Front Cover
 Table of Contents
 Conceptualization and measurement...
 Poor women in the urban econom...
 Women and transport: A case...
 Women's role in integrated urban...
 Appendix I. Transport pilot study...
 Appendix II. Additional tables...
 Appendix III. Project-related survey...

Group Title: plight of poor women in the Latin American metropolis
Title: The plight of poor women in the Latin American metropolis : an exploratory analysis of polciy issues
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00087141/00001
 Material Information
Title: The plight of poor women in the Latin American metropolis : an exploratory analysis of polciy issues
Physical Description: Book
Language: English
Creator: Schmink, Marianne
Publisher: Population Council
Place of Publication: New York
Publication Date: March, 1980
Copyright Date: 1980
 Record Information
Bibliographic ID: UF00087141
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: 233038023 - OCLC

Table of Contents
    Front Cover
        Front Cover
        Acknowledgement 1
        Acknowledgement 2
    Table of Contents
        Table of Contents 1
        Table of Contents 2
        Page A-1
        Page A-2
        Page A-3
        Page A-4
        Page A-5
        Page A-6
    Conceptualization and measurement of women's economic roles
        Page B-1
        Page B-2
        Page B-3
        Page B-4
        Page B-5
        Page B-6
        Page B-7
        Page B-8
        Page B-9
        Page B-10
        Page B-11
        Page B-12
        Page B-13
        Page B-14
        Page B-15
        Page B-16
        Page B-17
        Page B-18
        Page B-19
        Page B-20
        Page B-21
        Page B-22
        Page B-23
        Page B-24
    Poor women in the urban economy
        Page C-1
        Page C-2
        Page C-3
        Page C-4
        Page C-5
        Page C-6
        Page C-7
        Page C-8
        Page C-9
        Page C-10
        Page C-11
        Page C-12
        Page C-13
        Page C-14
        Page C-15
        Page C-16
        Page C-17
        Page C-18
        Page C-19
        Page C-20
        Page C-21
        Page C-21a
        Page C-21b
        Page C-21c
        Page C-22
        Page C-23
        Page C-24
        Page C-25
        Page C-26
        Page C-27
        Page C-28
        Page C-29
        Page C-29a
        Page C-30
        Page C-30a
        Page C-30b
        Page C-30c
        Page C-30d
        Page C-31
        Page C-32
        Page C-33
        Page C-34
        Page C-35
        Page C-36
        Page C-37
        Page C-38
        Page C-39
    Women and transport: A case study
        Page D-1
        Page D-2
        Page D-3
        Page D-4
        Page D-5
        Page D-6
        Page D-7
        Page D-8
        Page D-9
        Page D-10
        Page D-11
        Page D-12
        Page D-13
        Page D-13a
        Page D-13b
        Page D-13c
        Page D-13d
        Page D-13e
        Page D-13f
        Page D-13g
        Page D-13h
        Page D-13i
        Page D-14
        Page D-14a
        Page D-15
        Page D-15a
        Page D-15b
        Page D-15c
        Page D-15d
        Page D-15e
        Page D-16
        Page D-17
        Page D-18
        Page D-19
        Page D-20
        Page D-21
        Page D-22
        Page D-23
        Page D-24
        Page D-25
        Page D-25a
        Page D-26
        Page D-27
        Page D-27a
        Page D-28
        Page D-28a
        Page D-28b
        Page D-29
        Page D-30
        Page D-30a
        Page D-30b
        Page D-30c
        Page D-31
        Page D-31a
        Page D-31b
        Page D-32
        Page D-32a
        Page D-32b
        Page D-33
        Page D-33a
        Page D-34
        Page D-35
        Page D-35a
        Page D-36
        Page D-36a
        Page D-36b
        Page D-36c
        Page D-36d
        Page D-37
        Page D-37a
        Page D-38
        Page D-38a
        Page D-38b
        Page D-39
        Page D-39a
        Page D-39b
        Page D-39c
        Page D-40
        Page D-41
        Page D-42
        Page D-43
        Page D-44
        Page D-45
        Page D-46
        Page D-47
        Page D-48
    Women's role in integrated urban service programs
        Page E-1
        Page E-2
        Page E-3
        Page E-4
        Page E-5
        Page E-6
        Page E-7
        Page E-8
        Page E-9
        Page E-10
        Page E-11
        Page E-12
        Page E-13
        Page E-14
        Page E-15
        Page E-16
        Page E-17
        Page E-18
        Page E-19
        Page E-20
        Page E-21
        Page E-22
        Page E-23
    Appendix I. Transport pilot study methodology
        Page F
        Page F-1
        Page F-2
        Page F-3
        Page F-4
        Page F-5
        Page F-6
        Page F-6a
        Page F-6b
        Page F-7
        Page F-8
        Page F-9
        Page F-10
        Page F-11
        Page F-12
        Page F-13
        Page F-14
    Appendix II. Additional tables from the transport pilot study
        Page G
        Page G-1
        Page G-2
        Page G-3
        Page G-4
        Page G-5
        Page G-6
        Page G-7
        Page G-8
        Page G-9
        Page G-10
        Page G-11
        Page G-12
        Page G-13
        Page G-14
        Page G-15
        Page G-16
        Page G-17
    Appendix III. Project-related survey instruments
        Page H
        Page H-1
        Page H-2
        Page H-3
        Page H-4
        Page H-5
        Page H-6
        Page H-7
        Page H-8
        Page H-9
        Page H-10
        Page H-11
        Page H-12
        Page H-13
        Page H-14
        Page H-15
        Page H-16
        Page H-17
Full Text


I~iwt .1*



Preliminary draft.
pt for quotation or reproduction ,, .
t;hqt the author's permission. .
Thefrathor welomes commeutewtl cr1fticisms


, I



A great many people have assisted me in one way or another in the

preparation of this report. I have tried to acknowledge their specific

contributions throughout the following chapters, and I apologize for my

inability to do justice to the generous contributions of ideas and infor-

mation which have helped me in organizing the study. Here I will limit

my thanks to those individuals and institutions who contributed most

directly to my research endeavor, and who do not necessarily appear in the

remaining pages.

My conversations with researchers and planners which provide the back-

drop for the report have been carried out over a period of several years,

including during some earlier trips when I was employed as a consultant

for the Ford Foundation's office in Brazil. My work was also significantly

aided by participation in the IUPERJ-sponsored conference on "Women in the

Labor Force in Latin America," November, 1978, made possible through travel

support contributed by the International Communications Agency of the Ameri-

can Embassy, Brazil. Alice de Paiva Abreu provided assistance in reviewing

official data sources on women in Brazil. Dr. Thomas Merrick generously

provided me with tables from the 1972 PLAMBEL survey of Belo Horizonte.

My pilot study of urban transport was aided by the generosity of Dr.

Peter Watson of the World Bank, who provided me with the tables analyzed

in Chapter 4, section B. The Brazilian Urban Transport Firm (EBTU) and

Belo Horizonte's metropolitan planning agency (PLAMBEL) also provided me

with valuable information and insights into transport planning. The pilot -

study would not have been possible without the help of my local field

research team members, Herve, Mana, Nilmario, and Virgilio and the support

of the Jornal dos Bairros. Don and Diana Sawyer gave me the hospitality of

their home during the research period and Diana provided useful guidance on

sampling procedures. Lynne Foweraker and Chris Krueger helped me code the

results of the fieldwork, and Jack Dixon provided the expertise to facili-

tate computer analysis.

The Center for Latin American Studies, at the University of Florida,

cheerfully provided infrastructural, clerical, and administrative support

during the grant period. Finally, staff members of the Population Council

deserve my gratitude for their unfailing enthusiasm in all phases of the

work. Typing and details of production of this report were in the competent

hands of Philip Saltz. Axel Mundigo has provided crucial financial and moral

support at every turn. My biggest debt, however, is to Judith Bruce, who has

been a constant source of stimulation, inspiration, and direction. Her in-

fluence has marked every page of this report. All of its shortcomings however,

must remain my own responsibility.

Table of Contents


Chapter 1

Chapter 2

Chapter 3


Conceptualization and Measurement of Women's Economic Roles
A. Conceptual Problems
1. Concepts of economic activity
2. Concepts of the household
3. Concepts of the family
B. Collection of Data on Women's Economic Roles
1. Methods
2. Personnel
3. Respondents
4. Population census
5. Economic census
6. Continuing work statistics
C. National sources of Data on Women in Brazil
1. Population census
2. National Household Sample Survey (PNAD)
3. National Study of Family Expenditures (ENDEF)
4. Economic statistics
5. Annual Social Information Report (RAIS)
6. IBGE Department of Social Indicators (DEISO)
7. Other surveys
a. Transport surveys
b. Metropolitan household surveys
c. Non-official surveys
Notes to Chapter 2

Poor Women in the Urban Economy
A. Overview of Urban-Industrial Development in Latin America
B. A Framework for Understanding Women's Economic Behavior
C. Women's Access to Urban Services
1. Education
2. Child-care services
3. Health services
4. Housing and infrastructural services
Noes to Chapter 3

Women and Transport: A Case Study
A. Overview of Urban Transport Planning in Latin America
B. Sources of Data on Urban Transport
C. Pilot Study of Women and Transport
1. Overview of the study
2. The study neighborhoods in the Belo Horizonte urban context
3. Employment and unemployment
4. Trip to work and home again

Chapter 4

5. General transport problems
6. Transport and education
7. Transport and shopping
8. Transport and health
9. Women's travel patterns
D. Conclusions and Policy Recommendations
Notes to Chapter 4

Chapter 5

Appendix I

Appendix II

Appendix II

Women's Role in Integrated Urban Service Programs
A. Models of Community Participation
1. The modified "top-down" approach
2. Building bridges to local communities
3. Building community capacity
B. Women's Participation in Community Programs
C. Projects as Sources of Information on Women
D. Summary and Conclusions
Notes to Chapter 5

Transport-Pilot Study Methodology

Additional Tables from the Transport Pilot Study

I Project-Related Survey Instruments

SChapter 1


In the latter part of 1979 several regional meetings were held by the

United Nations to evaluate the progress which had been made in improving

the status of women in developing areas since the 1975 World Conference

for the International Women's Year. The conclusion of these meetings,

held in preparation for the 1980 World Conference of the United Nations

Decade for Women, was that in most countries the situation of women in

rural areas and in the so-called marginal urban sectors has worsened,

rather than improved, during the first five years of the United Nations

Decade for Women. In order to help remedy this deteriorating situation,

the meetings called for more adequate data collection, research, and

analysis of women's situation in these contexts.1

In Latin America, urbanization has proceeded more rapidly than in the

other developing areas, provoking particular problems for women in the ur-

ban milieu. Mexico City and Sao Paulo have doubled in size within the

last fifteen years and are becoming th largest metropolises in the world.

The impact of this urbanization process on women's roles, family composition,

household economic strategies, and the quality of urban life is largely

unassessed, impeding the formation of an adequate foundation for planning

of basic services in urban areas which can answer the needs of poor women

and their families. Yet certain trends suggest that special attention

should be paid to women's roles in Latin American metropolitan settings:

(1) women dominate in migration streams; (2) a high and increasing pro-

portion of households in urban areas are headed by women; (3) low in-

come women, partnered and unpartnered, migrant and non-migrant, play a crit-

ical role in the generation of income for family survival.


These factors draw attention to the fact that women may often predomin-

ate in disadvantaged urban groups. At the same time, their access to both

employment and services may be quite different from that of low income men's,

implying a need for special consideration of women's roles on the part of

planners. First, women in poor families have an important long-term role

in generating income, but their access to labor markets and to the training

associated with gainful employment is drastically different from men's.

Second, women also balance multiple roles in addition to their role in in-

come-generation. These include child-bearing and child-rearing, and house-

hold management and provisioning. Third, in their household management role

women manipulate extra-domestic groups and networks, and deal with official

agencies, in order to.negotiate access to services for themselves and their

families. All of these activities can be crucial both for the short-term

welfare of poor families as well as for the long-term integration of poor

populations into the urban system. For these reasons, women's access to

urban resources, services and employment deserve special attention in the

design of urban policies.

At the same time, adequate information for assessing women's access

to urban resources is rarely available. Even the evaluation of women's

work patterns is impeded by data sources which are often not sensitive to

the particular characteristics of women's employment patterns. And whereas

a number of studies have focused on patterns of women's employment, rela-

tively little is understood about the role played by poor women in a

variety of other activities which are also critical to family welfare.

These include unpaid labor inputs, the manipulation of extra-domestic net-

works, the general day-to-day management of household strategies, and the

utilization of collective urban services. In particular, next to nothing


is known about women's access to a broad range of urban services, and the

role they may play in facilitating income generation and improving welfare

in both short- and long-term.

This report is intended to contribute to filling in the gaps in know-

ledge about these crucial aspects of the situation of poor women in the

Latin American metropolitan areas. It is composed of a series of explora-

tory studies which deal with different aspects of the plight these women

confront daily and over their lifetime. For the most part, its chapters

represent an effort to piece together diverse sources of information in

order to draw out the essential outlines of the situation of low income

women. Analysis is intentionally focused on those aspects which are least

well-documented and understood, making it difficult to come to firm conclu-

sions. At the same time, the analysis is carried out within a coherent and

comprehensive framework which does not treat women as isolated individuals

but always as persons integrated into larger family and community units.

Overall, the multiple aims of the report might be summarized as follows:

to explore the outlines of women's role in the urban economy of Latin America

and provide an analytical framework for understanding this multifaceted role;

to draw together scanty information related to women's access to urban ser-

vices and the implications of these findings for the welfare of poor women

and their families; and to point to policy directions which arise out of

the tentative findings of these exercises. It is hoped that this report

will provide stimulating ideas and information for planners and researchers

concerned with practical ways of dealing with poor urban women's problems.

Chapter 2 provides an introductory discussion of the use of existing

data bases for the analysis of sex-differentiated patterns of behavior in -

urban areas. It begins with an overview of the major problems in concep-

- 4 -

tualization and data collection in official statistics in Latin America,

)which provides a background for the evaluation of studies of women's roles

relying on officially-produced statistics. The concluding section presents

a listing and brief description of the major sources of statistics relevant

to the analysis of urban women in Brazil.

Chapter 3 places in perspective the situation of poor women in the

large urban centers of Latin America. It begins with an overview of the

impact of urban-industrial development on women's employment opportunities

in the region. With this background, a conceptual framework is presented

within which women's economic activities may be more clearly understood.

It is argued that women's productive activities are best analyzed as one

component of complex economic strategies carried out at the household level.

The form of these strategies will vary according to the internal evolution

lof the domestic group and will be largely shaped by the household's position
in a specific socioeconomic and spatial context. Women's important roles

in household strategies determine not only their labor force behavior but

also other aspects of community and household-level manipulations of re-

sources crucial to family welfare.
The final section of the chapter focuses on women's utilization of

urban services and the role they potentially play in facilitating income

generation, improving household and community welfare in the short-term, and

furthering the long-term integration of poor populations. It is argued that

future policy interventions must be based on the analysis of how use patterns

differ by sex, what the implications of these differences are for women and

their families, what factors lead to differential use patterns, and what

kinds of policies might best serve to correct problems of essential service

delivery. The chapter presents an assessment of the state of information

on women's access to selected types of urban services in Latin America, in-


cluding education, child care, health, and housing and infrastructural ser-


The case of the urban transport sector is taken up in more detail in

Chapter 4. Despite the importance of transport services from the point of

view of planners as well as low income populations, this sector represents

relatively unexplored terrain with regard to women's special needs, and in-

deed the needs of poor populations in general. The discussion outlines the

most relevant issues in transport planning in Latin America, particularly

for women, and discusses existing sources of information on urban transport.

Some experimentation with special tabulations from Latin American survey

data (Bogota) is carried out.

The remainder of.the chapter presents the findings of a pilot study

of transport conditions in three poor neighborhoods in Belo Horizonte, Brazil,

Carried out in 1979. The objectives were: 1) to raise hypotheses about women's

transport patterns and how their urban transport needs might be better ad-

dressed by transport planners; and 2) to test a methodology for gathering

information from the perspective of the user on transport needs and patterns

of poor families, especially women. The multi-neighborhood research design

used in the study permitted the comparison of patterns of employment, service

utilization, and women's travel behavior in low-income contexts with distinct

transport conditions. The concluding section of the chapter draws on the

most important trends emerging from the pilot study to suggest concrete

steps which might be taken to improve conditions in poor, peripheral neigh-


The final chapter of the report is an exploratory essay more directly

concerned with the design of programs directed at low-income poor populations.

It focuses on models of community-oriented service programs, presenting a

- 6 -

typology of the programs encountered in the Brazilian context. The analysis

then turns to the question of women's potential contribution to such programs,

arguing that women represent an ideal target group for programs which seek

to build community participative capacity. Community-oriented projects can

also provide useful information about women's needs and behavior in low-income

urban neighborhoods. Examples of programs in Brazil are used to illustrate

these points.

The chapter concludes with the argument that community-oriented and

women-oriented service programs provide a convenient entry point in addressing

difficulties outlined in this report. By building on women's essential and

multiple roles in household community, responding to women's limited access

to employment, training, and services, and enhancing the potential for develop-

ment of women's technical and social skills, they can contribute to the short-

term and long-term prospects of women and their families. Furthermore, by

helping to build a more adequate base of information about poor women's needs

and capabilities, they can improve the understanding of appropriate policy

directions which treat women as contributing members of a developing society.

Notes to Chapter 1

1. United Nations, Division for Economic and Social Information, Department.of
Public Information, "Worsening situation of women will be main issue confronting
Commission on the Status of Women," DPI/DESI NOTE IWD/22, 13 February 1980 and
Women 1980, Newsletter No. 2.

Chapter 2

Conceptualization and Measurement of Women's Economic Roles

It is unfortunately true that any serious analysis of women's situation

in Latin America must begin with a discussion of the gaps and distortions in

existing data bases. Because a concern with sex-differentiated patterns of

behavior is relatively recent, there are still some questions for which in-

formation simply does not exist. In other cases data sources include infor-

mation on the two sexes but this variable is not tabulated.

In a country such as Brazil with well-developed statistical centers and

scientific communities, the problem in analyzing women's roles is not generally

the lack of data. Instead researchers are faced with the difficulty that the

data which does exist is poorly suited to capturing women's activities, since

the methods used to collect it were originally developed with men in mind. In-

ternational and national groups, including researchers interested in women's

issues, are now beginning to make suggestions about how these data problems can

be addressed.1

Scholars argue that the lack of adequate statistics impedes the inte-

gration of concerns about women into planning. An insufficient data base

makes it difficult to analyze the unintended aspects of industrialization

on women's employment and welfare, for example. In general, official statis-

tical centers in Latin America do not give clear priority to these issues

and hence are not concerned with producing the statistics appropriate to

address them.2 This problem is in part the result of a lack of dialogue

between the "producers" and "users" of statistics, which is only beginning

to be remedied.

This section of the report will present an overview of the major prob-

lems in conceptualization and data collection in official statistics in Lajin


America. The discussion provides a necessary background from which to

evaluate the available studies of women's roles, most of which rely on

officially-produced statistics. The chapter concludes with a preliminary

discussion of sources of information on women in Brazil.


2. A. Conceptual Problems

1. Concepts of economic activity.

The single largest conceptual problem in evaluating women's economic

roles is the definition of what is and is not "work." In general work has

tended to be male-defined, since men are generally regarded as the pro-

viders of their family's economic needs. This assumption is inadequate

on two counts: not only do a significant proportion of women provide the

main economic support for their families, but even those who do not, usually

play an important and often crucial role in supplementing household income.

Yet because of the way work is operationally defined, women's productive

activities tend to be "invisible" to data collectors even when women

generate monetary income.

The commonly used category of "housewife" often serves to disguise a

woman's role in myriad productive activities by associating them instead

with sex and marital status attributes. On the one hand housewives may /V,

represent a kind of "disguised employed" because of the importance of .

their activities in overall family welfare: providing for daily needs

and negotiating goods and services from the urban system in their role

as householdd managers." When domestic work tasks are included in the

definition of work, women's work hours and participation rates are even

higher than men's, and virtually all married women are in the economi-

cally active population (96% in an experimental study in Brazil ). At

the same time, women categorized as housewives may also be disguised un-

employed, or women who do not actively seek outside paid work only

because they are aware of their unlikely chances of finding acceptable

employment. This situation is particularly common for untrained women

in the labor surplus economies of most developing nations. One study


counted as disguised unemployed all those housewives of appropriate age

and educational categories in households where other persons were avail-

able to take over domestic chores. Those who fell into this category

made up 7% of all women, and their addition to the total unemployment
rate for women made it higher than that for men.

Aside from the invisibility of domestic work itself, the housewife

category often hides women who carry out income-earning activities based

in the home. Women are the majority of unpaid family workers who con-

tribute to family enterprises, even in urban areas,without receiving a

direct wage in return. They are also a high proportion of those in

"marginal" or "informal sector" jobs characterized by their small scale,

low productivity, and lack of legal registration and ties to prevailing

social institutions. Typically low-income married women combine flexible

inputs of labor into such activities as sewing, washing and cooking for

others with their own domestic chores, in order to generate small

amounts of income when needed. Women who are so engaged are often under-

renumerated in labor force statistics because of their unstable resi-

dence and employment patterns.5

The inadequacy of data collection techniques in representing women's

economic activities can have serious implications for development plan-

ing. "In developing countries the lack of an adequate data base may

result in planning that fails to take into account the economic activities

of substantial numbers of women. This can lead to the displacement of

women workers from traditional work patterns by modernization and large

scale industrialization. This displacement goes largely unnoticed in

national statistical systems resulting in great hardship and suffering

to the women concerned and consequently to their families."6

Some practical suggestions have been made as to how to improve the


collection of data regarding women's productive activities. In general,

more detailed and probing questions are required in order to overcome

the biases of both interviewer and respondent in automatically classi-

fying women as "inactive" housewives when they do not hold regular,

full-time jobs. Particular attention must be paid to part-time and

intermittent work; for example, a longer reference period (two weeks

to one month) may be necessary to capture women's irregular work acti-


2. Concepts of the household.

As will be demonstrated in later sections of this report, women's

economic roles are to a great extent mediated by their position in

specific household contexts. Hence it is important to'develop and use

appropriate measures not only at the individual but also the household

level in order to understand female behavior. Most standard data sources

incorporate household characteristics in some fashion, but more care

should be taken in defining key concepts and more use should be made

of data routinely collected,but less commonly tabulated,at the household


The concept of head of household has received the most attention in

relation to women, due to the growing recognition of the significant

proportions of poor households headed by women instead of men.8 Yet

household headshipp" may be defined according to very different criteria.

Commonly the household head is simply the person designated as such by

respondents, with no probing to determine what factors lie behind these

responses. Yet it is important to distinguish between head as decision-

maker and as principal income-provider in a household, since these func-

tions do not necessarily reside in the same individual,and yet both

designations may be important in determining individual behavior. While


in many parts of the world the problem of defining the concept of

household head has received substantial attention, in Latin America

the issue has not generally been seen as very important by data pro-

ducers.9 However the Brazilian 1979 National Household Sample Survey

included a special supplement devoted to verifying the components of

the concept; the results have not yet been released.

The household unit is commonly used as a principal unit of data

collection because it is considered to be the unit of consumption of

basic necessities. In practice, this basic consumption unit is identi-

fied using the residential unit as a proxy, on the reasonable assump-

tion that domestic groups sharing housing facilities typically share

access to most other basic necessities (food, clothing) as well.10

Despite this recognition of the household's importance as a unit of

income pooling and sharing of final consumption, it is rarely the focus p&-0-e

of data analysis. That is, there are few variables which are actually

analyzed at the household level, aside from occasional tables dealing

with characteristics of the household head. A few smaller-scale surveys

have begun to experiment with the development of measures at the house-

hold level by reconstituting domestic units.1 Household level measures

need to be developed for both compositional variables as well as economic


The most basic element of household composition in determining dif-

ferent patterns of individual behavior is the concept of life cycle, or

developmental cycle. Based on a model of the developmental cycle of the

nuclear family, life cycle is variously measured by the age of one or

both spouses and/or the age of dependent children. The latter is

probably preferable since it more directly addresses differences in


consumption demand and labor availability. When non-adult dependents

are defined only by their age and not by their relationship to the house-

hold head, life-cycle categories are applicable to both nuclear and non-

nuclear families.13 A second important measure of household composition

which is more commonly used is that of household size, which is related

to life-cycle changes. However, measures of size are less expressive of

the structural changes underlying variations in household composition.

More innovative measures might also be developed to capture the

potential and effective economic capacities of different kinds of

domestic groups. It has already been persuasively argued that realistic

measures of income must take into account differences in household

composition (i.e., -income per capital or income per consumer).14 Relatively

sophisticated techniques may be applied to analyze the extent to which

diverse households are able to utilize their potential stock of workers

effectively to generate sufficient income for their consumption needs.15

This effectiveness may be more simply expressed in dependency ratios,

particularly those which differentiate between child and adult dependents.

Simpler measures may also be devised which calculate sex and age-specific

participation rates at the household level for different sub-groups of

the population.16 Similarly, household-level measures of educational

attainment may be calculated by taking into account the schooling of

household members in relation to their age.17 Finally, it is possible

to devise measures of the source of livelihood of domestic groups

which take into account the internal heterogeneity of household workers

and the presence of complex labor force strategies for generating income.1
3. Concepts of the family.

It is important to conceptually separate the family unit from the

household; the latter refers to a unit of consumption of basic necessities


whereas the former is a social unit based on kinship relationships.

However both kinds of units are important in mediating individual

behavior, particularly for women. Thus whereas women behave differ-

ently (i.e., work more or less, and in distinct kinds of occupations)

according to the kind of household in which they are embedded, so too

do distinct family forms shape different responses. Scholars have

pointed out the importance of recognizing different kinds of families

(nuclear/extended/compound/single-parent) as well as of different life

cycle phases, here specifically designed to apply to evolving family forms.

While it has long been recognized that marital status is an important

determinant of women's labor force participation, more attention needs

to be paid to distinct categories of unmarried women: never-married,

divorced, separated, widowed, single mother. These categories determine

different contexts for women's behavior and welfare.19 In summary, more

attention should be paid to the specific family context in which women

operate, a topic which has been relatively neglected by standard data

collection agencies in Latin America.20

- 9 -

2..B. Collection of Data on Women's Economic Roles

1. Methods

Scholars and planners concerned about the inadequate understanding

of women's economic roles outlined above have made a series of suggestions

of ways to test and improve the collection of data related to women. It

is argued that those concerned with data collection must take a closer look

at the form and functioning of family productive units, at forms of re-

muneration for women, at household and family division of labor and their

variations, and at self-perceptions of women's economic roles and how

these affect the information given by respondents. Most agree that ex-

perimentation with these issues is appropriately carried out using studies

separate from the major data-collection procedures such as the census.

Qualitative studies, case studies, and in some cases household surveys

may be more appropriate vehicles for this task. However it is imperative

that their findings be integrated into the standard data sources in

some fashion, either through the refinement of concepts and their opera-

tionalization, the testing of new approaches for later use in the census,

or the use of sub-sampling procedures which would link smaller scale to

large-scale data sets.21 More specifically, experimentation with studies

of time-use (daily, weekly, monthly, and annual) might permit alternative

measures of economic activity more appropriate to women's activities.22

For example, a national-level survey of rural women's work roles in Colombia

was able to incorporate simple questionnaire items related to the use of

time in specific economic activities, after the technique had been tested

in intensive case studies.2

2. Personnel

Problems of conceptualization and measurement of women's economic

activities demand a greater sensitivity to the issues outlined above at

- 10 -

all stages of the data collection process. For this reason, official

statistical agencies are advised to employ more women as interviewers,

planners, and data analysts.24 Female interviewers are especially

important in order to probe respondents on crucial aspects of women's

economic activity, as well as on such sensitive topics as pregnancy

histories. Furthermore, women interviewers also have easier access

to young migrant women who are often under under-remunerated in the

census. In order for improvements in data collection to be maintained

through later stages of the data analysis process, personnel involved in

these tasks should also be sensitized to the major issues described here.

3. Respondents

In the same way that interviewers.of different sexes may elicit

different kinds of responses, male and female respondents can also be

expected to provide different kinds of information. More experimen-

tation in comparing the results is suggested in order to adequately

define the kind and degree of bias to be expected from relying on

respondents of different sexes. In general it is recommended that women

be directly interviewed about their own economic activities even when

the main respondent (for the household or family unit) is male, in

order to reduce this source of bias.25

4. Population census

In order to adequately capture the specific conditions of women's

income-generating activities, greater detail is needed in occupational

and industrial sector categories, including more attention to the quali-

fications of workers at different occupational levels. More detailed

tabulations of workers of each sex in different categories are needed.

For example, few census bureaus provide tabulations of industry by

11 -

occupation distribution for both sexes. Perhaps even more important is the

provision of additional tabulations based on family and household-level

variables, in order to analyze different patterns in women's economic be-

havior and welfare in different mediating contexts.26

5. Economic census

In the industrial census it is imperative to differentiate between

factory work and small-scale and/or home-based industrial activity, where

women predominate. These represent very distinct conditions of employment

which are poorly understood. In the Peruvian 1961 census home-based

female pieceworkers in the textile industry numbered 32,000 as opposed

to only 8,000 working in factories; by 1970 the two categories were lumped

together.27 The economic census could be a valuable source of informa-

tion if more detailed tabulations were provided of the size, type and
sector of workers in different age and sex sub-groups.2

6. Continuing work statistics

Data collected on a continual basis provides a unique opportunity to

monitor trends in development and their impact on women. More attention

should be paid to female patterns of employment, unemployment, and under-

employment (or disguised unemployment). In addition, continuing work

statistics can.provide on-going data related to remuneration, social

security availability and work stability for women if these statistics

are tabulated by sex.2

- 12 -

2. C. National Sources of Data on Women in Brazil

This section presents a listing and brief description of the major

sources of statistics relevant to the analysis of urban women in Brazil.30

1. Population census

The Brazilian Geography and Statistics Institute (IBGE) carries out

a decennial census from which information is available from the following

years: 1872, 1890, 1900, 1920, 1940, 1950, 1960, 1970. Of the recent

censuses, substantial published material is available for both the 1950

and 1970 surveys. Few data from 1960 have been published due to problems

in the census of that year. In addition to published tabulations, scholars

may petition the IBGE for special tabulations of their specification. Finally,

a 1% public use sample has recently been made available from the 1970 census,

permitting more detailed analyses using raw data.

The census survey is made up of two kinds of questionnaires. The first

consists of 10 basic questions applied to the whole population: sex, ane,

relationship to household head, nationality, literacy and school attendance.

The second consisted (in 1970) of 32 questions applied to one in four

households to obtain more detailed information on certain topics. These

include religion, length of residence in present location and previous resi-

dence, educational level, marital status, income and work status, occupation,

industry sector, and type of worker, and a series of questions on mortality

and fertility for women over 15 years of age. For the 1980 census, this

sample questionnaire will be expanded; the experimental census contained

56 questions. In particular more detail will be added to the questions

related to economic activity, which will be asked of all persons 5 years of

age or older (whereas before they were applied only to the economically

active population). The category of agricultural day worker (volante) will.

- 13 -

be added to the occupational list. The former global questionnaire item

on average monthly income will 'be disaggregated to include the following


average monthly income in cruzeiros
average monthly income in kind
average monthly income from secondary occupations
income from retirement
income from rent, etc.
income from capital gains
income from donations or monthly allowance

This greater detail will clearly provide much greater possibilities for

analyzing women's contribution to household total income.

