 172 60.85 48.95 190 60.05 49.11 −0.801 High SES (%) 172 30.41 46.14 190 21.27 41.03 −9.143 Age (%) 172 34.81 11.41 190 29.81 8.94 −4.996*** 15–19 years 172 0.70 8.33 190 3.72 18.97 3.024 20–24 years 172 17.94 38.48 190 30.88 46.32 12.944** 25–34 years 172 43.71 49.75 190 39.11 48.93 −4.605 35–44 years 172 14.59 35.40 190 20.91 40.77 6.318 45–64 years 172 23.07 42.25 190 5.39 22.64 −17.681*** Education (%) No education 172 0.00 0.00 190 1.34 11.53 1.339* Primary education 172 0.31 5.58 190 4.76 21.35 4.451** JHS education 172 2.25 14.88 190 22.63 41.95 20.376*** SHS education 172 33.01 47.16 190 48.84 50.12 15.831** Tertiary education 172 64.43 48.01 190 22.43 41.82 −41.997*** Received ECE 170 77.06 42.17 187 81.95 38.56 4.892 Language spoken at work (%) Akan 172 73.33 44.36 190 76.25 42.67 2.928 Ewe 172 4.93 21.71 190 7.95 27.12 3.018 Ga-Adangme 172 17.47 38.08 190 19.61 39.81 2.143 Mole-Dagbani 172 1.42 11.86 190 3.29 17.87 1.867 English 172 95.74 20.25 190 75.46 43.14 −20.279*** Others 172 8.68 28.24 190 7.60 26.57 −1.086 Language spoken at home (%) Akan 172 69.39 46.22 190 58.82 49.35 −10.570* Ewe 172 6.90 25.42 190 12.73 33.42 5.829 Ga-Adangme 172 10.73 31.04 190 11.16 31.57 0.424 Mole-Dagbani 172 3.40 18.18 190 7.64 26.63 4.233 English 172 4.25 20.22 190 3.07 17.31 −1.173 Others 172 5.33 22.53 190 6.59 24.87 1.256 Labor status (%) Employed 172 100.00 0.00 190 100.00 0.00 n.a. Unemployed 172 0.00 0.00 190 0.00 0.00 n.a. NEET 172 0.00 0.00 190 0.00 0.00 n.a. Inactive 172 0.00 0.00 190 0.00 0.00 n.a. Employment status (%) Formal employee 172 58.89 49.35 190 26.12 44.04 −32.774*** Informal employee 172 20.93 40.80 190 40.21 49.16 19.282*** Self-employed 172 20.18 40.25 190 33.67 47.38 13.492** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Mismatch of Skills and Unexploited Potential Tables 143 Table F.5  Unexploited Potential: Computer Skill (continued) Y1: Use computer at home Y2: Use computer at home and at work but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Occupation (%) Low-skilled occupation 172 7.31 26.11 190 28.23 45.13 20.921*** Mid-skilled occupation 172 28.27 45.16 190 39.73 49.06 11.458* High-skilled occupation 172 64.41 48.02 190 32.04 46.78 −32.380*** Economic sector (%) Agriculture, fishing, and mining 172 1.42 11.89 190 3.33 18.00 1.909 Manufacturing 172 4.25 20.22 190 8.52 27.99 4.270 Low- to mid-value-added 172 33.74 47.42 190 53.53 50.01 19.795*** High-value-added 172 60.59 49.01 190 34.62 47.70 −25.975*** Earnings Monthly earnings 163 1,360.81 4,175.78 184 514.12 703.04 −846.695* Socioemotional skills Extraversion (score) 171 2.72 0.56 182 2.62 0.54 −0.097 Missing extraversion 172 0.31 5.58 190 4.25 20.22 3.937** Conscientiousness (score) 171 3.40 0.46 182 3.34 0.50 −0.060 Missing conscientiousness 172 0.31 5.58 190 4.25 20.22 3.937** Openness (score) 171 3.27 0.48 182 3.18 0.58 −0.084 Missing openness 172 0.31 5.58 190 4.25 20.22 3.937** Emotional stability (score) 171 2.85 0.54 182 2.75 0.52 −0.091 Missing stability 172 0.31 5.58 190 4.25 20.22 3.937** Agreeableness (score) 171 3.19 0.57 182 3.12 0.60 −0.077 Missing agreeableness 172 0.31 5.58 190 4.25 20.22 3.937** Self-reported skills (%) Reading Don’t use reading skill 172 0.00 0.00 189 4.20 20.12 4.205*** Read with low intensity 172 20.51 40.50 189 36.49 48.27 15.976*** Read with medium intensity 172 32.86 47.11 189 32.65 47.02 −0.213 Read with high