Employment and Urban Poverty Urban Poverty: Lessons from Experience June 5, 2007 Pierella Paci 1 Why focusing on employment? Because: Growth is important for poverty reduction but it is NOT sufficient; Pro-poor growth requires a reduction in income inequality as well as an increase in average income; Inequalities in labor income are an important source of existing inequality and an important determinants of poverty; Thus employment is important transmission channels between growth and poverty reduction. 2 1
Evidence Pro-Poor Growth in the 1990s highlights employment as a crucial link between growth and poverty reduction and identifies amongst the priority areas for future research: o Labor market regulations and links between investment climate and employment opportunities; o Differential access of vulnerable groups to formal/informal employment. The forthcoming Moving out of Poverty report identifies access to productive employment as a major way out of poverty. Emerging evidence from Employment and Shared Growth program. 3 Employment and urban poverty The link between barriers to employment and poverty incidence is not surprising as labor is often the only asset of the poor providing jobs and increasing the efficiency of labor exchange is a crucial aspect of poverty reduction. The role of employment generation in poverty reduction is particularly evident in urban areas where access to safety nets, crop sharing and other informal survival mechanisms are limited. 4 2
Urban poverty and employment status Employment status Poverty Incidence Bangladesh Ghana 2005 2005 all urban rurall urban unemployed 15% 15% inactive 21% 12% employed 27% 9% 42% 27% of which paid employee 9% 6% self-employed non agric with employees 6% 1% self-employed non-agric without employees 15% 1% unpaid family workers non agric 24% 9% self-employed agric 32% 18% unpaid fam workers agric 49% 27% under-employed 46% 32% 5 But generating employment may not be enough Having a job is not sufficient to guarantee adequate living standards Over 500 million people are estimated to be working poor (18% of the those employed) and this number is not declining (outside China and India). What counts is not employment per se but the labor income derived from that employment; need not just jobs but good jobs. 6 3
Large earning differentials exist Employment status Median Earnings Bangladesh Madagascar 2005 large urban second ary urban rurall urban rurall employed 64.1 100.0 of which waged (no agr) 77.8 100.0 89.5 non-waged (no agric) 31.3 32.9 31.9 Agriculture 37.6 68.5 58.4 7 Main policy question 1? Higher returns in existing jobs? More Jobs? If job creation is not sufficient to ensure poverty reduction in LICs: should effort be targeted to increase the returns in available employment? 8 4
Additional challenge 1: urban population grows fast For a given increase in jobs available in urban areas, the corresponding increase in employment opportunities depends on Δ in working age urban population due to demographics and urbanization: e.g., if 10% in number of jobs but also 10% in urban labor force employment opportunities stay the same. In Bangladesh between 2000 and 2005: Share of urban population 5% points; 75% of income growth was in urban areas. In Ghana between 1999 and 2005: Share of urban population 5% points urban population 3 % points higher that rural population (4% v/s 1%) 9 Additional challenge 2: Beyond the single labor market The labor market does NOT exists as a single entity; It comprises a number of different segments offering qualitatively distinct types of employment to workers with similar endowments; All workers seek employment in the good segments but good jobs are rationed Not everyone gets access the good jobs Workers with similar endowments have different 10 earnings depending where they work. 5
And beyond the dual labor market E(Wu)>E(Wr) Decision to migrate Urban LM: Segmented Rural LM: Subsistence Agriculture Wr Bad Job Sector Wb=Wu+min Eb residual Good Job Sector Wg=Wc Eg determined labor demand 11 Implication of multi-segmented model When more than two sectors are allowed for Need to analyze: o The functioning of each labor market segment (wage setting mechanism) o The link between the different segments; The distribution of new jobs (across urban/rural, sectors and segments) matters to the poverty impact of growth The poverty impact of a given job increase depends also on: The distribution of the new jobs (geographical, sectoral) compared to the sectoral distribution of income/poverty; The ability of the poor to move to the sectors where the 12 new jobs are. 6
Main policy question 2 Low-paying sectors Better-paying sectors If the patterns of job generation matters: should efforts be focused on where the poor are or on where the poor are not, so that more of the poor can be drawn into the higher-earning parts of the economy? 13 Theoretical framework Two basic concepts: Structural change (Chenery and Syrquin): The economy consists of a number of different sectors and economic growth is to a large extent driven by the relative size and productivity of these sectors Creative destruction (Schumpeter): birth and death of firms and jobs is a natural process and a certain amount of churning is needed to generate economic growth Labor mobility play a key role The structure of the economy, labor institutions and regulations affect labor mobility 14 7
Zooming on some of the findings so far In the short run: overall employment intensity of growth does not matter for poverty reduction (and especially for reduction of urban poverty), but the sectoral pattern of employment growth and the relative impact on productivity and labor intensity is important employment in manufacturing urban poverty; employment shares in agriculture poverty in both employment and productivity in tertiary sector urban poverty (less than growth in 15 manufacturing) Check our website for more detailed results (including country studies) www.worldbank\employment 16 8