2. National Household Sample Survey (PNAD)

The PNAD has the substantial advantage over the census of providing

much greater detail on certain topics. However it presents a much shorter-

term series (since 1967) and since it is carried out on a sample population

cannot in most cases be utilized at municipal or state levels. The sampling

procedure is a four-stage process based on census sectors and including

only private households (not convents, penal institutions and other in-

stitutions). The sample is weighted using different fractions for each

region and is then expanded using the census as a base. Because of the

complexity of the sampling and data collection procedure, most scholars

agree that PNAD data should be treated with caution. The regions used

for the PNAD, and their fraction of the sample, are presented below:

Region States Fraction of Sample

I Rio de Janeiro 1/200

II Sao Paulo 1/300

III Parana, Sta. Catarina, Rio Gde. do Sul 1/300
Porto Alegre Metropolitan Area 1/100

IV Minas Gerais, Espirito Santo 1/200
Belo Horizonte Metropolitan Area 1/100

V All States of Northeast 1/200
Recife Metropolitan Area 1/100

- 14 -

Region States Fraction of Sample

VI Federal District (Brasilia) 1/20

VII North and Central-West (urban regions) 1/100

Data for all the regions are included beginning in 1972. In 1971, 1972,

1973 and 1976 the data are significant for the metropolitan regions of

Rio de Janeiro and Sao Paulo and in 1977 for Belo Horizonte, Porto Alenre

and Recife as well. In 1978 data exist for all the metropolitan areas

mentioned. Beginning in 1977 the Northeast Region is divided into three


The content of the PNAD has varied over the years. From 1967 to

1970 it was carried out four times per year, principally with the ob-

jective of analyzing labor force characteristics. Beginning in 1971 it

was carried out only in the last trimester of each year. The basic

questionnaire contains relatively detailed questions on economic activity,

and since 1976 these items have been similar enough to the census question-

naire to permit some comparison. In 1974 and 1975 a special PNAD called

ENDEF was carried out, which will be treated separately. Since 1973

special questionnaires have been appended to the basic instrument to col-

lect more detailed information on specific topics, as listed below.

1973 special supplement on occupational mobility and fertility
(mobility data were later used by the Social Indicators
department of IBGE in one of their projects; see section
2. C. 6)

1976 special supplement on race, applied to one-fifth of the
sample; certain 1973 questions on occupational mobility
were also repeated

1977 special supplement in the metropolitan areas with questions
of interest to the National Housing Bank (BNH); these data
should be released separately by the end of 1979

1978 research on vaccination

- 15 -

1979 Normal PNAD with a few items to explore the concept of
head of household

1980 to be carried out in January and February, with a special
supplement for ever-married women related to fertility.
Also planned is a trimestral survey of employment

The 1977 supplement which focused on housing could provide the

basis for a detailed analysis of the differences in access to adequate

housing between households headed by men and women. It also contains

a potentially interesting question about whether a household's sewing

machine is used for family consumption or for outside work. This iten

addresses the question of women's role in income generation and is also

linked to housing. A sociologist studying home-based piece-workers com-

ments that women take on the most outside work during the phase of house

construction, in order to help buy materials.31

PNAD data are released with some delay in published form (data

from 1976 are now becoming available) and can also be used by authorized

persons and institutions who wish to request special tabulations.

3. National Study of Family Expenditures (ENDEF)

The special PNAD carried out in 1974 and 1975 had three principal

objectives: 1) obtaining information on family consumption of food products,

on expenditures, and on inventories of durable goods; 2) collecting in-

dividual-level information on demographic characteristics, income and

occupation, and anthropomorphic characteristics; and 3) forming a nutri-

tional portrait at the family level. The study involved an enormous effort,

using between 700 and 800 interviewers to survey 55,000 families. The

sample was distributed over 40 survey periods in order to collect seasonal


Interviewers accompanied families over a period of seven days,

weighing food items two or three times per day, obtaining separate weight

measures of food before cooking, residual items (such as rinds and bones),

and waste. Respondents were also asked about how products were acquired,

16 -

in order to verify the role of monetary and non-monetary payments; non-

monetary payments were then attributed a monetary value. Family budgets

were distinguished from the expenses (,f non-family members with autonomous

budgets. For both, information on expenditures and income were collected

using weekly, monthly, trimestral and annual time bases (excluding the

costs of services provided by the state). Data on income (during the

past month and annual) and occupation (over the past 12 months) were also

collected for each person, including the number of hours worked. Finally,

in order to evaluate nutritional status, information was collected for

all persons on age, sex, weight, height, and circumference of biceps.

Data from the ENDEF study clearly present rich possibilities for

analysis although the complexity of the undertaking has meant considerable

difficulty in moving to the stage of generating publishable findings.

Data have begun to appear in published form but mostly in their most

aggregate state; for example, breakdowns by income are not yet available.

Further analysis is currently underway by the staff of IBGE's Department

of Consumption Studies (DESCO). They are attempting to create appropriate

nutritional measures for Brazil based on anthropomorphic measures from

persons in households in the higher income ranges. These measures will

then be used to compare the nutritional levels of different social qrouos,

including migrants vs. natives and male vs. female-headed households, as

well as different occupational groups for both sexes. Since ENDEF con-

tained no measures of distribution of food within the family, the important

question of differential access to resources within the family for men

and women cannot be addressed with this data set. However it would be

easy to generate a table comparing the locale of a meal (home, work,

visit) or its presence or absence, by sex and family relationships. The
survey also contained some interesting items related to domestic servants.

- 17 -

4. Economics statistics

The IBGE collects diverse types of statistics related to industry,

commerce and service sectors which might provide information about women's

economic roles. Some, like the economic censuses collected every five

years since 1970, gather information related to the broad structural

characteristics of the various sectors. Others, such as the annual in-

dustrial surveys, accompany sectoral changes or provide shorter term in-

dicators through the monthly industry studies. The following chart

synthesizes the surveys which exist for the three sectors. As it shows,

the industrial sector is the best covered, while the other two sectors

exhibit some gaps in coverage. This is particularly problematic in re-

lation to women, whose employment in industry is relatively minor compared

to the commerce and services sectors.

Type of Data: Industry Sectors
Industry Commerce Services

Structural Industrial Consumer goods Services
Data (Censuses):Construction Real Estate
Public Utilities

Sectoral Annual Industrial Highway taxes
Accompaniment: Survey Hospital resources
Telephone companies

Conjunctural Monthly Industrial National Price Survey
Information: Survey Survey of Prices of
Construction Materials

Data collected in these economic surveys generally refer to aggre-

gate value of production, sales and stocks, and energy consumption. Some

include information on the number of persons employed, both total and in

direct production, and on salaries paid. The only data discriminated by

sex, however, is that of the number of persons employed in 10 occupational

- 18 -

levels. Information on salaries, earnings and other remuneration is not

broken down by sex, nor is this discrimination foreseen in the future

since the data is allegedly difficult to obtain from employers. These

limitations clearly restrict the utility of an otherwise potentially use-

ful source of information on women employed in different sectors of the


5. Annual Social Information Report (RAIS)

Beginning in the 1970's, the system of collection of social data in

Brazil was divided among diverse institutions with specific objectives;

these included the National Housing Bank and other federal banks and

the Ministries of Social Welfare and Labor. These various efforts were

integrated in 1975 in the RAIS, which was designed to take the place of

earlier surveys on employment and earning. The first questionnaire of the

RAIS, using 1976 as a base year, was applied in 1977; data from 1976 and

1977 have already been verified and those from 1978 are currently under


The addition of the RAIS to the official data sources completes

the system of statistical information on the labor force, with the PNAD

covering the supply side and the RAIS concentrating on demand. In 1978

control of the RAIS was vested in a Coordinating Group composed of repre-

sentatives of the various agencies involved, under the control of the

IBGE. Although no data from the RAIS have yet been published, there is

a Basic Tabulation Plan and special tabulations can also be requested.

For example, the Labor Ministry has generated tabulations at the level

of municipios which show the number of enterprises and of employees and

the average salary by sex and nature of economic activity, by size of

establishment. The wealth of data contained in the RAIS would seem to
recommend it as an additional source of up-to-date information on women's

- 19 -

employment conditions.

The RAIS questionnaire is addressed to the employer, covering es-

pecially those in the private sector. While the public sector is also

entering the data system, the autonomous employer in the tertiary sector

is probably still marginal to the RAIS universe. This limitation is

particularly important in examining female employment. The question-

naire contains 10 questions directed at the employer himself, as well as

other items concerning all workers employed by this employer for at least

14 days during the past year. There are approximately 30 variables re-

ferring to each employee, of which the most potentially interesting are

those related to salary. The questionnaire form itself does not indi-

cate the sex of the worker, but this information can theoretically be

recuperated from other data. Information covered includes the following:

prior employment
marital status
reason for leaving employment (Quality of data questionable)
educational level
type of employment

contractual salary
hours worked per week
December remuneration

admission (Time series data on labor fluctuation by activity)

Social Security contributions
Employment Security Fund options (FGTS)
FGTS collection

birth; locale and date

work documents

Despite the official nature of the RAIS, there is no legal obligation

on the part of the employer to respond to the questionnaire. There is an

indirect obligation since if he does not respond, his employees do not re-

ceive certain benefits guaranteed by law (PIS). The employees can then com-

- 20 -

plain and force the employer to pay the PIS as well as a fine. In spite

of this problem, the coverage of the RAIS is judged to be better than that

of the previous more dispersed system of economic data collection.

6. IBGE Department of Social Indicators (DEISO)

This specialized department of the Census Bureau is responsible for

various on-going studies which draw on a variety of statistical sources.

Most of these projects are potentially of great utility in the analysis

of women's roles and conditions, although none had yet been published in

August of 1979. Personnel involved in the studies showed an interest in

discriminating information by sex whenever data permitted. Studies cur-

rently underway by the DEISO include the following:

a. Life conditions of the low income populations in the metropolitan
regions of Recife, Porto Alegre and Rio de Janeiro. The studies
of Rio and Porto Alegre have been finished and should be pub-
lished by the end of 1979, while that in Recife is still in

b. Employment in rural and urban areas. In initial phase.

c. Industrial workers. Using anthropological methodologies, case
studies were carried out of certain industrial occupations. This
study was stopped for some time and was to be taken up again at
the end of 1979.

d. Occupational Mobility. Using data from the 1973 PNAD, compares
present work with first job and father's work.

e. Social indicators report. Published end of 1979.

7. Other surveys

Aside from the wealth of data contained in the official sources listed

above, a series of other surveys also exist which could provide useful in-

formation on women's roles and conditions. By way of example, several are

mentioned in this concluding section; the list is by no means exhaustive

but merely suggestive of the many untapped resources which exist in Brazil.

a. Metropolitan household surveys. Local planning agencies often

carry out independent surveys of socioeconomic conditions in their metro-

- 21 -

politan areas which can provide rich sources of data. In Belo Horizonte,

an exhaustive socioeconomic survey was carried out in 1974 by PLAMBEL, which

provided the data base for a number of studies of that city.31 A second

survey was carried out in 1976 with an emphasis on problems related to

housing, which is to be followed by a second phase of intensive field re-

search with a sub-sample, to collect life history information related

to migration, employment and housing. Access to data from the second

survey had been delayed at the stage of analysis, but the first tabulations

should be available soon. It is likely that similar surveys exist in many

major cities.

b. Non-official surveys. In addition to official sources of informa-

tion, useful data has been collected by a variety of private research insti-

tutions in Brazil. By way of example, the household surveys conducted by

the Inter-Sindical Department of Statistics and Socio-Economic Studies

(DIEESE) have provided important basic information on the life conditions

of working class families.32 A new survey is projected for 1980 which

promises to gather even more detailed and up-to-date information for the

Sao Paulo metropolis. A massive study of-domestic groups in nine selected

regions of Brazil has also been conducted by the Brazilian Center for Analysis

and Planning (CEBRAP) in Sao Paulo, from which a number of studies specifi-

cally focussing on women are currently in progress.

c. Women's Studies in Brazil. In addition to using standard data

sources to analyze women's condition, a growing number of Brazilian scholars

are carrying out research specifically focused on women, using a combina-

tion of secondary data analysis and the collection of original information.

These studies constitute a growing body of scholarly work which can be use-

ful for those concerned with the practical problems faced by poor women.

In many cases research reports are available only in mimeographed form,

- 22 -

although some published volumes are also in preparation. One of the most

)important institutional bases for scholarly research on women is IUPERJ

in Rio de Janeiro, which sponsored the conference on Women in the Labor

Force in Latin America, November 1928. Secondly,the Fundacao Carlos

Chagas in Sao Paulo has administered a small-grants award program on

women in Brazil for the past two years, and results from these studies

are now becoming available. The Fundacao has also published the first

volume of a bibliography on Brazilian Women, of which the second volume

is currently in preparation.34

- 23 -

Notes to Chapter 2

1. These include various agencies of the United Nations as well as the
group of scholars and policy makers who gathered at the seminar entitled
"Women in the Labor Force in Latin America," sponsored by IUPERJ (Univer-
sity Research Institute of Rio de Janeiro), Rio, November 23-26, 1978.
The following discussion draws on three principal sources: Stan D'Souza,
"Sex Biases in National Data Systems," consultancy report for the UN
Statistical Office, 1978; Alice de Paiva Abreu, Maria Luiza de Carvalho,
Tania Salem, Maria da Gloria Ribeiro da Silva and Rosa Maria Ribeiro da
Silva, "Relatorio Geral do Seminario 'A Mulher na Forca de Trabalho na
America Latina,'" 1979; Susana Torrado, "Conclusiones y recomendaciones
relatives a las fuentes de information necesarias papa estudiar la partici-
pacion de la mujer en los mercados de trabajo en America Latina," 1979.

2. D'Souza 1978: 1.5.

3. Felicia Reicher Madeira, As Condicoes do Trabalho da Mulher e as
Condicoes de Vida da Familia, M.A. Thesis, University of Sao Paulo, 1979.

4. Madeira 1979.

5. D'Souza 1978: 4.3.

6. D'Souza 1978: 4.1.

7. D'Souza 1978: 4.5-4.6.

8. See, for example, Mayra Buvinic and Nadia H. Youssef, "Women-Headed
Households: The Ignored Factor in Development," Planning Report submitted
to AID/WID, 1978 and Thomas W. Merrick and Marianne Schmink, "Female-
headed households and urban poverty in Brazil," paper presented at the
ICRW conference on "Women in Poverty: What Do We Know?", Elkridge, Mary-
land, April-May 1978.

9. D'Souza 1978: 2.4.

10. D'Souza 1978: section 3.

11. The terms "household", "domestic group", "domestic unit" and "residential
unit" are used interchangeably in this report.

12. D'Souza 1978: 3.4-3.6.

13. See Marianne Schmink, Community in Ascendance: Urban Industrial Growth
and Household Income Strategies in Belo Horizonte, Brazil, Ph. D. Disserta-
tion, University of Texas at Austin, 1979: 188-193, for a discussion of life
cycle patterns in a Brazilian community.

14. Simon Kuznets, "Demographic aspects of the size distribution of income:
an exploratory essay," Economic Development and Cultural Change 25 (October):
1-94, 1976.

- 24 -

15. See Anna M. Sant'Anna, Thomas W. Merrick and Dipak Mazumbar, "Income
Distribution and the Economy of the Urban Household: The Case of Belo
Horizonte," Washington, D.C.: The World Bank, Working Pater No. 236, 1976
and Merrick and Schmink 1978.

16. See Brigida Garcia, Humberto Munoz and Orlandina de Oliveira, "La familiar
obrera y la reproduccion de la fuerza de trabaja en la cuidad de Mexico,"
paper presented at the conference on Development and Ineuality in Latin
America, Center for Latin American Sj esud~, s ni~idaG es-
vfieOct6ber 1979 andM igqracion, familiar y fuerza de trabajo en la
cTud dde-TexicoT'Cuadernos del CES, 26, 1979.

17. See Arthur D. Murphy, Urbanization, Development and Household Adaptive
Strategies in Oaxaca, A Secondary City of Mexico. Ph. D. Dissertation,
Temple University, 1979.

18. See Schmink 1979.

19. D'Souza 1978: 2.9; Merrick and Schmink 1978.

20. D'Souza 1978: 3.10; 5.3.

21. Abreu et al. 1979: 6; 17-18.

22. D'Souza 1978: 4.8.

23. Carmen Diana Deere, Jane Humphries and Magdalena Leon de Leal, "Class
Sand Historical Analysis for the Study of Women and Economic Change," paper
prepared for the Role of Women and Demographic Change Research Program,
I.L.O., Geneva, 1978.

24. D'Souza 1978: 5.11-5.14.

25. Abreu et al. 1979: 18.

26. Torrado 1979: 27-34.

27. Personal .communication, Violetta Sara-Lafosse.

28. Torrado 1979: 37.

29. Torrado 1979: 36.

30. This section is primarily taken from a report prepared for the author
by Alice de Paiva Abreu.

31. Personal communication, Alice de Paiva Abreu.

32. Thomas W. Merrick, "Employment and earnings in the informal sector in
Brazil: the case of Belo Horizonte," The Journal of Developing Areas 10
(April 1976): 337-354; Merrick and Schmink 1978; Sant'Anna et al. 1976.

L 33. DIEESE, Familia Assalariada: Padrao e Custo de Vida, Estudos Socio-
Economicos 1: 2TJan. 1974); Nivel Alimentar da Populacao Trabalhadora
da Cidado de Sao Paulo, Estudos Socio-Economicos 1: 1 (July 1973).

34. Fundacao Carlos Chagas, Mulher Brasileira: Bibliografia Anotada,
Volume 1. Sao Paulo: Editora Brasiliense, 1979.

Chapter 3

Poor Women In the Urban Economy

It is difficult and often misleading to generalize about Latin America,
given the great cultural, political, and economic diversity of the region.

Abstract discussions of women's roles and conditions are similarly over-

simplistic because of the crucial differences determined by socioeconomic

position, family status, and age and cohort differences. It is, however,

possible to draw some tentative region-wide conclusions about the situ-
ation of women in more defined contexts. In fact, when the analysis con-

centrates on the position of poor women in large urban centers in Latin

America, the commonalities are surprising. To a great extent these similar-

ities are due to the form of urban-industrial development which has taken

place in the last three decades, and which has shaped the urban environment

in most Latin American countries.

This chapter will therefore begin with a brief history of industrial

development and urbanization in contemporary Latin America, emphasizing

those aspects which have been most important in determining women's econ-

omic roles and welfare. Particular attention is paid to the impact of

recent development on women's labor force patterns. With this background,

a conceptual framework is presented within which women's economic activities

may be more clearly understood. It is argued that women's productive ac-

tivities are best analyzed as one component of complex economic strategies

carried out at the household level. The form of these strategies will vary

according to the internal evolution of the domestic group and will be large-

ly shaped by the household's position in a specific socioeconomic and

spatial context. Women's important roles in household strategies determine

not only their labor force behavior but also other aspects of community and

household level manipulation of resources crucial to family welfare.


3. A. Overview of Urban-Industrial Development in Latin America

Two general phases of industrial development may be distinguished

for the region. Through the beginning of this century, most Latin Ameri-

can countries supplied their internal markets with manufactured goods

through imports which were exchanged for raw materials or semi-processed

goods on the world market. With the Great Depression and the second World

War, supplies of these goods from abroad were greatly reduced, and Latin

America entered a phase of relative isolation from the international

market. At this time national governments, in conformity with new ECLA

perspectives, began to formulate conscious economic policies to promote

Import-Substitution-Industrialization (ISI) to supply national markets

with manufactured goods. Industrial establishments during this first

phase of development tended to be small-scale, labor intensive, and

nationally-controlled. Workers were recruited from the ranks of migrants

from rural to urban and metropolitan areas. These economic policies and

the resulting importance of the creation of an internal market for manu-

factured goods were consistent with populist political policies which

permitted a wider margin for labor union activity to achieve more adequate

wage levels, as well as an expanded state role in the provision of collec-

tive social services.

Beginning in the post-War period and mainly in the 1950's, the role

of Latin America economies in the world economic system entered a new

phase. In the developed countries, particularly the U.S., investors faced

a situation of falling rates of profit and began to actively seek oppor-

tunities for investments overseas. This second phase marks the appearance

of an important economic agent, the multinational corporation, on the

Latin American industrial scene. Facing the need to expand operations, and

the increasing costs of labor in developed economies, foreign investors


were attracted by the lower wages in overseas settings, where salaries

Averaged only one-tenth of those in the U.S.1

The penetration of international investment into the Latin American

industrial sector began to change the dominant form of industrial enter-

prise in the region. New firms were more likely to be large-scale, used

more productive technologies and were often controlled by multinational

corporations. Because of their more capital-intensive nature, these new

establishments were less labor absorptive. At the same time, their higher

productivity made it difficult for smaller firms to compete, so that

less productive national firms were increasingly driven off the industrial

market. Only those industries with competitive productive bases and a

relatively large scale of operation, particularly those controlled by

the state, could effectively compete.

At the same time, rural-urban migration streams were beginning to

far surpass the capacity of the industrial sector to absorb new workers.

The new form of industry, often based on intermediate and capital goods

production, selected its workers from among the "prime" skilled young

male laborers. These more fortunate urban workers were paid relatively

higher wages in comparison with those in other urban employment where the

capacity to pay was much lower. Thus while on the one hand multinational

corporations sought investment sites in Latin America because of the

lower prevailing wage rates, their impact within the region was to create

a kind of worker's "elite" based on relatively higher salaries than those

paid in other urban sectors.

With the increasingly selectivity of workers absorbed in the indus-

try sector, the majority of urban workers swelled the services and commerce

sectors, particularly those characterized as the "informal sector." These

workers, many of them women, have more erratic earnings and less stable


employment and are cut off from the benefits of the formal capitalist

sector. With no retirement benefits assured and facing uncertainty and

instability in both the short and long-term, the growth of informal sector

employment has not meant an effective integration into the urban inviron-

ment of the growing low-income populations. But it has been essential,

as this chapter will show, in permitting poor families to cope with pro-

longed financial crises.

National labor policies in Latin America began to shift with these

changes in industrial development patterns. Labor's bargaining position

was increasingly restricted, in some cases through overt repression of

labor union activities but also through more institutionalized means of

controlling wage increases to provide a "programmed decline" of wages.

As a result, in many Latin American countries high rates of growth in

industry have been associated with a tendency for increased concentration

of income; Brazil is a classic example of this model. The benefits of

increased productivity have benefitted mainly the highest level owners

and managers of production, (whose wages are higher than their counter-

parts in the United States) while wages of manual production workers have

risen relatively little.2

There are several interacting aspects of these changes whose impact

is particularly important in determining women's employment and welfare.

The first is the relatively restricted access to jobs in the expanding

industrial sector where wages are relatively higher and employment relatively

more stable. This restriction is linked to an impeded access to other

benefits such as essential social services, particularly for those not

employed in the formal sector. The most important of these benefits in-

clude health services, vocational training, workman's compensation and

retirement pensions, all of which provide an important "social wage" which


Functions as a safety net to underpin family financial uncertainty. At

the same time, insufficient and sometimes declining wages for working

populations have made additional income indispensable, even in the face

of unattractive employment options.

Given the diversity of national and regional conditions in Latin

America, the changes outlined here are not equally important in all con-

texts, but emphasize the most consistent tendencies which have accompanied

urban-industrial development in its most recent form. They are thus the

trends most noticeable in the more advanced metropolitan areas of the

region. They are linked to patterns of labor force activity for women

which also vary by region but overall appear to reflect responses to a

common set of circumstances. Despite the difficulties of using existing

data sources for information about women (see Chapter 2), there are some

patterns which have been reported and substantiated in various studies of

the topic and which are consistent with the structural conditions described


First, urban-industrial development signifies a shift of both male

and female labor out of the agricultural sector. This movement is accomp-

lished principally through the mechanism of migration from rural areas

and towns into metropolitan industrial centers. In Latin America, in con-

trast to other developing regions, rural-urban migration has tended to be

female-dominated.4 The degree of sexual imbalance in migration streams

and the causes of these differential patterns are probably highly variable

throughout the region. However a few remarks may be made about the probable

reasons for female migration into Latin American urban centers.

It is important to view rural-urban migration as a result of both

"pull" factors related to the growth of urban-industrial employment possi-
bilities as well as "push" factors associated with changes in the Latin


American countryside. The dynamics of the agrarian structure are in

themselves complex and variable, and fall beyond the scope of this anal-

ysis. But it is probable that in many parts of rural Latin America,

female urban migration has been associated with deteriorating opportuni-

ties for employment in rural areas. Some have suggested that traditionally

women's role in agricultural production in the region has been much less

important than in the African and Asian regions.5 Several recent studies

of women's participation in Latin American agriculture suggest that their

inputs have been underreported because of the inadequacies of standard

data collection procedures (see Chapter 2).6 While it still may be true

that agricultural employment for women is relatively less widespread than

in other regions, the magnitude of rural-urban migration streams suggest

that women are migrating in response to fundamental changes in agricultural

production systems and not merely to the insufficiencies of traditional


The large majority of women agriculturalists are employed in small

family farming systems, while the proportion employed as wage laborers is

relatively small. With the breakdown and subdivision of smallholdings

and sharecropping arrangements and the expansion of agrarian capitalism

in the region, women's employment opportunities may actually decline.

"Women's high rates of rural outmigration are attributed to their dis-

placement from subsistence agriculture as land consolidation, agricul-

tural mechanization, and the growth of wage employment reduce women's

productive role and leave them increasingly dependent on men's insecure

income."7 Population pressure often contributes to the fragmentation of

landholdings, so that the number of smallholdings may actually increase

although each productive unit is less capable of supporting all its poten-

tial workers. Farm families typically respond by allocating some members


to wage labor, either as seasonal day laborers or through temporary ur-

ban employment in order to maintain the diminishing family's holdings

while generating sufficient income to answer consumption needs.8

While some women do find wage-work in agriculture, their employment

opportunities are much broader in urban areas; rural-urban migration

movements are female-dominated, especially in youngest and oldest ane

groups (adolescence to early 20's and over 50).10 Young women respond

to the pull of employment in domestic and informal sector jobs, especially

in the largest urban centers. Older women probably migrate when left

(through widowhood or separation) as heads of household or alone; in the

city they either join their children's household or look for their own

employment. The economic necessity of heading a family may be, in it-

self, sufficient condition to induce rural-urban migration for a woman,

as may the status of single motherhood;1 in this sense, poor women

without a male partner find their way to urban areas and help to make

female headship a phenomenon associated with cities instead of the country-


Once in the city, women's employment possibilities in the expanding

industrial sectors are still limited. Female patterns of employment suggest
that the migratory pull is urban rather than industrial in nature.1 Some

women workers are incorporated into factory work, particularly in textile

and food processing industries where female labor has traditionally been

employed. But with increasing emphasis on the heavier industries associated

with the second phase of industrial development described above, women have

been progressively excluded from industrial employment. In Brazil, for

example, their proportion in manufacturing fell from 18.6% to only 11% dur-

ing the two decades from 1950 to 1970.13 Industrial workers account for

between 10 and 20% of the female urban labor force in most Latin American

- 8 -

countries. Furthermore, these remarks apply equally to migrant and native

women, whose economic characteristics show no significant differences once

they are all in the urban setting.

Some women are incorporated into the modern sector, as secretaries,

receptionists, store clerks, teachers and nurses. Much has been made of this

pattern and its contrast with other developing regions4 but overdrawn

assumptions about the possibilities for upward mobility for women should

also be avoided. These traditionally "female" occupations are usually reserved

for younger, single women with the benefit of some education or training.

In part this selectivity is a result of protective labor legislation which

prohibits women from some jobs considered dangerous and from the workshifts

usually favored by large modern firms (overtime and night work). Firms

typically hire only single women and dismiss them when they marry, arguing

That maternity leave and day-care requirements make women workers more ex-

pensive than men (although in many countries these costs are borne by the

state and not the firm).15 Even young single women with some education

find that their employment options are increasingly limited to the few

occupational categories described above; studies in Venezuela and Brazil

have shown that over two decades of rapid industrial growth women were in-

creasingly concentrated in these traditionally female occupations.16

The majority of urban women, whether migrants or urban-born natives,

are not absorbed directly into the capitalist or modern sector. Most are

concentrated in the domestic sector (whether paid or unpaid) which is the

female domain par excellence. Domestic service is consistently the largest

category of female urban wage workers in Latin America; furthermore, as men-

tioned in Chapter 2, many women who describe themselves as housewives may
) in fact be disguised unemployed. Domestic service is itself an occupation

linked to skewed income distribution patterns creating both the supply of

poor women for employment however demeaning and low-paying as well as


the demand for personal servants on the part of higher-income groups. Aside

from the domestic sector, women also dominate many occupations in the so-

called informal sector, where they typically work as street venders, seam-

stresses, beauty operators, laundresses and other similar self-emplyed

workers or unpaid family workers.17 Employment patterns for these women are

highly unstable and irregular and may also inhibit women's possibilities

for long-term mobility by interfering with their investment in education

and training for improved employment prospects.

This segregation of women in occupations distinct from those occupied

by male workers is linked to male/female earnings descrepancies.18 Even the

highest status female jobs, such as schoolteaching, are extremely low-paying,

despite the relatively large investment in education they require. In fact,

women's earning rise much less with education than do men's, so that salary

differences between the two sexes increase systematically with women's edu-

cational levels.19 Because of the dual nature of male and female labor mar-

kets, salaries in women's occupations do not reflect the higher prestige

levels of related but male-defined jobs; wages offered to women need not

be competitive on the male market since men do not directly compete for the

same jobs. In short, earnings for women workers do not reflect their human

capital investments and are generally little affected by occupational differ-

ences. Lower-status female occupations are particularly low-paid, despite

the bed-and-board benefits often included in the domestic servant's indirect

wage. Informal sector jobs, because they operate in a labor-surplus context

in response to financial necessity, are typically underpaid as well.

- 10 -

3. B. A Framework for Understanding Women's Economic Behavior

The preceding section presented a synthetic overview of the impact of

urban-industrial development in Latin America on the demand for female labor.
In order to understand the dynamics of women's labor force participation,

however, it is essential to view individual behavior as an outcome of de-

cisions mediated by the constraints of other social units. Changing economic

conditions at a variety of levels interact to determine the specific set of
factors which influence an individual woman's participation in different pro-

ductive activities. At the broadest level, her range of options is determined

by the impact of historical changes on particular nations or regions, such

as those outlined above for Latin America. However, these changes define

distinct options and constraints for women situated in different social

groups. At the same time, women's roles are most immediately mediated by the

needs of the domestic groups of which they are members, whose form is in

turn largely determined by their overall social position.20

Focussing specifically on economic activities, it is useful to analyze

women's roles in relation to the needs of the household unit, defined as

the unit of consumption of most basic needs, especially food and housing.21

The household or domestic unit is the most consistent locus of decision-

making about the allocation of labor and resources to meet consumption needs;

as such the unit carries out an evolving strategy for the generation of

income, serving as a redistributive unit which mediates between individual

income-earners and final consumption. Household "full income" or overall

standard of living will be comprised of some combination of collective ser-

vices (provided by the state and other agencies), monetary income, and non-
monetary inputs from home production and exchange.

- 11 -

Whereas adult men tend to specialize in the generation of monetary income,

women's roles are typically multiple in both household and community units.

Most women balance three roles over their lifetime: childbearing and child-

rearing responsibilities; income-generating work; and household management

and provisioning.22 This means that alongside their important role in bio-

logical reproduction women also play an active role in negotiating access

to all three components of household full income suggested above. Their in-

puts are particularly important in resource-poor households where the mone-

tary wage is insufficient. An inadequate wage creates a situation of income

stress which forces households to intensify income-generating strategies,

using available labor and resources as fully as possible. But at the same

time, poor women face much more serious constraints to their ability to carry

out necessary activities.

In urban settings the dependence on the monetary wage means that the

relative success of household strategies will depend on the "fit" between

household composition (available labor and consumption demand) and existing

labor market opportunities. Since both household and labor market struc-

tures are continually evolving, the fit is necessarily a changing one.

Households with a better fit to the labor market exhibit distinct patterns

of strategies from those facing "income stress." In order to illustrate the

dilemma faced by poor women, it is useful to contrast what are often termed

"survival" strategies from what might be called "mobility" strategies. The

former are associated with marginal populations in extreme poverty, whose

day-to-day needs may force all family members of working age to seek paid

employment despite losses in future opportunities (education for children)

or status (women in low status jobs), and despite the "double burden" of dom-

estic and paid work poor women face.23 More affluent households have the

- 12 -

luxury of placing their strategic priorities on long-term mobility through

a continuous and complete education for their children, with the wife choos-

ing to work only if she has the skills or training for an acceptable occu-

pation.24 In such cases, her domestic labor tasks can be allocated to a

paid domestic servant.

Strategies in the two types of households differ in the way their

components (unpaid labor, extra domestic exchanges, collective services,

and monetary earnings) are combined and in the relative importance of

each. Goods and services consumed on a collective basis include the nonwage

benefits associated with formal labor market employment and the infrastruc-

tural advantages available in more affluent neighborhoods. Wealthier house-

holds have better access to these collective goods, and also to an adequate

wage for the primary breadwinner. Households facing more severe income

stress, on the other hand, must intensify their strategies using the op-

tions open to them in order to stretch and supplement the insufficient prin-

cipal wage.25 Women play an essential role in this intensification effort.

While both poor and higher-income households use unpaid labor inputs

and inputs from exchange networks, the content and relative importance of these

activities in the overall strategy is different. Poor households seek to

reduce their consumption of purchased goods of all kinds, and unpaid labor,

the bulk of it women's, provides day-to-day material needs which can sub-

stitute for scarce monetary income allocations. It is in this sense that

these activities serve to stretch the insufficient wage. Such domestic

activities as child-care, home maintenance, clothing manufacture, cultiva-

tion of garden crops and care for small animals comprise a large proportion

of household labor time even in cities. In contrast, middle class women may

spend more time on domestic tasks such as planning and management of a child's

- 13 -

education,which have less immediate returns.26 Similarly, extra-domestic

exchange and cooperation principally among women help to substitute for

purchased goods and services in poor populations.27 Extra-domestic networks
among middle class households in Latin America contrast with those of poor

populations in functioning not to serve immediate material purposes, but
primarily to manipulate public and private administrative structures to

insure the long-term success of class and kin interests.28

Higher-income households are able to choose when and how household

members will enter the labor force, and typically maximize the completion

of children's education.29 Furthermore, women in these groups are more likely

to have completed secondary or higher education, giving them access to rela-

tively higher-status jobs; at the same time they are able to hire domestic

servants to take over some of their home responsibilities if they do enter

the labor force. Still, women in moderate to higher income groups tend

to have a relatively short labor force career consisting of the period between

the termination of schooling and marriage or their first child. As a result,

in the aggregate female labor force participation rates rise with income and

education, and are highly sensitive to marital status and child-bearing

patterns, lacking the "second peak" of participation typical of many devel-

oped countries.

Poor women, on the other hand, particularly in the older cohorts, have

lower educational levels than men and most are either illiterate or have
incomplete primary education. Despite the fact that they are often forced

to work in order to supplement family income, their labor force opportunities

are both limited and low-paying. In households facing the most severe in-

come stress, girls are often pulled out of school at an early age while their

brothers continue their education; they either enter the labor force or take

- 14 -

over the domestic chores which permit other household members (their mothers)

Sto work outside the home.30 A study of female-headed households in Brazil

found that girls missed school 30% more than boys, and that 80% of their

absences were due to responsibility for domestic chores.31 Similarly, stud-

ies of time use in Brazilian, Mexican and Venezuelan households found that

wives' labor force participation was in part determined by the age of their

eldest daughter.32 In the absence of other family members or a hired domestic
servant, women who must work must take on the "double burden" of housework

and income-generation.

Because of the importance of their supplementary income, however meager,

poor women in general have a much more permanent link to the labor market

than do higher-income women, despite their greater disadvantages. It is

for this reason that.some studies of poor populations find their participation

rates to be higher than the average, showing a reversal of the aggregate

trend for rates to increase with income.33 If official data collection

techniques were better suited to capturing women's irregular work activities

(see Chapter 2), this trend would undoubtedly be more commonly found. Yet

not only are their employment options extremely limited in the short-run

but they also do not provide the long-term mobility or security of training,

advancement, immediate nonwage benefits or social security rights.

Furthermore, if women's domestic work were considered as productive

activity, there is no question that poor women's work participation would be

higher than in more affluent households.34 These tasks take up more total

worktime than do income-generating activities in low-income households.35

The non-monetary inputs provided by poor women are essential in underwriting

low wages. In their role as household managers women strive to reduce financial

stress by limiting consumption of purchased goods, substituting unpaid labor

when possible, manipulating extra-domestic networks and patron-client re-

lationships, and negotiating access to collective services for themselves,

- 15 -

their families, and their community.

Alongside differences in patterns of economic activities of low and

higher-income households, they also exhibit differences in household compo-

sition. Studies in Belo Horizonte have shown that the poor have both higher

fertility rates and higher tendencies to aggregate additional household

members to the nuclear core, resulting in higher dependency ratios through-

out the life cycle.36 These patterns may themselves reflect the relatively

highermortality rates among the poor, and the need for additional potential

earners, even in urban areas. Furthermore, these dependency ratios in poor

households come an average of 10 years earlier than in more affluent groups,

at a time when lifetime earnings are likely to be lower.37 Poverty is to a

large extent the result of a domestic unit's inability to effectively utilize

their stock of potential workers, and many poor children are forced to play

San economic role from an early age. Finally, poor populations exhibit a

much higher and rising proportion of female-headed households, which are es-

pecially vulnerable to income stress.38 Not only are the heads of these

households disadvantaged on the labor market, but the other members are also

less likely to be prime-age male workers._ Women, because of their family

responsibilities, may be more likely to take on displaced relatives (es-

pecially other females) even when their financial situation is inadequate,

whereas young adult males may find it easier to escape the poverty trap of

such households.39

In summary, poor women are much more likely to be important income-gener-

ators over their lifetime, and also contribute significantly to household

well-being through their substantial non-monetary inputs. At the same time,

their work participation from an early age often inhibits their possibilities

of improvement, through training or job mobility. Their protection in times

of disability or old age is not guaranteed. Severe material deprivation

16 -

* (nutrition and education) can hamper the possibilities of future generations,
and in many cases these conditions are worse for women than for men. Girls

in poor families often exhibit lower rates of school attendance than boys,

and much lower rates than girls in higher-income families.40 Thus poor

women and poor families are threatened in both the short-and long-term.

- 17 -

3. C. Women's Access to Urban Services

Women balance a multiplicity of roles in household welfare and their

economic behavior is to a great extent determined by the needs and constraints

of the domestic unit. Women's productive activities are particularly intense

and essential in the poorest households. Whereas a number of recent studies

have focused on patterns of women's employment in the region, relatively

little is understood about the role played by poor women in a variety of

other activities which contribute to household full income. These include

unpaid labor inputs, the manipulation of extra-domestic networks, the

general day-to-day management of household strategies, and the utilization of

collective urban services. In particular, next to nothing is known about

women's access to a broad range of urban services, and the role they may

play in facilitating income generation and improving household welfare in

both short- and long-term.

The principal responsibility for household management and provisioning

typically falls not only to women who head households, but also to those who

are de facto heads during the time when the head is working or travelling

to and from work. In urban communities located in peripheral areas far from

the workplace, women often function as day-time managers of communities whose

male members leave their homes before daybreak and return after dark. As

a household's monetary income declines, the relative importance of activi-

ties to generate other sources of full income increases; it has already

been pointed out that the unpaid domestic labor performed by women can be

crucial in stretching available household income and helping to insure the

welfare of the domestic group. But urban women's socioeconomic role is far

from limited to these housewife functions. They also help to insure house-

- 18 -

hold welfare through the activation of social networks which provide an impor-

Itant source of emergency loans and other types of exchanges of goods and ser-

vices. Furthermore, women in their management function battle to secure

access to urban services such as education, health facilities, child-care,

housing and infrastructural benefits for themselves and their families. In

many cases this effort involves the cooperative action of community members

in mothers' clubs, housewives' groups and neighborhood associations, in

which women may play a dominant role. It also leads women to activate pat-

ron-client relationships with individuals and institutions and to negotiate

with government agencies and local financial institutions.41 These activi-

ties can provide an essential integrative element within an often unstable

and hostile urban environment. Urban women therefore potentially have an

important social role not only at the level of household welfare, but in

) broader concerns of the urban community as well.

Because of the multiplicity of women's roles in household and commun-

ity welfare, the analysis of their access to urban resources must focus not

only on productive employment but also on access to urban services more broad-

ly defined. Just as urban labor markets are sex-segregated, it is also likely

that access to other urban resources is differentiated by sex. It is there-

fore essential to direct research to the analysis of women's utilization of

urban services, and the role they potentially play in facilitating income

generation, improving household and community welfare in the short-term,

and furthering the long-term urban integration of poor populations.

Statistics examining sex-differentiated patterns of service use are

extremely scarce, making diagnosis difficult at this stage of research.

It is essential to know how use patterns differ, what the implications of

These differences are for women and their families, what factors lead to

- 19 -

differential use patterns, and what kinds of policies might best serve to

correct problems of essential service delivery for women and their families.

All of these tasks must precede any direct policy intervention aimed at

increasing women's access to urban resources.

Based on the scattered evidence which does exist from a variety of
studies of women in Latin America, the remainder of this chapter will present

an overview of the state of information on women's access to selected types

of urban services. This review is not intended to be definitive or complete

but only to draw together some indications of where further research might

be directed. The following chapter takes up in somewhat more detail the

analysis of women's access to urban transport services, focussing on a case

study in Brazil. This is followed by a more general discussion of the

importance and effectiveness of designing service delivery programs which

permit maximum community control (Chapter 6).

- 20 -

p 1. Education
Public education is the exceptional sector for which statistics com-
paring men and women do exist, and studies in a number of Latin American

countries have shown that in general female enrollment rates have become

similar to male during the last two decades of educational expansion in the

region.42 Women's illiteracy has fallen to levels which are not significantly

higher, and in some cases actually lower, than men's in urban areas. The

implications of this apparent equality of access are less clear, however.

The benefits of educational expansion have gone primarily to the "middle

sectors" in Latin America and not to the poor.48 In Colombia, for example,

women are actually a smaller proportion of illiterates than men, but most

still lack a completed primary education, while only 9% had completed middle

levels or higher.44

Even at higher educational levels, women are uniformly concentrated

in a few female categories, notably normal schools. Furthermore, it is

doubtful that increased education necessarily pays off in better employment

opportunities except at the higher levels of the occupational heirarchy,

and here women are limited to those few "female" fields described in Chapter

3. They are unlikely to find an adequate wage return to their investment

in education and mobility may in general be lacking for women in Latin Amer-

ica. In Peru and Colombia, parents exhibited higher mobility aspirations

for their sons than for daughters.45 Yet some studies show that young women

are less likely than young men to be kept out of school in order to work

to contribute to household income.46 This pattern is probably typical of

middle income levels where young women's schooling is, in effect, a form of

disguised unemployment which reflects their lack of job opportunities47 com-

bined with a family income sufficient to maintain nonproductive members.

- 21 -

In poor families, on the other hand, young women may be allocated to poorly-

paid productive work in order to support the more realistic mobility project

centered on their brother's educational attainment.

Households headed by women may be particularly subject to constraints

on educational availability and continuity. Survey data from Belo Hori-

zonte, Brazil showed that households headed by women were much more concentrated

in the lowest income group (61% as opposed to 35%) and by virtue of this skewed

distribution more likely to be lacking in access to all the services surveyed.48

That is, low monetary income in female-headed households is associated with

poor access to urban services. But even when only households in the two

lowest income groups were examined, differences in the utilization of services

between those headed by women and those headed by men still remained; that

is, aside from the "poverty syndrome" to which both types of households are

subject, households with female heads suffer additional relative deprivation.

While these differences were clearest with regard to health services

(see section 3. C. 3), differences in access to education were also suggestive.

Households headed by females in low-income groups were more likely to have no

children registered in school, and much more likely to cite financial prob-

lems as the reason for their absence than were male-headed households (Tables

3.1 and 3.2). They were also less likely to have their children in desired

schools (Table 3.3). Thus the need for multiple earners in female-headed

households appears to affect the extent to which future generations of

workers are able to take advantage of educational services in urban areas,

and female children may be particularly affected by these constraints.

Table 3.1

No. of Children Registered in School by Income
and Sex of Household Head, Belo Horizonte,

Students Females Males

0 46.0 34.8

1 19.8 17.8

2 15.9 19.2

3 or
more 18.2 28.2










3.9 13.5

Income Groups1

Females Males

78.6 54.9

7.1 14.2

10.7 14.8


Females Males

63.2 67.5

26.3 10.8

0.0 9.3

3.6 16.0 10.5


Females Males

63.0 51.8

17.0 15.7

9.7 14.7

12.3 10.3 17.8

99.9 100.0

126 351





99.9 100.0

162 19

99.9 100.0
194 300

1Income categories correspond to measures based on income-per-consumer.

Source: PLAMBEL Survey Data (see Sant'Anna, Merrick and Mazumbar 1976).





Table 3.2

Reasons for Children Not Frequenting School, by
Income and Sex of Household Head,
Belo Horizonte, 1972

Reasons Females--Males

problems 55.5

No school
nearby 5.5

like to
study 5.5

Need to
work -5.5

No vacan-
cy in
school 11.1

Others 16.7





Females Males

come Groups
Middle -
Females Males

15.8 12.5 26.7




- 33.3

10.5 50.0




99.8 100.0
18 38






Females Males

45.8 20.0

4.2 8.6

4.2 22.9

20.8 10.0

4.2 12.9

20.8 25.7

- 100.0




Source: PLAMBEL Survey Data.

Table 3.3

Access to Desired Schools for Children, by Income and
Sex of Household Head,
Belo Horizonte, 1972

All Children
Females Males

75.4 85.0

87.1 80.3

100.0 89.0

85.7 88.9

Some Children
Females Males

8.7 6.4

9.7 7.5
S 5.5

14.3 6.3

80.5 83.7 8.8

91 556 10

No Children
Females Males

15.9 8.6

3.2 12.2


Females Males

100.0 100.0

100.0 100.0

100.0 100.0


99.9 100.0

113 664

6.8 10.6

45 12

Source: PLAMBEL Survey Data.








- 22 -

A The difficulties women face in transferring the benefits of education

to labor force opportunities and earnings are even clearer when the analysis

focuses specifically on vocational training. In the Latin American region,

only Cuba and Costa Rica have official policies related to the training of the

female labor force.49 As a result, statistics on female participation in

vocational training programs are extremely scarce. Recent studies show that

women's participation is generally under 25% but appears to be growing in the

post-1970 period.50 As in the general educational system, women tend to be

concentrated in "female" fields which parallel their limited occupational

options: commerce and services represent more than half of those enrolled.

Within industrial training programs, 60-70% of women are in food and drink,

textiles, and leather and shoes industries.

The problem of women's access to and use of vocational training programs

is inextricably related to their labor force behavior, and policies to improve

women's opportunities must clearly be based on an examination of needs at

both levels. Women's concentration in a restricted range of fields is the

result of a combination of factors including prevailing attitudes about what

occupations are appropriate for women, hiring practices of potential employers,

labor legislation which can restrict women's possibilities compared with men's,

and the broader structural determinants of labor demand.

- 23 -

S2. Child-care services.