intensity 172 46.62 50.03 189 26.66 44.33 −19.967*** Writing Don’t use writing skill 170 2.46 15.53 189 9.50 29.40 7.040** Write with low intensity 170 50.99 50.14 189 67.14 47.10 16.153** Write with medium intensity 170 25.82 43.89 189 17.21 37.85 −8.609* Write with high intensity 170 20.74 40.66 189 6.15 24.09 −14.583*** Numeracy Don’t use numeracy skill 172 4.24 20.22 190 3.45 18.30 −0.795 Numeracy with low intensity 172 10.67 30.96 190 10.68 30.97 0.013 Numeracy with medium intensity 172 56.19 49.76 190 77.78 41.68 21.597*** Numeracy with high intensity 172 28.90 45.46 190 8.08 27.33 −20.815*** Computer Don’t use computer skill 172 0.00 0.00 178 0.00 0.00 n.a. Computer with low intensity 172 6.75 25.17 178 42.17 49.52 35.413*** table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 144 Mismatch of Skills and Unexploited Potential Tables Table F.5  Unexploited Potential: Computer Skill (continued) Y1: Use computer at home Y2: Use computer at home and at work but not at work Difference: Obs Mean SD Obs Mean SD Y2–Y1 Computer with medium intensity 172 8.47 27.92 178 26.07 44.02 17.602*** Computer with high intensity 172 84.78 36.02 178 31.77 46.69 −53.015*** Literacy assessment No answer on Reading Component or Core (%) 172 8.15 27.45 190 11.03 31.40 2.871 Reading Component (average) (%) Sentence incorrect answers 153 40.62 7.64 162 38.19 14.12 −2.439* Sentence correct answers 153 49.07 7.19 162 44.25 11.82 −4.825*** Sentence no answers 153 10.30 10.12 162 17.57 14.20 7.264*** Passage incorrect answers 153 51.48 6.76 162 50.63 12.24 −0.846 Passage correct answers 153 45.91 5.60 162 42.69 10.84 −3.214*** Passage no answers 153 2.62 5.33 162 6.68 12.35 4.060*** Vocabulary incorrect answers 153 73.63 3.87 162 71.35 9.03 −2.282** Vocabulary correct answers 153 23.30 1.85 162 22.19 3.90 −1.111*** Vocabulary no answers 153 3.06 4.20 162 6.46 11.99 3.393*** Core (average) Score Core test 153 6.63 1.72 162 5.65 2.54 −0.984*** Passed Core test 153 95.87 19.97 162 85.11 35.71 −10.763*** Literacy assessment Reading proficiency 172 233 190 199 −33.236*** Note: n.a. = not applicable; NEET = Not in Employment, Education, or Training; Obs = number of observations. *** p < 0.01, ** p < 0.05, * p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Appendix G Effect of Socioemotional Skills on Education and Labor Outcomes Table G.1  Years of Education, Controlling for Socioemotional Skills Time preference Risk Hostile bias Grit Big five Socioemotional skills Time preference (score) −0.4006*** (0.0964) Risk aversion (score) 0.2359** (0.0981) Hostile bias (score) −0.8940*** (0.1348) Grit (score) 0.3309* (0.1751) Openness (score) 0.6430*** (0.1853) Conscientiousness (score) 1.0076*** (0.1757) Extraversion (score) 0.6508*** (0.1502) Agreeableness (score) 0.4639*** (0.1668) Emotional stability (score) 0.2862* (0.1684) Controls Female (%) −0.4034** −0.3887* −0.3431* −0.3654* −0.0660 (0.1992) (0.2005) (0.1981) (0.2002) (0.1967) Age 0.0812*** 0.0869*** 0.0881*** 0.0841*** 0.0788*** (0.0096) (0.0094) (0.0095) (0.0096) (0.0094) Middle socioeconomic 0.7723** 0.7370** 0.6398** 0.7660** 0.7546** status (%) (0.3175) (0.3092) (0.2921) (0.3114) (0.2966) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1   145   146 Effect of Socioemotional Skills on Education and Labor Outcomes Table G.1  Years of Education, Controlling for Socioemotional Skills (continued) Time preference Risk Hostile bias Grit Big five High socioeconomic 0.8857** 0.8384** 0.7284** 0.8551** 0.6907** status (%) (0.3612) (0.3568) (0.3362) (0.3609) (0.3414) Mother’s education Primary −0.8607 −0.9151 −0.9787 −0.9161 −0.6848 (0.6032) (0.6176) (0.6097) (0.6239) (0.6082) Secondary 0.8415** 0.8650** 0.8621** 0.8787** 0.8963** (0.3500) (0.3517) (0.3433) (0.3537) (0.3489) Tertiary 1.9081*** 1.9352*** 1.9516*** 1.8942*** 1.8244*** (0.6603) (0.6487) (0.5916) (0.6696) (0.6760) Dummy for region x x x x X Number of observations 1,868 1,868 1,868 1,868 1,868 R2 0.144 0.137 0.163 0.135 0.209 Note: All models estimated using ordinary least squares. Robust standard errors are in parentheses. Controls include gender, age, socioeconomic status at age 15 years, mother's education, and dummies for region. Dependent variable equals years of education. ***p < 0.01, **p < 0.05, *p < 0.1 Table G.2 Linear Probability Model of Completing SHS or Tertiary Education, Controlling for Socioemotional Skills Time preference Risk Hostile bias Grit Big five Socioemotional skills Time preference (score) −0.0515*** (0.0137) Risk aversion (score) 0.0360*** (0.0124) Hostile bias (score) −0.1149*** (0.0193) Grit (score) 0.0293 (0.0238) Openness (score) 0.0783*** (0.0255) Conscientiousness (score) 0.1410*** (0.0247) Extraversion (score) 0.0850*** (0.0221) Agreeableness (score) 0.0518** (0.0234) Emotional stability (score) 0.0525** (0.0232) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Effect of Socioemotional Skills on Education and Labor Outcomes 147 Table G.2  Linear Probability Model of Completing SHS or Tertiary Education, Controlling for Socioemotional Skills (continued) Time preference Risk Hostile bias Grit Big five Controls Female (%) −0.0499* −0.0479* −0.0421 −0.0461 −0.0026 (0.0282) (0.0283) (0.0280) (0.0283) (0.0278) Age 0.0074*** 0.0082*** 0.0083*** 0.0079*** 0.0070*** (0.0013) (0.0013) (0.0013) (0.0013) (0.0013) Middle socioeconomic 0.0776 0.0729 0.0605 0.0758 0.0759* status (%) (0.0478) (0.0468) (0.0440) (0.0468) (0.0435) High socioeconomic 0.0809 0.0742 0.0606 0.0772 0.0557 status (%) (0.0537) (0.0532) (0.0497) (0.0536) (0.0504) Mother’s education Primary −0.0869 −0.0940 −0.1021 −0.0940 −0.0670 (0.0752) (0.0772) (0.0760) (0.0770) (0.0728) Secondary 0.0877 0.0903* 0.0904* 0.0927* 0.0926* (0.0541) (0.0544) (0.0535) (0.0546) (0.0542) Tertiary 0.1745* 0.1782* 0.1801** 0.1739* 0.1612 (0.0974) (0.0960) (0.0899) (0.0983) (0.1007) Dummy for region x x x x x Number of observations 1,868 1,868 1,868 1,868 1,868 R2 0.087 0.082 0.106 0.078 0.152 Note: All models estimated using ordinary least squares. Robust standard errors in parentheses. Controls include gender, age, socioeconomic status at age 15 years, mother's education, and dummies for region. Dependent variable equals 1 if highest level of education completed is senior high school (SHS) or tertiary, and 0 otherwise. ***p < 0.01, **p < 0.05, *p < 0.1. Table G.3 Linear Probability Model of Being Employed, Controlling for Socioemotional Skills Time preference Risk Hostile bias Grit Big five Socioemotional skills Time preference (score) −0.0108 (0.0150) Risk aversion (score) −0.0047 (0.0130) Hostile bias (score) −0.0023 (0.0201) Grit (score) −0.0194 (0.0240) Openness (score) −0.0006 (0.0281) Conscientiousness (score) 0.0254 (0.0280) table continues next page Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 148 Effect of