Key Findings

One of the aims of this project is to produce a detailed empirical description of labor market outcomes in low-income economies based on individual-level data from a number of Sub-Saharan African (SSA) countries. We focus on job creation in the private and public sector, labour earnings, and workers' mobility between them. We also look at the differences in some of these outcomes between men and women.

Below we document our main findings with some illustrative graphs.

Job Creation in Sub-Saharan Africa is Low

Most of the working population in least developed countries are working as own-account workers or unpaid family members. We refer to it as “subsistence sector” as most of these activities are low-productivity and low-pay, typically involving farming or petty retail. The subsistence sector comprises more than 80% of employment in most Sub-Saharan African countries (with a notable exception of South Africa and a few other countries).

While public sector employs only a small fraction of the whole workforce in the region, it comprises a considerable share of paid employees. Out of the workers working for a wage, about 25% on average are employed by the public sector. For some countries in our sample - such as Burkina Faso, Ethiopia, Nigeria, and Zambia -- this fraction exceeds 40%. For comparison, according to the ONS statistics, the current share of the public sector in the UK is about 17% of total employment.

Note: Subsistence sector consists of own account workers and unpaid family members. For South Africa, the subsistence sector also includes unemployed workers actively searching for a job. Source: Authors’ computations using household survey data.

Women tend work in subsistence, or in the public sector if working for a wage

For all countries in our sample, women are over-represented in the subsistence sector. This cannot be explained solely by differences in educational attainment between men and women: we find that even after controlling for education, age, marital status, and regions, women are less likely to be working for a wage than men.

This result can stem from women's choices as working in self-employment, family-owned farms or businesses may make it easier to combine work with household responsibilities (e.g. looking after children or elderly family members). On the other hand, it may also reflect differences in formal and informal institutions, including the lack of childcare support and societal norms.

Despite the fact that women are less likely to be working in a wage than men, once they are in employment they typically are more likely to be working in the public sector. These results hold in the regression controlling for education, age, marital status, and regions, with statistically significant differences between men and women for all countries with the exception of Benin, Guinea Bissau, Mali, and Uganda. In other countries in our sample female employees are more likely work in the public sector than their male counterparts with similar characteristics by at least 12% in South Africa and up to 48% in Malawi.

Compare to other parts of the world, Women in SSA work less in the public sector

With extra data from Jobs of the World Database (JWD), we compare women's participation in the public sector to other regions of the world. The figure shows that the share of public sector employment is much higher among women compared to men virtually everywhere in the world, with the exception of SSA.

The upper figure presents the global sectoral distribution of the workforce separately for men and women. The most visible gender segregation can be seen in the share of the public sector. It shows that women are increasingly more likely to work in the public sector as the level of GDP per capita increases.

However, the pattern does not hold in SSA. Women face persistently worse labour market conditions than men across the region, being more likely to work in self-employment and less likely to select into the public or private sector compared to men.

The lower figure further highlights the puzzle, showing SSA countries do not follow the global trend that public sector employees do not increase with the urbanisation rate.

Public Sector Hires More eDucated workers and Pays higher wages

Public sector hires predominately more educated workers – the majority of public sector employees have secondary education and above. Worker composition by educational attainment between the private sector employees and the self-employed is more similar – a vast majority of those employed in the subsistence sector and in private companies have primary education or less in four out of the six countries in our sample (see Education graphs below).

The public sector in least developed countries pays significantly higher wages than the private sector. For all countries in our sample the wage distribution in the public sector lies to the right of the private sector distribution and is typically more compressed, suggesting a lower degree of wage inequality in public sector jobs.

High Public sector premium, even After controlling for Worker Demographics

Note: Wage premium is expressed as the difference in log wages between the two sectors. For example, the wage premium of 0.5 is equal to exp(0.5)=1.649 or 64.9% higher wages in the public sector than in the private sector. The explanatory variables included in the regression are worker's gender, age, age squared, education level, rural or urban area of residence, as well as year, industry and region fixed effects. Source: Authors’ computations using household survey data.

Graph designed and embedded on the website by Sidharth Rony.

Some of the difference between public and private sector wages can be explained by differences in education levels of their workers. The graph below accounts for these differences in worker composition across the two sectors by running a linear regression of log labor earnings on observable worker characteristics, industry dummies and regions. We plot the public sector wage premium (expressed as the difference in log wages between the two sectors) in the raw data against the regression coefficient on public sector dummy variable (with 95% confidence interval).

Most of the regression coefficients are below the 45% degree line, showing that a significant share of the raw gap in wages can be explained by worker characteristics (e.g. education levels). However, with the exception of Ethiopia and Burkina Faso (where the coefficient is not statistically different from zero), most countries exhibit a positive wage premium for public sector employees compared to private firms, even if worker characteristics are similar.

We also find that public sector wage premium is much higher for women than for men.

Public sector Has Lower Gender Gap than private firms

We estimate female log-wage coefficients from country- and sector-specific wage regressions that control for educational attainment, age, marital status, industry, and location of residence. One feature that stands out is that the gender wage gaps are significantly larger (by absolute value) in the private sector than in the public sector. The figure shows that the orange points (female coefficients in the private sector) generally lie below the green points (female coefficients in the public sector).

Public sector offers higher job Security than private firms

Using the panel data from six SSA countries (Ethiopia, Niger, Nigeria, South Africa, Tanzania, and Uganda) we can study the mobility of workers between sectors. We find that the job separation rate into the subsistence sector is roughly twice as large in the private sector than in the public sector, suggesting that public sector jobs are more stable.

Average 2-year transition rates between sectors in Uganda

Note: SE stands for self-employment, P for the private sector, and G for the public sector. The transitions numbers show the average fraction of workers in a given sector in year t that report working in sector SE, P or G in year t + 2. The sample includes workers between 15 and 64 years old who are surveyed in at least two panel waves. Source: Authors’ calculations based on the Uganda National Panel Surveys 2010-2019.

Page updated

Google Sites

Report abuse