One of the most important components of women's household management
role is their function as "primary dispenser of health and nutrition care of
the family."51 Women remain responsible for virtually all tasks related to the
care of children, and financial pressures which force poor women to work
create the serious dilemma of providing alternative arrangements for the care
of children of pre-school age. For mothers, access to some form of day-care
is probably the single most important factor facilitating participation in
income-earning activities. One study in urban Brazil found that over half the
women surveyed were unemployed and the presence of young children was the most
frequently cited reason for unemployment. "Although children may create the
need for a woman to enter the labor force, they may require her to stay home
and thus limit her alternatives for work."52 Adequate day-care facilities are
woefully lacking for poor women in Latin American metropolitan centers, and
virtually nonexistent in smaller urban centers. In Chile, for example, despite
a national commitment to the provision of services to children under the age
of six, existing programs reached less than 5% of that population in 1972,
rising to 12.3% by 1976.53
Brazil lacks a pre-school educational priority at the level of the national
government, to attend the approximately 35 million children in this category in
1980.54 The only legislation dealing with this population is the Labor Law
of 1943 which requires that firms employing at least 30 women provide a day
nursery where female employees may keep their children during the nursing
period. However the law permits the contracting of services to facilities
distant from the workplace, creating a serious problem of transportation
for working mothers. Furthermore, the law provides for child-care only during
the first six months of the child's life. Finally, the law is not enforced

- 24 -

nd the fine for transgressing the norm is so miniscule as to be insignificant.55
As a result, demand for day-care far exceeds supply. In Sao Paulo, for example,
in approximately 150,000 businesses there are only 33 nurseries, while there
is an immediate need for 330,000 vacancies for children under six years old.56
One study found day-care centers existed for only one out of every 28 children
of working mothers in Sao Paulo.57 A 1975 study in Rio Grande do Sul revealed
that only 27% of the state's firms with more than 30 women employees were ful-
filling their legal obligations for child-care provisions. A survey by the
press in 1969 counted "no more than 200 creches in all of Brazil, concentrated
in the three principal cities (Rio, Sao Paulo, and Belo Horizonte)."58 Thus
despite the paucity of comprehensive data on child-care services, there seems
little doubt that their provision is deficient.
Furthermore, since many day-care centers are private institutions which

Dmay be expensive, they are accessible only to middle and upper class families,
while the majority of working mothers must fight for access to a state-sup-
ported center or find individual solutions to their work-home responsibilities.
Government centers are provided by a variety of federal, state and municipal
programs whose responsibilities often overlap. Because of the lack of integrated
planning among the various institutions, programs and resources sometimes overlap
and lack necessary continuity.59 The leading program for child-care provision
is the Brazilian Assistance Legion, with a projected 3100 centers planned for
the end of 1978 to attend nearly 100,000 children.60 Services are most con-
centrated in Sao Paulo, where between 200 and 300 centers exist.61 However,
in 1970 only 1% of working mothers surveyed by the state left their children
in day-care institutions; nearly half (46.6%) left them with relatives, while
21.6% left them unattended in the home.62
Recent studies of child-care arrangements have found that social networks

- 25 -

composed of female kin and neighbors provide working women's main assistance

lin child care as well as other domestic responsibilities.63 None of the working
women surveyed in poor urban neighborhoods of the Dominican Republic and Brazil in 1979
used day-care institutions; by far the most common solution was to leave child-
ren with relatives in their own home or in that of the caretaker. The second
most common alternative was to leave young children at home in the care of older
siblings, an option which potentially interferes with young women's school
or work participation. However a large proportion (one-third in Brazil; nearly
two-thirds in Dominican Republic) of women surveyed cited that they would
prefer to leave their children in a day-care center, if one acceptable to them
were available. Furthermore, institutional facilities for day-care might also
be an important means of improving children's nutritional level. The Brazil-
ian study found children of working mothers more likely to be malnourished,
not apparently because of the time spent away from the child, but because of
Ithe low economic level of the family which forced the mother to work in the

first place.64 Furthermore, in the majority of households whose incomes are
insufficient to purchase necessary foods, most low-income mothers reserve
larger rations for working household members, thus depriving growing children
of necessary nutrition.65
Because of the severity of the problem of day-care availability and the
stimulus provided by International Year of the Child, several institutions
have directed their efforts to outlining the elements needed for appropriate
and effective day-care programs.66 Appropriate child-care may be defined as
"an integrated system of services for preschool age children including health,

nutrition and education, and custodial care, which is responsive to the child's
social, economic and cultural context" and which is usually provided in the
absence of the mother who is involved in income-generating activities (for
cash or kind, inside or outside the home).67 Perhaps the highest priority
target group for the provision of these services is the urban poor with child-

26 -

ren under the age of seven.

) One of the most important characteristics of a successful child-care
program is its community basis, both in terms of location and participation.
In the Dominican Republic, the overwhelming majority (90%) of women preferred
neighborhood locations for day-care centers, in order to avoid transportation
problems. The location of day-care in low-income urban neighborhoods allows
their integration into multisectoral efforts of combined services (such as
training, education, health and nutrition) which can have a far greater impact
on the community's basic necessities. Furthermore, they permit a higher degree
of local participation and authority, an essential element for the continuing
success of the endeavor. Day-care centers may be based on already-existing
reciprocity networks, such as the "spontaneous" creche: women who take in
children of working neighbors in their own home, for monetary payment or other
forms of reciprocal payment. In particular, the community orientation of day-
Icare and other services can provide essential involvement of women at the com-
munity level in positions of authority, as well as sources of training and
direct employment close to home.
Other criteria for evaluating day-care programs for the urban poor in-
clude the necessity for providing full-day service. Part-time care does not
answer the needs-of the full-time working women who also face long hours spent
on transportation to and from work. Such community centers should be based
on small initial investments provided by the government and supplemented by low
cost inputs (in labor or money) by program recipients. Small investments are
advisable in order to spread services as broadly as possible among poor urban
communities. Furthermore, local day-care centers should not confront poor
children with elaborate structures and materials which will create a disjunc-
ture between the home and day-care environments; rather, an effort should be..
made to utilize materials readily available in the local environment which

- 27 -

allow children to continue their activities in the home with younger siblings.

really, because each local community is distinct and evolving, child-care
programs should be approached flexibly, with a constant reassessment of needs

in conjunction with community participants.

Some models of day-care programs exist which incorporate the elements

outlined here, in several Latin American countries. The Inter-American Foun-
dation has financed several small-scale programs in Sao Paulo which emphasize

the importance of community control and participation and which have been

at least initially successful.68 UNICEF has developed a prototype of combined ser-

vices, including day-care, for low-income neighborhoods.69 The Associaci6n Peru-

.Mujer in Lima is designing a child care program which would also deliver health,

education, job training and job placement services in poor neighborhoods.70

Finally, the Overseas Education Fund has described experimental programs which

exist in a variety of Latin American countries.71 In Brazil, several models

exist including the "cocoon units" directed by the Brazilian Assistance Legion,

which use existing buildings and local entities, and require the participation

of mothers, as well as others run by federal and state institutions (COBES,

CEAPE, CSU). These programs are supported by the 1971 Presidential Decree

which provided financing from federal, state and municipal sources for mater-

nal-health care programs under the direction of the Health Ministry in cooper-

ation with local entities. In Colombia financing of day-care services has

been provided by a 2% tax on monthly salaries paid by public and private firms

(regardless of their employment of women) and dispensed to local entities

by the Colombian Institute of Family Welfare. Both this program and that of

the Venezuelan Child's Foundation "Day Care Homes Program" use neighborhood

homes and local non-professional women who are trained and supervised for the

Purposes of the program. While all of these programs present problems, particu-
larly in effectively permitting local control and participation, they are

28 -

significant attempts to respond to an essential service need of poor women
and their families.

- 29 -

3. Health services

f While data on women's demand for and use of health services is fragmentary,

they incontestably predominate over men as clients due to their role in bio-

logical reproduction as well as mediation in the health of other family members.

A study in Colombia, for example, found women's demand for outpatient services

to be twice that of men's due to pregnancy, a proportion which was even higher in

urban areas. Furthermore, women in childbearing ages (15-44) were three times

more common in hospitals than were men.72 At the same time, women's direct

access to government-subsidized health services is much less than that for men,

because of the link between social security health programs and employment in

the formal sector. In Lima, 87.8% of manual workers eligible for social security

benefits are men, and 61.8% of eligible white-collar workers are men.73 Similarly,

80% of Colombian women-are not directly affiliated with social security benefits,

compared with 61% of men.74
While many women have access to health benefits indirectly through their

working husbands, this lack of direct access is critical for the many women

whose husbands are also ineligible or who temporarily or permanently find

themselves without the protection of a male partner. For this reason the

analysis of differences between male and female-headed households is a useful

proxy for the direct comparison of sex differences in use of services. Analysis

of Belo Horizonte survey data comparing households headed by males and females

showed striking differences in access to health services.
Households headed by women were much less likely to have access to the

main government-sponsored social security program (INPS) in all income groups;

in the lowest category, male-headed households were more than four times as

likely to have access (see Table 3.4). This striking difference is clearly a

reflection of employment patterns which relegate poor women to informal sector

jobs. As a consequence, male-headed households were about twice as likely

Table 3.4

% Contributing for Government Programs, by Income
Groups and Sex of Household Head,
Belo Horizonte, 1972

Poor Low Middle High
Females Males Females-Males Females-lales Females Males

13.5 57.0 37.8 69.0 50.0 78.4 52.6 82.5

0.8 5.7 4.7 4.6 8.0 21.1 6.7

6.3 4.6 13.4 5.0 21.4 7.4 15.8 12.9

Females Males

29.7 69.0

3.7 5.6

11.3 6.3

Source: PLAMBEL Survey Data.




30 -

to use INPS medical, laboratory and hospitalization services, whereas those

headed by women were twice as likely to use services provided by religious

and charity organizations or government health posts (Tables 3.5, 3.6, and


Furthermore, aside from institutional access, households headed by

women showed distinct patterns of use of other types of health services

(Table 3.8). They were less likely to turn to a pharmacist and more likely to

resort to a friend or relative, or ritual curer. This pattern probably re-

flects the differential cost of services for poor households. It may also

suggest a preference for community-based health specialists. A study of

low-income women in Campinas, Brazil found that doctors are often distrusted as

representing part of the alien urban environment, although the efficacy of

*patent medicines and prescription drugs is not questioned. Instead, the great

majority prefer to visit the neighborhood pharmacist or community healer,

who offer credit and are familiar with the medical history of family members.75

Evidence therefore suggests that health services, like day-care, can

more effectively reach those who need them if offered on a community basis.

The decentralization of some health services could provide many of the advantages

of local control and participation described above, including the potential

recruitment of local women as non-professional and para-professional personnel.

These measures could improve both the cost and the appropriateness of health

services used by poor women and their families.

Table 3.5

Medical Services Used, by Income Groups and
Sex of Household Head,
Belo Horizonte, 1972

Low Poor Middle Hi h
Females-Tales Females Males Females -Wales Femaleii ales

36.5 67.2 58.3 70.0 53.6 54.3 42.1 29.4

Females Males

47.7 61.6

Posto de Saude
or rel.

56.3 23.6 18.9 13.4

7.1 5.6

- 3.1 32.3 13.6

7.1 8.3 22.8 16.7 39.3

39.5 57.9 67.0

0.9 0.6 0.5

100.0 100.0 100.0 100.0

(126) (351) (127) (659)



Source: PLAMBEL Survey Data.


None -


20.0 24.4







100.0 100.0

(300) (1366)

Table 3.6

Type of Medical Laboratory Services Used, by Income Groups
and Sex of Household Head,
Belo Horizonte, 1972

Poor Low Middle High
Females Males Females Males Females Males Females Males

35.7 67.2 57.5 69.7 57.1 50.6 47.4 26.3

52.4 22.5 19.7 14.0

8.7 8.5 19.7 15.8

3.2 1.7 3.1 0.6









3.6 3.7

- 2.6

39.3 44.4 52.6 71.1







Females Males

47.7 60.6

30.7 13.3

19.0 25.2

2.7 0.9

100.0 100.0

(300) (1366)



Source: PLAMBEL Survey Data.


Postos de




Table 3.7

Type of Hospitalization Used, by Income Groups and
Sex of Household Head,
Belo Horizonte, 1972

Poor Low Middle High
Females Males Female Males Females Males Females Males

INPS 37.3 70.7 69.3 75.7 60.7 69.8 57.9 52.1

Rel. Charity





54.0 22.2 17.3 11.1 3.6 3.7

- 3.1

4.0 6.0 11.0 11.8 28.6 25.3 42.1 44.3

4.8 1.1 2.4 1.4 7.1 1.2

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

(126) (351) (127) (659) (28) (162) (19) (194)

Females Males

54.3 70.4

30.3 11.9

11.7 16.5

3.7 1.2

100.0 100.0

(300) (1366)

Source: PLAMBEL Survey Data.

Table 3.8

Other Health Services Used, by Income Groups and
Sex of Household Head,
Belo Horizonte, 1972

Poor Low Middle High
Females- ales Females-ales FemalesMales Females Males

Pharmacist 29.4 45.9 34.6 44.0 17.9 38.3 15.8 18.6

Midwife 0.8 2.6 2.4 1.4

Friend or


9.5 4.6 3.9 3.9

- 1.2


Females Males

29.7 40.2

- 0.5

- 4.1

1.3 1.5

5.7 3.8

15.1 5.7 8.7 5.5 7.1 3.1 5.3 3.6

Nobody 45.2 41.3 50.4 45.2









75.0 56.2 78.9 73.2



11.0 5.0

52.3 49.5

100.0 100.0

(300) (1366)



Source: PLAMBEL Survey Data.

31 -

4. Housing and Infrastructural services.

Assessment of women's access to basic infrastructural services in urban

areas is particularly difficult on two counts. First, whereas women's special

needs with regard to services such as health and child-care are readily percep-

tible, women's issues in housing and transport are much less obvious. Second-

ly, almost no studies exist which have focused on sex differences in demand for

and use of such services. For many service sectors, exploratory research is

needed before the relevance of specific service provisions to women as a sub-

group can be evaluated. The case study of the transport sector reported in

the next chapter of this report is one such preliminary attempt. Experimenta-

tion with data on households headed by males and females revealed no clear

differences with respect to such infrastructural services as water, sewage

and electricity; for these services the head-of-household proxy is not useful

in the analysis of sex differences. It may be true that these basic services

are not sex-discriminated because of their spatial distribution to whole

communities rather than to individuals or families.

Urban housing is a crucial element in the welfare of poor families, and

one which is provided directly to families or households within target com-

munities. Studies of how women fare relative to men in housing programs are

scarce, but a few sources exist which can suggest some important considerations

for planners seeking to reach poor populations.76 Both the need for and ac-

cess to sites and services projects may be differentiated by sex. That is,

in general in evaluating socioeconomic characteristics of poor target popula-

tions, attention should also be directed to the extent of female domination of

migrant streams, the proportion of female-headed households, the existing

sex differentiation of service and employment access in the local or national

- 32 -

context. Particularly with regard to housing, life-cycle characteristics

can be an important variable affecting demand, determining both the motiva-

tion for moving within the urban system as well as the capacity to pay for

housing and infrastructural services.77 Whereas more mature families are more

likely to have sufficient income to afford their own housing, adequate facil-

ities for families with young children may be beyond their means; these families

might then receive special attention in housing projects. Similarly, house-

holds headed by women may require special consideration in making sites and

services programs accessible, such as more flexible construction scheduling

in self-help programs, and special credit arrangements. The latter would be

based on an analysis of the requirements imposed by official institutions

in relation to collateral, the need for a marital guarantor, and the impact

of marital status on property rights and access to credit, as well as the

possibility of reinforcing alternate sources of credit such as informal sav-

ings associations.

Furthermore, there are good reasons to consider women's special roles

and needs at the design stage of sites and services programs. First, the

location of such projects should take into consideration residents' access

to sources of employment (for women as well as men), to marketing, and to

such necessities as water and domestic fuels. In designing the layout of

planned communities, women's patterns of utilization of household and community

space should be crucial considerations. These may dictate, for example, the

need for locally-based cottage industries, for culturally appropriate public

gathering places for women, for sufficient privacy from neighbors and within

the household itself, and particularly for attention to the needs of women's

domestic activities which might be facilitated by design modifications of

housing layouts. Women as household managers should be consulted about how -


33 -

housing designs can improve the ease of supervision and care of children,

the amount of upkeep differently designed houses require, the size and place-

ment of kitchens, the placement of water and bathing facilities, and the space

available for kitchen gardens and care of small domestic animals.

- 34 -

Notes to Chapter 3

1. Michel Chossudovsky, "Transnationalization and the development of peri-
pheral capitalism," Research Paper no. 7903, Faculty of Social Sciences, De-
partment of Economics, University of Ottawa, 1979.

2. Marianne Schmink, Community in Ascendance: Urban Industrial Growth and
Household Income Strategies in Belo Horizonte, Brazil. Ph. D. Dissertation,
University of Texas at Austin, 1979.

3. The remainder of this section draws largely on earlier works by the author
(Schmink 1979, and Marianne Schmink, "Women and Urban Industrial Development
in Brazil," revised version of paper presented at the Latin American Studies
Association meeting, 1980) and on the papers and discussion from two Latin
American conferences: "Women in the Labor Force in Latin America," sponsored
by IUPERJ, Rio de Janeiro, November 23-26, 1978; and "Life Conditions of the
Urban Popular Sectors," sponsored by CEDES, Buenos Aires, December 4-7, 1979.

4. Ester Boserup, Woman's Role in Economic Development, New York: St. Martin's
Press, 1970; International Center for Research on Women (ICRW), "Women in
Migration: A Third World Focus," Report submitted to AID/WID, June, 1979.