Socioemotional Skills on Education and Labor Outcomes Table G.3  Linear Probability Model of Being Employed, Controlling for Socioemotional Skills (continued) Time preference Risk Hostile bias Grit Big five Extraversion (score) −0.0228 (0.0267) Agreeableness (score) −0.0402* (0.0237) Emotional stability (score) −0.0083 (0.0257) Controls Female (%) −0.1072*** −0.1067*** −0.1065*** −0.1086*** −0.1048*** (0.0302) (0.0302) (0.0301) (0.0303) (0.0304) Age 0.0068*** 0.0068*** 0.0068*** 0.0069*** 0.0070*** (0.0015) (0.0015) (0.0015) (0.0015) (0.0015) Middle socioeconomic 0.0677 0.0677 0.0673 0.0654 0.0690 status (%) (0.0483) (0.0483) (0.0479) (0.0469) (0.0480) High socioeconomic 0.0795 0.0794 0.0785 0.0794 0.0816 status (%) (0.0518) (0.0517) (0.0512) (0.0517) (0.0526) Mother’s education Primary −0.1510* −0.1518* −0.1520* −0.1536* −0.1592* (0.0895) (0.0906) (0.0904) (0.0902) (0.0893) Secondary −0.0908 −0.0889 −0.0892 −0.0892 −0.0929* (0.0568) (0.0564) (0.0564) (0.0563) (0.0558) Tertiary −0.2269* −0.2259* −0.2256* −0.2238* −0.2247* (0.1359) (0.1360) (0.1334) (0.1337) (0.1365) Dummy for region x x x x x Number of observations 1,362 1,362 1,362 1,362 1,362 R2 0.091 0.091 0.091 0.092 0.096 Note: All models estimated using ordinary least squares. Robust standard errors in parentheses. Controls include gender, age, socioeconomic status at age 15 years, mother's education, and dummies for region. Dependent variable equals 1 if employed and 0 if not employed and not in education. ***p < 0.01, **p < 0.05, *p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Effect of Socioemotional Skills on Education and Labor Outcomes 149 Table G.4 Linear Probability Model of Being Self-Employed, Controlling for Socioemotional Skills Time preference Risk Hostile bias Grit Big five Socioemotional skills Time preference (score) −0.0007 (0.0190) Risk aversion (score) −0.0408*** (0.0158) Hostile bias (score) 0.0878*** (0.0223) Grit (score) 0.0077 (0.0313) Openness (score) 0.0308 (0.0314) Conscientiousness (score) −0.0882*** (0.0342) Extraversion (score) −0.0712** (0.0282) Agreeableness (score) −0.0328 (0.0297) Emotional stability (score) −0.0516* (0.0307) Controls Female (%) 0.2829*** 0.2805*** 0.2765*** 0.2840*** 0.2621*** (0.0363) (0.0361) (0.0362) (0.0366) (0.0367) Age 0.0056*** 0.0053*** 0.0053*** 0.0056*** 0.0068*** (0.0018) (0.0018) (0.0017) (0.0018) (0.0017) Middle socioeconomic −0.0159 −0.0187 −0.0093 −0.0160 −0.0099 status (%) (0.0478) (0.0479) (0.0470) (0.0478) (0.0468) High socioeconomic 0.1039* 0.1070* 0.1049* 0.1026* 0.1139** status (%) (0.0554) (0.0554) (0.0547) (0.0557) (0.0535) Mother’s education Primary 0.0343 0.0332 0.0401 0.0350 0.0272 (0.1002) (0.0993) (0.0979) (0.0999) (0.0979) Secondary −0.0321 −0.0377 −0.0323 −0.0324 −0.0219 (0.0660) (0.0644) (0.0676) (0.0659) (0.0658) Tertiary 0.0096 0.0167 0.0057 0.0083 0.0264 (0.1073) (0.1037) (0.1030) (0.1070) (0.1025) Dummy for region x x x x x Number of observations 1,089 1,089 1,089 1,089 1,089 R2 0.121 0.128 0.137 0.121 0.144 Note: All models estimated using ordinary least squares. Robust standard errors in parentheses. Controls include gender, age, socioeconomic status at age 15 years, mother's education, and dummies for region. Dependent variable equals 1 if selfemployed and 0 if wage employed. ***p < 0.01, **p < 0.05, *p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 150 Effect of Socioemotional Skills on Education and Labor Outcomes Table G.5 Linear Probability Model of Working in a Medium- to High-Skilled Occupation, Controlling for Socioemotional Skills Time preference Risk Hostile bias Grit Big five Socioemotional skills Time preference (score) −0.0113 (0.0179) Risk aversion (score) 0.0144 (0.0153) Hostile bias (score) −0.0244 (0.0232) Grit (score) 0.0388 (0.0286) Openness (score) 0.0366 (0.0346) Conscientiousness (score) 0.0351 (0.0331) Extraversion (score) 0.0261 (0.0282) Agreeableness (score) 0.0481 (0.0305) Emotional stability (score) 0.0722** (0.0317) Controls Female (%) 0.2295*** 0.2314*** 0.2323*** 0.2358*** 0.2571*** (0.0357) (0.0355) (0.0357) (0.0361) (0.0353) Age −0.0004 −0.0002 −0.0003 −0.0004 −0.0011 (0.0018) (0.0018) (0.0017) (0.0018) (0.0018) Middle socioeconomic 0.0951* 0.0957* 0.0929* 0.0944* 0.0898* status (%) (0.0492) (0.0492) (0.0495) (0.0491) (0.0509) High socioeconomic 0.0988* 0.0970* 0.0979* 0.0916* 0.0784 status (%) (0.0546) (0.0545) (0.0550) (0.0550) (0.0565) Mother’s education Primary −0.0877 −0.0841 −0.0861 −0.0822 −0.0718 (0.0907) (0.0918) (0.0898) (0.0894) (0.0848) Secondary −0.0226 −0.0179 −0.0199 −0.0223 −0.0289 (0.0548) (0.0541) (0.0549) (0.0545) (0.0532) Tertiary −0.1785 −0.1808 −0.1772 −0.1846 −0.1919* (0.1160) (0.1146) (0.1149) (0.1148) (0.1071) Dummy for region x x x x x Number of observations 1,089 1,089 1,089 1,089 1,089 R2 0.091 0.091 0.092 0.092 0.113 Note: All models estimated using ordinary least squares. Robust standard errors in parentheses. Controls include gender, age, socioeconomic status at age 15 years, mother's education, and dummies for region. Dependent variable equals 1 if working in medium- or high-skilled occupation and 0 if working in low-skilled occupation. ***p < 0.01, **p < 0.05, *p < 0.1. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, we leverage electronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. We follow the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on Forest Stewardship Council (FSC)–certified paper, with nearly all containing 50–100 percent recycled content. The recycled fiber in our book paper is either unbleached or bleached ­ using totally chlorine-free (TCF), processed chlorine–free (PCF), or enhanced elemental chlorine–free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://www.worldbank.org/corporateresponsibility. Stepping Up Skills in Urban Ghana  •  http://dx.doi.org/10.1596/978-1-4648-1012-1 The Skills Toward Employment and Productivity (STEP) survey is an initiative of the World Bank in cooperation with other development partners and nongovernmental agencies. STEP was carried out in more than 14 countries globally. In Ghana, the first phase of the survey, focusing on adults in urban communities, was undertaken in partnership with the University of Ghana’s Institute of Statistical, Social, and Economic Research; the Ministry of Education; the Council for Technical and Vocational Education and Training; and the Ghana Statistical Service. Stepping Up Skills in Urban Ghana provides detailed insights from the survey for policy makers. These insights cover areas including investments in early childhood education, the role of improvements in the quality of education, and the creation of incentives for economic actors to invest in on-the-job training to improve Ghana’s competitiveness and the well-being of its citizens. ISBN 978-1-4648-1012-1 SKU 211012