Women and the economy
- Written by Gary Watkins
- Published in articles001-100
Women and the economy, Neva Seidman Makgetla, August 2004
Source: http://www.genderstats.org.za/
In the early 2000s, compared to other groups and especially to white men, black women in South Africa faced higher unemployment, lower incomes from work, and relatively poor access to training and promotions. Employed women remained largely concentrated in relatively poorly paying occupations and industries. For African women, domestic work was still the largest single employer.
Yet discrimination based on race and gender had been banned for ten years. Several laws reinforced the Constitutional requirements, and the Employment Equity Act required employers to take positive measures to improve prospects for historically disadvantaged employees, including women.
So why did most women continue to face economic disadvantage?
To start with, for the first decade of democracy, extraordinarily high overall unemployment by world standards combined with slow economic growth. These circumstances made employment equity appear a zero-sum game, and very difficult to enforce.
Second, the laws on equity essentially did not reach beyond economic transactions and employment. They did not directly address the economic context of high levels of unemployment and women’s lack of economic assets. Nor did they engage persistent inequalities in homes, communities and schools. In part, this reflects the pervasive belief that the state should not intervene directly in family relations except in extreme cases. But indirect methods of influencing household and community relations, for instance through provision of assets directly to women, were also neglected.
The problem was aggravated by the fact that African women were over-represented in the ex-homeland areas. Even ten years after the end of apartheid, these regions were still extraordinarily poor and lacking in basic infrastructure.
Government policies geared to improving government services and supporting black ownership in the economy should have helped address these problems. Unfortunately, despite substantial improvement, they were hampered by inadequate funding and, in some cases, inappropriate policies.
In sum, improving women’s position in the economy requires structural transformation to increase overall equity and growth, rather than just better enforcement of anti-discrimination measures. Key steps would aim
- To shift the formal sector toward more labour-intensive sectors that could provide employment on a large scale.
- To improve the ability of poor households to engage in the economy by providing productive assets, skills and access to marketing and financial networks – amongst others, through large-scale land reform as well as improved access to credit, infrastructure and training in black communities, with a focus on women.
- To reduce the burden of reproductive labour on women both by improving household infrastructure and by influencing the division of labour in the household through education and by enhancing women’s economic independence.
- To ensure genuine equity in the education system in terms of class, race and region as well as gender, which would be reflected at least in representative pass rates for matric and in greater representivity in the universities.
This paper first describes women’s overall economic position. It then locates the key factors behind continuing differentiation in a broader structural understanding of unemployment and poverty in South Africa. A final section outlines some of the policy implications.
1 Women’s economic position
In 2003, almost a decade after the achievement of democracy, women as a whole still had lower incomes, higher unemployment, and less access to assets than men. But racial differences were larger than gender inequalities within racial groups. We can only understand the position of women in the economy, then, if we also take race into account.[1]
As the following table shows, black women were far less likely to have paid employment than any other group. To start with, they were more likely to be counted as “economically inactive,” that is, to report neither having an earned income of their own nor to be seeking one. Obviously, virtually all these women were active in unpaid, mostly reproductive labour, and many also received childcare grants or old-age pensions.
In addition, black women faced far higher rates of unemployment. The unemployment rate for African women was almost ten times as high as for white men. African women made up 42% of the labourforce, but only 30% of the employed and 51% of the unemployed.
Table 1. Employment by race and gender, 2003a
|
Women |
men |
||||
African |
Coloured/ Asian |
White |
African |
Coloured/ Asian |
White |
|
Not economically active |
5,567,000 |
877,000 |
781,000 |
3,825,000 |
506,000 |
416,000 |
Employed |
3,556,000 |
840,000 |
855,000 |
4,405,000 |
1,024,000 |
1,104,000 |
Unemployed |
4,284,000 |
383,000 |
92,000 |
3,200,000 |
318,000 |
67,000 |
Total |
13,408,000 |
2,100,000 |
1,728,000 |
11,430,000 |
1,848,000 |
1,588,000 |
unemployment rate |
55% |
31% |
10% |
42% |
24% |
6% |
employment as % of adult population |
27% |
40% |
49% |
39% |
55% |
70% |
Note: a. These data use the expanded definition of unemployment, which includes workers who would take paid work immediately but who are too discouraged actively to seek it. Source: Calculated from, September 2003 Labour Force Survey. Statistics South Africa. Pretoria. Database on CD-Rom.
Unemployment was particularly high for young people. African women under the age of 30 faced an unemployment rate of 75%. They constituted 17% of the labourforce – that is, the employed plus the unemployed - but 31% of the unemployed.
Unemployment rose for the black population as a whole at least through the early 2000s. The data are not entirely reliable, however. The official employment surveys – the October Household Survey (OHS) until 1999 and the Labour force Survey thereafter – may not be fully comparable over time. (See Makgetla 2004c)
The main problem is that the surveys gradually redefined unpaid labour, especially subsistence farming and work in family enterprises as employment. The OHS enumerators generally failed to identify this work as employment and labelled those involved as "economically inactive.” In contrast, the Labourforce Survey made a greater effort to capture it as informal employment. The inclusion of subsistence farming alone as “employment” accounted for 20% of reported employment growth between 1997 and 2002. (Devey et al, 2002)
In addition, with soaring unemployment, the subjective nature of “unemployment” became increasingly obvious. Standard statistics in South Africa define a person as “unemployed” only if they are actively seeking work. The line between economically inactive and unemployed may reflect only an individual’s discouragement. The broader definition of unemployment, used here, tries to reduce the subjective element by asking only whether a person would take a paying job.
Both these factors disproportionately affected the data on black women. Women are most likely to be engaged in unpaid labour such as subsistence work or support for family enterprises. Moreover, since black women face the highest levels of unemployment, they are more likely to be discouraged from actively seeking paid work.
Despite these caveats, as the following table shows, reported unemployment soared through the 1990s and early 2000s. It rose relatively slowly for black women, but they started with much higher joblessness than other groups.