5. Boserup 1970.

6. Carmen Diana Deere, "Changing Social Relations of Production and Peruvian
Peasant Women's Work," Latin American Perspectives IV: 1 and 2 (Winter and
I Spring, 1977): 48-69; Helena Lewin, Jaqueline Pitanguy and Carlos Manuel
Romani, Mao-de-Obra no Brasil: Um Inventario Critico, Petropolis: Vozes, 1977.

7. ICRW 1979: 88.

8. Lourdes Arizpe, "La migracion por relevos y la reproduccion social del
campesinado." Mimeo, 1978.
9. Verena Martinez-Alier, "As mulheres do caminhao de turma," Debate & Critica
5 (March): 59-85, 1975.

10. ICRW 1979.

11. ICRW 1979: 73.

12. ICRW 1979: 89-90.

13. Maria Christina A Bruschini, "Sexualizacao das ocupacoes: o caso brasileiro,"
presented at the IUPERJ conference on "Women in the Labor Force in Latin America,"
Rio de Janeiro, November, 1978.

14. Boserup 1970.

15. Brazil, Ministerio do Trabalho, Formacao Profissional da Mulher Trabal-
hadora ho 3rasil, Rio: SENAI/DN, 1976, Part IV; Marly Cardone, "Subsidios do
Direito do Trabalho para um debate sobre a situacao da mulher," Fundacao Carlos
Chagas, Cadernos de Pesquisa 15 (Dec. 1975): 124-131.

- 35 -

16. Bruschini 1978; Marianne Schmink, "Dependent Development and the Division
of Labor by Sex: Venezuela," Latin American Perspectives IV: 1 and 2 (Winter
and Spring, 1977): 153-179.

17. Lourdes Arizpe, "Women in the Informal Labor Sector: The Case of Mexico
City," Signs 3: 1 (Autumn, 1977): 25-37.

18. Schmink 1977.

19. Alice de Paiva Abreu, "Mao-de-obra feminine e mercado de trabalho no
Brasil," SENAC, Boletim Tecnico 3: 1 (Jan/April, 1977): 5-19; Manuel Barrera,
"Diferencias salariales entire hombres y mujeres en America Latina," paper pre-
sented at the IUPERJ conference on "Women in the Labor Force in Latin America,"
November 1978; Maria Helena da Cunha Rato, "A participacao feminine na popu-
lacao ativa frente as necessidades do sistema produtivo no Brasil," paper
presented at the IUPERJ conference on "Women in the Labor Force in Latin
America," November 1978; Heleieth I.B. Saffioti, A Mulher na Sociedade de
Classes, Mito e Realidade. Sao Paulo: Quatro Artes, 1969.

20. Brigida Garcia, Humberto Munoz and Orlandina de Oliveira, "La familiar
obrera y la reproduccion de la fuerza de trabajo en la ciudad de Mexico,"
paper presented at the conference on "Development and Inequality in latin
America," Gainesville, Florida, October, 1979 and "Migracion, familiar y fuerza
de trabajo en la ciudad de Mexico," Cuadernos del CES, 26, 1S79; Lucia Ribeiro
de Souza, "0 trabalho feminine e a estrutura familiar," paper presented at the
Meeting of the National Association of Post-Graduation and Research in Social
Sciences, Belo Horizonte, Brazil, October, 1979.

21. The following discussion of household strategies is taken from Schmink
1979 and Schmink 1980.

22. See Elise Boulding, The Underside of History.

23. Joaquin Duque and Ernesto Pastrana, "Las estrategias de supervivencia
economicadelas unidades familiares del sector popular urban: una investi-
gacion exploratoria," FLACSO, Santiago, mineo, 1973; Ana Maria Quiroga Fausto
Neto, Familia Operaria e Reproducao da Forca de Trabalho, Master's Thesis,
Universidade de Brasilia, 1978; Larissa Adler Lomnitz, Networks and Marginality,
New York: Academic Press, 1977.

24. Elisabete Doria Bilac, Familias de Trabalhadores: Estrategias de Sobre-
vivencia. A Organizacao da Vida Familiar em una Cidade Paulists, Sao Paulo:
Simbolo, 1978.

25. Carmen Diana Deere, Jane Humphries and Magdalena Leon de Leal, "Class
and Historical Analysis for the Study of Women and Economic Change," paper
prepared for the Role of Women and Demographic Change Research Program, ILO,
Geneva, 1978.

26. Joann Vaneck, "Time spent in housework," Scientific American 233: 5:
116-120, 1974.

- 36 -

27. Fausto Neto 1978; Grupo de Estudos e Trabalho em Educacao Comunitaria
(GETEC), Mutirao, Belo Horizonte: Cadernos GETEC, 1978; Lomnitz 1977; Francisco
de Oliveira, "A economic brasileira: critical a razao dualista," Selecoes
CEBRAP, 1: 5-78, 1975; Paul Singer, Economia Politica do Trabalho, Sao Paulo:
Hucitec, 1977.

28. Anthony Leeds, "Brazilian careers and social structure: an evolutionary
model and a case history," American Anthropologist 66: 1964; Larissa Adler
Lomnitz, "Reciprocity of favors in the urban middle class of Chile," Pp. 93-
106 in George Dalton (ed.), Studies in Economic Anthropology, Anthropological
Studies No. 7, Washington, D.C.: American Anthropoligical Association, 1971;
Charlotte Ingrid Miller, Middle Class Kinship Networks in Belo Horizonte,
Minas Gerais, Brazil: The Functions of the Urban Parentela. Ph. D. Disser-
tation, University of Florida, 1976.

29. Bilac 1978.

30. Rolando Franco, Agustin Llona and Irma Arriagada, "La mujer en situation
de extrema pobreza: el caso de Chile," in Paz Covarrubias and Rolando Franco
(eds.), Chile: Mujer y Sociedad, Santiago: UNICEF, 1978: 337; Fausto Neto
1978; Felicia Reicher Madeira, As Condicoes do Trabalho da Mulher e as Con-
dicoes de Vida da Familia, Master's Thesis, Universidade de Sao Paulo, 1979;
Zahide Machado Neto, "As meninas sobre o trabalho da crianca e da adoles-
cente na familiar proletaria," presented at the IUPERJ conference on "Women
in the Labor Force in Latim America," Rio de Janeiro, November, 1978; Arakcy
Martins Rodrigues, "0 padrao de distribuicao de papeis em families operarias,"
presented at the IUPERJ conference on "Women in the Labor Force in Latin
America,"Rio de Janeiro, November, 1978.

31. Machado Neto 1978.

32. E. A. Cebotarev, "La Organizacion del Tiempo de Actividades Domesticas
y No-Domesticas de Mujeres Campesinas en Latinoamerica," presented at the
IUPERJ conference on "Women in the Labor Force in Latin America," Rio de
Janeiro, November, 1978.

33. Nidia Aylwin de Barros, "El costo social del actual modelo de desarrollo
en un sector urban de extrema pobreza," Escuela de Trabajo Social, Universidad
Catolica, Santiago, Mimeo, 1979; Sonia Bittencourt, Child Care Needs of Low
Income Women: Urban Brazil, Washington, D.C.: Overseas Education Fund, 1979;
Garcia et al. 1979a; 1979b; Schmink 1980.

34. Madeira 1979.

35. Unpublished time budget surveys by the author.

36. Anna M. Sant'Anna, Thomas W. Merrick and Dipak Mazumbar, "Income Distribu-
tion and the Economy of the Urban Household: The Case of Belo Horizonte," Wash-
ington, D.C.: The World Bank, Working Paper no 236, 1976.

37. Sant'Anna et al. 1976.

- 37 -

038. Carmen Barroso, "Sozinhas.ou mal acompanhadas a situacao da mulher
chefe de familiar presented at the IUPERJ conference on "Women in the Labor
Force in Latin America," Rio de Janeiro, November 1978; Mayra Buvinic and Nadia
H. Youssef, Women-Headed Households: The Ignored Factor in Development, plan-
ning report submitted to AID/WID, 1978; Thomas W. Merrick and Marianne Schmink,
"Female-headed Households and Urban Poverty in Brazil," presented at the ICRW
conference on "Women in Poverty: What Do We Know?", Elkridge, Maryland, April-
May, 1978.

39. Merrick and Schmink 1978.

40. Franco et al. 1978: 353.

41. Parry Scott, "A producao domestic e a mulher no Recife," presented at
the meeting of the National Association of Post-Graduation and Research in
Social Sciences, Belo Horizonte, October, 1979.

42. This discussion is based on the following basic sources: Ofelia Machado
Bonet, Hacia la Revolucion del Siglo, Montevideo: Editorial Goes, 1972; Brazil,
Ministerio do Trabalho, 1976; Nestor Campiglia, "Algunos problems de al
formacion professional de la mujer en America Latina," Boletin Cinterfor,
Montevideo, 37-38 (Jan/April, 1975): 63-84; Maria Angelica Ducci, "Formacao
professional feminine; repercussoes socials no mercado de trabalho," Boletin
Tecnico SENAC, Rio de Janeiro 3: 1 (Jan/April, 1977): 45-70; Hernando Ochoa
Nunez, "La mujer en el sistema educativo," Chapter III, pp. 71-122 in Magda-
lena Leon de Leal (ed.), La Mujer y el Desarrollo en Colombia, Bogota: ACEP,
1977; Ernesto Schiefelbein, "La mujer en la education primaria y media," pp.
693-713 in Paz Covarrubias and Rolando Franco (eds.), Chile: Mujer y Sociedad,
Santiago: UNICEF, 1978; Rosa M. Ribeiro da Silva, Formacao professional
da mulher trabalhadora/America Latina, Recomendacoes do seminario regional
do Cinterfor projeto 102," Boletim Tecnico SENAC, Rio de Janeiro, 3: 1 (Jan/
April, 1977): 33-44; Gabriela Villalobos de Urrutia, Diagnostico de la Situ-
acion Social y Economica de la Mujer Peruana, Lima: Centro de Estudios de
Poblacion y Desarrollo, 1975.

43. Urrutia 1975: 1-2.

44. Nunez 1977: 88; 109.

45. Nunez 1977: 117; Urrutia 1975: 1-11.

46. Schiefelbein 1978: 697, 708; Brazil, Ministerio do Trabalho 1976: 46.

47. Madeira 1979.

48. Merrick and Schmink 1978.

49. Ducci 1977: 54.

50. Ducci 1977: 50-52.

51. Bittencourt 1979: 1.

52. Bittencourt 1979: 21.

- 38

53. Josefina Aragoneses, Carmen Fischer, Dina Alarcon and Alicia Navarro,
"El desarrollo do la education parvularia en Chile y su influencia en el
desarrollo de la mujer," in Paz Covarrubias and Rolando Franco (eds.), Chile:
Mujer y Sociedad, Santiago: UNICEF, 1978: 491.

54. Maria Theresa Egger-Moellwald and Vera Lucia Soares Raucci, "A guard
dos filhos de 0 a 6 anos: um problema a resolver: 1 e 2 relatorios," re-
search report submitted to Carlos Chagas Foundation, Sao Paulo, 1979; Bit-
tencourt 1979: 31.

55. Maria Machado Malta Campos, "As creches no Brasil," deposition presented
to the Mixed Parliamentary Inquiry Commission on the Situation of Women, April
28, 1977; Egger-Moelwald and Raucci 1979.

56. Bittencourt 1979: 31.

57. Study by Sao Paulo Social Welfare Secretariat, cited in Nos Mulheres
2: 10, 1976.

58. Campos 1977: 5.

59. Egger-Moelwald and Raucci 1979.

60. Bittencourt 1979: 34.

61. Egger-Moelwald and Raucci 1979.

62. Secretaria do Trabalho e Administracao, Sao Paulo, "Mao-de-Obra Feminina -
Levantamento Socioeconomico em Industrias da Capital," 1970.

63. Bittencourt 1979; Vivian M. Mota, The Burdened Women: Women's Work and
Child Care in the Dominican Republic, Washington, D.C.: Overseas Education
Fund, 1979; Machado Neto 1978.

64. Bittencourt 1979: 44.

65. Escola Paulista de Medicina, Instituto de Medicina Preventiva, Univer-
sidade de Sao Paulo, Instituto de Pesquisas Economicas, "Estado Nutricional
de Criancas de 6 a 60 meses no Municipio de Sao Paulo," vol. II, Data Analysis,
Sao Paulo 1975, cited in Bittencourt 1979; see also Daniel R. Gross and Bar-
bara A. Underwood, "Technological Change and Caloric Costs: Sisal Agricultural
in Northeastern Brazil," American Anthropologist 73 (1971): 725-740.

66. This discussion is based on the following sources: Bittencourt 1979;
Egger-Moelwald and Raucci 1979; Mary Racelis Hollnsteiner, "Basic services
for children of the urban poor in developing countries," United Nations,
I/ICEF/L.1371, April 1978; Katherine Douglass Massel, "1979: International
Year of the Child," Americas 31: 8 (August): 9-16, 1979; Mota 1979; "The
urban child in IYC," The Urban Edge 3: 9 (November, 1979): 1-5.

67. Bittencourt 1979: ii.

68. Inter-American Foundation, Internal Program Analysis and Evaluation
69. Hollnsteiner 1978; "The UNICEF experience," The Urban Edge 3: 9 (Novem-
ber, 1979): 5-6.

- 39 -

)70. Jeanine Anderson, personal communication.

71. "The urban child in IYC," The Urban Edge 3: 9 (November, 1979): 1-5.

72. Franz Pardo Tellez, "Condiciones de salud de la mujer," Chapter IV, pp.
123-182 in Magdalena Leon de Leal (ed.), La Mujer y el Desarrollo en Columbia,
Bogota: ACEP, 1977.

73. Urrutia 1975: III-2.

74. Tellez 1977: 176.

75. Sumerlynne Solop Harrison, "The family and working-class culture: a
study of industrial workers in Campinas, Brazil," research proposal, 1979.

76. The following discussion is based on Barbara Furst, "The economic parti-
cipation of low income urban women in the developing countries," excerpt from
a draft report submitted to the Office of Urban Development, Bureau for
Development Support, USAID, June 1979 and Lucy Conger, Foundation for Cooper-
ative Housing, "Women-in-development issues for incorporation into the shelter
sector assessment guidelines," submitted to USAID, 1979.

77. See Schmink 1979, Chapter 6.

Chapter 4

Women and Transport: A Case Study

Transport facilities are an essential integrative element in the

urban system. From the point of view of the planner, urban transport is

an important mechanism for inducing growth and incorporation of the city's

diverse snatial ones. From the point of view of noor populations, it is

fundamental in permitting movement between place of residence and iob

site, and also facilitates access to basic services including medical

facilities, schools, and shopping locales.

For women, transport expedites access to income-oeneratino activities

and a variety of other tasks such as shopnina, health care and bureaucratic

errands which women carry out in their household management roles. How-

ever, the extent to which transport needs and patterns of utilization are

differentiated by sex has not been a focus of analysis by planners. In-

deed, research on women and transport is extremely rare and limited to

developed countries.1

Because of the importance of transnort services, both from the point

of view of planners and also of low-income populations, and because the

transport sector represents relatively unexplored terrain with renard

to women's special needs, research efforts were directed to a preliminary

case study of urban transport. This chapter outlines the most relevant

issues in olanninn, focuses attention on the importance of transport for

noor women, and discusses existing sources of information on urban trans-

port. Some experimentation with special tabulations from Latin American

survey data is carried out. Finally, the chanter ends with a presentation

of findings from a pilot study of transport conditions in three noor


neighborhoods in Belo Horizonte, carried out in 1979. The method-

ology of the study is presented in more detail in an appendix to this


- 3-

4. A. Overview of Urban Transport Planning in Latin America

From the second half of the 19th century through the first two

decades of this century, Latin American cities were served by urban

transport systems largely dependent on streetcars and commuter trains.2

Beginning in the 1920's, these systems were supplemented by motor buses in

metropolitan areas. After 1950, the motorized bus rapidly supplanted

the earlier transport modes, so that "today it scarcely seems possible

that only a few decades ago the streetcar was the most important trans-

portation system in many Latin American cities."3

The emerging reliance on gasoline-powered vehicles is the most

prominent feature of contemporary Latin American urban transport

planning. This change is attributed to a number of factors related to

the spurt of urban industrial growth experienced in the region beginning

after the 1940s. "Cheap petroleum, the greater rate of technological

evolution of the motor bus, the development of congestion and the inflexi-

bility of the tracked right-of-way all contributed to the decline and

fall of the tramways in Latin America."4 Also cited as factors are

the relative lack of electric energy and the prevailing image of the

streetcar as the transport mode of the poor, compared to the more modern

image of the motorized bus and automobile.5 Furthermore, the shift was

congruent with the expansion of industrialization, particularly the

auto industry, in Brazil and other countries during this period.

