Photo : © Haja Faniry Razafimahenina MAURITIUS Inclusiveness of Growth and Shared Prosperity GPVDR AFRICA September 2015 Standard Disclaimer This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank 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 information 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. Copyright Statement The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750- 4470, http://www.copyright.com/.A ll other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. MAURITIUS GOVERNMENT FISCAL YEAR July 1 – June 30 CURRENCY UNIT (Exchange Rate Effective as of August 15, 2014) Currency Unit = Mauritian Rupees US$1.0 = Rs.30.60 WEIGHTS AND MEASURES Metric System Rivière du Rempart Pamplemoussses Port Louis Flacq Moka Plaines Wilhems Rivière Noire Grand Port Savanne Vice President : Makhtar Diop Senior Director : Ana L. Revenga Country Director : Mark R. Lundell Practice Manager : Pablo Fajnzylber Task Team Leader : Victor Sulla ABRREVIATIONS AND ACCRONYMS ABBREVIATIONS AND ACRONYMS BRP Basic Retirement Pension CPE Certificate of Primary Education CPI Consumer Price Index CSO Civil Society Organizations CSR Corporate Social Responsibility DCP Decentralized Cooperation Program EAP Eradication of Absolute Poverty EBIT Earnings Before Interest and Taxes ECD Early Childhood Development ECCE Early Childhood Care and Education GDP Gross Domestic Product GIC Growth Incidence Curve GNI Gross National Income HBS Household Budget Survey HIV/AIDS Human Immuno-deficiency Virus/  Acquired Immuno-Deficiency Syndrome ILO International Labor Organization IMF International Monetary Fund IP Invalid Pension ISIC International Standard of Industrial Classification IT Information Technology LSMS Living Standard Measurement Surveys MIC Middle Income Countries MUR Mauritian Rupee NEET Neither in Education nor in Training and Unemployed  NEF National Empowerment Fund NISP New Income Support Program NPF National Pension Fund NSF National Savings Fund OECD Organization for Economic Cooperation and Development  OLS Ordinary Least Squares PISA Progress for International Student Assessment PMT Proxy Means Test PPP Purchasing Power Parity ROA Return on Assets Rs Rupees SA Social Assistance SACMEQ  Southern and Eastern Africa Consortium for Monitoring Education Quality SAP Social Aid Program SBTC Skill Biased Technical Change SI Social Insurance SME Small and Medium-sized Enterprises SOC Standard Occupational Classification SP Social Protection SRM Social Registry of Mauritius SSA Sub-Saharan Africa STEM Science, Technology, Engineering and Math TFSIVG  Trust Fund for the Social Integration of Vulnerable Groups TIMSS Trends in International Mathematics  and Science Study UNESCO  United Nations Educational and Scientific and Cultural Organization USD United States Dollar WCP Widows and Children Pension iv MAURITIUS | Inclusiveness of Growth and Shared Prosperity TABLE OF CONTENTS TABLE OF CONTENTS Executive Summary xi A. Reforms to sustain growth.............................................................................................. xi B. Fast growth but low shared prosperity............................................................................... xi C. Resolving inefficiencies and looking forward....................................................................... xii 1 Chapter 1 INTRODUCTION A. Background................................................................................................................ 2 B. The concept of inclusive growth ...................................................................................... 2 C. Inclusive growth approach adopted in this report.................................................................. 3 5 Chapter 2 Mauritius Economic Development A. Economic reforms and economic outcomes.......................................................................... 7 B. Economic challenges................................................................................................... 11 C. Moving forward ........................................................................................................ 13 15 Chapter 3 How Growth and Economic Reforms Translates into Income Distribution of the Households A. Introduction.............................................................................................................. 17 B. Consumption expenditure and income.............................................................................. 18 C. Poverty trends in Mauritius............................................................................................ 18 D. Inequality trends in Mauritius......................................................................................... 20 E. Shared prosperity – why stayed behind.............................................................................. 23 F. International comparison of poverty, inequality, and shared prosperity ...................................... 24 27 Chapter 4 Poverty, Vulnerability and the Middle Class A. Introduction.............................................................................................................. 29 C. Who are the poor in Mauritius........................................................................................ 32 D. Who are the vulnerable and middle class in Mauritius............................................................ 41 47 Chapter 5 Causes of Poverty and Vulnerability Changes A. The role of growth and inequality in poverty changes............................................................ 49 B. Drivers of changes in poverty—decomposing poverty reduction................................................. 52 C. Linking growth, inequality, and poverty changes—poverty trace analysis..................................... 55 D. Looking ahead: how to tackle poverty while boosting the middle class ....................................... 57 61 Chapter 6 Social Protectionin Mauritius A. Mauritius’ social protection system.................................................................................. 63 B. Social assistance........................................................................................................ 65 C. Social insurance: contributory pensions............................................................................. 72 D. Areas of focused attention - Social protection .................................................................... 74 MAURITIUS | Inclusiveness of Growth and Shared Prosperity v TABLE OF CONTENTS 75 Chapter 7 Labor Market Characteristicsand Challenges A. Introduction.............................................................................................................. 77 B. Labor market outcomes................................................................................................ 77 C. Tightening of the Mauritian labor market and sectoral changes................................................ 78 D. The role of the high-tech sector...................................................................................... 82 E. Wages and earnings..................................................................................................... 84 E. Increasing inequality following the deteriorating of low-skilled wages........................................ 85 F. Rigidity of labor regulations........................................................................................... 88 G. Rising skills mismatches in Mauritius................................................................................. 89 H. Human capital is growing but not intergenerational mobility in human capital ............................. 90 I. Disadvantaged position of women in the labor market .......................................................... 93 J. Disadvantaged position of young workers........................................................................... 98 K. Mauritius labor market -areas of focused attention ............................................................ 102 105 Chapter 8 Evidence from Firm-Level Analysis A. Introduction............................................................................................................ 107 B. Challenges and advantages of the Mauritian private sector.................................................... 107 C. Creation of new firms................................................................................................ 111 D. Size and profitability of firms....................................................................................... 113 E. Financial structure and access to credit ......................................................................... 115 F. Relationship between profitability and financial structure..................................................... 119 References and Appendix 122 vi MAURITIUS | Inclusiveness of Growth and Shared Prosperity LIST OF FIGURES List of Figures Figure 1:   Doing Business ranking, 2007-13............................................................................................ 8 Figure 2:  Budget deficit and public debt (percent of GDP), 2005-1............................................................... 8 Figure 3:  Current account deficit and FDI (percent of GDP), 2005-1.............................................................. 9 Figure 4:  Real GDP growth rate, 2001-13.............................................................................................. 9 Figure 5:  Macroeconomic performance.............................................................................................. 10 Figure 6:  Decomposition of per capita value-added growth, 2000-12.......................................................... 12 Figure 7:  Factors in per capita value-added growth, 1990-2013................................................................. 12 Figure 8:  Consumption distribution in 2007 and 2012............................................................................. 19 Figure 9:  Income distribution in 2007 and 2012.................................................................................... 19 Figure 10:  Poverty in Mauritius over time........................................................................................... 19 Figure 11:  Lorenz curve over time................................................................................................... 20 Figure 12:  Growth incidence curve of household income......................................................................... 21 Figure 13:  Growth incidence curve of household consumption expenditures................................................. 21 Figure 14:  Median monthly earnings (Rs)............................................................................................ 21 Figure 15:  Poverty and inequality across countries................................................................................ 25 Figure 16:  Shared prosperity in Mauritius, international comparison........................................................... 26 Figure 17:  Middle class in Mauritius, 2007 and 2012............................................................................... 32 Figure 18:  Poverty incidence and the share of the poor by household size.................................................... 32 Figure 19:  Poverty by gender of head................................................................................................ 33 Figure 20:  Gender of head and marital status...................................................................................... 33 Figure 21:  Age pyramid and poverty, 2007.......................................................................................... 34 Figure 22:  Age pyramid and poverty, 2012.......................................................................................... 34 Figure 23:  Poverty by age groups..................................................................................................... 35 Figure 24:  Poverty by age of head.................................................................................................... 35 Figure 25:  Distribution of ethnic groups............................................................................................. 36 Figure 26:  Ethnicity and poverty...................................................................................................... 36 Figure 27:  Education of head by income quintiles in 2012....................................................................... 38 Figure 28:  Poverty by education of head............................................................................................ 38 Figure 29:  Poverty rates, by status of employment................................................................................ 39 Figure 30:  Distribution of poor, by status of employment........................................................................ 39 Figure 31:  Poverty by sector of activity............................................................................................. 40 Figure 32:  Poverty by occupation..................................................................................................... 40 Figure 33:  The middle class by selected demographic characteristics, 2012.................................................. 41 Figure 34:  The middle class by labor force and employment characteristics, 2012.......................................... 43 Figure 35:  The middle class by labor force status, 2007 and 2012.............................................................. 44 Figure 36:  The middle class sector of employment, 2007 and 2012............................................................ 45 Figure 37:  Occupation by income group, 2007 and 2012.......................................................................... 46 Figure 38:  Education by income group, 2007 and 2012........................................................................... 46 Figure 39:  Elasticity of poverty to consumption growth, 2007-12............................................................... 50 Figure 40:  Elasticity of poverty to inequality growth, 2007-12.................................................................. 50 Figure 41a:  Growth inequality decomposition : Income poverty change,2007-12........................................... 50 Figure 41b:  Growth inequality decomposition : Consumption poverty change, 2007-12.................................... 50 Figure 42:  Contribution to poverty reduction in percent, 2007-12.............................................................. 52 Figure 43:  Contribution to inequality increase in percent, 2007-12............................................................ 52 Figure 44:  Contribution to reduction in economic vulnerability reduction in percent, 2007-12............................ 53 Figure 45:  Contribution to poverty reduction by groups in percent, 2007-12................................................. 54 Figure 46:  Mauritius poverty trace curves (PTC) (consumption poverty)....................................................... 56 Figure 47:  Poverty and inequality projections, baseline scenario............................................................... 58 Figure 48:  Poverty simulations based on selected policy scenarios............................................................. 59 Figure 49:  Coverage of social protection, social insurance, and social assistance........................................... 63 Figure 50:  Simulated poverty and inequality impacts in the absence of SP, SI, and SA programs.......................... 64 Figure 51:  Distribution of elderly population, BRP old-age pension beneficiaries and benefits across deciles of income per equivalent adult, 2012................................................................................................. 66 MAURITIUS | Inclusiveness of Growth and Shared Prosperity vii LIST OF TABLES Figure 52:  Generosity of BRP old-age pension by decile of income per equivalent adult, 2012............................ 66 Figure 53:  Coverage of Social Aid beneficiaries and benefits across pre-transfer per AE income deciles, 2012......... 68 Figure 54:  Distribution of Social Aid beneficiaries and benefits across pre-transfer per AE income deciles, 2012...... 68 Figure 55:  Profile of Social Aid beneficiaries and non-beneficiaries, 2012..................................................... 69 Figure 56:  Distribution of benefits (targeting accuracy) of widows and children, disability, and other social pensions, 2012........................................................................................................ 70 Figure 57:  Share of Social Aid beneficiaries who also receive benefits from other........................................... 72 Figure 58:  Coverage of contributory pensions by post-transfer income decile, 2007 and 2012............................ 73 Figure 59:  Mean benefit amount of contributory pensions by post-transfer income decile, Rs. in constant 2006 prices.............................................................................................................. 73 Figure 60:  Labor market:  main indicators.......................................................................................... 77 Figure 61:  Tightening labor market in Mauritius................................................................................... 78 Figure 62:  Role of the foreign workers in Mauritius............................................................................... 80 Figure 63:  Sectoral composition of employment shifts toward tertiary sector................................................ 81 Figure 64:  Public vs. private employment, by shares............................................................................. 82 Figure 65:  High-tech vs. overall employment and wage changes................................................................ 83 Figure 67:  Change in average real wages 2001-12................................................................................. 85 Figure 68:  Labor characteristics by consumption quintiles 2007-12............................................................ 86 Figure 69:  Returns to educational investment difference from no education................................................. 87 Figure 70:  Change in log real monthly wage by percentile, 2001 vs. 2012.................................................... 87 Figure 71:  Smoothed changes in employment by occupation. 2001-09......................................................... 87 Figure 72:  Measuring skills mismatches in Mauritius, skills mismatch index................................................... 89 Figure 73:  Educational attainment 2001-12......................................................................................... 90 Figure 74:  Labor-force status by highest educational level—2012............................................................... 90 Figure 75:  Importance of family background for schooling completed......................................................... 93 Figure 76:  Main indicators:  gender differences.................................................................................... 94 Figure 77:  Females inactivity probability........................................................................................... 95 Figure 78:  Explained and unexplained gender wage gap......................................................................... 97 Figure 79:  Main labor indicators by age group..................................................................................... 100 Figure 80:  Youth unemployment rates, international ............................................................................ 101 Figure 81:  Mauritius competitiveness indicators, country rating (lower is better), 2013/14.............................. 108 Figure 82:  Obstacles of doing business in Mauritius, 2009 ...................................................................... 110 Figure 83:  Number of new incorporations over Time............................................................................. 111 Figure 84:  Ease of starting a business............................................................................................... 111 Figure 85:  Industry compositions for new incorporations and other firms, 2007-12......................................... 112 Figure 86:  Distribution of sales (2001-12).......................................................................................... 113 Figure 87:  Profitability density of firms (2007-12 average)...................................................................... 115 Figure 88:  Institutional framework for getting credi............................................................................. 116 Figure 89:  Access to financial services.............................................................................................. 117 Figure 90:  New credit for firms over time.......................................................................................... 118 List of Tables Table 1:  Gini inequality decomposition by income sources (Shapley value approach)....................................... 22 Table 2:  Shared prosperity within Mauritius, selected groups.................................................................... 24 Table 3:  Occupation and education................................................................................................... 39 Table 4:  Macro projections, baseline scenario...................................................................................... 57 Table 5:  Composition of SA benefits, 2013........................................................................................... 65 Table 6:  Probability of accessing further education—selected variables........................................................ 92 Table 7:  Main indicators by gender, 2001-12 change............................................................................... 95 Table 8:  Marginal effects of background characteristics on probability of being inactive................................... 96 Table 9:  Probability of Being NEET (15-24)......................................................................................... 102 Table 10:  Firm size, age, and profitability by industry (2007-12 average).................................................... 114 Table 11:  Firm size and financial structure by industry (2007-12 average)................................................... 116 Table 12:  Firm size, financial structure, and profitability (2007-12 average)................................................. 119 viii MAURITIUS | Inclusiveness of Growth and Shared Prosperity AKNOWLEDGEMENTS Acknowledgements This report was prepared by a core team comprising Victor Sulla (Task Team Leader) , Rafael Munoz Moreno, Carlos Da Maia, Leora Klapper, Peter Van Oudheusden, Melis U. Guven, Denis Nikitin, Virendra Polodoo, Patrick Leon Randriankolona, Jacopo Mazza (The University of Manchester) and Timothy Heleniak (University of Maryland). The report was prepared on request of Mauritius Ministry of Finance and coordinated by Mr. Gerard Bussier. The report benefited from discussions with government officials, development partners, and practitioners outside the government. The analyses were conducted in close collaboration with Statistics Mauritius. A number of officials of Mauritius Statistics Office were instrumental in supporting and facilitating the work. The team would like to thank Mrs. Aimee Kai Suet (Director of Statistics Mauritius) and Mrs. Meera Ganoo (Deputy Director of Statistics Mauritius) for their fruitful collaboration, and especially the poverty statistics team. The peer reviewers for the report are Pedro Olinto and Sailesh Tiwari. The team is thankful to Pablo Fajnzylber (Practice Manager), Julio E. Revilla (Program Leader), Thomas Buckley (Country Program Coordinator), John Panzer (Practice Director) and Mark R. Lundell (Country Director) for their continued support through the preparation of the report. The ADePT Software developed by the World Bank was used to produce most of the statistical outputs presented in the report. Madeleine Chungkong and Richard Alm edited the report. MAURITIUS | Inclusiveness of Growth and Shared Prosperity ix Photo : © Photo : © Haja Faniry Razafimahenina x MAURITIUS | Inclusiveness of Growth and Shared Prosperity EXECUTIVE SUMMARY Executive Summary A. Reforms to sustain growth B. Fast growth but low shared prosperity 1. Mauritius is a high middle-income country with low 4. he economic changes of the 2000s led to increasing T levels of poverty and inequality. The headcount income inequality and deterioration in the shared- poverty level was 6.9 percent in 2012; measured by the prosperity indicators. The economy’s polarization international standard of US$2 per day (PPP), poverty was associated with a structural transformation from was less than 1 percent. On inequality, Mauritius labor-intensive industries to services and knowledge- also fared well compared to its peer middle-income intensive industries. Declines in agriculture and countries, with a Gini coefficient 0.39 in 2012. On the traditional textile industries led to a deterioration negative side, Mauritius’ growth has not been equally of the primary and secondary sectors, while shared, despite the general improvement in welfare in accommodations and wholesale trade have been the 2000s. In terms of per capita income growth rates at the forefront of a booming tertiary sector. The of the bottom 40 percent of the population, Mauritius financial and construction sectors have also expanded. ranked 63rd among 84 developing countries. High-tech industries have grown in recent years, but they are still marginal. The demand for traditional 2. ince independence in 1968, Mauritius’ economic S and low-skilled occupations has declined, and migrant performance has been strong, associated with diligent workers have taken many blue-collar domestic jobs, economic policies, productivity growth, and human- filling vacancies in unattractive occupations that no capital accumulation. Despite the general success longer appeal to Mauritian job-seekers. in the early 2000s, the country developed many inefficiencies, including restrictive regulations in trade 5. Growing demand for highly skilled workers, combined and labor and deficiencies in macro management. GDP with insufficient supply, led to an increase of almost growth reached its long-run potential, and labor- 30 percent in the skills-mismatch index between 2001 market indicators started to deteriorate. Structural and 2010. The Enterprise Survey points to inadequate and institutional challenges led the Government to skills as a major challenge for the larger enterprises. liberalize its industrial, trade, and labor policies in Foreign workers also substituted for Mauritians in the mid-2000s. It removed many bureaucratic and many low-skill occupations. Some workers who lost regulatory obstacles and introduced an array of their jobs were forced to look for employment in improvements in the business environment. Along more advanced sectors, where higher education is with these structural reforms, the Government took at a premium, but their skills were not necessarily a bold approach to dealing with high public deficits adequate. Unmet demand led to a disproportionate and rising public debt, removing the medium-term increase in relative wages for skilled workers. The threat that an unsustainable fiscal course posed to highest salaries are in the services sector, and the macroeconomic stability. This was complemented by a trend remains upward. Compared to agriculture, prudent monetary policy and flexible exchange rates, tourism and the tertiary sector paid around 40 which helped build considerable foreign reserves. percent more in 2012, while manufacturing salaries were 30 percent higher. At the same time, high skills 3. T he reforms had an immediate, palpable impact and high-tech jobs growth were important sources of on Mauritius’ economic performance. GDP growth employment growth, starting in the second half of the accelerated, associated with improvements in decade. The STEM1 and high-tech occupations also pay exports and the current account, increases in FDI, and considerably higher salaries. Labor-market outcomes improvements in the main labor market indicators. are worse among the poor, and their situation has However, the reforms did not resolve the issue of deteriorated, leading to widening disparities. relatively low and falling productivity, and TFP’s contribution to growth was limited. 6. Reforms boosted job creation and the entrance of new firms in the mid-2000s. However, SMEs face challenges to being profitable and raising their market share, and they report difficulties in finding qualified employees. Mauritian firms are relatively small, not 1 STEM is an acronym referring to the academic disciplines of science, technology, engineering and mathematics. MAURITIUS | Inclusiveness of Growth and Shared Prosperity xi EXECUTIVE SUMMARY very profitable, and generally lack growth potential. 10. Poverty is especially high among the unemployed, but New firms face even more severe challenges. They the inactive group makes up the largest share of the have been more likely to enter the construction poor. In addition, the working poor are a relatively and services industries. Access to financing is still large group, representing 26 percent of Mauritians a major obstacle. A majority of new firms generate living in poverty. White-collar occupations are little revenue, and they are severely leveraged and associated with lower poverty, while poverty among more risky. Around 70 percent of small firms and blue-collar workers is high and has tended to increase roughly 55 percent of medium and large firms are over time. Better educated individuals have better highly leveraged. Small firms are more likely than chances than the poorly educated to get the best other businesses to be unprofitable. Compared to jobs. Poverty also varies widely across occupations. the services industry, firms are more likely to be unprofitable in agricultural and textiles industries. 11. ncreasing economic vulnerability is a worrisome I trend in Mauritius. The share of the population 7. D espite some improvements in labor regulations in considered economically vulnerable increased from the 2000s, wage determination in Mauritius depends 10.2 percent in 2007 to 12.7 percent in 2012. The heavily on non-market regulations and collective share of the population in the middle class has also bargaining. A puzzling aspect of the Mauritian economy declined—although the majority of the population is is a disproportionate increase in real wages in the still considered middle class. Our analysis suggests public sector. Increases of 23.5 percent have been that skilled employment and quality tertiary education observed in the public sector, compared with only 7 are the main attributes for reaching the upper middle percent in the private sector. class in Mauritius. In addition, employment in public administration or public enterprises is key, with 75 8. Rising income inequality and lagging shared prosperity percent of those employed in these occupations had adverse impacts on relative poverty and inequality making it to the upper middle class, compared in Mauritius. Although absolute poverty fell from 8.5 with 53 percent in private enterprises, 43 percent percent to 6.9 percent in 2007-12, relative poverty2 in export-oriented enterprises, and 32 percent in rose from 8.5 percent to 9.8 percent. Income household services. Vulnerability, however, is growing inequality, measured by the Gini coefficient, increased in agriculture and industry and is an attribute of those from 0.36 to 0.39.3 According to our analysis, the with lower labor-force participation. reduction of absolute poverty in Mauritius would be almost twice as large if growth were better shared, 12. As opposed to receiving most income from transfers and inequality would not have worsened. Economic provided through various forms of government growth and declining inequality are equally important assistance, working leads to higher shares of the for the reduction and possible eradication of poverty population becoming middle class and lower shares in Mauritius. being poor or vulnerable. Highly skilled occupations are also key to gaining middle-class status. More than 9. The poor are generally trapped in poverty due to a 70 percent of managers, professionals, technicians/ weak connection to the labor market, demographic associates, and clerical workers are upper middle issues, low education, and health challenges. The class, but less than 44 percent of those in skilled poor tend to live in large households, often headed agricultural, trade, and elementary occupations are by a single parent. Poverty has a predominantly young upper middle class. face, increasing among households headed by younger people. Overall, Sino-Mauritians are the least poor ethnic group, and they have experienced a large C. Resolving inefficiencies and looking forward decline in the incidence of poverty. People living in households headed by more highly educated people 13. Moving from middle-income to high-income status will tend to earn higher incomes than their less educated require a careful review of an economic model that counterparts. As a result, poverty is highest among has worked in the past. When Mauritius will be able people living in households with heads who did not to become a high-income country will depend on its complete any education level. ability to improve the labor force’s skill set, develop infrastructure, and further improve the business environment to attract FDI and generate domestic investment. Inclusiveness remains the main challenge 2 Relative poverty defined as 50 percent of median consump- for the current growth pattern. tion per adult equivalent. Gini coefficients presented in this report are estimated on an 14. Rapid poverty and vulnerability reduction requires 3 income or consumption per capita basis, while official inequa- lity figures are estimated on total income or consumption. For more inclusive growth. Micro-simulation analysis this reason, official estimates of inequality are higher. suggests that reducing and eventually eradicating xii MAURITIUS | Inclusiveness of Growth and Shared Prosperity EXECUTIVE SUMMARY Photo : © Haja Faniry Razafimahenina poverty in Mauritius will depend on a two-fold 16. Being employed is obviously a key factor in achieving combination of policies—first, improved targeting middle-class status, while being unemployed is among efficiency in social protection and, second, lower the most telling vulnerability factors. There is a clear unemployment and greater productivity. Targeted correlation between increased education and higher policy interventions could boost poverty reduction in shares in the middle class, especially for those with Mauritius. Investment in the following areas should a secondary or higher education. Tertiary-education boost shared prosperity in Mauritius: expansion needs to focus on innovation and R&D. a) Long-term productivity improvement b) Fix inadequate labor regulations 15. Low productivity remains a major challenge for 17. The labor market needs to foster flexibility and reward private-sector development in Mauritius. Policies higher productivity. Annual salary compensation and designed to upgrade infrastructure, support R&D and remuneration orders are designed to reduce disparities, innovation, advance public-sector efficiency, and but they rarely impact wage determination in the further improve the business environment will boost intended way (see labor section of the report). The productivity. A new wave of public-sector reforms thresholds are set at very low levels by international could raise accountability at all levels and improve standards—on average, 22 percent of the wage. planning, procurement, and management processes In addition, the national tripartite negotiations across the system. Efficient country-level monitoring set up in 2010 make it more difficult to maintain and evaluation (M&E) systems should be developed, competitiveness. In the longer term, Mauritius has supporting evidence-based policymaking. Public to find an appropriate balance between worker utilities need to become more efficient and have their protection and labor-market flexibility. infrastructure upgraded. Reforms in public enterprises will create fiscal space for more productive spending. Improved road infrastructure and further development of public transport are also suggested. MAURITIUS | Inclusiveness of Growth and Shared Prosperity xiii EXECUTIVE SUMMARY c) Reversing growing skills mismatches f) Addressing youth unemployment and and boosting education vulnerability 18. Demand for highly educated workers has not been met, 21. Young people between ages 15 and 24 experience resulting in mismatches between the supply of available substantially worse labor-market outcomes than the skills and the demand for skills. The mismatches rest of Mauritius’ population. Youth unemployment put upward pressure on unemployment rates. This rates are especially high, and young workers are report finds that skills mismatches grew by 30 percent particularly vulnerable to labor-market fluctuations. during 2000s, signaling an urgent need for policies to Compared to the rest of the population, young people reduce the mismatches and support the transition to display a more volatile pattern of employment, high-tech and services-oriented industries. Resolving reflecting a higher sensitivity to the economic cycle. the problem of skills mismatches remains the main On a positive side, the portion of young individuals challenge for Mauritian development. Education is who are neither in education nor in training and a fundamental prerequisite for individual economic unemployed (NEET) has decreased considerably success. The share of employed workers with tertiary since 2005. In addition, the number of young people education more than doubled in Mauritius. However, in education has increased, reaching a high point in the country has considerable scope to improve its the past two years. educational system, and educational reforms are needed to provide people with appropriate and g) Resolving challenges of the social relevant skills. Both the SCD and this report find that protection system the lack of adequate skills has a negative impact on the inclusiveness of growth, with the more vulnerable the 22. The expansion of social protection (SP) programs has most affected by educational deficiencies. Education not been sufficient to prevent an increase in income and skills should be improved and realigned toward inequality. Higher incomes from work and self- the needs of the business sector. employment among initially better-off groups led to the greater inequality. Mauritius has operated a wide d) Overcoming lack of intergenerational range of labor-force activation programs for some mobility time, but they are small in coverage, fragmented, lack mutual coordination, and have few robust linkages 19. A lack of intergenerational mobility has adverse effects to SP programs. The Government has undertaken for the overall economy’s growth potential. Our meaningful steps toward greater coordination in SP analysis finds a strong influence of family background programs, but further improvements are needed. The on post-secondary education. Parents of tertiary Social Aid program—the only program in Mauritius that educated individuals are better educated and richer specifically targets the poor—has been the leading than the rest of the society and these differences are contributor to poverty reduction; however, it could not disappearing with time. The offspring of well- be scaled up and significantly improved. educated and rich families will invest in education, increasing their probability of preserving their favorable economic position. Meanwhile, poorer and worse-educated parents will not be able to offer the same opportunities to their children, perpetuating the social structure over time. e) Fixing gender disparities 20. Major gender disparities are evident in the Mauritian labor market. Women experience substantially lower employment levels and higher unemployment and inactivity rates than their male counterparts. These gaps have been falling, an encouraging sign of convergence in Mauritius. The gender wage gap in Mauritius is severe and, unlike the gaps related to labor force status, shows no sign of decreasing. In fact, it widened in recent years. Even when comparing men and women with the same education level, age, potential work experience, and sector of employment, women still earn significantly less than men. xiv MAURITIUS | Inclusiveness of Growth and Shared Prosperity Chapter 1 INTRODUCTION INTRODUCTION A. Background B. The concept of inclusive growth 23. By the mid-2000s, the Mauritian economy was facing 27. The need for “inclusive growth” has now been structural challenges and stagnating exports. The recognized in many countries. During the past decade, investment rate, which had peaked at more than 30 China, India, and Mozambique and many other percent of GDP in 1994, slumped to an average of just developing nations had stable and often high rates of 22 percent in 2001-05, diminished by a retrenchment economic growth. However, the extent to which this of both public and private investments. Declining growth has been shared differs greatly, and income investments and labor-market rigidities led to a rise inequality has increased in many countries. Regional of the unemployment rate from less than 3 percent in inequalities have tended to increase sharply in such 1991 to 9.5 percent in 2005. In addition, deteriorating places as Ghana and Nigeria, where the northern parts external conditions and a “triple trade shock”—the of the country have been traditionally left behind.5 losses of textile and sugar preferences and soaring oil Yet, many Latin American countries have successfully prices—put pressure on the balance of payments and reduced income inequalities. Cash transfers played slowed economic growth. an important role, along with increases in other government investments and macroeconomic 24. The Government confronted this situation in the mid- stability. 2000s by implementing a set of bold reforms—for example, opening the economy to further competition, 28. The World Development Report 2013 focuses on eliminating distortions between the EPZ and the rest labor-market institutions. The report highlights the of the economy, significantly eliminating tariffs and importance of looking beyond macroeconomic growth trade barriers, simplifying the tax system with low and paying attention to distributional concerns and income-tax rates set at 15 percent, and simplifying the extent to which people feel they participate labor and business regulations. in making the decisions that shape their life. The micro dimension captures the role of structural 25. The structural transformation accelerated the transformation in economic diversification and economy in Mauritius, and the rewards came quickly competition. Inclusiveness and shared prosperity are in the form of an increased FDI, reinvigorated growth the essential ingredients of any successful growth in high value added sectors (i.e., ICT, finance), strategy. Inclusiveness is a multidimensional concept and lower unemployment. In addition, there has that encompasses equity, equality of opportunity, and been a noticeable progress on measures of human protection in market and employment transitions. It development, including life expectancy, maternal entails changes in market structure, access to finance, and infant mortality, school enrollment measures, discrimination in labor and product markets, and and access to primary education for both genders. conditions in the informal sector. The inclusive growth Mauritius is one of the few African countries that approach takes a longer-term perspective, focusing has accomplished remarkable results on the MDG on increasing incomes for traditionally excluded indicators in just 15 years, with six of eight specific groups through productive employment rather than goals more or less achieved. Extreme poverty is redistribution. almost negligible, the net primary school enrollment ratio has risen 97 percent (2011), life expectancy has 29. A wide range of literature covers various aspects of increased, and infectious diseases such as malaria, inclusive growth and shared prosperity. As stated above, polio, diphtheria, typhoid, and cholera have been inclusiveness is a multidimensional process related virtually eradicated.4 to economic factors, human capital, and political and social dimensions. It combines improvement in 26. The 2005 economic and trade reforms led to a average level of various indicators with a distributional substantial reallocation of resources, with clearly component. This multidimensional approach takes positive effects on economic growth and human into consideration inequality in economic factors, development indicators. Our report looks into the human-capital accumulation, and political and social inclusiveness of these positive changes to determine dimensions in the relatively long period of time and whether all groups of population have benefited in a sustainable manner. For growth to be inclusive, equally from the recent growth. The report explores productivity must be improved, new opportunities for how economic changes have affected households, employment created, and the gains should be shared workers, and firms. across population groups. 4 “Do children in Mauritius have Equal Opportunities in Educa- 5 Regional disparities might not be the case in Mauritius, while tion?” 2012. Statistics Mauritius with the support of the World ethnic disparities should be explored if data were to become Bank. available. 2 MAURITIUS | Inclusiveness of Growth and Shared Prosperity INTRODUCTION ii. Opportunities created for employment C. Inclusive growth approach adopted in this report 33. Employment creation has become a priority for the Government, which sees it as the best way to ensure broad-based economic growth and social cohesion. 30. The overall situation has improved in Mauritius on For a particular group of individuals—those with a various fronts, but additional analysis is needed to limited supply of certain types of labor skills—the measure the distribution of these positive changes. constraints are related to the capacity of individuals The IMF studied the inclusiveness of growth in Mauritius rather than the business or labor environment. This in the 2000s using household survey data.6 It was a situation calls for an in-depth analysis of labor-market period of profound structural change in the Mauritian challenges that determine individuals’ resources. economy, linked to the loss of sugar preferences and This section analyzes the main labor-market issues in the phase-out of textile trade preferences (i.e., Mauritius, such as employment, sectorial movement dismantling of the Multi-Fiber Agreement). As the and changes over time, and unemployment rates services sector emerges as a new engine of growth, the and wages. Finding work is challenging for many question is whether the benefits of economic growth youths, either due to lack of the skills demanded by continue to be widely shared by various segments of a modern economy or labor-market rigidities. The the population. The authors find evidence pointing to report systematically looks at the issues of female a more skewed distribution of the benefits of growth, job participation, intergenerational mobility, skills possibly because of fundamental structural changes mismatches, and other issues and disparities in the in the Mauritian economy. labor market. 31. To enhance the policy relevance of the analysis, iii. How reforms have accelerated economic growth the uneven distribution needs greater emphasis; at firm level especially now, with Mauritius entering the club of developed economies. The report looks into the 34. In comparison to studies using only aggregate data, inclusiveness of growth in Mauritius, taking into analyses of firm-level data have the potential to more consideration its three main dimensions: (i) gains credibly identify in more detail the effects of certain and shared prosperity associated with the growth of policies and describe the mechanisms behind the incomes, (ii) opportunities created for employment, effects of the policies. This would serve to identify and (iii) inclusiveness in firms’ profitability. the main systemic factors behind the lagging sectors and how government policies can adequately support i. Gains and shared prosperity associated with them. The objective of this section is to improve the growth of incomes understanding of firms’ performance and inclusiveness of growth by analyzing the determinants of firm 32. This section includes an analysis of decade-long trends profitability, size, and sectoral dynamics. in income distribution, focusing on economic growth and its inclusiveness in Mauritius. It looks at the 35. To assess the inclusiveness of growth in all these poverty, inequality, and shared prosperity indicators. dimensions, the report systematically looks at The report discusses the sources of changes in poverty patterns of household incomes and consumption and vulnerability in Mauritius, focusing on how prices, growth (chapter 3), analyses the causes of changes jobs, income, and social-protection efficiency have in poverty and vulnerability (chapter 4), defines and impacted poverty and economic vulnerability. The analyzes economic vulnerability and middle class trends report defines the scope of the middle class in (chapter 5), looks at the role and efficiency of the Mauritius and follows the evolution of the middle class social-protection system (chapter 6), conducts detailed over time. It analyzes the impact of the economic analysis of labor market (chapter 7), and examines firm changes on the size and characteristics of the poor profitability and challenges (chapter 8). and vulnerable. 6 Antonio C. David and Martin Petri, 2013 “Inclusive Growth and the Incidence of Fiscal Policy in Mauritius — Much Progress, But More Could be Done.” MAURITIUS | Inclusiveness of Growth and Shared Prosperity 3 INTRODUCTION Photo : © Haja Faniry Razafimahenina 4 MAURITIUS | Inclusiveness of Growth and Shared Prosperity Chapter 2 Mauritius Economic Development Since independence in 1968, Mauritius’ economic performance has been strong, associated with efficient government, diligent economic policies, human capital accumulation, and fast growth in FDI, tourism, and exports. However, structural and institutional challenges led the Government of Mauritius to liberalize its industrial, trade, and labor policies in the mid-2000s. The reforms had an immediate, positive impact on the Mauritian economy’s performance but brought increasing income inequality. Traditional textile and agriculture sectors contracted while tertiary sectors expanded. Among the economic challenges are relatively low productivity that stunts growth, widening skills mismatches, and rigid labor regulations. Moving forward, Mauritius’s GDP growth rate appears to be close to its long-run potential. Achieving high- income country status will depend on Mauritius’ ability to improve the labor force’s skill set, develop infrastructure, and further improve the business environment to attract FDI and generate domestic investment. Inclusiveness remains the main challenge of the current growth pattern and will be analyzed in later chapters. CHAPTER 2 - Mauritius Economic Development led economies, such as China, the reform program A. Economic reforms and economic outcomes expanded to the entire economy the favorable tax and regulatory environment that was previously provided 36. Mauritius has been characterized by strong economic exclusively to Export Processing Zone (EPZs), It also performance. In the 1970s and 1980s, it diversified eliminated 95 percent of tariff lines on a phased from a mono-crop economy dominated by sugar cane basis over three years, lightened regulatory burdens, to a more diversified one based on sugar, textiles and and developed and implemented sector strategies to garments, and tourism. This economy has expanded reduce costs and increase competitiveness in existing continuously since the 1990s. From 1992 to 2005, it and new sectors. grew at an annual average of 5.1 percent. By the mid- 2000s, however, the rapid growth in low-wage, low- 40. Removing bureaucratic obstacles improved the business skill, and labor-intensive exports that had powered environment. According to the Doing Business survey the Mauritian miracle in the 1980s ran out of steam. for 2007, it took 49 days to start a business in Mauritius, compared to 27 days in Mexico, nine days in Turkey, 37. Mauritius’ economic development, successful since and an average of 16.6 days in the OECD countries independence, confronted structural challenges and (Figure 1) Mauritius’ score in the Difficulty of Firing stagnating exports in the mid-2000s. Labor shortages Index nearly doubled the OECD average and was even had emerged in the early 1990s, driving up real wages above average for the SSA region. To respond to this, and undermining competitiveness in low-skill sectors. the Government passed the Business Facilitation Act The investment rate, which peaked at over 30 percent of 2006 to spur investments and creation of new of GDP in 1994, slumped to an average of just 22 businesses. As a result, Mauritius’ ranking in the percent of GDP in 2001-05, reflecting a retrenchment World Bank’s Doing Business Index improved to 17th of both public and private investment. Rigidities in in 2010. the economy made it difficult to transfer resources to emerging sectors. After 2000, exports stagnated, 41. Along these structural reforms, the Government also investment slumped and, reflecting also labor-market took a bold approach to dealing with high public deficits rigidities, unemployment rose to 9.6 percent in 2005, and a rising public debt, removing the medium-term up from less than 3 percent in 1991. Then in the middle threat to macroeconomic stability of an unsustainable of the decade, the country suffered a “triple trade fiscal course. The Government sought to broaden the shock” with the loss of textile and sugar preferences tax base by introducing a flat tax rate of 15 percent and soaring oil prices, further hurting economic for both personal and corporate income, with no growth and putting the balance of payments under exemptions. In addition, a unified and strengthened pressure. Mauritius Revenue Authority enlarged the tax base and simplified tax procedures. As a result, tax revenue 38. Economic reforms have accelerated since 2005. A new increased from 17.9 percent of GDP in 2005 to 18.9 Government that came to power in 2005 confronted percent in 2012. Many low-income tax payers actually this situation and implemented a bold set of saw their tax payments dwindle. Public revenues, reforms, including opening the economy to further which had averaged 19.7 percent of GDP between competition. The Government focused on halting the 2000 and 2005, increased to 21percent in 2008, and a slide by raising competitiveness, promoting higher prudent fiscal stance reduced the public deficit from value-added exports, investing in infrastructure an average of 5.8 percent of GDP between 2000 and and education, and reforming industrial relations. A 2005 to 2.7 percent in 2008 (Figure 2). As a result structural reform program was implemented to raise of these efforts, public debt was reduced from 65 the efficiency of the private sector and modernize percent of GDP in 2005 to 52 percent in 2008. A new the public sector for a post-regulatory world. The Public Debt Management Act enshrined the medium- Government significantly reduced custom tariffs and term sustainability of the public finances, mandating trade barriers, simplified the tax system with low a public-debt threshold of 60 percent of GDP, with a income tax rates set at 15 percent, and streamlined reduction to 50 percent by 2018. labor and business regulations. 39. The Government liberalized its industrial and trade policies. Interventions and regulations had created a Tax revenue increased biased structure of incentives, with trade protection favoring domestic production rather than exports, from 17.9 percent inflexible regulations deterring new sectors, and complex incentive schemes and high compliance costs of GDP in 2005 to 18.9 percent in 2012. favoring large rather than small firms. In an effort reminiscent of those made by successful export- MAURITIUS | Inclusiveness of Growth and Shared Prosperity 7 CHAPTER 2 - Mauritius Economic Development 42. Fiscal reform was complemented with prudent monetary 43. In general, the reforms had immediate and positive policy and flexible exchange rates, which helped build impacts on the performance of the Mauritian economy. considerable foreign reserves. After current account A more favorable business environment, a rise in surpluses averaging 2 percent of GDP in 2000-04, FDI, and stronger macroeconomic policies led to a deficits averaged 7.5 percent of GDP between 2005 progressive improvement in economic growth as and 2008, reflecting negative contributions from well as the formation of new sectors. GDP growth declining sectors. The impact of the external sector on rose from 1.5 percent in 2005 to 5.5 percent in 2007 real domestic income was even larger after taking into (Figure 4); private investment increased from 8.0 account the 9.2 percent deterioration in the terms percent of GDP in 2005 to 17.7 percent in 2007; FDI of trade between 2005 and 2008 because of lower tripled from 1 percent of GDP in 2005 to 3 percent in prices of textile and sugar products and higher food 2007. As a result, net job creation accelerated, and and oil import prices. The worsening current account the unemployment rate fell from 9.6 percent in 2005 was financed through a doubling of FDI to Mauritius to to 7.2 percent in 2008. The services and construction 2.6 percent of GDP and net short-term capital inflows sectors were the main contributors to growth between averaging close to 6 percent of GDP, related to the 2005 and 2008. The tertiary sector’s share of the of development of Mauritius as an international financial the economy rose by more than 3 percentage points center (Figure 9). As a result, balance of payments between 2000 and 2004 and again between 2005 and turned positive and international reserves increased 2008, reaching an average of 60.2 percent of GDP. 30 percent to around US$1.76 billion (3.4 months on The diversification served to compensate for slower imports) in 2008. However, inflation picked up, rising growth in such traditional sectors as agriculture, from an average of 4.9 percent in 2000-04 to 8.1 which underwent annual contractions of 1.4 percent percent in 2005-08 as a result of an accommodating between 2005 and 2008. monetary policy, double-digit increases in import prices, and increases in some excises. Figure 1: Doing Business ranking, 2007-13 60 50 40 30 20 10 0 2007 2008 2009 2010 2011 2012 2013 Source: World Bank, Doing Business indicators. Figure 2: Budget deficit and public debt (percent of GDP), 2005-1 60 50 40 30 20 10 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 Public Debt Budget Deficit Source: Statistics Mauritius. 8 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 2 - Mauritius Economic Development Figure 3: Current account deficit and FDI (percent of GDP), 2005-1 16 Current Account Deficit FDI 14 12 10 8 6 4 2 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source: Statistics Mauritius. Figure 4: Real GDP growth rate, 2001-13 7 6 5 4 3 2 1 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source: Statistics Mauritius. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 9 CHAPTER 2 - Mauritius Economic Development Box 1: Outline of the reforms in Mauritius In the mid-2000s, a bold package of policies and institutional reforms deepened the efforts initiated in the preceding years and it aimed at addressing some politically sensitive reforms as well. The ambitious reform program was structured around four pillars: (i) fiscal consolidation and public sector efficiency; (ii) trade competitiveness; (iii) improving the investment climate, and (iv) widening the circle of opportunities. The empowerment program, which included a workfare scheme emphasizing training and skill-building, supported the reforms. The list below outlines main reforms introduced in Mauritius in this period: A. Consolidating Fiscal Performance and Improving Public-Sector Efficiency • Fiscal rule (public debt legislation) • Public financial management reforms • Revamping of tax system (single flat tax on personal and corporate income) B. Enhancing Competitiveness • Tariffs reduced • Regulations for export processing zone (EPZ) and non-EPZ firms unified • Improving telecommunication services C. Improving the Business Climate • Business registration, regulation, and insolvency revamped through new legislation • Restrictions on land acquisition by foreigners eased • New labor-market legislation for widening the circle of opportunity through greater participation D. Inclusion and Sustainability • The National Empowerment Foundation as umbrella institution minimizing social costs of economic transformations • Education reform launched • Background analytical work to improve efficiency of the SP system Figure 5: Macroeconomic performance 12 35 Unemployment & GDP growth rates 10 30 Investment % GDP 25 8 20 6 (%) 15 4 10 2 5 0 0 1990 1995 2000 2005 2010 2015 Year GDP growth rate Unemployment rate Investment % GDP Source: Adapted using data from Statistics Mauritius. 10 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 2 - Mauritius Economic Development The economic reforms have reversed some of the negative trends. Between 1990 and 2005, Mauritian B. Economic challenges GDP growth was extremely volatile. Exports stagnated. Investment as a share of GDP peaked at 46. Moving from middle-income to high-income status 30.4 percent in 1994 and gradually fell afterwards, will require a careful review of an economic model that The unemployment rate tripled from 3.3 percent in worked in the past. The Government is expecting to 1992 to 9.6 percent in 2005. achieve high-income status in the medium term while This deterioration in the economic performance was ensuring inclusive growth. However, as presented in in part associated with the loss of sugar preferences the SCD analysis, the economy’s performance since and the phase-out of textile trade preferences. As 2008 has been less robust than expected, and the a result of the mid-2000s economic reforms, the country is facing challenges on several fronts. GDP investment rate showed a positive trend since 2005, growth is losing steam as the positive impact of unemployment subsided, and GDP growth became less reforms wanes. Job creation remains slow, income volatile, averaging 4.4 percent per year between 2006 inequality is increasing (as will be shown later in the and 2012. report), and economic vulnerability is not falling. 44. These reforms facilitated a proactive approach to the 47. The economy is showing some signs of fragility, extraordinary challenge of the 2008-10 global economic reinforced in large part by an uncertain external crisis. With a small domestic market, dependence on environment. Investment has been on a downward Europe for exports and FDI, and heavy reliance on trend while unemployment rose from 7.8 percent in imports, Mauritius was exposed to the financial and 2010 and 8.0 percent in 2013. This provides further economic downturn that hit the world economy and evidence of the labor-market rigidities that have particularly Europe. Real GDP growth fell from 5.5 not been addressed and the increasing difficulties in percent in 2008 to 3.1 percent in 2009. The tourism absorbing unskilled and semi-skilled workers as the sector was severely hit, with earnings falling from economy transitions to services and knowledge-based US$997 million in 2008 to US$763 million in 2009. industries. The gap in firm efficiency has widened After the reforms implemented in 2006-08, however, both across and within sectors. While Mauritius hosts Mauritius was in a relatively strong position to cushion the leading regional firms in many sectors, they often the effects of the global crisis. Two sets of policy coexist with less efficient firms that seem unable to actions were at the core of Mauritius’ resilience. First, fully acquire the technology and market knowledge the resolute implementation of the mid-2000s reform of the leading companies. agenda fostered investor confidence and reinforced economic diversification, helping to sustain overall 48. Since 2010, reforms have faltered, and relatively economic activity as more traditional sectors faltered. accommodative monetary and fiscal policies have been Second, the fiscal space achieved during previous difficult to rein in. Current public expenditures have years allowed the Government to adopt a stimulus remained relatively high, going from 24.9 percent of package in 2008 to counter the impact of the global GDP in 2010 to 24.8 percent in 2013. The debt-to-GDP crisis, accelerating infrastructure investment projects ratio has actually increased to reach 57.9 percent and providing timely, targeted, and temporary social in 2013. Implementation of reforms has slowed assistance to cushion the crisis’ impact on workers substantially, further accelerating the decline in gross and the most vulnerable citizens. national saving to below 15 percent of GDP and leading to a stagnation of private investment at around 18 45. Overall, the current macro-fiscal framework fosters an percent of GDP . As a result, economic growth has been environment conducive to economic growth. Despite on a slowly declining trajectory—from 4.1 percent in recent slowdown in reforms, the successful policies 2010 to 3.2 percent in 2013. Contributing factors have implemented prior to 2010 built a resilient and been difficulties in the tourism sector, where growth thriving economy that has diversified over the years. slowed to an average of 3.8 percent over the period, Government decisions with regard to spending and and construction, which declined -3.4 percent. Banks saving also contributed to this increased resilience. As remain well capitalized, with adequate provisions, a result of significant fiscal consolidation, the public and loans have increased substantially, particularly sector in Mauritius has become a net saver, reducing to construction and real state, which today represent the demand for external financing. Mauritius’ economy 20.4 percent of GDP, up from 14.1 percent in 2010. is still performing well in a difficult global context. Its However, non-performing loans have slowly increased 3.2 percent growth in 2013 (Figure 4) was reasonably from 0.95 percent of GDP to 1.67 percent of GDP. solid, despite a high unemployment of 8.0 percent. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 11 CHAPTER 2 - Mauritius Economic Development 49. Mauritius has been characterized by relatively low 51. auritius has always relied largely on its human M productivity growth. Figure 6 shows that total factor resource to sustain economic growth, increasing the productivity contributed less than 20 percent to importance of access to education and its quality. annual value-added growth between 2000 and 2010. At 15 percent, public spending on education as a Productivity’s impact on growth was fairly steady percentage of government expenditures was above throughout the period. Most of the value-added the global average. As a share of GDP, however, growth was driven by the tertiary sector (Figure 9). public spending was below average at 3.5 percent The primary sectors, mainly agriculture, contributed (WDI databases). Free public schools provide all the least. children with full access to education, although the fact that many households finance private schooling 50. Challenges and inefficiencies exist in the labor market, translates into highly unequal educational outcomes. associated with rigid institutional arrangements. The proportion of pupils starting primary school and As will be described in more detail in the labor reaching its last grade is very high at 98.7 percent, chapter, the disproportionally high wages set by according to official data. The literacy rate is collective bargaining and labor regulations affect relatively high, increasing from 85 percent in 2001 to the competitiveness of certain sectors and lower 89.7 in 2011. Literacy rates are almost 100 percent employment creation. The question of when Mauritius for children and youths and growing for the elderly. will be able to achieve high-income country status will However, overall literacy remains below the median depend on its ability to improve the labor force’s skill for comparable countries. set and infrastructure quality. In addition, the speed of technology adoption and further improvements in business environment will be essential to attracting FDI and generating domestic investment. Figure 7: Factors in per capita value-added Figure 6: Decomposition of per capita value-added growth, 1990-2013 growth, 2000-12 100% 90% 3.0 Primary, Annual Change (percentage points) 80% 0.15 2.5 Secondary, Share of Real GDP Growth 70% 0.90 2.0 60% 50% 1.5 40% Tertiary, 30% 1.0 1.48 20% 0.5 Intersector 10% Reallocation Effect, 0.33 0.0 0% 2000-2004 2004-2008 2008-2012 Source : Mauritius SCD report (forthcoming). Data sources: various, including WDI, Statistics Mauritius (2015). The periods do not match, and the figures Total Factor Productivity Labor should be used for illustrative purposes. Human Capital per Labor Capital Stock Source: Authors’ calculations. 12 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 2 - Mauritius Economic Development Photo : © Petter Thorden 52. Educational quality remains a challenge. The education a faltering recovery in the euro zone would further system is failing to equip many young Mauritians with undermine economic growth by cutting both tourism adequate skills.. Not only do too many children fail earnings and FDI inflows. to acquire the minimum level of education, partly because of failing the Certificate of Primary Education, 54. The current outlook recognizes possible downsides that but the overall quality of learning does not compare could cause growth to deviate from these projections. well with other upper middle-income countries. The outlook assumes that measures will be taken to stimulate private investment, utilize public investment, expand market share in emerging C. Moving forward economies, and support growth in emerging sectors while consolidating traditional sectors. It remains 53. Mauritius’ GDP growth rate appears to be close to subject to downward revision should various risks its long-run potential. Supported by an improving materialize. Domestically, the main threat to the external environment, the economy is projected outlook is the slow pace of the structural reforms to grow by between 3.7 percent and 4.0 percent in needed to support growth, chiefly the need to increase 2014. The fishing, ICT, and financial services sectors the efficiency of the public sector. On the external are expected to drive near-term growth, more than front, sluggish growth in external demand and the offsetting slow or even negative growth rate in pressures that may build on the current account construction. However, these forecasts are subject remain matters of concern. Nevertheless, In light of to significant downside risks, and current projections the resilience exhibited by the Mauritian economy depend, inter alia, on successful implementation of in recent years, and given the means available for the Government’s public-investment program. While coping with external uncertainty, these risks should Mauritius continues to be resilient to external shocks, be manageable. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 13 CHAPTER 2 - Mauritius Economic Development Photo : © Matti Mattila 14 MAURITIUS | Inclusiveness of Growth and Shared Prosperity Chapter 3 How Growth and Economic Reforms Translates into Income Distribution of the Households Earlier studies found noticeable progress on human- development indicators in Mauritius, including life expectancy, maternal and infant mortality, enrollment measures, and access to primary education for both genders. This study confirms that the general well-being of the population has improved between 2007 and 2012. Household consumption and income have both increased in real terms, and absolute poverty has declined. However, the growth was not equally shared. When it comes to growth, people in the middle of the distribution and the rich have benefited more than the bottom 40 percent. Income inequality and relative poverty have increased. Poverty depth has also increased. Growth of the bottom 40 percent was much less than the average growth rate. Professionals and generally skilled workers benefited the most from growth, while the unemployed and inactive population lagged. In terms of the levels of poverty and inequality, Mauritius is comparable to many other middle-income countries, but it is behind in terms of the shared-prosperity indicators. CHAPTER 3 - How Growth and Economic Reforms Translates into Income According to the report, Mauritius is one of the few A. Introduction African countries achieving remarkable results over the past 15 years in meeting MDG indicators, with six 55. Earlier studies have found noticeable progress on out of eight specific goals more or less accomplished. human-development indicators in Mauritius, including The overwhelming majority of the population has life expectancy, maternal and infant mortality, access to safe drinking water. Primary education is enrollment measures, and access to primary education universal. The population’s general state of health is for both genders. A recently published Statistics good. Life expectancy increased from 62 years at the Mauritius Human Opportunity Index report7 found time of independence in 1968 to 72 years in 2010, and that the “overall social picture is quite impressive infectious diseases such as malaria, polio, diphtheria, and encouraging, as demonstrated by good progress typhoid, and cholera have been virtually eradicated. on the Millennium Development Goals (MDGs).” 7 Quotation is taken from the World Bank report “Do Children in Mauritius Have Equal Opportunities in Education?” 2012. Box 2: Overview of inclusive growth concept 8 The need for more “inclusive growth” has been recognized in many countries. During the past decade, many developing nations had stable and often high rates of economic growth, including China, Vietnam, India, Mozambique, and Bangladesh. But the extent to which this growth has been shared differs greatly, with many of the countries experiencing higher income inequality. In some places, the greater inequality takes on a regional dimension—as in Ghana and Nigeria, where the northern parts have traditionally been left behind.9 However, many Latin American countries have successfully reduced income inequalities. Cash transfers played an important role, along with increases in other public investments and macroeconomic stability. Inclusiveness of growth and shared prosperity are essential ingredients in any successful growth strategy. Inclusiveness is a multidimensional concept that encompasses equity, equality of opportunity, and protection in market and employment transitions. The inclusive growth approach takes a longer-term perspective because it focuses on productive employment, rather than on direct income redistribution, as a means of increasing incomes for excluded groups. Inclusiveness covers a broad range of issues—for example, changes in the market structure, access to finance, discrimination in labor and product markets, and conditions in the informal sector. The micro dimension captures the importance of structural transformation for economic diversification and competition. The World Development Report 2013 emphasizes the role of labor- market institutions. It highlights the importance of looking beyond macroeconomic growth and taking into account distributional concerns and the extent to which people feel they take part in the decisions that shape their lives. Extensive research addresses various aspects of inclusive growth and shared prosperity. As stated above, inclusiveness is a multidimensional process related to economic factors, human capital, and political and social dimensions. It combines improvement in average level in various indicators as well as distributional component. The following is a table map of different measures generally associated with inclusiveness of growth. Inclusive growth method is a multidimensional approach taking in the consideration inequality in economic, human capital accumulation, social dimensions on the relatively long period of time and in a sustainable manner. For growth to be inclusive productivity must be improved, new opportunities for employment created, and the gains should be shared across population groups. 8 See Box 2 describing definition of the inclusive growth (IR). 9 Regional disparity might not apply in Mauritius, while ethnic disparities should be explored if data availability. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 17 CHAPTER 3 - How Growth and Economic Reforms Translates into Income 56. The IMF has analyzed the distributional impact of the and income as the welfare aggregates. Household Mauritian economic reforms, using household survey consumption expenditure is the value of goods data10 to e 00s; however, the evidence points to a and services acquired during the reference period, skewed distribution of the benefits of growth, possibly regardless of whether they were paid for or received associated with fundamental structural changes in the for free. Household income is the total receipts of Mauritian economy. Inequality has increased, largely members who received employment income, property explained by variations in employment income. The income, transfer income, income from own produced largely untargeted SP system played an important role goods, and imputed rent for non-renting households. in successfully combating poverty, but reforms are needed to ensure that resources are spent in the most 59. Between 2007 and 2012, both household consumption cost-effective way. On the revenue side, Mauritian and income increased in real terms, but poorer people income taxes are relatively progressive, although benefited less. In both cases, better-off individuals they have a negligible impact on the overall income benefited more than their less advantaged counterparts distribution. The analysis also indicated that the VAT (Figure 8 and Figure 9). For low consumption and is relatively progressive, even if its impact on overall income levels, the probability density functions income distribution was small. corresponding to the periods 2007 and 2012 trace each other. For higher consumption and income levels, 57. We look at inclusiveness of growth in Mauritius based however, the 2012 probability density functions are on recently available HBS data for 2007 and 2012, to the right of the 2007 ones. The finding that the labor force surveys (LFS) for 2001-12, and firm-level rich benefited more than the poor is confirmed by data from registry of companies. As described in the comparing mean and median changes. Over the period Chapter 1, economic reforms resulted in a substantial studied, mean per capita consumption expenditure reallocation of resources, which clearly had positive increased by 16 percent while the median increased effects on economic growth. This report addresses by 7 percent. For per capita income, the figures are several questions: How has economic growth affected 17 percent and 6 percent. This suggests that richer workers, firms, and households? What is the role of individuals pushed up incomes and expenditures. The the SP programs and improved employment? Has poor showed little change. vulnerability increased, and is the middle class better- off or worse than eight years ago? With respect to firms, how much of the structural change took place C. Poverty trends in Mauritius in recent years? 60. To understand the effect of growth on poverty, this section focuses on two concepts of poverty: absolute B. Consumption expenditure and income and relative. Absolute poverty compares per adult equivalent household income to a fixed poverty line 58. Mauritius’ HBS data are of good quality and generally over time. In other words, the poverty line is the same comparable over time. This section benefits from the in 2012 as it was in 2007. In relative poverty analysis, use of more recent data to analyze the effect of the poverty line is allowed to vary with income—i.e., growth on household well-being. It uses the latest per adult equivalent income is compared to a relative HBS, implemented in 2012, making comparisons to poverty line. The relative poverty line is defined as the preceding one, the HBS 2006-7. The surveys half median monthly household income per adult cover a similar set of variables, follow the same equivalent. Figure 10 presents absolute and relative sampling procedures, and are generally comparable poverty estimates over time. for analysis.11 To understand the effect of growth on well-being, we used both household expenditures 10 Antonio C. David and Martin Petri, 2013 “Inclusive Growth and the Incidence of Fiscal Policy in Mauritius— Much Progress, But More Could be Done.” 11 Although the surveys are generally comparable over time, some variables’ definitions have changed, leading to difficulties in comparing some results. For example, the definition of em- ployment sectors is among the indicators that have changed. 18 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 3 - How Growth and Economic Reforms Translates into Income Figure 8: Consumption distribution in 2007 and 2012 0.00014 Probability density function 0.00012 0.0001 0.00008 0.00006 0.00004 0.00002 0 0 10 20 30 40 50 60 2006-7 2012 Per capacita household expenditure, thousands Source: Authors’ calculations using HBS 2007 and 2012. Figure 9: Income distribution in 2007 and 2012 Probability density function 0.00006 0.00005 0.00004 0.00003 0.00002 0.00001 0 0 20 40 60 78 100 120 2006-7 2012 Per capacita household expenditure, thousands Source: Authors’ calculations using HBS 2007 and 2012. Figure 10: Poverty in Mauritius over time 2006-7 2012 12.0 61. Absolute poverty declined in Mauritius from 8.5 10.0 9.8 percent in 2007 to 6.9 percent in 2012. Keeping 8.5 8.5 poverty line in constant prices over time suggests 8.0 improvement in the well-being of the population. 6.9 62. Relative poverty has increased over time—from % 6.0 8.5 percent to 9.8 percent. This reflects the effect of increased inequality, despite the observed 4.0 economic growth. Pushed by the income gains of the richer households, median income per adult 2.0 equivalent has grown faster than the incomes of those in the lowest quintiles. Had inequality not 0.0 increased, relative poverty would have remained at Absolute Relative least the same, and the decline in absolute poverty Income Poverty would have been greater. Source: Authors’ calculations using HBS 2007 and 2012. Poverty measurement is based on officially adopted methodology. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 19 CHAPTER 3 - How Growth and Economic Reforms Translates into Income 63. Mauritius’s poverty is low, but its depth has increased. inequality index and the Lorenz curve corroborate The headcount poverty rate does not distinguish this finding. Between 2007 and 2012, the Gini index between those with consumption just below the has increased from 0.34 to 0.37.12 In the same period, poverty line and those deeper in poverty. Policies the gap between the Lorenz curve and the 45o degree designed to improve the well-being of those at the line has widened (Figure 11). This means that, over bottom of the consumption distribution will result time, the rich grabbed a larger share of income than in poverty reduction only if benefits are sufficient the poor. A similar increase in the income inequality is to cross the poverty line. The poverty gap measures observed when we measure the changes of inequality “the depth of poverty, or how far the poor are from by other indices. the poverty line.” The poverty gap, measured against the relative poverty line, was 1.9 percent in 2012, 65. The growth incidence curve (GIC) offers further evidence an increase of 0.3 percentage point from 2007. The of increased inequality in Mauritius. The GIC looks at increase in the poverty gap is associated with the how growth in income or consumption expenditures general deterioration of the poor population in the is distributed among various quintiles and shows the income distribution. Even with this increase, the interaction between growth, poverty, and inequality. poverty gap and poverty headcount are significantly As national income or expenditures rise, the curve lower in Mauritius than in other African countries. helps to address the policy question of whether the income or expenditures of the poor are increasing more or less quickly than the country overall. This D. Inequality trends in Mauritius is particular interest in Mauritius because of the increased inequalities observed over time. 64. Income inequality has increased in Mauritius. Initially better off individuals have benefited more from economic growth than their less-advantaged 12 Gini coefficients presented in this report are estimated on counterparts. Figure 9 showed that the spread of an income or consumption per capita basis, while official ine- the probability density function has become larger quality figures are estimated on total household income or between 2007 and 2012, indicating increased income consumption. For this reason, the official inequality estimates inequality over time. The observed changes in the Gini are higher. Figure 11: Lorenz curve over time 1 Cumulative proprtion of income 0.8 0.6 0.4 0.2 0 0 0.2 0.4 0.6 0.8 1 Cumulative population proportion 2006-7 2012 Source: Authors’ calculations using HBS 2007 and 2012. 20 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 3 - How Growth and Economic Reforms Translates into Income Figure 12: Growth incidence curve of household income 8 Household income Annual growth rate, % 6 4 2 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 90 93 95 99 -2 Income percentiles -4 Source: Authors’ calculations using HBS 2006-7 and 2012. Figure 13: Growth incidence curve of household consumption expenditures 4 Household Consuption Expenditure Annual growth rate, % 3 3 2 2 1 1 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 90 93 95 99 -1 -1 -2 Source: Authors’ calculations using HBS 2007 and 2012. Expenditure percentiles Figure 14: Median monthly earnings (Rs) 25,000.0 250000.0 15,000.0 Rs 10,000.0 5,000.0 0.0 Lowest 2 3 4 Highest quintile quintile 2006-7 2012 Source: Authors’ calculations using HBS 2007 and 2012. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 21 CHAPTER 3 - How Growth and Economic Reforms Translates into Income Photo : © Stephane Calvet 66. Those in the middle of the distribution and the rich for this group had declined by 11.6 percent. Those in benefited more from growth. Figure 13 and Figure the second income quintile also experienced falling 14 present GICs for household income and household earnings. But earnings have increased over time for consumption expenditures. The first curve shows the richer population groups. For the highest income that most of the population has experienced positive quintile, median monthly earnings rose 15.3 percent income growth. Incomes rise less quickly among the between 2007 and 2012. poor than in the country as a whole. The poor have experienced a decline in consumption expenditures. 68. The Gini index indicates that earnings differences were the main source of inequality in Mauritius. As presented 67. Inequality was high among the employed, and it has in Table 1, employment income has been the main tended to increase. Figure 12 presents median monthly source of inequality. Furthermore, employment’s earnings across income quintiles over time. Those in relative contribution to inequality has increased 69.4 the lowest quintile (the poorest) received a median percent in 2007 to 73.2 percent in 2012. 5,400 Mauritian rupees (Rs) in 2007 from employment or self-employment. By 2012, median monthly earnings Table 1: Gini inequality decomposition by income sources (Shapley value approach)13 Y ea r 2007 2 0 12 I n c om e Absolute R e l at i v e Absolute R e l at i v e s ou r c e s contribution contribution (%) contribution contribution (%) Employment income 25.73 69.41 28.54 73.17 Self-employment 5.84 15.77 5.89 15.11 Property income 1.70 4.59 0.88 2.26 Transfers 3.77 10.17 3.67 9.40 Own production 0.03 0.07 0.03 0.07 Total economy 37.07 100.00 39.01 100.00 Source: Authors’ calculations using HBS 2007 and 2012. 13 The table uses total household income as the living standards indicator. Thus, total absolute income inequality will differ from the initial Gini index figures, which use per adult equivalent household income. 22 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 3 - How Growth and Economic Reforms Translates into Income bottom 40 percent of the population in every country. E. Shared prosperity – why stayed behind The promotion of shared prosperity requires a growing economy and equal redistribution of the gains. It 69. Reducing extreme poverty and fostering shared requires both growth and reduction of inequality. This prosperity are twin core goals Mauritius should pursue section discusses how Mauritius fared with respect to to achieve the inclusive growth advocated by the World this goal between 2007 and 2012. As discussed earlier, Bank. The percentage of people living on less than the nation’s economy has been growing continuously US$1.25 a day is literally zero, putting Mauritius since the 1990s. But the GICs presented in Figure 13 among the world’s relatively developed countries. and Figure 14 suggest that the well-being of bottom The second goal focuses on shared prosperity. It 40 percent grew slower than the overall population pledges to foster real income growth among the in Mauritius. Box 3: World Bank’s twin goals of ending extreme poverty and promoting “shared prosperity” The World Bank Group has established ambitious goals to reduce international poverty and boost shared prosperity. These two goals and their respective indicators can be summarized as: 1. End extreme poverty: the percentage of people living with less than $1.25 a day to fall to no more than 3 percent globally by 2030. 2. Promote shared prosperity: foster income growth of the bottom 40 percent of the population in every country. Ending extreme poverty and promoting shared prosperity are also unequivocally about progress in non-monetary dimensions of welfare, including education, health, nutrition, and access to essential infrastructure as well as enhancing voice and participation of all segments of society in economic, social, and political spheres. Ending extreme poverty within a generation and promoting shared prosperity must be achieved in ways that are sustainable over time and across generations. This requires promoting environmental, social, and fiscal sustainability. The shared prosperity indicator implies a direct focus on the incomes of the less well-off—a departure from the common practice of focusing only on growth in GDP per capita and implicitly relying on the “trickle down” impact to benefit the bottom of the distribution. To analyze shared prosperity in Mauritius, this section compares the mean annual growth rates of the poorest 40 percent of the population and the total population over five years—i.e., for 2007 to 2012. The annual growth rate is measured using the following formula: , where idot is the annual growth rate of per adult equivalent income, F is the final value for income, S is the initial value of income, and y is the number of years over which the annual growth rate has been calculated. 70. Growth was not shared equally in Mauritius between and services had higher shared prosperity than 2007 and 2012. At the national level, the real income those in agriculture and industry. Professionals had of the bottom 40 percent of the population grew at an the highest shared prosperity. As for the gender of annual average of 1.8 percent (Table 1). In contrast, household heads, real income for the bottom 40 annual average growth for the Mauritian population percent lagged the whole population for both men and as a whole was 3.1 percent—more than 1 percentage women. But the bottom 40 percent living in female- point faster. The gap between the better off and the headed households experienced smaller increases in worse off has increased over time, indicating that the income gap. It grew by 1.6 percent across female- prosperity has not been shared in Mauritius. headed households, compared with 1.8 percent for the nation at large. The comparable figures for male- 71. P rofessionals and generally skilled workers have headed households were 2 percent for the bottom 40 benefited the most from growth (Table 2). Shared percent and 3.4 percent for the country. prosperity was higher among the employed than among the unemployed and inactive. Those in trade MAURITIUS | Inclusiveness of Growth and Shared Prosperity 23 CHAPTER 3 - How Growth and Economic Reforms Translates into Income Table 2: Shared prosperity within Mauritius, selected groups Annual growth rate Bottom 40 percent All (2007 to 2012) (%) (%) N at i o n a l 1.8 3.1 Economic status E m p l oy ed 1.9 3.0 U n e m p l oy ed 1.3 2.1 O u t of l a b o r f o r c e 1.8 2.8 Sector of activity Ag r i c u lt u r e 0.4 0.3 I n d u s t ry 1.7 2.2 Trade 2.9 9.7 S er v i c e 2.0 3.6 Occupation M a n ag e r s 1.9 2.5 P r of e s s i on a l s 3.0 1.4 T ec h n i c i a n s 1.7 1.6 C l er i c a l wo r k e r 1.8 1.4 S er v i c es / s a l e s wo r k e r s 2.0 1.7 Ag r i c u lt u r e ( s k i l l e d ) 1.1 -0.02 T r a d e s wor k e r s 1.6 1.5 O p er ator s a n d a s s e m b l e r s 2.3 1.9 E l em en ta ry o c c u pat i o n s 1.7 1.7 Gender of head Male 2.0 3.4 F em a l e 1.6 1.8 Source: Authors’ calculations using HBS 2007 and 2012. Bold type indicates groups with growth of bottom 40 percent above the average. chart). Mauritius’ “neighbors,” such as Madagascar, F. International comparison of poverty, report over 80 percent of their populations below the inequality, and shared prosperity international poverty line. In having virtually zero $1.25 headcount poverty, Mauritius compares with Eastern European countries, such as Ukraine, Bosnia 72. To make international poverty comparisons, this and Herzegovina, and Belarus. section uses the US$1.25 a day per capita poverty line, evaluated at 2005 purchasing power parity. Using 74. In terms of Gini inequality, Mauritius also compares this absolute poverty line rather than the national well to its peer middle-income countries. It does thresholds permits making meaningful comparisons of much better than its African “neighbors.” Of the well-being. For comparisons of inequality, the section 74 countries in Figure 15 (right chart), only 17 are uses Gini coefficients. more equal than Mauritius. The 56 other countries show greater inequality than Mauritius, particularly 73. Mauritius belongs to the group of low poverty and Seychelles, South Africa, Comoros, Botswana, and low inequality countries. It has virtually zero $1.25 Namibia. dollar-a-day poverty, a rarity in SSA (Figure 15, left 24 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 3 - How Growth and Economic Reforms Translates into Income Figure 15: Poverty and inequality across countries 70 B. Low Poverty C. High Poverty High Inequality High Inequality Seychelles 65 Comoros South Africa Botswana 60 Namibia Zambia 55 Lesotho Swaziland 50 Mozambique Inequality Gini 45 Malawi Madagascar 40 35 Mauritius 30 25 D. High Poverty 20 Low Inequality 0 10 20 30 40 50 60 70 80 90 Poverty $1.25 90 80 70 Poverty gap Poverty rate Poverty rate, gap 60 50 40 Mauritius 30 20 10 0 Congo, Dem. Rep. Liberia Nigeria Malawi Tanzania Micronesia Angola Guinea-Bissau Benin Niger Gambia, The Togo Namibia India* Côte d'Ivoire Lao PDR Sudan Belize Colombia Venezuela, RB Djibouti Paraguay China* Ecuador Pakistan Costa Rica Peru Tajikistan Bhutan Dominican Republic Algeria China--Urban Hungary Albania Serbia Gabon Uruguay Slovak Republic Slovenia Poland Maldives West Bank and Gaza Iraq Source: POVCALNET database and authors’ calculations. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 25 CHAPTER 3 - How Growth and Economic Reforms Translates into Income 75. Despite positive real income growth among the bottom the best performing country, was roughly seven times 40 percent, Mauritius did not compare well at the greater than the comparable figure for Mauritius. On international level. Looking at the 81 countries Figure average, the real income growth of the bottom 40 16, Mauritius’ bottom 40 percent grew faster than the percentile in Mauritius was 1.8 percent, compared bottom 40 percent in just 19 countries; 61 countries with 4.1 percent across the comparable countries performed better than Mauritius. The bottom 40 with available data. percent’s real income growth in the Slovak Republic, Figure 16: Shared prosperity in Mauritius, international comparison Consumption annual growth of the bottom 40% of the population, 2006-12 16.00 14.00 12.00 10.00 Mauritius (1.8) 8.00 6.00 4.00 2.00 0.00 -2.00 -4.00 Fiji Iraq Mali Peru Chile India Togo country Slovak Republic Poland Russian Federation Belarus Panama Uruguay Malaysia Bhutan China Argentina Nepal Romania Cambodia Uganda Azerbaijan Brazil Tajikistan Latvia Kazakhstan Paraguay Kyrgyz Republic Moldova Vietnam Bolivia Costa Rica Estonia Lithuania Ukraine Turkey Rwanda Mozambique Botswana Colombia Bulgaria Nicaragua Hungary Ecuador Honduras Kosovo Czech Republic Tunisia Thailand South Africa El Salvador Pakistan Nigeria Sri Lanka Jordan Burkina Faso Slovenia Albania Montenegro Dominican Republic West Bank and Gaza Lao PDR Mauritius Bangladesh Malawi Philippines Mexico Senegal Ethiopia Egypt, Arab Rep. Armenia Mauritania Indonesia Côte d'Ivoire Georgia Macedonia, FYR Serbia Croatia Guatemala Central African Republic Zambia Source: World Bank databases. 26 MAURITIUS | Inclusiveness of Growth and Shared Prosperity Chapter 4 Poverty, Vulnerability and the Middle Class Close to 80 percent of Mauritius’ population could be classified as middle class. A worrisome trend has been the increase in the relative size of the vulnerable population. Demographically, being in a smaller family or one headed by a male provides better odds of being middle class. Having more education—the key to employment in higher-paying sectors—is a path to middle- class status; so is employment in some of the growing sectors of Mauritius’ economy. The relative numbers of poor remained the same between the two HBS surveys and appear to be well covered by various social protection schemes. The highest growth in the vulnerable was among female-headed households, those who receive large amounts of transfer income, and those employed in household services, other low-skilled occupations, and the unemployed. CHAPTER 4- Poverty, Vulnerability and the Middle Class 77. To enhance the policy relevance of this analysis, A. Introduction however, more emphasis should be put on the uneven income distribution, especially now that 76. In recent years, the concept of a middle class has Mauritius is close to entering the club of developed been broadly discussed in socio-economic literature economies. The issue of distribution will become even and policy debates. Empirical evidence shows that more challenging because it will require sustained countries with faster growth in the middle class are economic growth and shared prosperity, associated associated with better governance, reforms, and with reduction of economic vulnerability and the even better infrastructure. As people gain middle- rise of the middle class achieved through substantial class status, they tend to accumulate savings and productivity gains. While the percent of people who acquire secondary and tertiary education—i.e., make are middle class has been increasing in Mauritius, a investments in the future. Members of the middle class worrisome sign is that there has also been an increase are likely to support accountable government, the in the number considered vulnerable. rule of law, property rights, and better infrastructure, education, and economic stability. Faster growth and 78. This chapter defines and analyzes the middle class poverty reduction is associated with the appearance in Mauritius based on two recent household budget and growth of the middle class. Mauritius has had surveys. It is structured as follows. The first section considerable economic growth in recent decades, reviews the literature on the middle class, including accompanied by significant improvement in many various definitions of the middle class, characteristics social indicators and a growing middle class by any of the middle class from other studies, and a discussion definition. Policy reforms in the mid-2000s helped of the importance of the middle class for economic alleviate structural challenges and led to continued growth. This is followed by the main section that broad-based growth even during the global financial defines and profiles the middle class in Mauritius, using crisis of the late 2000s. a vulnerability-to-poverty approach. The chapter concludes by providing policy recommendations regarding the middle class in Mauritius. Box 4: Literature on the middle class A long and growing literature focuses on the middle class, its characteristics, and its importance to economic growth and stability. The emergence of many of today’s high-income countries has often been attributed to the development of a middle class. This group is dominated by people with a vested interest in a stable society, who accumulate savings, invest in education for themselves and their children, and in other ways make investments in human and social capital. They advocate good governance, rule of law, and economic stability. Considerable evidence points to links between faster economic growth and an expanding middle class. While these intellectual and policy concepts have a long history, no consensus has been reached on how to define and measure the middle class. Easterly (2000),14 for example, takes a relativist approach, defining the middle class as those between the 20th and 80th percentile of the consumption distribution. Bhalla (2009)15 takes an absolute approach, defining the middle class as those with annual incomes over US$3,900 in purchasing power parity terms. Banerjee and Duflo (2007)16 use two alternative absolute measures—those with daily per capita expenditures of US$2 to US$4 and those with daily per capita expenditures between US$6 and US$10. Ravallion (2009)17 takes a hybrid approach, defining a “developing world middle class,” as with a range of incomes between the median poverty line of developing countries and the US, and a “Western world middle class.” Absolute definition the middle class: Many have tried to define and characterize a global middle class and to distinguish it from the global poor. The upper and lower bounds are often defined in an ad hoc manner.18 Banerjee and Duflo based their two ranges—a lower middle class at US$2 and US$4 a day per capita (adjusted at purchasing power parity) and an upper middle Easterly, W. (2000), “The Middle Class Consensus and Economic Development,” Policy Research Working Paper 2346, World Bank, Washington, 14 DC. 15 Bhalla, S. (2009), “The Middle Class Kingdoms of India and China,” Peterson Institute for International Economics, Washington, DC. Abhijit V. Banerjee and Esther Duflo, “What is middle class about middle classes around the world?”, Massachusetts Institute of Technology, 16 Department of Economics, December 2007. Martin Ravallion, The Developing World’s Bulging (but Vulnerable) “Middle Class,” Policy Research Working Paper No. 4816, Development 17 Research Group, The World Bank, January 2009. 18 Ibid. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 29 CHAPTER 4- Poverty, Vulnerability and the Middle Class class between US$6 and US$10—on a sample of household surveys in 13 low- and middle-income countries.19 The lower bound of US$2 a day reflects developing-country standards—few people who have incomes above the threshold are considered poor in these economies. However, the upper bound of US$10 a day would still be regarded as poor in developed countries, where the threshold for poverty is about US$13 a day. With economic growth over the past decade and a half, 1.2 billion people in the developing world have joined the ranks of the middle class—by its standard. However, many of them would still be considered poor in the developed world—thus the need for separate middle-class thresholds for the developed and developing worlds. Relative definitions: In developed countries, a different approach is often adopted—defining poverty on a relative basis, such as people who have incomes less than 60 percent of the mean. In less developed countries, others have adopted a relativist approach—for example, Easterly defined middle class as consumption between the 20th and 80th percentiles.20 Others define the middle class as those with per capita incomes between 75 percent and 125 percent of median per capita income.21 Subjective definitions of the middle class: A long tradition in the sociological literature involves subjective definitions of the middle class, where people are asked how they rank in the income distribution.22 Much of this research has been done in developed countries, where much more data is available.23 Characteristics of the middle class: Once middle class is defined, it is important to analyze the group’s characteristics and determine how they differ from the poor in terms of occupation, consumption patterns, family size and household composition, place of residence, education, health, and other variables. Then the direction of causality needs to be determined. Do households with certain characteristics become middle class or do households with “middle class” characteristics adopt different behaviors to stay middle class? In Mauritius, the middle class exhibits some quite distinct social and economic characteristics that are important for policy. Most studies define the middle class based on income or consumption and then compile characteristics of those falling into that class. This study will do the same. Why the middle class is important: Historical studies have pointed to the importance for overall economic growth of having a middle class that earns a larger share of a country’s income. For example, an expanding middle class was a driving force behind the growth of many of today’s high income countries in Western Europe. A number of studies have shown that economic growth is higher in countries with larger middle classes. Three reasons are often given for this. First, new entrepreneurs who have delayed consumption create employment and productivity for the rest of society. Second, the middle classes are a source of vital inputs for the entrepreneurial class. Third, the middle classes are willing to pay more for better quality goods, increasing investment levels and raising income levels for the entire society. Another study by Easterly defines a middle-class consensus as a national situation with neither strong classes nor ethnic differences. He shows that lower ethnic polarization and higher income shares held by the middle class are associated with a range of desirable development outcomes—higher incomes, faster growth, better health and education, more political stability, less civil war, and more democratic societies.24 The opposite of middle-class economies are unequal ones, where wealth flows to a small number of people, and several studies show that high inequality is associated with poor growth outcomes. 19 Banerjee and Duflo (2007). 20 William Easterly, The Middle Class Consensus and Economic Development, Journal of Economic Growth, Vol. 6, No. 4, 2001, pp. 317-335. 21 Nancy Birdsall, Carol Graham, and Stefano Pettinato, “Stuck in the Tunnel: Is Globalization Muddling the Middle Class?” Brookings Institu- tion, Center on Social and Economics Dynamics Working Paper. No.14, 2000. 22 C. Wright Mills, The Sociological Imagination, 1959, 2000, Oxford University Press. Reynolds Farley, editor, State of the Union: America in the 1990s. Volume One: Economic Trends, Russell Sage Foundation, New York, 1995. 23 Reynolds Farley, editor, State of the Union: America in the 1990s. Volume Two: Social Trends, Russell Sage Foundation, New York, 1995. 24 Easterly (2000). 30 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 4- Poverty, Vulnerability and the Middle Class Photo : © Werner Bayer MAURITIUS | Inclusiveness of Growth and Shared Prosperity 31 CHAPTER 4- Poverty, Vulnerability and the Middle Class 81. In a worrisome trend, those classified as vulnerable rose B. Scope of the vulnerability and middle class in from 10.2 percent in 2007 to 12.7 in 2012. Overall, it seems the income distribution stretched a bit over Mauritius this period, with people from the lower middle income group moving in equal numbers into the upper 79. Absolute definition of the middle class using a middle income and vulnerable groups. That the upper vulnerability approach: In Mauritius, the definition middle class now encompasses more than half the of vulnerability and middle class is based on two population is a promising trend, but the increased household budget surveys, one conducted on 2007 share of those classified as vulnerable is worrisome and the other in 2012. The framework for defining and the next section looks at the characteristics of middle class according to households’ vulnerability this group. to poverty follows a regression-based approach to estimate an income threshold associated with a low probability of falling into poverty. The methodology C. Who are the poor in Mauritius introduced by Lopez-Calva and Ortiz-Juarez envisions a three-stage process for defining the middle class. The first stage identifies the actual characteristics of iv. Demographics of poverty those moving in and out of poverty. The second stage constructs probabilities of falling into poverty. The 82. Poverty incidence and the share of the poor tended to third stage identifies the income level associated with increase with household size. As shown in Figure 18, that probability. people in single-member households were poorer than those living in households with two members. 80. The vulnerability of the middle class and other income But people living in bigger households experienced a groups is calculated for Mauritius for 2007 and 2012. greater incidence of poverty, particularly those living Between 2007 and 2012, the overall size of the middle in households with seven or more members. Bigger class declined slightly from 79.9 percent to 77.2 households made up a larger share of the poor—but percent of the population (Figure 17). The relative only up to households with four members. People in sizes of the rich and poor groups remained roughly households with five or more members accounted for the same, with a slight reduction in poverty. The a smaller share of poverty. upper middle class grew from 49.4 percent to 52.3 percent of the population, while the lower middle 83. Households with larger age-dependency ratios. In both class declined significantly from 30.5 to 24.9 percent. 2007 and in 2012, the poor, when compared to the Figure 17: Middle class in Mauritius, Figure 18: Poverty incidence and the share of the 2007 and 2012 poor by household size 100% 90% 30.0 80% 26.7 Share of total population 70% 49.4 52.3 25.0 22.2 Poverty headcount (%) 60% 20.0 50% 15.0 13.6 40% 13.7 12.0 30% 30.5 24.9 10.0 8.4 20% 10.2 12.7 5.0 3.5 10% 8.2 7.9 0% 0.0 2007 2012 1 2 3 4 5 6 7 or more rich middle class high low middle class poor vulnerable Share of the poor (2012) 2006-7 2012 Source: Authors’ calculations. Source: Authors’ calculations, using HBS 2007 and 2012. 32 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 4- Poverty, Vulnerability and the Middle Class non-poor, lived in bigger households and with a larger the gap relative to male-headed households (Figure number of children, households with a smaller number 19). As will be seen later, labor participation rates of working-age adults, and households with younger among females was significantly low. This fact, in members. In other words, the poor tended to live conjunction with the mentioned widowed marital in households with larger age-dependency ratios. status of female heads, has limited their access to In 2012, the share of female heads was 31 percent labor income, which is the main source of household among the poor and 15 percent among the non-poor. income in Mauritius. In Mauritius, most male heads of households were married, but female heads were 84. The incidence of poverty was higher among people living mostly widowed. Figure 20 compares male and female in female-headed households and, for them, poverty household heads by marital status over time. In 2007, has increased sharply over time. Among those living 92 percent of male heads were married. By 2012, the in male-headed households, poverty remained steady figure was still large at 91 percent. Only 10 percent of at 8 percent between 2007 and 2012. For female- the female heads of households were married in 2007 headed households, however, poverty increased from and 2012. In both periods, over 60 percent of female 13 percent in 2007 to 18 percent in 2012, increasing heads were widowed. Figure 19: Poverty by gender of head 20.0 18.2 18.0 16.0 13.2 Only 10 percent of 14.0 12.0 the female heads % 10.0 of households were 7.8 8.1 8.0 6.0 4.0 2.0 married in 2007 0.0 Male Gender of household head Female and 2012. 2006-7 2012 Source: Authors’ calculations using HBS 2007 and 2012. Figure 20: Gender of head and marital status 100.0 90.0 2.2 2.9 80.0 70.0 Single 60.0 50.0 62.5 Separated 91.6 62.5 40.0 64.3 90.7 Divorced Widowed 30.0 Married 20.0 10.0 9.8 9.8 10.1 0.0 Male Female Male Female 2006-7 2012 Source: Authors’ calculations using HBS 2007 and 2012. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 33 CHAPTER 4- Poverty, Vulnerability and the Middle Class 85. Mauritius’ age structure is changing, with predominance 2012, the share of this age cohort had declined by 1 of young people gradually diminishing. Figure 21 and percentage point and became roughly at par with the Figure 22 show that the base of the age pyramids share of older five-year age cohorts—all the way up to became narrower over time. In 2007, 0-5 year olds the 55-60 age cohort. constituted about 4.5 percent of the population. By Figure 21: Age pyramid and poverty, 2007 2006-7 90-95 80-85 70-75 60-65 Age in years 50-55 40-45 30-35 20-25 10-15 0-5 5 4 3 2 1 0 1 2 3 4 5 Share in total population, % Poor females Poor males Females Males Source: Authors’ calculations using HBS 2007 and 2012 Figure 22: Age pyramid and poverty, 2012 2012 90-95 80-85 70-75 60-65 Age in years 50-55 40-45 30-35 20-25 10-15 0-5 0-5 5 4 3 2 1 0 1 2 3 4 5 Share in total population, % Poor females Poor males Females Males Source: Authors’ calculations using HBS 2007 and 2012 34 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 4- Poverty, Vulnerability and the Middle Class Figure 23: Poverty by age groups 18.0 16.0 14.0 12.0 10.0 % 8.0 6.0 4.0 2.0 0.0 0-5 6-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65+ 2006-7 2012 Source: Authors’ calculations using HBS 2007 and 2012 Figure 24: Poverty by age of head 45.0 40.0 35.0 30.0 25.0 % 20.0 15.0 10.0 5.0 0.0 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65+ Age of household head 2006-7 2012 Source: Authors’ calculations using HBS 2007 and 2012 86. Poverty has a predominantly young face. Figure 21 87. Sino-Mauritians were the least poor ethnic group and and Figure 22 also show that the share of poor males experienced a large drop in the incidence of poverty. and females declines with age. This can also be seen Mauritius has four main ethnic groups: Hindus, in Figure 23. Children, those aged under 15 years, Muslims, Sino-Mauritians, and the general population experienced the highest incidence of poverty in both (Figure 25). In 2012, the Hindus constituted the 2007 and 2012, and their increase in poverty incidence biggest share of the population (50 percent), followed over time was larger. Poverty incidence was also more by the general population (32 percent), the Muslims evident among individuals living in households headed (17 percent, and the Sino-Mauritians (1 percent). by younger people, but it tended to subside as heads Figure 26 shows that poverty among Sino-Mauritians grew older. In 2012, for instance, the incidence of fell from 3.7 percent in 2007 to 1.6 percent in 2012. poverty among people living in households with heads The Hindus and the general population became aged 15 to 19 years was above 40 percent. But for slightly poorer over time and had the largest share those living in households headed by people aged 30 of poor households. Roughly 49 percent of the poor and above, poverty was below 15 percent. in 2012 were part of the general population and 36 percent were Hindus. Muslims’ share of poverty was only 15 percent; the figure for Sino-Mauritians was roughly zero. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 35 CHAPTER 4- Poverty, Vulnerability and the Middle Class 88. The relatively low poverty of the Sino-Mauritians might v. Education and poverty be partly explained by their educational attainment and labor-market prospects. Sino-Mauritians had the 89. A clear relationship exists between household heads’ highest share of people with secondary and tertiary education and their income. Those living in households educations; few had no education. They also had the headed by more-educated people had greater incomes highest proportion of people employed and the lowest than their less-educated counterpart. Figure 27 shows proportion unemployed or out of the labor force. the income quintile composition of four educational Finally, Sino-Mauritians had the highest proportions levels in 2012. Among people living in households of people in managerial and professional occupations, headed by someone woho did not complete any and a negligible proportion in elementary occupations. education level, 33 percent were in the poorest As expected, median monthly earnings among this income quintile. Only 5 percent were part of the group were the highest, with the largest percentage top quintile. Outcomes were more dramatic at the increase over time. These characteristics have put tertiary-education level: 83 percent of people living Sino-Mauritians in an advantageous position to better with heads who completed tertiary education level benefit from economic growth. were part of the richest quintile. Figure 25: Distribution of ethnic groups 60.0 51.4 50.0 49.9 40.0 31.9 30.0 30.7 2006-7 2012 20.0 17.1 16.0 10.0 1.9 1.1 0.0 Hindu Muslim General population Sino Mauritian Source: Authors’ calculations using HBS 2007 and 2012. Figure 26: Ethnicity and poverty 60 50 48.6 40 35.8 30 20 15.4 10 0 0.2 Hindu Muslim General population Sino Mauritian Shar of poor (2012) 2006-7 2012 Source: Authors’ calculations using HBS 2007 and 2012. 36 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 4- Poverty, Vulnerability and the Middle Class Photo : © Cisco MAURITIUS | Inclusiveness of Growth and Shared Prosperity 37 CHAPTER 4- Poverty, Vulnerability and the Middle Class 90. As a result, poverty was highest among those living in vi. Labor and poverty households headed by someone without any education. Where household heads are uneducated, the poverty 91. Poverty is highest among the unemployed, but the headcount rate increased from 14.7 percent in 2007 largest share of the poor is among the inactive group to 15.5 percent in 2012; however, the group’s share in Mauritius. In 2012, the poverty rate was highest of poverty was under 7 percent and tended to decline among the unemployed, reaching 15.6 percent. over time (Figure 39). There was virtually no poverty Employed individuals had the lowest poverty rate among people living with a university-educated at 4.2 percent (Figure 29). In addition, the biggest household head. The rate was under 0.5 percent improvement was among the inactive population, in both 2007 and 2012; however but their share of with poverty rates falling from 10.1 in 2007 to 8.3 in poverty increased from 6.1 percent in 2007 to 8.5 2012. The faster poverty reduction among the inactive percent in 2012. Although it declined over time, the is most certainly associated with the growth of social share of poverty was highest among those living with protection benefits, but also reduce incentive to join household heads that completed primary or secondary the labor force, which in turn calls for additional education. benefits to compensate for lost labor income. As Figure 27: Education of head by income quintiles in 2012 90.0 82.5 80.0 70.0 60.0 50.0 % 40.0 32.7 30.0 27.3 26.6 23.5 21.5 20.1 20.0 16.2 25.0 22.8 12.8 23.8 11.9 10.0 18.2 8.0 4.8 16.8 4.1 0.2 1.4 0.0 Without Primary Secondary Tertiary Education of household head Poorest Quintile Q2 Q3 Q4 Richest Quintile Source: Authors’ calculations using HBS 2007 and 2012 Figure 28: Poverty by education of head 50.0 46.1 43.9 41.2 42.2 40.0 Share of population 2006-7 30.0 Poverty in 2006-7 20.0 % Share of 8.5 population in 2012 10.0 6.5 5.4 6.1 Poverty in 2012 0.0 Without Primary Secondary Tertiary Education of household head Source: Authors’ calculations using HBS 2007 and 2012 38 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 4- Poverty, Vulnerability and the Middle Class Figure 29: Poverty rates, by status of Figure 30: Distribution of poor, by status employment of employment 16.1 15.6 100.0 80.0 10.1 68.2 65.3 8.3 60.0 5.1 40.0 4.2 7.3 8.3 20.0 24.5 26.3 Employed Unemployed Inactive 0.0 2007 2012 2007 2012 Employed Unemployed Inactive Source: Authors’ calculations using HBS 2007 and 2012. Children below age 15 are excluded from this analysis. mentioned earlier, close to 75 percent of the poverty 92. Better-educated individuals tend to get the best reduction was associated with the social protection jobs, which often provide higher pay than the jobs benefits. In 2012, the inactive population represented employing largely less educated people. Some 65.3 percent of the poor. The group’s share of poor evidence of this hypothesis is provided by the HBS fell between 2007 and 2012 (Figure 30). Employment data provides. As shown in Table 3, 56 percent of those does not necessarily eliminate poverty—about 26 who had no education were employed in elementary percent of the poor were had jobs in 2012. The share occupations in 2012. In contrast, 56 percent of those of the poor increased for both the employed and who completed tertiary education were in managerial unemployed. and professional occupations, with only 1 percent in elementary occupations. This job data help explain why households headed by better educated individuals Table 3: Occupation and education Employment occupation Without Primary Secondary Tertiary Total Managers 0.27 0.69 4.37 15.67 4.70 Pro fes s io nals 0.00 0.31 5.11 40.06 8.31 Technicians /as s ociate 1.63 2.90 12.26 21.50 10.41 Cler ical wo r k er s 0.00 1.04 12.14 12.97 8.56 Serv ice/s ales wo rkers 18.05 13.26 25.95 6.50 19.14 Sk illed agr icu ltural 11.09 7.75 1.94 0.16 3.67 Trades wo r k er s 10.04 28.32 16.36 1.65 18.02 Operators and assemblers 2.98 11.88 9.30 0.47 8.80 Elementary o ccu pation 55.94 33.86 12.56 1.02 18.39 Total 100 100.00 100.00 100.00 100.00 Source: Authors’ calculations using HBS 2007 and 2012. commanded greater incomes than households headed 42). Among managers, professionals, technicians, by someone with less education (Figure 38). and clerical workers, poverty was under 2 percent in both 2007 and 2012 (Figure 32). But for those in 93. Poverty also varies widely across occupations. Among other occupations, poverty was higher. Among those those holding white-collar jobs, poverty was low; in skilled agriculture, for example, poverty was 14.4 poverty was quite high and tended to increase over percent in 2007, increasing 17.5 percent by 2012. In time among those in blue- collar occupations (Figure terms of worsening poverty, those in agriculture were MAURITIUS | Inclusiveness of Growth and Shared Prosperity 39 CHAPTER 4- Poverty, Vulnerability and the Middle Class Figure 31: Poverty by sector of activity 40 38.0 34.1 35 30 25 % 20 15.2 15 17.5 9.2 7.7 8.5 10.4 10 6.1 5 4.3 3.5 4.0 0 Agriculture Industry Trade Services Sector of activity 2006-7 2012 Share of poor (2012) Source: Authors’ calculations using HBS 2007 and 2012 Figure 32: Poverty by occupation Poverty by occupation 40 35 30 25 % 20 17.5 15 11.0 10.0 2006-7 10 2012 4.3 3.9 Share of poor (2012) 5 2.7 0.7 0.6 0.6 0.8 1.2 14.4 7.3 4.2 9.6 0.1 0.3 0.4 0 s ls te rs rs rs s ns rs er er na ke ke ke ia tio ke ag bl oc sio or or or or m pa an ss lw lw w sw se es cu M /a es ica ra as of oc le ns ad tu Pr er sa nd ry cia ul Tr Cl ta sa / ric ni ice en ch or ag em rv at Te d Se er ille El Op Sk Source: Authors’ calculations using HBS 2007 and 2012 40 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 4- Poverty, Vulnerability and the Middle Class hit harder than those working in trade and elementary households earning upper-middle class incomes. Of occupations. all middle-class households, 84 percent are headed by males and 16 percent by females. D. Who are the vulnerable and middle class in 96. Aging decreased the tendency to be either poor or vulnerable and increased the likelihood of being Mauritius middle class or above. Like many other countries, Mauritius has relatively high unemployment among 94. A number of different socio-economic characteristics young people aged 20 to 24, delaying their entry into of the middle and other income classes can be the labor market and their achieving middle-class identified, many of which have important policy status. Between 2007 and 2012, all age groups had implications. a similar pattern—declining shares among the lower middle class and increasing shares among the upper 95. Males and females are similar in vulnerability and middle class or vulnerable groups. The increased middle-class status, but there are significant differences shares of the upper middle class always exceeded between male- and female-headed households. In the share becoming vulnerable. For example, young 2012, over 30 percent of female-headed households people aged 15 to 24 had a decline of 8.7 percentage were either poor or vulnerable, compared to just points in the lower middle class, an increase of 4.0 19 percent for male-headed households (Figure 33). percentage points in those vulnerable, and a gain of In the middle class, 78 percent of households were 4.9 percentage points in those in the upper middle male-headed and 67 percent were female-headed class. All age groups except for those under age 15 households—with a much higher share of male-headed had decreases in the percent poor. Figure 33: The middle class by selected demographic characteristics, 2012 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 Male Female Male Female 0-15 15-24 24-35 35-64 64+ 1 person 2-3 persons 4-5 persons 6+ Head Head Gender Female Head Age Group Household size poor vulnerable low middle class middle class high rich Source: Authors’ calculations. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 41 CHAPTER 4- Poverty, Vulnerability and the Middle Class Photo : © Carla Schnetler 42 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 4- Poverty, Vulnerability and the Middle Class 97. Middle-class status fell with increased household and 2012, the only discernible trend of middle class size, especially for households exceeding six or more status by household size was an overall reduction in people. A high share of households with six or more average size, especially for large households with six people is poor or vulnerable, and only a small share or more people. This structural change likely explains is rich. The major difference between the vulnerable part of the overall increase in the size and share of and the middle class was household size. Of all poor the middle class. Over time, households with six or households, 49 percent had four to five persons, and 33 more persons had the largest decline in the share percent had six or more. Of all vulnerable households, among the lower middle class (4.7 percentage points) 51 percent had four to five persons, and 25 percent and the largest increase in the share vulnerable (4.2 had six or more. Large households make up a much percentage points). smaller portion of the middle class. Between 2007 Figure 34: The middle class by labor force and employment characteristics, 2012 Source of transfers as main source informal Income Main Work as main source Formal Employement Informal Employement Household services Private enterprise Employer Export oriented enterprise Public enterprises Public administration Elementary occupations Operators and assemblers Trades workers Skilled agricultural workers Occupation Service/sales workers Clerical workers Technicians/associate Professionals Managers Out of LF LF Status Unemployed Employed 0 10 20 30 40 50 60 70 80 90 100 poor vulnerable low middle class middle class high rich Source: Authors’ calculations. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 43 CHAPTER 4- Poverty, Vulnerability and the Middle Class 98. Being employed is obviously a key factor in achieving 100. Employment in public administration or public middle-class status; the unemployed are more likely enterprises is a key to middle-class status, with to be poor or vulnerable. In 2012, 80 percent of 75 percent of those working in these sectors being the employed were in the middle class, joined by upper middle class, compared with 53 percent in a surprising 64 percent of the unemployed (Figure private enterprises, 43 percent in export-oriented 34). This could be explained the country’s low rate firms, and 32 percent in household services. Looking of unemployment and lack of long-term joblessness. at type of employer finds a much larger portion Among the poor or vulnerable, only 17 percent of of the poor or vulnerable working in low-paying employed, while 35 percent are unemployed. Between household services. Among those holding these 2007 and 2012, employment increased and more jobs in 2012, 16 percent were poor and 20 percent people were in the labor force, which contributed were vulnerable. By comparison, only 1 percent of to the growth in the middle class. The employed, those in public administration or public enterprises unemployed, and those out of the labor force all were poor, and only 4 percent were vulnerable. had declines in the lower-middle class. The share of However, employment in public administration or unemployed among the vulnerable grew since 2007 public enterprises constitutes only a small portion of (Figure 35) Mauritius’ labor force. Roughly 85 percent of all jobs were in the private sector in 2012, an even higher 99. As opposed to receiving the bulk of income in the form share than in 2007. In this five-year period, household of transfers (various forms of government assistance), services saw an increase of 4.4 percentage points in working definitely leads to higher shares being middle the poor, an increase of 2.8 percentage points in the class and lower share being poor or vulnerable. Among vulnerable, and a decline in the middle class. All told, those receiving transfers as their main source of people employed in household services seem to be a income in 2012, 15 percent were poor and 17 percent particularly vulnerable group, and their vulnerability were vulnerable. With work as the main source of seems to be increasing. income, only 7 percent were poor and 12 percent were vulnerable. The shares of the lower middle class 101. M auritius is undergoing a long-term structural were roughly the same for those working and receiving transformation of its economy away from such primary transfer income. However, 54 percent of those who sectors as sugar and textiles and toward the tertiary worked were in the upper middle class, compared sector, mostly the financial services and tourism with 41 percent of those with transfer income as their industries that now dominate the economy. Vulnerability main source of income. is growing in agriculture and industry (Figure 36). Being Figure 35: The middle class by labor force status, 2007 and 2012 100 90 80 70 60 50 40 30 20 10 0 2007 2012 2007 2012 2007 2012 Employed Unemployed Out of LF poor vulnerable low middle class middle class high rich Source: Authors’ calculations. 44 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 4- Poverty, Vulnerability and the Middle Class employed in the growing sectors is a key to middle- 24 percent were either poor or vulnerable (9 percent class status. Eighty-one percent of those employed in poor, 15 percent vulnerable). Between 2007 and 2012, trade and 84 percent of those in services are middle those employed in trade made the greatest progress class, compared with 75 percent in agriculture or in rising out of poverty, with poverty rates declining services. Only 16 percent of those employed in trade 9 percentage points. In addition, trade led in those and only 12 percent of those in services were poor or moving into the upper middle classes, with a decline vulnerable. Among those employed in agriculture, 25 of 14 percentage points in those in the lower middle percent were either poor or vulnerable (12 percent classes and an increase of 20 percentage points for poor, 13 percent vulnerable); among those in industry, those in the upper middle class. Figure 36: The middle class sector of employment, 2007 and 2012 100 90 80 70 60 50 40 30 20 10 0 2007 2012 2007 2012 2007 2012 2007 2012 Agriculture Industry Trade Service poor vulnerable low middle class middle class high rich Source: Authors’ calculations. 102. Having a highly skilled occupation is also key to 103. A clear correlation exists between increased education middle-class status, with more than 70 percent of and making it into the middle class, especially for those managers, professionals, technicians/associates, and with a secondary education or higher (Figure 38). Among clerical workers being upper middle class, compared those not completing any education level, 70 percent with less than 44 percent of skilled agricultural workers, are middle class and only 41 percent upper middle class. trade workers, and those in elementary occupations Among those with a primary education, 71 percent are (Figure 37). Twenty percent of managers and 10 middle class. Moving up to the next levels, the middle- percent of professionals are upper middle class. class shares are 82 percent for secondary education Evidence points to a structural transformation of and 85 percent for tertiary. The evidence points to the the economy between 2007 and 2012. The percent influence of structural factors in reshaping the middle of those employed in elementary occupations and class, with decreases among those without much trade declined. Professionals increased—another education or just a primary education and increases structural factor possibly contributing to the growth among those with a secondary or tertiary education. of the middle class. The labor market is becoming The importance of increased education to middle- bifurcated, with a portion of the lower-skilled and class status will become even stronger in the future less-educated losing their jobs in such sectors as as the country continues its transformation from an textiles and labor shortages in the growing IT sectors. economy based on the primary sectors to one based This has implications for the size of the middle class, on tertiary sectors, such as financial services and IT. with some people moving into the upper middle class The rising sectors require a more educated workforce, while others with less skills becoming vulnerable. often with specific knowledge and skills. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 45 CHAPTER 4- Poverty, Vulnerability and the Middle Class Figure 37: Occupation by income group, 2007 and 2012 100 90 80 70 60 50 40 30 20 10 0 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 2012 2007 MANAGERS PROFESSIONALS TECHNICIANS CLERICAL WORKERS SERVICE/ SKILLED AGRICULTURAL TRADES WORKERS OPERATORS AND ELEMENTARY AGRICULTURE /ASSOCIATE SALES WORKERS WORKERS ASSEMBLERS OCCUPATIONS poor vulnerable low middle class middle class high rich Figure 38: Education by income group, 2007 and 2012 100 90 80 70 60 50 40 30 20 10 0 2007 2012 2007 2012 2007 2012 2007 2012 WITHOUT PRIMARY SECONDARY TERTIARY Education poor vulnerable low middle class middle class high rich Source: Authors’ calculations. 46 MAURITIUS | Inclusiveness of Growth and Shared Prosperity Chapter 5 Causes of Poverty and Vulnerability Changes Lagging shared prosperity had an adverse impact on poverty in Mauritius. The reduction of poverty would have been almost twice as large and stronger if inequality had not worsened. The most important contributor to poverty reduction was social protection incomes and subsidies, contributing 74.1 percent. Labor Incomes also reduced poverty, but their contribution was 18.2 percent. Demographic changes associated with decreasing household size and lower dependency ratios contributed 17.2 percent. Rising labor and self-employment incomes among initially better- off groups contributed to the increase in the inequality. The deterioration of the traditional agriculture and textile industries and improvement of the trade and services sectors had the most prominent impact on the increase in inequality. Economic growth and diminished inequality are equally important for poverty reduction and its possible eradication in Mauritius. Assuming a neutral growth scenario, 40 percent cumulative growth in consumption per capita would be required to halve the poverty rate from 7.9 percent to 4 percent. Based on these assumptions, it will take close to 15 years to completely eradicate poverty. However, our projections do not suggest inequality will decline, and poverty is not expected to fall fast. Micro- simulation analysis suggests that improved targeting of the social protection, combined with a decline in unemployment could result in significantly lower poverty in Mauritius, measured against a baseline scenario. CHAPTER 5 - Causes of Poverty and Vulnerability Changes This chapter looks at the factors behind the observed poverty composition of employment? Finally, we look forward and reduction and distributional changes in Mauritius. The project changes in poverty and inequality based on macro incidence of poverty fell but inequality increased between scenarios. 2007 and 2012. We used several methods to quantify the contributions of different factors in poverty reduction. The start with the Datt-Ravallion (1992) standard decomposition A. The role of growth and inequality in poverty method, which analyzed the role played by the growth and redistribution factors in poverty reduction. This analysis changes determines how growth vs. redistribution affected poverty. Later, we analyze the sources of income and economic forces 104. From the previous analysis, we have concluded behind the observed changes. Was the reduction in poverty that inequality has increased in Mauritius; the growth a result of higher employment, higher earrings, or higher incidence curve had a regressive pattern (downward public transfers and remittances? And finally, we attribute shape), suggesting relative deterioration among the the changes to the sectoral wage premiums and social poorest households. Now, we decompose the absolute protection. Were these changes a result of improved human poverty changes into growth and redistribution capital characteristics or higher returns to education and components to quantify their impact on poverty. how are the results associated with changes in the sectoral Box 5: Measuring growth to poverty elasticity Traditionally, poverty economists project the incidence of poverty as a function of economic growth, using the consumption-to- poverty elasticity, an empirically measured index. It quantifies how much poverty reduction occurs for each 1 percent increase in per capita consumption. The responsiveness of poverty reduction to growth is a function of the number of people living just above poverty line. If the elasticity is high, poverty responds strongly to economic growth. If it is low, even strong growth will be relatively ineffective in reducing poverty. The impact of growth on poverty can also be captured by estimating the elasticity of growth to poverty, a measure of the reactivity of poverty with respect to changes in the average per capita expenditure. Following Duclos and Araar (2006), we used the following general formula for the elasticity : where P(a) is the Foster-Greer-Thorbecke poverty measure with parameter α, f(z) and F(z) denote, respectively, the probability density function and the cumulative density function of per capita expenditure and z is the absolute poverty line. Growth to poverty elasticity in Mauritius Mauritius’ relatively moderate inequality is associated with relatively high elasticities to growth and inequality, and the elasticity to growth fell over time due to increasing inequality. Mauritius has a relatively high elasticity of poverty to consumption expenditures—at -3.35 in 2007 and -3.10 in 2012 (Figure 41). This corresponds to the elasticity levels of Eastern Europe’s middle-income countries (Ukraine and Russian Federation); it is much higher than other middle-income countries in Africa. For example, Botswana’s elasticity is about 1 percent. Elasticity is twice as high in Mauritius than in other countries. This means that a relatively high proportion of the population is living close to the poverty line, and small growth rates lead to rapid poverty reduction. The opposite is true as well; relatively small adverse changes in growth have a strong impact on poverty. The worrisome sign, however, is that elasticity fell between 2007 and 2012 due to the increase in inequality. Mauritius has even higher elasticity of poverty to inequality, and it has remained almost unchanged over time (Figure 40). Inequality really matters, pointing to the importance of developing social and economic policies that foster pro-poor growth. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 49 CHAPTER 5 - Causes of Poverty and Vulnerability Changes Figure 39: Elasticity of poverty to consumption Figure 40: Elasticity of poverty to inequality growth, growth, 2007-12 2007-12 7.3 7.3 6.4 1.0 0.2 0.2 0.2 6.3 0.0 -1.0 5.0 5.0 -2.0 -3.0 -4.0 -3.1 -3.1 -3.2 -3.3 -3.2 -3.4 -5.0 -6.0 -7.0 -8.0 Poverty Poverty Gap Squared Poverty 0.1 0.0 Headcount Gap 0.0 2007 2012 Change Poverty Poverty Gap Squared Poverty Headcount Gap Source: Authors’ calculations. 2007 2012 Change Figure 41: Growth inequality decomposition a. Income poverty change,2007-12 b. Consumption poverty change, 2007-12 4.0 4.00 3.0 2.8 3.00 Change in absolute poverty 2.0 Change in absolute poverty 2.00 1.0 1.00 0.55 0.32 0.0 0.00 -1.0 -0.35 -1.00 - 0.9 -1.22 -2.0 -1.6 -2.00 -3.0 -3.00 -4.0 - 3.5 -4.00 Total Growth Inequality Interaction Total Growth Inequality Interaction Source: Authors’ calculations. Changes in poverty can be decomposed (i) the growth component (GC), which identifies the poverty change into growth and inequality (or redistribution) components. We used due to the growth of mean per capita expenditure, and (ii) the growth-inequality decomposition method introduced by Datt and inequality component (IC), which identifies the poverty change due Ravallion (1992), which quantifies the relative contributions of to a more equal distribution of income. It is important to emphasize economic growth and redistribution (e.g., changes in inequality) to that the redistribution component is not necessarily associated changes in poverty. The method decomposes the observed changes with expansion of government transfers; it measures the impact of in poverty into two components: general inequality changes on poverty. 50 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 5 - Causes of Poverty and Vulnerability Changes Photo : © Haja Faniry Razafimahenina MAURITIUS | Inclusiveness of Growth and Shared Prosperity 51 CHAPTER 5 - Causes of Poverty and Vulnerability Changes 105. Poverty reduction in Mauritius would have been much poverty headcount. Similar results were obtained for stronger if inequality had not worsened. As illustrated the poverty gap and squared poverty gap measures. in Figure 41, the income-based absolute poverty headcount decreased by about 1.6 percentage points between 2007 and 2012. However, the reduction in B. Drivers of changes in poverty—decomposing poverty would be almost twice as large if inequality had not worsened. Focusing on the growth component, poverty reduction25 the decomposition analysis suggests poverty would have fallen by 3.5 percentage points if other things 106. This section will further analyze the observed were unchanged. The redistribution (inequality) poverty changes by applying decomposition component was 2.8 percentage points, suggesting techniques. Was the reduction in poverty a result of how much poverty would have increased if not for higher employment, higher earrings, or higher public economic growth. Because the growth component transfers? Several methods provide an accounting was stronger that the redistribution one, the of how much of the total change in poverty can be overall effect was a moderate reduction in poverty. attributed to different groups or factors. We used Similar conclusion could be reached for absolute the methodology developed by Azevedo et al. to consumption poverty (Figure 41B). The growth calculation income poverty decompositions and component reduced consumption poverty, while quantify the contribution of different factors to the redistribution component tended to increase it. changes in poverty and inequality (Figure 42 and As a result, consumption poverty was lowered only Figure 43). slightly between 2007 and 2012. In both cases, the redistribution component associated with changes 25 Azevedo et al. method was used in this study to decompose in inequality had a strong adverse impact on the poverty changes. Figure 42: Contribution to poverty reduction in percent, 2007-12 Government Transfers 74.1 Employment income 18.2 Demographics 17.8 Own produced 6.3 Property income 3.1 Others incomes - 5.4 Self employment - 14.2 -40.0 -10.0 20.0 50.0 80.0 % contribution to poverty reduction Figure 43: Contribution to inequality increase in percent, 2007-12 Employment income 88.0 Self employment income 28.0 Others incomes 20.0 Own produced 0.0 Demographics 0.0 Property income - 12.0 Government Transfers -20.0 -40.0 -10.0 20.0 50.0 80.0 % contribution to inequality increase 52 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 5 - Causes of Poverty and Vulnerability Changes Figure 44: Contribution to reduction in economic vulnerability reduction in percent, 2007-12 Employment income 46.3 Transfer 30.6 Demographic 17.1 Self imployment income 7.6 Own produced 1.2 Property income 1.1 Others incomes -3.9 -40.0 -10.0 20.0 50.0 80.0 % contribution to poverty reduction Source: Authors’ calculations based on micro-decomposition analysis (Azevedo et al.). 107. Social protection (SP) had the largest impact on 108. Disproportionate increases in labor and self- poverty reduction in Mauritius; employment incomes employment incomes among the better-off population and demographic factors had significant impacts on increased inequality in Mauritius. As shown in Figure poverty reduction. National poverty declined by 1.6 43, the rising Gini coefficient (inequality increase) was percentage points between 2007 and 2012. The associated with employment income (88 percent of the most important contributor was social protection change) and self-employment income (28 percent of incomes and subsidies, which accounted for 74.1 the change). However, government transfers worked percent of poverty reduction. Labor incomes also in the other direction, reducing inequality. In other reduced poverty, with a contribution of 18.2 percent. words, wages grew among the better off, increasing Demographic changes associated with decreasing inequality; meanwhile, social transfers benefited the household size and reduced dependency ratios were poorest, but the magnitude was not large enough to at 17.2 percent. Other sources of income had very offset the impact of higher wages. small impacts on poverty, while self-employment incomes actually increased poverty. Photo : © Haja Faniry Razafimahenina MAURITIUS | Inclusiveness of Growth and Shared Prosperity 53 CHAPTER 5 - Causes of Poverty and Vulnerability Changes 109. Relative contributions to poverty reduction differed for the poor rely on the SP system, and the expansion the poor and vulnerable groups. As presented in Figure of the transfers significantly improved their welfare 44, changes in employment incomes and transfers and reduced poverty. The vulnerable group, which is both played important roles in vulnerability trends. a little better off, relied significantly on transfers and Demographic changes had a very significant impact labor incomes. In fact, labor incomes had the largest on both poverty and vulnerability. In other words, impact on the group’s poverty reduction. Figure 45: Contribution to poverty reduction by groups in percent, 2007-12 70.0 80.0 Sector 71.7 Labor Status 60.5 Hours of work Contribution to change in poverty 80.0 71.6 Contribution to change in poverty Contribution to change in poverty 60.0 60.0 45.8 50.0 60.0 40.0 40.0 40.9 19.7 40.0 33.4 20.0 20.0 30.0 6.1 0.0 0.0 20.0 -20.0 -20.0 10.0 -22.9 1.2 -40.0 -40.0 - 28.1 0.0 less 100 100-150 150-200 200+ hours Agriculture Industry Trade Service Employed Unemployed Out of LF hours hours hours Household's size 30.0 Age group Education 41.0 Contribution to change in poverty 24.4 Contribution to change in poverty 45.0 Contribution to change in poverty 45.0 40.1 25.0 22.4 40.0 21.2 40.0 35.0 20.0 35.0 30.0 29.9 30.0 25.0 23.6 15.0 12.0 25.0 20.0 10.0 20.0 16.2 16.5 15.0 7.0 15.0 10.0 5.0 10.0 5.0 1.3 5.0 0.0 0.0 0.0 1 person 2-3 persons 4-5 persons 6+ 0-15 15-24 24-35 35-64 64+ -5.0 - 0.7 Without Primary Secondary Tertiary Contribution to change in poverty 40.0 33.3 35.0 Occupation 28.0 30.0 25.0 20.0 Ethnicity 15.0 10.7 45.0 41.6 10.0 2.5 5.0 0.5 0.8 40.0 0.0 -5.0 35.0 32.3 30.2 -1.5 -0.4 Contribution to change in poverty -10.0 -3.8 30.0 s als s w te rs sw s s rs ns er r er ke ke ble ale cia tio n k ag sio or or or o m 25.0 pa an s lw se es as cu M ra as of er ns/ oc e ad ltu Pr 20.0 nd ry l/s ia Tr icu nic ta sa ica en gr ch or da 15.0 em at Te Cl er ille El Op Sk 10.0 4.5 5.0 Contribution to change in poverty 30.0 25.8 District 25.0 0.0 20.0 16.6 19.2 u lim n an 15.0 12.8 nd io i 10.4 rit us at Hi 8.7 8.1 ul au M 10.0 7.3 p M Po - 5.0 no 0.2 l ra Si ne 0.0 Ge -5.0 -10.0 -8.4 -15.0 s es rt q rt e s a r s ive ui em ue ok nn ac pa Po ss Lo Fl M ig kR va m ilh ou d dr rt Sa Re an sW ac em Ro Po Gr Bl du pl ne m re ai Pa vie Pl Ri Source: Authors’ calculations based on micro-decomposition analysis (based on Ravallion and Huppi (1991) methodology). 54 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 5 - Causes of Poverty and Vulnerability Changes 110. Labor-market changes associated with the primary role. The small island of Rodrigues was the only location sector’s deterioration and improving returns in the with negative impact on poverty in Mauritius. secondary and tertiary sectors had important impacts on poverty. Another way to decompose changes in poverty is based on a methodology developed by Linking growth, inequality, and poverty C. Ravallion and Huppi (1991), which attributes poverty changes to population groups. We decomposed changes—poverty trace analysis poverty by households’ labor demographic, education, and location characteristics. Figure 44 presents the 111. The poverty trace curve (PTC) provides an evocative contribution of each factor to total poverty changes. graphical summary of projected poverty dynamics. It is important to emphasize that this analysis looks The method associates economic growth and changes only at changes in poverty and not at the entire in income inequality to poverty reduction.26 PTC income distribution. Already poor groups have a analysis demonstrates how different combinations of much stronger impact on poverty, while changes economic growth and inequality reduction will affect among initially non-poor groups do not impact poverty poverty in Mauritius. PTC analysis has an additional statistics. For example, highly educated and skilled dimension as a way to gauge how proposed policies workers are not among the poor in the first place, and would impact poverty. It could be used in various their relative income gains affect income inequality areas of distributional analysis. In conjunction with but not poverty reduction. It is important to keep this micro simulation, for example, PTC could measure the observation in mind when interpreting the results of degree to which a proposed social assistance subsidy the group decomposition analysis. The findings are could reduce poverty. summarized in the following bullet points: • Labor factors. A decomposition analysis by employment Box 6: Poverty trace curve (PTC) sector shows that a larger share of poverty reduction analysis is a poverty approach to could be attributed to the trade and services sectors, shared prosperity while agriculture had an adverse impact on poverty. Gains by the employed and those out of the labor force PTC analysis is based on the iso-poverty curve approach contributed to poverty reduction, while unemployment that links changes in inequality to the shared- prosperity led to a worsening of poverty. Working more hours indicators—particularly, the growth rate of the bottom contributed to overall poverty reduction. Gains by those 40 percent. The common issue related to iso-poverty is in the elementary occupations played a significant role the specific mechanism by which the Gini index is linked to in poverty reduction. However, the impact of wage the transformation of the income distribution. A reduction premiums for the better educated is not reflected of the Gini index can be caused by different changes in in this analysis because highly skilled people are not the distribution. For example, a transfer of incomes can among the poor (income inequality picks up the impact take place between the extremes of the distribution or of these higher incomes). between the middle-income groups of the population. • Demographic factors. Demographic factors played an The impact on poverty will be much greater in the first important role in poverty reduction. Gains by larger case than in the second. PTC model implicitly postulates households and relatively younger people had a positive a strong relationship between changes in the Gini index impact. However, the elderly group contributed less and their poverty effects by changing the relative income to poverty reduction. Improvements in the general of the bottom 40 percent of the distribution. population had the most important role. • Education. Tertiary-educated people are generally not among the poor, and their advances did not show in the poverty statistics. However, improvements among the group with primary education played an important 112. PTCs for Mauritius are presented in Figure 46. The set role in poverty reduction between 2007 and 2012. of charts illustrates how much the projected growth The reduction would have been even larger if labor rate in consumption per adult equivalent will reduce productivity and wages would have increased more poverty based on the different assumptions of Gini equally among the less educated. The wage is low distributional changes associated with various growth because the productivity is very low. Better education, rates among the lowest 40 percentile of the income skills and opportunities should help to raise wages for distribution. The official poverty datum line was used those less educated. • Location and poverty. Mauritius is a small island, and 26 PTC analysis is based on the widely used iso-poverty ap- proach, a statistical decomposition of the economic changes geography did not play a significant role in poverty required to achieve a target poverty rate (P*). The iso-poverty changes. As the charts show, most districts contributed breaks down the required changes in mean growth (b) and ine- to poverty reduction, with Port Louis taking the leading quality (a) to achieve the target poverty reduction. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 55 CHAPTER 5 - Causes of Poverty and Vulnerability Changes for this analysis. Based on consumption poverty, the (i.e., no change in the aggregate consumption current situation corresponds to 7.9 percent poverty distribution), with an annual growth rate of 4 percent, and 0.33 inequality. The PTC curves are drawn based would take from seven to nine years to halve poverty. on various inequality levels that are associated with The relationship between neutral growth and poverty, alternative growth paths of the bottom 40 percent. however, is not linear, so it would take close to 15 years The central red line depicts the poverty projections to further reduce and finally eradicate poverty, based associated with the neutral growth scenario— on the neutral-growth scenario. If growth were pro- cumulative changes in the consumption per adult poor and associated with reduced inequality, however, equivalent with constant inequality. The blue lines the poverty could be brought down at a much faster below the neutral growth scenario are related to pace. For example, poverty could be halved in four reduced inequality and are pro-poor growth. The lines years if growth is associated with a reduction of 2 above the neutral growth scenario are associated with percentage points in the Gini coefficient. Similarly, the increasing inequality and are anti-poor growth. growth associated with increasing inequality would result in a deterioration of the poverty situation. PTC analysis suggests that the range of poverty change is 113. The PTCs indicate that economic growth and reduced wide and depends on inequality changes. Pro-poor inequality are equally important for alleviating poverty. growth is essential for fast poverty eradication. As presented in Figure 46, the neutral-growth scenario Figure 46: Mauritius poverty trace curves (PTC) (consumption poverty) 40 Poverty Trace Curves (PTC) 35 Estimated Headcount Poverty Rate in % 30 Ver 1 &=0.8 Gini=.37 Ver 2 &=0.9 Gini=.35 25 Ver 3 &=1.0 Gini=.33 (neutral) Ver 4 &=1.1 Gini=.32 Ver 5 &=1.2 Gini=.30 20 2012 situation (poverty 7.9 and Gini 0.33) 15 10 5 0 -50 -30 -10 10 30 50 70 90 110 130 Cummulative Consumption Per adult Equivalent Growth Rate, in % Source: World Bank staff calculations. In the analysis, a is the disproportionate ratio of consumption growth of bottom 40 percentile of the consumption distribution; a =1 indicates neutral growth, a<1 indicates increasing inequality, and a>1 suggests reduced inequality. The neutral growth scenario is depicted in the red line. Higher inequality shifts the PTC lines upwards, while lower inequality pushes them downwards. The higher the inequality, the more consumption growth required to reduce poverty. 56 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 5 - Causes of Poverty and Vulnerability Changes 114. The main conclusion of the PTC analysis is that the of poverty and distributional changes under a distribution of income plays an important role in the number of policy scenarios. The simulations show distribution of wealth, and policies geared toward that generating more growth would significantly reducing inequality are required to eradicate poverty. accelerate the pace of poverty reduction, although The extent of poverty reduction depends equally on the income distribution may become more unequal economic growth and inequality changes. The PTC due to sectoral transformations. This is why both analysis presented considerable evidence that the human development and redistributive policies may distribution of income has a significant influence on have to be put in place. Such transfers should further poverty. More rapid poverty reduction requires more reduce poverty and improve the income distribution. growth and a more pro-poor pattern of growth. Policies geared toward reducing income inequality will result 116. Macro projections suggest further expansion of in greater poverty reduction. Some examples of the the tertiary sector. We analyzed the poverty and types of policies that can promote poverty reduction distributional changes associated with the macro by reducing inequalities follow in the next section. projections presented in Table 4: Macro projections, baseline scenario. We used the medium fertility variant from UN population projections to simulate changes in D. Looking ahead: how to tackle poverty while the age-gender structure of the population. We used boosting the middle class IMF’s GDP projections published in the October 2014 World Economic Outlook. We also used time series historical data to model sectorial changes. The macro 115. This section looks at the possible poverty and projections suggest expansion in the tertiary sector inequality outcomes associated with sectoral growth and further deterioration in primary and secondary trajectories, changes in demographic characteristics, sectors. and labor dynamics. We also simulate trajectories Table 4: Macro projections, baseline scenario 20 1 5 2 01 9 20 1 2 (pro je cte d ) (proj ec t ed ) GDP Per Capita (constant prices) 149,222 162,775 187,793 Cumulat ive GDP Per Ca pita Gr owth r ate (201 2=1 0 0 ) in % base 9.1 25.8 Sectoral GDP Growth Rates in % Primary base -0.7 1.1 Secondary base 8.3 23.7 Tert iary base 9.7 27.3 Sectorial Employment Growth Rates in % Primary base 6.7 5.9 Secondary base 25.5 23.3 Tert iary base 67.8 70.9 Populat ion Growth rate, % base 0.9 2.2 Unemployment rate in % 8.0 8.0 8.0 Source: Authors’ calculations. The information on sectoral economic growth comes from the World Bank’s Lesotho economic projections. The age-gender population growth rates used for the microsimulation are not presented. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 57 CHAPTER 5 - Causes of Poverty and Vulnerability Changes Box 7: ADePT microsimulation approach used to project poverty and inequality We used a micro-simulation approach to project the distributional impacts of macro shocks in Mauritius through 2019. Micro- simulation is the process by which information on aggregate projected changes in output, employment, and transfers are used to generate changes in labor and non-labor income at the micro level through structural models. The microsimulation is conducted in four main steps. First, we used population growth projections to adjust for demographic changes between the base year of 2012 and 2015 and 2019. This allowed us to explicitly take into account changes in dependency ratios and adjust for changes in the size of the working-age population. Second, we used the projections from labor-force status and labor earnings models to replicate projected changes in aggregate total and sectoral output and employment. We calculated the total number of individuals that needed to be reassigned between employment and non-employment and across employment sectors to match projected aggregate changes in total and sectoral employment. Initially non-employed individuals could become employed, employed individuals could become non-employed, and individuals could change sectors. Third, we used the earnings model to predict earnings for two groups of workers: those with no previous earning history and those who change sectors. Earnings are a function of both observable and unobservable individuals. Once all workers were assigned positive labor earnings, total earnings in a sector were adjusted to match aggregate projected changes in output. Fourth, we simulated the distributional impact to changes in social protection and in sources of non-labor income. We assumed that capital and financial income would grow at the same rate as real GDP, international remittances would remain constant in real terms at 2012 levels, and domestic remittances would change at the same rate as labor income. The simulation module of ADePT software was used to conduct microsimulation exercise. The software documentation can be /spark.worldbank.org/groups/poverty/projects/micro-simulations. found through this link: https:/ Figure 47: Poverty and inequality projections, baseline scenario 10 40 Poverty 39 9 7.9 Poverty headcount (consumption) 8 7.1 38 6.5 37 7 Inequality (Gini) 6 36 5 Inequality 35 4 35.2 34 3 34.0 33 2 33.3 32 1 31 0 30 2012 (actual) 2015 2019 Source: Authors’ calculations. Poverty (left axis) Gini (right axis) 58 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 5 - Causes of Poverty and Vulnerability Changes Figure 48: Poverty simulations based on selected policy scenarios 9.0 Poverty reduction, % points 8.0 7.0 2.5% 6.0 5.0 4.0 3.0 2012 2015 2019 Baseline 7.9 7.1 6.5 SP Expansion 7.9 6.1 4.5 Unemployment 7.9 6.8 6.3 reduction Sectoral 7.9 6.7 6.3 Productivity Combined effect 7.9 5.5 4.0 Source: Authors’ calculations. 117. In the baseline scenario, poverty is expected to gradually impact of various policy interventions. We analyzed three decline while inequality increases. The results of the policy interventions: (1) a gradual 30 percent expansion micro-simulation exercise are presented in Figure of social-protection spending without improvements in 47. Consumption poverty is expected to decline from targeting efficiency; (2) a 20 percent reduction in the 7.9 percent in 2012 to approximately 7.1 percent unemployment rate by; and (3) an increase in primary in 2015, falling further to 6.5 percent in 2019.27 and secondary sectors’ labor productivity. The three The baseline scenario points to an increase in Gini- proposed scenarios should be used for illustrative measured consumption inequality from 33.3 in 2012 purpose. None of them is an ideal one that could fully to 35.2 in 2019. The estimated reduction in poverty respond to the challenges of the increase in inequality is primarily driven by cumulative growth in per capita the Mauritian economy is currently facing. Figure 48 consumption, associated with rapid growth of tertiary presents the poverty figures associated with the above sectors. The main driver of the expected increase scenarios. The analysis suggests that the policy measures in inequality is the rising disparity in consumption could almost double the speed of poverty reduction in between sectors and continued relative deterioration Mauritius. The most significant direct impact on poverty of the primary sectors. The projected increase in could be attributed to the improvement in targeting of inequality will have adverse effects on the pace of the social-protection, but the other measures also have poverty reduction. To reverse the trend, policies to positive impacts on poverty.28 reduce income inequality will be required. 118. Targeted policy interventions could boost poverty reduction in Mauritius. The micro-simulation exercise first enables us to simulate the baseline scenario associated with macro projections and then evaluate the distributional 28 The analyzed policy interventions are definitely not exhaus- 27 Similar trend is observed in case of income poverty. tive. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 59 CHAPTER 5 - Causes of Poverty and Vulnerability Changes 119. Rapid poverty reduction requires more growth and a back into poverty. While the upper middle class has more pro-poor pattern of growth. Mauritius has one of grown to 55 percent of the total population, the lower Africa’s highest elasticities of poverty to consumption middle class is struggling to retain its status and some expenditures. This means that a relatively high are falling back into the vulnerable group. Ensuring proportion of the population is close to the poverty that the progress achieved in reducing poverty and line and consumption growth rates translate into creating a middle class is not reversed will require rapid poverty reduction. This elasticity, however, fell unlocking the potential to accelerate economic between 2007 and 2012 due to increasing inequality, growth, including policies to benefit the bottom 40 diminishing the impact of economic growth on poverty percent of the population. reduction. Inequality-reducing polices in conjunction with fast economic growth would accelerate poverty 121. The next chapters explore in more detail the potential reduction in Mauritius. for moving forward on an agenda for accelerated economic growth that translates into poverty 120. A backlash cannot be ruled out if broadly shared reduction and an expanding middle class, assessing economic growth slows down. The high poverty to the macro-environment for sustainable growth and consumption expenditures elasticity in Mauritius the micro-level environment for individuals to take an means that relatively small adverse changes in advantage of that growth for productive employment. consumption growth may also have strong impacts on poverty. The majority of households who have escaped poverty remain vulnerable—at risk of falling Photo : © Haja Faniry Razafimahenina 60 MAURITIUS | Inclusiveness of Growth and Shared Prosperity Chapter 6 Social Protection in Mauritius Mauritius’ social protection (SP) programs played a significant role in ensuring the poor benefit from economic growth through redistribution of resources. Convergence, average transfers, and generosity of the social protection system all significantly increased between 2007 and 2012. Close to 75 percent of the poverty reduction could be linked to SP expansion in Mauritius. Without the SP system, poverty would be almost three times higher, and inequality would be 12 percent higher. While the majority of poverty reduction is attributable to Social Assistance (SA) schemes, social insurance (SI) programs were responsible for a minor improvement in poverty outcomes. The weak poverty focus of individual SA schemes and a lack of coordination across programs undermine the effectiveness of social safety nets. Proxy means tested SA programs should be developed to reduce inequality and improve the programs’ efficiency. CHAPTER 6 - Social Protection in Mauritius 125. The Government of Mauritius invests considerable A. Mauritius’ social protection system resources in SP. Expenditures were nearly MUR20.3 billion30 in 2013, accounting for more than 122. Mauritius has a comprehensive SP system that 20 percent of total government spending and 5.5 addresses key risks individuals face over the lifecycle percent of GDP.31 Excluding public-servant pensions, as well as exogenous shocks—price shocks, natural the Government spends MUR12.8 billion (3.5 percent disasters, occupational hazards, etc. The system of GDP) on the remaining schemes, with MUR8 billion consists of a diverse mix of contributory social (2.2 percent of GDP) allocated to the BRP. insurance (SI) programs and non-contributory social assistance (SA) schemes. 126. The Government’s SP commitment has led to an expansion of coverage of both SI and SA schemes. From 123. Contributory SI schemes include National Pension 2007 to 2012, SI coverage grew from 12.3 percent to Fund (NPF), the National Savings Fund (NSF), and 18.8 percent while the SA coverage expanded from voluntary retirement pensions for employees, 40 percent to 40.5 percent. As a result Mauritius civil servants and local government workers. has an extensive SP system that reached nearly half Unemployment benefits are provided through the of the population in 2012—46.6 percent, counting Transitional Unemployment Benefit program for direct and indirect beneficiaries (i.e. recipients of up to one year. The NSF also provides lump-sum benefits and their household members). unemployment payments. Empirical analysis of SI schemes focuses on benefits received through the 127. In the absence of the existing SI and SA schemes— NPF and from former employers (EF).29 assuming no other changes— poverty in Mauritius would be significantly higher. The absolute poverty 124. SA covers several broad types of programs: cash headcount would be 16.4 percent, rather than the transfers, in-kind transfers, labor-market activation, current 6.9 percent. The poverty gap would have community development, and CSO support programs. likely be quadruple the actual rate—6.5 percent and The empirical analysis of SA covers only the main not 1.7 percent. The Gini coefficient—an aggregate programs—the non-contributory Basic Retirement measure of the economy’s income inequality—would Pension (BRP), Widows and Children Pension (WCP), be 0.41 rather than the actual 0.37. Invalid Pension (IP), Social Aid (SAP), other social pensions, and scholarship grants and subsidies on examination fees, textbooks, etc. The full list of Mauritius’ SA programs is provided in Appendix A. Figure 49: Coverage of social protection, social insurance, and social assistance Total Coverage Coverage by Income Decile 50 46.6 45.5 45 60 40.9 40.0 40 50 35 30 40 Percent 25 30 20 18.8 15 12.3 20 10 10 5 0 0 Y2007 Y2012 Y2007 Y2012 Y2007 Y2012 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 SP SI SA Decile of income per AE SI 2007 SI 2012 SA 2007 SA 2012 Source: Authors’ calculations 30 Authors’ calculations based on data provided by the Mauritius Ac- countant General’s Office. 31 Total government expenditures were MUR102.9 billion in 2013. GDP is estimated at 366.5 billion in 2013 (based on the national accounts 29 Health risks are primarily addressed through the public health sys- data published by Statistics Mauritius at http://statsmauritius.gov.mu/ tem used by approximately 85 percent of Mauritians. English/StatsbySubj/Pages/qna-4qtr13.aspx). MAURITIUS | Inclusiveness of Growth and Shared Prosperity 63 CHAPTER 6 - Social Protection in Mauritius 128. The effectiveness of Mauritius’ SP system in reducing 129. While SA schemes account for most of the poverty poverty has increased over time. A comparison of reduction, SI programs also produce minor the simulated SP poverty impact in 2007 and 2012 improvements in poverty outcomes. The coverage indicates that the contribution of SI and SA programs of SI schemes among the poorest decile increased has increased. In 2007, SP programs were associated from 2007 to 2012, although it remained significantly with a decline of 8.4 percentage points in poverty; lower among the poorer population than among the in 2012, poverty would by higher by 9.5 percentage better-off groups. points in the absence of SP schemes. Over the same period, the poverty-reducing impact of SA schemes alone has increased from 6.5 to 6.9 percentage points. Figure 50: Simulated poverty and inequality impacts in the absence of SP, SI, and SA programs Poverty Headcount, 2012 Poverty Gap, 2012 7% 6.5% 18% 16.4% 6% 16% 14% 13.8% 4.9% 5% 12% 4% 10% 8.2% 8% 6.9% 3% 2.0% 6% 2% 1.7% 4% 1% 2% 0% 0% Without Without Without Without Without Without SP SI SA SP SI SA Actual Counterfactual Actual Counterfactual Reduction in poverty headcount attributable to Gini, 2012 SP in 2007 and 2012 (percentage points) 0.45 0.41 0.40 0.4 0.37 0.37 6.9% Without SA 0.35 6.5% 0.3 0.25 1.3% Without SI 0.2 1.2% 0.15 0.1 9.5% Without SP 8.4% 0.05 0 0% 2% 4% 6% 8% 10% Without Without Without SP SI SA Y2007 Y2012 Actual Counterfactual Source: Authors’ calculations. 64 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 6 - Social Protection in Mauritius 130. The improving SP poverty-impact performance substantial proportion of their benefits to relatively suggests that reforms implemented in 2007-12 well-off groups; (c) they also remain largely have been effective, but further gains can be fragmented and lack coordination, which leads to achieved through continued modernization of both gaps in coverage and/or overlapping coverage. SI and SA schemes. The discussion below reviews the performance of key SP programs and their 132. By far, the largest SA scheme is the BRP, a universal contribution to poverty reduction; it then discusses non-contributory social pension paid by the what can further promote SP’s role in inclusive Government to persons over age 60 (Table 8).32 growth. In addition to the retirement benefits, the BRP provides universal invalidity and survivor benefits, plus a host of untargeted allowances for recipients B. Social assistance of basic pensions, including career allowances, child allowances, and other benefits. The BRP is intended 131. The Government allocated MUR13.9 billion (3.8 to provide a basic minimum of protection to the percent GDP) to SA spending—one of the highest elderly, while contributory schemes provide the “top totals for low- and middle-income economies. The up.” Universal benefits under the BRP account for SA expenditure has paid off in terms of poverty 80 percent of all social assistance spending, with 58 reduction; however, the poverty-reducing effect percent going to retirement benefits alone. could be significantly higher for the amount of financial resources spent. There are several main reasons for it: (a) Mauritius spends majority of its 32 A retirement age of 65 is being phased in. Currently, the new retire- SA budget on programs that do not target the poor; ment age applies to contributory pensions only. The analysis refers to (b) even those interventions intended to benefit 2012 and therefore it does not include pension increases granted in the poor tend to be small in coverage and extend a 2015. Table 5: Composition of SA benefits, 2013 Composition of SA benefits, 2013 Amount, billion Percent BRP 11.23 81% R etirem ent benef its 8.03 58% Other benefits 3.20 23% Non-BRP 2.64 19% SA – So cial Aid benefits 1.26 9% Other benefits 1.39 10% Total SA 13.87 100% Source: Authors’ calculations based on data provided by the Mauritius Accountant General’s Office. 133. As a universal benefit, the BRP achieves broad households, the old-age pension contributed 44 coverage and delivers substantial benefits to the poor. percent of their per capita income. Assessed against By virtue of being universal, the BRP has a very broad the poverty line, old-age pensions account for 27 coverage, and 98 percent of all persons over age 60— percent of the poverty line for the poorest decile.33 or 34 percent of the population—receive the old-age No other scheme come close to this coverage and pension. An average beneficiary received MUR3,732 benefit level. per month in 2012, the equivalent of 15 percent of household per capita income in beneficiary households. For the poorest decile of beneficiary 33 Poverty line of MUR3821 in 2006 prices MAURITIUS | Inclusiveness of Growth and Shared Prosperity 65 CHAPTER 6 - Social Protection in Mauritius 134. The BRP lacks a focus on the poor because it with 34 percent nationwide. This group receives 7 disproportionately favors well-off households. The percent of all BRP old-age pension benefits because BRP old-age pension extends similar coverage and a smaller share of the elderly reside in poorer benefit levels to all elderly, whether well-off or poor. households—e.g., 7 percent in the elderly are in Since the BRP is a universal benefit, it would be the bottom decile. Furthermore, beneficiaries in inappropriate to view its transfers to the non-poor as the poorest decile received an average benefit that leakage, yet the fact remain that BRP lacks a poverty was 16 percent lower than that of the wealthiest focus and in fact allocates a greater portion of its half of the population—MUR3,372 versus MUR3,901. benefits to better-off households. The coverage of the bottom decile34 is only 24 percent, compared 34 This refers to the decile of post-transfer income per equivalent adult. Figure 51: Distribution of elderly population, BRP old-age pension beneficiaries and benefits across deciles of income per equivalent adult, 2012 14 12 10 8 6 4 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 Share of 60+ y.o. in population Old age benefits Old age beneficiaries Source: Authors’ calculations. Figure 52: Generosity of BRP old-age pension by decile of income per equivalent adult, 2012 50 45 44.1 40 35 33.9 Percent 30 27.1 25 24.0 20.6 20 14.9 15 12.7 10 5 0 D1 D2 D3 D4 D5 D6-10 Total Decile Source: Authors’ calculations. 66 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 6 - Social Protection in Mauritius Photo : © Cercle2Confusion MAURITIUS | Inclusiveness of Growth and Shared Prosperity 67 CHAPTER 6 - Social Protection in Mauritius 135. Its sheer size makes the old-age pension the leading could not reach more than 55 percent of the poor contributor to poverty reduction among SA programs, and 38 percent of the poorest decile even if it were but it is not efficient at reducing poverty. In the perfectly targeted. For eligible households, Social absence of the BRP, poverty would be 11.1 percent Aid paid MUR555 per capita in 201236 (or 11 percent rather than 6.9 percent. Yet as a consequence of of the poverty line), contributing a meaningful 22 universal targeting and dilution of benefits, each percent to the poorest decile’s household budgets. rupee spent on the old-age pension translates into only 0.28 rupees of poverty-gap reduction. By 138. Despite the limited coverage, Social Aid has a relatively contrast, each rupee spent in Social Aid, a program low inclusion error, meaning it is efficient in allocating specifically targeting the poor, reduces the poverty benefits to the poor. Each rupee spent in the program gap by 0.66 rupee. translates into a 0.66 rupee reduction in poverty gap, Social Aid has a progressive coverage (Figure 136. With the bulk of SA spending dedicated to the BRP, 53) and is well targeted by international standards, funding is low for programs specifically intended to with 62 percent of beneficiaries in the bottom benefit the poor, which limits SA’s overall impact on quintile of the pre-transfer income distribution37 and poverty reduction. Programs more closely linked 68 percent of benefits going to this group (Figure to poverty or its markers—e.g. social pensions 56). Conversely, this implies that 38 percent of that don’t fall under the BRP, Social Aid, disability Social Aid’s beneficiaries come from outside of the assistance, certain subsidies (on rice, textbooks), poorest quintile. school feeding, income support to temporarily unemployed, etc.—receive only less than a fifth of 139. Social Aid’s small size and use of categorical the SA budget. targeting offset its relative efficiency, restricting its impact in reducing poverty. If Social Aid were not 137. Social Aid—the only program in Mauritius that implemented, the poverty headcount would only specifically targets the poor based on a means test— rise to 7.3 percent, up from 6.9 percent. The low is small in terms of budget and coverage. It accounts overall impact is a function of small size and gaps for only 9 percent of SA budget, and it covers 3.8 in targeting. Along with the means test, Social Aid percent of population and 15 percent of the poor uses categorical eligibility criteria,38 which lead as direct and indirect beneficiaries.35 Social Aid’s small size is the main reason for its high exclusion 36 In 2012 prices. error. Given its low total coverage, the program 37 Fifty-three percent of Social Aid beneficiaries come from the bottom quintile of the post-transfer income distribution. 35 Coverage of pre-transfer poor is 19 percent. 38 Categorical targeting is used to extend benefits to the chronically Figure 53: Coverage of Social Aid beneficia- Figure 54: Distribution of Social Aid beneficia- ries and benefits across pre-transfer per AE ries and benefits across pre-transfer per AE income deciles, 2012 income deciles, 2012 20 60 18 17.1 16 50 14 40 12 10 30 Percent 8 6.8 20 6 3.8 4.5 4 3.5 10 2 1.5 1.0 0 0 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D1 D2 D3 D4 D5 D6-10 Total Deciles Beneficiaries Benefits Source: Authors’ calculations. 68 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 6 - Social Protection in Mauritius to the acceptance of beneficiaries who would not disincentives to work and build linkages between qualify based on the means test alone and exclude Social Aid and active labor-market programs to some of those who could conceivably be eligible for facilitate the graduation of households. Beneficiary benefits on the basis of the means test. As a result, profiles indicate that a larger share of Social Aid the categorical filters contribute to inclusion and beneficiaries of working age are out of labor force exclusion errors. The New Income Support Program or unemployed. They can likely be transitioned to (NISP), introduced in 2014, aims to fill the gaps left unemployment-based income-support programs— by Social Aid’s categorical targeting. The fact that Unemployment Hardship Relief, Temporary the NISP has been introduced as a separate program Unemployment Benefit, and other schemes that will further contribute to the complexity of the SA promote labor-force reintegration. system. 141. One constraint facing Social Aid beneficiaries is lower 140. While targeting errors make it difficult for the poor education levels than non-beneficiaries. Among to obtain Social Aid, challenges prevent the exit of Social Aid beneficiaries, 4.5 percent are without households that could potentially graduate from the schooling, compared with 1.8 percent for non- program. For instance, 66 percent of beneficiaries beneficiaries (Figure 57). Forty-eight percent of who join Social Aid on the basis of a temporary loss Social Aid beneficiaries have only primary education, of the ability to work stay in the program for more compared with 30 percent for non-beneficiaries. than 19 months. The same holds for 60 percent Vocational training or re-training programs as well of “abandoned women” participants. Improved as entrepreneurship programs can useful labor-force recertification of beneficiaries would be a key step in activation tools, especially if they include a focus improving the exit of those no longer eligible. At the on women, who are overrepresented among Social same time, steps should be taken to reduce possible Aid beneficiaries. sick and their caregivers, abandoned women and children, single mo- thers, and dependents of inmates in government institutions. Figure 55: Profile of Social Aid beneficiaries and non-beneficiaries, 2012 By sex (percentage By level of educational (direct and of women) By labor-force status (direct and indirect beneficiaries of working indirect beneficiaries of working age, 16-60 years old) 70 age, 16-60 years old) 60 58.6 70 60 55.4 61.6 50.8 60 50 48.1 50 43.0 50 49.6 40 40 43.8 40 29.9 30 33.2 30 30 20 20 20 13.0 10 10 4.5 4.4 10 5.2 6.6 1.8 0 0 0 Non-benef. Beneficiaries Non-benef. Beneficiaries Non-benef. Beneficiaries Without Primary Employed Unemployed Secondary Tertiary Out of LF Source: Authors’ calculations. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 69 CHAPTER 6 - Social Protection in Mauritius 142. Mauritius has operated a wide range labor-force Figure 56: Distribution of benefits (targeting accu- activation programs for some time, but they are limited racy) of widows and children, disability, in coverage, fragmented, lack robust linkages to Social and other social pensions, 2012 Aid and have little coordination between each other. Implemented by multiple agencies, these programs 35 32.6 32.0 include pre-service training, on-the-job training, job- search assistance, support to micro-entrepreneurs, 30 and life-skills training (see Appendix A). 25 143. The Widows and Children’s Pension, the Invalidity 19.1 Pension, and other social pensions39 are pro-poor in 20 terms of the welfare profile of their beneficiaries 14.3 15.0 and the flow of their benefits to the poorer deciles. 15 Nearly one-half of their beneficiaries come from 10.5 the bottom quintile and a quarter to a third come 10 from the bottom decile (depending on the program). The distribution of benefits is roughly comparable. 5 These schemes make a meaningful contribution to the income of recipient households, especially poor 0 ones. Among beneficiaries from the poorest decile, Invalid pension Widow and children pension Other pension the women and children’s program contributes 33 percent to household income, compared to 14 Total Poorest Decile percent for the nation as a whole (Figure 58). For the Source: Authors’ calculations. disability pension, it is 32 percent versus 15 percent for the nation. The other social pensions, the figures are 19 percent and 10 percent. 146. Among the three education-related programs, the means-tested textbook subsidy does best in targeting 144. Room for improvement in these programs’ targeting the poor. Based on post-transfer income, it delivers remains. Approximately 28 percent of Widows and 55 percent of its benefits to the poorest decile. Children’s Pension benefits, 19 percent of Invalidity However, the monthly benefit is small at MUR180 Pension benefits, and 32 percent of other social per capita.42 The subsidy is a book loan program that pension benefits flow to the wealthiest half of the supplies free textbooks to secondary school students population. Among the main reasons is reliance who live in needy households.43 Eligibility is based on categorical eligibility criteria and delivering on a means test, with automatic eligibility extended a large share of the social pensions (widows and to the beneficiaries of Social Aid and Unemployment children’s benefit, disability, and survivor benefits) Hardship Relief. as components of the BRP—i.e., to those households that are beneficiaries of the BRP old-age benefits.40 147. The scholarship grants 44 and exam subsidies provide more generous benefits and they are still pro-poor, 145. The Government operates several education-related but the leakage toward the non-poor is high. Both transfer schemes, but they have small coverage schemes target low-income students and involve a and—with the exception of textbook subsidies—are means test, with Social Aid-beneficiary households only moderately pro-poor in the distribution of their automatically qualifying for the examination-fee benefits. A majority of the transfers go to the non-poor. subsidy. The scholarship grants proved a per capita The most significant education-related programs are monthly benefit of MUR918, and the average for the the scholarship grants program, exam subsidies, and exam subsidies is MUR382. However, 41 percent of textbook subsidies. The scholarship grants program scholarship funds and 29 percent of exam subsidies provides funding for tertiary schooling. The shares go to the wealthiest half of the population. of benefits going to the general population and the poorest decile are: 0.2 percent and 0.4 percent for the scholarship grants, 2.2 percent and 4.5 percent for the exam subsidies, and 0.5 percent and 2 percent for the textbook subsidies.41 sal for primary students. The program provides a loaf of bread to more 39 These include Unemployment Hardship Relief, Food Aid, Fishermen’s than 100,000 primary students free of charge. Allowance, etc. 42 In 2012 prices. 40 The limitations of the HBS survey do not allow distinguishing fully 43 Textbooks are provided free of charge to all primary students, irres- between social pension benefits received as part of the BRP scheme and pective of income. those delivered outside the BRP. 44 The scholarship grants scheme covers the cost of tuition for tertiary 41 Mauritius also implements a school-feeding program that is univer- schooling. 70 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 6 - Social Protection in Mauritius Photo : © Haja Faniry Razafimahenina MAURITIUS | Inclusiveness of Growth and Shared Prosperity 71 CHAPTER 6 - Social Protection in Mauritius 148. The Government provides a number of other untargeted 150. The Government has taken meaningful steps toward subsidies, most of which tend to leak benefits toward greater coordination, but further improvements are the non-poor. Specifically, it subsidizes the price of needed. Establishment of the National Empowerment rice, flour, cooking gas, and transportation for the Fund (NEF) in 2009 was an important step toward elderly, disabled, and students. With the exception coordinating several programs: the Program for the of the rice subsidy, these subsidies disproportionately Eradication of Absolute Poverty (EAP), the Trust benefit the wealthier households (Government of Fund for the Social Integration of Vulnerable Groups Mauritius, 2010). (TFSIVG), the Decentralized Cooperation Program (DCP), and the National Committee Corporate 149. In addition to the weak poverty focus of individual Social Responsibility (CSR).46 To further improve the SA schemes, fragmentation and lack of coordination coordination of social safety nets, the Government across programs undermine the effectiveness of social introduced the Social Registry of Mauritius (SRM), safety nets. The implementation of cash-transfer a unified SP program database and targeting schemes alone involves five different government platform, in 2012. SRM is capable of coordinating entities, while seven agencies participate in and harmonizing social assistance. It currently links implementing in-kind transfers. 45 Many more Social Aid and the NEF databases; in the future, it entities are engaged in the delivery of programs in is expected to store beneficiary data and provide community development, labor-market activation, a proxy means test (PMT)-based platform for all and CSO support. Fragmented safety nets run by targeted SP programs in Mauritius. multiple agencies are difficult to coordinate, leading to the system-wide loss of efficiency due to overlaps and gaps. Some households may receive C. Social insurance: contributory pensions multiple transfers while others—no less deserving of assistance—may be missed by all or most of these 151. Demographic aging is the most significant long- schemes. For the country as a whole, 47 percent term risk facing Mauritius’ SP system. The country’s of the absolute poor (post-transfer) remained not population is aging. The fertility rate is well below covered by any of the main social assistance schemes replacement level, and life expectancy is expected in 2012; under-coverage of the poorest quintile is 48 to rise to 78 year for men and 81 years for women percent. At the same time, households in the Social by 2050.47 Old-age dependency—expected to rise Aid program include beneficiaries of various other from its current level of 18 percent to 55 percent schemes, as illustrated in Figure 59. by 2050—will increase the burden on the working- Figure 57: Share of Social Aid beneficiaries who also receive benefits from other Other subsidy 1.4 Subsidy on textbook 2.7 Subsidy on examination fees 3.6 Scholarship grant 0.2 Other social pension 0.8 Invalid pension 31.8 Widow an children pension 12.7 Old age pension 35.4 Retirement pensions from former employer 14.2 NPF 1.7 0 5 10 15 20 25 30 35 40 Source: Authors’ calculations. 45 Government agencies implementing cash transfers: MOSS, MOAI- FPS, MoWRCDFW, PMO, municipal governments. Government entities involved in in-kind transfers: MOECHR, MOFEE, MOHQL, MOPILTS, 46 See http://nef.mu/historique.php. MOHL, MOSS, MOAIFPS 47 UN Population Projections, 2010. 72 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 6 - Social Protection in Mauritius age adults of supporting the elderly. Currently five 153. In this context, the admittedly modest expansion working-age adults support each person over 65 of SI coverage among the poorest segments of the years of age; in 35 years, fewer than 2 persons will population constitutes an important development—a be supporting each elderly individual. potential move toward a more sustainable SP system. Contributory schemes are not designed to address the 152. The structure of old-age protection, with the non- issue of poverty; however, SI is one pillar of SP, and contributory social pension as the main means of weak SI systems are likely to put additional burdens protecting the elderly, is vulnerable to the fiscal on other programs. Over the years, contributory pressures associated with an aging population. pension schemes have improved their coverage, The pressures increase in proportion to the rising including their coverage of the poorer segments dependency ratio. In Mauritius, the risks associated of society. In the poorest quintile, the coverage of with aging are addressed through a multi-pillar old-age contributory pensions increased from 7 percent in pension structure. The universal, non-contributory 2007 to 12 percent in 2012, but benefit value for BRP is the primary pillar, and contributory retirement this quintile remained nearly constant in real terms. schemes constitute a secondary pillar intended to complement the BRP. As the old-age dependency 154. The contributory pension schemes’ coverage remains ratio expands, the already high BRP budget would low, especially among the poorer population. These have to expand proportionately until it exceeds pensions benefit 53 percent of those in retirement fiscally sustainable levels or forces a reduction in age and deliver benefits to 19 percent of population. the BRP benefit level. Either way, the protection of Among the poorest quintile, contributory retirement the poor will be at risk. The social old-age pension schemes cover only 32 percent of the elderly and might decline or expanding BRP scheme might draw 12 percent of population. The coverage of the funds away from other SA programs. Within this NPF alone—excluding former civil servants and system, well-functioning contributory retirement beneficiaries of optional occupational schemes paid schemes are key to containing the Government’s by relatively large employers in the formal private cost of protecting the growing number of elderly and sector—is less than 5 percent of the total population allowing the Government to mobilize fiscal resources (13 percent of all elderly) and less than 4 percent for other tasks, such as poverty reduction. of the bottom quintile (12 percent of the elderly Figure 58: Coverage of contributory pensions by Figure 59: Mean benefit amount of contributory post-transfer income decile, 2007 and 2012 pensions by post-transfer income decile, Rs. in constant 2006 prices. 30 6000 25 5000 20 4000 15 3000 10 2000 5 1000 0 0 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 y2007 y2012 y2007 y2012 Source: Authors’ calculations. Source: Authors’ calculations. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 73 CHAPTER 6 - Social Protection in Mauritius in the poorest quintile). Employer pensions that the social safety nets. Proxy means tested social include civil servants and government employees assistance programs should be developed to reduce at all levels extend coverage to 43 percent of all inequality and improve SP efficiency. elderly and 20 percent of the elderly in the poorest quintile, reaching 15 percent of population and 8 157. Mauritius has operated a wide range labor force percent of the bottom quintile. activation programs for some time, but these programs are small in coverage, fragmented, lack robust linkages to Social Aid, and poorly coordinated. D. Areas of focused attention - Social The Government has undertaken meaningful protection steps toward greater coordination, but further improvements in SA coordination are needed. An aging population is the most significant risk facing 155. Mauritian social protection (SP) programs played Mauritius’ SP system in the long-term. a significant role in ensuring that the poor benefit from economic growth through redistribution of 158. Despite being the leading contributor to poverty resources. Convergence, average transfers, and reduction, the Social Aid program could be scaled generosity of the social protection system all up and significantly improved. The share of SP significantly increased between 2007 and 2012. programs intended to specifically benefit the poor Close to 75 percent of poverty reduction has been is low in Mauritius, which limits the overall impact attributed to SP expansion. Without the SP system, on poverty reduction. Social Aid is the only program poverty would be almost three times higher and that targets the poor based on a means test—and it is inequality would be 12 percent higher. While the small in terms of budget and coverage. Social Aid is majority of the reduction in poverty is attributable efficient in allocating benefits to the poor due to its to social assistance (SA) schemes, social insurance relatively low inclusion error, but its small size and (SI) programs also made minor contributions to use of categorical targeting restrict the magnitude poverty reduction. of its poverty impact. Social Aid could deliver better results if the program were scaled up and its relative 156. However, this study has found a weak poverty focus efficiency were further improved. The Government in individual SA programs in Mauritius, and the provides a number of other untargeted subsidies, fragmentation associated with lack of coordination most of which tend to leak benefits toward the non- across programs undermines the effectiveness of poor. 74 MAURITIUS | Inclusiveness of Growth and Shared Prosperity Chapter 7 Labor Market Characteristics and Challenges I. Mauritius Labor Market characteristics and recent trends Major transformations in the labor market led to shifts in employment patterns and increases in wage disparities. The tightening of the Mauritian labor market was associated with growing skills mismatches and a lowering demand for traditional employment. The Mauritian labor market has undergone a shift from such labor-intensive industries as textiles to more knowledge-intensive ones, including finance and tourism. We can already see the increasing, but somehow still marginal, importance of these industries and occupations for the overall economy. From 2007 to 2012, Mauritius’ real wages surged by more than 8 percent. Workers in the services sector receive the highest salaries. Tourism and the tertiary sector highlight the ongoing upward trend, with pay around 40 percent higher than agriculture in 2012. STEM and high-hech occupations pay considerably higher salaries. CHAPTER 7 - Labor Market Characteristics and Challenges 160. In this section, we will show and analyze how A. Introduction some commonly used labor-market indicators have evolved throughout a period that straddles the 159. As discussed by McDonald and Yao (2003) for the economic reforms of the 2000s. The indicators are 1991–2002 period, steady economic growth does divided in three categories. The first includes the not necessarily combine with a contemporaneous most common measures of labor-market health: improvement of labor market conditions in the unemployment rate, employment rate, and activity Mauritian economy. Quite the opposite, the steady rates. We will refer to these as the main indicators. economic growth since the beginning of the The second category will focus on the segmentation 1990s has translated in a sluggish labor market of the Mauritian labor-market structure along that saw unemployment constantly increasing, three indicative dimensions: high-tech vs. low- reaching a peak in 2005. The reasons for this tech, agriculture vs. non-agriculture and public vs. underwhelming performance have been identified private. We will refer to these as the labor-structure as: (a) tight regulations on the wage bargaining indicators. The last set of indicators will consider process, characterized by highly centralized wage mean and median total and hourly wages paid in the determination; (b) excessive bureaucratization of market—i.e., the wage indicators. the government departments responsible for this bargaining, leading to the de facto impossibility of employers relocating their employees; and (c) B. Labor market outcomes the extremely high costs companies face for job termination (Porter, 2004). All of these factors 161. The main indicators that will be discussed in this were causing a detachment of labor productivity section are: employment rate, unemployment rate, from remuneration. Having realized that these and inactivity rate.48 Three measures are widely bottlenecks were plausibly responsible for economic disseminated, discussed, and analyzed by policy growth and job creation below potential, the makers and media, but they suffer from well-known Mauritian government reacted in the mid-2000 with shortcomings. Keeping these caveats in mind, these a series of labor-market reforms. measurements can serve as a useful preliminary check on the performance of a labor market, and they will be analyzed here. Figure 60: Labor market: main indicators Employment, Inactivity and Hours of work per week unemployment rates 42 60 11 41 50 10 Employment rate, incactivity rate 41 Unemployment rate 9 40 40 Hours of work 8 30 40 7 39 20 6 39 10 5 38 0 4 38 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Employment rate Hours of Work Inavtivity rate Unemployment rate Linear (Hours of Work) 48 The exact definitions given for these three rates somehow vary Source: World Bank staff calculations, based on CMPHS data. across labor statistics institutions. The one that we have adopted here defines every individual at least 15 years old as economically active. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 77 CHAPTER 7 - Labor Market Characteristics and Challenges 162. The analysis of the main labor-market indicators inactivity rate49 is not necessarily worrisome. The suggests changes in structure, falling demand, rate is a composite, pooling together individuals in and general tightening. Mauritius’ workforce has very different circumstances. It is therefore very been shrinking, both in terms of labor-market important to understand the composition of the participation and hours supplied. Beside the blip economically inactive. in 2003, the employment rate was fairly constant until 2009 (Figure 592). A light downward trend was detectable from 2004 to 2007, and the trend went C. Tightening of the Mauritian labor market and just as gently in the opposite direction from 2007 and 2009. Since 2009, the employment rate has sectoral changes decreased more decisively, reaching its minimum at the end of 2012. The employment decrease has 164. The Mauritian labor market’s tightening is associated not been matched by simultaneous movements of with skills mismatches and diminishing demand for the unemployment rate, which peaked in 2005. It traditional employment. Since 2006, the vacancies constantly and markedly diminished in the next statistics have gone down steadily, while the number three years, reaching a plateau at just below 8 of unemployed and the number of permits for percent. On average, each worker worked almost foreign workers rose. The continuous change in labor three hours a week less in 2012 than in 2001. demand reflects the construction and real estate boom on one end and the reduction in demand for 163. Inactivity rate started picking up in 2009, and apparel on the other. it has increased ever since. The movement out of employment after 2009 has not been into unemployment; rather, it has been out of the 49 The ILO defines an individual as economically inactive if he is within active labor force and into inactivity. An increasing the working age but does not participate in the labor market. The pos- sible reasons for inactivity are: caring for family, retirement, sickness or disability, school enrolment, discouragement, or no intention to work. Figure 61: Tightening labor market in Mauritius Vacancies Work Permits and Vacancies composition 55000 Unemployed 50000 Other 45000 Textile 40000 Financial intermediation 35000 Hotels & restaurants 30000 Information technology 25000 Wholesale & retail trade 20000 Construction 15000 Real estate 10000 Wearing Apparel 2004 2005 2006 2007 2008 2009 2010 2011 2012 Unemployed (official) Vacancies Work permits 2005 2013 Source: World Bank staff calculations, based on official data from Mauritius Ministry of Labor. 78 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges Photo : © tedesco57 MAURITIUS | Inclusiveness of Growth and Shared Prosperity 79 CHAPTER 7 - Labor Market Characteristics and Challenges Box 8: Surge of foreign workers in Mauritius Mauritius has experienced a substantial increase in foreign workers—from 3.5 percent of the labor force in 2003 to 4.5 percent of labor force in 2013 (Figure 62). According to Statistics Mauritius, the number of foreign workers has increased from around 17,000 in 2003 to more than 26,300 in 2013. The number of foreign workers has increased by 54 percent during this period, while the labor force grew only 11.3 percent. In addition, the number of work permits has grown significantly in recent years Figure 62: Role of the foreign workers in Mauritius Share of foreign workers in labor force Characteristics of foreign workers 5.0 4.5 4.0 96.6 3.5 81.4 3.0 70.3 2.5 62.7 2.0 1.5 1.0 0.5 0.0 Share of males Share of Share of working Share of working employed in in in textile Source: Digest of Labor Statistics, 2002-13 large enterprises Manufacturing Most of the foreign workers take jobs in lower-skilled occupations, but some are high-skilled. The most frequent professions are machine operators (more than 60 percent), sewers, fish cutters, spinners, thread and yarn Masons, general site managers, production technicians, concrete plant machine operators, bakery product workers, bartenders, stonemasons, construction plasterers, general carpenters, general public relations officers, cooks, and hotel/restaurant employees. Forty percent of the foreign workers come from Bangladesh, with a large increase since 2009. The number of migrant workers has also increased from India, Madagascar, Nepal, Morocco, Lebanon, Korea, Japan, Israel, Indonesia, Haiti, Algeria, Argentina, Burundi, Congo, and Denmark. The share of foreign workers who are males has been steadily growing, reaching close to 63 percent. In 2013, 81.4 percent of the foreign workers were employed in the manufacturing sector, with 70.3 percent in textiles. Almost all (96.6 percent in 2013) work in the relatively large enterprises. Migrant workers are often filling vacancies for unattractive jobs that no longer appeal to Mauritian job seekers. The increase in demand for foreign workers is associated with skills mismatches and deteriorating wages in the primary sectors, including textiles. Companies turn to foreign workers to alleviate the shortage of skilled labor in the manufacturing and construction sectors. 80 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges 165. The Mauritian labor market has undergone a shift 166. Data confirm the migration from labor-intensive to from such labor-intensive industries as textiles to knowledge-intensive activities. Between 2001 and 2012, more knowledge-intensive ones, including finance employment decreased 22 percent in the primary and tourism. The migration from labor-intensive to sector and 9.3 percent in the secondary sector; in knowledge-intensive has been noted by McDonald the tertiary sector, meanwhile, it increased roughly and Yao (2003) and David and Petri (2013). It the 43 percent. The sectoral composition of employment declining share of employment in the manufacturing also shifted toward tertiary sector. In 2012, close to and agriculture sectors and the increased share of 65 percent of employees worked in the tertiary sector, labor employed in the tertiary sector and, to a lesser compared with 53 percent in 2002. The primary and extent, in tourism. secondary sectors decreased proportionally (Figure 63). Figure 63: Sectoral composition of employment shifts toward tertiary sector Employment by sector, 2012/2002, % -22.0% Primary -9.3% Secondary -25.8% - Manufacturing 25.3% -Construction 42.8% Tertiary 40.2% - Wholesale trade 62.2% 35.6% -Transportation 27.0% -Public Administration 21.1% - Education, Human… 16.7% Total Employment, Sectoral composition 100% 90% 80% 52.7% 70% 64.5% 60% 50% 40% 30% 35.8% 20% 27.8% 10% 11.5% 7.7% 0% 2002 Axis Title 2012 Primary Secondary Tertiary Source: World Bank staff calculations, based CMS data. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 81 CHAPTER 7 - Labor Market Characteristics and Challenges 167. The deterioration in the primary and secondary sectors is associated with the decline in the D. The role of the high-tech sector agriculture and manufacturing, while booming tertiary sector has been led by accommodations Box 9: Role of high-tech sector in Mauritius and wholesale trade. The secondary sector would In the past decade, the Mauritian labor market experienced a have contracted even more if not for growth in substantial employment shift from manufacturing to the service construction. Employment in accommodations grew the most, rising by more than 62 percent between sector. In a first approximation, this transition might be thought 2002 and 2012. Wholesale trade grew 40.2 percent of as a movement from labor-intensive to knowledge-intensive and transport 35.6 percent. In addition, employment industries. This is not necessarily the case because both sectors in public administration grew by 27 percent. Over are a mix of high and low knowledge-intensity industries. If all, the share of the Tertiary sector increased from we are interested in which type of inputs are required in the 52.7 percent in 2002 to 64.5 percent in 2012, while economy’s most dynamic segments, a better distinguishing shares shrunk for the primary and secondary sectors. concept would be the high-tech sector. It transcends the usual distinction between secondary and tertiary sectors and 168. In 2013, around 80 percent of the Mauritian workforce includes both high-technology manufacturing activities and was employed by privately owned enterprises (Figure knowledge-intensive services.51 64).50 This number is constant throughout the period. It is common to think of these highly specialized sectors as the fundamental engine of growth in modern economies because Figure 64: Public vs. private employment, by shares they generate the high value added products and services that 0.9 command higher salaries. For this reason, understanding the size and development trajectories of these sections is fundamental 0.8 to getting a sense of whether growth is sustainable in the near 0.7 future and whether the sectoral dynamics are in fact hiding a movement along the technological curve. 0.6 0.5 High Tech employment could be defined in two ways in Mauritius. One method is based on sector of employment 0.4 and another one is on occupational attainment. Appendix 0.3 C describes in detail the definition of high tech based on the NSIC4 industry classification. 0.2 0.1 The NSIC-based definition of high-tech industry combines both workers in high-tech occupations and workers within the 0 high-tech sector who fill low-tech positions. For a complete Pablic sector Public enterprises Private Sector picture, it is useful to analyze a different aspect of high-tech 2003 2012 employment—that of high-tech occupation. Source: World Bank staff elaboration on CMPHS data. 170. Analysis of both dimensions of high-tech industries 169. In the past 13 years, the Mauritian labor market point toward an increasing, but somehow still marginal, structure has evolved toward increased importance importance of these industries and occupations for of the services sector. This sector mainly comprises Mauritius’ overall economy. Figure 65 shows the professions requiring medium to high skill levels. On cumulative percentage change in high-tech and STEM the other hand, occupations traditionally reserved to (science, technology, engineering and math) jobs as low-skilled individuals have either maintained their a share of total employment.52 The base years for skill levels (agriculture), or recently increased them calculating gains differ for the two series. The figure (manufacturing). Quite worrisome, the share of clearly shows that high-tech has been an important individuals occupied in the highly skilled intellectual source of employment growth, starting in 2006. The sector has recently shown a diminishing trend. most recent data show general employment was at 51 See the appendix for a detailed description of the high-tech sector definition adopted in this study. 52 The high-tech employment series starts in 2004 because we do not have access to detailed four-digit NCIS coding for 2001 to 2003. For this reason, we were unable to reconcile the coding for these early years with the classification applied in the later ones. The year 2010 is mis- 50 No information is available prior to 2003. sing for the same reason. 82 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges nearly the same level it was 11years earlier, but 171. We also show trends in high-tech occupation high-tech employment grew a staggering 70 percent employment. We use the concept of STEM in eight years. As a result of this growth, high tech’s occupations, which is receiving increasing attention share of total employment increased 3 percentage in the scientific literature53 and among policymakers. points in eight years; in 2012, it constituted around 7 To define STEM occupations, we have followed US percent of the economy. Another factor worth noting Bureau of Labor Statistics guidelines.54 We find that is that this sector has displayed a very consistent after an initial but feeble downward trend, these growth with the exception of a slight ebb between high-knowledge industries have expanded their 2008 and 2009—at the height of world economic share of the workforce by almost 30 percent since recession. Without the phenomenal performance of 2001. The downside is that STEM workers’ share to the advanced sector, Mauritian employment level total employment is still minor—only 4.6 percent in would have trended downward over this period, 2012, compared to 3.6 percent in 2001. instead of being substantially stagnant. 53 See, for example, Goos, Hataway, Konings and Vandemeyer (2013). 54 Bureau of Labor Statistics (2012), “Options for defining STEM (Scien- ce, Technology, Engineering, and Mathematics) occupations under the 2010 Standard Occupational Classification (SOC) system.” The two-di- gits NASCO occupations defined as STEM are: 21, 22, 31 and 32. Figure 65: High-tech vs. overall employment and wage changes Cumulative employment growth and share if High-Tech 100 8.0 % growth in employment Share of high-Tech in total 7.0 in total 80 employment (right scale) 6.0 60 Share of high Tech - 5.0 40 4.0 3.0 20 2.0 0 1.0 04 05 06 07 08 09 10 11 12 -20 0.0 20 20 20 20 20 20 20 20 20 Total Employment High-Tech STEM Employment Share of High-Tech Real wages, STEM and total 23,000 21,000 19,000 17,000 15,000 Real wages 13,000 11,000 9,000 7,000 5,000 3,000 01 02 03 04 05 06 07 08 09 10 11 12 20 20 20 20 20 20 20 20 20 20 20 20 Total Wages STEM Wages Source: World Bank staff elaboration on CMPHS data. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 83 CHAPTER 7 - Labor Market Characteristics and Challenges 173. Workers in the services sector earn the highest salaries, E. Wages and earnings and the constant influx of workers in this sector has not slowed the upward trend shown in Figure 63d. This 172. Since 2001, real mean wages in Mauritius have sector’s contemporaneous increase in employment surged by more than 8 percent. The evolution of this share and wages indicates an unmet expansionary increase can be divided in four periods (Figure 63a). potential that could absorb even more workers if the A strong hike characterized 2001-05, followed by a right type of human capital were available. In the substantial retrenchment of more than 10 percent same figure, we can see that salaries in manufacturing in 2005-07. After 2007, real wages started galloping and tourism are stagnating, while the primary sector, upward again, more than making up for the previous already the lowest one in 2001, saw a decrease in slump. However, this growth has halted in the past its wage. These different dynamics obviously have two years. After all the ups and downs, median many possible explanations. In tourism, stagnating wages have increased by 8 percent in these 11 years. wages coupled with a stagnating employment share (Figure 63c) might be symptomatic of a mature sector with little expansionary potential in the immediate future. On the other hand, stable wages and shrinking employment levels in the secondary sector might indicate an ongoing transition of labor force from this sector to services. Figure 66: Sectoral composition of employment changes towards tertiary sector a. Average and median wages (2005 rupees) b. Sectorial wage differential controlling for demographical background .5 11,000 % Wage Gap - Agricolture 10,000 .4 9,000 Real wages 8,000 7,000 .3 6,000 5,000 .2 4,000 3,000 .1 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Median Wages Average Wages Manufacturing Services Turism c. Share of labor force by sector d. Wages by sector (2005 rupees) 60 12,000 11,000 50 10,000 9,000 Share in total labor force 40 Real wages 8,000 30 7,000 20 6,000 5,000 10 4,000 0 3,000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Agriculture Maufacturing Agriculture Maufacturing Services Tourism Services Tourism Source: World Bank staff elaboration on CMPHS data. 84 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges 174. The tourism and tertiary sectors paid around 40 percent Skills mismatches are associated with rising levels more than agriculture in 2012, while manufacturing of unemployment and weak job creation in Mauritius. salaries were 30 percent more. Figure 63b clarifies Lack of inter-generational mobility has potentially that the differences in payoffs between the very adverse effects for the overall economy’s growth primary sector and the rest of the economy are not potential. Mauritian women experience substantially immediately linked to “quality” differences in human lower employment levels and higher unemployment capital employed in various sectors. In this figure, we and inactivity levels than their male counterparts. plotted the behavior of wage premiums from 2001 to The gender wage gap is severe and shows no sign of 2012 in the secondary, tertiary, and tourism sectors, decreasing. In fact, it widened in recent years. If any compared to agriculture. The wage premiums are difference exists, in fact, suggests more, not less, obtained from a regression55 in which the log of human-capital accumulated by Mauritian women. Young wages is regressed on sectorial dummies and various workers between ages 15 and 24 are disadvantaged in individual-level demographic data. The results can terms of unemployment and particularly vulnerable to be interpreted as the percentage wage difference labor-market fluctuations. between the plotted sector and agriculture, keeping constant education, age, potential work experience, gender, and geographical location. Compared to E. Increasing inequality following agriculture, wages were higher in 2012 by around 40 percent in the tourism and tertiary sectors and the deteriorating of low-skilled wages 30 percent in manufacturing. All three differentials have been expanding. 175. Demand shifts centered on skills are a likely source of changes in the Mauritian wage structure. These STEM and high-tech occupations pay considerably changes are rewarding highly skilled individuals, and higher salaries (Figure 67). Mean wages in STEM their wages are rising. The source of this different occupations have consistently doubled other dynamics at the two extremes of the schooling occupations’ payoffs. This reflects both the scarcity distribution is to be attributed to advancements in of highly skilled workers and the high value that mechanization and information technology which these types of workers add to production. are complements of highly skilled individuals and substituting lower skilled ones. II. Challenges of the Labor Market in Mauritius 176. As discussed in the previous chapter, less-educated individuals suffered a substantial decrease in their The polarization of employment and wages is associated mean wages. As shown in Figure 67, individuals with with rapid changes in demand for skilled labor. A rigid post-secondary education are the only group that system of determining pay increases and complex labor saw an improvement in wages. At the other extreme, regulations reduces competitiveness and limits the the average wage of those with no formal education ability of the economy to undergo structural changes. plummeted by more than 30 percent. 55 Full regression is displayed in the appendix. Figure 67. Change in average real wages 2001-12 No Edu. -.301 Primary Edu. -.065 Secondary Edu. -.027 Post-sec. Edu. .151 Private Empl. .071 Public Empl. .235 Females .157 Males .119 -.35 -.3 -.25 -.2 -.15 -.1 -.05 0 .05 .1 .15 .2 .25 .3 Source: World Bank staff elaboration on CMPHS data. % Change MAURITIUS | Inclusiveness of Growth and Shared Prosperity 85 CHAPTER 7 - Labor Market Characteristics and Challenges 177. Main labor-market outcomes are worse among improvement among the better off and deterioration the poor, and their situation has deteriorated. The or no change among the poor. In sum, the poorer charts in Figure 70 cover the period of 2007 and population clearly experienced less favorable 2012, showing that the lowest quintile had higher changes in their labor incomes. unemployment rates, deteriorating real wages, and a larger share of low-wage earners. In general, 178. For the Mauritian economy, a puzzling sign is a unemployment rates declined among all groups, disproportionate increase in real wages in the but the reduction was largest among the richest public sector. The deterioration of low-skilled group. Wages fell by 4.2 percent among the lowest wages has been accompanied an unbalanced wage quintile, but they increased 11.4 percent for the determination in the public sector. The public sector top quintile. The proportion of low-wage earners had a 23.5 percent increase in wages, compared increased among the lowest quintile and fell among with only 7 percent in the private sector, where the richest, indicating a deterioration in the relative tradable goods are produced. Female workers’ 15.7 position of the poor. The shares of low earners due to percent increase outperformed the 11.9 percent short hours or voluntary low earners have increased increase of their male counterparts. among the rich, another sign of the disproportionate Figure 68. Labor characteristics by consumption quintiles 2007-12 18.0 Unemployment rates Wage Earnings (2012/2007), % 16.0 14 14.0 12 11.5 11.4 12.0 2007 10 10.0 8 7.2 2012 8.0 6 4 6.0 2 0.3 4.0 0 2.0 -2 0.0 -4 -6 - 4.2 Lowest 2 3 4 Highest quintile quintile Lowest quintile 2 3 4 Highest quintile 25.0 Low earners rates Share of Low Earners Who Have Low 100.0 Earnings due to Short Hours 90.0 20.0 80.0 70.0 2007 15.0 2007 60.0 2012 2012 50.0 10.0 40.0 30.0 20.0 5.0 10.0 0.0 0.0 Lowest 2 3 4 Highest Lowest 2 3 4 Highest quintile quintile quintile quintile Source: World Bank staff estimates, based on HBS data. 86 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges Figure 69: Returns to educational investment difference from no education 1.60 1.40 1.20 Returns on eeucation (wage differences) 1.00 0.80 0.60 0.40 0.20 0.00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2012 Primary Secondary Post-secondary Source: Author’s calculations. Coefficient from Mincer regressions, no education is the base. The evidence clearly points toward education’s sizable Evidence from developed countries (Autor et al., 2006 and economic advantages in terms of higher payoffs. In Figure 2008; Acemoglu and Autor, 2011) indicates a secular raising 69 we report the coefficients from a wage regression56 for trend in the demand for skills due to their complementarity primary, secondary and post-secondary graduates. These with new technologies introduced in the workplace. This coefficients report the percentage difference in wages for now popular theory is usually referred to as skill-biased the three categories with respect to the least educated technical change (SBTC), and it attributes increasing wage individuals over an 11-year period. They are independent of inequality to technological progress and its differential gender, geographic location, age, and sector of activity. The impact on demand for certain occupations associated with advantages of education are clearly increasing. In fact, the certain skills, according to their technological content premium for primary education hovers around 20 percent, and their complementarity or substitutability with new and the post-secondary premium varies between 100 percent technologies. and 140 percent. Wage premiums for secondary and primary education are constant throughout the period, while those 179. Wages grew significantly more among the relatively for post-secondary degrees are increasing, especially in better off, widening wage disparities. We begin by recent years. The data point toward the growing importance laying out some key facts about Mauritius’ wage of human capital in determining wages and suggests a structure. In Figure 70, we plotted the change in corresponding upturn in the relative demand for skills. log real monthly wages by percentile for Mauritian workers between 2001 and 2012. It reveals two 56 Full regression results are reported in the appendix. interesting results. First, an almost linear spreading Figure 70: Change in log real monthly wage by Figure 71: Smoothed changes in employment by percentile, 2001 vs. 2012 occupation. 2001-09 .4 .6 100 x Change in Employment Share .2 .4 Log Earning Change 0 .2 -.2 0 -.4 -.2 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 Occupation's Decile in 2001 Education Distribution Monthly Wage Percentile kernel = epanechnikov, degree = 0, bandwidth = .71 Source: Authors’ calculations. Source: Authors’ calculations. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 87 CHAPTER 7 - Labor Market Characteristics and Challenges out of the entire wage distribution took place—the be attributed to labor market institutions—strong higher the earnings in 2001, the steeper the salary unions, centralized bargaining, and minimum wage increase over the next 11 years. Second, the tails of legislation—that are serving as effective tools for the wage distribution show diverging evolutions: the protecting the most vulnerable workers from falling top half has seen an increase in real wages, while too far behind. the bottom half has seen real wages deteriorate. The discontinuity occurs almost exactly at the 50th percentile. F. Rigidity of labor regulations 180. A polarization of employment has been accompanied A rigid system of determining pay increases and by a rapid increase in employment share among top- complex labor regulations reduce competitiveness skill occupations. In Figure 71, we plot the trends and limit the ability of the economy to undergo in skill content of Mauritian jobs for 2001-09. As a structural changes. proxy for skill content, we first calculate the each occupation’s mean years of education,57 and then 183. Despite changes in labor regulations, wage we then sort these occupations into deciles of the determination in Mauritius depends heavily on the average years of education. Finally, we plot the non-market forces and collective agreements. Before change in employment share of each occupation 2005, complex labor regulations tended to limit the between 2001 and 2009 against the skill content of labor market, and wage-setting was due to non- those same occupations. The figure clearly shows a market bargaining power, hampering the ability polarization of employment, with modest growth at of the economy to create new jobs. Under the old the very bottom of the skill distribution, declining system, wage increases were linked to the CPI, and employment in the middle, and a nearly constant, real wages grew much faster than labor productivity monotonic, and rapid increase in employment share in 2000–06. Since 2006, the government has adopted in the top-half. The pattern of job growth fits well several structural changes that gradually relax with the previously discussed evidence on increasing labor regulations. However, many regulations still wage inequality between top and median wages and affect Mauritian wage determination and working increasing post-secondary wage premiums. conditions. 181. The facts emerging from our analysis of Mauritius’ 184. There are two complementary minimum-wage wage dynamics over the past decade seem to fit the support systems in Mauritius: (i) the annual Salary SBTC framework. Demand shifts driven by skills Compensation and the (ii) Remuneration Order are a likely source of changes in the Mauritian system: wage structure. These changes are rewarding i. The annual Salary Compensation is designed as highly skilled individuals with rising wages. The a cost-of-living adjustment mechanism. Each source of the different dynamics at the schooling year, the Government issues a decree stipulating distribution’s extremes is advances in mechanization increases in minimum wages. It sets a series of and information technology that complement highly thresholds based on the level of earnings, and skilled individuals. They substitute for lower-skilled increases are generally higher for those earning workers. This ongoing tendency has not been met lower wages. by a sufficient and contextual expansion of the ii. The Remuneration Order system provides pay supply of highly qualified individuals—as indicated increases for workers in certain types industries by the number of post-secondary graduates in Figure and occupations under the supervision of the 73. Even if overall inequality in Mauritius does not National Remuneration Board (NRB), a part of assume dramatic proportions when measured by Gini the Ministry of Labor, Industrial Relations, and coefficients and the other statistics, a trend toward Employment. The NRB defines minimum wages and growing inequality at the top end of wage and skill other working conditions in the private sector. It distribution seems to be emerging, especially in issues the 30 Remuneration Orders and regulations recent years. currently in force and applicable to different occupational categories in specific economic 182. At the other end of income and skills distribution, activities. The orders apply to around 50 percent inequality between low income earners and the median of the workforce—excluding the civil service, earner has been kept in check. This can probably to which is governed by separate provisions. 57 We would have liked to extend our analysis to 2012; unfortunately, the CMPHS classification of occupations adopts two different standards 185. The labor market needs to reward higher productivity. The Salary Compensation and Remuneration Orders for the years before (NASCO) and after 2010 (NASCO-08). With the two- digit occupation classification we were provided, it is impossible to re- are designed to reduce disparities, but they hardly concile the two series. impact wage determination in the intended way. 88 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges The thresholds are set at very low levels by the 187. Skills mismatches can be measured by an index recently international standards—on average, 22 percent developed by the ILO. It focuses on the differences of the wage. In addition, the national tripartite in educational attainment of the employed in negotiations set up in 2010 make it more difficult comparison with the unemployed.58 According to to maintain competitiveness. In the longer term, an ILO publications, the index can be interpreted as a appropriate balance between worker protection and summary measure of the relative position of labor- labor market flexibility has to be found in Mauritius. market groups with different levels of education. If primary, secondary, and tertiary graduates all have the same unemployment rate, the index will G. Rising skills mismatches in Mauritius have a value of zero; the index would reach a value of 1 (complete dissimilarity) if, for example, all A sharp increase in skills mismatches characterizes the those with primary and tertiary education were labor market in Mauritius. The mismatches are growing employed and all those with secondary education with increasing demand for skilled labor. The mismatches were unemployed. are associated with rising levels of unemployment, a sign of weak job creation in Mauritius. 188. Between 2001 and 2010, skills mismatches in Mauritius increased by almost 30 percent. Figure 72 186. In Mauritius, the contraction of traditional sectors and shows that the skills mismatch index grew from 0.09 higher unemployment resulted in workers looking for to 0.13 points. In recent years, the increase was jobs in other sectors and occupations. Some workers much stronger. Rising mismatches in the late 2000s who lost their jobs were forced to seek employment were associated with rising unemployment. The in sectors more advanced in terms of educational trends underline the need for policies that ensure attainment. As a result, the employed share of the best possible matches in the labor market to those with tertiary educations more than doubled curb the negative trend of rising unemployment. in Mauritius. The supply of highly educated workers The skills mismatch index’s level corresponds to that has not met demand, creating mismatches in the of many developed countries; it is higher the level labor market. The mismatches put upward pressure of the developing economies. According to the ILO on unemployment rates. In Mauritius, the issue of skills mismatches has received renewed attention 58 It should be emphasized that this index captures mismatches between the employed and the unemployed in terms of level of educa- in the recent years. tion. It does not capture mismatch at more detailed levels of skills. The index also does not capture mismatches between job requirements and labor supply. Figure 72: Measuring skills mismatches in Mauritius, skills mismatch index Skills mismatch index, Mauritius Contribution to skills mismatch index 14.0 100% 90% 12.6% 15.6% 12.0 Contribution to skills missmatch 80% -match index 10.0 70% 60% 50.0% 50.0% 8.0 50% 6.0 40% Skills miss 4.0 30% 2.0 20% 37.4% 34.4% 10% 0.0 0% 2001-2006 2007-2010 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Primary and below Secondary Tertiary Source: Authors’ calculations. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 89 CHAPTER 7 - Labor Market Characteristics and Challenges report (2011), the index is less than 10 for most of and less-educated parents will not be able to offer these the developing countries. opportunities to their children, reproducing the same social structure over time. 189. Growing demand for highly skilled workers leads to increases in the skills mismatch index. The right- 191. Recognizing the central role played by education hand chart of the Figure 72 presents contribution both in terms of economic development and political of the index’s three main education levels. The main and social inclusion, the Mauritian Constitution increase is associated with tertiary education. As of 1968 clearly mandates the public provision of tertiary education increased its composition, low education and stresses non-discriminatory access education decreased. to education. In this section, we will first describe how educational attainments have evolved over 190. The growing skills mismatches emphasize growing, time; then we will describe the factors that favor unmet demand for skilled workers. This sharp increase human capital accumulation. In the last part, we will is a sign of weak job creation and growing risks of describe the dynamics of the sectoral differences long-term structural changes in the labor market in returns to education and wage premiums over due to growing skills mismatches. It underscores 11 years. the need for policies that ensure the best possible matches in the labor market. 192. We observed a general trend toward increasing in educational attainments in Mauritius. Figure 73 shows the shares of the population by highest level H. Human capital is growing but not of education competed. We include only individuals intergenerational mobility in human capital who have already finished their studies; i.e., young people still in school are excluded. The vast majority Gaps in access to education do not seem to diminish over (75 percent) of individuals earned either primary time. The offspring of well-educated and rich families or secondary degrees, but the two categories are will invest more in education, increasing the probability moving in opposite directions. Primary school is of preserving their favorable economic positions; poorer steadily declining while secondary school is steadily Figure 73: Educational attainment 2001-12 Figure 74: Labor-force status by highest educatio- nal level—2012 .45 .9 0.878 .4 .8 .35 .7 Share of the population 0.646 Share of group .3 .6 0.532 .25 0.509 .5 0.460 0.416 .2 .4 .15 .3 0.295 .1 .2 0.115 .05 .1 0.053 0.058 0.031 0.006 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 0 No Education Primary Secondary Above secondary No Education Primary Secondary Post-Secondary Employed Unemployed Inactive Source: Authors’ calculations 90 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges Box 10: Educational attainments in Mauritius 2001-12 Education in Mauritius is organized in four cycles (UNESCO, 2010): pre-school, primary, secondary, and higher education. Pre- school (ages 0-5) is organized in two separate phases to meet different needs: toddlers (ages 0-3), known as early childhood development (ECD), and early childhood care and education (ECCE, ages 3-5). With the 2008 reform, primary education now lasts six years, divided into three two-year cycles. Primary education is mandatory from age 5. At the end of primary school, all pupils take a standardized national test, the Certificate of Primary Education (CPE). It serves two purposes: measuring achievement and verifying eligibility for admission to secondary education. Secondary education consists of two cycles. The first has two stages: Forms 1-3 provide students a more or less common curriculum and Forms 4-5 prepare students for the O level, covering both core subjects and a wide range of optional modules. The second cycle prepares students for the A level. Higher education is offered by polytechnics and universities. Polytechnics generally offer two-year programs in information technology or business and administration, while universities offer three-year bachelor’s degrees, potentially followed by two-year master’s degrees in various subjects. Educational attainment lowers the probability of being inactive and raises the probability of being employed. Figure 74 shows additional evidence of this beneficial link. Looking at the 2012 labor-force status for people in each of the four educational categories allows us to identify two clear points of discontinuity—the first at the end of primary school, the second at post-secondary graduation. Completing the lowest educational level increases the probability of being employed from 11.5 percent to 50.9 percent and almost halves the probably of being inactive from 87.8 percent to 46 percent. The jump between secondary degree and post-secondary education is less abrupt but still considerable, with an increase of 11 percentage points in the share of employed and a decrease of 12 percentage points in the share of inactives. Surprisingly, the proportion of unemployed rises with education, but this is due to the very different share of inactives among the four categories. The results displayed in Figure 76 suggest three clear messages. First, completing any education level, even the lowest one, substantially diminishes the probability of being out of the labor force. Second, the benefits of obtaining a post-secondary degree have increased over time; between 2001 and 2012, the gap between the probabilities of being inactive with a post- secondary degree and with no degree declined by 10 percentage points. Third, moving from finishing primary school to a secondary degree does diminish the probability of inactivity, but the difference is negligible and fairly constant. rising, and the gap of almost 10 percentage points percent for the mother’s side. Equally evident is the in favor of primary in 2001 flipped to a gap of 5 effect that fathers’ income plays on sons’ education. percentage points in favor of secondary in 2012. When compared to having a father in the bottom The general trend toward an increase in educational of the earnings distribution, a father in the second attainments is also visible in the slow but constant quartile increases the probability of schooling by 5.4 increase in the share of people obtaining post- percent to 15.6 percent. The probability for offspring secondary qualifications. This number jumped from of the richest families over the most economically 8.3 percent to 11.6 percent. disadvantaged families is even higher—an additional 12.9 percent in 2012. 193. Given the tangible benefits of education, it is important to understand who is investing in it and 194. This lack of inter-generational mobility has potentially the factors influencing the decision to acquire it. adverse effects for the growth potential of the overall People earning tertiary degrees usually belong to economy. Established social hierarchies are likely to families with more advantaged backgrounds. Their endure—given that education is the quickest and parents are better educated and richer than the safest ticket to better-paid positions and family rest of the country, and these differences are not background plays such a decisive role in accessing disappearing over time. Residents of Rodrigues higher education. The continuation of education Island, the poorest part of the nation, tend to drop and income from one generation to the next might out after the first educational level and very rarely waste valuable human resources by placing able but receive tertiary degrees. Time is not mitigating this disadvantaged individuals in low-skilled occupations, trend. In going from primary to secondary education, dissipating the potential value they could add to the the increase in the probability of tertiary education economy if prepared for professions better suited to varies between 5.1 percent and 10 percent for their talents. the father’s side and between 7.4 percent and 11 MAURITIUS | Inclusiveness of Growth and Shared Prosperity 91 CHAPTER 7 - Labor Market Characteristics and Challenges Table 6: Probability of accessing further education—selected variables 2001 2003 2005 2007 2009 2012 Father’ s Educat ion 0.060*** 0.103*** 0.051*** 0.089*** 0.073*** 0.092*** (0.014) (0.014) (0.010) (0.011) (0.012) (0.011) Mother’ s Educat ion 0.091*** 0.074*** 0.084*** 0.110*** 0.101*** 0.080*** (0.013) (0.014) (0.011) (0.011) (0.012) (0.012) Father’s Income Quartile: Second 0.081*** 0.106*** 0.156*** 0.070*** 0.054** 0.074*** (0.023) (0.022) (0.018) (0.018) (0.019) (0.019) Third 0.192*** 0.130*** 0.202*** 0.076*** 0.061*** 0.089*** (0.024) (0.023) (0.019) (0.017) (0.018) (0.019) Fourt h 0.271*** 0.223*** 0.306*** 0.129*** 0.182*** 0.129*** (0.026) (0.025) (0.022) (0.020) (0.021) (0.022) Note: Standard errors robust to heteroskedasticity in parentheses. */**/*** for significance levels at 10%, 5% and 1% respectively. Photo : © Haja Faniry Razafimahenina 92 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges 195. Family background is a stronger influence on post- 196. The measure for inter-generational mobility secondary education than on primary or secondary indicates that social status tends to be preserved education. Figure 75 sheds some light on the past and increasingly so. This is particularly true at the decade’s evolution of social mobility in Mauritius. highest educational level, the one granting access to We plotted the percentage variance in children’s the best occupations and where the highest rewards educational attainment explained by parents’ for investments in human capital can be reaped. education and fathers’ incomes over time. A This is bad news if a fully open society is considered coefficient of 0 implies that parents’ social status has optimal. no bearing on kids’ educational attainment, while a coefficient of 1 signifies that school achievement is perfectly predictable by parents’ socioeconomic I. Disadvantaged position of women in the status. The numbers on the vertical axis are hard to interpret; rather than looking at the absolute labor market values for each year, attention should be paid to how Mauritian women experience substantially lower this measure behaves over time and on its relative employment levels and higher unemployment and importance in explaining school success for the three inactivity levels than their male counterparts. Even if educational categories. The graph clearly shows we control for other characteristics, ample differences family background has a stronger influence on post- in participation rates persist between men and women, secondary education than on primary or secondary but the downward trend is encouraging. Unlike the gaps education. It is also noticeable how family status related to labor-force status, which are all on downward increasingly explained tertiary-education success, trends, the gender wage gap is severe and shows no sign especially after 2007. For primary and secondary of decreasing. In fact, it has widened in the most recent education, family background plays a minor role years. and very little differences are discernible between the two trends, reflecting the fact that these two schooling categories are compulsory. Figure 75: Importance of family background for schooling completed 0.03 0.025 % of variation explained 0.02 0.015 0.01 0.005 0 2001 2006 2012 Primary education Secondary education Post secondary Source: Authors’ calculations. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 93 CHAPTER 7 - Labor Market Characteristics and Challenges 197. Whether Mauritian labor market has offered sufficient final levels of the three rates and their percentage opportunities to women for the full deployment of changes for both genders. The employment rate is their talents is the important question we address substantially higher for men, and unemployment and in this section. It is well known (Jaumotte, 2003; inactivity rates are substantially lower. However, Pissarides et al. 2003) that high levels of female the gaps have been closing—with the exception of labor-force participation are often related to better the unemployment rate. This was mainly due to economic performance on a number of indicators— the steep decrease in the men’s rate, which was from GDP growth to welfare systems’ sustainability not been matched by the women’s rate.59 Almost and poverty reduction. half of Mauritius’ female labor force is still outside the formal market. If these women decided to 198. Males outperform women, but we see an encouraging participate in the labor market, their probability of tendency toward convergence. Using the indicators finding unemployment would be significantly lower from the previous section, we focus specifically than that of men. on women, comparing their performance to men. Figure 7Figure 76 shows that males outperform 59 If we were to use 2005 as base year, this indicator would also show women in employment, unemployment, and a tendency toward a narrower gap: the change in men’s unemployment inactivity rates over the 11 years: however, the gaps rate was -11.84, while the change for women was -31.22 over this shor- have been closing. In Table 7, we report initial and ter period. Figure 76: Main indicators: gender differences Employment rates Unemployment rates 100 3 100 0.8 90 90 0.7 2.5 Ratio male/Female 80 80 Unemployment rate Employment rate 0.6 Ratio male/Female 70 2 70 60 60 0.5 50 1.5 50 0.4 40 40 0.3 30 1 30 0.2 20 0.5 20 10 10 0.1 0 0 0 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Female Male Ratio male/female Female Male Ratio male/female Inactivity rates Real Wages 100 3 12,000 1.8 90 11,000 1.7 2.5 Ratio male/Female 80 10,000 1.6 Inactivity rate Ratio male/Female 70 9,000 Real Wage 2 60 8,000 1.5 50 1.5 7,000 1.4 40 6,000 1.3 30 1 5,000 1.2 20 0.5 4,000 10 3,000 1.1 0 0 2,000 1 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Female Male Ratio male/female Female Male Ratio male/female Source: World Bank staff elaboration on CMPHS data. 94 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges Table 7: Main indicators by gender, 2001-12 change Main Indicators 2001 2012 % Change Males 71.72 68.06 -5.1 Employment Females 31.72 36.51 15.1 Males 8.14 5.2 -36.12 Unemployment Females 12.14 11.83 -2.55 Males 16.4 22.54 37.44 Inact ivit y Females 48.38 47.33 -2.17 Source: World Bank staff calculations. 199. Several factors explain women’s low labor-force inactive—but to a lesser extent. Comparing 2001 participation in Mauritius. Both cultural and economic to 2012, the differential between married men factors play important roles. As described in the and women decreases steeply, largely because the Box 13, a married woman had an almost 59 percent difference between highly educated individuals now higher chance of being economically inactive than favors women. For the remaining characteristics, a married man in 2001, a gap that fell to 40 percent the differential probabilities persist and are fairly in 2012. Larger family size and, specifically, a higher constant. number of children increase the chance of being Box 11: Females’ inactivity rate explained Low female labor supply is a phenomenon common to many economies at all stages of economic development.60 Labor-force participation is influenced by short- and long-term factors. In the short run, it responds to wages and general unemployment levels; in the long run, it fluctuates with cultural expectations and roles, incentives and institutional rules set in the labor market, returns to education, and long-run productivity. The meager women’s participation rate might result from lower human capital or social norms, consequent divisions of tasks within the family, or other factors. Simple descriptive statistics such as those in Figure 76 and Table 7 do not capture all these effects. To show obtain more informative results and start to understand the high inactivity rate, we will resort to a simple estimation of the probability of being inactive, given a series of control variables.61 In Figure 77, we show the difference in probability of inactivity between men and women for 11 years, holding constant educational levels, age, marital status, and child-rearing duties. We see that women are still substantially less likely to participate in the labor market, but we can also appreciate the dramatic fall in the inter-gender gap. Figure 77: Females inactivity probability .35 Probability Inactive .3 .25 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Source: World Bank staff elaborations. 60 For an international survey of trends in female labor-market participation, see Mincer (1985). 61 The full specification for the probit model from which we obtain the coefficient for females’ inactivity probability is shown in the appendix. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 95 CHAPTER 7 - Labor Market Characteristics and Challenges In Table 8, we clarify what drives both level and rate differentials between the two genders.62 The coefficients should be interpreted as the difference in probability of being inactive between men and women with the same characteristics.63 In terms of levels, getting married exhibits the highest probability differential at both the beginning and end of the period. A married woman has an almost 59 percent higher chance of being economically inactive than a married man in 2001 and a 40 percent higher chance in 2012. An increase in family size and, specifically, the number of children increases the chance of being inactive—but to a lesser extent. The probability differential monotonically decreases with educational levels. In 2001, female primary-school graduates had an 11 percent higher chance of being inactive; the difference disappeared for post-secondary graduates. In 2012, post- secondary trained women actually had a lower probability of being inactive. In terms of changes, the differential between married men and women has decreased steeply, largely because the gap between highly educated individuals now favors women. For the remaining characteristics, the differential probabilities persist and are fairly constant. Table 8: Marginal effects of background characteristics on probability of being inactive 2001 2012 Difference (1) (2) (1)-(2) Married 0.588*** 0.403*** 0.184*** (0.011) (0.010) (0.011) Number of kids 0.014*** 0.011*** 0.003 (0.003) (0.003) (0.003) Education Primary 0.109*** 0.097** 0.012 (0.019) (0.018) (0.016) Secondary 0.073*** 0.093*** -0.019 (0.019) (0.017) (0.016) Pos t -Seconda ry 0.012 -0.066*** 0.078*** (0.023) (0.020) (0.021) Notes: World Bank staff elaborations. Standard errors in parentheses; * stands for 10 percent level of significance; ** stands for 5 percent level of significance; *** stands for 1 percent level of significance. Marginal effects at the mean. Reference category: men. 62 The coefficients show here are taken from the regression specified and reported in the appendix. They are the interaction terms between a dummy variable for female and the corresponding characteristic for 2001 and 2012 63 96 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges 200. Unlike the gaps related to labor-force status, which Blinder—from the names of the two economists who are on a downward trend, the gender wage gap in first proposed this methodology. The Oaxaca-Blinder Mauritius is severe and shows no sign of decreasing. procedure separates the existing gaps into what In fact, it widened in the most recent years. To can be “explained” by differences in two groups’ gain a better sense on the causes of the gender productivity characteristics and a residual part—the gap, Figure 78 shows the coefficients of a female “unexplained”—that cannot be accounted for by dummy regressed on the log of wages, separately these differences. This “unexplained” component estimated for each survey year (a so-called Mincer is often regarded as a measure of discrimination regression).64 This line is the percentage wage because its existence cannot be justified by difference between women and men for each individual characteristics that could influence survey year, keeping fixed a series of demographic labor productivity and compensation. We apply characteristics, such as education, potential work this procedure to our data with the intention of experience, and geographical location. looking for possible discrimination against women in the Mauritius’ labor market and determining a Figure 78: Explained and unexplained rough quantification of its extent. It must be noted gender wage gap Gender gap based on Mincer regression Oaxaca-Blinder decomposition 1.1 1.05 % Unexplained 1 1 0.9 .95 .9 Gender gap premia 0.8 0.7 .55 0.6 0.51 0.525 .45 0.478 0.507 Log Wage Difference .35 0.5 .25 0.4 .15 0.3 .05 0.2 -.05 2001 2004 2007 2012 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2012 Source: World Bank staff elaborations. Unexplained Endowment 201. Even when comparing men and women with the same that the unexplained component cannot be taken education levels, age, potential work experience, and as an exact measure of discrimination because the sector of employment, women still earn between 48 validity of the decomposition and its interpretation percent and 53 percent less than men, depending on hinges on the assumption that all factors influencing the year. Looking at the raw differentials, no traces labor market productivity have been included in the of real convergence are emerging—a contrast to linear regression. It is easy to understand how this the encouraging signs regarding women’s labor- assumption is seldom respected in reality; however, market participation. Nonetheless, the dynamic of the numbers in the remainder of this section should women’s pay might be less bleak than a superficial be interpreted more as a tendency than a proper look at the data suggests. As more women enter the quantification of discrimination. labor force, the average working woman in 2012 will probably be less able and productive than the 203. The negative endowment effect is capturing the fact average working woman in 2001, so the constant that in recent years Mauritian women have become gender wage gap might be hiding diminishing labor- more educated, on average, than Mauritian men, market discrimination against female. but this fact has not translated into better wages. Figure 78 shows the results of the Oaxaca-Blinder 202. A commonly used methodology to study labor- decomposition for a period of 11 years. The height market differentials that might be ethnic or gender of the bars in the right panel (bottom) quantifies related is to decompose mean differences in log the log of hourly wage gap between male to female wages into two components: the explained and salaries. The raw differential is basically constant the unexplained. The procedure is called Oaxaca- over time. It can be decomposed into what is justified by differences in endowments between the average 64 The full regression is shown in the appendix. working man and the average working woman and MAURITIUS | Inclusiveness of Growth and Shared Prosperity 97 CHAPTER 7 - Labor Market Characteristics and Challenges what remains after the endowment effect has been a post-secondary degree. taken into account—i.e., what cannot be attributed to differences in characteristics between the groups. 208. Even after taking into account all the shortcomings In this graph, the endowment component is marked of the method for detecting gender discrimination, in red, while the unexplained component is in green. this section’s striking numbers should raise serious concerns among policymakers about how women 204. Women have the same observable characteristics as seem to be considerably disadvantaged in the men, but their wages are much lower. It is evident that labor market. In the medium to long run, this the unexplained part of wage differential strictly persistent undervaluation of females’ labor input dominates the justifiable part in all of 11 years. The might discourage Mauritian girls from maintaining right panel of Figure 78 (top) highlights the trends in the favorable secondary and tertiary education the ratio of the explained and not explained in the enrollment rates they have reached in recent gender wage differential. It reveals an interesting years. The evidence regarding the gender gap and phenomenon—namely, that the unexplained the performance of women in the Mauritian labor component exceeds 100 percent of the total wage market is ambivalent. On one hand, the main differential starting in 2008. This is also evident from indicators point to a manifest gender disparity. the right panel (bottom), where the endowment Compared to men, women were still 22 percent less effect turns negative starting in 2008. A negative likely to be employed in 2012, 6 percent more likely endowment effect indicates that if women had the to be unemployed, and 25 percent more likely to same observable characteristics as men—in terms be out of the labor market. But all these gaps are of educational level, age, potential experience, closing. The estimated probabilities indicate that geographical location, and sector of employment— the main factors behind remaining disparities are their wages would in fact be lower than what we the traditional and deep-rooted norms of family observe in the raw data. roles. Married women tend to stay home as the main caretakers for children. On the positive side, these Females’ educational norms seem to be evolving toward increased gender achievements are not different parity. The acquisition of human capital might serve as an incentivizing mechanism, pushing women to be than those of males. economically active, and should be encouraged. In conclusion, the existing gender gaps favoring male 205. In this section, we look at women’s educational workers in unemployment and participation rates achievements and consider whether the existing do not seem justifiable by differences in acquired gender gap might be justified by women’s investment, skills. In fact, these differences, to the extent they or lack thereof, in their own human capital. exist, point toward more, not less, human-capital accumulation by Mauritian women. 206. The graduation rates for individuals with primary and post-secondary education are fairly similar for men and women. However, the small post- J. Disadvantaged position of young workers secondary graduation gap that favors men has been diminishing, a reflection of the improved conditions 209. Young workers between ages 15 and 24 are particularly for Mauritian women. At the zero level, enrollment vulnerable to labor-market fluctuations. This is not rates are equal for men and women, and a positive surprising. Economic theory suggests that employers difference indicates a higher incidence of graduation are reluctant to lay off more experienced workers who for that schooling category among women and vice have acquired both general and firm-specific on-the- versa for a negative difference. job-training and whose severance costs are usually higher. 207. The genders do not differ on graduation probabilities. To further corroborate our findings of minimal 210. A vast body of literature (Arulampalam, 2001; gender educational gaps, we have estimated a linear Holzer and LaLonde, 2000; Khan, 2010; Neumark, probability model for the likelihood of completing 2002) has highlighted how early labor-market each of the four education levels, controlling for experiences shape individuals’ future paths. This geographical location, family background, and age.65 phenomenon, which has been called the “scarring Women display a lower probability of belonging to effect,” would imply that difficulties in the school- the two lowest educational categories, a fairly to-work transition would assign affected individuals similar probability of being secondary graduates, to a suboptimal path on which earnings would be and a higher, and increasing, probability of obtaining lower and the probabilities of unemployment and slipping out of the labor force altogether would be 65 A complete description of the estimated regression can be found in higher. These individuals are often at risk of poverty, the appendix. and their separation from the labor market should 98 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges Photo : © Haja Faniry Razafimahenina MAURITIUS | Inclusiveness of Growth and Shared Prosperity 99 CHAPTER 7 - Labor Market Characteristics and Challenges not be minimized. Easing this transition seems to be rates could hide a contemporaneous increase in a necessary condition for a well-functioning labor school enrollment or formal training for this age market. How the Mauritian labor market treats its group. Whether young individuals are out of the youngest participants is the subject of this section. labor force as a necessity or by deliberate choice obviously has very different implications. If the rise 211. In Mauritius, young people experience substantially in inactivity rates can be at least partly attributed worse labor market outcomes than the rest of the to gains in education, we could conclude that this population. Figure 79 shows the main indicators for pattern should be encouraged more than feared. two subgroups: young workers (ages 15 to 24) and the rest of the population (over age 25). As expected, 213. The portion of young individuals who are neither in young people have lower employment rates and education nor in training and unemployed (NEET) higher inactivity and unemployment rates. The decreased considerably after 2005, reaching a magnitudes of these differences are fairly constant minimum in the past two years. The trend reflects from 2001 to 2012. The employment rate gap has an increase of young people in education. Figure decreased slightly in the past two years, mainly 74 showed the pattern of school enrollment for the because of a decrease among the older group rather 15-24 age group, which allows us to deduce that than any gains among the young. The unemployment the inactivity rate’s post-2006 increase has been rate gap has been constant, while the inactivity rate primarily driven by more people in education. In gap has been increasing since 2004, mainly due to nine years, the percentage of people between 15 the uptick in the youth rate. The common element and 24 in education increased from just above 35 among these three indicators is young workers’ percent in to about 50 percent (Figure 81).66 greater volatility, reflecting their higher sensitivity to the economic cycle. 212. The increase in the young workers’ inactivity rate might reflect a more complex interplay between the labor market and education. In fact, rising inactivity 66 No data available prior to 2003. Figure 79: Main labor indicators by age group Employment rate, by age cohorts Unemployment rate, by age cohorts 70 70 60 60 50 50 Employment rate Unemployment rate 40 40 30 30 20 20 10 10 0 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Agre 25+ Age 15-24 Agre 25+ Age 15-24 100 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges Inactivity rate, by age cohorts Percent of 15-24 age cohort neither in education nor in training and unemployed (NEET) 70 70 % of Youth (15-24) population 60 60 50 50 Inactivity rate 40 40 30 30 20 20 10 10 0 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Age 25+ Age 15-24 Neet In School Source: World Bank staff elaboration on CMPHS data. Figure 80: Youth unemployment rates, international 70 Unemployment Youth Ratio Unemployment rates, youth and total 60 rate Unemployment World Average 9.1 20.6 2.4 Mauritius 7.9 21.7 2.7 50 40 Mauritius 30 20 10 0 Qatar Macao Bhutan Norway Korea, Rep. Peru Austria Ecuador Paraguay Netherlands Taiwan, China Panama Luxembourg Mexico Kazakhstan Samoa Uruguay New Zealand El Salvador Czech Rep. Philippines Chile Argentina Romania Costa Rica Sweden Cyprus Ukraine Venezuela Morocco France Turkey Hungary Bulgaria Jamaica Jordan Croatia Bahamas Dominican Rep. Lithuania Latvia Tunisia Montenegro Spain South Africa Macedonia Unemployment rate (total) Youth Unemploymnet Rates MAURITIUS | Inclusiveness of Growth and Shared Prosperity 101 CHAPTER 7 - Labor Market Characteristics and Challenges 214. Youth unemployment is slightly higher in Mauritius than in other counties, but the rates are not extraordinary. K. Mauritius labor market -areas of focused Figure 80 shows Mauritius is 54th among the 92 countries available in the World Bank’s databases attention on youth unemployment. The ratio of youth-to-total unemployment is slightly higher in Mauritius than in i. Boost productivity other countries. 216. The country growth model should be changed to boost 215. Young individuals aged 15 to 24 experience worse productivity. It is evident that wage increases above labor-market outcomes in terms of employment, productivity gains eroded the competitiveness of unemployment, and inactivity rates. On the other traditional sectors and lowered private investment hand, the increasing number of inactives is due and employment creation. Developing a new growth primarily to a dramatic increase in young people model for Mauritius will require steps to raise the in education over the past 10 years. More than 50 country’s competitiveness by improving productivity percent of the inactives in this age group are enrolled at the firm level and easing access to financing. in some kind of formal education or training. Equally An additional research is required to develop impressive is the decline of NEETs among the 15-24 comprehensive model boosting productivity. age group, now at a low point. Box 12: Determination of NEET employment—empirical model This box analyses the characteristics affecting youths’ probability of being NEETs. In Table 9, we show the impact of selected family and individual demographic characteristics on the probability of falling into this particularly disadvantaged group in 2003 and 2012 and the differences between the two periods. Clearly, parents’ education and family income have the largest influence. In 2003, fathers’ education decreased the probability of being NEETs by 2.6 percent and mothers’ by 3.3 percent. The figures for 2012 suggest a change—now, only fathers’ education is negatively related (4.7 percent) to the probability of falling into the NEET category. Family income, represented by fathers’ wages, is expressed in log terms and should be interpreted as the percentage change in the probability of being NEETs associated with a 1 percent increase in wages. The coefficient for females reflects the improved conditions discussed in the previous section. Compared to boys, they had a 10 percent higher chance of being NEETs in 2003, but their comparative situation dramatically improved in 2012, when the coefficient was down to 3.9 percent. The only other factor increasing the probability of being NEETs is the presence of other siblings. This is true only for 2003, while the coefficient for 2012 is quite precisely estimated at 0, expressing a lack of correlation between these two factors in that year. Only two characteristic showed statistically significant changes— the number of siblings (not influencing NEET probability in 2012) and the female covariate. All other variables are fairly constant. Table 9: Probability of Being NEET (15-24) 2003 2012 Difference Main Indicators (1) (2) (1)-(2) Sibling s 0.023*** 0.005 0.018*** (0.004) (0.003) (0.005) Fat h er Educat ion -0.026** -0.047*** 0.022 (0.012) (0.008) (0.014) Mot h er Educat ion -0.033* -0.012 0.012 (0.011) (0.008) (0.022) Lo g Fat her Wage -0.017** -0.008** -0.009 (0.006) (0.004) (0.007) Mot h er Employed -0.023 -0.035*** 0.012 (0.019) (0.011) (0.022) Fe male 0.104*** 0.039*** 0.065*** (0.0138) (0.009) (0.017) Notes: World Bank staff elaborations. Standard errors robust to heteroskedasticity in parentheses; * stands for 10 percent level of significance; ** stands for 5 percent level of significance; *** stands for 1 percent level of significance. 102 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 7 - Labor Market Characteristics and Challenges Photo : © Haja Faniry Razafimahenina ii. Raising incentives and boosting competiveness fully reap the benefits of economic growth because some employment opportunities are diminishing and 217. The most far-reaching phenomena shaping the many workers lack adequate skills for today’s labor Mauritian economy is the ongoing transition from an market. The Government has invested substantially economy based on low value-added manufacturing in providing widespread and equitable access to to one in which technology and innovation will basic infrastructure and free health and education play a major role in future growth. This process is for the entire population. Across the board, however, still at an early stage, but enough has happened to issues related to the quality of these public services suggest that this transition will affect an increasing explain the diminishing prospects among the most number of jobs and workers. Unfortunately, the vulnerable and intergenerational inequality. current labor supply seems to be only partly fit for the emerging economy. Even though education iii. Easing regulatory burdens achievements are markedly improving and the most common qualification is now a high school degree, 219. At the lower end of wage distribution, inequality the lack of highly qualified labor force might hinder seems to be rising at a slower pace. In this regard, the development of high-tech industries. the changes in the minimum-wage regulations might contribute to sustaining the incomes of the lowest 218. The changing labor-market structure tends to reward segment of the wage distribution. They should skills, especially those acquired in tertiary education, be kept in place and re-evaluated periodically to and reduce the payoffs of workers without sufficient maintain the minimum wage’s real value. However, qualifications. Our analysis finds an increase in the current system of the wage setup and minimum- earnings inequality, especially at the top of the wage determination only hardly shortchanges the wage distribution. A plausible cause is the upsurge of poorest but also imposes other constraints on high-tech jobs and the contemporaneous decline of development. traditional manufacturing occupations. In this regard, promoting tertiary education has the triple virtue of 220. The wage-settling mechanism fails to keep salary assisting the development of high-tech industries, increases in line with sector productivity. As presented setting the right environment for future growth, and in this chapter, labor-market institutions constrain guaranteeing that inequality stays in check as the the economy’s capacity to create jobs. Key factors high-tech sector takes a larger share of the labor are spillovers from the more dynamic sectors and force. Wage income is the main driver of prosperity large public-sector salary increases as well as inertia in Mauritius, yet the most vulnerable struggle to in determining wage growth in relation to inflation MAURITIUS | Inclusiveness of Growth and Shared Prosperity 103 CHAPTER 7 - Labor Market Characteristics and Challenges rather than productivity. As a result, unit labor costs first factor should be encouraged, and the second is in certain sectors rise too quickly, undermining a result of extension of pension coverage. However, competitiveness and employment creation. A rigid the high women’s inactivity rate needs to be reduced. system of determining pay increases and complex Their feeble participation cannot be attributed to labor regulations tend to limit the ability of the lower human capital accumulation; on the contrary, economy to undergo structural changes. Our report educational attainment is higher among women than also suggests a need for further reform of labor men. In light of the evidence, it is probable that regulations and wage determination in Mauritius. the reasons behind the unfavorable treatment of women in the labor force are deep-rooted and hard 221. Dedicated efforts will be needed to raise the quality to influence by norms. The large salary gap and the of the entire education system, including a vocational extremely high estimated discrimination parameter stream responsive to private-sector demands. also reflect the status of women. Employment is vital in shaping household income in Mauritius. If not adequately corrected through 223. Policies with the potential to activate female labor training later on, inequity in education outcomes market participation include: implementation of results in income inequality, ingraining substantial a special fiscal regimes favoring women’s labor, intergenerational poverty. In the short term, large affirmative action measures to discriminate in favor targeted-training programs could be envisaged to of women in the labor market, and public provision retool the Mauritian labor force in line with current of child care. market requirements, boosting employment and income generation. 224. Employment policies for young people ages 15 to 24 deserve further support. The number of individuals iv. Improving conditions for women and the youth within this age group enrolled in some form of education or training has been increasing in recent 222. The labor-market participation rate is unsatisfactory years, while the number of individuals neither in in general and dramatically low for women. The low school nor working has decreased considerably. activity rate is mainly driven by people in training, Several youth-related policies have been already retirees, and women occupied in family care. The implemented in Mauritius. 104 MAURITIUS | Inclusiveness of Growth and Shared Prosperity Chapter 8 Evidence from Firm-Level Analysis 67 This chapter was prepared by Leora Klapper (DECRG-FDP) and Peter van Oudheusden (DECRG-FDP) as material for the 67 analysis “Mauritius Inclusiveness of Growth and Shared Prosperity: Micro Analysis and Labor Productivity Gains.” We thank the Mauritian Company Registrar for providing the data used in this analysis. Mauritius’ institutional framework has improved in recent years, and the country performs relatively well compared to other SSA countries when it comes to access to financial services. However, it has mixed ratings on global competitiveness indicators. Small enterprises reported infrastructure and informal sector practices as major impediments, and they continue to face challenges to increasing market share and employment. Small firms are more leveraged than large ones in Mauritius. The large firms indicate shortages of skilled workers and labor regulations are their major challenges. The entrance of new firms has been stagnant since the global financial crisis. The number of new firms almost quadrupled between 2002 and 2008 but remained unchanged in the subsequent years. Business creation stagnated despite the introduction of the reforms at the end of the 2000s, which made it easier and cheaper to start a company. In 2007-12, the construction and services industries saw the most new firms, while startups were relatively low in textiles and manufacturing. The inequality in firms’ revenues has widened in recent years, especially among older firms. On average, SMEs are unprofitable and disadvantaged in terms of growth prospects. Firms in agriculture and textiles are less profitable than those in trade, construction, and services. The number of firms with new loans has increased since beginning of the 2000s, but it dropped after 2009, with the lingering fallout from the financial crisis. Credit increased in trade and services and fell in manufacturing. Financial vulnerability is especially high for small firms and new corporations. CHAPTER 8 - Evidence from Firm-Level Analysis A. Introduction firms’ financial structure. Unprofitable firms are more likely to have more liabilities relative to their assets, 225. This chapter’s objective is to improve the and their current liabilities are more likely to exceed understanding of the private sector’s performance their short-term assets, indicating they may not be and inclusiveness by looking at firms’ composition, able to satisfy their short-term obligations. characteristics, and dynamics. It looks at profitability, access to credit, and vulnerabilities to identify the 229. The rest of this chapter discusses these findings in potential of supportive government policies. more detail. Section B talks about the challenges and advantages of the Mauritian labor market. Section 226. After being relatively stable up to the beginning C discusses the creation of new firms and their of the 2000s, new incorporations almost quadrupled characteristics. Section D discusses the sales and between 2002 and 2008. In 2007-12, the construction profitability of firms. Section E discusses the financial industry saw he most new incorporations. Differences structure of firms and their access to credit. Section between new firms in terms of turnover, or sales, F discusses the financial structure of firms and its are large. The smallest 60 percent of new firms only relationship to profitability. produce a fifth of all sales created by new firms. In general, inequality in sales runs high. Moreover, growth in sales is not concentrated in particular firms, B. Challenges and advantages of the Mauritian so that the distribution of sales was stable in 2007-12. Sales are higher for older, more established firms. private sector 227. Access to credit is widespread in Mauritius and 230. Mauritius has mixed ratings on global competitiveness not concentrated in particular firms. The country’s indicators. The general rating places Mauritius institutional framework has improved in recent years, 39th among 144 countries worldwide—a relatively and Mauritius performs relatively well compared good position. However, Mauritian private-sector to other SSA countries when it comes to access to development and competitiveness have their financial services. The number of firms with new loans advantages and disadvantages (FigureFigure 82). has increased since the beginning of the 2000s but the According to Global Competitiveness Index’s historical number dropped in 2009 and subsequent years, with database, Mauritius has many indicators usually the lingering fallout of the financial crisis. associated with highly competitive economies, while other indicators fare much worse. Mauritius has high 228. Profitability also varies starkly between firms. ratings in trade tariffs, business rules of law, investors’ Very small firms are more likely to be unprofitable. protection, accountability regulations on security, Compared to the average firm, profitability is banking, and generally good market efficiency. relatively low for small firms in agricultural and However, the country is far below average on wage medium and large firms in the textiles. Although being determination, labor regulations, and position of an older firm or being new incorporation is related women in the labor market. The country also lags in to profitability, the relationship is much stronger for R&D and capacity for innovations. Photo : © Haja Faniry Razafimahenina MAURITIUS | Inclusiveness of Growth and Shared Prosperity 107 Figure 81: Mauritius competitiveness indicators, country rating (lower is better), 2013/14 108 Relatively good rating Relatively worse rating Trade tariffs, % duty 4 Flexibility of wage determination 120 Business impact of rules… 7 Women in labor force 118 Effect of taxation on… 9 Strength of investor… 12 Availability scientists & engineers 116 Accountability 14 HIV prevalence, % adult pop. 113 Soundness of banks 15 Foreign competition 15 Capacity for innovation 112 Domestic competition 17 Hiring and firing practices 112 Control of international… 18 MAURITIUS | Inclusiveness of Growth and Shared Prosperity Market Flexibility 18 Efficient use of talent 101 CHAPTER 8 - Evidence from Firm-Level Analysis Protection of minority 18 University collaboration in R&D 101 Organized crime 20 No. days to start a business 21 PCT patents, applications/pop. 94 Efficiency of legal… 22 24 Quality of scientific institutions 91 Regulation of securities… Intensity of local… 24 Individuals using Internet, % 85 Goods market efficiency 25 Strength of auditing and… 25 Innovation 76 Source: World Economic Forum database, Global Competitiveness Index, historical database 2013-14 CHAPTER 8 - Evidence from Firm-Level Analysis Photo : © Luke Manhood MAURITIUS | Inclusiveness of Growth and Shared Prosperity 109 CHAPTER 8 - Evidence from Firm-Level Analysis 231. Access to financing is a major obstacle for all firms medium and large firms. Small and medium companies in Mauritius. Small enterprises reported infrastructure report a lack of ability to grow and increase their and informal-sector practices as major obstacles. Large employment. firms indicate lack of skills and labor regulations as their major challenges. The analysis of the main 233. Limited access to financing may be part of the private- obstacles facing Mauritian enterprises is based on sector challenge. Mauritian firms face challenges in data from the Enterprise Surveys. Figure 82 presents a accessing credit and financing investments. The snapshot of the top six business obstacles as identified structure and incentives in Mauritius’s financial by small and large firms in Mauritius. Access to sector creates biases favoring larger companies. The financing is the main obstacle for both small and large Mauritian Government is taking steps to address some enterprises. However, larger enterprises identified of the challenges that Mauritian companies face. For skills mismatches and inadequately educated the past years, the Government earmarked over MUR6 workforce as the second obstacle. In addition, larger billion through various types of lending and non- enterprises gave labor regulations as one of the lending instruments. Credit is provided at subsidized top most-problematic factors for doing business in rates, with partial state guarantees. While these Mauritius. schemes are in line with the Government strategy of Figure 82: Obstacles of doing business in Mauritius, improving the SME operating environment, areas of 2009 68 duplication as well as some gaps continue. Small firms (5-19) Large firms (100+) 40,0 40,0 35,0 29,1 35,0 27,2 30,0 30,0 25,0 25,0 17,4 15,4 18,1 20,0 20,0 14,4 13,8 15,0 10,7 7,7 15,0 6,5 10,0 5,7 10,0 4,9 5,0 5,0 0,0 0,0 ce or y ... n e ce e n or n s te cit rc rc tio io io nd ct ct an an fo fo at pt ra tri rta se se ta fin fin ks ks ul ru x ec po Ta al al eg ef r or or to to El Co m m th ns rr w w s s r r fo fo bo ce ce a of of e, Tr In In im Ac Ac La n n tio tio Cr ca ca u u Ed Ed Source: Mauritius enterprise survey, 2009. 232. SMEs continue to face challenges in increasing market 234. Tertiary-education expansion needs to focus on share and employment. Small establishments that innovation and R&D. Mauritius ranks 54th in higher employ less than 10 people represent 90 percent education and training. These ratings reflect of all businesses in Mauritius, but they employ just low tertiary-education enrollment rates, weak 54 percent of the workforce. The top 10 percent collaboration between universities, research, of firms account for 40 percent of all sales, while and industry, and low availability of scientists and around 60 percent of SMEs generate only a fifth of engineers. Mauritius needs to attract and retain more all sales. This distribution has remained unchanged talent to meet the need to improve the availability, since 2001 despite efforts to liberalize the economy. quality, and relevance of skills. The need to enter into Furthermore, around 70 percent of small firms are new markets and sectors and increase the knowledge highly leveraged,69 compared to roughly 55 percent of content of existing products will require attracting overseas talent. Employment surveys reveal that it 68 It is important to emphasize that Mauritius enterprise Survey was is becoming more difficult to find employees with conducted in 2009 during the global economic crisis. Access to finan- appropriate experience and proper attitudes. The ICT cing was a major challenge during this period. The situation has impro- and the financial sectors report especially large labor ved since then and the results on access to financing should be treated accordingly. and skills shortages. 69 Highly leveraged firms are defined as those with a liabilities-to-as- sets ratio above two-thirds. In addition, a distinction is made for firms Firms with short-term liquidity problems have a current ratio below with either short-term liquidity problems or short-term liquidity risks. one, meaning their current-liabilities exceed their current assets. 110 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 8 - Evidence from Firm-Level Analysis C. Creation of new firms 236. In 2008, Mauritius made starting a business faster by implementing a centralized database linking the company 235. New incorporations were relatively stable up to registry with tax, social security, and local authorities.70 the beginning of the 2000s and start to increase Before 2008, it took 46 days to start a business; after in the middle of the decade. Figure 83 shows the that, it dropped to six, greatly reducing the time barrier number of new incorporations of non-financial firms to register a business (Figure 84). At the same time, in the Registrar of Companies over a 15-year period, the actual cost of starting a business—here measured starting in 1997. Average annual new incorporations as a share of income per capita—in 2013 was only a were around 1,300 before 2002, followed by a third of that in 2005. In terms of cost and time, it steady increase in the mid-2000s, when the number was much easier to start a business at the end of the quadrupled to more than 5,200 in the year 2008. 2000s than it was in the middle of the decade. The Information for 2009 is incomplete, but the numbers reduction in barriers coincides with the high level of for subsequent years show that new incorporations new incorporations illustrated in Figure 84. remain high but slightly below 2008. 70 See http://www.doingbusiness.org/reforms/overview/economy/ mauritius for an overview of other reforms. Figure 83: Number of new incorporations over Time 6 000 Source: Mauritian Company Registrar and authors’ 5 000 # o f N ew In co rpo ra t ion s calculations. Notes: The average yearly number of 4 000 new incorporations 1992-96 3 000 was 860. The year 2009 is not shown because data are 2 000 incomplete. The number of new incorporations in 2009 1 000 was 1,699, but no data are 0 available after May of that year. Figure 84: Ease of starting a business 50 12% (% of income per capacita) # of days to start a business 10% Cost to start abusiness 40 8% 30 6% 20 4% 10 2% 0 0% 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Days to start a business Cost to start a business (% of income per capita) Source: Doing Business (World Bank) and authors’ calculations. a procedure with minimum follow-up with government agencies Notes: More information is available on the website of Doing and no extra payments. “Cost to start a business” is recorded as a Business at http://www.doingbusiness.org/data/exploreeconomies/ percentage of the economy’s income per capita. It includes all official mauritius. “Days to start a business” is defined as the total number fees and fees for legal or professional services required by law. The of days required to register a firm. The measure captures the median costs exclude bribes. duration that incorporation lawyers indicate is necessary to complete MAURITIUS | Inclusiveness of Growth and Shared Prosperity 111 CHAPTER 8 - Evidence from Firm-Level Analysis 237. New firms are more likely to be in the construction the MUR80 million or more in sales to qualify as large and services industries; construction had the largest firms. The remaining 10 percent are medium firms, relative increase—from 6 percent to 9 percent. In with sales falling between those of small and large 2007-12, 35 percent of all firms in the Registrar of firms. The size of new firms, measured by turnover/ Companies were new incorporations.71 Figure 85 sales, does not differ across industries. shows the firms by industry over this period, where a distinction is made between “new” firms that are new 239. Majority of new firms hardly generate much revenue; a incorporations and established firms, designated as small number of firms account for a large share of overall “old.” Although the overall share of firms in textiles is sales. Although the majority of new firms are small, small, this industry saw the largest relative decline— they do differ in terms of sales. The solid blue line in from 3 percent to 2 percent. The manufacturing Figure 86 shows the distribution of sales for new small industry also saw a large relative decline. firms in 2007-12. Sixty percent of these firms generate only a fifth of all sales, while 80 percent generate 40 238. Around 88 percent of new firms are defined as percent of all sales. Hence, a small number of firms small, meaning that they have sales of MUR10 million are responsible for a large share of sales, even when Mauritian or less. Barely 2 percent of new firms have looking only at small firms. Indeed, the top 10 percent of firms are good for almost 40 percent of all sales. 71 Most of the analyses in this chapter are restricted to 2007-12 due The corresponding GINI coefficient of sales for small to data availability. On average, there are almost 11,000 yearly observa- tions over this period; the total number of observations in 2001-06 was new firms is a relatively high 54. only 191. Figure 85: Industry compositions for new incorporations and other firms, 2007-12 "New" firms "Old" firms Agriculture/Extractive Construction Manufacturing 9% 6% 6% 8% Services Textiles Trade 37% 42% Source: Mauritian Company Registrar and authors’ calculations. Notes: “New” firms are defined as being a new incorporation in 2007-12, and “old,” or established, firms are defined as any other firm in the Registrar of Companies in the period. A caveat is that industrial classifications are based on textual descriptions of the firm, which may not be precise. A firm can belong to multiple industries. 112 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 8 - Evidence from Firm-Level Analysis Figure 86: Distribution of sales (2001-12) (2001-06 average) (2007-12 average) 100% 100% 80% 80% Cumulative share of sales Cumulative share of sales 60% 60% 40% 40% 20% 20% 0% 0% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Cumulative share of firms Cumulative share of firms Small New Firms Old Firms Small New Firms Old Firms Source: Mauritian Company Registrar and authors’ calculations. Notes: “New” firms are defined as new incorporations in 2001-12, and “old” firms are defined as any others in the Registrar of Companies in the period. A “small” firm is defined as having MUR10 million in sales or less. Old firms’ sales are winsorized at the 5 percent level. 240. The inequality of sales is lower for small firms, their greatest possibilities to exploit economies of scale. sales are lower, and they are generally similar. The These industries also have the oldest firms. Table 10 inequality in sales increases among larger firms. Figure illustrates that size is positively associated with age. 86 shows the same information for 2001-06 and 2007- The typical (median) small firm was nine years old 12, facilitating a comparison of the distributions of in 2014, while longevity was 12 years for the typical sales over time. In terms of the distribution of sales, medium and 19 years for the median large firm. the two periods show very few differences. The GINI coefficients of sales for small new firms are 54 for 242. Although profitability is higher for medium and both periods. The number of unique firms for which large firms, it is not necessarily related to firms’ age. information is available increased from roughly 11,000 Regression analyses show that almost 10 percent of all to 18,000, capturing the increase in new incorporations variation in firm size, based on sales, can be explained as shown in Figure 83. Based on the distribution of by firm age alone. For profits, firms’ age explains only sales and corresponding GINI coefficients, inequality slightly more than a percentage point of the variation.72 in sales remained stable over time for small new firms The variation of average profitability across industries and increased slightly for other firms. is limited. A notable exception is textiles, where the average small or medium-sized firm has negative profits. Although the textile industry’s typical firm D. Size and profitability of firms performs slightly better, its return on assets (ROA) is much lower than those of firms in the other industries. 241. In all industries, at least 70 percent of all firms Overall, textile industry profitability falls short of the are small, and roughly 20 percent are medium-sized. average for all firms in Mauritius. A notable exception is the services industry, where almost 80 percent of firms are small, with slightly more than 15 percent classified as medium. Manufacturing 72 Regressing size on age and a constant gives an R2 of 0.09, while regressing ROA on age and a constant gives an R2 of 0.01. The coef- and textiles have the biggest share of large firms ficient of age enters positively with statistical significance in both re- at 10 percent, suggesting these industries have the gressions. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 113 CHAPTER 8 - Evidence from Firm-Level Analysis Table 10: Firm size, age, and profitability by industry (2007-12 average) Construction Manufacturing Services Textiles Trade Small 76% 70% 78% 72% 77% Med i um 19% 22% 16% 18% 18% Large 5% 9% 6% 10% 5% Small Firms Age (average) 8.25 11.02 10.26 11.68 10.34 ROA (average) -0.05 -0.06 -0.04 -0.09 -0.05 Age (median) 7 8 8 9 8 ROA (median) -0.02 -0.02 -0.01 -0.06 -0.02 Medium Firms Age (average) 12.87 17.81 14.73 18.54 14.99 ROA (average) 0.08 0.06 0.07 -0.02 0.05 Age (median) 12 15 12 17 12 ROA (median) 0.09 0.07 0.07 0.00 0.05 Large Firms Age (average) 24.36 30.44 24.47 23.54 25.15 ROA (average) 0.07 0.07 0.08 0.01 0.08 Age (median) 21 26 18 27 21 ROA (median) 0.05 0.06 0.06 0.01 0.07 Source: Mauritian Company Registrar and authors’ calculations. Notes: A “small” firm is defined as having MUR10 million in sales or less, a “medium” firm as having between MUR10 million and MUR80 million in sales, and a “large” firm as having more than MUR80 million in sales. “Age” is the age of the firm in years in 2014. Return on assets (ROA) is defined as the earnings before interest and taxes (EBIT) divided by total assets. The ROA is a normalized measure of profitability and facilitates a comparison of firms of different size. The ROA is winsorized at the 10 percent level. 243. Unlike larger firms, small businesses have negative the much larger absolute number of small firms. For profits. Indeed, slightly more than half of all small small firms, the pattern persists when new firms are firms in Mauritius have negative profits. In contrast, excluded from the analysis, suggesting that factors only 26 of medium-sized firms and 17 percent large other than age play a role in explaining the variation firms have negative profits. Figure 87 shows the in profitability. Moreover, differences across industries density plots of profitability for small firms (solid blue are limited. Small firms also have more outliers with line) and medium and large firms (dashed green line). large negative profits. For example, only 7 percent of The variation among small firms is much larger than medium and large firms have ROAs of -0.125 or less, among other firms, which can be partly explained by while more than 20 percent of small firms do. Photo : © Fernando LORENTE 114 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 8 - Evidence from Firm-Level Analysis Figure 87: Profitability density of firms (2007-12 average) 0,04 0,03 0,02 0,01 0,00 -0,500 -0,375 -0,250 -0,125 0,000 0,125 0,250 0,375 0,500 Return on Assets (ROA) Small Firms Medium & Large Firms Source: Mauritian Company Registrar and authors’ calculations. Notes: A “small” firm is defined as having MUR10 million in sales or less, a “medium” firm as having between MUR10 million and MUR80 million in sales, and a “large” firm as having more than MUR80 million in sales. Return on assets (ROA) is defined as the earnings before interest and taxes (EBIT) divided by total assets. The ROA is a normalized measure of profitability and facilitates a comparison of firms of different size. The ROA is winsorized at the 10 percent level; these observations are not shown. 244. SMEs are disadvantaged in Mauritius in terms of growth prospects; over longer periods, larger firms E. Financial structure and access to credit are more likely to grow faster. This is consistent with the increase in inequality of sales over the period. 245. Small firms are more leveraged and more risky in Profitability not only differs a lot among firms, but Mauritius. Access to finance by firms is captured by the also for specific firms over time. Firms’ age is not leverage ratio, or liabilities-to-assets ratio, defined strongly related to profitability, on average; however, as at the ratio of total liabilities to total assets.73 In a larger share of medium and large firms than small 2007-12, small firms and new incorporations are the firms are profitable. This holds in all industries except most highly leveraged, probably explained by these textiles, where average profitability is very low. Other firms’ strategy of financing expansion with debt (Table characteristics of firms, such as its financial structure, 11). Debt financing is largest for the average small may provide additional information on profitability. firm in the textiles industry. Leverage ratios generally decline with firm size—but it varies by industry. For example, debt financing for large firms in construction remains higher than other industries. Differences between average and median firms are largest for small firms, which likely have the largest variation in access to finance. For medium and large firms, this difference is relatively small. 73 Total liabilities are the sum of current liabilities and total non-cur- rent liabilities. Total assets are the sum of total current assets and total non-current assets. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 115 CHAPTER 8 - Evidence from Firm-Level Analysis Table 11: Firm size and financial structure by industry (2007-12 average) Construction Manufacturing Services Textiles Trade Small Firms Liabilit ies-to-Asset s (average) 1.01 1.01 0.97 1.07 1.01 Current Rat io (average) 2.22 2.39 2.43 2.08 2.57 Liab i lit ies-to-Asset s (median) 0.92 0.93 0.93 0.95 0.94 Current Rat io (media n) 1.18 1.17 1.21 1.21 1.39 Medium Firms Liab i lit ies-to-Asset s (aver age) 0.82 0.74 0.75 0.86 0.80 Current Rat io (average) 1.73 2.14 2.22 1.86 2.09 Liab i lit ies-to-Asset s (median) 0.80 0.71 0.72 0.76 0.78 Current Rat io (media n) 1.24 1.38 1.33 1.39 1.39 Large Firms Liab i lit ies-to-Asset s (aver age) 0.73 0.62 0.62 0.69 0.67 Current Rat io (average) 1.54 1.68 1.96 1.56 1.77 Liab i lit ies-to-Asset s (median) 0.77 0.61 0.65 0.67 0.67 Current Rat io (media n) 1.15 1.31 1.26 1.28 1.27 New Incorporations Liab i lit ies-to-Asset s (aver age) 0.96 1.00 0.97 1.02 0.94 Current Rat io (average) 2.34 2.43 2.49 2.24 2.71 Liab i lit ies-to-Asset s (median) 0.92 0.94 0.94 0.93 0.91 Current Rat io (media n) 1.24 1.13 1.22 1.21 1.40 Source: Mauritian Company Registrar and authors’ calculations. Notes: A “small” firm is defined as having MUR10 million in sales or less, a “medium” firm as having between MUR10 million and MUR80 million in sales, and a “large” firm as having more than MUR80 million in sales. The liabilities-to-assets ratio is defined as total liabilities divided by total assets. The current ratio is defined as total current assets divided by total current liabilities. The top 10 percent of observations are winsorized. Figure 88: Institutional framework for getting credit 6 100% 5 Public registry coverage 80% Depth of credit information (% of adults) 4 (0-6) 6 is maximum 60% 3 index 40% 2 20% 1 0 0% Depth of credit information index (0-6) Public registry coverage (% of adults) Source: Doing Business (World Bank) and authors’ calculations. Notes: More information is available on the website of Doing Business at http://www.doingbusiness.org/data/exploreeconomies/mauritius. “Depth of credit information index” measures rules and practices affecting the coverage, scope, and accessibility of credit information available through either a public credit registry or a private credit bureau; an index value of 6 is the maximum. “Public registry coverage” reports the number of individuals and firms listed in a public credit registry with information on their borrowing history over the past five years. 116 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 8 - Evidence from Firm-Level Analysis 246. In the past eight years, credit information increased data. At the same time, coverage of individuals and rapidly. Figure 88 provides background information on firms increased sevenfold. This improvement in the the institutional framework related to access to credit. institutional framework facilitates the provision of Currently, both positive and negative information is credit to firms, creating an environment that fosters available for individuals and firms—including retailers, business expansion and growth. utility companies, and financial institutions. The information is available for more than two years, while borrowers have the right to access their Figure 89: Access to financial services Mauritius Mauritius Botswana Botswana Kenya Kenya South Africa South Africa Tanzania Tanzania Rest of Rest of Sub-Saharan… Sub-Saharan… 0% 20% 40% 60% 80% 100% 0% 50% 100% Account: Adult Credit: Adult Account: Formal Businesses Credit: Formal Businesses Account: Informal Businesses Credit: Informal Businesses Source: Global Findex (World Bank) and authors’ calculations. Notes: More information is available on the Global Findex website at http:// www.worldbank.org/Globalfindex. “Account” measures the share of respondents who have an account at a bank or other type of formal financial institution. These are the adult population ages 15 and older, formal business owners, or informal business owners. “Credit” measures the share of respondents that borrowed from a formal financial institution in the past 12 months. All data are from 2011. For more information, see Demirguc-Kunt and Klapper (2013). 247. Access to financial services, including account 248. Despite improvements in information and access, ownership and obtaining credit from formal financial amount of credit has declined in the recent years. The institutions, is relatively broad in Mauritius, compared to number of firms that obtained loans from financial some other SSA countries (Figure 89). Reported account institutions steadily increased from 1999 to 2004 and ownership by businesses, both formal and informal, stabilized in the following years (Figure 90). In 2009 is high relative to the region and comparable to, for and 2010, the number of firms that obtained credit example, Kenya and South Africa. Access to credit declined relative to 2008. This pattern coincides from formal financial institutions is high for Mauritian with the start of the financial crisis at the end of adults but relatively low for formal businesses, 2008 and the subsequent drop in global demand in especially compared to Kenya. the following years. Recent years have seen a small recovery, although the number of firms that obtained new credit is roughly the same as it was in 1999. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 117 CHAPTER 8 - Evidence from Firm-Level Analysis 249. Credit has increased in trade and services and for trade. Over time, the share of manufacturing declined in manufacturing. Looking at the industry firms with new credit declined, while the share in composition of firms with new loans, we find it similar the trade industry increased. The share of new firms to that of all firms in the past five years or so.74 In receiving credit—those incorporated within two years both the services and trade industries, for example, of receiving one or multiple loans—is fairly stable over the share of firms with new credit averaged around time at around 20 percent. This number corresponds 42 percent. These numbers are similar to the share to the average share of new firms in the economy over in total firms—38 percent for services and 44 percent the past five years or so. 74 Information is only available for some of the firms that obtained credit in a year. For earlier years, information is available for more than 70 percent of the firms; this share declines to around 50 percent for the more recent years. Figure 90: New credit for firms over time 450 400 350 300 # of firms with new loan(s) 250 200 150 100 50 0 00 01 02 03 04 05 06 07 08 09 10 11 12 99 20 20 20 20 20 20 20 20 20 20 20 20 20 19 Source: Mauritian Company Registrar and authors’ calculations. Notes: Firms may obtain multiple loans within a year and from more than one financial institution. 118 MAURITIUS | Inclusiveness of Growth and Shared Prosperity CHAPTER 8 - Evidence from Firm-Level Analysis 250. The similarity in structures for all firms and firms that obtained credit, whether by industry F. Relationship between profitability and or new firms, suggest that access to new credit is widespread and not concentrated in particular firms. financial structure A similar picture emerges when looking at the stock 251. Although access to credit is good for business of liabilities. The average total amount of liabilities development, too much debt financing may create 2007-12 is around MUR17 billion. Around 83 percent considerable vulnerability, especially if borrowers resides with small firms, 13 percent with medium are unable to meet their short-term obligations. A firms, and the remainder with large firms. Moreover, measure of this short-term liquidity risk is the current 35 percent of liabilities reside with firms incorporated ratio, defined as the ratio of current assets to current in the period. These numbers roughly correspond to liabilities. the overall composition, confirming that access to credit is widespread. At the same time, it shows 252. The current ratio is especially high for small firms that a disproportionally large share of liabilities is and new incorporations, which on average have more concentrated in small and new firms, raising the issue than two times the coverage to meet short-term debt of vulnerability. payments (Table 11). The typical firm, regardless of size and sector, has a current ratio slightly above one. These lower ratios suggest that the typical firm is vulnerable to default in case of unexpected downturns in economic activity or increases in short- term interest rates. Table 12: Firm size, financial structure, and profitability (2007-12 average) Small Firms Highly leveraged firms (9,942) Normal leveraged firms (4,239) Profitable Unprofitable Profitable Unprofitable Sh ort -t erm liqui dity problem 13% 39% 6% 3% Sh ort -t erm liqui dity risk 13% 12% 13% 4% Ot hers 10% 14% 58% 16% Medium & Large Firms Highly leveraged firms (2,221) Normal leveraged firms (1,666) Profitable Unprofitable Profitable Unprofitable Sh ort -t erm liqui dity problem 19% 22% 10% 2% Sh ort -t erm liqui dity risk 36% 10% 27% 2% Ot hers 10% 4% 55% 4% Source: Mauritian Company Registrar and authors’ calculations. Notes: A “small” firm is defined as having MUR10 million Rs in sales or less, a “medium” firm as having between MUR10 million and MUR80 million in sales, and a “large” firm as having more than MUR80 million in sales. Highly leveraged firms have a liabilities-to-assets ratios above two-thirds. Firms with short-term liquidity problems have current ratios below one, those with a short-term liquidity risk are between one and two, and others above two. The liabilities-to-assets ratio is defined as total liabilities divided by total assets. The current ratio is defined as total current assets divided by total current liabilities. The top 10 percent of observations are winsorized. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 119 CHAPTER 8 - Evidence from Firm-Level Analysis 253. Around 70 percent of small firms and roughly 55 255. Compared to the services industry, firms are more likely percent of medium and large firms are highly leveraged to be unprofitable in agriculture and textiles. However, in Mauritius. Table 12 shows how size, financial this only applies to small firms for the agricultural structure, and profitability are related. First, we industry, while it holds primarily for medium and large make a distinction between firms that are highly firms in the textiles industry. In these sectors, ROA leveraged and those that are not. Highly leveraged is on average 4 to 8 percentage points lower than in firms are defined as those with liabilities-to-asset the services industry. Some evidence points to lower ratios greater than two-thirds. Second, we identify profitability in manufacturing and trade, buy the firms that have either short-term liquidity problems or differences are relatively small. short-term liquidity risks. Firms that have short-term liquidity problems have current ratios below one— 256. Small firms are more likely to be unprofitable than their current liabilities exceed their current assets. other firms. While medium-sized firms are also more Firms with current ratios between one and two have likely to be unprofitable than large firms, the extent short-term liquidity risks. Unexpected downturns in to which this is the case is small compared to small economic activity or increases in short-term interest firms. Some evidence suggests that profitability rates may leave these firms vulnerable to default on increases with age for small firms and decreases with their short-term obligations. Other firms have current age for medium and large firms, but the economic ratios above two. For profitability, a distinction is significance of these relationships is negligible. made between firms that are profitable and those Firms incorporated between 2007 and 2011 are also that are not, ignoring the extent of their profitability. less likely to be profitable and have a lower ROA. The table show that around 70 percent of small firms Although almost 90 percent of these firms are small, and roughly 55 percent of medium and large firms are the relationship to profitability is less strong than for highly leveraged. Among highly leveraged firms, only small firms in general. 10 percent are profitable with current ratios above two, regardless of the size of the firm. In contrast, 257. The strongest relationships are related to the financial firms with these profit and current ratio characteristics structure of the firms. Compared to firms with current make up more than half of all normally leveraged ratios of at least two, firms with short-term liquidity firms. For firms that are unprofitable and face short- problems—i.e., current liabilities exceed current term liquidity problems, a clear distinction emerges assets—are almost 20 percent less likely to be profitable between those that are highly leveraged and those and, on average, have ROAs that are 10 percentage that are not. Among normally leveraged firms, around points lower. These numbers are higher for firms that 2 percent to 3 percent have losses and short-term are highly leveraged—i.e., with liabilities-to-assets liquidity problems. When looking at highly leveraged ratios greater than two-thirds. Among small firms, the firms, these numbers rise to 39 percent for small firms highly leveraged are more than 30 percent less likely and 22 percent for medium and large firms. to be profitable than the normally leveraged. For medium and large firms, it is 20 percent. The ROA for 254. More formally, we use regression analysis to explore the average highly leveraged small firm is almost 15 the relationships between profitability and firm percentage points lower than its normally leveraged characteristics, including financial structure, age, counterpart; for medium and large firms, the gap is size, and industry or sector (Table A in the appendix). slightly less than 9 percentage points. Some evidence The analysis looks as both the extensive margin of suggests that a too high current ratio is associated profitability—i.e., whether a firm is profitable or with lower profitability, especially for small firms, but not—and the intensive margin, or return on assets. the economic significance is rather small. Note that the results only highlight relationships and cannot establish causality. 120 MAURITIUS | Inclusiveness of Growth and Shared Prosperity references and Appendix APPENDIX - REFERENCES REFERENCES Acemoglu, D. and D. Autor (2011), “Skills, Tasks and Technologies: Implications for Employment and Earnings”, in: Card D., and Ashenfelter, O. eds., Handbook of Labor Economics, Elsevier, Volume 4, Part B, pp. 1043-1171. Altonji J. G. and R. B. Blank (1999), “Race and Gender in the Labor Market,” in O. Ashefelter and D. Card (eds.), Handbook of Labor Economics, Vol. 3C, Amsterdam: North Holland. Arulampalam, W. (2001), “Is Unemployment Really Scarring? Effects of Unemployment Experiences on Wages”, Economic Journal, vol. 111(475), pp. 585-606. Autor D., L. Katz, and M. Kearney (2006), “The Polarization of the U.S. Labor Market,” American Economic Review P&P, vol. 96(2), pp. 119-124. Autor D., L. Katz, and M. Kearney (2008), “Trends in U.S. Wage Inequality: Revising the Revisionists,” The Review of Economics and Statistics, vol. 90(2), pp. 300-323. Becker, G. S. (1964), “Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education”, New York, NY: National Bureau of Economic Research. Bureau of Labor Statistics (2012), “Options for defining STEM (Science, Technology, Engineering, and Mathematics) occupations under the 2010 Standard Occupational Classification (SOC) system.” David, A. C. and Petri M. (2013), “Inclusive Growth and the Incidence of Fiscal Policy in Mauritius – Much Progress, but More Could be Done,” IMF Working Paper, WP/13/116. Demirguc-Kunt, Asli and Leora Klapper (2013). “Measuring Financial Inclusion: The Global Findex Database,” Brookings Papers on Economic Activity. Goos, M., I. Hataway, J. Konings, and M. Vandemeyer (2013), “High-Technology Employment in the European Union,” Vives Discussion Paper, 41. Hanushek, E. A., and L. Woessmann (2011), “The Economics of International Differences in Educational Achievement,” In E. A. Hanushek, S. Machin, and L. Woessmann (Eds.), Handbook of the Economics of Education, Vol. 3, pp. 89–200, Amsterdam: North Holland. Hanushek, E. A., and L. Woessmann (2012), “Do Better Schools Lead to More Growth? Cognitive Skills, Economic Outcomes, and Causation,” Journal of Economic Growth, vol. 17 (4), pp. 267–321. Holzer, H and R. LaLonde (2000), “Job Stability and Job Change Among Less Skilled Workers,” in D. Card and R. Blank, eds., Labor Markets and Less-Skilled Workers, New York: Russell Sage Foundation. Kahn, L. B. (2010), “The Long-Term Labor Market Consequences of Graduating from College in a Bad Economy,” Labour Economics, vol. 17(2), pp. 303-316. Jaumotte, F. (2003), “Labour Force Participation of Women: Empirical Evidence on the Role of Policy and other Determinants in OECD Countries,” OECD Economic Studies, vol. 37, pp. 51-108. 122 MAURITIUS | Inclusiveness of Growth and Shared Prosperity APPENDIX - REFERENCES McDonald, C. and J. Yao (2003), “Mauritius: Unemployment and the Role of Institutions,” IMF Working Paper, WP/03/211. Mincer, J. (1970), “The Distribution of Labor Incomes: A Survey with Special Reference to the Human Capital Approach”, Journal of Economic Literature, vol. 8(1), pp. 1–26. Mincer, J. (1974). Schooling, experience, and earnings. New York: NBER. Neumark, D. (2002), “Youth Labor Markets in the United States: Shopping Around vs. Staying Put,” Review of Economics and Statistics, vol. 84(3), pp. 462-482. Olivetti C. and B. Petrongolo (2008), “Unequal Pay or Unequal Employment? A Cross-Country Analysis of Gender Gaps,” Journal of Labor Economics, vol. 26, pp. 621-654. Pissarides, C., P. Garibaldi, C. Olivetti, B. Petrongolo, and E. Wasmer (2003), “Women in the Labour Force: How Well is Europe Doing?” Fondazione Rodolfo De Benedetti. Porter, N. (2004), “Wage Compensation, Employment Restrictions and Unemployment: the Case of Mauritius,” IMF Working Paper, WP/04/205. UNESCO (2010), “World Data on Education”, VII ed. 2010/11. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 123 APPENDIX - REFERENCES APPENDIX A. SOCIAL PROTECTION Appendix A1: Selected Social Assistance Programs in Mauritius Implementing Tar get Group Targ etin g Rs . U S$ % of N u mb er Benefits Agency Mech anism Million Million G DP Ser ved Pro v ided 2 008/ 09 2 008/ 09 2 008 Cash Transfers Poor and Rs 1,055/ Social A id MOSS Means Tested 371.4 11.9 0.15 44899 Indigent month 190,000 Elderly, N on-C ontri butory (as of Maximum of MOSS invalids, Universal 7729.6 248.6 3.08 Pensions August Rs. 2945 widows 2009) Poor and Electricity Rs 115/person/ Income Support MOSS 130 4.2 0.05 96000 Indigent consumption month Unemployment Hardship MOSS Unemployed Means Tested 1.9 0.1 0 372 Rs 324 /month R elief Bad W eather Allowance MOSS/ Fishermen Universal 60 1.9 0.02 N.A. Rs 200/day for Fis h er men MOAIFPS Vulnerable Rs 17,684/ N ational Solidarity Fu nd NSF/MOSS Means Tested 12.8 0.4 0.01 722 Families person Fam ilies in D istress Vulnerable MoWRCDFW Means Tested 0.4 0 0 10 Rs3,000-5000 Sch eme Families Accidents/ Prim e Minister’s R elief severe Rs 25,000 per PMO N.A. N.A. N.A. 0 N.A. and Support Fund hardship applicant as a victims one-off grant Sm all Planters Welfare MAIFPS Planters Universal 3 0.1 0 1000 Various benefits Fu nd Fis her men’s Welfare MAIFPS Fishermen Universal 3.5 0.1 0 60 Various benefits Fund Various A ssistances to Municipal Various N.A. N.A. N.A. N.A. N.A. N.A. Vulnerable Grou ps Governments In-Kind Assistance Sch ool Feeding Primary MOECHR Universal 52 1.7 0.02 119000 Loaf of bread Prog ramme Students Poor pre- Support for E AP Pre -pri mary S upport EAP/MOFEE school Means Tested N.A. N.A. N.A. 517 pre-schools and students enrolment Persons Cost of medical Overseas Medical Care MoHQL requiring Universal 32 1 0.01 665 care and travel medical care Poor T extbook Loan S cheme MOECHR secondary Means Tested 6.1 0.2 0 N.A. Textbooks students 124 MAURITIUS | Inclusiveness of Growth and Shared Prosperity APPENDIX - REFERENCES Rs3,000 - Grant Schem e for Tertiary Council/ Poor tertiary Rs8,000/month Means Tested 5.7 0.2 0 N.A. Tertiary Education MOECHR students plus Rs10,000 one-off Poor primary 11,000 Up to School Supplies TFSIVG/ MOFEE & secondary Means Tested 0.4 0.01 students Rs100,500 students Rs.60,000 Corrug ated Iron S heet 3,000 to TFSIVG/ MOFEE Indigent Means Tested N.A. N.A. N.A. for housing Housing date materials Casting roof Poor and slab; sites Social Ho using MHDC/MOHL Means Tested 537.4 17.3 0.2 5000 indigent and services; housing Students, Free Bus Subsidy MOPILTS elderly, Universal 792 25.5 0.32 N.A transportation disabled NEP/ Model Village - Low income I ntegrated Community Geographic 0.2 0 0 200 Housing families Develop ment MOFEE Poor pre- Pre -pri mary sc hool 517 in Pre-primary EAP/MOFEE primary Geographic N.A. N.A. N.A. Project 2009 expenses students Starter k its to poor MAIFPS Poor farmers Farm acreage N.A N.A 0 N.A farmers Parking Support for Persons MOSS Disabled Categorical N.A N.A 0 N.A coupons, Bus with Disabilities fare Francois S ock alingum Disabled Rs. 500 - Rs MOSS Categorical N.A N.A N.A. N.A Scholarship students 1,500 monthly Active-Labor Market Programs 1,107 (as Based on salary MOLIRE , MOSS, Redundant Workfare Self-targeted N.A. N.A. N.A. of August with minimum NEP(1) workers 2009) of Rs 3,000 National Trade Vulnerable Certificate Fou ndation IVTB Self-targeted 37.4 1.2 0.01 1025 Skills training youth Project National Trade Vulnerable Certificate Lev el 3 IVTB Self-targeted 58.3 1.9 0.02 Skills training youth 2,039 Course Vulnerable A pprenticeship S cheme IVTB Self-targeted 13.1 0.4 0.01 Skills training youth 778 Second Chance Vulnerable Remedial IVTB Self-targeted 1.5 0 0 Program m e youth 302 (2009) education Remedial Remedial and Vocational Vulnerable 200 to TFSIVG/MOFEE Self-targeted N.A. N.A. N.A. Education and Education youth date Training Empowerment 6,000 On-the-job Placement for Training Program/ Job seekers Self-targeted 36.8 1.2 0.01 since 2006 training MOFEE(1) MAURITIUS | Inclusiveness of Growth and Shared Prosperity 125 APPENDIX - REFERENCES Support to S m all and NEP/ Medium Enterprises and Entrepreneurs Self-targeted 99.1 3.2 0.04 N.A. Microcredit MOFEE(1) Booster Loans Microenterprise S u pport TFSIVG/MOFEE Entrepreneurs Self-targeted N.A. N.A. N.A. N.A. Microcredit Youth Entrepreneu rship MOY Youth Self-targeted N.A. N.A. N.A. N.A. Training Education and PARS MOY Youth Self-targeted N.A. N.A. 0 N.A. counseling Community Based Programs Poor Community Community D ev elop m ent EAP/MoFEE Geographic N.A. N.A. N.A. N.A. communities infrastructure Community Poor Community TFSIVG Geographic N.A. N.A. N.A. N.A. infrastructu re communities infrastructure Sugar Welfare Community Sug ar W elfare Centres Communities Geographic 114.6 3.7 0.05 N.A. Fund activities Community MOSS C ommu nity Centres MOSS Communities Geographic 8.7 0.27 0 N.A. activities Support to Civil Society Organisations NGO capacity N GO Trust Fund MOSS NGOS Not targeted 16 0.5 0.01 30 building D ecentralised MoFEE NGOs N.A. 190 6.1 0.08 N.A. Project support Prog ramme T FS I VG TFSIVG/MoFEE NGOs N.A. N.A. N.A. N.A. N.A. Project support EAP EAP/MoFEE NGOs N.A. N.A. N.A. N.A. N.A. Project support PM’s W omen ’s and PMO NGOs Not targeted 7.6 0.2 0 8 projects Project support Ch ildren’s R elief Fu nd Notes: NEP expenditure data for each of its subprograms include the proportional share of the administrative NEP budget, estimated at a total of Rs. 43.1 million in 2008/09. Expenditure figures are estimates because actual expenditures were not available for all years and all programs in Government’s Estimates of Expenditure. Some expenditure data was provided by program managers. Source: Government of Mauritius (2010). “Mauritius Social Protection Review and Strategy: Final Report,” March. 126 MAURITIUS | Inclusiveness of Growth and Shared Prosperity APPENDIX - REFERENCES APPENDIX B: GOVERNMENT EXPENDITURE ON SOCIAL PROTECTION Actual Government Expenditure on Social Protection*, 2013 Category Amount, MR % SP % SA % GDP All S P 20,265,917,537 100.0% - 5.5% SA ( S P w/o public service pensions) 13,873,820,603 68.5% 100.0% 3.8% BRP 11,230,636,616 55.4% 80.9% 3.1% Old age pension under BRP 8,027,384,985 39.6% 57.9% 2.2% Disabilit y benefits under BRP 1,161,511,066 5.7% 8.4% 0.3% Su rv iv or benefits under BRP 855,894,974 4.2% 6.2% 0.2% Fam ily and children benefits under BRP 254,575,623 1.3% 1.8% 0.1% Other benefits under BRP 931,269,968 4.6% 6.7% 0.3% Other S A ( non-BRP ) 2,643,183,988 13.0% 19.1% 0.7% Fam ily and children (non-BRP ) 260,842,092 1.3% 1.9% 0.1% Social exclusion (non-BRP ) 1,817,411,211 9.0% 13.1% 0.5% Other non-BR P 564,930,684 2.8% 4.1% 0.2% Notes:* Refers to expenditures classified as “Social Protection” in government accounts. Source: data provided by the Mauritius Accountant General's Office MAURITIUS | Inclusiveness of Growth and Shared Prosperity 127 APPENDIX - REFERENCES APPENDIX C: LABOR ANALYSIS C1. HIGH-TECH SECTOR In Chapter 8, we introduced the definition of high-tech sector. This classification exploits data on the sector of economic activity of the individual employer, collected in the CMHPS. The definition of sector of activity follows the NSIC standard, a national adaptation of the ISIC (International Standard of Industrial Classification of All Economic Activities) consisting of a coherent classifications of all economic activities based on a set of internationally recognized concepts and classification rules. This classification is subject to periodical updates to capture the cyclical transformations of world economy. For this reason, two different revision of the ISIC classification, revision 3.1 and the revision 4, corresponding to revision 1 and 2 respectively in the national adaptation, are implemented in the 12 CMPHS waves analyzed. Revision 1 is used between 2001 and 2010, while the newest revision 2 standard is adopted in the last two waves. For specific purposes, researchers often need to modify the aggregation provided by the ISIC structure to capture alternative concepts, such as the high-tech sector. Alternative but standardized aggregations have been created. In our analysis, we have exploited the OECD definition of high-tech industries. We have decided to adopt the ISIC revision 4 definition and adapt the earlier waves to the later classification by consulting the UN correspondence tables freely accessible on Internet. Even though we have tried to apply the utmost care, some discrepancies might still occur between the two classifications. TABLE C.1.1: HIGH-TECH INDUSTRIES OECD CLASSIFICATION NSIC Rev. 2 NSIC Rev. 1 Division Division High and Medium Technology Manufacturing High and Medium Technology Manufacturing 20 Chemicals and Chemical Products 24 Chemical and Chemical Products 21 Pharmaceutical Products 29 Machinery and Equipment n.e.c. 26 Computer, Electronic and Optical Products 30 Office Accounting and Computing Machinery 27 Electrical Equipment 31 Electrical Machinery and Apparatus n.e.c. Radio, Television and Communication Equipment 28 Machinery and Equipment n.e.c. 32 and Apparatus Medical Precision and Optical Instruments, 29 Motor Vehicles 33 Watches and Clocks 30 Other Transport Equipment 34 Motor Vehicles, Trailers and Semi-Trailers Railway and Tramway Locomotives and Rolling 352 Stock 353 Aircraft and Spacecraft 359 Transport Equipment n.e.c. Knowledge Intensive Services Knowledge Intensive Services 58-63 Information and Communication 64 Post and Telecommunications 64-66 Finance and Insurance 65-67 Financial Intermediation 69-75 Professional Scientific and Technical Activities 72 Computer and Related Activities 73 Research and Development 128 MAURITIUS | Inclusiveness of Growth and Shared Prosperity APPENDIX - REFERENCES C2. LABOR MARKET PARTICIPATION We have estimated a multiple regression model via Probit. The estimated equation is: P (yi=1|x) = f (ai +b1 female + b2 married + b3 kids + b5 age2 + b6 Rodrigues + b7 primaryedu + b8 secondaryedu + b9 tertiaryedu + Ei Where yi is a dummy variable assuming value 1 if the individual is inactive, female is a dummy variable for being female, and b1 is the coefficient of interest. The other variables included as controls are: • Married = 1 if individual is married 0 otherwise; • Kids = number of kids in the family; • Age = age of the individual; • Age2 = age squared; • Rodrigues = 1 if individual resides on the island of Rodrigues, 0 otherwise; • Primary_edu = 1 if individual’s highest educational level is primary education, 0 otherwise; • Secondary_edu = 1 if individual’s highest educational level is secondary education, 0 otherwise; • Tertiary_edu = 1 if individual’s highest educational level is tertiary education, 0 otherwise. The coefficients are the marginal effects at the mean for the covariate female in each of the 11 survey years and can be interpreted as the difference in probability of being inactive attributable to gender only.. Photo : © Haja Faniry Razafimahenina MAURITIUS | Inclusiveness of Growth and Shared Prosperity 129 APPENDIX - REFERENCES TABLE C.2.1: MARGINAL EFFECTS AT THE MEAN FOR INACTIVITY PROBABILITY 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2012 Female 0.350*** 0.329*** 0.303*** 0.310*** 0.293*** 0.285*** 0.290*** 0.279*** 0.269*** 0.259*** 0.256*** (0.007) (0.007) (0.006) (0.006) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) Married -0.063*** -0.111*** -0.084*** -0.102*** -0.143*** -0.154*** -0.174*** -0.177*** -0.184*** -0.184*** -0.194*** (0.009) (0.009) (0.008) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) N umber of 0.010*** 0.011*** 0.011*** 0.008*** 0.014*** 0.007*** 0.007*** 0.013*** 0.016*** 0.012*** 0.006*** Kids (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) Age 0.004*** 0.006*** 0.005*** 0.005*** 0.007*** 0.006*** 0.006*** 0.005*** 0.004*** 0.003*** 0.003*** (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) Ag e2 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) R odrig ues -0.007 0.007 -0.013 -0.006 0.002 -0.015 -0.020 -0.059*** -0.043*** -0.052*** -0.049*** (0.017) (0.017) (0.011) (0.010) (0.010) (0.010) (0.010) (0.010) (0.011) (0.011) (0.011) Primary 0.083*** 0.060*** 0.095*** 0.073*** 0.070*** 0.070*** 0.084*** 0.081*** 0.103*** 0.089*** 0.075*** (0.013) (0.013) (0.013) (0.011) (0.010) (0.010) (0.011) (0.011) (0.011) (0.011) (0.012) Secondary 0.192*** 0.163*** 0.190*** 0.192*** 0.199*** 0.217*** 0.225*** 0.225*** 0.252*** 0.238*** 0.222*** (0.013) (0.014) (0.013) (0.012) (0.010) (0.011) (0.011) (0.011) (0.011) (0.012) (0.012) Above 0.270*** 0.239*** 0.239*** 0.134*** 0.100*** 0.107*** 0.143*** 0.170*** 0.196*** 0.186*** 0.162*** Secondary (0.016) (0.016) (0.016) (0.015) (0.013) (0.013) (0.013) (0.013) (0.013) (0.013) (0.014) N 22,204 22,240 21,746 29,257 37,196 37,072 37,511 36,714 35,639 35,715 36,006 Note: * p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses. Data for 2011 missing. Labor discusses the results of a linear probability model for the probability of falling into the NEET group given a series of covariates. The model estimated takes the form: Where yi is a dummy variable assuming value 1 if the individual is a NEET are two dummy variables indicating the respective years interacted with the following controls: Pr (yi=1|xi) = 2003 * f (b1 + b2 siblings + b3 fatheredu + b4 motheredu + b5 fatherlog (wage) + b6 mother active + b7 female + b8 Rodrigues ) + 2012 * (b9 + b10 siblings + b11 fatheredu + b12 motheredu + b13 fatherlog (wage) + b14 mother active + b15 female + b16 Rodrigues ) + Ei • Siblings = number of siblings; • Father_edu = highest educational level for the individual’s father; • Mother_edu = highest educational level for the individual’s mother; • Father_log(wage) = individual’s father log of monthly wage; • Mother_active= 1 if individual’s mother participates in the labor market, 0 otherwise; • Female = 1 if individual is a woman 0 otherwise; • Rodrigues = 1 if individual resides on the island of Rodrigues, 0 otherwise. 130 MAURITIUS | Inclusiveness of Growth and Shared Prosperity APPENDIX - REFERENCES TABLE C.2.2. PROBABILITY OF NEET, AGES 15-24. 2003 2012 Year Constant 0.416*** 0.411*** (0.064) (0.044) interacted with: S iblings 0.023*** 0.005 (0.004) (0.003) Father_edu -0.026* -0.048*** (0.012) (0.008) Mother_edu -0.034** -0.012 (0.011) (0.008) Father _log(w) -0.018** -0.008* (0.006) (0.004) Mother_empl -0.023 -0.035** (0.019) (0.011) Fem ale 0.104*** 0.039*** (0.014) (0.009) R odrigues 0.047 -0.037 (0.025) (0.021) r2 0.174 N 8248 Note: Standard errors robust to heteroskedasticity in parentheses. */**/*** for significance levels at 10%, 5% and 1% respectively. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 131 TABLE C.2.3. LINEAR PROBABILITY MODEL - INACTIVITY PROBABILITY 132 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2012 Female 0.406*** 0.378*** 0.369*** 0.359*** 0.343*** 0.331*** 0.332*** 0.314*** 0.307*** 0.293*** 0.284*** (0.007) (0.007) (0.007) (0.006) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) Married 0.087*** 0.073*** 0.059*** 0.053*** 0.015* 0.011 0.007 -0.008 -0.015* -0.006 -0.005 (0.009) (0.009) (0.009) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) No kids 0.004 0.002 0.008** 0.001 0.011*** 0.001 -0.000 0.008*** 0.011*** 0.006** 0.001 (0.002) (0.002) (0.003) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.001) Age -0.052*** -0.054*** -0.054*** -0.051*** -0.052*** -0.052*** -0.054*** -0.054*** -0.056*** -0.057*** -0.054*** APPENDIX - REFERENCES (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) Age2 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Rodrigues -0.062*** -0.056** -0.088*** -0.066*** -0.041*** -0.065*** -0.073*** -0.120*** -0.107*** -0.117*** -0.103*** (0.018) (0.018) (0.012) (0.010) (0.010) (0.011) (0.010) (0.011) (0.011) (0.011) (0.011) MAURITIUS | Inclusiveness of Growth and Shared Prosperity Primary -0.056*** -0.093*** -0.036* -0.071*** -0.081*** -0.061*** -0.049*** -0.041** -0.028 -0.035* -0.019 (0.015) (0.016) (0.017) (0.014) (0.013) (0.013) (0.013) (0.013) (0.014) (0.015) (0.014) Secondary -0.074*** -0.136*** -0.057** -0.096*** -0.093*** -0.066*** -0.073*** -0.067*** -0.053*** -0.063*** -0.061*** (0.016) (0.016) (0.018) (0.015) (0.013) (0.014) (0.014) (0.014) (0.015) (0.015) (0.015) Above_sec -0.073*** -0.137*** -0.084*** -0.195*** -0.226*** -0.208*** -0.195*** -0.169*** -0.158*** -0.169*** -0.176*** (0.019) (0.018) (0.020) (0.017) (0.015) (0.015) (0.015) (0.015) (0.016) (0.016) (0.016) Constant 1.135*** 1.259*** 1.151*** 1.196*** 1.206*** 1.231*** 1.289*** 1.269*** 1.300*** 1.340*** 1.316*** (0.032) (0.032) (0.033) (0.028) (0.025) (0.025) (0.024) (0.025) (0.026) (0.026) (0.024) r2 0.321 0.311 0.295 0.297 0.301 0.293 0.299 0.295 0.298 0.289 0.280 N 16193 16345 15639 21527 27039 27334 27980 27457 26740 27036 27954 Note: * p<0.05, ** p<0.01, *** p<0.001. Robust standard errors in parentheses. Data for 2011 missing. APPENDIX - REFERENCES C.3.1 WAGE REGRESSION In section 2.2, we have estimated a wage regression. The estimated equation is: Where is the monthly wage for the individual i, is a dummy variable equal 1 if the individual’s highest degree is for primary log (wi ) = ai + bn xi + bn+1 primaryedu + bn+2 secondaryedu + bn+3 tertiaryedu + Ei school, is a dummy variable equal 1 if the individual’s highest degree is for secondary school, is a dummy variable equal 1 if the individual’s highest degree is for post-secondary school. , and are the coefficient of interest indicates a vector of control variables including: • Married = 1 if individual is married 0 otherwise; • Kids = number of kids in the family; • Age = age of the individual; • Age2 = age squared ; • Rodrigues = 1 if individual resides on the island of Rodrigues, 0 otherwise; • Construction = 1 if the individual works in the construction sector, 0 otherwise; • Trade & Trans = 1 if the individual works in the trade and transport sector, 0 otherwise; • Tourism = 1 if the individual works in the tourist sector, 0 otherwise; • Manufacturing = 1 if the individual works in the manufacturing sector, 0 otherwise; • Finance = 1 if the individual works in the financial sector, 0 otherwise; • Real Estate = 1 if the individual works in the real estate sector, 0 otherwise; • Public Service= 1 if the individual works in the public sector, 0 otherwise; • IT & Com = 1 if the individual works in the IT and communication sector, 0 otherwise (only for 2012); • Prof. Service = if the individual works in the professional service sector, 0 otherwise (only for 2012); • Other = 1 if the individual works in the residuals sectors, 0 otherwise; • Female = 1 if the individual is a female, 0 otherwise. C3.2 SCHOOLING EQUATION In Table 6, we discussed the results of a schooling equation obtained from an estimated linear probability model for the probability of accessing one of the four increasing educational levels given a set of covariates. The exact specification of the model is the following: Where s = 1,…,4 indicates one of the four possible educational categories (no education, primary education, secondary Pr (si =1|xi ) = ai + bn xi + Ei education, post-secondary education) that we have created. i indicates individual i and a vector of control variables including: • Female = 1 if individual is a woman 0 otherwise; • Siblings = number of siblings; • Age = age of the individual; • Rodrigues = 1 if individual resides on the island of Rodrigues, 0 otherwise; • Father’s education = father’s educational category; • Mother’s education = mother’s educational category; • Father’s income quartile 2/4 = 1 if the individual father’s work income falls in the second, third or fourth quartile of the distribution, respectively, 0 otherwise: • Father (mother) unemployed = 1 if the individual’s father (mother) is unemployed, 0 otherwise; • Father (mother) employed = 1 if the individual’s father (mother) is employed, 0 otherwise; • Father (mother) inactive = 1 if the individual’s father (mother) is inactive, 0 otherwise. Table 6 displays the full specification of this regression for each of the 10 years considered. It only displays a subset of covariates for six selected years. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 133 APPENDIX - REFERENCES C3.3 INTERGENERATIONAL MOBILITY In section 2.3 Figure 75, we have displayed and discussed the evolution of family background and its impact on offspring schooling achievement. The graph shows a plot of the R2 of a linear probability model for each of the four educational categories we used and for each of the 10 years for which we have data. The specification of the model is: Pr (si =1|xsi ) = asi + bsn xsi + Esi Where s = 1,…,3 indicates one of the three possible degrees (primary education, secondary education, post-secondary education) that we have created. i indicates individual i and a vector of control variables including: • Father’s education = father’s educational category; • Mother’s education = mother’s educational category; • Father’s income quartile 2/4 = 1 if the individual father’s work income falls in the second, third or fourth quartile of the distribution, respectively, 0 otherwise. C3.4 FEMALES GRADUATION PROBABILITIES In section 0 we have discussed Mauritian’s women educational achievements and compared it to those of men. These graduation probabilities are obtained via a linear probability model taking the form: Pr (si =1|xsi ) = asi + b1 female + bs1 + n xsi + Esi Where s = 1,…,4 indicates one of the four possible educational categories (no education, primary education, secondary education, post-secondary education) that we have created. i indicates individual i, female is a dummy variable for being female and the associated slope parameter, b1, is the coefficient of interest. xi a vector of control variables including: • Siblings = number of siblings; • Age = age of the individual; • Rodrigues = 1 if individual resides on the island of Rodrigues, 0 otherwise; • Father’s education = father’s educational category; • Mother’s education = mother’s educational category; • Father’s income quartile 2/4 = 1 if the individual father’s work income falls in the second, third or fourth quartile of the distribution, respectively, 0 otherwise; • Father’s (mother’s) unemployed = 1 if the individual’s father (mother) is unemployed, 0 otherwise; • Father’s (mother’s) employed = 1 if the individual’s father (mother) is employed, 0 otherwise; • Father’s (mother’s) inactive = 1 if the individual’s father (mother) is inactive, 0 otherwise. 134 MAURITIUS | Inclusiveness of Growth and Shared Prosperity Table C3.1: Wage regression 2001 2002 2003 2 004 2 005 2 006 2 007 2 008 2 009 2 010 2 012 P ri m ary 0.200*** 0.145** 0.188*** 0.256*** 0.162*** 0.234*** 0.203*** 0.159*** 0.086* 0.103 0.278*** (0.046) (0.045) (0.046) (0.045) (0.036) (0.041) (0.040) (0.041) (0.042) (0.159) (0.049) S econdary 0.517*** 0.466*** 0.503*** 0.574*** 0.484*** 0.580*** 0.551*** 0.495*** 0.424*** 0.464** 0.644*** (0.047) (0.046) (0.047) (0.045) (0.037) (0.041) (0.040) (0.041) (0.042) (0.159) (0.050) P ost -secondary 1.079*** 1.029*** 1.093*** 1.212*** 1.123*** 1.212*** 1.149*** 1.143*** 1.128*** 1.241*** 1.340*** (0.052) (0.049) (0.051) (0.048) (0.040) (0.044) (0.043) (0.044) (0.045) (0.162) (0.052) Age 0.065*** 0.061*** 0.067*** 0.067*** 0.068*** 0.070*** 0.062*** 0.063*** 0.070*** 0.172*** 0.069*** (0.004) (0.004) (0.004) (0.004) (0.004) (0.003) (0.004) (0.004) (0.004) (0.012) (0.003) Ag e2 -0.001*** -0.001*** -0.001*** -0.001*** -0.001*** -0.001*** -0.001*** -0.001*** -0.001*** -0.002*** -0.001*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) R odri gu es -0.441*** -0.328*** -0.335*** -0.271*** -0.278*** -0.299*** -0.383*** -0.264*** -0.293*** -1.550*** -0.368*** (0.059) (0.050) (0.031) (0.026) (0.030) (0.030) (0.031) (0.030) (0.032) (0.102) (0.032) Married 0.156*** 0.151*** 0.182*** 0.135*** 0.153*** 0.132*** 0.137*** 0.156*** 0.165*** 0.103* 0.130*** (0.017) (0.017) (0.017) (0.014) (0.014) (0.013) (0.014) (0.013) (0.014) (0.044) (0.014) N of Kids -0.012* -0.006 -0.004 -0.003 -0.006 -0.004 0.005 0.005 0.009 -0.037* -0.001 (0.005) (0.005) (0.005) (0.004) (0.004) (0.004) (0.004) (0.004) (0.005) (0.015) (0.003) C onstru ction 0.250*** 0.211*** 0.306*** 0.300*** 0.260*** 0.298*** 0.225*** 0.330*** 0.301*** 0.786*** 0.285*** (0.027) (0.026) (0.026) (0.023) (0.021) (0.021) (0.022) (0.021) (0.022) (0.062) (0.028) T rade & T rans 0.321*** 0.315*** 0.371*** 0.384*** 0.384*** 0.430*** 0.374*** 0.438*** 0.435*** 0.651*** 0.382*** (0.024) (0.024) (0.024) (0.021) (0.020) (0.020) (0.021) (0.020) (0.021) (0.066) (0.027) MAURITIUS | Inclusiveness of Growth and Shared Prosperity APPENDIX - REFERENCES T o u ris m 0.390*** 0.313*** 0.434*** 0.410*** 0.405*** 0.455*** 0.378*** 0.465*** 0.485*** 0.585*** 0.416*** (0.034) (0.032) (0.031) (0.026) (0.025) (0.025) (0.025) (0.024) (0.025) (0.086) (0.030) 135 136 Man u fact u rin g 0.224*** 0.201*** 0.274*** 0.242*** 0.284*** 0.344*** 0.204*** 0.312*** 0.283*** 0.783*** 0.348*** (0.022) (0.022) (0.023) (0.020) (0.019) (0.019) (0.020) (0.020) (0.020) (0.060) (0.027) F inance 0.583*** 0.632*** 0.673*** 0.676*** 0.692*** 0.720*** 0.669*** 0.790*** 0.857*** 1.512*** 0.767*** (0.058) (0.054) (0.054) (0.038) (0.039) (0.038) (0.039) (0.037) (0.038) (0.081) (0.040) R eal Estate 0.371*** 0.389*** 0.402*** 0.404*** 0.427*** 0.484*** 0.404*** 0.466*** 0.490*** 1.141*** 0.742*** (0.041) (0.039) (0.039) (0.033) (0.030) (0.028) (0.030) (0.028) (0.029) (0.070) (0.097) P u blic S er v ices 0.434*** 0.400*** 0.494*** 0.562*** 0.558*** 0.563*** 0.512*** 0.621*** 0.671*** 1.382*** 0.625*** (0.026) (0.025) (0.026) (0.022) (0.021) (0.021) (0.021) (0.022) (0.022) (0.059) (0.028) APPENDIX - REFERENCES Ot h er 0.286*** 0.287*** 0.287*** 0.292*** 0.306*** 0.326*** 0.306*** 0.354*** 0.398*** 0.895*** -0.018 (0.036) (0.039) (0.042) (0.033) (0.030) (0.032) (0.032) (0.029) (0.031) (0.085) (0.034) I T & Com . 0.668*** (0.043) P rof S er v ices 0.401*** MAURITIUS | Inclusiveness of Growth and Shared Prosperity (0.032) F em ale -0.478*** -0.478*** -0.548*** -0.507*** -0.529*** -0.517*** -0.510*** -0.534*** -0.545*** -1.000*** -0.525*** (0.015) (0.014) (0.015) (0.012) (0.012) (0.011) (0.012) (0.011) (0.012) (0.038) (0.012) C onstant 6.480*** 6.665*** 6.522*** 6.432*** 6.605*** 6.474*** 6.714*** 6.768*** 6.759*** 4.476*** 6.786*** (0.089) (0.088) (0.088) (0.091) (0.078) (0.075) (0.077) (0.077) (0.079) (0.279) (0.080) r2 0.444 0.453 0.471 0.492 0.449 0.461 0.439 0.464 0.480 0.239 0.461 N 8,267 8,475 8,560 11,349 13,953 14,128 14,230 14,246 13,917 14,831 14,305 * p<0.05, ** p<0.01, *** p<0.001. Robust standard errors in parentheses. Data for 2011 missing. TABLE C3.2: SCHOOLING EQUATION 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2012 Female 0.094*** 0.075*** 0.118*** 0.116*** 0.065*** 0.098*** 0.117*** 0.110*** 0.100*** 0.136*** 0.131*** (0.017) (0.018) (0.017) (0.014) (0.013) (0.013) (0.013) (0.013) (0.013) (0.013) (0.014) Rodrigues -0.176*** -0.177*** -0.161*** -0.171*** -0.094*** -0.129*** -0.134*** -0.108*** -0.137*** -0.096** -0.187*** (0.044) (0.045) (0.030) (0.026) (0.026) (0.028) (0.028) (0.028) (0.029) (0.030) (0.030) Siblings 0.052*** 0.026*** 0.045*** 0.051*** 0.055*** 0.060*** 0.069*** 0.083*** 0.062*** 0.059*** 0.045*** (0.008) (0.007) (0.008) (0.007) (0.006) (0.006) (0.006) (0.007) (0.007) (0.007) (0.006) Age individual 0.094*** 0.088*** 0.099*** 0.094*** 0.096*** 0.093*** 0.093*** 0.090*** 0.091*** 0.088*** 0.080*** (0.002) (0.002) (0.002) (0.001) (0.002) (0.001) (0.001) (0.002) (0.002) (0.002) (0.002) Father’s Edu. 0.060*** 0.065*** 0.103*** 0.058*** 0.051*** 0.081*** 0.089*** 0.076*** 0.073*** 0.098*** 0.092*** (0.014) (0.014) (0.014) (0.012) (0.010) (0.011) (0.011) (0.011) (0.012) (0.012) (0.011) Mother’s Edu. 0.091*** 0.070*** 0.074*** 0.095*** 0.084*** 0.098*** 0.110*** 0.143*** 0.101*** 0.113*** 0.080*** (0.013) (0.014) (0.014) (0.012) (0.011) (0.011) (0.011) (0.012) (0.012) (0.011) (0.012) Father’s Income Quartile: Second 0.081*** 0.065** 0.106*** 0.069*** 0.156*** 0.104*** 0.070*** 0.056** 0.054** 0.041* 0.074*** (0.023) (0.023) (0.022) (0.020) (0.018) (0.018) (0.018) (0.018) (0.019) (0.018) (0.019) Third 0.192*** 0.126*** 0.130*** 0.142*** 0.202*** 0.125*** 0.076*** 0.075*** 0.061*** 0.058** 0.089*** (0.024) (0.023) (0.023) (0.020) (0.019) (0.019) (0.017) (0.019) (0.018) (0.018) (0.019) Fourth 0.271*** 0.207*** 0.223*** 0.228*** 0.306*** 0.195*** 0.129*** 0.157*** 0.182*** 0.139*** 0.129*** (0.026) (0.026) (0.025) (0.022) (0.022) (0.022) (0.020) (0.021) (0.021) (0.021) (0.022) Father’s Labor Force Status: Unemployed 0.572*** 0.427*** 0.510*** 0.400*** 0.467*** 0.345*** 0.521*** 0.392*** 0.600*** 0.363*** 0.368*** (0.066) (0.070) (0.089) (0.076) (0.062) (0.055) (0.064) (0.067) (0.060) (0.067) (0.054) Employed 0.502*** 0.428*** 0.506*** 0.474*** 0.390*** 0.458*** 0.517*** 0.492*** 0.508*** 0.393*** 0.322*** (0.042) (0.040) (0.043) (0.035) (0.035) (0.035) (0.033) (0.034) (0.036) (0.035) (0.030) Mother’s Labor Force Status: Unemployed 0.059 0.016 0.083* 0.081* 0.100*** 0.024 -0.006 0.037 0.048 0.036 0.006 (0.041) (0.039) (0.043) (0.034) (0.023) (0.027) (0.025) (0.027) (0.027) (0.026) (0.033) Employed 0.153*** 0.135*** 0.152*** 0.149*** 0.152*** 0.131*** 0.126*** 0.105*** 0.091*** 0.081*** 0.104*** (0.017) (0.017) (0.016) (0.014) (0.013) (0.013) (0.013) (0.013) (0.013) (0.013) (0.013) Constant 0.705*** 1.028*** 0.570*** 0.720*** 0.801*** 0.692*** 0.612*** 0.598*** 0.799*** 0.841*** 1.210*** MAURITIUS | Inclusiveness of Growth and Shared Prosperity APPENDIX - REFERENCES (0.084) (0.094) (0.073) (0.063) (0.070) (0.066) (0.059) (0.070) (0.073) (0.073) (0.069) r2 0.499 0.458 0.533 0.520 0.532 0.505 0.507 0.489 0.491 0.489 0.445 N 9,100 9,074 8,909 11,770 14,938 14,782 14,864 14,631 14,056 14,067 13,866 * p<0.05, ** p<0.01, *** p<0.001 137 TABLE C3.3: DETERMINANTS OF SCHOOLING—FAMILY BACKGROUND 138 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2012 Primary Education Father Edu. -0.033*** -0.022** -0.046*** -0.027*** -0.030*** -0.025*** -0.022*** -0.040*** -0.024*** -0.021*** -0.032*** (0.008) (0.008) (0.008) (0.007) (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) Mother Edu. -0.008 -0.007 -0.020* -0.023*** -0.025*** -0.022*** -0.023*** -0.019** -0.024*** -0.013* 0.000 (0.007) (0.007) (0.008) (0.007) (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) Second 0.012 -0.004 0.005 -0.020 0.003 0.013 0.000 0.025* 0.022* 0.061*** 0.035** (0.014) (0.014) (0.014) (0.012) (0.011) (0.011) (0.011) (0.011) (0.011) (0.011) (0.011) Third -0.004 -0.028 0.002 -0.005 -0.017 -0.012 -0.021* 0.004 0.008 0.008 0.007 (0.015) (0.014) (0.015) (0.012) (0.011) (0.011) (0.010) (0.011) (0.011) (0.010) (0.010) Fourth -0.072*** -0.085*** -0.027 -0.067*** -0.044*** -0.042*** -0.061*** -0.033** -0.035** -0.039*** -0.026* (0.015) (0.015) (0.015) (0.013) (0.012) (0.012) (0.011) (0.011) (0.011) (0.012) (0.011) Constant 0.526*** 0.489*** 0.592*** 0.533*** 0.536*** 0.496*** 0.485*** 0.515*** 0.468*** 0.401*** 0.387*** (0.026) (0.026) (0.027) (0.023) (0.021) (0.021) (0.022) (0.022) (0.022) (0.023) (0.022) R2 0.012 0.010 0.013 0.012 0.011 0.009 0.010 0.012 0.009 0.010 0.006 APPENDIX - REFERENCES N 9100 9074 8909 11770 14938 14782 14864 14631 14056 14072 13866 Secondary Education Father Edu. -0.014 -0.027*** -0.021** -0.043*** -0.032*** -0.022*** -0.030*** -0.034*** -0.031*** -0.043*** -0.028*** (0.008) (0.008) (0.008) (0.007) (0.006) (0.007) (0.006) (0.007) (0.007) (0.007) (0.007) Mother Edu. -0.058*** -0.058*** -0.042*** -0.035*** -0.047*** -0.050*** -0.041*** -0.033*** -0.051*** -0.044*** -0.059*** (0.007) (0.007) (0.008) (0.007) (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) (0.007) Second 0.006 0.007 0.018 0.028* 0.034** 0.015 0.015 0.001 0.007 -0.033** 0.008 (0.013) (0.014) (0.014) (0.013) (0.011) (0.011) (0.012) (0.011) (0.012) (0.011) (0.012) Third 0.030* 0.028* 0.025 0.036** 0.064*** 0.050*** 0.029** 0.003 0.016 0.014 0.013 (0.014) (0.014) (0.014) (0.012) (0.011) (0.011) (0.011) (0.012) (0.011) (0.012) (0.012) MAURITIUS | Inclusiveness of Growth and Shared Prosperity Fourth 0.072*** 0.050** 0.053*** 0.104*** 0.116*** 0.070*** 0.061*** 0.042** 0.054*** 0.043** 0.022 (0.015) (0.016) (0.015) (0.014) (0.013) (0.013) (0.013) (0.013) (0.013) (0.013) (0.013) Constant 0.564*** 0.640*** 0.542*** 0.618*** 0.615*** 0.632*** 0.651*** 0.641*** 0.699*** 0.742*** 0.750*** (0.025) (0.026) (0.026) (0.023) (0.021) (0.022) (0.022) (0.023) (0.023) (0.024) (0.024) R2 0.012 0.014 0.008 0.011 0.011 0.009 0.008 0.007 0.011 0.012 0.013 N 9100 9074 8909 11770 14938 14782 14864 14631 14056 14072 13866 Post-Secondary Education Father Edu. 0.012* 0.021*** 0.027*** 0.031*** 0.027*** 0.021*** 0.031*** 0.043*** 0.036*** 0.045*** 0.052*** (0.005) (0.005) (0.005) (0.004) (0.004) (0.004) (0.004) (0.004) (0.004) (0.005) (0.005) Mother Edu. -0.000 -0.002 -0.002 -0.009* -0.002 0.002 0.009* 0.011** 0.016*** 0.013** 0.016*** (0.004) (0.005) (0.005) (0.004) (0.003) (0.003) (0.003) (0.003) (0.004) (0.004) (0.005) Second -0.025** -0.032*** -0.027*** -0.024*** -0.047*** -0.048*** -0.040*** -0.068*** -0.057*** -0.082*** -0.071*** (0.008) (0.008) (0.008) (0.006) (0.006) (0.006) (0.006) (0.006) (0.007) (0.007) (0.008) Third 0.010 -0.012 0.003 -0.002 -0.025*** -0.026*** -0.005 -0.032*** -0.033*** -0.038*** -0.035*** (0.009) (0.009) (0.009) (0.007) (0.007) (0.007) (0.007) (0.008) (0.007) (0.008) (0.009) Fourth 0.093*** 0.086*** 0.089*** 0.074*** 0.044*** 0.063*** 0.062*** 0.050*** 0.060*** 0.045*** 0.024* (0.011) (0.012) (0.011) (0.010) (0.009) (0.009) (0.009) (0.010) (0.010) (0.011) (0.011) Constant 0.050*** 0.038* 0.008 0.005 0.012 0.014 -0.039** -0.065*** -0.052*** -0.045** -0.061*** (0.015) (0.017) (0.016) (0.013) (0.012) (0.013) (0.013) (0.013) (0.014) (0.016) (0.017) R2 0.023 0.025 0.028 0.025 0.021 0.026 0.027 0.039 0.035 0.036 0.028 N 9100 9074 8909 11770 14938 14782 14864 14631 14056 14072 13866 * p<0.05, ** p<0.01, *** p<0.001 APPENDIX - REFERENCES APPENDIX D. OAXACA-BLINDER DECOMPOSITION Given two groups of individuals—group a and group b—an outcome variable Y, and a set of predictors x, the Oaxaca-Blinder decomposition decomposes variation in Y between group A and B to a part explained by the set of predictors and a residual part that is unexplained. Formally, the question is establishing how much of the group difference is accounted for by the group difference in the predictors: R = E(Ya ) - E(Yb ) Where E(Y) denotes the expected value of the outcome variable, based on the linear model: Yi = x' bi + ei with E(ei ) = 0 and l Î (a;b) X is a vector containing the regressors and a constant, b contains the slope parameters and e is the error term. The mean outcome difference can be expressed as the difference in the linear prediction at the group specific means of the regressors: If we want to identify the contribution of group differences in predictors to overall outcome differences, the previous R = E(Ya ) - E(Yb ) = E(Xa )’ ba - E(Xb )’ bb equation can be rearranged as: Where b* is the nondiscriminatory coefficient vector. This formulation can be thought of as being the sum of two components: Is the part of the outcome differential that is explainable by group differences in the predictors, while the second term of R = {E(Xa ) - E(Xb )}’ b* + {E(Xa )’ (ba - b* ) + E(Xb )’ (b* - bb ) } the equation: Is the unexplained part that in labor economic literature on group differential is usually attributed to discrimination. An Q = {E(Xa ) - E(Xb )}’ b* important disclaimer is that the term U captures also all the potential effects that unobservable characteristics play in explaining wage differentials and for this reason estimates of discrimination based on the Oaxaca-Blinder decomposition U = {E(Xa )’ (ba - b* ) + E(Xb )’ (b* - bb ) } should be taken with caution. MAURITIUS | Inclusiveness of Growth and Shared Prosperity 139 APPENDIX - REFERENCES APPENDIX E. FIRM LEVEL REGRESSION ANALYSIS Table E1: Regression Results for Firm Profitability (2007-2012 average) (1) (2) (3) (4) (5) (6) Dependent Variables Profitable firm Return On Assets Small Medium Small Medium All firms firms & Large All firms firms & Large Firms Firms Agricu ltu re / Extractive -0.120*** -0.139*** -0.045 -0.053*** -0.059*** -0.028* (0.030) (0.036) (0.055) (0.013) (0.016) (0.016) C onstru ction 0.005 -0.007 0.043* 0.003 0.000 0.016* (0.013) (0.015) (0.024) (0.006) (0.007) (0.008) Man u facturing -0.034*** -0.040** -0.009 -0.017*** -0.019*** -0.005 (0.013) (0.016) (0.021) (0.005) (0.006) (0.007) Textiles -0.129*** -0.098*** -0.202*** -0.054*** -0.046*** -0.073*** (0.021) (0.024) (0.040) (0.008) (0.010) (0.013) Trade -0.023*** -0.030*** -0.000 -0.012*** -0.013*** -0.009** (0.007) (0.008) (0.013) (0.003) (0.003) (0.005) Short-term liq uidit y proble m -0.198*** -0.183*** -0.256*** -0.100*** -0.098*** -0.103*** (0.008) (0.009) (0.017) (0.003) (0.004) (0.005) Short-term liq uidit y risk 0.058*** 0.066*** 0.017 0.010*** 0.018*** -0.012*** (0.008) (0.010) (0.013) (0.003) (0.004) (0.004) Small firm -0.221*** -0.079*** (0.012) (0.004) Mediu m Fir m -0.035*** -0.062*** -0.002 -0.016*** (0.013) (0.013) (0.004) (0.004) Age in 20 14 -0.000 0.001** -0.001*** -0.001*** 0.000 -0.001*** (0.000) (0.001) (0.000) (0.000) (0.000) (0.000) New Incorporations -0.088*** -0.079*** -0.071*** -0.037*** -0.036*** -0.010* (0.008) (0.009) (0.017) (0.003) (0.004) (0.006) Hig h ly lev eraged firm -0.287*** -0.320*** -0.190*** -0.130*** -0.145*** -0.086*** (0.008) (0.009) (0.013) (0.003) (0.003) (0.004) C onstant 1.020*** 0.795*** 1.020*** 0.192*** 0.112*** 0.187*** (0.015) (0.012) (0.019) (0.005) (0.005) (0.006) Ob servations 21,243 16,383 4,860 21,243 16,383 4,860 R-squared 0.223 0.181 0.180 0.262 0.227 0.238 Source: Mauritian Company Registrar and authors’ calculations. Notes: OLS regressions. Coefficient is significantly different from 0 at the *** .01, **  .05, and * .10 level. Robust standard errors are in parentheses. The omitted categories are (large) firms that are normally leveraged, have current ratios above two, and were incorporated before 2007, and are in the services industry. Profitable firms have positive earnings before interest and taxes (EBIT), and the return on assets (ROA) is defined as EBIT divided by total assets. Columns (1)-(3) deal with the extensive margin of profitability and estimate a linear probability model. Columns (4)-(6) deal with the intensive margin of profitability. A “small” firm is defined as having MUR10 million in sales or less, a “medium” firm as having between MUR10 million and MUR80 million in sales, and a “large” firm as having more than MUR80 million in sales. Highly leveraged firms have liabilities-to-asset ratios above two-thirds. Firms with short-term liquidity problems have current ratios below one, those with short-term liquidity risk have one between one and two, and others have one above two. The liabilities-to-assets ratio is defined as total liabilities divided by total assets. The current ratio is defined as total current assets divided by total current liabilities. New incorporations are firms that got incorporated between 2007 and 2011. 140 MAURITIUS | Inclusiveness of Growth and Shared Prosperity Photo : © Haja Faniry Razafimahenina Designed by World Bank / AFREC Under the supervision of Erick Rabemananoro Layout and graphic design RAKOTOMANANA Andriantoavina Photo : © Sandra Salerno Mauritius is a high middle-income country with low levels become a high-income country will depend on its ability to of poverty and inequality. The headcount poverty level was improve the labor force’s skill set, develop infrastructure, 6.9 percent in 2012; measured by the international standard and further improve the business environment to attract FDI of US$2 per day (PPP), poverty was less than 1 percent. On and generate domestic investment. Reduction in inequality inequality, Mauritius also fared well compared to its peer and boost of shared prosperity will require more growth middle-income countries. On the negative side, Mauritius’ and a more pro-poor pattern of growth. An increase in growth has not been equally shared, despite the general female labor force participation, reduction of high youth improvement in welfare. The economy’s polarization was unemployment rates, improving the efficiency of the social associated with a structural transformation from labor- protection system will reduce growing skills mismatch intensive industries to services and knowledge-intensive facilitating inclusive growth and eradicating poverty in industries. Inclusiveness remains the main challenge for Mauritius. the current growth pattern. When Mauritius will be able to