Table 2. Employment status by race and gender, 1996 and 2003
|
1996 |
2003 |
Women |
|
|
African |
51% |
55% |
Coloured/Asian |
22% |
31% |
White |
6% |
10% |
Men |
|
|
African |
35% |
42% |
Coloured/Asian |
14% |
24% |
White |
4% |
6% |
Source: Calculated from, Statistics South Africa. 1996 October Household Survey and September 2000 and 2003 Labour Force Surveys. Pretoria. Databases on CD-Rom.
High unemployment for black women went hand in hand with lower pay. Table 3 shows that in 2003 almost two thirds of black women earned under R1000 a month, compared to 3% of white men. The available data suggest that the share of women earning under R1000 a month had remained virtually unchanged since 1996. Since purchasing power dropped by over half in this period, this figure suggests that real earnings declined. Presumably, however, the increasing inclusion of unpaid work explains part of the fall in real pay.
Table 3. Incomes by race and gender, 2003
|
women |
|
men |
|
||
Monthly income |
African |
Coloured/ Asian |
White |
African |
Coloured/ Asian |
White |
up to R1000 |
64% |
31% |
5% |
40% |
23% |
3% |
R1001 to R2500 |
18% |
32% |
14% |
35% |
30% |
8% |
R2501 to R4500 |
9% |
19% |
27% |
15% |
23% |
14% |
R4501 to R8000 |
9% |
17% |
42% |
9% |
20% |
47% |
over R8000 |
1% |
2% |
11% |
2% |
5% |
29% |
Total |
100% |
100% |
100% |
100% |
100% |
100% |
Source: Calculated from, September 2003 Labour Force Survey. Statistics South Africa. Pretoria. Database on CD-Rom.
Income differentials appear largely to reflect the concentration of women in lower-paid industries and occupations. While unequal pay for equal work certainly persisted, it was illegal and generally camouflaged by differences in job title and status.
In 2003, 5% of black women were employed as managers and senior professionals, compared to 33% of white men. Some 25% of African women were employed as domestic workers, and 27% were elementary (that is, “unskilled”) workers.
Table 4. Occupation by race and gender, 2003
|
Women |
men |
||||
African |
Coloured/ Asian |
White |
African |
Coloured/ Asian |
White |
|
Legislators, senior officials and managers |
2% |
4% |
15% |
4% |
10% |
29% |
Professionals |
3% |
3% |
14% |
2% |
4% |
14% |
Technical and associate professionals |
11% |
12% |
20% |
6% |
9% |
15% |
Clerks |
8% |
23% |
36% |
5% |
9% |
7% |
Service and sales workers |
12% |
13% |
10% |
13% |
10% |
8% |
Skilled agricultural and fishery workers |
4% |
0% |
1% |
4% |
1% |
3% |
Craft and related trades workers |
5% |
5% |
2% |
19% |
20% |
17% |
Plant and machine operators and assemblers |
3% |
8% |
1% |
19% |
13% |
4% |
Elementary occupations |
27% |
20% |
1% |
27% |
25% |
3% |
Domestic workers |
25% |
11% |
0% |
1% |
0% |
0% |
Total |
100% |
100% |
100% |
100% |
100% |
100% |
Source: Calculated from, September 2003 Labour Force Survey. Statistics South Africa. Pretoria. Database on CD-Rom.
African women made little progress in occupational terms between 1996 and 2003. Their share in senior management and professional category rose only from 11% to 12% in this period. In contrast, Coloured and Asian men and women gained substantially in this period, while the share of African men reportedly declined.[2]
Table 5. Occupation by race and gender, 1996 and 2003
|
women |
Men |
|
||||
African |
Coloured/ Asian |
White |
African |
Coloured/ Asian |
White |
Total |
|
senior management and professionals |
|
|
|
|
|
|
|
1995 |
11% |
3% |
14% |
22% |
9% |
42% |
100% |
2003 |
12% |
5% |
18% |
20% |
11% |
35% |
100% |
technical and associate professionals |
|
|
|
|
|
|
|
1995 |
30% |
5% |
16% |
22% |
5% |
21% |
100% |
2003 |
33% |
8% |
14% |
23% |
8% |
14% |
100% |
clerks, service and sales workers |
|
|
|
|
|
|
|
1995 |
21% |
9% |
22% |
29% |
8% |
10% |
100% |
2003 |
28% |
12% |
15% |
31% |
7% |
6% |
100% |
skilled production |
|
|
|
|
|
|
|
1995 |
8% |
3% |
1% |
54% |
14% |
19% |
100% |
2003 |
14% |
4% |
1% |
62% |
12% |
9% |
100% |
elementary occupations |
|
|
|
|
|
|
|
1995 |
39% |
7% |
0% |
44% |
9% |
1% |
100% |
2003 |
37% |
6% |
0% |
45% |
10% |
1% |
100% |
domestic workers1 |
|
|
|
|
|
|
|
2003 |
86% |
9% |
0% |
4% |
0% |
0% |
100% |
total employees |
|
|
|
|
|
|
|
1995 |
23% |
6% |
9% |
39% |
9% |
14% |
100% |
2003 |
30% |
7% |
7% |
37% |
9% |
9% |
100% |
Note. 1. Domestic workers were not aggregated separately from other elementary occupations in 1995. Source: Calculated from, Statistics South Africa. 1995 October Household Survey and September 2000 and 2003 Labour Force Surveys. Pretoria. Databases on CD-Rom.
Overall, between 1996 and 2003, African women moved from the lower level professions – essentially nurses and teachers – to clerical and retail work. This, in turn, largely reflected the freezing of public service positions after 1996, which closed off what had previously been a major source of higher-level employment for black women.
As the following table shows, changes over time emerged as generational differences. African women under 30 were more likely to be unemployed and, if employed, to work in retail and clerical occupations. They were distinctly less likely to have jobs as teachers or nurses.
Chart 1. Women’s occupations by age, 2003
Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
Women were also predominantly found in relatively poorly paid sectors, as Chart 2 demonstrates. In 2003, half of all women were employed in domestic and sales work, which are poorly paid. Some 8% were in farming, with the vast majority subsistence farmers and farmworkers. A further 17% were in education and health – sectors requiring considerable skills, but paying relatively little given the high education level.