Scholars writing three decades later have pointed out the negative

side effects of this reliance on gasoline-powered modes, even before

rising petroleum prices restricted their long-term feasibility. The

priority placed on the motorized vehicle amounted to a "rejection of

the mass public transport option."6 The decision to stimulate bus


transport as the principal public mode and the automobile as the

private mode has been postulated to have a potentially negative

impact on the distribution of income with the urban system.7 Com-

petition from these transport modes, and the fiscal abandonment of

streetcar and commuter train systems, have virtually eliminated

cheaper alternative modes for low-income users. Furthermore, car

ownership has been called a "regressive influence on the distribu-

tion of the fruits of economic progress" since the affluent make

more trips, especially in private cars, and especially during peak

traffic hours. The added congestion of these private vehicles slows

bus service to poorer populations.

This problem is compounded by the spatial distribution of Latin

American cities, where the poor are typically relegated to distant

Peripheral areas.8 While middle and higher-income populations have

cheaper collective transport means at their disposal, and also can use

private transport, lower income groups pay more elevated costs for an

inefficient and precarious transport system. In Brazil, differentiated

tariff policies mean higher prices for the poor who live in more distant

areas (and often have to transfer buses to reach their destination).9

Similarly, studies of time expenditures in Sao Paulo urban transport

show they tend to be higher for lower-income populations.10 The net

effect is to transfer the highest transport costs to low-income popu-

lations which can ill afford them. Thus some Brazilian poor families

pay 20-25; of their income for urban transport.1

The role of the state in urban transport planning in Latin America

has historically been indirect but is gradually expanding in the region.12

For this reason, the impact of the transport problems outlined above

- 5 -

differs from patterns in developed countries "largely due to the fact

that bus services are rarely operated by the public authorities with

certain social objectives in mind, but rather are, for the most part,

operated by private companies and individuals under loose public

supervision."3 In metropolitan Brazil, as in many other Latin American

cities, transport services have been delegated to monopoly firms whose

goals are profit maximization rather than service provision to the needy.

Planners typically view transport services as potentially inducing

urban development and redistributing income through the optimization of

time and distance in workplace-home movements.15 However these potential

beneficial effects may be attenuated by the dependence on criteria of

"cost returns" which lead to a distribution of vehicles and routing

in favor of middle and higher income strata, to the detriment of poor

populations. The skewed income distribution patterns characteristic

of Latin America exacerbate these problems both directly and indirect-

ly. The existence of high-income groups creates a market for special

elite-oriented collective transport services (i.e., more comfortable

air-conditioned buses on special routes) which planning authorities

allow firms to operate, in order to avoid raising fares already too

high for the poorest urban populations.16 Transport firms are there-

fore led to invest in these special services while tariff controls on

regular collective services reduce the incentive for improvement.

These factors reinforce the tendency for low-income populations to pay

higher costs in time expenditure for their travel needs. In Sao Paulo.

for example, average travelling time increased more than 30 from

1970 to 1976, and workers commonly spend three to four hours per day

travelling to and from work.17

- 6 -

In some cases, transport improvements may actually have the reverse

of the intended beneficial effect. Transport investments usually act

as a stimulant to speculation and rising land values.1R In the long

run, poor families are unable to afford to stay in improved areas, and

are pushed to even more distant neighborhoods. To transport planners

who are unaware of the adverse effects of technical investments, this

pattern appears as an irrational response by poor populations which

prefer to profit from speculation and continue to pay high costs in

time for transport. In fact, poor families are pushed out by rising

rents, and financial crises force them to sell their property in the

inflated market.

The focus of planning and services is on the home-work trajectory,

which accounts for the bulk of urban travel during compressed periods of

the day. Working class transport is reduced to the "squeeze" hour when

"thousands of workers dispute a place on public transport to reach the
workplace where they will earn their daily bread."19 However, the deter-

ioration of urban transport systems in relation to the growing demand in

large urban centers has meant severely inadequate service and great hard-

ship to workers.

Queues, overcrowding, delays, losing a day's work, and, at
times the anger which explodes into the stoning of trains and
buses, cannot be dismissed as simple 'traffic problems.' The
long hours spent waiting or travelling when one could be resting,
whether before or after a long and arduous day's work, add daily
to the exhaustion of those who have to use buses and trains to
get to their jobs.20

If transport from home to job is inadequate for many workers,

planning for urban travel for other purposes is virtually nonexistent,

despite the importance of travel to school, health services, and shop-

ping. Indeed, the lack of adequate transport facilities can make the

- 7 -

provision of other services useless to populations which cannot effectively

make use of them. After the rush hour, transport firms typically withdraw

many vehicles from circulation, so that time investments in non-work

trips are also increased.

In short, most planning of urban transport systems is based on only

one kind of trip: the journey to work of formally employed workers. It

is possible that a closer analysis of patterns of use by different types

of travellers would suggest ways in which transport systems could im-

prove poor populations' access both to income-generating activities and

to necessary urban services which will enhance overall household and

community well-being. The differences in women's and men's needs and

use of urban transport are largely unknown. Yet women's employment is

likely to differ in both localization and scheduling from men's. Lack

of attention to these needs may mean that deficiencies of transport sys-

tems further restrict women's labor force opportunities. In European

countries, studies have shown that women weigh transport availability as I

an important consideration in their choice of paid employment.21 In Latin

America, emerging studies have found the problem of transport impeding

employment opportunities further complicated by the restricted availa-

bility of day-care facilities in accessible locations. In short, where-

as male workers must spend large amounts of time moving to and from

the workplace, women who balance their income-generating activities with

other essential roles in family well-being simply cannot afford to spend

the same time in transport.

Even non-working women play an important role through other kinds

of trips mentioned here. They are often in charge of household pro-

visioning but may be restricted to more expensive local shops because

of transport difficulties. Travel related to health care, either for

themselves or accompanying children, is also most often women's responsi-

- 8 -

ability. In fact the clearest symbol of the transport problem for poor

women is that of the woman who gives birth inside the bus, a story common

to most poor urban neighborhoods in Brazil.22

Few studies exist which examine the broader role of transport service'

for low-income urban families. Planners in the transport sector are

aware of the importance of their services and often eager for insights

which can be of use in designing urban transport systems. Although data

are routinely collected on travel patterns in most Latin American cities,

their analysis is rarely disaggregated sufficiently to allow a diagnosis

of the differences in transport use by different household members. In

any case these data may be inadequate to address some of the questions

raised here. The analysis in this chapter is intended to begin exploring

the uses to which existing data sets may be put, and to experiment with

additional methods of data collection which were tested in an original

pilot study.

- 9 -

4. B. Sources of Data on Urban Transport

A variety of potential sources of information on urban transport

patterns exist in Brazil. Few of these were consulted directly for this

study, but they will be mentioned by way of introduction. Some sources

deal with specific transport systems in certain metropolitan areas. Thus

studies of the subways under construction in Rio and Sao Paulo have

been carried out by the metro authorities in each city. In Brasilia,

CODEPLAN carried out two recent studies of the cost of transport to

residents of the capital's satellite cities. A study of urban rail

transport was commissioned by the Federal Railway Network (Rede Ferro-

viaria Federal) to ENGEVIX, a Rio consulting firm, which published 22

volumes of data. These data were later consulted by the IBGE Department

of Social Indicators for their study of the conditions of the urban poor.

Questions related to transport expenditures were also included in the

ENDEF survey.

A few studies have focused more specifically on the impact of

transport on the user population, developing new methodologies which go

beyond the standard measures of time and money expenditures. One is a

longitudinal impact study of the Sao Paulo subway being carried out by

the Urban Development Program (PRODEUR) of the Arquitecture and Urbanism

Faculty of the University of Sao Paulo. PRODEUR researchers chose to

focus on one urban neighborhood located at the end of one of the metro

lines, which had also been the target of a package of community develop-

ment measures. The study aims to accompany the long-term impact of these

changes through the application of an initial questionnaire to residents,

with a follow-up on neighborhood population turnover. For a period of

- 10 -

several years they will study those families which enter and leave the


A second innovative methodological approach to transport questions

is that of an impact study of the Rio subway, carried out by the Brazilian

Institute of Municipal Administration (IBAM) in Rio.23 The study focused

on the impact Of transport changes on diverse social groups. Of particuli

interest is their development of open-ended techniques for the analysis

of transport "perceptions" by different social actors in the communities

studied. This approach permits the differentiation of transport-related

behavior among groups with different age, sex, and occupational character-

istics. While the sex variable was not systematically analyzed in the

published reports from the IBAM study, such an analysis would be possible

through a more formal consultation of the raw interview data on file.

For the purposes of this study it was decided to concentrate on trans-

port studies of more general applicability to explore their usefulness in

addressing the problem of women's transport needs and behavior. This

criterion led to a focus on the widely-used "origin-destination" studies

of urban transport movements. These studies permit a "user approach"

because they are surveys carried out at the user's residence rather than

at transport sites such as bus stops. They typically include information

collected at both individual and household levels. That is, basic demo-

graphic information is collected for all household members, as well

as housing and infrastructural characteristics. "Trip data" containing

information on all movements by household members on the day preceding

the survey is also collected. This information includes the time of

arrival and departure, the mode(s) of transport used, the motivation for

the trip, and often more detailed information on the occupation and em-

- 11 -

ployment status of the traveller.

While the precise formulation of the questionnaire differs, this

basic household survey is standard fare in large cities all over the

world. Through the Transport Ministry and local planning agencies, most

urban centers of any importance in Brazil have carried out household

transport surveys in recent years. Preliminary inquiries with trans-

port agencies in Brasilia unearthed recent surveys in the following

Brazilian cities of very different characteristics and regions: Rauru,

Belo Horizonte, Brasilia, Campina Grande, Campinas, Caruaru, Caxias,

Cuiaba, Florianopolis, Fortaleza, Maceio, Pilotas, Porto Alegre, Recife,

Ribeirao Preto, and Salvador.

The household survey therefore provides a promising tool for examin-

ing transport patterns in many cities. In most, but not all cases, trip

data collected is differentiated by sex, but apparently never tabulated

separately for men and women. These surveys therefore seemed a useful

focus of experimentation in analyzing sex-differentiated patterns of

urban transport behavior. However, access to the data proved to be more

problematic than expected. Most data sets are owned by government trans-

port agencies which will release special tabulations after requests

are made through official bureaucratic channels. Without a large

budget to finance the solicited tables, such requests are generally not

feasible. A request for 10 tables comparing men and women was finally

approved in one city after two earlier attempts and the collaboration

by the author with transport authorities at local and national levels.

Despite the approval of the request and the intervention of a responsible

transport agent on the author's behalf, after six months the tables still

Shad not materialized.

Attempts through another channel were somewhat more successful.

- 12 -

Contacts with the Urban Projects division of the World Bank revealed

4 that the Bar.l had in its possession four household transport surveys

from Latin American cities: Santiago, Chile; Salvador, Brasil; Brasilia,

Brasil; and Bogota, Colombia. The data set from Bogota was in active

use at the time of the inquiries, making it relatively inexpensive to

generate a few additional tables comparing women and men. These were

graciously provided free of charge to the author, although the hoped-

for additional tables never materialized due to changes in programmers

and priorities within the furnishing agency. As a result, the utility

of transport survey data was put to only a partial test. Because of

the unique nature of the attempt, it was nevertheless judged worthwhile

to report its results. The remainder of this section will report findings

from the Bogota study, in order to pinpoint ambiguities and difficulties

in the data itself as well as to suggest a few hypotheses about dif-

Sferential transport behavior among disaggregated groups of travellers.

1. Trip purposes

In order to simplify tabulations and focus analysis on the most

significant kinds of urban travel from the standpoint of poor popu-

lations, categories of trip purpose were collapsed into four: work,

shopping, school, and other. The "other" category included trips

for personal business, social, recreation, passenger service, to

change mode, and home. As can be seen from Table 4.1, this collapsing

of categories obscured the specific purpose of more than half of all

trips for travellers in all sex and occupation categories. The "other"

category is consistently larger than any other, including the trip to

work. Unfortunately, it was not possible to retabulate this variable

using disaggregated categories. However, the data do seem to indicate

that women, especially housewives, are more likely than men to travel

for shopping purposes in Bogota.

- 13 -

2. Mode of Transport

Some preliminary findings with regard to differential use of dif-

ferent modes of transport did emerge from the tabulations.

a. Low-income women virtually do not use private transport,

whereas their male counterparts do so to a limited extent (Table 4.2).

b. Low-income women are, on the other hand, more likely to

use non-motorized forms of transport (i.e., walking) than low-incor.C

men (Table 4.2).

c. The greater use of private transport by males appears to be

in their trip to work (Table 4.3). Women's journeys to work, on t(-

other hand, are more likely to be on public transportation. In all

other types of trips, they are more likely than men to use non-motorized

modes. These findings could be clarified by tabulations using more

disaggregated categories of trip purposes.

3. Frequency and Duration of Trips.

Figures 4.1 and 4.2 show that, except in the very highest income

groups, women travel less frequently than do men. When they do travel,

they spend an average of 30 minutes less than do men in each income group,

with the gap narrowing in higher income groups (Figure 4.3). The average

probability of their having to transfer buses is also slightly lower than

men's (Figure 4.4). However, these patterns are somewhat less clear when

the analysis is restricted to the lowest five income groups in their

trips to work (Figures 4.5 and 4.6). While men's travel time and num-

ber of transfers show a rapid decline with rising income in these cate-

gories, women travelling to work show more erratic patterns with respect

to these variables.

4. Disaggregation of Travellers.

Table 4.4 shows the distribution of all travellers by income, work

Table 4.1

Trip Purposes by Occupation and Sex


Prof., tech., manage.

White collar















females 14.5
males 28.5




























100.0 422
100.0 1634

100.1 867
100.1 1219

100.1 273
100.0 767

100.1 532
100.0 2091

100.1 2688
100.0 2964

100.0 26
100.0 89

100.1 1103

100.0 5911
100.0 8764

* excluding "no reply"

Table 4.2

Mode of Transport by Sex and Income Groups


Females Males
Non- Non-
Income Groups Private Public Motor Total Private Public Motor Total

Less than 500 1.2 73.8 25.0 100.0 82.8 17.2 100.0

501 1000 69.9 30.1 100.0 2.9 78.5 18.7 100.0

1001 1500 0.6 65.3 34.1 100.0 3.8 73.7 22.5 100.0

1501 2000 0.3 73.0 26.7 100.0 4.4 75.9 19.7 100.0

2001 3000 0.8 67.3 31.9 100.0 5.1 76.5 18.5 100.0

3001 5000 5.0 77.9 17.1 100.0 11.7 72.2 16.1 100.0

5001 15000 23.7 65.6 10.7 100.0 32.8 57.9 9.3 100.0

15000 30000 45.3 46.7 8.0 100.0 .50.6 41.1 8.4 100.0

30000 + 60.0 32.0 8.0 100.0 55.8 41.9 2.3 100.0

TOTAL 10.8 68.1 21.1 100.0 14.8 69.2 16.0 100.0

(1253) (5871) (1359) (8483)

(616) (3897) (1214) (5727)

Table 4.3

Trip Purposes by Mode

Mode Work Shop School Other Total

Private 10.7 19.8 4.2 12.6 10.8

Public 84.4 66.8 57.4 68.1 68.0
Non-Motor 5.0 13.4 38.4 19.3 21.2

TOTAL 100.1 100.0 100.0 100.0 100.0
N (864) (298) (1330) (3526) (6018)

Private 21.1 10.2 4.0 15.1 15.0

Public 70.4 79.7 62.6 69.7 68.9

Non-Motor 8.4 10.2 33.4 15.2 16.1

TOTAL 99.9 100.1 100.0 100.0 100.0

(2499) (118) (1400)

(4928) (8945)

Mean days per week utilize transport

males (8484)
females (5727)

. -...- -----(ales


Income Groups

Figure 4.1

males (8484)
females (5727)

Mean Trips per Day


Figure 4.2

S; females *
i i males



Income Groups

9 Total



University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs