A.I.D. PMoganm Evaluation DiscuLssion Paper No. 8
Assessing the Impact of Development
LuL V111 L J 77Z\--1C--JD
ELLZI 'G' iU ZD!L] \X 1L
Bureau for Program and Policy Coordination
Agency for International Development 0
A.I.D. EVALUATION PUBLICATIONS
PROGRAM EVALUATION DISCUSSION PAPERS
No. I: Reaching the Rural Poor: Indigenous Health Practitioners Are There Already (March 1979)
No. 2: New Directions Rural Roads (March 1979)
No. 3: Rural Electrification: Linkages and Justifications (April 1979)
No. 4: Policy Directions for Rural Water Supply in Developing Countries (April 1979)
No. 5: Study of Family Planning Program Effectiveness (April 1979)
No. 6: The Sociology of Pastoralism and African Livestock Development (May 1979)
No. 7: Socio-Economic and Environmental Impacts of Low-Volume Rural Roadsa-A Review of the Literature (February 1980)
No. 8: Assessing the Impact of Development Projects on Women (May 1980) EVALUATION REPORTS
No. 1: Family Planning Program Effectiveness: Report of a Workshop (December 1979)
No. 2: A.I.D.'s Role in Indonesian Family Planning: A Case Study
With General Lessons for Foreign Assistance (December 1979) No. 3: Third Evaluation of the Thailand National Family Planning
Program (February 1980)
No. 4: The Workshop on Pastoralism and African Livestock Development (May 1980)
PROJECT IMPACT EVALUATIONS
No. 1: Colombia: Small Farmer Market Access (December 1979) No. 2: Kitale Maize: The Limits of Success (December 1979)
No. 3: The Potable Water Project in Rural Thailand (Forthcoming)
No. 4: Philippine Small Scale Irrigation (Forthcoming)
No. 5: Kenya Rural Water Supply: Program, Progress, Prospects
No. 1: Afghanistan Basic Village Health (Forthcoming)
PROGRAM DESIGN AND EVALUATION METHODS
Manager's Guide to Data Collection (November 1979)
ASSESSING THE IMPACT OF DEVELOPMENT PROJECTS ON WOMEN
Ruth B. Dixon
A.I.D. Program Evaluation Discussion Paper No. 8
Office of Women in Development .and
Office of Evaluation Bureau for Program and Policy Coordination
U.S. Agency for International Development
The views and interpretations in this publication are those of the author and should not be attributed to the Agency for International Development.
TABLE OF CONTENTS
About the Author
II. CURRENT EVALUATIONS OF WOMEN IN DEVELOPMENT PROJECTS:
A REVIEW AND ANALYSIS 4
Participation of Women in Project Decision Making 19
Access of Women to Project Benefits 28
Effects on the Status of Women 41
III. ASSESSING THE IMPACT OF DEVELOPMENT PROJECTS ON WOMEN:
AN ANALYTICAL FRAMEWORK 52
Indicators of Physical Wellbeing 57
Indicators of Economic Wellbeing 64
Indicators of Social Wellbeing 71
IV. METHODOLOGICAL ISSUES IN COMPILING COMPARATIVE DATA 81
Sources of Data on the Participation of Women in Project
Decision Making, Access to Project Benefits, and
Immediate or Long-term Consequences 81
A Comparative Approach for Evaluating the Effectiveness
of Different Project Strategies 90
List of Tables
Table 1. Women in Development Projects, Classified by
Purpose and Sector 7
Table 2. Women in Development Projects, with Intended
Beneficiaries, Major Inputs, and Participation
of Women 11
Table 3. Framework for Evaluating the Distribution of
Project Benefits 55
Table 4. Actual and Alternative Designs for an
Integrated Rural Development Project 80
Table 5. Major Project Characteristics Affecting
Inducement and Ability to Participate 93
Table 6. Basic Sets of Factors in Task Environment
Affecting Participation in Rural Development
This is one of a series of papers presented to encourage discussion on
the evaluation of important development issues. The A.I.D. Program Evaluation Discussion series reviews findings from evaluation and proposes issues for evaluation of current and future development activities. The series is coordinated by A.I.D.'s Office of Evaluation.
We are particularly pleased to present this paper to a worldwide development audience. Since 1974 A.I.D. has attempted to enhance the role of women in development (WID). With the recent emphasis within A.I.D. on knowing the impacts of development assistance, particularly the benefit incidence of projects and programs, we believe it important to look at the impacts of assisted activities on poor women, the sub-group of the poor which has been the Agency's most clearly selected target for help. Directing the fruits of development to the poor is difficult and we have no illusions about the difficulty of directing positive impacts to a sub-group of the poor. A.I.D. has been aware that the U. S. legislative policy regarding women in development is intended to apply to as much of the program as possible.
Last year we commissioned a data collection effort on evaluations involving women'ss projects" (Elliott and Sorsby, 1979).* The authors of that study charged that the evaluations they identified appeared scanty and could not clearly tell impacts. We pursued this by commissioning the present paper to analyze, in depth, evaluations on women in development projects and to recommend lessons learned and ways of better evaluating such projects.
This paper verifies the charge. It puts forth a considerable amount of
data on what is now known about bettering the effect of all development projects on women. We are publishing this paper as a direct challenge to colleagues to
*Available in limited supply from the Office of Evaluation, A.I.D.
take the impact of their projects on women more seriously. As an institution, A.I.D. should be at the stage of learning fully from its "WID" projects and should be well into the stage of making far larger investments to better the condition of women.
Thankfully, Dr. Dixon has compiled an impressive array of information on
what is known about impacts of development activities on women. Her recommendations should give confidence to developers (planners, designers and evaluators) to put more significant funding into activities intended directly or indirectly to better the lives of women.
Comments on this report and contributions to the evaluation findings regarding women in development will be welcome by A.I.D. and the author.
Arvo aserRobert J. Berg
Coordinator Associate Assistant Administrator
Office of Women in Development Office of Evaluation
Bureau for Program and Policy Coordination
About the author
Ruth B. Dixon is Associate Professor of Sociology at the University of California at Davis. A social demographer, her research interests center on the social aspects of reproductive behavior, the relation between economic development and demographic change, and the role of women in development. She is the author of Rural Women at Work: Strategies for Development in South Asia (1978) which draws on field investigations of women's economic activities in Bangladesh, India, Pakistan, and Nepal. Professor Dixon has been a consultant to the United Nations Branch for the Promotion of Women, to the Office of Women in Development at AID, and to the World Food Programme, and is currently on the Board of Directors of the Population Association of America.
This report was commissioned by the Office of Evaluation of the Bureau for Program and Policy Coordination of USAID, in consultation with the Office of Women in Development, to accomplish four objectives:
1. Examine current evaluations of Women in Development
projects in the Third World for the adequacies and
inadequacies of their measures of social impact, drawing on evaluations of projects funded by AID and by other
donors classified in the report by Veronica Elliott and
Victoria Sorsby, "An Investigation into Evaluations of
Projects Designed to Benefit Women" (1979);
2. Identify, from these same documents, institutional and
substantive lessons learned regarding what works and what does not in specific socioeconomic environments;
3. Propose a more comprehensive framework for measuring
the impact of development projects on women; and
4. Suggest a strategy for collecting additional information
from the field, and for coordinating evaluations with
other agencies for comparative purposes.
The report that follows differs in several ways from the original intent. First, it quickly became apparent that the question of "what works and what does not" in specific socioeconomic environments can be answered only tentatively and partially from existing documents. Many Women in Development projects are too new to have completed a formal evaluation process, even though questions of potential impact on women are addressed in a preliminary way in Project Papers or other documents. Among those with completed evaluations, both the description of socioeconomic environment and the measures of social impact are usually inadequate for these purposes. Identification of institutional and substantive lessons learned beyond those few included here, must await a more systematic series of project evaluations based on a comparative methodological approach.
Second, the framework proposed here for measuring the impact of development projects on women represents a general type of social assessment that looks at the differential effects of projects on a variety of social groups: females relative to males, landless peasants compared with landowners, ethnic minorities compared with the numerically or socially dominant ethnic group, and so on. Although the examples in this paper refer to women, the approach is essentially the same regardless of social category of interest.
Third, although the original objective was to look at evaluations of Women in Development projects, I am proposing instead an approach to evaluating the impact on women of an development project -- whether or not women are specifically cited as beneficiaries -- ranging from small-scale local training projects to large-scale integrated land settlement schemes. The identification in agency files of Women in Development projects is undoubtedly helpful in assessing the extent to which projects explicitly recognize women as intended clients and in documenting the level of funding of such projects, although the problems of classification are considerable.-1 But whether a project is classified as "women-specific," "women's component," or general (beneficiaries identified without regard to gender, such as "small farmers,"~ "out-of-school youth," "urban slum dwellers"), the methodological issues are similar. We need a technique that will help us to answer the question: under what conditions are women least likely to be disadvantaged by development projects, both in absolute terms
1L/ For discussions of this point, see Elliott and Sorsby (1979); Staudt (1979); U.S. Agency for International Development (1978).
and relative to men?-/ Or, to put it more positively, how can we be sure that women as well as men, girls as well as boys, take advantage of new opportunities? How can we minimize the differential impact on males and females of project benefits and costs, and, where sexual inequalities prevail, ensure that females are able to catch up to males in their access to material and social resources?
- The negative wording of the question is intended as a reminder of the uneven and sometimes disastrous consequences to women of a variety of development projects and policies in the past. For an overall perspective, see (among others) Boserup (1970); Chaney and Schmink (1976); Tinker (1976); Van Allen (1974). Annotated bibliographies on Women in development include Buvinic (1976), Nofiformal Education Information Center (1978), Rihani (1978).
II. CURRENT EVALUATIONS OF WOMEN IN DEVELOPMENT PROJECTS: A REVIEW AND ANALYSIS
The Elliott and Sorsby report includes profiles of 43 projects
funded by AID or by private voluntary organizations (PVOs) in which
women are specified as beneficiaries (in whole or in part) and, for AID
projects, that have been evaluated. A number of these do not comply
with the more restrictive definition proposed by AID's Office of Women
in Development, however, which identifies a WID project as one designed
for women only to help them "catch up" with men, or one with a strong
women's component as an integral part of a general project. AID's
definition, which emphasizes women's economic roles, excludes activities
such as maternal and child health or family planning that offer goods
and services in the absence of training or other assistance aimed at
increasing productivity or earnings. According to the Special Concerns
Code for classifying AID projects, the WID category should:
CODE CONCERN DEFINITION
WID WOMEN IN Include activities which will help integrate
DEVELOPMENT women into the economy of their respective
countries, thereby both improving their status as well as assisting the total development effort. (See Section 113 of the Foreign Assistance Act.) Programs and projects which
are in whole or part specifically designed to afford women the
opportunity to participate in the development process in a
significantway are to be included in this category. Not all
projects which include women as beneficiaries are to be included.
For instance, population projects in which women are merely
recipients of goods, such as contraceptives, or health projects where mothers receive food and services for their children, are to be excluded. However, where, in addition to the provision of
goods and services, women receive training or other assistance
designed to-increase their earning capacity or enhance their
economic productivity, include the relevant portion of the funding
for the women's component in this category. Where a specific
women's component is designed into an integrated project, include the proportion of that component as a women in development effort.
Of the 43 projects included in the Elliott and Sorsby report, 32 are reviewed here. Of the 11 that were excluded, 6 are AID agency-wide grants to U.S. institutions for training or research in which the analysis of social impact on intended beneficiaries is not possible;one is an urban water project with no clear women's component;/ one is a day-care center for nutritionally deprived children;- one is a grant for the evaluation of an educational TV system;-/ and two are clear women's projects for which documentation could not be obtained.7/ Evaluations were obtained from AID files and from the July 1978 and January 1979 Directory of Projects Involving Women compiled by the Secretariat for Women in Development of the New TransCentury Foundation in Washington,
-/ Grants to Stanford University for the design of radio prototypes for teaching elementary math in LDCs; to Stanford University for curriculum development in low-cost communication; to Florida State University for the development of educational technology useful for LDCs; to the University of Massachusetts for training in non-formal education; to Oklahoma University for research in low-cost methods of water and waste treatment in LDCs; and to the American Home Economics Association for training third world home economists in family planning.
- Taiz Water Rehabilitation in Yemen Arab Republic (AID/NE).
- Day care center for hardship children in Chile (AID/LAC), 6-/ Evaluation of education TV projects in El Salvador (AID/LAC). 7/ Poultry development in Yemen Arab Republic (AID/NE); Caritas de Honduras housewives clubs (Inter-American Foundation). 8/ I am indebted to Debra DeWitt for collecting and summarizing these documents. Additional compilations of Women in Development projects can be found in Mickelwait, Riegelman, and Sweet (1976) and U.S. Agency for International Development (1978), among others.
The remaining 32 Women in Development projects are listed in Table 1 according to their purpose and sector. Purposes can be classified according to a number of criteria. A recipient agency
-- a Ministry of Education, a marketing cooperative, a women's association -- receives aid on the "promise" that it will provide goods or services (directly or indirectly) to a targeted class of beneficiaries within given parameters such as budget and time constraints and through designated procedures. These goods or services are intended to achieve specific project purposes, e.g., "to increase annual net cash farm incomes over 50% in a five-year period, by adopting modern sericulture practices." One useful approach is to distinguish between projects designed primarily to increase the productivity of beneficiaries (classes of individuals, households, or localities) and those designed to improve their welfare.2-'
Projects defined as increasing productivity try to raise production, employment, or incomes among the target group, either directly through skills training (formal or nonformal), group mobilization, technical assistance, credit, or job creation, or indirectly through the expansion of physical infrastructures such as electricity, irrigation, and roads, or through other means. They may focus on productivity in food and agricultural production and distribution, on nonagricultural production and services, or on some mix of these. Included in the agricultural
9/ "tWelfare," for lack of a better term, refers here to the fulfillment of basic human needs. The distinction between productivity and welfare is somewhat artificial, since material prosperity (deriving from productivity) is a basic element of human welfare, while health, education, and organizational capacity can all increase productivity.
WOMEN IN DEVELOPMENT PROJECTS, CLASSIFIED BY PURPOSE AND SECTOR
Purpose Sector Projects a/
Increase Food and 1. Palm oil purchasing/food marketing in Cameroon
production/ agricultural (4)
employment/ production & 2. Sericulture in northeast Thailand (17) incomes distribution 3. Pig raising in the Republic of Korea (47)
4. Land conservation in Lesotho (9) 5. Rural water systems in Kenya (8)
Non- 6. Commercial sewing & baking in Costa Rica (22)
agricultural 7. Vocational skills training in El Salvador (24)
production & 8. Sewing center in the Philippines (49)
Mixed 9. Mohair, tie & dye, poultry in Lesotho (10)
10. Food production & processing in Upper Volta
11. Food processing, toy making, silk screen & crafts in Kenya (44)
12. Income-generating schemes for rural women in Bolivia (20)
13. Sewing, candy production, floriculture in Paraguay (29)
14. Rural market women's co-ops in Nicaragua (28) 15. Handicrafts, animal raising, food crops in Dominican Republic (52)
16. Women's co-ops in Bangladesh (45) 17. Handicrafts and poultry raising in Fiji (46) 18. Income-generating projects in the Philippines
Improve Education 19. Secondary school for girls in Uganda (12)
welfare (formal) 20. Primary schools and teachers' hostels in
(nonformal) 21. Literacy & community self help in Ethiopia (5) 22. Radio programs for highland Indians in Guatemala (26)
23. Community development project in Sri Lanka (30)
Health 24. Audio cassettes on health and nutrition in
TABLE 1 (continued)
Purpose Sector Project a/
Improve Community 25. Research on women in Ghanaian development
welfare organization (6)
(continued) 26. Leadership training for PVOs in Ghana (7)
27. Seminars on women in Upper Volta (13) 28. National Women's Development Academiy projects in Bangladesh (16) 29. Citizenship training for girls in Thailand
30. Leadership training for volunteers in Latin Amierica (9)
31. Leadership training for volunteers in Costa Rica (23)
Promote All 32. Pespire Valley Integrated Developmient in
integrated sectors TFonduras (27)
a/ Numbers in parentheses refer to pages in Section MI of Elliott and
sector in Table I are a palm oil purchasing and food marketing cooperative for women in Cameroon, a sericulture project for farm families in northeast Thailand, a pig raising project in the Republic of Korea, a land conservation scheme in Lesotho, and a rural water systems project in Kenya. The nonagricultural sector includes skills training and production centers in commercial sewing and other activities in Costa Rica, El Salvador, and the Philippines. Mixed income-generating projects -- the most frequently occurring category -- include a variety of small-scale agricultural and nonagricultural production schemes for low-income rural and urban women such as handicrafts, poultry raising, and food production and processing.
Projects defined as improving welfare include general education in the formal or nonformal sectors (school construction or staffing in Uganda and Afghanistan, functional literacy and self-help activities in Ethiopia, Guatemala, and Sri Lanka); health programs (health and education in Tanzania); and support for national or local organizations for leadership training, research, program planning, community outreach, and other activities intended to promote citizen participation. This latter category, the second most frequent on the list, includes seven projects in Africa, Asia, and Latin America.
Some projects overlap the boundaries because of the diversity of their purposes or the phased sequence of activities. Several projects in the community organization category, for example, have as their eventual goal the expansion of income-generating opportunities for women although they are still in the formative stages of leadership training or research. An integrated rural development scheme (listed separately)
will, by definition, try to reach all sectors. But the framework provides a useful starting point for viewing the Women in Development projects under discussion here.
Table 2 indicates for the 32 projects the intended beneficiaries as stated in Project Papers and other documents, the major inputs or means by which project purposes were to be achieved, and a summary of the participation of women in project decision making, their access to project benefits, and the major social and economic impact of the project on women.
The column labelled intended beneficiaries describes the characteristics of the intended project clientele, or target group. Twenty-two of the 32 projects are women-specific, that is, are intended to serve an exclusively, female clientele (e.g. "all women of Bui division who want to join Nso women's co-op" in Cameroonj "50 women from home improvement clubs in 11 low-income villages" in the.Republic of Korea). The remaining 10 include an assumed or explicit women's compnent (e.g. "1500 farm families in settlement areas of 8 provinces" in Thailand; "about 300,000 remote rural dwellers in Kenya, especially women."). Twenty of the projects are based in rural areas, four urban, and eight mixed. The focus on low-income people is clear, although the immediate beneficiaries of several leadership training projects are primarily middle or uppermiddle-class women rather than the poor whom they are ultimately expected to serve.
The size the the target group (where it can be determined) varies enormously, ranging from the 50 women of home improvement clubs in a PVO-sponsored pig-raisingI~rojedt in Korea to the 300,000 rural dwellers
TABLE 2. W(MEN IN DEVELOPMENT PROJECTS, WITT INTENDED BFNeFICIARIS, MAJOR INrPUTS, AND PARTICIPATION OF WCr4FN
Participation of Access of Effects of Source of
.women in project women to project on information
Major a decision making project benefits status of women
inputs d 0 r .
11 *i N d r .+ +0 C
a/ Intended tb 4 -0 o e J
Pro ject beneficiaries H o _I.Palm oil purchasing/ All women of Bui Nigh local part. All members of 62 Improved nutrition Report by
food marketing in Division who want X X X in initial village co-ops at lower cost of Development
Cameroon (4) to join Nso stages are women palm oil Alternatives
Women's Co-op Inc. 1979
(AID grant to Nso Some transfer of Some poor women Vigher incomes (Donald R.Jackson)
Co-op Union) d-m to male mgt. cannot afford from food sales
of Nso Union to buy shares
Some loss of social
contact at market
.Sericulture in 1500 farm families Active in project 85% of farm Increases in net New TransCentury
no ~theast Thailand in settlement X X design at local participants family income January 1979
l') areas of 8 level; women will be female; average $150/year
(AID) provinces train other 50% of 210 trainees
women in 1978 were female
5Pig raising in 50 women from ( )( X Village women All participants Increases in income, New TransCentury
Rep. of Korea (4T) home improvement decided on are women pigs bougitfor $35 January 1979
clubs in 11 activity, most sold for $198 aver.
(Save the Children low-income donor staff male Poorest women
Federation, Inc.) villages participate less Initial lack of
confidence turns to
enthusiasm and skill
+Land conservation 12,000 subsistence X K X Women "will be Women "play a Increases in income Project paper
for small farmers farmers of Thaba deeply involved predominant role not expected until
Sin Lesotho C) Bosiu area ...in the imple- in the field of 1982/83 Independent
mentation of agriculture" )a evaluation
(AID/IBRD) this project Oct. 1975
.Water systems for about 300,000 )( X )x ( Village women Project behind Reduction in women's New TransCentury
rural communities remote rural participate in schedules wmmen drudgery of water June 1978
in Kenya C(8) dwellers, esp. org'n and mgt. will benefit collection S
I women of self-help Evaluation 1976
_(AID/CARE) _community efforts
ADetails on women's roles are reported in Annex 2 of IBRD Appraisal Report S L pa e / afJEvaluation of impact on women's daily activities can be found in a CARE
NumbersTreferto pages in Section III of Elliott and Sorsby tp. cit.) and IBRD report, October 1977. Neither of these was obtained for
TABLE 2 (continued)
inputs 0 4) r
S o 3 ..- I 0 Participation of Access of Effects of
Intended 0 W X women in project women to project on Source of
Project beneficiaries -H 0 E 0 0 H decision making project benefits status of women information
6. Commercial sewing About 1,500 low- )( Trainees play All 181 trainees Increased incomes in Project
and baking, crafts, income women from major role in are women industrial sewing Evaluation
cosmetology in Costa barrios of San Jose identifying group Statement
Rica (22) problems, pro- Req't of 6th grade Nov. 1978
posing solutions educ. dropped to Improved housing
(AID) recruit poorer woeA
exteem of husbands
L Group efforts
7. Skills training Approx. 1,000 poor 2
in El Salvador (24) urban residents in Not stated 2,400 trainees Major empl't gains Project
automechanics, weld- informal economic include 1,700 grads in auto mechanics & Evaluation
ing, sewing, tailor- sector; about 45% female not stated welding, least in Statement
ing, radiorTV repair, are women cooking & cosmetology April 1978;
cooking, cosmetology Courses are offered Final report
AID) ifor women May 1978
8. San Marcos Sewing Poor rural women in Center in Philippines Province of Balucan Local women All 20 workers at Increased incomes New TransCentury
(49 1 oa oe l 0wresa nrae noe New TransCentury
for sewing center women; 45 more to children's clothing
( sto build on local be added.
S(P1) skills and need Sense of pride &
for income accomplishment
; Innovation of work; ing outside home
'fohair, kit ring ooii-rmiin T i omen i m
tie & dye cloth, poul- ages 10-60 in 3 vil- Village women Presumably all Generated income New TransCentury
try raising in lagesi pred. wives initiate activi- women I January 1979
Lesotho (10) of men working in ties, request tech- Eagerness of village
S. Africa mines nical assis. from women to participate i
(US Embassy) 50-100 women
wIolunteers Improved nutrIon of
children in nurseries
T. Food production Ruiral women in -60
& processing in Upper villages with female 'X (N X Village women All women Too soon to tell Carolyn Barnes,
Volta (14) extension agents encouraged to de- project is behind "Strengthening
cide indiv. or group schedule Voltaic Women's
(AID/DC) activities (veg. Roles in Dev't."
gardens, peanut oil January 19781
....___.___ _presses, grain mills) New TransCentury
TABLF 2 (continued)
In td ed4 Participation of Access of Effects of Source of
It. o 4 0 women in project women to information
I o- -0 11 .r 0 X project on
project beneficiaries k O 1. 0 decision making project benefits
.Poet eeiiais- 0 13 E- -. cd U C4sau o1oe
11. Food processing, Approx. 220 women v Self-help projects All beneficiaries IIncome generated New TransCentury
toy making, silk in 4 villages decided by rural are women by four groups January 1979
screen, crafts in women's groups
Kenya (44) Growing self-confidance; opportunity
(FVOs) to work together
12. Agricultural Poor Aymara & Quecha Promoters are All beneficiaries Increased incomes New TransCentury
production, animal spiking peasant women peasant women: are women but limited market June 1978
raising, sewing, 18 years or older staff all women
crafts, co-op stores Women interested in
in Bolivia (20) organizing co-ops
13. Sewing, candy ist year, 300 women Not stated, al- 192 prof'Is trained; Not stated by sex Project paper:
production, flori- &160 men from Asun- though most staff! 62 volunteers, 520 Project
culture in Paraguay cion & poor rural are female beneficiaries Participants income Evaluation
(29) communities: 2nd % female not stated increased in train- Statement
year double with ing; projects not Nov. 1978
(AID) 70% rural yet self-supporting
14. Rural market rural market women: v Women occupy 4-6- 55% women in 12 Not stated Project paper:
women's co-ops in agric'l producers, of 6 top position! credit unions with Project Annual
Nicaragua (28) poultry, food proces- in each co-op: 1,400 members Review Dec. 1975
sera, artisans (many evaluation &
(AID) are women) advisory team
4dit & social ser- Rural women In the XVillage women All beneficiaries Improved access New TransCentury
vices in Dominican Sur & Region Central: A initiate projects are women to credit January 1979
Republic (52) 70 groups after 5 almost all staff
handicrafts, goat years are female Awareness of
raising, food crops social services
IP')Village women All beneficiaries Increased incomes New TransCentury
16. Women's co-ops Rural women in 4 of ) I / ) initiate projects: are women
in Bangladesh (45) thanks, esp. wives of t es within purdah January 1979
rice, seeds, oil pro- small landowners
ceasing, animals & Group efforts
poultry, silkworms _(CARE)
TABLF 2 (continued)
Major o, )
inputs Cd r e .
CO4 W d 'H
17H -H 14 4cr4tS&o Participation of Access of Effects of
Project Intended 0 4 11 0) 0 "woeinpoct oentprject on Source of
beneficiary 0 CD W 4J 0 deiinmkn project benefits status of women inomtn
-1 andicrafts & 85 girls in Metho- Girls perform all !All beneficiaries Not stated New TransCentury
poultry raising in dist Wandicraft & functions of are female January 1979
Fiji (46) Farming School school farm, which
is self-sustain(PVOs) ing. All female
18. Food crops & poor rural women in X > )( ) Trained leaders Village benefi- Increased incomes New TransCentury
floriculture, co-op 32 villages in from villages ciaries are women; iJanuary 1979
stores in Philip- Cavite Province consult local also 76 leadership Leadership skills
pines (48) women about trainees
project plans Attitude change
(Pvus) from indifference
19. Comprehensive girls of highschool ) I )ot stated 540 students are Increase in number
seodrNots g nras nnme Project paper
secondary girls' age all female; % of of female graduates revision 1969;
school in Uganda I teachers female from secondary Project Annual
(12) (AID) not known school Review 1971-72
20. Construction of Rural children and Not stated
17 rrl leetay teachers; 15% of o ttdQuota of 15% Not stated Project Paper 1975;
170 rural elementary girls met in Evaluation cable
schools & 40 teachers' places in newly con- Parwan but not in May 1976
hostels in Afghanistan structed schools re- southern zones; %
(31) (AID) served for girls teachers female
21. Literacy & com- Adults aged 20-40 Women involved in 80% of 148 partic. Not differentiated Internal -evaluation
munity self-help in in selected urban X self-help projects! in 6 communities by sex February 1977
Ethiopia (5) & rural areas; staff mostly male responding to
majority female (4 of 24 group questionnaires Changing attitudes New Transcentury
(AID) leaders female) were female toward women's roles June 1978
practices & diet
22. Radio programs 16,000 illiterate food production
for highland Indians highland Indians j Role of women % female not stated Not differentiated John R. Davidson,
in Guatemala (26) engaged in subsis- not mentioned among 246 promot- by sex "The Basic Village
tence agriculture ers, 9,000 recip- Education Project
(AID) ients of info. Some change In agri- in Guatemala," AID
cultural knowledge, Sept. 1976;,
attitudes, practice Project evaluation
summary Jan. 1979.
TABLE 2 (continued)
4. 0H 4.
inputs 0~ 4J ) ~ .
Wo -l ;9 4. W Participation of Access of Effects of
n. k X W women in project women to project on ource of
Project beneficiaries decision making project benefits status of women nformation
23. Community devel- Girls from slum X >( Most staff All beneficiaries Improved health of New TransCentury
opment in Sri Lanka areas of Colombo. are female are female children in day January 1979
(50) esp. school dropouts; care center
mothers and preschool Some needlework
(PVOs) children skills with possible incomegenerating potential
- R wm Village women All beneficiaries Initiated self-help Joyce Stanley with
24. Audio cassettes Rural women in 2 2 selected group are female activities in health, lisa Lundeen, "Audio
on health nutrition villages leaders for project nutrition, gardening, Cassette Listening
in Tanania (i) crafts Forums" (AID, n.d.)
(AID) Self-awareness &
25._ Research on Nat'lCou-c-l on Research conducted All beneficiaries Not evaluated Project Paper
25. Research on Nat'l Council on byatlouclnwilewmn
women in Ghanaian Women & Development by Nat'l Council on will be women
Development (6) ultimately, urban & women's orgs. get
rural poor women
(AID) grants for pilot
womenspta-Kym-e-f-m-sWorkshops planned 11l beneficiaries Some participants New TransCentury
ing for PVOs in organizations in a 9 by Ghana Associat- will be women organized local January 1979
Ghana (7) regions, ultimately, ion of Women workshops
urban & rural poor Some cottage indus(AID) women _tries and handicraft
27.... Seminars of V Women organized 2 11l beneficiaries Not evaluated New TransCentury
27. Seminars on women Fed'n of Voltaic Women X( seminars on role will be women June 1978
in Upper Volta (13) and other prof'l women of women
in Ouagadougou ult(AID) imately, all Voltaic
Project planned by 11l beneficiaries 154 women trained by New TransCentury 28. National Women's 3600 village wen X Nat'l Women's Org.: ill be women NWDA earn salaries Jan. 1979in Bangladesh (16) 16,000 rural women i managed & staffed as motivators Project Evaluation
beneficiaries by women 180 women placed in Statement Jan. 1979;
(AID) jobs thru exchange Sallie Craig Ruber,
I About 10,000 rural "Personal impressions.."
:women received some AID/Dacca Jan. 1979
TABLE 2 (continued) .
Major 4 a1 0
inputs 0 94)~
-H is 0 Participation of Access of women Effects of
Intended women in project to project project on Source of
Project beneficiaries A. V 0 1E decision making benefits status of women information
Planned by Girl Almost all girls; Not evaluated New TransCentury
29. Citizenship Girls aged 15-24 Guides Assoc. & some boys added by June 1978
training for girls in from low-income staffed by women local request
Thailand (51) (PVOs) rural & urban families
STraining organ All beneficiaries More benefits to End of Project
30. Leadership train- Middle & upper-class ized & managed by will be women middle class volun- Evaluation d/
ing for volunteers women; ultimately, League of Women ees than to poor
in Latin America (19) low-income people Voters women
31. Leadership train- Middle & upper-class Training organ- All beneficiaries Increased Evaluation memo
Ing for volunteers in citizens, mainly ized & managed by will be women consciousness & 1978 d/
Costa Rica (23) women$ ultimately, women's org. in political parti(AID) low-income people Costa Rica cipation
32. Pespire Valley Small scale, near X N ot clear: Participation in Increased incomes Project Evaluation
integrated delo subste fa ers women "will be" diverse activities for 19 women a- Summary d/
ment in ronduras (27) in 10 village clusters about 50% of com- not specified by ployed in mango
in Pespire Valley munity council sex, although puree co-op, tho
n Pespire Valley members for local women assumed to future uncertain
(AID) decision making participate
d/ These evaluations could not be obtained for this reviews information drawn from these reports is summarized in the Elliott and Sorsby project profiles
of the CARE/AID water project in Kenya. In general, however, the womenspecific projects are significantly smaller in scale than are the mixed projects that have the potential, at least, of reaching larger numbers of female beneficiaries.
The treatment of women's roles assummarized in Table 2 is analyzed under three headings: the extent of women's participation in project decision making, either as project staff or as members of the client group; the extent of women's direct access to project benefits; and the immediate and long-term effects of the project on women's social and economic status. 10/
Participation in decision making refers to women' s involvement in project design, implementation, and evaluation. It is useful to distinguish here between the role of women as staff members in donor agencies arid recipient institutions (service providers) that administer the project, and the role of women as intended beneficiaries. First, to what extent and in what capacity are women represented among staff members in donor and recipient agencies responsible for decisions regarding project planning and management? Second, does the decision-making process within donor and recipient agencies reflect an institutionalized concern for the interest of girls and women as beneficiaries? Third, to what extent and in what capacitydo women from the class of intended beneficiaries participate in decisions of what to do and how to do it? How are their needs and priorities determined and reflected in project design and
10 For other approaches to measuring women's participation (or
popularr participation") in development projects, see (among others) American Council of Voluntary Agencies for Foreign Service (1975); Cohen and Uphoff (1977), pp. 27-58; Dulansey (1977); Mickelwait, et al. (1976); Palmer (1979).
implementation? To what extent do women from the client group participate in project monitoring and evaluation?
Access to project benefits refers to the extent to which girls and women are represented among direct recipients of goods and services such as vocational training, agricultural credit, or membership in cooperative societies. It is important to distinguish here between direct access to project benefits, and indirect access in which it is assumed that females benefit as members of families from activities in which males participate. Numbers and proportions of females among beneficiaries can generally be obtained from project records pertaining to direct client contacts: farmers contacted by extension agents, children enrolled in school, workers on a construction project, recipients of bank loans, and so on. The identification of beneficiaries within this context does not prejudge the question of whether they actually benefited,. however, nor does it assume that those without direct access to project benefits did not gain (or lose) in some obvious or subtle ways.
The effects of the project on the status of women refer to the various ways in which women's position in the family and the community may be affected by the projects, either in absolute terms or relative to males. Some changes may be positive and others negative; some intended and others unplanned. Immediate effects should be distinguished from long-term impacts where possible, although most formal evaluations are undertaken too soon to pick up long-term impacts.
In the discussion that follows, the Women in Development project evaluations included in Table 2 are scrutinized for the adequacy of their treatment of these three dimensions of female participation and
for the "lessons learned" regarding the achievement of their goals.
Participation of Women in Project Decision Making
The role of women staff and beneficiaries in design, implementation, and evaluation is easier to deduce from the New TransCentury
project profiles than from AID documents. In compiling its directory of
Women in Development projects, New TransCentury asked donor agencies
the following questions, among others:i-/
13. We would like information about the staff people involved
with this project. First, please list the title -- such as Project Director, nutrition educator, secretary, etc. Then,
f or each position you have listed, record the total number of
people who hold that position, how many of these people are women, how many of these people are host country nationals,
and how many of these people are expatriates.
19. We are interested in learning how women were involved in the planning, initiation and design of this project. (Please
describe in detail and distinguish between female staff and
female beneficiary involvement.)
20. We are interested in learning how women were involved in
the management and control of this project; that is, in
determining the direction of the project and in making the decisions involved with implementing the project. (Again,
please distinguish between female staff and female beneficiary
As an example of the type of information elicited, the New TransCentury directory profile of an AID-funded project aimed at modernizing
sericulture practice and raising farm incomes among 1500 families in
settlement areas of northeast Thailand (#2 in Table 2) includes the
-Directory of Projects Involving Women (Washington, D.C.: New
TransCentury Foundation, Secretariat for Women in Development, July 1978 and January 1979).
Target group involvement: In rural areas, besides the role of
hard-working agricultural laborers, women carry out most of the principle tasks connected with sericulture. ... Females participating in this project are playing a key role, first
as recipients of the training received and subsequently as
trainers in assisting in upgrading the skills of women
members of their village group.
Female staff: Project design team included women officials from
the governmental Public Welfare Department and USAID woman
Assistant P'roject Officer with women mission/Women in Development
Coordinator active in project review. Efforts of female staff
include recommendations to bring women more fully into local
level project discussions, using women already in sericulture
training to persuade other women to take more action roles,
employing outstanding women to work in Settlement management
positions. Additionally, three out of ten on-site sericulture
supervisors in the project are women. Female staff go out to
the selected sites during the feasibility study.
Staff profile: All nationals except for AID Project Officer
(Male); 1 male Government of Thailand Assistant Project Director;
1 male Government of Thailand Project Manager; 2 Governmental
Project Coordinators (1 Female); 1 female AID Assistant Project
Officer; 10 Governmental Project Supervisors (3 Females); 70 Governmental Agricultural Extension Workers (1 Female) (New
TransCentury, Jan. 1979)
AID documents are less likely to include such detailed information,
at least in the summaries that are most accessible, because questions
regarding female participation are usually not asked.
The typical project model in Table 2, particularly the village
women's income-generating projects and nonformal education schemes,
tries to achieve a high degree of participation in decision making by
intended beneficiaries on a self-help basis. Women are encouraged -in consultation with staff members who act as catalysts -- to identify
their needs and priorities in group discussions, to decide jointly on
activities that will address their most pressing problems, and to
request appropriate training, credit, technical assistance, or other
goods or services from the administrative agency.- Beneficiaries are also frequently encouraged- to participate in project evaluations. Two projects providing physical infrastructures for agricultural production (land conservation for small farmers in Lesotho #4, water systems for rural communities in Kenya #5) involve female beneficiaries in project implementation but not in identification or design. Projects categorized as improving welfare through community organization (#25-31) elicit very high levels of female participation throughout the project cycle because they are planned and implemented by women's organizations. Among the remaining projects, the role of women in decision making cannot be ascertained.-3
What can we learn about women's participation in project decision makin g from a review of available documents? In the absence of a systematic approach to the question in the various reports, and in view of the small number of projects under review, the "lessons" must remain tentative:
1. Female participation in decision making at both staff and beneficiary levels is higher when projects are administered through women's sections of government ministries or PVOs or through national women's associations than through general_.PVOs or government agencies
The village-based income-generating activities for peasant women in Bolivia, for example (#12), are administered by the Rural Women's 12 Project numbers 1, 2, 3, 6, 8, 9, 10, 11, 12, 15, 16, 18, 21, 24 in Tables 1 and 2.
13 Skills training in El Salvador #7, income-generating activities in Paraguay #13, school construction in Uganda #19 and Afghanistan #20, radio programs for farmers in Guatemala #22, integrated rural development in Honduras #32.
Promotion Division of the Bolivian National Community Development Service. "The Women's Division works to assure that women are active participants in these projects at both the decision-making and action levels" (New TransCentury, June 1978). The community organization project in Bangladesh (#28), implemented by the National Women's Development Academy, was "completely planned by the National Women's Organization [Bangladesh Jatiy Mahila Sangstha] which is all women, and approved by the Bangladesh Government Women's Affairs Division (comprised of some women) and the Planning Commission (comprised mostly of men)" (New TransCentury, Jan. 1979). Virtually all staff members are female in those projects extending funding to women's organizations, many of which rely extensively on volunteers.
In contrast, women are less fully represented among the paid
staff of some PVOs or government agencies without a specific focus on women, even when women are the primary or sole beneficiaries. The Ethiopian nonformal community education project (#21), with approximately 80 percent of beneficiaries female, had a largely male staff which included only four women among 24 group leaders (New TransCentury, June 1978). The Thailand sericulture project (#2), with a majority of intended training recipients female, included only three women among 10 government sericulture supervisors and only one woman among 70 government agricultural extension officers working with the project (New Trans14/ e.g., Lesotho National Council of Women #9, Mujeres en Desarrollo Inc. in Dominican Republic #15, Sri Lanka Women's Conference #23, Ghana Assembly of Women #26, Federation of Voltaic Women #27, Bangladesh Jatiyo Mahila Sangstha #28, Girl Guides Association of Thailand #29. The Lesotho project, in particular, relied extensively on volunteers for training rural women in income-generating activities --.from 50 to 100 depending on their availablity.
Century, Jan. 1979). A woman-specific income-generating scheme in Kenya (#11) included only seven women among 20 consultants, even though specific attempts were made to utilize women wherever possible; the Korean women' s pig-raising scheme (#3) included one female staff member among six.
Although women's participation in project planning and implementation could in most cases be increased by channeling funds through woment s organizations or women's components of government ministries or PVOs, this strategy bears a cost. As noted earlier, women-specific projects tend, on the whole, to be smaller in scale than those aimed at classes of beneficiaries such as farm families, out-of-school youth, or other social categories not limited to females.
2. When women's programs are affiliated with larger male-dominated institutions, decision making on major policy issues tends to be
-transferred to men in the parent institution. The Women's Cooperative for palm oil purchasing and food marketing in the Cameroon (#1), for example, affiliated with the Nso Cooperative Union -- a secondary level cooperative with sole marketing rights for cocoa and coffee (men's cash crops) -- in order to obtain capital funds and technical assistance. According to one evaluation,
The Cooperative Union auditors perform yearly audits for the
Women's Cooperative, and have also given it financial support, in the form of three staff salaries. This relationship, plus the cultural issue of strong male domination, has resulted in what appears to be a transfer of a certain amount of decisionmaking authority to the union management.
A particular example of this was the decision to purchase a
pickup truck for the transport of member produce from the
villages to the cooperative warehouse. The women were against
acquiring the vehicle because they had no experience with one and
felt renting these services would be cheaper. The men from the
union, however, were in favor of the purchase and talked the
women into buying it. (Jackson, 1979, p. 7).
A similar decision-making pattern can occur when the women's
component is formed from the top down by a male-dominated institution. In Bolivia (#12), although the method used to integrate women in Bolivia National Community Development Service projects was determined and implemented by female staff at the local level, "Overall guidelines in terms of cooperative projects and infrastructure [are] set by men at [the] national level" (New TransCentury, June 1978). To combat this tendency, special efforts will have to be made in all projects to ensure that women play a major decision-making role in the parent institution as well as in day-to-day project implementation.
3. The participation of female beneficiaries in project planning and implementation-is higher when projects are located in communities with indigenous formal or informal women's associations, or with a strong tradition of community self-help. Even given these preconditions, however, women may be hesitant to express their concerns. Although the Korean pig-raising project (#/3) worked with women who were already members of home improvement clubs engaged in a rice savings scheme and other community ventures, "The women lacked confidence in the early stages of the project" (New TransCentury, Jan. 1979). In the audio-cassette listening forum project in Tanzania (1#24) which recruited leaders and participants from village meetings of national women's association (Umoja wa Wanawake Tanzania), the first group discussions of women' s priorities ". ..were stiff and formal, but as the sessions continued each participant seemed to realize that the problems she was identifying were also being heard and felt by others. The
atmosphere became increasingly open and informal" (Stanley, n.d., p. 29). A local tradition of community cooperation (Harambee) enhanced the participation of women in village committees of the Kenya rural water project (#5) as well as legitimizing their contributions of cash and labor. Women's opinions were also actively solicit ed in this project as a component of the evaluation process. In general, then, working through indigenous community groups in which women play an active role offers an excellent means of mobilizing female decision making in project design and implementation.
4. Within village women's associations or cooperatives,, the pattern of female decision making tends, to reflect the male power structure of the community. Although most evaluations do not address this issue directly, a report on the Nso Women's Cooperative in Cameroon states that "The wives of the chief and his elders were the leaders of both village groups interviewed. These leaders are often chosen on the basis of thoir respect within the community and their ability to control a group of women" (Jackson, 1979, p. 6). (For an additional discussion of this issue in another cultural context, see Dixon, 1978, pp. 139-145.) This tendency may be useful in acquiring support for the project from village elites, but disfunctional from the point of view of involving the poorest women in project decisions.
5. When rural or urban women from the target group of lowincome families play an active role in group discussions to set project priorities, they are most likely to identify economic need as their most pressing problem. In the Cameroon (#1), this concern focused on the, irregular supply and high cost of palm oil (a staple of local diets)
and on the uncertain market and low returns on marketing food crops. In other projects, women requested assistance in finding jobs or in undertaking individual or group income-generating (or saving) activities:-5
Korea #3 (pig-raising): "The women's home improvement clubs had expressed a great interest in increasing income in other than the traditional areas of farming and seaweed raising and
processing" (New TransCentury, Jan. 1979)
Costa Rica 116 (vocational training in San Jose): "The women in the Leon XIII program defined their most pressing need as economic, and requested training which would lead to actual
employment" (PES Nov. 20, 1978, p. 7)
Philippines #8 (sewing center): "Target group initiated project
idea through expressed need for additional sources of income to upgrade families' lives and recognition that skills in needlecraft already existed among them" (New TransCentury, June 1978) Lesotho #9 (mohair, tie & dye, poultry): "Because essentially
the majority of the able-bodied males are out of the country, Lesotho women feel both an urgent need and a strong sense of
duty to ... sustain themselves economically" (New TransCentury,
A second concern, expressed particularly strongly in several African projects, is women's heavy domestic and agricultural burdens. The Cameroon Nso Women's Palm Oil Cooperative (#1) originated in an earlier experience with group purchasing of mechanical corn grinders to solve the problem of extremely labor-intensive hand grinding methods. In early meetings of the nonformal education groups in Tanzania (#24) some women expressed particular anger at their heavy burdens compared with men and-their lack of control over money:
Women work just as hard as the men. Yet, when we return from
our work in the field, the men rest and we must care for the
-/ See also Bolivia #12, .Dominican Republic #15, Bangladesh #16, among others.
children, wash the clothes and prepare the man's food.
Where is our time for resting?
The money is spent on drinking, not on us or on the
children. We share the work, or do more of it, but he
takes all the money telling us it's his -- that he earned
it. It is a joke. (Stanley. n.d., p. 30)
These consciousness-raising sessions resulted in concrete plans for income-generating activities and community improvements that formed a basis for a sophisticated nonformal educational program using audiocassettes in group sessions.
Access of Women to Project Benefits The second dimension of female participation refers to the extent to which girls and women have direct access to the goods and services provided by the project. New TransCentury asked the following questions in compiling its directory:
16. How is this project designed to benefit women? Rather,
what were the unique needs of women that this project was
designed to address?
17. Please describe the major functions or activities of
this project. That is, how is this project structured to carry
out your objectives?
18. With which of these project activities are women actually
Among the 22 women-specific projects in Table 2, women by definition have direct access to benefits. The question of access in these cases relates both to constraints on the overall participation of women as reflected in the numbers of clients, and to the selectivity of those who do participate according to their socioeconomic characteristics.
For those projects with or without a specific women's component, we need to know in addition what percentage of those who have direct access to goods and services such as loans or training are female.
In three of the four projects for which the percentages can
be ascertained at least approximately, there appears to be some slippage between the anticipated representation of females as defined in project papers and their actual participation as measured in project annual reviews or other documents. In the Thailand sericulture scheme
(#2), 85 percent of expected participants were reported in one document to be female, whereas another reported actual enrollments in a 1978
sericulture training course of 103 men and 107 women.- The Nicaragua credit cooperatives for market women and agricultural producers, food processers, and artisans included 55 percent women among their 1,400 members at the time of review -- a significant accomplishment, but less than anticipated in a project titled "Rural Market Women's Cooperatives." The primary school construction project in Afghanistan (#20) was unable in some regions to fulfill its quota of 15 percent of spaces in new schools reserved for girls. Only in the nonformal education project in Ethiopia (#21) did the actual percentage of female participants (80 percent of those responding to a questionnaire) appear to meet or exceed expectations.
Six projects did not report percentages of female beneficiaries in documents retrieved for this analysis. Two are large rural infrastructure projects aimed at increasing agricultural productivity -- soil conservation in Lesotho (#4) and rural water supply in Kenya (#5) -for which figures on female participation may be available in World Bank documents cited in Table 2 but are not reported in AID summaries. Two are vocational training programs. In El Salvador (#7), although women were 45 percent of the target group of marginal urban service workers, data from the followup questionnaires of trainees showing their current employment status and incomes are not disaggregated by sex. Similarly, the Paraguay project (#13) included specific targets for intended trainees (300 women and 160 men in the first year) but not actual female participation rates in the Project Evaluation Summary (Nov. 1978).
6/ USAID, Thailand Sericulture/Settlements Project: Joint Mid-Term Evaluation, Jan. 1979.
The final two projects are the agricultural radio programs for peasant farmers in Guatemala (#22) and the integrated rural development project of Pespire Valley in Honduras (#32). The Guatemala radio campaign was aimed at two populations: illiterate highland Indians in subsistence agriculture and Spanish-speaking Latinos in the southeast. Symbolized by the radio messages called "Let's Talk, Mr. Farmer," the complex experimental scheme to change agricultural knowledge, attitudes, and practices did not once identify in a major summary document whether women participated as farmers, radio listeners, forum discussants, community change agents, or questionnaire respondents (Davidson, 1976). Women's role as agricultural producers appears to be completely ignored. The Pespire Valley scheme in Honduras included explicit references to a women's component in the Project Paper (women were to make up 50 percent of community council members in the 10 village clusters, for example), but according to the Elliott and Sorsby summary of the evaluation, the major income-generating activity for women was a cooperative mango puree plant employing 19 women during harvest, with an uncertain future. 17/
The tentative lessons that can be gleaned from the documents
regarding conditions that facilitate or impede woens direct access to project activities include the following:
6. Women have more direct access to project benefits when
planners explicitly recognize the prevailing sexual division of labor and design activities that build on women's work and enable them to control their earnings. In the Cameroon, for example (#1), men
-/ The Project Evaluation Summary cited in Elliott and Sorby p. 111-27 could not be obtained for this review.
traditionally cultivated coffee and cocoa while women cultivated food crops. Whereas an effective cooperative marketing system was available for both coffee and cocoa, food crop marketing (mostly corn, beans, and potatoes) was left to small itinerant traders (Jackson, 1979, p. 3). It was to solve this problem of unreliable markets and low returns that the members of the women's palm oil purchasing cooperative decided to organize a marketing system of their own, selling food primarily to the plantations from which they purchased palm oil and using the same transport for both transactions. A plan for marketing food crops through the male-dominated coffee or cocoa cooperatives would undoubtedly result in lower food crop production as women lost control over their earnings. The Thailand sericulture project (#2) was explicitly designed
to reach girls and women who carried out most of the principle tasks connected with traditional methods of sericulture. In the Philippines sewing project (#8), women recognized that their needlecraft skills formed a natural basis on which to build income-generating activities.
The strategy of building on women's work reduces the likelihood that resources will bE? co-opted by men, but the danger always remains that once an enterprise becomes profitable, men will be interested in taking it over. In the Gambia, for example, about 4,000 women were successfully growing onions as a cash crop when male farmers decided for the first time to ask the government for similar assistance (Tinker, 1979, p. 13). Although their onion schemes were apparently not successful (in part because their wives refused to work on them), the importance of control over earnings as an incentive to production remains central.
7. Project activities that f it with prevailing cultural norms and the allocation of household responsibilities attract higher rates of female participation by reducing resistance from the women themselves, from their husbands or fathers, and from the community at large.
Although this approach is a conservative one, it appears to offer a valuable means of providing larger numbers of girls and women with direct access to benefits, while at the same time acting as a sort of wedgeg" with which to introduce more comprehensive changes in a phased sequence. The Bolivian income-generating project for peasant women (1#12) is a good example:
Women are ... receptive to certain types of participation.
What we attempt to do is build upon those areas of participation
which are acceptable in order to generate others. Their husbands tend to feel the same, in that they support activities
which don't detract from normal time usage, chores, etc. (New
TransCentury, June 1978).
Similar experiences are reported in Bangladesh, the Philippines, and Costa Rica. In Bangladesh,
One of the problems of the rural woman in Bangladesh is that she
has no opportunity to earn an independent income and thus contribute financially to the welfare of her family. Because of
the system of Purdah, she is often isolated in her compound with
l-ittle chance to learn income generating skills, gain educational qualifications or participate with other women in social
activities (New TransCentury, Jan. 1979).
In its organization of women's cooperatives, the Bangladesh project (#16) worked within these restrictions by offering women new possibilities for earning independent incomes in rice cultivation, seed distribution, oil pressing, etc. while providing an acceptable mode of association with other women. Women in the Philippines (#18) expressed similar concerns:
Income-generating women's projects to be engaged in [poultry,
pig fattening, mushrooms, floriculture] are home-based and small scale entrepreneurial pursuits since the majority of
rural Filippino women express a desire to combine their familyhousehold roles with new, compatible means of attaining higher
levels of living (New TransCentury, Jan. 1979).
In San Jose', Costa Rica (#6), women reported that "... their
husbands did not want them to leave the immediate area to seek employment or training because of family and household obligations" (Project Evaluation Summary, Nov. 1978). Participants decided to locate an industrial sewing facility and baking cooperative within their residential area; evaluators recommended that "An orientation of the project be provided to husbands and fathers of participants."
Some constraints can be overcome simply by redesigning projects to include child-care or dormitory facilities or by adapting the timing or duration of training and employment to women's daily and seasonal round of work. The Costa Rican women, identifying the complete lack of child care as a major impediment to their participation in training sessions, established a temporary facility with plans for a permanent one. Village women in Bangladesh were able to attend training classes in silkworm production when a dormitory for non-commuting students was constructed. Evaluators of the El Salvador vocational training program recommended that certain courses be offered in smaller communities by mobile teams rather than in larger towns or cities requiring extended absences from home. Policies such as these would certainly facilitate women's access to services. In the long run, of course, efforts would be directed toward distributing domestic responsibilities among all household members to reduce the weight of women's double burden in the home.
8. Women's direct access to project goods and services is
frequently limited by customary-or legal restrictions on their right to resources such as land, credit, or schooling. Forming the core of a complex system of social stratification based on age and sex (among other attributes), these structural and cultural barriers can sometimes be penetrated with the permission of higher status individuals or groups if the latter can be mobilized to support the project.
Lack of independent land rights can be a severe impediment.
Women in the Cameroon cooperative, for instance, receive permission for land use through their husbands who petition on the women's behalf to the village chief. In the land conservation scheme for subsistence farmers of Thaba Bosiu in Lesotho (#4), women's access to training in animal husbandry or to decision-making positions on range management committees is restricted by the traditional practice of granting grazing land at the age of majority to every Basotho male. In Upper Volta (#10),
At the village level men must be consulted and support gained if women are to participate [in new income-generating activities], because a male head of household usually has authority
over the allocation of time and labor of its female members.
Also, if the land is needed for an activity since males control access rights to land their consent must be obtained. Further, the involvement of men may be necessary to undertake activities,
such as construction of buildings and fences, since according
to social norms certain functions are performed by men
(Barnes, n.d., pp. 6-7)5.
Projects frequently require that credit be granted to males
'even when production is based on women's work. In the Thailand sericulture project, for example, each family is entitled to a loan through the Bank for Agriculture and Agricultural Cooperatives to get the project started. Although the available documents do not say
whether women can apply as individuals, the family unit is usually represented in formal transactions by the male household head.
Even the basic right to receive certain kinds of training may be denied to women.by local tradition. In the Upper Volta project, female regional coordinators and national staff in the domestic economy units are under the direct supervision of men; "hence, the consent of these men is essential for project success. For instance, these officers have to be willing to permit the female staff to participate in training sessions" (Barnes, n.d., p. 7; emphasis added).
The most extreme example, however, comes from Afghanistan (#20), where only 10 percent of primary school students in rural areas are girls. Most rural parents refuse to send their daughters to a coeducational school, yet two provinces have no village schools for girls at all. According to the Project Paper, "... traditional Afghan values dictate segregation of the sexes. ... Segregation is more strongly enforced as children approach puberty, resulting in more female dropouts" (Jan. 1975, p. 43).
In sum, tactics may be needed in many projects to persuade those in power -- village leaders, employers, parents -- to "permit" girls or women to transcend the traditional restrictions on their mobility, with the long range goal of building an independent power base from which to press for more basic cultural and structural transformations.
9. The poorest women in the community typically havq less access to project goods and services than those who are relatively better off. Although this is not surprising in view of the findings from most development projects, the situation is sometimes exacerbated by setting unnecessary criteria for eligibility. Although the Costa
Rican vocational training project was aimed at women in low-income families of San Jose', the original requirements were set at a 6th grade education. Realizing that this excluded many poor women, organizers subsequently dropped the requirement. In the Korean pig-raising project, although women were selected from home improvement clubs on the basis of their low income and their expressed interest in the project, "Not enough of the very poor families in the communities were involved. The initial groups participating tended to be a little better off and more highly motivated" (New TransCentury, Jan. 1979). By working through formal women's groups, organizers were able to capitalize on the women's decision-making skills but bypassed the poorest, most isolated women who remained outside the structure of community associations. The poorest village women in the Cameroon were also largely excluded from thepalm oil and food marketing cooperatives, in part apparently because some could not buy a share or the membership fee (Jackson, 1979, p. 9). Reaching the poorest of the poor may require bypassing traditional community associations or informational networks, an approach that raises additional difficulties of recruitment and communication.
10. Shortages of volunteers and of trained female staff pose major obstacles to the recruitment of more women as beneficiaries.
Although such shortages are likely to pervade most types of development projects, women-specific projects are particularly hard hit because they tend to rely more heavily on volunteers due to funding limitations. In addition, women working in the field as volunteers or as paid project staff members face special problems.
The first issue is the heavy reliance of some projects on volunteers. Although several evaluations report high levels of enthusiasm among volunteers, others hint at their lower incentive to work (e.g., Lesotho #9). In either case, the capacity of projects to reach large numbers of beneficiaries is constrained by the number of volunteers. The Costa Rican vocational training program, for example, "Currently has more requests for training from government agencies than it can provide, and thus the possibility of reaching the desired number of women exists, but is limited by the current number of available volunteer trainers (9) and research aides(3)" (Project Evaluation Summary, No. 1978, p.. 2).
Although the use of volunteers permits women' s projects to
operate on a larger scale than would be possible if all staff positions were paid, the practice raises serious questions regarding the motivation of governments or donor agencies to provide serious funding for women's projects. It is doubtful that a large-scale project aimed at training men in agricultural techniques or industrial skills would expect male staff members to contribute their labor. The funding of staff positions on women' s projects would also permit more active recruitment of women trainees from the target population of beneficiaries, rather than relying on volunteers from a different social stratum. Grants to PVOs for leadership training of volunteers (e.g., Ghana #25 and #26, Upper Volta 1127, Latin America #30, Costa Rica #31) tend to support middle and upper-class women who may or may not translate their training into genuine advocacy for the poor.
The second issue relates to the working conditions of paid
female staff, especially those working as promoters in rural areas. The
model of training rural women to teach other rural women, clearly a cost-effective approach to community development, is followed by a number of the projects in Table 2. In Bolivia (#12), promoters are peasant women trained by the National Community Develo*pment Service and assigned to communities based on the type of project solicited. The promoters are all literate although the female community leaders may not be. One of the major problems with the program, according to one evaluation, is that the personnel live under difficult, isolated conditions with low salaries.
The social isolation of outsiders in villages is often acute. In Afghanistan, for example,
Qualified teachers have been reluctant to work in remote rural
areas where motivation is difficult to maintain and urban incentives are non-existent. ... There have been no houses
for rural teachers. Social isolation from the community and local politics, and reinforcement of social distance between
villages and teacher, has been commonplace (Project Paper,
Jan. 1975, p. 4).
Women teachers in Afghanistan, who are more likely to come from towns or from Kabul than from rural backgrounds, are especially reluctant to seek rural jobs. "Women are unlikely to be allowed, nor would they voluntarily seek to live alone without relatives to protect them, and thereby diminish the possibility of marriage." Women's reluctance to tea-], in rural areas increases the reluctance of parents to send their daughters to school, for most parents outside Kabul will not send their girls to school unless they are taught by a woman.
Even when workers are selected for training by women within their own village and return there to work, conditions are often difficult. Although promoters in a community organization project in Bangladesh (#28) were trained at district centers near their homes,
Lack of appointment letters for field workers and irregular receipt of salaries is shattering the morale of the workers.
Perhaps this is one cause of the resignations of field workers
who are seeking more job security in other development
programs (Huber, 1978, p. 1)
Lack of communication of field workers with central offices is an additional source of frustration for many. Whereas the Bangladesh income-generating cooperative project (#16) cites as one basis of its success the close supervision and contact with beneficiaries deriving from the location of field offices in the same geographic areas, other evaluations refer to the low levels of communication between central offices and rural areas (e.g., Dominican Republic #15, Bolivia #12) which result in a lack of "fit" between centrally made policy decisions and local needs, as well as in the sense of isolation of field workers. In many cultures, women field workers face additional restrictions on
their ability to travel freely either between villages or to urban centers.
In addition, female extension agents are often expected to
perform too many diffuse functions with inadequate training in any of them. Domestic economy workers in Upper Volta, for example (#10), are supposed to promote and supervise both agricultural and nonagricultural production activities for women, as well as home economics, hygiene, health, literacy, and other functions. Village promoters in Bangladesh 028) learn nutrition, family planning, home management, cooperative principles, leadership skills, and'adult literacy in the district training centers. Combined with a general negligence of follow-up training, the lack of a sharp focus to the promoter's role appears to contribute to feelings of inadequacy. In many cases, these problems could be overcome by promoting improved working conditions such as
higher salaries (at least equivalent to what male workers earn in similar capacities), more followup training permitting workers to discuss their experiences and problems with one another and with their supervisors, and safe transportation and housing facilities for women workers. All of these could be built into the project design.
Effects on the Status of Women
This dimension refers to the immediate and long-term consequences of the *project -- both positive and negative -- to women in the project area, whether or not they are (or were) direct participants. Both absolute changes in the status of women, and changes in their position relative to men in the family or household and the community, are key issues here.
Adequate answers to these questions depend of course on the
completion of systematic evaluations. Although the Elliott and Sorsby report intended to include those Women in Development projects that have been evaluated, it is clear both from their project profiles and from further examination of the documents that many "evaluations" are drawn from sources outlining expected benefits. In the case of AID, these are sometimes End of Project Status (EOPS) statements in Project Papers; in the case of PVOs, these are sometimes informal observations in which the method of review or substantiating evidence is not described.
In its questionnaire regarding project impacts on women, New Trans Century asks:
21. Has any project evaluation been done to date? If so, what results did you expect? What results were achieved?
How did you go about measuring this?
22. In order to participate or benefit from the project,
what sacrifices or accommodations did the women beneficiaries
have to make? What tasks did not get done? Was their routine complicated by additional tasks?
23. In retrospect, so far, what parts of this project have
been especially successful? What has worked out best?
For 13 of the 33 projects listed in Table 2, evaluations had either not yet been conducted or were not cited in available
summaries,-8 although one did include plans for doing so (#15). Most of these summaries nevertheless refer in a general way to project effects such as increased earnings or a growing sense of enthusiasm and self-confidence among participants.
Among the rest, almost all of the evaluation documents examined are inadequate for our purposes in two respects. First, they tend to concentrate overwhelmingly on measuring project outputs in relation to purposes -- the number of persons enrolled in a training program, the number of members of a cooperative, the number of households connected to water supplies -- while ignoring the larger issue of how the project affects the everyday lives of beneficiaries. As such, they really measure participation rather than impact. Second, among those projects that are not women-specific, data on participation or on impact are rarely disaggregated by sex. Thus we have little idea of the consequences to females either in absolute terms or relative to males.
Several vocational skills and nonformal education projects offer good examples of these approaches. The vocational skills program for low-income people in Paraguay (#13) provides data on the number of coordinators, promoters, course participants, and income-generating projects started, without reference either to numbers of men and women or to the impact of these activities on their lives (Project Evaluation Summary, Nov. 1978). Three projects with sophisticated evaluation techniques that include baseline data and followup measures do not disaggregate their data by sex. El Salvador (#7) reports the employment and income status of 60 percent of over 1,000 graduates from vocational training classes 18/ Project numbers 2, 3, 8, 9, 10, 11, 17, 23, 25, 26, 27, 29.
a year later according to subject matter, but not by sex; Ethiopia (#21) includes indicators of changes in attitudes and behavior in literacy, nutrition, health, family planning, etc. without reference to gender; and Guatemala (#22) includes a wide range of logistical data on numbers of promoters trained and communities reached by radio programs, including measures of changes in agricultural knowledge and practices of farmers in experimental and control villages, but without specifying differences between men and women.
Two women-specific projects do include interesting information
on project impacts. Costa Rica (#6) reports "a measurable increase" in the economic level of women trained in industrial sewing, an improvement of housing conditions, increased access to social services, and important social effects:
Participants in the motivation/adaptation training repeatedly
expressed that they now have higher aspirations, more selfawareness, a better self image; are able to identify problems
and work on solutions; enjoy working in a group; and that
there is a definite change of attitude on the part of their
husbands (PES Nov. 1978, p. 9).
(Conclusions are not substantiated by quantitative data, however, which were to be obtained from follow-up questionnaires). The Tanzania audiocassette nonformal education project (#24) includes both pre- and posttests of women in participating and control villages which, combined with evaluation seminars and unstructured observations, provides information on changes in a variety of attitudes and behaviors (Stanley, n.d.)
Others are more general in their discussion of project impacts, at least in their report summaries. An evaluation of the comprehensive Lesotho soil conservation scheme (A), noting among other things that "Experience thus far has shown no indication of any groups either
resisting or being harmed by the project," adds that "With the possible exception of profits to some villages from fish ponds, it has not yet caused any increase in rural income and with the exception of a few households, will not bring about any appreciable increase during the projected six-year life of the Project" (Evaluation Report, Oct. 1975, Annex V). The CARE-assisted Kenya rural water project evaluation (#5) focuses primarily on the administrative and managerial problems of meeting the target numbers of water beneficiaries, pointing in addition to the general absence of a significant self-help component in most community efforts. Although the EOPS of the Project Paper anticipates major benefits to women (less time spent fetching water, more time for child care, family betterment, agricultural production), the evaluators conclude that "CARE's overweening preoccupation with the benefits of these projects upon women and ... upon agricultural productivity and health is unjustified," in part because the assumption that women will use the time productively is untested, and in part because additional technical inputs would be required to achieve these purposes that are not forthcoming (Biggs and Schott, 1976:32).19 1
Given the uneven quality and quantity of impact data, the task of gleaning lessons from evaluation summaries becomes even more precarious. Nevertheless, the following generalizations are offered on the basis of the evidence available for these 32 projects:
11. The social impact of projects is magnified when women are organized for group action, particularly when they were previously
9/ A report by CARE (Oct. 1977) on project impacts on women was not available for this review.
confined to their households or were unused to collective activity The process of getting together appears to have stimulated considerable enthusiasm among women across the range of projects under consideration here. Seven evaluations referred specifically to the group process as a vital ingredient of project success, while one referred to the loss of social contact as a project "cost" (women in Cameroon selling food crops to cooperative buyers rather than in the marketplace, as before), and one mentioned the lack of social cohesiveness as an obstacle to achieving project goals (community development in the slums of Colombo, Sri Lanka).
Group activities tend to engender feelings of pride, self-confidence, and skill (6 projects), of eagerness and enthusiasm (2 projects), of awareness of leadership skills (2 projects), and of higher aspirations (1 project). When limited to women only, the group may offer the only culturally acceptable means for participants to leave their homes for collective action (e.g. for purdah-observing women in Bangladesh). The San Marcos sewing center in the Philippines, for example, is described as having the following impact on the women who participated:
Women have achieved an ongoing additional source of income
from an idea they originated, thus not only are the results tangible in the growing business but intangible as well in their well deserved sense of pride and competence at their
The fact that the project started within its own building
provided women with the opportunity to literally "go to work."
The women responded enthusiastically to this innovation in
their lives. (New TransCentury, June 1978)
12. More lasting effects may be experienced when women unfamiliar with organized cooperative efforts begin with a single activity that carries clear and immediate benefits, then move into other activities
as their skills and confidence increase. The Cameroon women's cooperative provides a good example of a phased sequence of events. In the first stages, women mobilized for the purchase of corn mills to reduce the daily drudgery of grinding corn at home: subsequently, and after a considerable time lapse, they organized a palm oil purchasing scheme, then a food marketing scheme, with future plans for fertilizers and consumer stores. The project description concludes with a clear lesson: "Starting slowly and answering one problem or issue at a time has allowed the leaders to grow at their own pace and ability" (Jackson, 1979, p. 9).
13. The achievement of concrete economic benefits is a key
motivating factor responsible for maintaining group activities. Fifteen projects cited higher incomes (or savings) as indicators of success, e.g. additional net incomes per family of $150 a year for the Thai farm families engaged in sericulture (#2), and profits from the pig fattening project of a women's group in the Republic of Korea (113). Evaluators of the Kenya project for women (#11) report that "The incomegenerating part of the project has been the most successful area of endeavor," while those of the Ethiopian nonformal education project (#21) recommend greater emphasis on income-generating projects and areas of community development. The absence of clear economic benefits may reduce the motivation for some women to participate. Although the linkage is not explicit, evaluators of the Sri Lanka project consisting (among other activities) of lectures on nutrition, sanitation, and home gardening for mothers reported that "The program for the mothers had a limited success as the attendance was sometimes below expectation' (New TransCentury), Jan. 1979).
Evaluations of the Lesotho land conservation scheme doubted that rural households would willingly alter their patterns of production in the desired direction (controlled livestock grazing, etc.) unless yields could be increased by at least 50 percent or even doubled.
14. Marketing proves to be one of the most difficult obstacles to creating viable economic enterprises based on the small-scale production of most rural women's projects. Not only do female beneficiaries frequently lack the skills needed to establish regular markets, but they face additional problems such as cultural disapproval or personal reluctance to carry goods to market (Bolivia #12), lack of access to transport (Fiji #17), and difficulties of competing with established enterprises. Profits from the palm oil sales of the Cameroon women's cooperative, for example, helped subsidize the marketing of food crops:
The cooperative structure is a constraint to making food marketing
profitable -- particularly when competing with private dealers,
who have lower transportation costs, sophisticated marketing
contacts. In addition, members do not view the cooperative as
their own but rather as an outside force with unlimited
finances. This often results in members demanding unreasonable
prices for their produce at the institutional level (Jackson,
1979, p. 8).
Project designs should pay explicit attention to marketing requirements to ensure that income-generating schemes for women can become genuinely self-supporting, rather than being maintained as ficharitable" enterprises based on a welfare mentality. By the same token, of course, vocational training projects should be based on a thorough analysis of local labor market conditions to ensure that trainees can find employment at least at prevailing wages. Although the El Salvador vocational training courses (#7) reduced the overall
level of unemployment of trainees from 68 to 48 percent, students graduating from the cooking and cosmetology classes (presumably almost all female) maintained the highest levels of subsequent unemployment (over 60 percent), while those graduating from automechanics and welding classes (presumably male) had the lowest rates(13 and 36 percent, respectively) (Final Report/Evaluation, May 1978, p. 7). Evaluators recommended revising the types and locations of courses offered in order t o assure that they respond to real needs for skills training applicable to local labor market conditions.
15. If girls and women--are not specifically identified in project papers as intended beneficiaries, they are likely to remain invisible in planning and evaluation documents. This general observation relates to two processes. First, when clients are defined as pooro, unskilled, urban and rural students," "poor children ages 6-14 and their families," or "small scale, near subsistence farmers," the evaluator is not reminded to look at the distribution of project benefits among male and female unskilled students, or male and female poor children, or men and women within farm families.
Second, in the absence of a clear identification of females
among clients, project administrators are less likely to consider women's needs and priorities during design and implementation, and less likely to train female workers to serve a female clientele (e.g., the women of farm families who would benefit from direct assistance in agricultural production and marketing of their own crops).
This shortcoming derives from the role of social soundness
analysis in project design. Rather than forming the basis for deciding what types of interventions are best suited to particular socioeconomic
situations -- that is, rather than actively shaping decisions about project type, location, and design -- most social analysis appears as an afterthought, a justification (required for AID project approval) of decisions based on other criteria. Yet, as Heli Perrett points out in her perceptive report,
Such design stage social analysis prepares the way for social
analysis during and following inplementation through identifying
critical assumptions about the process of development and flow
of benefits which can be monitored; identifying possible
negative consequences and distributional questions which should be looked at during project evaluation; identifying longer term
social changes which might come about as a result of the
project and the early signs which precede them (Perrett, 1978,
Perrett's comments are particularly relevant to the question of women. AID Country Development Strategy Statements rarely pay specific attention to women's social and economic roles, and more rarely link these to planning strategies.0/ Critical assumptions appearing in the Logical Framework rarely refer to social issues such as the distribution of benefits within the household. Although the project development process outlined in Handbook 3 requires specification in Project Identification Documents and Project Papers of identities of intended beneficiaries and the extent of their proposed participation, few explicitly refer to women.-/ The lack of attention to the differential impact of projects on males and females during project monitoring and evaluation is consequently not surprising.
-0/ Memo from Patrick Fleuret, PPC/PDPR/HR, "The treatment of developmentlinked male/female roles in 46 current CDSSs" (March 26, 1979) and response from Kathleen A. Staudt, PPC/WID, "Women in development: the CDSSs" (May 18, 1979).
-1/ The situation should be improved considerably with the adoption of Handbook 3 revisions that pay special attention throughout to the roles of women. See memo from Kathleen A. Staudt, PPC/WID, "Recommended additions for revision of Handbook 3" (February 28, 1979).
Preliminary social soundness analysis plays at least two key roles.-2 One is to protect people from harmful consequences. As William Sif fin points out,
The basic aim is negative -- to minimize the likelihood of error
the kind of error that has too often diverted project benefits
from intended recipients, that has produced success at the cost of non-replicability, that has damaged those who were intended to be helped. Social soundness analysis is protective, aiming
to avoid certain kinds of outcomes as well as foster certain
general values (Siffin, n.d., p. 6).
This protective role relates directly to the question raised in the introduction: under what conditions are women least likely to be disadvantaged by development projects, either in absolute terms or relative to men?
A second role of social analysis is to prevent expensive
failures that demoralize AID donors and recipients alike. A preliminary analysis that leads to (among other things) the "discovery" of women's often hidden social and economic roles will improve the chances of meeting project purposes and goals. In the absence of baseline information on the sexual division of labor in the production of goods and services for domestic consumption and for sale or exchange, on the sexual distribution of rewards, and on the participation of women and men in household and community decision making, assistance may be misdirected. .A program aimed at increasing agricultural production by organizing marketing cooperatives for (mostly male) household heads, for example, is likely to fail if the crop was traditionally grown and
22 Guidelines for conducting social soundness analysis are outlined in USAID Handbook 3, Part III, Annex F, and in Ingersoll (1977) and Siff in (n .d.).
marketed by women. As men gain control over the economic returns to their wive's labor, the women lose their incentive to produce, and output falls. 13/
The first step in designing an evaluation system with an institutionalized concern for the interests of girls and women as beneficiaries, then, is to make sure that questions regarding their participation are raised throughout the entire cycle of project identification, design, implementation, monitoring, and evaluation, leaving room for corrective action at every stage. The second step is to develop a procedure for evaluating the differential impact of projects of all types on males and females so that the "lessons learned" can be compiled and compared in a systematic fashion. Some suggestions for doing so are outlined in Sections III and IV.
23/ A number of examples of unintended negative effects are presented in the discussion of project impacts in Section III.
III. ASSESSING THE IMPACT OF DEVELOPMENT PROJECTS ON WOMEN: AN ANALYTICAL FRAMEWORK
The discussion in Section II focused on three dimensions of female participation: their role in project decision making, their direct access to benefits, and the effects of the project on their position in the family and the community. Information on decision making and on access to benefits is far easier to collect from project records and interviews with personnel at various levels than is information on the immediate or long-term effects of the project on the status of girls and women. Evaluation documents, on the whole, are addressed primarily to the logistical questions of whether the project is "on time" and functioning as planned rather than to social impacts. Although this preoccupation with institutional support and technical transfers (the means of development) is understandable, it obscures the more important questions of the wellbeing of recipients (the ends of development). What happened to trainees and their families after they completed their courses? Did they find jobs? Where? Did cooperative members become more productive, use the credit for intended purposes, participate in community decision making? Who were the primary beneficiaries of large-scale rural water supply systems, and how did their lives change? How were benefits distributed between males and females? Who gained, and who lost? In what ways?
This section describes an approach to performing both mid-stream evaluations of the immediate social effects of ongoing projects, and ex-post evaluations of long-term impacts following project termination. By choosing projects rather than sector-level or country-level analysis,
more qualitative as well as quantitative methods of assessing changes within and among households can be considered. Questions raised in the evaluation framework regarding differential effects on males and females can shape the inquiry of the preliminary social soundness analysis and baseline data collection. Such questions can also form a basis for project monitoring, especially if formal evaluations are scheduled too late for major revisions of project policies and practices affecting women.
Whether the project is a women-specific vocational training program ("low-income women from barrios of San Jose"'), a poultry development scheme with a women's component ("small farmers in the Yemen Arab Republic of whom women are expected to assume a major role"), or a radio education program aimed at a general population ("16,000 illiterate highland Indians engaged in subsistence agriculture"), the process of assessing its social impact on women and men, in both absolute and relative-terts, would address two major questions.
First, what are the direct effects on primary beneficiaries
(male and female), i.e., now has their access to important resources changed in absolute terms as a result of the project? This target population may be defined as classes of individuals such as malnourished
children or unemployed youth; classes of households such as tenant farm families or members of an agricultural cooperative; or whole localities such as villages or districts within integrated development schemes.
Second, what are the distributional effects of projects across different categories of persons both within and outside the targeted population? Of particular importance to the analysis of women's
situation is the allocation of resources within households. How has the project altered the division of labor within the household, or the distribution of resources such as food or schooling, on the basis of t he sex, age, or relationship to household head of family members? When families or households rather than classes of individuals are defined as the primary beneficiary unit, do benefits accrue to all household members or is the domestic balance of resources significantly altered depending on who has direct access to project goods and services?
We are also interested in analyzing distributional effects of projects across households, both withinand between the targeted and non-targeted populations. Are women-head households bypassed when titles are granted in land distribution schemes, for example? Do benefits from membership in a women's marketing cooperative accrue disproportionately to those from better-off households or from dominant racial, religious, or ethni c groups? Does a project that introduces new technology in agricultural or nonagricultural production to one class of persons (say, men employed in modern fish freezing facilities) result in the displacement of another (e.g., women engaged in traditional fish drying and marketing activities)?
Table 3 summarizes the framework for evaluating the absolute and relative effects of projects on people's access to a wide range of resources that determine their physical, economic, and social wellbeing that is, their position in the household and community social structure. In the following discussion, some examples are given of how projects can affect males and females differently, in some cases narrowing the gap between them in their access to important resources, and in other
FRAMEWORK FOR EVALUATING THE DISTRIBUTION OF PROJECT BENEFITS
Direct effects Distributional effects Resources that on primary nl within across
determine beneficiaries- households households
1. food, water, fuel
3. environmental quality
4. medical care
5. personal safety
6. rest and leisure
7. income/cost of living
9. land and water 10. technology 11. other assets/debts
12. knowledge 13. power
-/ Primary beneficiaries or target populations may be defined as classes of individuals, of households, or of localities.
cases widening it. Sometimes, of course, improved access to one type of resource may impede access to another. Full-time wage employment for women, while offering badly needed cash incomes, can result in the elimination of home garden production with severe nutritional disadvantages for the entire family, the early cessation of infant breast feeding, and a critical shortage of agricultural labor during periods of peak demand, among other consequences.
The likelihood of mixed results poses a major dilemma to policy makers who must decide just what economic or social costs recipients may be expected to bear in order to achieve some other (presumably more valuable) economic or social benefit. The question of "what works and what does not in specific socioeconomic environments" is thus complicated by the fact that a project can work in one way but not another. Purposes may even be contradictory: for instance, a project might increase agricultural production more rapidly -- at least in the short run -- through authoritarian leadership that contradicts the goal-of consciousness raising and participatory decision making.
The framework summarized in Table 3 and elaborated below includes a number of resources that determine people's physical, economic, and social wellbeing. Implied in the framework is a definition of development that includes 11(l) a general improvement in levels of living, together'with (2) decreasing inequalities of income distribution, and
(3) the capacity to sustain continuous improvements over time" (Kocher, 1973, p. 5). This latter element, which derives in large part from the acquisition of important social resources such as knowledge, power, and prestige, is of particular relevance to the analysis of women's situation.
Indicators of Physical Wellbeing
1. Access to food, water, and fuel. How has the project affected the overall quantity, quality, and security (e.g., severity of seasonal shortages) of food, water, and fuel consumed by primary beneficiaries? What are the distributional effects within and across households?
A project designed to increase food production that does not take into account the sexual division of labor in agriculture, for example, can have the unintended consequences of reducing household food consumption. In the MWEA land settlement scheme of Kenya, women who worked long hours with their husbands (or in their husbands' extended absence) in the new irrigated rice fields had less time than before to grow traditional foods on their own garden plots, which were too small in any case to provide for the family's subsistence. While rice production increased and total incomes among participating households rose, nutritional levels fell (Hanger and Moris, 1973). If the importance of women's food production had been recognized, provision could have been made for agricultural inputs such as credit, fertilizers, and seeds for the garden plots as well as the rice plots, perhaps with some experiments in labor-saving collective production and marketing of traditional foodstuffs.
Similarly, even a project that increases total household food consumption can worsen the nutritional status of certain household members. In many cultures, women and children eat only when the men have finished, and male children are given preference over female children when food is scarce. Although family food rations paid to male laborers on food-for-work construction schemes in Bangladesh are
intended for six persons, the heavy labor increases men's caloric requirements and leaves women and children in some families with less to eat than before. _2/
Some projects attempt to alter the pattern of food allocation within the family directly. Feeding programs for children in primary schools, for example, benefit children over adults (assuming that children's home rations are not reduced by an amount equivalent to that consumed at school) and also boys over-girls, to the extent that daughters are kept out of school more than sons. Feeding programs for pregnant and nursing mothers attempt a similar reallocation in favor of nutritionally vulnerable women, assuming that beneficiaries actually consume the full portions meant for them rather than sharing with other family members (Singer, 1978). It is clear that project planners must understand fully the role of girls and women in food production, processing, distribution, and consumption if overall targets are to be met in a manner that distributes benefits equitably.2?3/
2. Housing. How has the quality of beneficiaries' housing changed since the project began, as measured by the adequacy of protection against the elements, the amount of space per person, household amenities, and other indicators such as cultural acceptability? Do home improvements (e.g., the acquisition of electricity) benefit one sex more than the other, or one type of household more than another? Are there negative
-2/ Interview with Penny Satterthwaite, United Nations Fund for Population Activities, Dacca, Bangladesh, Feb. 1976; see also Chen and Ghuznavi (1977) on women in food-for-work schemes. 23/ See, for example, the two volumes of Proceedings and Papers of the International Conference on Women and Food (1978).
aspects that might affect females more than males, such as increased crowding due to immigration to an employment scheme, or insufficient temporary shelter for migrant workers on construction projects?
Some projects specifically designed to improve the housing supply might unintentionally bypass certain classes of households. A community development project, for example, offers low-cost loans for building .material in a self-help housing scheme dependent on volunteer labor from participating households. Women-headed households, i.e., those without adult males present, are less likely to qualify because they cannot provide the necessary labor (or are not included in the relevant training programs) and are too poor to pay back the loans. Or a resettlement scheme may construct housing without consultation with future residents, with the result that the design and location of units are inappropriate to women's needs. Given that in most societies women spend more hours per day on average in the home than men do, any positive or negative change in housing quality is likely to affect them more intensely.
3. Environmental quality. This component of physical wellbeing is measured by the existence of sanitary facilities, the adequacy of drainage, the degree of air and water contamination, and the prevalence of environmentally based diseases, among other indicators, as measured in the home, the workplace, and the community at large.
The relationship of persons to their physical environment is
determined in part by the sexual division of labor and by cultural norms determining the nature and location of various patterns of daily or seasonal activity. Because males and females within any cultural setting interact somewhat differently with their physical environment, projects
that have a strong environmental impact (either positive or negative), or that alter significantly the sexual division of labor, may affect the health of males and females differently. In villages where women rarely leave their household compounds, for example, the provision of latrines or of bathing facilities near the home will carry greater benefits to females than males, who work in the fields and bathe in the rivers. At the same time, a program to eradicate water-borne diseases such as schistosomiasis would benefit disproportionately those in most contact with the water, i.e., boys and men. Certain shifts in the sexual division of labor may also carry unintended negative consequences for women. Successful agricultural projects that increase yields and household incomes in traditional purdahobserving societies may lead to the withdrawal of girls and women from the agricultural labor force to the more highly valued (and now affordable) practice of female seclusion. With darkness and inactivity, the incidence of vitamin D deficiency and associated diseases (sometimes including obesity) increases significantly.
4. Medical care. How has the project affected the accessibility of the target population to preventative and curative health care and to family planning information and services? As for food, access to medical resources is likely to be distributed unevenly within the household as well as across households, with preference given to certain members of the family such as highly valued male children or the adult male breadwinner. In addition, women's access to medical care is constrained in some societies by cultural restrictions on contact between males and females. A project designed to deliver health services will require in almost all settings a clear women's component, with
special efforts to train female medical practitioners to reach a female clientele in an environment sensitive to their concerns. Failure to do so can result in outright rejection of the program, or in unnecessary suffering for women who are desperate to seek help.
5. Personal safety. This aspect of physical wellbeing is rarely discussed in the evaluation literature, yet it can be central to the concerns of project participants: the degree of exposure to, and protection from, personal violence or accident and injury. Have risks of certain types increased since the project began? For whom? Projects may unwittingly expose girls and women to personal dangers that they either must endure, if there is no choice, or that contribute to project failure. The protection of girls of marriageable age, in particular, is a matter of extreme concern in societies placing a high value on female chastity as a symbol of the honor of the kin group. A training program that requires young women to walk long distances
from their villages to the training center, thus exposing them to the risk of sexual harassment, leads to the termination of the program by administrators who mistakenly decide that the high incidence of dropouts represents a lack of interest among participants. The provision of safe transport for trainees (or safe places to stay overnight, where necessary) would solve this problem.
In another example, the spread of more lucrative coffee and cocoa plantations outward from a village in Cameroon forced women off the cleared land to more distant fields for their food crops:
Food fields are anywhere from one to ten kilometers from the village with three to six kilometers most often cited. This distance implies a one-half to one and one-half hour walk to the food fields over rough forest paths, often with slippery
stream and marsh crossings. The worst aspect of the trek
comes during the return -- a woman is often carrying the daily food supply of cassava, plantain, and corn, plus firewood, and
often her baby as well. The weight is anywhere from 30 to 80 pounds. Injuries from falls or scrapes are common, and much
spontaneous abortion and persistent backache is blamed on this
aspect of women's work (Henn, 1976, quoted in Tinker, 1979,
Again, careful attention needs to be paid to measures that will increase the safety of working conditions for women, such as more accessible locations and access to means of transport that men currently control.
6. Rest and leisure. This component is measured by the intensity (energy) and extensiveness (time) of labor required for production for household consumption or for sale or exchange, and by the number of hours available to household members for rest, leisure, and sleep. Daily, weekly, and seasonal fluctuations are important. Studies of time allocation within and across households in different settings reveal differences not only by household composition, landholding status, occupation, and seasonality, but by age and sex. When production for household consumption is considered in combination with production for sale or exchange, women are frequently shown to have less time for rest and leisure than men. Often they are the first to rise in the morning and the last to sleep at night.
How does a project affect the intensity and extensiveness of female labor, both in absolute terms, and relative to men and other classes of women? This depends heavily on the current division of labor by class, age, and sex, and on the nature of the project. If it is aimed at increasing agricultural production, for example, are the new expanded crops or activities primarily women's work or men's work? Do new technologies reduce the time male farmers spend ploughing the fields, but
not the time women farmers spend in weeding and harvesting? Do changes in cropping patterns require heavier seasonal labor commitments from women that detract from their ability to care for older children or breastfeed their infants? In an evaluation of the effect on women's workload of a hypothetical land settlement scheme based on a composite ,of several actual projects, Palmer (1979, p. 50) summarizes:
Given the traditional sex-typing of agricultural tasks, the
effect of new labor requirements has been the continued yearround work of landed women accompanied by greater intensity of daily work schedules at seasonal periods. For men, the effect has been more days of the year worked, through double-cropping,
but no greater intensity of work than previously.
Women face conflicts between work on the subsistence and rice crops, and between child care and productive work at seasonal
peak periods. Women are unable to use exchange or hired labor
to ease their burden (except for planting) because they have
insufficient influence on either the organization of labor or the
use of the profits from rice. Women in polygamous households
and landless women probably do not work as hard as other women, but even landless women have a greater work load than men when
household and child care responsibilities are taken into account.
Although this aspect of the project reflects a significant cost to women (increased agricultural workload) with little compensating gain in independent access to resources such as cash returns from rice, other aspects of the project reduced the drudgery of women's work. Time spent fetching water was reduced from an average of 1-1/2 hours per day to a half hour with access points now as close as the irrigation streams (Palmer, 1979, pp. 54-55). Fuel, too, which used to be gathered from nearby forests or prepared from cattle dung cakes, was now more likely to be purchased in the form of firewood or -oal. Although this new development reduced women's burdens, it created new pressures for cash to buy fuel. A careful social soundness analysis during the preliminary stages of project design that takes into account the intensity and
extensiveness of labor inputs of all household members could help to ensure that female workloads are reduced, if they are now excessive, or that they are increased only in exchange for some other valued resource and with appropriate social supports.
Indicators of Economic Wellbeing
7. Income in cash, kind, or trade in relation to cost of living. Household income can be measured by its amount, its security (extent of seasonal fluctuations, short-term unemployment, long-term prospects, etc.) and its source; economic wellbeing depends in part on the relation between income and those expenses required to maintain an adequate standard of living.
Not only the level of household income, but the question of who earns the income and in what proportion to the total, is crucial to understanding the domestic economy. Although the household is often considered the basic unit of production and consumption, in some societies the household economy is highly segmented. Women in many African societies, for example, are expected to provide for their own and their children's subsistence, while men's earnings go for larger cash outlays (land, cattle, marriage exchanges) and for their personal needs. Which household members earn direct economic returns to their labor (surplus production for sale or exchange), and which members engage in unpaid labor (production for family consumption)? Do women whose labor contributes to surplus production have direct access to (and control over) earnings, or are their husbands paid for women's labor? Who pays for what household expenses? Does a project increase women's unpaid labor but not their paid labor?
A review of the literature reveals a number of cases in which
development efforts have altered the distribution of incomes within the family in such a way as to leave women absolutely or relatively worse off than before. In some cases this loss results in the failure of the project to meet its production targets. A classic example comes from the reorganization of a marketing cooperative in the pyrethrum industry in Kenya. Whereas women had been growing the crop and selling the dried flowers directly, the new cooperative returned payments only to formal members, who were mostly men. The women became discouraged by the loss of personal incomes which they had formerly controlled; "rationally and realistically in the circumstances, their output fell" (Apthorpe, 1971, p. 73). The expansion of sugarcane cash cropping in northern Belize offered new employment opportunities for men but reduced local cultivation of corn, upon which the women had depended not only for food for the family and for exchange, but for chicken and pig feed. Pigs were the main independent source of income for women; the decline in pig production thus represented a significant loss of women's economic independence as well as a nutritional logs to the family (Stavrakis and Marshall, 1978).
Even where women's personal incomes are increased by access to new employment opportunities within or outside agriculture, the gains can be counteracted by increasing financial obligations. Kikuyu women in a coffee-growing region of Kenya who earned independent incomes in agricultural labor lamented the increasing tendency of their husbands to spend money on imported beer and hotel food, while failing to provide clothes and school fees for the children (Stamp, 1975-76, p. 28) Under some conditions, men may decide to work less hard as their women work more. Project planners clearly have little control over this aspect of
household decision making, but it is crucial that they understand the current pattern of control over earnings and expenditures within beneficiary households before introducing activities that may undermine women's position and perhaps subvert the project.
8. Access to credit. How has the project affected the supply and cost of credit or loans available to households in the target group? Are female as well as male family members eligible for loans? Are women-headed households eligible? Most projects designed to provide new sources of credit consider the household as a unit and the male as its head. Women are consequently denied independent access to credit that could increase their own productivity in agriculture, handicrafts, or other activities. In a rice project in Senegal, women were the main paddy growers but only men could obtain credit for agricultural inputs from the project. Because husbands were reluctant to go into debt on their wive' s behalf, the project had to be redesigned to reach women with credit directly before it could achieve its production goals (World Bank, 1978, pp. 26,28). In addition, new banking institutions -- even those for low-income rural populations -- may set collateral requirements that women are unable to meet. The People's Bank of Indonesia, for example, with branches throughout the country, makes low-interest loans to farmers against collateral such as land or cash crops (rice, maize, soybeans, sugar, cotton) but not fruit or vegetable crops or household equipment which are typically women's assets (Milone, 1978, pp. 107-112). Evidence from some projects suggests that innovative methods could be designed to reach this traditionally ineligible group of borrowers, with important productive consequences (Buvinic, Sebstad and Zeidenstein, 1979).
9. Land and water. This component of economic wellbeing is
measured by the amount and quality (productivity) of land legally owned; by the amount and quality of land available for use (leasing, sharecropping, tenancy, etc.); and by the security of land and water use rights.
Most projects involving land redistribution, such as new settlement schemes, are concerned with creating a more equitable distribution of resources between the landless or near landless and the landed classes. Less attention is paid to the distribution of land rights within the household, as well as between male-headed and female-headed households. Does the project expand or contract women's legal or traditional rights to the ownership or use of land? Does a scheme designed to secure tenants' rights, for example, address the question of what happens to women of tenant families when the husband dies? Do. land resettlement schemes provide women with their own plots, or legally recognize their joint ownership or use rights with their husbands? Or are titles granted to the male household head, undermining women's current access and inheritance rights! Does a project reduce landlessness among male-headed households but not among female-headed households?
A resettlement scheme in Nigeria, for example, distributed fivehectare plots to families for soybean, corn, and bean production, but did not provide for kitchen gardens in which women traditionally raised vegetables for family consumption and for the local market (Dulansey, 1977). And although some settlement schemes will allot land to womenheaded families (in the Rahad area of the Sudan, about 7 percent of resettled land to be farmed with cotton and groundnuts was allotted to
women who were heads of formerly nomadic families), it appears that in most voluntary resettlement schemes, women without husbands are usually excluded. 24/
10. Technology and technical assistance. Access to these resources increases productivity and the potential for higher economic returns, depending on their quality, appropriateness to local conditions, and frequency of use.
Which household members, and what classes of households, have the greatest access to labor-saving technology permitting them to reduce time or energy inputs and/or increase outputs? What are the distributional effects of the introduction of new technology? Who benefits, who is bypassed? Do planners consider the development of appropriate technology for domestic consumption as important as technology for the production of surplus in agriculture or industry?
The history of development efforts frequently points to negative impacts on women of technologies that widen rather than narrow the productivity and earnings gap between the sexes. A typical example is the introduction in many regions of the world of community milling machines. Although increasing the speed and reducing the cost of grain processing, they do not benefit poor landless women who earn their living by traditional hand-pounding methods. Not only are the new machines usually operated by men, but the women can no longer compete and their earnings fall. Similarly, new technologies are frequently applied to large-scale cash cropping but not to small-scale
1-4/ For recommendations on the role of women from the World Conference on Agrarian Reform and Rural Development, see Food and Agricultural Organization of the United Nations (1979).
food cropping, and to the work men do in the crop cycle (ploughing, perhaps harvesting) but not to women's work (planting, weeding).
The provision of technical assistance to beneficiaries may also bypass women unless specific steps are taken to ensure that they are reached. Does the project being evaluated attempt to contact women directly with technical assistance in agriculture or small industry, for example, or is most assistance extended to the male household head who is expected to pass on his knowledge to other household members? Does the assistance narrow or widen the gap in productivity across different classes of households?
Women's access to technical assistance is constrained by at least two factors: as workers they are concentrated in sectors of the economy such as subsistence agriculture, handicrafts, or small-scale marketing that are bypassed by most technical assistance programs; and as women they tend to be less "visible" within these sectors than are men. As a consequence, their productivity suffers from constraints on access to good quality raw materials, marketing assistance, advice on more effective methods of organizing production, knowledge about credit sources, management and bookkeeping assistance, and so on.
In many countries, agricultural extension programs have no female agents in some or all of the districts they serve. Male agents deliver services only to male farmers, ignoring women's contributions within these households as well as women farmers who head their own households. For instance, in an area of small-scale farming in Western Kenya in which 40 percent of a sample of farm managers (as distinct from owners) were women alone rather than women managing jointly with men, femalemanaged farms were significantly less likely to have been visited by
government extension agents, to include a household member who had visited a demonstration plot, to include a household member trained in short specialized programs of crops or farm technology, or to have applied for or received a loan (Staudt, 1978, p. 443; see also Staudt, 1975-76). Although the effect of lower service on agricultural productivity was mediated by a number of other factors, including a strong informal network of information exchange among women farmers, the benefits of technical assistance clearly accrued disproportionately to farm families with an adult male present. As a first step to correct these inequities, women must be trained in all aspects of technical assistance in order to reach women producers with appropriate information and services.
11. Other assets in relation to debts. These can be measured by the ownership or use rights of buildings and other capital goods (e.g., household goods, animals); amount of savings; amount of debts. How are assets distributed within the household? Who owns or uses them, and who has the right to sell or trade them, give them away, or bequeath them? In particular, what independent assets do women control? Does the project improve women's ability to accumulate assets, or
As with land, women's legal and traditional rights to assets are usually constricted; they may be positively or adversely affected by development efforts. Frequently a woman's only personal assets are moveable property such as jewelry, a few household items, and perhaps some small animals, all obtained through inheritance or dowry. Yet, small as they may be, these assets are of critical importance in determining women's sense of independence and self worth. Among the Tuareg
and Fulani herders of the Sahel, for example,
It is the obligation of the bride's family to send her to her
new home with a dowry consisting of household goods and animals
- usually 5 or 6 donkeys and 10 to 40 goats. Sometime after
marriage it is customary for a husband to give his wife a gift
of animals according to his means -- a few goats, one or two
camels (Cloud, 1978, p. 74).
Following the Sahelian drought, the lack of understanding among program over
administrators of the sexual division of control/resources seriously damaged the nomadic women's economic and social positions:
*** (among') both Fulani and Tuareg herders, one of the major
concerns expressed was that the government's program to reconstitute herds lost in the drought was replacing cattle only
for the men. Women's stock was not being replaced. This was crippling their social system -- animals were unavailable for dowry and bridewealth payments, women had lost their independent property. This was apparently the unintentional result
of the government program that issued a card to the head of
each family, and replaced animals only to the family head
In this case, a fairly simple shift in project policies and procedures would have corrected the distortion in the distribution of benefits and long-term social impacts.
Indicators of Social Wellbeing
12. Knowledge. This component includes skills specific to the performance of certain tasks, as well as more general knowledge about the world at large. It can be measured along several dimensions: literacy and numeracy; vocational skills; and a broader level of understanding of the interplay of socioeconomic and political forces in the household, the community, and the larger society.
How are literacy skills distributed within the household and across households of different classes? Are females able to take advantage of formal or informal educational programs to the same extent as males?
Failure to design incentives to ensure that girls as well as boys are able to attend school can severely undermine a project goal of expanding primary education, as in the Afghanistan project in which the quota of 15 percent of new school places reserved for girls could not be met "where it is not accepted culturally." The expansion of women's workload, whether in unpaid agricultural labor or in paid employment, sometimes results in daughters being kept home from school to help in the fields or at home. Project administrators should be prepared to counteract this tendency with alternative plans for labor allocation. With the altered rice cropping patterns of a land resettlement scheme, for example,
... girls are now performing more weeding and assist their
mothers in preparing meals for the exchange labor. They can also be seen working on the subsistence plots. Boys, on the
other hand, are now freer of labor commitments than previously.
This has a recognized effect on the relative attendance of
boys and girls in the schools (Palmer, 1979, p. 54).
The benefits of vocational skills training can also be unequally
distributed within and across households. Has the project opened up new opportunities for women as well as men, extending their vocational skills into areas formerly closed to them? Is the skills training likely to result in paid employment? How can constraints against female partic ipation be overcome? Many training programs funded by international agencies perpetuate sexual stereotypes. Typically, community extension programs train local women in tasks related to home maintenance and child care, or in handicrafts, failing to prepare them for employment in modernized agriculture, industry, services, sales, or clerical/ administrative and technical/professional occupations. Thus women's skills are largely lost to these sectors and the benefits of training accrue disproportionately to males.
one could also consider the impact of the project on women's general knowledge, or world view. In societies placing severe restrictions
on women's physical mobility, the limits of their social world are likely to be narrow and their social and political dependence on men is intensified, flow does a project affect their access to knowledge and experience of the outside world ? Are there deliberate efforts at "1consciniain among beneficiaries, that is, of stimulating a critical
understanding of the dynamics of househol d and community social structures and of possible strategies for change? Do these reach women as well as men?
13. Power. The capacity of directing one's own and others'
behavior can be considered as the sum of separate elements: autonomy, as measured by the degree of self-reliance, belief in the efficacy of one's actions, freedom from coercion by others, and freedom of physical movement; participation in household and community decision making, including an ability to produce desired outcomes; and mobilization for group action, measured by the number and strength of linkages among individuals and groups, the degree of shared self-interest or group consciousness, and the capacity for effective group action.
Women's autonomy in the household and the community differs widely across and within societies according to social class, ethnic or tribal background, age, and other social factors. The question here is
whether development projects expand or constrict women's capacity for autonomous action in either absolute terms or relative to men. An income-generating program can increase women's social and economic independence, while another type of program may undermine it. A family planning project that requires the husband's consent for his wife's
sterilization or abortion, for instance, greatly reduces her capacity for autonomous action. Depending on the type of project, investigators could select appropriate questions from among the following possibilities: How does the project affect the pattern of decision making within the household about important choices such as the timing and number of births, the education of children, the timing of children's marriages and the choice of a mate, what to produce for sale or exchange, how to market it, when to plant and harvest, whether to buy or sell land, and so on? Similarly, what role do women play in the formal and informal decision-making processes in the community at large? In what proportions are they foundin formal governing bodies, and with what effect? What social class do they represent?
Knowledge of women's role in household and community decision mak ing should form a central ingredient of development planning (see Mickelwait, Riegelman, and Sweet, 1976), both at the point of baseline data collection to determine project design, and at the point of social impact assessment. Projects can undermine women's current role in decision making, or fail to take advantage of structural opportunities to expand it. In the large Aiuul Dairy Cooperative in India, for instance, the majority of shareholders are low-income male household heads (owners of one or two milch animals, usually buffalo) who attend meetings, form the board of directors of village societies, and receive the shares of yearly profits (Dixon, 1978, pp. 50-56). Women, who typically care for and milk the animals, do receive daily cash payments when they take the milk to village collection centers. Although this is a valuable source of independence and prestige, women are denied the opportunity
for cooperative leadership roles that serve as a training ground for higher level political office in the community and district. Changes in membership rules requiring each household to enroll a female as well as male member, and a quota of half the places on boards of directors reserved for women, could help to correct this inequity.
The third dimension of power refers to the mobilization of groups for collective action. How do males and females in th e household differ in their access to informal and formal social networks based on kinship, landholding status, occupation, residential location, caste or ethnic or tribal identity, religious affiliation, personal friendship, and other bases of identity and loyality? How does membership in such groups differ across social classes and other categories? To what extent does the development project facilitate or impede the capacity of women to organize formal or intormal groups engaged in collective action?
At the most basic level, failure of planners to recognize the importance of indigenous patterns of group activity can prevent the completion of a project. Women in one Indian community subverted a scheme for piping water into each housing compound because it would deny them their only legitimate opportunity for visiting together around the village well. In Nepal, failure to organize workers into a viable producers cooperative doomed some women who had gone through a training program in machine sewing to non-competitive hand sewing or inactivity because they could not individually afford to buy sewing machines (Dixon, 1978, pp. 155-6).
Even assistance directed specifically to the support of woe's organizations can have adverse distributional effects. The evaluation of one program to train middle- and upper-class women volunteers in
leadership skills and organizational development concluded that "there are more benefits to volunteers than to poor people," thus widening the resource gap between the classes. Project designers will need to pay special attention to the means by which women denied access to most material and social resources can be mobilized to form an independent power base for articulating their needs and working toward collective solutions.
14. Prestige. Prestige refers to the honor or esteem accruing to persons as individuals or groups members, of which both subjective and objective aspects are important. Self-esteem can be measured by the person's subjective belief in his or her own value as a person; the esteem of others is measured by the degree to which a person or group is "objectively" valued by other household or community members.
Who within the household, and in the community, appears to have the highest levels of self-esteem? Do women value their own contributions as highly as men, or do they tend to be self-deprecating? Preliminary talks with poor women by one community organizer elicited a typical response: "We have no special skills; nothing we do is worth selling. -/ How does the project affect women's feelings about themselves?
Although most development efforts are likely to increase women's pride and self-esteem, especially if women are defined as primary beneficiaries of training or income-generating projects, situations such as the failure of a handicraft center to find a steady market for
21Interview with Sister Michael of Holy Cross, Dacca, Bangladesh, Feb. 1976.
its goods, or the arrogance of some extension workers, can have the opposite effect. Young girls in a newly opened school who are told repeatedly by the male teacher that they are not as smart as boys quickly become discouraged from learning and drop out, a self-fulfilling prophesy. The loss of self-esteem can act as a major impediment to future risktaking behavior.
Apart from women's feelings about themselves, do other members of the household and community value their contributions more highly? Of course the two are closely interrelated. since self-esteem is so dependent on the esteem of others. Vocational training and incomegenerating projects for women should -- at least in the long run -visibly raise their status within the household and community. But some jobs are viewed as "status-depressing" rather than "status-enhancing" even if they provide badly needed cash incomes, especially when the jobs expose women to possibly dishonoring contact with men. In one project in Bangladesh, although the young women themselves were extremely proud of their skills and ability to earn money as government agents visiting moments cooperatives throughout the district, conservative residents disparaged them loudly for their immorality in breaking purdah restrictions (Dixon, 1978, pp. 47, 162-3). In the absence of strong program efforts to counter such criticism, female workers and their families were under considerable personal strain. Other families, fearing their daughters would be unmarriageable if they engaged in such work, refused to permit them employment.
The shift in northern Belize from corn production to sugarcane
has had some interesting implications not only in nutritional standards
and women's ability to earn independent incomes from raising chickens and pigs (as discussed previously), but also in the sources of prestige:
Because food production has always been woven socially and
ideologically into the culture, the change in production
patterns has ramifications throughout society. One of the
most significant has been the removal of the source of male prestige from the domestic household activities of producing
good food and good children and the creation of a new "public"
sphere of activity which is western and modern in structure,
characterized by material goods.
The women's activities are relegated almost totally to [the]
domestic sphere, whereas, the acquisition of prestige and
status, once accessible to both men and women, is now largely
relegated to the public sphere where women cannot compete.
*** The women were left with domestic activities of devalued
social status and responsibility (Stavrakis and Marshall, 1978,
This case clearly cries out for some strong programming to create new income-generating opportunities for women that will increase their social value and prestige as well as their economic contribution.
Although this discussion of the differential effects of projects on males and females may appear to have dwelled unnecessarily on the negative, it illustrates how even the best intentioned projects can have unanticipated deleterious consequences if women's social and economic roles in the household and community are not fully understood. The more positive side is that projects can almost always be designed to achieve their goals -- expanded agricultural output, for instance,
-- in a way that maximizes women's access to resources such as control over the economic returns to their labor, technical assistance to increase their productivity, or active participation in community decision-making bodies, while minimizing the costs. Palmer's analysis of a land resettlement scheme suggests an alternative design for meeting production goals and satisfying basic human needs of the entire settlement
population that allocates productive resources between men and women more equitably at the outset, resulting in a cumulative sequence of more beneficial consequences (Palmer, 1979, pp. 78-80). The actual and alternative strategies she outlines, which are presented in Table 4, are applicable in principle to a wide range of integrated development projects where such comprehensive decisions are made.
Some tradeoffs are probably inevitable in any development project; one cannot usually move effectively on all fronts at once. The major issue here is to ensure that the costs are not paid disproportionally by those least able to pay them.
TABLE 4. ACTUAL AND ALTERNATIVE DESIGNS FOR AN INTEGRATED RURAL DEVELOPMENT PROJECT
Ehvirotent: irrigated black soil good for rice growing,
subsistence crops possible on both red
and black soils.
ACITUAL Tradltional sexual division of labor: women work subsis- ALTNATIVE
tence crops, wanen and men work rice crop.
soceialte statusoldofn's aner
a ale ownership of rice land and Tradiiol land inheritance: daughters inherit one-half sipt o ri and on
asi w nnnerhorie lad than leel needs at tosh ofie lael, nce ontsx%
membership in Fanrmers' Association; i ship; with wen continuing tradiw cm in chrge of loaned subsis- w ttinal role in o*ed subsistence
tegncep land with no access to crop land; both sexes have direct
credit, technical advice membership of Parners' Association
I I II
noinres higher yields equal control HMr
in yields of subsistence plots of higher on
subsistence with more cash income role of
crops surplus traded from rice wmen in
er cash Association
oe of / little change
man of trading surplusI
of subsistence crop
rautritimnal more rational legal and
wonen have weak and benefits sexual division Boda status
derived position in -aximized of labor of wanen
Residents' Assoclation enhanced
ineffective channels of chil laore
cc jreaic fr ctespecially especially
to a n corrctiem n girls, five landless
for school- wanen, a
-of e ing ployed
ntial to vulnerable legal and women's and I
sy all social status of wcmen infants' better income
c needs i m- distribution
ling pvdbetween landed lihe wasiclead and landless
uneven gains in the satisfaction of enhanced ability to meet basic humn
basic human needs at the family level, needs at the family level, Incentives
no incentives to limit family size, Sources Palmer (1979), p. xi. to limit family size to consolidate
no improvenent and possibly sane economic gains transferable to the
decline in women's roles between next generation, maintenance and
generations possibly sane enhancement of women's
roles and status
IV. METHODOLOGICAL ISSUES IN COMPILING COMPARATIVE DATA
If an evaluation is to assess the degree of absolute and relative change in people's access to material and social resources, baseline data should reveal their access to resources before the project began. Such information should have been collected or reviewed in the project identification and design stages, but, as we have seen, preliminary social soundness analyses are frequently based on inadequate data and attempted only after major project decisions have been made.
These inadequacies should not subvert the entire process of social impact assessment, however, for evaluations of change can be made in the absence of good baseline data if we are willing to rely on people's statements of how their lives differ now from the way things were before the project began, and especially if we can interview people in experimentala" and "control" settings (see Staudt, 1978, for example). Since most mid-stream evaluations take place within two or three years of the project's inception, memory loss should not be great. Although people's perceptions of change may well differ from whatever objective indicators one could collect, and may be colored by a variety of intervening events, one could argue that it is the perceptions of participants more than the objective indicators that are likely to determine the ultimate social success or failure of the project.
Sources of Data on the Participation of Women
in Project Decision Making, Access Project to Benefits,
and Immediate or Long-term Consequences
General baseline information on access to many of the resources discussed in Section III, particularly those relating to physical and
economic wellbeing, is available from a variety of sources. National population censuses and more detailed socioeconomic surveys will usually permit broad regional breakdowns with aggregate data on housing conditions, employment status, individual income, household income, household size, land ownership and use, literacy, completed years of schooling of adults, current school enrollment rates of school-age youth, and other variables, although they are frequently not disaggregated by sex, nor by other relevant social categories such as age, marital status, race, and ethnic or tribal identity that would permit an assessment of distributional effects.-6
Special interest surveys conducted in the region can offer
additional data on such factors as nutritional status or the prevalance of particular diseases, reproductive behavior, labor conditions and wage rates, and other items of interest to the evaluator. Many will permit differentiation by sex, as will most public records such as vital statistics, school attendance registers, hospital or clinic records, membership lists of cooperatives, or bank loan statements. Ethnographic studies are also a rich source of insights even if not conducted in the precise area of project location.
For mid-stream evaluations of the sexual distribution of project benefits, however, we need additional sources. The methods of data collection suggested below are primarily informal and qualitative, even impressionistic, on the assumption that outside evaluators of most projects can spend only a few days or weeks in the area. (If project
- -On the collection of aggregate data on women in development, see Biocentric (1978).
personnel and representatives of client groups continuously monitor the dispersal of goods and services to ensure that females have direct access with males, then much of the evaluator's work will have already been done.) The assumption thus precludes the use of large, systematic
random samples of households with standardized questionnaires require ng extensive coding and computer analysis. It suggests, instead, more flexible techniques of rapid appraisal. 27/
Interviews with project workers, personnel in related institutions, and community leaders.
In interviewing women and men at all levels of project administration, evaluators should pay special attention to those involved in the direct delivery of services to clients on a face-to-face basis (agricultural extension agents, village "motivators," vocational teachers, construction supervisors, etc.). Are they actually reaching women and in what capacity? A special meeting of project workers could be called to address collectively questions of how the project is working, who are its primary beneficiaries, what are the major constraints to women's participation, and how these might be overcome. Key women and men in related local institutions (schools, cooperatives, and marketplace, other service institutions) as well as community leaders can provide, individually or in group meetings, contrasting political perspectives on the project as a whole and the role of women. Although their reactions may be exaggerated in either a positive or negative direction because
A workshop on Rapid Rural Appraisal held at the Institute of
Development Studies in Sussex, England, included several papers that were particularly helpful; see Jackson, Mandal, and Carruthers (1978); Richards (1978); Wood (1978).
of personal sta kes in the outcome, their opinions as influential community members are crucial to the success of the project.
Direct observation of project operations and settings
Direct observation of training sessions for project personnel,
cooperative meetings, health clinics, village literacy classes, handicraft production centers, and other activities can reveal important information about the number and identity of clients (including the proportions female), the nature of interaction between clients and project personnel and among one another, adequacy of the facilities, and other items that may not be revealed in project documents. At a community meeting, for example, even though women are present, they may rarely if ever speak out, or their opinions may be ignored. This information is helpful in understanding the extent of women's actual rather than reported participation in project decision making. The outside observer may be struck by other visual discrepancies that documents cannot reveal: that few women'are actually attending adult literacy classes, for example, even if they are formally enrolled; that women working on construction sites must carry their infants on their backs and leave other children unattended; that girls who should be in school can be seen carrying water and collecting fuel throughout the day.
Visits to the site of service delivery (farms, small industries, health clinics, etc.) offer a natural setting for interviewing clients. The limits of this approach are obvious: not only is the on-site availabilityt" sample frequently unrepresentative of all beneficiaries,
but it excludes those who are not direct beneficiaries, including other members of clients' households.
But the advantages are also compelling. People are engaged in activities related to project purposes, which adds a concreteness and timeliness to the questions. An interested observer can usually stimulate lively debate about what is good and bad about the project and how the everyday lives of clients and their families have changed, allowing plenty of room for participants to air their grievances and tell stories in their ownL termsL.-8
Although the more vocal members are likely to dominate the discussion, a skilled interviewer should be able to get some idea of the degree of consensus or disagreement. Organizers of the audio-cassette listening forum project in Tanzania trained leaders to survey respondents in group discussions, finding the method preferable to individual questionnaires that elicited conflicting information:
If the ... meetings are conducted according to the methodology
developed in the training sessions, it is very difficult to
receive unreliable information. With an open climate for
discussion, untrue information is corrected by the participants.
A number of lively discussions centering around differing
opinions about the village practices took place and the discussions and consensus conclusions verified the belief in the
information that was finally recorded. The use of group
discussions for collecting information also permits any outsiders to question immediately what they might see as discrepancies in the given information (Stanley, n.d., p. 31).
Perhaps most important, the group setting can allay suspicion
about the motives of the interviewer, permitting those who wish to say nothing to remain quiet, yet paving the way for later interviews with
--Note, for example, Perdita Huston's interviews with village women in many developing countries (Huston, 1979) and Kusum Nair's observations of village communities in India (Nair, 1961).
individual participants and their families. This is relevant not only to mid-stream evaluations but to the preliminary project identification and design stages as well, especially when the topic is a sensitive one:
Gathering a group of farmers together at the start of an
enquiry to ask them collectively to explain, e.g., the
ways in which men can get land, to give local terms for
land quality/soil etc. can be very productive and is fast.
This is not a threat to anyone particular individual because one is asking the collectivity. To ask X about land quality
immediately raises the suspicion "Is he going to take away
my land" which does not occur directly with a group gathering together as men who are knowledgeable about.farming (Jackson,
Manual, and Carruthers, 1978).
The group interviews need not (should not) be limited to on-site samples. It would be extremely helpful to interview groups of people representing different sectors of the community, such as persons in the target group who have not participated in the project (non-members of agricultural marketing cooperatives, for example) and persons assumed to be indirect beneficiaries (women in families whose male heads are members). Frequently they can be interviewed collectively in their places of work, gathering in the village square or at the well, or in their neighborhoods. A female interviewer will of course have far easier access to groups of women in their homes, especially in areas practicing female seclusion. These group discussions with persons who are not direct beneficiaries can elicit information about their awareness of project goals, the spread of project benefits, and the ways in which project participants are viewed by others in the community.
Selecting two or three key informants from female project personnel who work directly with clients and from women in the client group is one
short-cut, along with the group discussions, to collecting information on women's access to project benefits. The researcher can work with informants throughout the evaluation, checking back for clarification of issues, requesting additional information that is inaccessible to the outsider. Knowledgeable informants can describe typical patterns of food and fuel consumption, treatment of illnesses, access to land, group affiliation, the sexual division of labor, time use according to season, and even patterns of household decision making. These findings can be used as a basis for additional questioning of samples of participants and nonparticipants that would focus on exceptions to the rule, rather than on repetitive and detailed questioning of everyone on the same topic.
Interviews with samples of households
Ultimately, the evaluator of a large project will need a systematic sample of participants and nonparticipants to interview on the absolute iand relative changes in their access to resources. Sample size and design Depend on a number of factors such as the type of project, its scope, and the time available. In almost all cases, it will be necessary to forego large-scale random sampling'procedures and standardized questionnaires that permit greater statistical elegance in favor of small-scale, less random samples with more flexible interviewing techiniques.
If an entire locality is defined as the target area, e.g., in a rural roads project, then a stratified sample of households within the project area is preferable. In the likely absence of an adequate sampling frame, selection may have to rely heavily or. directly observable
traits such as housing conditions and more visible socioeconomic characteristics of residents, and on the advice of informants. Special efforts should be made to include women-headed households. Although the non-randomness of the sample will cause some problems in interpretation, we are interested essentially in subjectively perceived matters of process and change rather than in establishing firm parameters of population characteristics.
If classes of individuals or households are defined as beneficiaries, a stratified sample of participant households should be approximately matched with a sample of nonparticipants of similar socioeconomic status. Women in each household should be interviewed regardless of whether they are defined as primary beneficiaries: ideally, a male adult (preferably the husband) should be interviewed as well.
An interviewer could not possibly ask all of the questions stated or implied in the discussion of access to resources in Section III. The approach will have to be flexible. Wives can answer questions about changes in their husbands' access to resources, investments in children, food consumption, literacy skills and vocational training, the sexual division of labor, and household decision making, that do not also need to be asked of men. Some information on typical behavior gained from group discussions and from informants can be passed over, or briefly clarified. Specific questions relating to the direct access of beneficiaries to project services, however (visits by agricultural extension agents, etc.) will have to be asked of everyone.
Questions considered controversial or sensitive can usually be dropped with little loss of information by asking them of informants instead, or by making educated guesses. Detailed information on
household income and expenses, on time use, and on other complex issues could be collected from a small subsample of households rather than from the entire sample. In any case, the value of household interviews is to promote discussion as well as collect facts, so that respondents have a chance to raise new points and to challenge the assumptions of the researcher.
Although ideally the interviews should be conducted in private, this practice often gives rise to suspicion or is simply not feasible. Most often the interviews will include not only'all members of the household but inquisitive neighbors as well. This means that more articulate (more powerful) members are likely to dominate the discussions. If women do not speak out, the interviewer should return at another time when the men are absent so that women -- individually or in groups -have a chance to tell their story.
As in any project evaluation, the general approach described in this section is vulnerable to serious methodological problems. First, is the project responsible for changes, or are they due to other factors? In the real world of development it is hardly ever possible to set up a classic research design with experimental and control groups. Even if we could match households or communities on key variables, there is always too much else going on socially, economically, and politically even to approach the conditions of centers paribus. Participants may attribute changes in their lives to the effects of the project when the two are related only coincidentally. It is the task of the evaluator, then, to try to sort out the causal sequence of events and to suggest changes in implementation procedures in the absence of clear guidelines for doing so.
Second, evaluating a project within a year or two of its
inception --- although valuable if it is early enough to permit corrective feedback -- may be too soon to pick up genuine change in important attributes such as people's capacity to organize for collective action. Or, early evaluations may pick up indicators of change that appear promising -- increases in household income as a result of particular inputs -- that are not maintained in the long run. The researcher must be aware of the implications of short-term vs. long-term effects, and of the possibility that the timing of the evaluation can affect the outcome in fundamental ways. This caveat is especially relevant to the assessment of new development projects for women.
A Comparative Approach for Evaluating the Effectiveness of Different Project Strategies
In order to answer the question of "what works and what does not in specific socioeconomic environments," we need a method of classifying projects according to their purposes (what works in facilitating access of low-income urban residents to employment, for example); a method of classifying the organizational strategy by which the recipient agency, or service provider, hopes to achieve its objectives; and a method of classifying the socioeconomic environment in which it functions. To these, we add an additional specification: "What works in reaching women? What strategies appear to be most effective in, at the least, protecting women from further disadvantage, and, at the most, offering genuine improvement in their physical, economic and social wellbeing? With projects classified along these different dimensions, we can begin to work out an approach to comparing different types for their impact on women.
What are the project purposes?
In Table 1 we proposed a classification of projects according to purposes into three large categories: those designed to increase production, employment, or incomes within the sectors of food and agricultural production and distribution, nonagricultural production and services, and mixed activities; those designed to improve welfare within the sectors of formal or nonformal education, health, or community organization; and integrated development programs incorporating all sectors. This preliminary classification suggests at least two questions regarding what works. First, do broader, integrated programs or more concentrated sector-specific programs appear to be most effective in bringing about concrete changes in women's lives? Second, which sectors have the greatest spin-off effects? Critics have challenged the practice, common in the past, of subsuming most women's programming under the rubric of "family welfare" while ignoring their roles as producers (e.g., Germaine, 1976-77). Will expanding women's earning capacity directly, for example, have more pronounced long-term effects than subsidizingg food, medical care, family planning, education, or housing, by creating an effective demand for these? What are women's own priorities? "Most rural women express economic needs before others," argues one critic. "Why not build on identified self-interest?" (Bruce, 1977, p. 42).
What is the most effective organizational strategy for achieving project purposes?
Given a particular objective, we should be able to weigh the
relative effectiveness of different project designs in reaching beneficiaries under different socioeconomic conditions. The introduction of
these two new sets of variables greatly complicates the comparative strategy, however, because the criteria for classifying both project designs and environmental settings are virtually limitless.
One such attempt at classifying major project characteristics affecting client participation (and the distribution of consequent benefits) is presented in Table 5. Most of the examples of the effects havedirect relevance to women. Projects requiring technical skills such as complex accounting or marketing practices (characteristic 1) that exclude less educated persons from leadership roles are particularly likely to exclude poor women. Those requiring resources such as land, collateral, or share capital (2 and 6) will probably have similar effects unless innovative procedures are designed to reach individuals or groups usually considered ineligible (encourage collective leasing of land, for example and forego usual collateral requirements for groups of landless women).
Women are less likely to increase agricultural productivity if benefits accrue to their husbands as cooperative members and not themselves (3), or to risk new agricultural techniques if they are primarily responsible for growing food for their families (4). If administrative inaccessibility discourages participation in general (9), female clients can be especially discouraged when male administrators pass them by. If women are mobilized into groups (e.g., handicraft producer's cooperatives) they may gain greater control over administrative decision making. The central question here is, Which type of organization strategy within a given project purpose will ensure that low-income women also benefit?