Global Poverty Research Group

The Nature of Unemployment in South Africa: Voluntary or Involuntary

Overview

Unemployment in South Africa is so widespread that it demands an explanation. What is unique and unusual about the South African labour market, vis a vis most developing and middle-income countries, is the smallness of the size of its informal sector and its corresponding large rate of open unemployment. South Africa is an international outlier in terms of the size of its informal sector (Tables 1 and 2 below). The research under this programme examines a central question about South African unemployment. Why do the unemployed not enter the informal sector, as is common in other developing countries? And is South African unemployment largely voluntary? The data do not support the idea that unemployment is largely voluntary. The policy implications – that government should diminish labour market segmentation and the obstacles to entering the productive informal sector – may be relevant also to other developing countries with high unemployment.

The findings so far provide little support for the idea that unemployed people choose in any meaningful sense to be unemployed. We find that there is sharp earnings segmentation between the formal sector and the relatively small informal sector, and that the unemployed are substantially worse off even than the informally employed, in terms of both income and expenditure. This contradicts the luxury unemployment interpretation of joblessness, whereby higher household income reduces the incentive to become employed in the informal sector and increases the incentive to consume more leisure. It might be contended that, given the disutility of work, some people prefer to substitute leisure for higher monetary income, so that their apparent deprivation cannot be used to argue that they are constrained to be unemployed. However, if their unemployment is to be interpreted as voluntary, such people should be happier (or less unhappy) than if they were employed. The research shows that households with a high proportion of unemployed persons are very substantially and significantly less satisfied with their quality of life than households with a high proportion of informally employed. They suggest that unemployment arises through impediments to entry into informal work, and they are at odds with the notion that unemployment is anything other than the better of two awful choices. Although this important issue deserves more research, we find various plausible reasons why the informal sector has been inhospitable to newcomers in South Africa.

Table 1

Unemployment and informal employment

 
Urban unemployment rate
Employment rate in the informal sector (3)
Ratio of informal sector employment to unemployment
 
South Africa
29.3
18.9
0.7
 
 
 
Other Sub-Saharan Africa (1)
16.0
74.8
4.7
Benin
10.1
92.8
Burkina Faso  
--
77.0
Chad  
--
74.2
Guinea
12.3
71.9
Kenya
16.2
71.6
Mali
9.9
78.6
Mauritania
31.6
75.3
Mozambique
--
73.5
Zambia
--
58.3
 
 
 
 
 
Latin America (1)(2)
8.1
56.9
7.0
Argentina
18.8
53.3
Bolivia
3.6
63.6
Brazil
4.6
57.6
Colombia
9.0
55.5
Ecuador
6.9
53.5
Mexico
6.3
59.4
Paraguay
5.6
65.5
Venezuela
10.3
46.9
 
 
 
Asia (1)
5.3
63.0
11.9
India
--
73.7
Indonesia
7.2
77.9
Pakistan
6.1
64.6
Philippines
7.4
66.9
Thailand
0.4
51.4
Iran
--
43.5

 

Table 2

Percentage share of informal employment in total labour force, by definition of labour force

 

OHS 1997
OHS 1998
OHS 1999
LFS Feb 2000
LFS Sep 2000
LFS Feb 2001
LFS Sep 2001
LFS Feb 2002
Using Narrow Definition of LF
IS employment/total narrow LF
19.0
17.0
20.3
27.4
26.6
27.7
21.4
22.5
Using Broad Definition of LF
IS employment/total broad LF
15.2
14.2
16.9
24.1
23.0
23.7
17.8
18.8

Note: IS is short form for informal sector; LF is short form for labour force.

Source: Computed from Statistical Releases of Statistics South Africa.

 

In attempting to understand a crucial issue for South Africa, we have imperfect data. The quality of life questions are thus far only available at the household level; this information is more relevant at the individual level. Data on the distribution of consumption, and the sources of income, among members of the household would help to clarify the options available to the unemployed. A regular national panel household survey would help to overcome the problems of unobserved heterogeneity that have qualified our analysis, and would provide the longitudinal information on workers needed to understand more about the nature of unemployment and of informal sector employment, for instance by measuring the income changes resulting from transitions between employment states. A more precise reservation wage question is needed that collects information on expected hours of work per period, maximum acceptable distances to work, past wages, and past wage offers rejected, and which makes a dedicated attempt to obtain data on the minimum wage that would be acceptable for work rather than the expected, fair or bargaining wage. Finally, and most importantly, our arguments pinpoint the need to understand the potential barriers to entry into the informal sector in South Africa: surveys of unemployed people that specifically address this issue are required.

It is likely that most currently unemployed workers in South Africa are involuntarily unemployed in the sense that they would accept formal sector jobs at the going wages. Although each unemployed worker voluntarily chooses not to enter free-entry activities, this may well be because incomes in the free-entry part of the informal sector are extremely low. However, there is no real choice. For as long as barriers to entry continue to restrict opportunities in much of the informal sector, this sector will be unable to absorb significantly more of the currently jobless. Unemployed workers face a high probability of remaining unemployed, whatever their search activity. The need for policies that would reduce unemployment in South Africa is compelling. Our diagnosis yields two main policy implications. Government should try to diminish labour market segmentation and to overcome the obstacles to entering the productive informal sector.

Although South African unemployment is extremely high, a number of developing countries have comparable rates. For instance, in various years in the 1990s, countries with high national unemployment rates in Africa included Algeria (30%), Botswana (22%), urban Ethiopia (39%), urban Mauritania (32%), Morocco (22%), Zambia (25%) and Zimbabwe (22% in 1992 and estimated at 50% in 1999); in Latin America, Argentina (19%) and Colombia (21%); and in Southern Europe, Armenia (36%) and Macedonia (39%)1. It is possible that our diagnosis applies also to some of these countries. There is a case for research to compare the South African labour market with those of similar economies which have either suffered or avoided high unemployment.

1Source: ILO (2000, 2001 – from labour force surveys) except Ethiopia (Krishnan et. al., 1998); Mauritania (Charmes, 2000); Zambia (SADC quoted in www.germanchamber.co.za/sadc.htm) and Zimbabwe (CSO, 1994, p99) and SADC quoted in www. germanchamber.co.za/files/sadc.htm.

Recent publications

Kingdon, G. and J. Knight. “Unemployment in South Africa: The Nature of the Beast”, World Development, 32, No. 3: 391-408, March, 2004.

Researchers to contact for this project

Geeta Kingdon & John Knight