Chart 2. Industry and incomes by race and gender, 2003
Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
There is a strong statistical correlation between the percentage of African women in the labourforce of a sector and the percentage of workers in the sector earning under R1000.[3]
For African women, domestic labour remained a critical source of paid labour. Some 96% of domestic workers were black women. This was the worst-paid industry, with 93% of workers earning under R1000 a month. Even farmworkers enjoyed slightly better conditions.
In short, the data point unambiguously to the fact that, a decade after the end of apartheid, black women were less likely to have access to any paid work at all. Even if they had employment, they were typically engaged in lower-level occupations and worse-paid industries. The following section explores some of the reasons that this situation persisted.
2 Factors affecting women’s economic engagement
The two dominant economic approaches to understanding unemployment and poverty in South Africa generally neglected gender. One view essentially reflected standard human-capital theories, arguing that the unemployed were essentially unemployable because of low skill levels. (Bhorat, 2002; PCAS 2003) The other approach, in contrast, blamed the structure of the economy, which was associated with rising capital intensity in the formal sector and the impoverishment and marginalisation of the majority. (See, for instance, Makgetla 2004b; DTI 2002; Altman 2003; de Swardt, 2003)
These two approaches generate very different research agendas. The first tends to emphasise an investigation of training levels and fields of study. The second focuses attention on institutions inside and outside of the labour market and how they reproduce poverty and joblessness. We here explore each in turn.
2.1 Human capital and women
There is no question that low levels of education hamper employment creation. But the data simply do not support the argument that low levels of education formed the main cause of rising unemployment. The persistence of this argument seems to reflect assumptions about African skills dating back to the apartheid era, rather than the realities of South African society in the early 2000s.
As the following table shows, there was virtually no difference in education levels between African women and men. Indeed, African women under 30 had slightly more formal education than African men. Moreover, unemployed African youth have virtually the same levels of education as the lucky few with jobs. In contrast, the economically inactive tended to have substantially lower levels of education, especially for those aged over 30 years.
Table 6. Average years of education by race, gender age and employment status, 2003
|
African |
Coloured/Asian |
White |
|||
|
women |
men |
women |
men |
women |
men |
not economically active |
|
|
|
|
|
|
30 or less |
9.8 |
9.4 |
10.5 |
10.3 |
11.3 |
11.1 |
over 30 |
4.4 |
4.8 |
7.9 |
7.6 |
12.1 |
12.1 |
employed |
|
|
|
|
|
|
30 or less |
11.1 |
10.5 |
11.9 |
11.3 |
13.1 |
13.0 |
over 30 |
8.7 |
8.6 |
10.1 |
10.2 |
13.1 |
13.1 |
unemployed |
|
|
|
|
|
|
30 or less |
10.7 |
10.3 |
10.6 |
10.4 |
12.4 |
12.1 |
over 30 |
8.4 |
8.2 |
9.0 |
9.7 |
12.4 |
12.3 |
difference between employed and unemployed in months of schooling |
||||||
30 or less |
5 |
2 |
15 |
11 |
8 |
11 |
over 30 |
3 |
5 |
13 |
7 |
8 |
10 |
Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
In addition, education levels for unemployed African women rose rapidly from the mid-1990s. As the following table shows, the education level of African women rose by over a year between 1996 and 2003, as more educated people joined the ranks of the jobless.
Table 7. Average years of education of African women by age and employment status, 1996 and 2003
|
Average years of education |
change in months |
|
1996 |
2003 |
||
economically inactive |
|
|
|
30 years old or less |
9.6 |
9.8 |
2 |
over 30 |
4.2 |
4.4 |
2 |
employed |
|
|
|
30 years old or less |
10.4 |
11.1 |
9 |
over 30 |
8.4 |
8.7 |
3 |
unemployed |
|
|
|
30 years old or less |
9.6 |
10.7 |
14 |
over 30 |
7.0 |
8.4 |
17 |
Source: Calculated from, Statistics South Africa. October Household Survey 1996, and Labour Force Survey September 2003. Pretoria. Databases on CD-Rom.
The argument that low skills are the main reason for high unemployment also fails to explain differences in unemployment for university graduates, which are strongly related to race and gender. As the following table shows, African women with university degrees faced an unemployment rate of 13%, compared to 1% for white men. Unemployment for African women with other tertiary degrees was even higher.
Table 8. Unemployment amongst tertiary graduates, 2003
|
African |
Coloured/Asian |
White |
Women |
|
|
|
tertiary other than university |
29% |
6% |
8% |
university degree |
13% |
10% |
3% |
degrees as % of total tertiary |
22% |
31% |
42% |
Men |
|
|
|
tertiary other than university |
26% |
7% |
3% |
university degree |
9% |
4% |
1% |
degrees as % of total tertiary |
27% |
35% |
49% |
Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
In addition, as Table 9 shows, black women earned less than other groups at every income level. A white man with matric was twice as likely as an African woman with a university degree to earn over R8000 a month.
Table 9. Education and incomes by race and gender, 2003
|
African |
Coloured/Asian |
White |
Women |
|
|
|
primary |
89% |
75% |
0% |
some secondary |
73% |
36% |
12% |
matric |
45% |
14% |
6% |
tertiary other than university degrees |
10% |
5% |
3% |
university degree |
3.4% |
4.4% |
2.6% |
Men |
|
|
|
primary |
61% |
61% |
9% |
some secondary |
41% |
25% |
3% |
matric |
23% |
9% |
3% |
tertiary other than university degrees |
7% |
2% |
2% |
university degree |
2.8% |
3.8% |
1.4% |
Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
It is possible to argue that figures on overall education do not adequately take into account systematic gender differences in fields of study. As the following tale shows, women in general, and particularly Africans, tended to take degrees in “softer” subjects – education, culture, social and business studies. They remained poorly represented in the “harder” subjects like science and engineering, as well as in the professions.
Table 10. Degrees by field, race and gender, 2003
|
Total number |
% of degrees in field for: |
|||
Field |
men |
women |
African women |
White men |
|
university degrees as % of total tertiary degrees |
37% |
44% |
32% |
23% |
57% |
tertiary ex university |
|
|
|
|
|
soft subjects (education, social studies, culture) |
549,000 |
27% |
42% |
47% |
11% |
business and communications |
360,000 |
21% |
25% |
21% |
22% |
professionals (health, law, services) |
279,000 |
14% |
21% |
19% |
18% |
hard subjects (engineering, science, agriculture, construction) |
342,000 |
37% |
12% |
13% |
49% |
Total |
1,531,000 |
100% |
100% |
100% |
100% |
university degrees |
|
|
|
|
|
soft subjects (education, social studies, culture) |
296,000 |
22% |
44% |
49% |
13% |
business and communications |
248,000 |
30% |
24% |
21% |
33% |
professionals (health, law, services) |
202,000 |
22% |
23% |
21% |
19% |
hard subjects (engineering, science, agriculture, construction) |
172,000 |
27% |
9% |
8% |
34% |
total |
918,000 |
100% |
100% |
100% |
100% |
Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
Moreover, the majority of black graduates still attend historically black universities, which employers considered to be of lower quality. Both these arguments, however, point to deeper institutional factors as a problem, rather than simply levels of education.
In addition, except at the highest levels, blacks and women had less access to training than white men, as Chart 3 illustrates. Again, this points to structural problems. After all, both the Employment Equity Act and the National Skills Strategy required that employers prioritise training for black women.
Table 3. Reported access to training by race and gender, 2003
Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
In short, the available evidence points to rising levels of education for African women even as unemployment increased. That suggests that low skills in themselves did not form a primary cause of joblessness. Certainly higher education levels as well as a shift to higher-paid fields would help solve the problem. In themselves, however, they would not generate a qualitative increase in the demand for labour or permit equal participation by black women in the economy.
2.2 Structural interpretations
A wide variety of structural interpretations of women’s status in the economy exist. We here add a gender dimension to the argument.
From around 2003, the government began to argue that the South African economy was heavily dualist. It had a “first” economy comprising essentially the formal sector and a “second” economy including everyone else. (See SARPN 2004)
The obvious weakness in this argument is that the second economy does not actually constitute a coherent economic system. Rather, it comprises a heterogeneous grouping of the unemployed and informally employed – that is, all those who have effectively been marginalised from the formal economy.[4] The model is particularly weak in dealing with domestic workers. They are employed in a subordinate and impermanent position in households that, in turn, are mostly integrated into the formal sector.
If we gender this characterisation, it becomes clear that women are disproportionately represented in the so-called second economy. As the following table shows, the profile of African women’s employment status is dramatically different from that of other groups defined by race and gender. They are far less likely to have formal jobs, and far more likely to be unemployed, informally employed, or domestic workers.
Table 11. Informal employment and unemployment by race and gender, 2003
|
women |
men |
|
||||
|
African |
Coloured/ Asian |
White |
African |
Coloured/ Asian |
White |
TOTAL |
% of population group |
|
|
|
|
|
|
|
Formally employed |
20% |
44% |
53% |
34% |
49% |
52% |
33% |
Informally employed |
11% |
3% |
4% |
12% |
5% |
3% |
9% |
Domestic worker |
10% |
6% |
0% |
0% |
0% |
0% |
4% |
Unemployed |
60% |
48% |
43% |
54% |
46% |
45% |
54% |
Total |
100% |
100% |
100% |
100% |
100% |
100% |
100% |
% of employ-ment grouping |
|
|
|
|
|
|
|
Formally employed |
21% |
8% |
9% |
38% |
11% |
12% |
100% |
Informally employed |
40% |
2% |
2% |
50% |
4% |
3% |
100% |
Domestic worker |
86% |
9% |
0% |
4% |
0% |
0% |
100% |
Unemployed |
39% |
6% |
5% |
38% |
7% |
7% |
100% |
Total |
35% |
6% |
6% |
37% |
8% |
8% |
100% |
Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
The structural divisions in the South African economy were closely linked to spatial disparities. Even in 2003, the former homeland areas continued to be the poorest regions in the country. As the following table shows, they were characterised by extraordinarily high levels of unemployment and poverty – and a predominance of African women.
Women made up 56% of the population of the former homeland areas, compared to 52% in the rest of the country. Almost half of all African women adults lived in the former homeland regions, and over 40% of African men, but under 5% of whites, Coloureds and Asians.
Table 12 shows that African women in the former homelands were far less likely to be formally employed and far more likely to be economically active, if employed, to earn under R1000 a month.
Table 12. Employment status, incomes and gender in the former homeland areas,1 2003
|
African women |
African men |
||
|
predominantly homelands |
other areas |
predominantly homelands |
other areas |
not economically active |
51% |
33% |
43% |
26% |
unemployed |
30% |
34% |
27% |
29% |
total employed |
19% |
33% |
30% |
45% |
Of which: |
|
|
|
|
- formal |
7% |
18% |
18% |
36% |
- informal |
8% |
6% |
11% |
9% |
- domestic |
4% |
9% |
0% |
1% |
Total |
100% |
100% |
100% |
100% |
% earning under R1000/month |
81% |
53% |
60% |
30% |
% of adults population by area |
49% |
51% |
44% |
56% |
Note. 1. The data are from the rural areas of KwaZulu Natal, Mpumalanga, Limpopo, the Eastern Cape and Northwest. These areas are not identical to the former homelands, but are largely contiguous with them. Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
The structural approach throws up four main explanations for the persistence of women’s economic subordination.
First, the colonial and later the apartheid state initially impoverished the black majority, and especially women. Once the state ended those interventions, however, the market did not spontaneously transfer productive resources to the poor or to deprived regions in general. Government efforts since 1994 have not sufficed to overcome the backlogs created over centuries of oppressive rule.
Second, growth in the first economy focused on relatively capital-intensive sectors – minerals, heavy chemicals, auto manufacturing, telecommunications and the financial sector. These industries traditionally employ primarily men. In contrast, light industry and services, where women have historically found employment, tended to stagnate.
Third, women with jobs were still less likely than men to belong to unions. Yet labour law, including the Employment Equity Act, relies heavily on workers' own organisations to monitor conditions and help bring about improvements.
Fourth, most women kept responsibility for household labour and childcare, even if they earned their own income. This burden was aggravated by shortcomings in infrastructure in black communities, as well as by the impact of the AIDS pandemic.
We here review how developments in each of these areas affected women s economic position.
2.2.1 Access to productive assets and infrastructure
Historically, the state largely denied black communities access to land, education and skills, basic infrastructure, financial services and marketing networks. The Reconstruction and Development Programme (RDP) argued that providing infrastructure, land and financial services would provide the basis for increased household productivity and incomes, as well as raising living standards directly.
In the event, these expectations have not been borne out.
In part, the shortcomings reflect inadequate funding. As a result, although improvements have been made in access to basic infrastructure and education, the low level and relatively high cost of services limits the economic impact. Provision of services to black communities proved particularly slow in the late 1990s, when the government cut the budget in real terms by around 1% a year. In contrast, since 2000, when the budget has consistently grown, service provision has accelerated.
In part, the limited employment impact of government anti-poverty programmes arises out of the relative ineffectiveness of programmes to support small and micro enterprise. This, in turn, results both from inadequate funding and from the broader failure to restructure the formal sector toward job-creating growth.
As the following table shows, substantial progress was made in extending basic household services, especially in the early 2000s.
Table 13. Access to basic infrastructure, 1996 and 2003
|
Percentage of households with access to service |
Average annual increase in share with access |
|||
Type of infrastructure |
1996 |
2000 |
2003 |
1996-2000 |
2000-'03 |
Electricity for lighting |
64% |
71% |
79% |
2.1% |
3.6% |
Electricity for cooking |
51% |
51% |
59% |
0.0% |
5.0% |
Piped water |
82% |
83% |
86% |
0.3% |
1.2% |
Flush toilet |
52% |
54% |
57% |
0.6% |
1.6% |
Source: Calculated from, Statistics South Africa. 1996 October Household Survey and September 2000 and 2003 Labour Force Surveys. Statistics South Africa. Pretoria. Databases on CD-Rom.
Often, however, these services were provided at a very low level, and sometimes at unaffordable costs. Thus, new electricity connections to poor households generally sufficed only for light, not for cooking or for refrigeration. That ruled out cooking and spaza shops as a way to make money. As the following table shows, the cost of services often proved high for the low-income group. In 2000, while most poor households did not pay for electricity or water, those that did paid much more relative to their income than better-off households.
Although the government subsidised over a million houses after 1994, progress was less impressive in terms of improving housing, largely because of substantial rural-urban migration. The share of informal housing remained virtually stable at 13%, formal housing expanded from 69% to 76%, and traditional housing declined.
Table 14. Housing by type, 1996 and 2003
|
% of total |
|
|
Type of housing |
1996 |
2003 |
% change |
formal over three rooms |
54% |
50% |
24% |
formal three rooms or less |
15% |
26% |
126% |
Informal |
12% |
13% |
42% |
traditional dwelling |
19% |
12% |
-18% |
Total |
100% |
100% |
34% |
Source:Calculated from, Statistics South Africa. October Household Survey 1996 and Labourforce Survey September 2003. Pretoria. Databases on CD-Rom.
A particular problem with the housing programme was that, largely to reduce the costs of land, most new settlements were located far from economic centres. As a result, they reduced access to economic opportunities and employment.
Government also improved access to social grants considerably. These are particularly important to women, who are more likely to suffer from unemployment. Moreover, they were increasingly provided to mothers, which presumably went some way toward strengthening these women’s position within their households.
Table 15. Share of population receiving state pensions and social grants, 1996 and 2003
|
African |
Coloured/Asian |
White |
1996 |
|
|
|
State pensions (includes civil service pensions) |
15% |
13% |
8% |
Social grants |
2% |
8% |
1% |
% earning under R1000 |
46% |
36% |
8% |
2003 |
|
|
|
State old-age pension |
19% |
18% |
16% |
Disability grant |
5% |
6% |
5% |
Child support rant |
14% |
12% |
14% |
% earning under R1000 |
51% |
26% |
4% |
Source:Calculated from, Statistics South Africa. October Household Survey 1996 and Labourforce Survey September 2003. Pretoria. Databases on CD-Rom.
The state had far less success in supporting small and micro enterprise, which formed a major source of income especially for poor women. The following table demonstrates that self-employment was important for women’s survival strategies, especially in the former homeland areas. But they earned very little. Most self-employed women were in retail – essentially hawkers and spaza shop runners – and the vast majority made under R1000 a month.
Table 16. Self-employment by gender, region and sector, 2003
|
women |
men |
||
|
other areas |
primarily HL |
other areas |
primarily HL |
% of total |
|
|
|
|
Wholesale and retail trade |
63% |
73% |
45% |
38% |
Manufacturing |
10% |
13% |
9% |
9% |
Community, social and personal services |
14% |
5% |
10% |
7% |
Agriculture, forestry and fishing |
1% |
5% |
4% |
16% |
Construction |
1% |
1% |
12% |
20% |
Transport, storage and communication |
1% |
1% |
7% |
8% |
Financial and business services |
8% |
1% |
13% |
1% |
Total |
100% |
100% |
100% |
100% |
self employed as % of employed |
10% |
22% |
11% |
14% |
% of self employed earning under R1000 |
67% |
92% |
36% |
67% |
Note. 1. The data are from the rural areas of KwaZulu Natal, Mpumalanga, Limpopo, the Eastern Cape and Northwest. These areas are not identical to the former homelands, but are largely contiguous with them. Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
The RDP called for the transfer of 30% of land to smallholders. This commitment was repeated by the ANC in every elections manifesto. Still, by 2003, only 2,3% of land had been redistributed. Estimates suggested that the budget for land reform would have to expand tenfold to achieve the 30% target by 2014. (PBC 2004)
It is harder to summarise the impact of other programmes for smaller enterprise. Still, most observers argued that they were not very effective. (See PCAS 2003, p. 40)
Finally, until the early 2000s, the state did very little to compel financial institutions to serve small and micro enterprise. In 2002, at the insistence of community groups and organised labour, together with the formal financial institutions it engaged in the Financial Sector Summit. Ultimately, that process should lead to improved access for poor households to financial services.
However, the model proposed by the financial institutions under the Financial Sector Charter to extent facilities remained geared to wage earners rather than the self-employed. In particular, it would provide limited deposits and payments, primarily through electronic transfers. It would therefore effectively discriminate against black women, who are less likely to have formal jobs and more likely to be self-employed in remote rural areas.
Generally, government anti-poverty programmes were hampered by worries about cultivating a culture of dependency. In particular, the housing, infrastructure and land programmes increasingly insisted that, unless they were destitute, beneficiaries must contribute to costs. Obviously, this type of co-payment approach hits hardest at women, who tend to have the lowest incomes. When a co-payment was introduced in the housing programme in the early 2000s, it led to a severe slowdown in delivery. Most eligible households simply could not afford the down payment.
After 2000, government ameliorated this policy by requiring that municipalities provide free minimum services for households with incomes under R800 a month. As the following table indicates, however, the numbers of poor households paying for water, at least, appeared very little affected by this strategy.
Table 17. Percentage of households paying for water, 1996 and 2003
|
% paying for water |
|
Income group |
1996 |
2003 |
Others |
55% |
58% |
Indigent households (R500 p.m. in 1996, R1000 in 2003)* |
55% |
52% |
Total |
55% |
56% |
Note: R1000 in 2003 was about 25% higher than R500 in 1996. Source: 1996 October Household Survey and 2003 Labour Force Survey
In short, while government substantially improved access to basic services between 1994 and 2003, its efforts did not go far enough to overcome the marginalisation of most poor households. In particular, they did not suffice to raise incomes substantially in the former homeland areas.
2.2.2 The formal sector and women
With the failure to improve household productivity, the formal sector remained the main source of employment. But formal employment rose only about 1% a year from 1994, or around half the rate of growth in the population. The main reason for slow employment creation was the low level of growth overall combined with a long-term shift toward relatively capital-intensive sectors.
As the following table shows, compared to other middle-income developing countries, overall growth in the economy was slow and investment was very low through the late 1990s and early 2000s.
Table 18. Growth, investment and unemployment compared to other countries
|
GDP growth |
GDP per capita1 |
investment as % of GDP |
unemploy-ment rate |
|
1990-2001 |
2001 |
2001 |
1998-20012 |
South Africa |
2.1% |
10,910 |
15% |
23% |
Middle-income countries |
3.4% |
5,390 |
24% |
5% |
of which: |
|
|
|
|
Malaysia |
6.5% |
7,910 |
29% |
3% |
Chile |
6.3% |
8,840 |
21% |
10% |
South Korea |
5.7% |
15,060 |
27% |
4% |
Egypt |
4.5% |
3,560 |
15% |
8% |
Brazil |
2.8% |
7,070 |
21% |
10% |
Notes: 1. The GDP per capita is here calculated in terms of purchasing power parity, which seeks to measure actual output without taking exchange rate fluctuations into account. 2. The unemployment rate is given for one year between 1998 and 2001. Source: World Bank, Development Indicators 2003. Washington, D.C.
The main reasons for slow growth appear to have been restrictive fiscal and monetary policies, especially in the late 1990s, combined with substantial restructuring. In particular, the opening of the economy internationally with the transition to democracy led to a decline in sections of manufacturing. (See Roberts 2003)
The opening of the economy also increased pressure to maintain conservative macro-economic policies. Typically, this emerged in terms of demands that developing countries limit government spending, taxes and borrowing, and maintain high interest rates. The South African government argued in the late 1990s that international pressures required it to adopt these kinds of pressures, which it articulated in the “Growth, Employment and Redistribution” (GEAR) policy. (See Finance 1996)
These policies generally depressed economic expansion. The slowest economic growth occurred in the late 1990s, when government was cutting the budget and interest rates soared to over 20%. The economy picked up somewhat in the early 2000s, as the state relaxed its macroeconomic policies.
Slow growth in government spending affected women in two ways. First, it limited limiting job opportunities in the public service – especially education and health. Second, it reduced government’s scope for improving infrastructure and housing as well as support for SMEs and land reform.
In terms of structural policies, the state essentially embarked on an export drive, rather than concentrating on sectors that would create employment. Yet the dominant export sectors were all heavily capital intensive, and therefore unlikely to create either large number of jobs or opportunities for new enterprise. Between 1994 and 2002, the fastest growth was experienced in auto manufacturing and heavy chemicals, as well as platinum mining. The share of gold mining declined strongly.
The benefits of the export focus proved particularly limited for women. Employment in the main export sectors was strongly male dominated, as the following table shows. The top five export industries accounted for just over two thirds of all exports. But they accounted for only 10% of total employment, and only 2% of women’s employment.
Table 19. Exports and employment by sector
Sector |
% of exports |
change in export share |
women as % of em-ployment1 |
2003 percentage of: |
||
1994 |
2002 |
women's employment |
total em-ployment |
|||
mining and quarrying |
41% |
30% |
-11% |
5% |
0.03% |
0.1% |
machinery and equipment |
7% |
16% |
8% |
25% |
0.2% |
0.4% |
metals and metal products |
12% |
12% |
0% |
12% |
1% |
3% |
basic chemicals |
5% |
5% |
1% |
19% |
1% |
3% |
coke and petroleum |
2% |
5% |
3% |
27% |
0% |
8% |
subtotal |
67% |
68% |
1% |
12% |
2% |
14% |
All other industries |
33% |
32% |
-1% |
48% |
98% |
91% |
Note: Figures for 2003. Source: Figures on exports downloaded from TIPS EasyData on www.tips.org.za in March 2003. Figures on employment calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
In short, government policies toward the formal sector effectively encouraged heavy industry in a context of restrictive macroeconomic policies. The slow growth in employment that resulted meant that the marginalisation of women persisted. In these circumstances, anti-discrimination policies geared to the labour-market alone could do very little to improve women s economic position.
2.2.3 Union membership
Overall, women were less likely to belong to unions than men. Within industries, union density was somewhat lower for women than for men. Moreover, except for the social services, women generally worked in less organised sectors, including domestic work and retail. As a result, in 2003, only 28% of women in formal and domestic jobs[5] belonged to a union, compared to 36% of men.
Chart 4. Union membership by industry and gender, 20031
Note: 1. Excludes the self-employed. Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
This situation made it harder to enforce labour laws designed to protect workers. Labour legislation essentially depends on workers themselves, organised in unions, to monitor minimum standards and ensure improvements through collective bargaining. The state itself does not have capacity to enforce standards by inspecting workplaces. Moreover, beyond some minimum requirements, it is risky for the government to set conditions of employment, since it cannot evaluate the economic circumstances facing individual enterprises.
This system means that where unions are weak, the laws become much harder to enforce. Workers must rely on complaints to state inspectors, who are badly overstretched. The Employment Equity Act requires employers to consult with workers on their equity plans; if unions are weak, this consultation is harder to design and less likely to mobilise strong worker inputs.
2.2.4 Reproductive and productive labour
All over the world, women’s participation in paid labour has been hindered by their role in household labour. Most women take on the role of primary family care-giver, with the associated work of caring for children and other household members, cooking and cleaning.
In developing countries like South Africa, the prevalence of unemployment, especially for women, and the associated cheap labour mean that higher-income women can employ other women to undertake these tasks. For a privileged minority, this situation reduces the time pressure from household labour, as they employ (primarily African) women at a pittance. Given very high unemployment, as in South Africa, even relatively low-income households can often rely on help from un- and underemployed women relatives.
But the burden of household labour was aggravated by inadequate infrastructure in poor communities, particularly in rural areas. In the absence of water on site and electricity, women must spend far more time on cooking and cleaning. Almost 40% of African women in the former homelands, or 13% of all women, spent over an hour a week fetching water. On average, the chore took them around five hours a week. Some 25% of African women in the former homelands spent a similar amount of time, on average, fetching wood.
Table 20. Time spent fetching wood and water, 2003
|
African women in former homeland areas1 |
all others |
water |
|
|
under 1 |
63% |
93% |
1 to 7 |
31% |
6% |
8 to 14 |
5% |
0% |
over 14 |
1% |
0% |
wood |
|
|
under 1 |
75% |
97% |
1 to 7 |
21% |
3% |
8 to 14 |
4% |
0% |
over 14 |
1% |
0% |
% of population |
22% |
78% |
Source: Calculated from, Statistics South Africa. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
The HIV/AIDS pandemic added to the hours spent on household labour. Typically, women ended up caring for sick children and partners. That, in turn, made it harder for them to keep paid employment or run their own enterprises. Much of this labour would be made unnecessary if people with AIDS had access to anti-retrovirals, which permit most to remain healthy for much longer.
Finally, the housing programme also increased time spent on care giving. As the following table shows, subsidised housing in the urban areas was substantially more likely to be distant from basic amenities, including clinics, schools and welfare offices. That in itself added to the time women had to spend to obtain services for themselves and their children.
Table 21. Distance of urban subsidised housing from amenities, 2003
|
Clinic |
Hospital |
Primary School |
Secondary School |
Welfare Office |
Postal Services |
up to 30 minutes with subsidy without subsidy |
84% 87% |
60% 66% |
92% 94% |
86% 90% |
70% 78% |
78% 84% |
30 minutes to an hour with subsidy without subsidy |
14% 12% |
32% 30% |
8% 6% |
14% 9% |
27% 21% |
20% 15% |
60 minutes or more with subsidy without subsidy |
2% 1% |
8% 4% |
0% 0% |
1% 1% |
3% 1% |
2% 1% |
Source: Calculated from, September 2003 Labour Force Survey. Statistics South Africa. Pretoria. Database on CD-Rom.
3 Policy implications
The analysis provided here points to the need for policy initiatives in four directions.
First, all anti-poverty measures should be reviewed to ensure that they contribute as far as possible to higher productivity and incomes at the household level. That, in turn, requires affordable access to services at a level sufficient to engage with the economy.
Second, and most important, the government should do far more to support the growth of light industry and services. That means developing a broad vision for growth in specific industries, and co-ordinating supply-side measures, skills development and infrastructure to achieve it. Critically, these industries should be geared to meeting basic needs and reducing the cost of living as well as increasing exports and replacing imports.
Industries that could create employment especially for women include the public services; home-based personal services such as child care, hairdressing or catering; the retail industry and tourism; transport; light manufacturing (assembly of appliances, food processing, clothing and textiles, furniture, plastics and so on); and industries downstream from metal and chemicals production.
Third, in the labour force, employment equity and decent conditions seem unlikely to work for most in the absence of stronger organisation. In sectors that are difficult for unions, it might be possible to consider mobilisation through service or community-based organisations.
In addition, existing anti-discrimination legislation, in particular the Employment Equity, Skills Development and Broad-Based Black Economic Empowerment Acts, should be reviewed systematically. The Employment Equity Act appears to have had little real impact on employment patterns, especially for lower-level workers. As noted above, too, working women still have noticeably less access to skills development.
The new sectoral BEE Charters, arising from the Broad-Based Black Economic Empowerment Act, set more ambitious targets. Moreover, since they are linked to government procurement, they have a stronger enforcement mechanism. However, the charters drafted so far set very low targets for black women in senior management – just 4% in the Financial Sector Charter. Moreover, they typically do not provide for promotions for lower-level workers.
Finally, an effective policy to advance women in the economy must look beyond the labour market. In particular, more should be done to ensure women have access to productive assets, skills and education and adequate housing and infrastructure. That means above all extending and strengthening programmes to support micro enterprise and land reform.
Ultimately, women cannot be empowered unless the economy as a whole is restructured toward more equitable employment-creating growth. Given the inherited economic structure, anti-discrimination legislation necessarily ends up benefiting only the small high-level group.
References
Altman, M. 2003. “Employment Trends and Policy Implications.” Paper presented to TIPS/DPRU Annual Forum, Johannesburg, August.
ANC. 1994. Reconstruction and Development Programme. Johannesburg. Downloaded from www.anc.org.za. 2001.
Bhorat, H. 2002. “Employment Trends: Has the Economy Created Jobs Since GEAR?” in, South African Labour Bulletin XXVI.1 (February). Johannesburg.
COSATU. 2003. Employment Creation and Investment. Draft Labour Position Paper for the Growth and Development Summit. Downloaded from www.cosatu.org.za. August 2003.
COSATU. 2003. Labour Position Paper for the Growth and Development Summit. Downloaded from www.cosatu.org.za. August 2003.
De Swardt, Cobus. 2003. “The Shadow of the Rainbow Nation: Chronic Poverty a Decade After Liberation.” Draft paper for CPRC. Available from cdeswardt@uwc.ac.za
Devey, Richard, Caroline Skinner and Imraan Valodia. 2002. “The Informal Economy in South Africa: Who, Where, What and How Much?” Paper presented to DPRU Second Annual Conference on Labour Markets and Poverty, October 2002. Johannesburg.
DTI. 2001. Driving Competitiveness: An Integrated Industrial Strategy for Sustainable Employment and Growth. www.dti.gov.za. Downloaded 2001.
Finance. 1996. Growth, Employment and Redistribution: A Macroeconomic Strategy. www.finance.gov.za. Downloaded 2000.
Lewis, Jeffery D. 2001. “Policies to Promote Growth and Employment in South Africa.” World Bank Southern Africa Informal Discussion Paper No. 16. (Pretoria)
Makgetla, Neva. 2004a. “Employment and Economic Structure,” in, NALEDI Policy Bulletin May 2004. Available on www.naledi.org.za
Makgetla, Neva. 2004b. “The Post-Apartheid Economy,” in, Review of African Political Economy, forthcoming 2004.
Makgetla, Neva. 2004c. “Unpacking the Unemployment Data.” Downloaded from www.cosatu.org.za
National Treasury. 2004. National Budget Review. Pretoria. Downloaded from www.treasury.gov.za in March 2004.
SARPN (SA Poverty Research Network). 2004. Readings on the Second Economy. Pretoria. Downloaded from www.sarpn.org.za July 2004.
StatsSA. 2000. Labourforce Survey, September. Electronic database. CD-Rom. Pretoria.
StatsSA. 1996. October Household Survey 1995. Pretoria.
StatsSA. 2001. A Survey of Time Use. Pretoria. www.statssa.gov.za. Downloaded in September 2003.
StatsSA. 2002. Earning and Spending in South Africa. Pretoria. www.statssa.gov.za. Downloaded in December 2002.
StatsSA. 2003a. Labourforce Survey, September. Electronic database. CD-Rom. Pretoria.
StatsSA 2003b. Survey of Earnings and Employment. Long-term data series. www.statssa.gov.za. Downloaded September 2003.
StatsSA. 2003b. Key findings of Census 2003. www.statssa.gov.za. Downloaded August 2003.
Terreblanche, S. 2002. A History of Inequality in South Africa 1652-2002. University of Natal Press. Pietermaritzburg.
Wittenberg, Martin. 2002. “Formal and Informal Work in South Africa: Evidence from the Time Use Survey.” Paper presented to DPRU, UCT, Second Annual Conference on Labour Markets and Poverty in South Africa.
World Bank. 2003. World Development Indicators.Washington, D.C. USA
[1]We here use “black” for Africans, Coloureds and Asians.
[2]The Employment Equity Commission also published data on occupation by race and gender from 2000. They also found very little substantial progress, but their overall figures vary substantially from the LFS findings. Their data derive, however, from employment equity reports filed only by larger formal companies. This process means the companies themselves essentially define occupational levels. In contrast, the LFS is a survey of around 30 000 households with more consistent occupational categorization. For this reason, we here use only the LFS data.
[3]The r2 is 35% and the regression is significant at higher than 95%. That means the correlation is strong and explains 35% of the difference between sectors in the share of workers earning under R1000 a month. The structure of the equation was Y = k + AW, where Y = % of workers in sector earning under R1000, k is a constant, and AW = % of African workers in the sector. The final result was Y = 0.084 + 0.844AW. The relationship is substantially stronger if the two professions traditionally dominated by women – healthcare and education – are excluded. The regressions used data from, Statistics SA. Labour Force Survey September 2003. Pretoria. Database on CD-Rom.
[4]The concept of dualism originally arose to describe colonial economies where pre-colonial economic systems persisted, although distorted by colonialism. Many economists contest the view that pre-colonial modes of production survived the colonial experience anywhere. In any case, it is clear that in South Africa, pre-colonial economies were destroyed by colonialism and apartheid. The former homeland areas maintain some nominally traditional trappings in their governance structures, but the economies were shaped by landlessness and overcrowding imposed by the state.
[5]Excluding the self-employed.
Warning: count(): Parameter must be an array or an object that implements Countable in /home/gwdhmoih/public_html/templates/gk_news2/html/com_k2/templates/default/item.php on line 176
Gary Watkins
Gary Watkins
Managing Director
BA LLB
C: +27 (0)82 416 7712
T: +27 (0)10 035 4185 (Office)
F: +27 (0)86 689 7862
Website: www.workinfo.comRelated items
- Don't blame women for leaving fields like engineering. Blame bad attitudes
- Anger over anti-Indian quotas for Durban's top jobs
- Training out sexism in the workplace by Devan Moonsamy, CEO of the ICHAF Training Institute
- 2017 EE online and manual reporting season
- Public shaming of workplace harassers may force employers to stop protecting them
Latest from Gary Watkins
- National and Regional Economically Active Population Profile QLFS Q3:2021
- National and Regional Economically Active Population Profile QLFS Q2:2021
- Consolidated Directions on Occupational Health and Safety Measures in certain workplaces as at 11 June 2021
- COVID19TERS Benefits as at 20 July 2021
- Adjusted Level 3 Lockdown - 25 July 2021