This page summarizes findings from several earlier sets of country-level data analyses that considered the average earnings and/or the poverty risk of different segments of the labour force, both formal and informal.
Data Sources
1. The first set is the analysis of national data in five countries — Egypt, El Salvador, India, Russia and South Africa — commissioned by the Economic Policy Institute-Global Policy Network (EPI-GPN) for a comparative workforce development project funded by the Ford Foundation. Following a common framework of questions, these analyses all studied the links between macroeconomic processes and labour force development (though they varied in the measures used). Most importantly, for our purposes here, they also disaggregated the labour force by formal and informal employment, men and women.
All of the data for the EPI-GPN studies were from the late 1990s and early 2000s. The data sources and years for the three countries were as follows:
- Egypt: Egyptian Labor Market Survey (ELMS 1998) and Labor Force Sample Survey (LFSS) 1990 and 2001
- El Salvador: Multi-Purpose Household Survey 1991-2002
- South Africa: various sources before 2000, October Household Survey 2000, and Labor Force Survey 2000-3
2. A second relevant analysis is a compilation of data for 14 countries by Jacques Charmes. For all countries, Charmes compares data on the average monthly income of micro-entrepreneurs (i.e., informal employers who hire others) and the average monthly wage of employees of micro-enterprises, both expressed as multiples of the legal minimum wage level in those countries.
The 14 countries whose national data was compiled by Jacques Charmes include: Morocco and Tunisia in Northern Africa; Benin, Burkina Faso (street vendors only), Chad, Ethiopia (urban), Gabon, Kenya, Mali, and Niger in sub-Saharan Africa; Brazil, Colombia, and Mexico (all urban) in Latin America; and India and Indonesia in Asia. All of the data was from the late 1990s with the exception of Morocco, where the data was from 1992 (Charmes n.d.).
3. A third pair of relevant analyses from India is of the National Sample Survey 1999-2000 (55th Round) by N.S. Sastry (Sastry 2004) and of the National Sample Survey 1987/88 and 1993/94) (Dubey et al. 2001).
4. A fourth relevant analysis is of the 1997 national labour force survey in Tunisia cited in Charmes and Lakehal (2003).
Data Findings
Average Earnings
The links between employment, gender, and poverty can be seen by comparing a) average earnings in formal and informal employment and b) average earnings of different categories of informal employment.
Average Earnings in Formal and Informal Employment -
A first comparison is the contrast between average wages or earnings in formal and informal employment, taken as a whole. In three of the five countries in the EPI-GPN set of studies, the analysts were able to compare average wages or earnings data. The results confirm that, on average, wages or earnings are higher in formal than in informal employment, as summarized below:
- Egypt: Average real wages of the formal and informal workforce, both sexes, were measured at two points in time (1988 and 1998). The results suggest a large gap between formal and informal real wages in both years and for both sexes, but a narrowing of the gap by the second point of time as formal real wages declined more rapidly than informal real wages. However, between the two points in time, female informal wages declined faster than female formal wages (El Mahdi and Amer 2004).
- El Salvador: Earnings from formal and informal employment in relation to the minimum wage were compared for 2002. A relatively small share (14%) of the formal workforce earns below the minimum wage. Within the informal workforce, a higher share of rural workers (77%) than urban workers (49%) earn below the minimum wage (Lara 2004). It should be noted that the minimum wage is set at a level that would not cover the cost of “basic goods.”
- South Africa. The income of formal and informal sector workers for 2001 was compared. While the majority of formal workers earn above 1,000 rand per month, the majority of informal workers earn less than 1,000 rand. The estimated minimum level of income needed for a family of five is set at 1,777 rand per month (NALEDI 2003).
Comparative Earnings within Informal Employment
A second comparison is the difference in average earnings within informal employment. As noted earlier, the informal economy is diverse and segmented. The different segments are associated with different earning potentials that would be concealed by the average for the informal economy as a whole. For example, an analysis of 1997 data on employment in the informal sector (small unregistered enterprises) in Tunisia found that the employers who hired others – the micro-entrepreneurs – were not poor. Indeed, the average income of micro-entrepreneurs was found to be four times as high as the legal minimum salary and 2.2 times the average salary in the formal sector (Charmes and Lakehal 2003).
Although micro-entrepreneurs may have relatively high earnings in Tunisia — and elsewhere — most workers in informal employment do not fare so well. For example, the micro-entrepreneurs in Tunisia paid their employees on average roughly the legal minimum wage of 200 dinars per month. The Tunisian study also included information on earnings in jobs outside informal enterprises – notably for homeworkers. Homeworkers, who are paid by the piece, earned an average of 60 dinars per month, which is only 30 per cent of the minimum wage (Charmes and Lakehal 2003).
The national data from 14 countries compiled by Jacques Charmes show the disparities in earnings within informal employment (Charmes n.d., table, presented in Chen and Vanek 2004). In every case except Kenya, the average monthly income of micro-entrepreneurs is higher than the average monthly wages of the employees of micro-enterprises. Generally, the wages of employees tend to hover around the minimum wage — which in itself may be less than the minimum needed for survival (ibid).
Another important comparison is between the average earnings of micro-entrepreneurs and of own account operators. Among the 14 countries studied by Charmes, only two — Colombia and India — report separate earnings data for employers and own account operators. In India, the lowest multiple of average monthly income to legal minimum wage (1.34) was for own account operators in India. In marked contrast, the average monthly income of employers in India was 5.4 times the legal minimum wage. A similar contrast was found in urban Columbia, where employers earn 4.2 times the legal minimum wage and own account operators earn only 1.6 times. In fact, in urban Columbia the employees of micro-enterprises earn nearly as much as own account operators: 1.5 times the legal minimum wage. In sum, in both Colombia and India, micro-entrepreneurs/informal employers earn higher monthly average income than own account operators, and own account operators have only slightly higher average earnings than employees of informal enterprises (ibid).
Household Poverty
Two recent labour force surveys — the 2002 Labour Force Survey in South Africa and the 1999-2000 National Sample Survey of India on employment and unemployment — provide unique data that begin to answer questions regarding the relationship between employment and poverty. Both surveys collected household expenditure data as well as data on employment, including informal employment; and both studies tried to link these variables in a meaningful way by classifying households by sources of income and by expenditure categories. Analyses of these data sets all found an overlap between depending on informal employment and being poor at the household level.
South Africa was one of the three countries studied in both the EPI-GPN set of country cases. The EPI-GPN study looked at the relationship between monthly household expenditure categories and the sources of employment income in households in South Africa: that is, by whether a household had one or more persons in permanent employment, in informal employment, in domestic work or unemployed (NALEDI 2003). The higher the monthly expenditure category, the higher the percentage of households with persons in permanent employment. Moving down the expenditure categories, the percentage of households with persons in informal employment (including domestic services) increases. Not surprisingly, the lowest expenditure category had the highest percentage of households with an unemployed person or persons (using an expanded definition of unemployment). It should be noted that the unemployment rate is very high in South Africa (ibid).
An analysis of the 1999-2000 India data looked at poverty rates among urban Indian households that sustain themselves on informal employment income by broad industrial sector and employment type (Sastry 2004). In marked contrast to South Africa, unemployment in India is not high, and the vast majority of workers — 92 per cent — are in informal employment (using the expanded definition of informal employment). Households that depend on “regular” (as opposed to casual) informal wage employment have lower poverty rates relative to households that rely on self-employment, and households that depend on casual labour as their primary source of income are the most likely to be poor. This hierarchy of poverty risk — households depending on “regular” informal wage employment having the lowest, self-employment the next highest, and casual wage employment the highest risk — is robust across industrial sectors in urban India.
Another study in India, also using data from the National Sample Survey but from two earlier surveys (1987/88 and 1993/94), found a similar relationship between poverty and the nature of employment. Dubey et al analyzed the probability of urban households being poor according to their main source of income — classified as regular salary, self-employment and casual wage labour — and by the size of the city or town in which they were located (Dubey et al. 2001). Their analysis shows that, for cities or towns of all sizes and both points in time, households with regular salaried employees (both formal and informal) have the lowest probability of being poor, while those that depend on casual day labour have the highest probability, and households that depend on self-employment falling roughly half-way in between. All employment groups fared better in larger cities. And, between the two rounds of the survey, the probability of being poor declined for all groups.