Printer Friendly

Annex 3: Gender and labor markets in Zambia and Ghana: an analysis of household survey data.

1. This note provides an overview of characteristics of labor markets in Zambia and Ghana based on the Household Surveys1. It also examines the links between labor market outcomes and gender. Section I deals with the supply of labor, and the determinants of labor force participation. Section II is concerned with the absorption of labor--the sectoral pattern of employment--in household activities and market activities. Section HI discusses the issue of gender discrimination in the labor market and Section IV concludes with a brief summary of major findings and implications.

1. The Supply of Labor

2. Table 1 examines the labor force participation rates in Ghana and Zambia by age groups, gender, education level, and rural/urban sectors. The rates were more or less the same for men and women. In Zambia, the rates are considerably higher in the 25-60 year age group (productive age group) for males, but in Ghana they were almost identical. As observed in other countries, the participation-age profile for both genders is concave, as younger and older individuals have lower participation rates than their prime aged counterparts. Participation rates in urban areas are slightly higher than in rural areas. Similarly the participation rates among the poor are somewhat less than among the non-poor--though not significantly so.

II. The Demand for Labor

3. This section analyzes the absorption of labor in different segments of the labor market, the distribution of formal and informal workers, the public-private composition of wage workers--especially formal workers. The key questions are: Which sector is the main provider of wage employment? Does female employment distribution differ from that of males? Is agriculture still the main source of employment in rural areas? What proportion of individuals are employed in the formal and informal sectors? What proportion of the labor force is unemployed?

4. Table 2 examines the distribution of workers by sector of employment. In Ghana, over 50 percent of workers are employed in agriculture while another 28 percent each are employed in services. There are significant differences by gender. For example, 21 percent of women are engaged in trading compared to only 4 percent of men. 30 percent of males work in services, a sector which employs over 23 percent of women.

5. Table 3 examines the distribution of the workforce by sector of employment. The largest sectors are self-employment (79 percent) followed by the public sector (9 percent). Again, not surprisingly, there are significant gender differences. Women are much more likely to be self-employed than men.

6. Men are more likely to be employed in the formal (2) sector while women are more likely to be self-employed in agricultural or services sector (Table 2). In Ghana, over 95 percent of female workers are employed in the informal sector compared with 90 percent in Zambia (Table 4). Most of the female workers are engaged in agriculture, trading and services sector--small-scale business activities in the informal sector.

7. This section examines earnings (3) by sector of employment, area and education levels. On average, males earn close to 48 percent more than females (Table 5). Looking at the mean earning of males and females by sector of employment, public sector workers earn significantly higher incomes than those in any other sector. Through the regression functions below we shall see whether public sector workers are in fact actually overpaid, or whether their higher earnings are a result of their human capital characteristics--higher levels of education and experience.

8. If we look at the earnings of workers by level of education, we see the familiar pattern of increasing returns to education (Table 6). Individuals with no education are the lowest paid workers while those with university and post-graduate education are the highest paid. Women earn significantly less than their male counterparts with the same levels of education, though the difference narrows with increasing levels of education. This is partly because females predominantly work in the informal sector where the pay is much lower than in the formal sector, and also because of discrimination (see below).

III. Discrimination in the Labor Market

9. As in most other developing countries, female labor is paid less than male labor in Ghana and Zambia. While a portion of this male-female wage differential may be explained due to differences in individual characteristics (human capital accumulation and labor market experience) and employment characteristics (relative to women, males predominate in the high-paying formal sector and work in different industries and occupations), a portion of the differential may still remain unexplained. This portion measures an upper bound on wage discrimination against women. Using Oaxaca's (1973) technique we can decompose the pay gap between males and females into these two components.

10. Assuming the male and female earnings regressions to be:

ln[W.sub.m] = [C.sub.m] + ([X.sub.m]) [b.sub.m] + [[epsilon].sub.m], (1)

ln[W.sub.f] = [C.sub.f] + ([X.sub.f]) [b.sub.f] + [[epsilon].sub.f], (2)

where the subscripts'm' and 'f refers to males and females respectively; In (W)'s are the log of earnings, C's are the constants terms, X's are a vector of characteristics, b's are the coefficients and e's are the error terms. The difference in the average log of earnings is equivalent to the percentage difference between male and female pay. Given that the error term in the male and female earnings function are mean zero, we can show that:

ln[W.sub.m] - ln[W.sub.f] = ([C.sub.m] - [C.sub.f]) + [([X.sub.m])[b.sub.m] - ([X.sub.f])[b.sub.f]] (3)

where [X.sub.m] and [X.sub.f] are the average values of male and female characteristics in the sample. Re-arranging this equation we get:

ln[W.sub.m] - ln[W.sub.f] = ([C.sub.m] - [C.sub.f]) + [X.sub.f]([b.sub.m] - [b.sub.f] + [[b.sub.m]([X.sub.m]) [X.sub.f])] (3)

