Contact person
Dr Neil Rankin (neil.rankin{ A T }wits.ac.za)
Description
This project, funded by the
International Development Research Centre, Canada, brings together evidence from four African countries:
Ghana,
Madagascar,
South Africa, and
Tanzania to investigate two key questions:
1) How and where are jobs created that can be accessed by the poor?
2) What types of jobs create trajectories out of poverty, and what is constraining the growth of these jobs?
Collaborating institutions
African Micro-Economic Research Umbrella, University of the Witwatersrand;
University of Dar es Salaam, Tanzania;
Ghana Statistical Services;
National Institute of Statistics Madagascar (INSTAT);
Centre for the Study of African Economies, University of Oxford;
Lafayette College;
Cornel University;
University of Legon, Ghana;
University of Cape Town.
Research Outputs
Ghana
1) Manufacturing Firms in Ghana: Comparing the 1987 and 2003 Censuses - Anthony Krakah, Moses Awoonor-Williams, Nicholas Nsowah-Nuamah and Francis Teal. Presented at the 2009 CSAE Oxford conference. The conference paper can be found here.
This paper presents a comparison of the 1987 and 2003 censuses of manufacturing firms in Ghana. The study shows that the number of industrial firms increased from 8,000 in 1987 to 26,000 in 2003. However, the increases were predominantly amongst small-sized firms which more than tripled, and medium-sized firms that doubled. Large firms remained about the same in number but firms employing 500 persons and more actually contracted from 52 to 40.
With regards to wage levels in the manufacturing sector, the findings from the two censuses indicate that wages in large firms more than doubled for all categories of workers between 1987 and 2003. Average wage per employee per month in large firms rose from $53 in 1987 to $139 in 2003. It is also observed that wages and salaries of administrative, clerical and accounts staff are twice as much as what the production workers (up to the foreman level) receive. However, in spite of the remarkable increases in the number of medium-sized firms over the reference period, average wage per employee generally increased from $38 in 1987 to only $50 in 2003 (i.e. 32%). The increases among these size class firms also varied across employment categories as production workers benefited by 13.0% compared to 33.0% amongst the administrative, clerical and accounts staff.
The paper discusses the implications of these changes in the size distribution of firms and the structure of wages.
2) Jobs, Skills and Incomes in Ghana: How was poverty halved? - Moses Awoonor-Williams, Nicholas Nsowah-Nuamah and Francis Teal. Presented at the 2009 CSAE Oxford conference. Published as a CSAE working paper.
Poverty has halved in Ghana over the period from 1991 to 2005. Our objective in this paper is to assess how far this fall was linked to the creation of better paying jobs and the increase in education. We find that earnings rose rapidly in the period from 1998 to 2005, by 64% for men and by 55% for women. While education, particularly at the post secondary level, is associated with far higher earnings there is no evidence that the increase in earnings that occurred over the period from1998 to 2005 is due to increased returns to education or increased levels of education. In contrast there is very strong evidence, for all levels of education, that the probability of having a public sector job approximately halved over the period from 1991 while the probability of having a job in a small firm increased very substantially. In 1991/92 a male worker with secondary education had a 7 per cent probability of being employed in a small firm, by 2005/06 this had increased to 20 per cent which was higher than the probability of being employed by the public sector. Employment in small firms, which is the low paying occupation within the urban sector, increased from 2.7 to 6.7 percent of the population, an increase from 225,000 to 886,000 employees. Jobs in total have been increasing in line with the population but the proportion of relatively low paying ones increased markedly from 1998/99 to 2005/06. The rises in income that occurred over this period were due almost entirely to increases in earnings rates, for given levels of education, across all job types particularly among the unskilled. Why unskilled earnings rates rose so rapidly is unclear.
3) Does doing an apprenticeship pay off? Evidence from Ghana - Courtney Monk, Justin Sandefur and Francis Teal. Published as a CSAE working paper.
In Ghana there is a highly developed apprenticeship system where young men and women undertake sector-specific private training, which yields skills used primarily in the informal sector. In this paper we use a 2006 urban based household survey with detailed questions on the background, training and earnings of workers in both wage and self-employment to ask whether apprenticeship pays off. We show that apprenticeship is by far the most important institution providing training and is undertaken primarily by those with junior high school or lower levels of education. The summary statistics indicate that those who have done an apprenticeship earn much less than those who have not. This suggests that endogenous selection into the apprenticeship system is important, and we take several measures to address this issue. We find a significant amount of heterogeneity in the returns to apprenticeship across education. Our most conservative estimates imply that for currently employed people, who did apprenticeships but have no formal education, the training increases their earnings by 50%. However this declines as education levels rise. We argue that our results are consistent with those who enter apprenticeship with no education having higher ability than those who enter with more education.
Madagascar
1) The Demand for Hired Labor in Rural Madagascar - Jean Claude Randrianarisoa, Christopher B. Barrett and David Stifel. Presented at the 2009 CSAE Oxford conference. A version of this paper can be found here.
This paper estimates structural labor demand equations separately for farm and non-farm enterprises in rural Madagascar. It adapts recent labor supply estimation methods that address the general unobservability of both wage rates – due to widespread self-employment – and employers’ non-wage costs of hiring workers in order to fill a significant void in the existing literature. Labor demand in rural Madagascar appears strongly increasing in enterprise owners’ educational attainment, in enterprises’ capital stock, and in community-level public goods. Furthermore, labor demand appears wage inelastic, especially in the non-farm sector where government labor market policies, such as minimum wage laws, are more commonly enforced.
2) Firm diagnostics and constraints to growth of formal sector labor demand -
Harivelo Rajemison, Faly Hery Rakotomanana, Patrick Randriankolona, Rachel Ravelosoa, Tiaray Razafimanantena and David Stifel. The final version of this paper is available here.
This study analyses the determinants of the labor demand formal sector enterprises in Madagascar. An important result is that demand for labor in the private formal sector in Madagascar is inelastic. This result consistently emerges for separate estimates for SQS’s and EI’s whether it is for total labor demand or demand for only hired labor. Further, it is consistent with the results of Randrianarisoa et al. (2009) who, using a different data source (a household survey), found that hired labor demand is inelastic among informal enterprises in rural Madagascar. This suggests that concerns about the employment-reducing effects of wage growth appear to be unfounded in these data
South Africa
1) Determinants of Job Matching in South Africa - Neil Rankin, Gareth Roberts and Volker Schöer. Presented at the 2008 DPRU conference and at the 2009 CSAE Oxford conference. The conference paper can be found here.
Job creation is a key policy objective in South Africa. But, whilst job creation would allow various unemployed workers to enter employment, the actual allocation of vacant jobs among the unemployed is determined by the matching process of job seekers and recruiting firms through their search strategies and recruitment methods respectively.
Empirical evidence from various South African surveys suggests that most matching is done via social networks (referrals) which implies that job seekers without access to such recruitment networks are systematically disadvantaged (Wittenberg, 2001; Schöer & Leibbrandt, 2006). However it is not clear what factors determine this equilibrium outcome between the search methods of the unemployed and the recruitment methods of the firms with the data available on the South African labour market.
Using a unique data set of unemployed and employed youth who also report on characteristics of the firm they work for, this paper investigates the matching process and its impact on the allocation of jobs among the job seekers.
Our findings suggest that trying to understand labour market outcomes using only supply side characteristics can yield misleading results. Given the sample, individual characteristics only explain the difference between workers who got employed through formal channels compared to other channels. However, employees who got employed through either networks or informal channels exhibit the same individual characteristics. In their case, firm characteristics are crucial in the matching process. Job seekers are more likely to get employed through networks by small firms. Large firms and their wider range of recruitment methods allow for a greater range of unemployed job seekers to be included in the matching process. Furthermore, firm size seems to proxy for a range of unobservables including the employer’s ability to carry the cost of recruitment screening and training, impact of labour legislation on the recruitment process as well as the visibility of the firm to the job seeker and therefore the arrival rate of job applicants.
2) Export destination, product quality and wages in a middle-income country. The case of South Africa - Neil Rankin and Volker Schöer. Presented at the 2008 DPRU conference and at the 2009 CSAE Oxford conference. The final version of the paper can be found here.
A robust finding in the firm-level literature is that exporting firms pay higher wages. However, in many cases these results are at the firm-level and do not control for worker specific characteristics. In this paper we use data from South Africa and four other SADC countries to expand on the growing body of research that uses matched employer-employee data to investigate the relationship between exporting and wages at a worker level. South Africa, a middle-income country, has two distinct main export markets – a regional market in SADC where per capita incomes are lower than at home and an international market, predominantly the United States and Western Europe, where per capita incomes are higher than at home. We exploit this distinction to investigate whether export destination impacts on wages. Our estimates show that workers in South African firms that export to the region earn less than those that produce for the domestic market. Those in firms that export outside the region earn more than either domestic producers or SADC-only exporters. These results support previous theoretical and empirical work which suggests that export destination is related to product quality which in turn is related to worker quality and therefore wages.
3) Youth unemployment, firm size and reservation wages in South Africa – Neil Rankin and Gareth Roberts. Presented at the 2009 CSAE Oxford conference. The paper can be found here.
Youth unemployment in South Africa is high. We investigate whether one of the reasons may be that the wages young people want or need are above those that they could reasonably expect to earn given their characteristics. Unlike previous work on the relationship between reservation wages and unemployment we differentiate between wages in different sizes of firms. Larger firms pay more and thus, even if reservation wages are similar to average predicted wages, they may be above wages that young people could expect to earn in smaller firms. We find that this is the case.
Tanzania
1) Private Sector Development and Income Dynamics: A Panel Data Study of the Tanzanian Labour Market - Simon Quinn and Francis Teal. Published as a CSAE working paper.
In this paper, we use a three-period panel of Tanzanian households to explore the determinants of earnings and earnings growth from 2004 to 2006. In doing so, we draw particular attention to the role of education and to the importance of heterogeneity between more and less formal occupations. Several important conclusions emerge. Education is found to have a significant convex effect upon earnings levels, but to have had no significant effect upon earnings growth (indeed, there is some suggestion that education may have had a negative impact). This suggests that recent Tanzanian growth may have reflected an ‘unskill-biased technological change’, providing relative reward to informal skills rather than to formal education. Further, there are interesting insights into the age-earnings relationship: the relationship is found significantly to be concave in levels, yet age is not found significantly to have affected earnings growth. This suggests that the concave levels relationship is driven by workers’ participation decisions, rather than by a concave earnings trajectory at the level of the individual worker. Finally, we find significant evidence of variation between formal and informal enterprises, and between sizes of enterprises within these different employment sectors.
2) Does Higher Education Influence Productivity of Graduates? Evidences from Tanzanian Manufacturing Enterprises – Godius Kahyarara. The paper can be found here.
This paper analyses the effects of higher education on worker’s performance in the labour market. It does so by estimating the firm level productivity effect of higher education in Tanzania manufacturing enterprises. The estimates are obtained from a Cobb-Douglas production function that treats higher education variable as one of the determinants of productivity. The analysis drawn from such estimate contributes knowledge to two empirical questions; 1) are there gains from investment in higher education to the employers? 2) What is the influence of productivity in explaining higher wages to graduates? Data used is employer-employee matched data from Tanzania Enterprises from 1994 till 2008. A time dimension of the data allows controls for time invariant individual characteristics that are potential sources omitted variable bias in micro analysis like the one provided in this paper. The findings strongly support the hypothesis of positive correlation between higher education and worker productivity. The OLS coefficient estimate on the proportion of higher education is 0.027 suggesting that 1 percent increase in the proportion of the workforce with higher education increases observed productivity by 0.27 percent. But GMM coefficient estimates are much lower at 0.0009 and statistically significant showing that our OLS estimates were biased. The results also confirm a positive correlation between productivity and a graduate manager. Using such results, the paper concludes that there is robust evidence of productivity effect of higher education in the data. Hence, employers gain from utilizing higher education graduates through increased firm performance especially productivity. There is a justification for employers to support higher education.
3) Estimates of the Size and Determinants of Formal Informal Wage Gap in Tanzanian labor market - Godius Kahyarara. The paper can be found here.
This paper measures the size and determinants of the wage gap between formal and informal employment in the Tanzanian labour market. Specifically it estimates a microeconomic model of wage gap that provide the actual size of the gap, then it analyses the determinants of the gap using rich information of individual and establishment level information to assess the determinants of the gap. The findings of the paper indicate the formal-informal real wage gap at between 24 to 28 percent. The variations of the gap are induced by time dimension, human capital characteristics of education, experience, training and other job characteristics. Such correlation is robust even when fixed effects and omitted variable bias are controlled for via instrumental variable and fixed effects models. However, the estimates that control for fixed effects indicate that the OLS estimates are between 20-30 percent higher suggesting the significant effects of firm fixed effects on the observed wage gap. The paper concludes that policy interventions to reduce wage gap should take the form of equalizing levels of human capital in both sectors especially more training and education.
4) Job creation in the Tanzanian labour market: What jobs are being created, where? - Godius Kahyarara. The paper can be found here.
This paper assesses the process of job creation in Tanzania by addressing the question: what types of jobs are being created and where? To address these questions the paper’s conceptual framework and strategies consider the interrelation between the number of jobs that can be created in a labor market segment, the quality of jobs in terms of human capacity requirement, productivity and remuneration. The specific method for capturing job creation used by the paper are estimates of the labour market, indexes of job creation and job destruction at sectoral level. The paper findings are that job creation in Tanzania is mostly happening in traditional agricultural sector and urban informal sector as both employ over 90 percent of the working population. The absorption capacity in the formal sector jobs is very small and there are all indications that it will remain so even in the near future. Findings also suggest that the fastest growing sectors such as mining and tourism have low levels of employment. Tourism has the highest total employment effect in the economy than any other sector. Specifically if we consider the results in the row corresponding to tourism it is indicated that 1% rise in output of tourism sector will require increase in direct employment of 0.25%. The forward and backward linkage for this sector seems to be the highest. The results further show that agriculture has less employment multiplier effect although it provides the largest direct employment opportunity. Mining sector employment has low multiplier effects.
Collaborative papers
1) The returns to formality and informality in urban Africa? - Paolo Falco, Andrew Kerr, Neil Rankin, Justin Sandefur and Francis Teal. Published as a CSAE working paper.
This paper addresses the question as to why we observe such large differentials in earnings in urban African labour markets after controlling for observable human capital. We first use a three year panel across Ghana and Tanzania and find common patterns for both countries assuming that movement between occupations is exogenous. Unobserved individual market ability is by far the most important factor explaining the variance of earnings. Sector differences do matter even with controls for ability and the sectoral gap between private wage employment and civil servants is about 50 per cent, once we control for unobserved time-invariant factors. Wage earners earn the same as the self-employed in both Ghana and Tanzania. An additional important aspect of formality is enterprise size. At most half of the OLS effect of size on earnings can be explained by unobservable ability. Workers in largest firms are the high earners with wage rates which exceed those of civil servants. We then use an extension of the Ghana panel to five years to assess the extent of possible biases from the assumption of exogenous movement. We find evidence that this is important and that OLS may be understating the extent of both the size effect and the private sector wage (negative) premium. The implications of our results for understanding the nature of formal and informal employment in Africa are discussed.
2) Learning and earning in Africa. Where are the returns to education high? - Neil Rankin, Justin Sandefur and Francis Teal. Published as a CSAE working paper.
This paper investigates the role of learning - through formal schooling and time spent in the labor market - in explaining labor market outcomes of urban workers in Ghana and Tanzania. We investigate these issues using a new data set measuring incomes of both formal sector wage workers and the self-employed in the informal sector. In both countries we find significant, convex returns to education and large earnings differentials between sectors when we pool the data and do not control for selection. In Ghana there is a particularly steep age-earnings profile. We investigate how far a Harris-Todaro model of market segmentation or a Roy model of selection can explain the patterns observed in the data. We find highly significant differences across occupations and important effects from selection in both countries. The data is consistent with a pattern by which higher ability individuals queue for the high wage formal sector jobs such that the age earnings profile is convex for the self-employed in Ghana once we control for selection. The returns to education are far higher in the large firm sector than in others and in this sector they are linear not convex. In both countries there is clear evidence of convexity in the returns to education for the self-employed and here the average returns are low.
Other Outputs, Conferences and Workshops
Many of the papers from this project were presented at the CSAE's Annual Conference in March 2009.
The papers from this project will be presented at a workshop in Johannesburg during October 2009. More details are available on the AMERU events page.
Ghana
Policy workshop - 17th September 2008, Accra, Ghana. Hosted by the Ghana Statistical Service and the University of Oxford.
Madagascar
Policy workshop - 19th-20th June 2008, Antananarivo, Madagascar. co-hosted by INSTAT and the World Bank.
Policy workshop - 20th January 2009, Antananarivo, Madagascar. co-hosted by INSTAT and the World Bank.
South Africa
Research and policy workshop "Firm level data analysis"- 19th-20th February 2008, Johannesburg, South Africa, co-hosted by Economic Research Southern Africa.
Two Masters Scholarships for 2009 - Arabo Ewinyu and Bongani Daka
Tanzania
Policy brief - Academic versus Vocational Training: Maximising Tanzania's Educational System - Godius Kahyarara, University of Dar es Salaam
Policy workshop - "Economic development, labour markets and education" - 3rd December 2008, Dar es Salaam, Tanzania.
Policy workshop - "Improving labour markets in Africa" - 22nd July 2009, Dar es Salaam, Tanzania. A report can be found here.