11. The difference in pay comes from two different sources. The first term represents the differential rewards to male and female characteristics in the labor market, while the second term represents the differences in the quantities of these characteristics. The portion of the wage gap arising out of differences in quantity of characteristics can be thought of as not being discriminatory or as "justified discrimination." However, the portion of the wage gap arising out of different rewards to male and female characteristics can be thought of as the upper bound of "unjustified" wage discrimination.

12. The male-female earnings gap is around 26 percent in Ghana. In Table 7, we examine the Oaxaca decomposition. Our results are somewhat startling. The nondiscriminatory part of the earnings actually reduced differentials by 69% but the discriminatory portion of the gap increased differentials by 169 percent. This can be seen to be the upper bound of discrimination. This means that if both male and female workers had the same characteristics (i.e. [X.sub.m] = [X.sub.f]), the earnings differential would be 1.69 times higher than the present level due to discrimination. If there was no discrimination ([b.sub.m] = [b.sub.f]), me earnings differential would only be 69 percent of the present level, in other words the differential would be reduced by 31 percent. This is mainly because men's and women's characteristics are not identical. We find similar results in Zambia, but the unjustified component of earnings differentials is slightly smaller.

IV. Conclusions

13. Briefly summarized, some of the main conclusions that have emerged are:

* Labor force participation rates of women are almost the same as that of men. Female labor force participation rates are about the same as male participation rates in Ghana. In Zambia, female labor force participation rates are lower than those of males because a substantial number of women are unpaid workers. If the activities most commonly associated with women, especially poor women--unpaid household work--are formally classified as labor force activities, their participation rates are similar to those of males. However, these low participation rates may also be due to the significant discrimination against women in the labor market (see below).

* Female workers are disproportionately employed in the informal sector. In Ghana over 95 percent of female workers are employed in the informal sector (90 percent in Zambia). Most female workers are engaged in agriculture, trading and the services sector--small-scale business activities in the informal sector. These workers often lack access to basic infrastructure such as water, power, telecommunications and transport and their access to credit--which is crucial in enabling them to utilize more efficient technologies--is limited.

* There is discrimination against women in the labor market. There is significant discrimination in the labor market. Women earn significantly less than do their counterparts and this leads to inefficient allocation of resources. These groups are less likely to invest in education and human capital accumulation, and less likely to participate in labor market activities. Furthermore, as these groups are among the poorest in the economy--efforts to reduce poverty will be hindered until this issue is addressed more thoroughly.

C. Mark Blackden

Chitra Bhanu

in collaboration with the Poverty and Social Policy Working Group of the Special Program of Assistance for Africa

(1) This annex is based on analysis of the 1992 Ghana LSMS data and the 1996 LCMS survey of Zambia, and was prepared by Saji Thomas, Poverty Reduction and Social Development Group, Africa Region. World Bank.

(2) A person is classified as employed in the formal sector if s/he is a wage worker, receives benefits, or works in a firm which has a onion.

(3) The earnings were adjusted for regional price differences to make them comparable across regions. Earnings include income from wages, self-employment, food, housing, clothing, and transportation allowance. Average earnings are only for those employed.
Table 1: Labor force participation rates (in percent)

 Ghana Zambia
Category Male Female All Male Female

Area
Urban 63.5 64.8 64.2 58.5 59.6
Rural 79.1 80.0 79.2 61.7 38.1
Poverty Groups
Poor 68.6 71.7 70.3 57.4 43.1
Non-poor 75.4 75.0 75.2 61.5 42.3
Age-groups
15-19 yrs 46.8 44.8 45.9 23.9 28.1
20-24 yrs 59.6 66.3 63.2 58.5 47.7
25-29 yrs 79.7 81.8 80.9 78.2 55.4
30-39 yrs 89.2 85.5 87.1 75.2 54.6
40-49 yrs 89.8 87.8 88.7 67.1 51.4
50-59 yrs 89.9 79.2 83.7 60.4 50.4
60-69 yrs 82.7 72.6 78.3 87.8 74.2
Education level
No education 94.4 62.5 79.4 12.1 22.5
Primary 71.8 75.3 73.7 88.7 64.8
Secondary 56.2 54.9 55.7 92.7 62.4
Higher 77.1 76.6 76.3 97.5 89.3

Total 73.5 74.1 73.8 59.6 47.4

Category All

Area
Urban 48.38
Rural 60.37
Poverty Groups
Poor 50.1
Non-poor 51.9
Age-groups
15-19 yrs 26.1
20-24 yrs 52.8
25-29 yrs 66.6
30-39 yrs 64.7
40-49 yrs 59.6
50-59 yrs 55.3
60-69 yrs 81.0
Education level
No education 17.5
Primary 75.2
Secondary 79.7
Higher 94.8

Total 53.4

Table 2: Distribution of workers by industry
and gender (in percent)

 Ghana

Industry Male Female Both

Agriculture 54.0 48.2 50.9
Mining & 1.01 0.13 0.54
Utilities
Manufacturing 10.7 7.73 7.46
Trade 3.85 20.8 13.0
Services 30.4 23.1 28.1
All 100.0 100.0 100.0

 Zambia

Industry Male Female Both

Agriculture 22.2 25.9 24.1
Mining & 3.2 0.33 1.74
Utilities
Manufacturing 5.74 1.88 3.78
Trade 8.26 7.9 8.12
Services 60.6 63.9 62.3
All 100.0 100.0 100.0

Table 3: Distribution of workers by sector and gender (in percent)

 Ghana Zambia

Industry Male Female Both Male Female Both

Public 14.1 4.9 9.3 24.5 10.3 18.1
Private 14.0 5.5 6.5 17.9 6.13 12.6
enterprise
Self-employed 70.0 86.9 79.0 46.9 46.5 46.7
Unpaid labor 1.8 2.7 2.2 10.6 37.0 22.6

All 100 100 100. 100. 100.0 100.0

Table 4: Distribution of workers by formal/informal
sector and gender (in percent)

 Ghana

 Male Female Both

Informal 84.6 95.4 90.4
Formal 15.4 4.6 9.6
All 100.0 100.0 100.0

 Zambia

 Male Female Both

Informal 73.1 89.5 81.5
Formal 26.9 10.5 18.5
All 100.0 100.0 100.0

Table 5: Mean Yearly Earnings by sector and gender

 Ghana (1992 cedis) Zambia (1996 Lws)

Education level Male Female All Male Female

Public 354,442 319,440 344,534 134,422 107,587
Private 189,393 70,876 151,832 57,740 21,845

All 271,310 184,322 245,201 102,853 37,137

 Zambia (1996 Kws)

Education level All

Public 127,718
Private 38,179

All 73,610

Table 6: Mean yearly Earnings by education
levels (cedis)

 Ghana (1992 cedis)--Yearly

Education level Male Female All

None 7134 -- 7134
Primary 168514 60994 141193
Secondary 237632 175315 219801
Higher 353229 314223 339473

All 271310 184322 245201

 Zambia (1996 Kws)--Monthly

Education level Male Female All

None 42848 10937 22055
Primary 46346 16807 32172
Secondary 126732 81288 112508
Higher 392636 198342 332349

All 102853 37137 73610

Table 7: Decomposition of Gender Earnings Differential

 Ghana

Differential Justified Unjustified Total

Differences in log wages -0.14 0.44 0.26
in % terms 69 169 100

Table 7: Decomposition of Gender Earnings Differential

 Zambia

Differential Justified Unjustified Total

Differences in log wages -0.58 0.92 0.34
in % terms 58 158 100
COPYRIGHT 1999 The World Bank
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Gender Growth and Poverty Reduction: Special Program of Assistance for Africa, 1998 Status Report on Poverty in Sub-Saharan Africa
Publication:Gender, Growth and Poverty Reduction
Date:Mar 1, 1999
Words:2336
Previous Article:Annex 2: The interface between time allocation and agricultural production in Zambia: a case study 1.
Next Article:Annex 4: Gender inequality and growth: note on data and methodology.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters