N o v e m b e r1 9 9 8

nn uu mm bb ee rr 88

Gender

Providing the poor with financial servicesincreases their income and productivity,thereby reducing poverty. But in many coun-tries traditional financial institutions—whichrequire physical collateral worth three to fourtimes the amount of their loans—have failedto reach the poor. Microcredit programs havebeen developed to fill this gap. Many of theseprograms use social mechanisms, such asgroup-based lending, to reach the poor andother groups, especially women, who lackaccess to traditional financial institutions.With increasing assistance from the WorldBank and other donors, microfinance isemerging as a powerful tool for reducingpoverty and improving access to financial ser-vices in low-income countries.

Although both men and women benefitfrom financial services, women generally haveless access to such services. This is partlydue to women’s limited social and economicmobility and their lack of physical collat-eral. Women in developing countries are ata particular disadvantage because of poverty.With their program design and focus onthe poor and women, microcredit programshave helped women and the poor becomeself-employed and increase their income.When wage employment is unavailable ordifficult for women to access, credit canfinance income-generating activities that helpwomen become self-employed. Credit alsosmoothes consumption fluctuations due toseasonality in income and employment. More

important, it increases women’s options foroptimizing the use of available resources,including time, across activities.

This note analyzes the outreach and impacton the poor (defined as members of house-holds owning less than half an acre of land)and women of three well-known microcre-dit programs in Bangladesh: the GrameenBank, Bangladesh Rural AdvancementCommittee (BRAC), and Bangladesh RuralDevelopment Board’s Rural DevelopmentProject 12 (RD-12). The findings are drawnfrom an in-depth analysis of household andcommunity survey data collected in 1991–92by the World Bank and the BangladeshInstitute of Development Studies.

Bangladesh’s programsBangladesh is a leader among low-incomecountries offering microcredit. Using a group-based approach, the country’s small-scalemicrocredit programs provide more creditin rural areas than do traditional financialinstitutions. The Grameen Bank, founded asa project in 1976 and transformed into a bankin 1983, is the best-known microcredit pro-gram. The bank believes that the most press-ing need of the poor is credit, which they canuse to create and expand self-employmentopportunities. Its group-based lendingscheme has several attractive features: • The assetless poor find it easier to access

institutional credit through groups andto work with groups that they select.

Credit increases

women’s options

for optimizing the

use of available

resources

Using microcredit to advancewomenThis year’s floods have endangered the substantial achievements of Bangladesh’smicrocredit programs. Recent studies nevertheless bring out many importantlessons from such programs. Bangladesh shows that microcredit programscan raise living standards and alleviate poverty—particularly for women andtheir households.

PREMnotesThe World Bank

from the development economics vice presidency and poverty reduction and economic management network

• Loan contracts are enforced throughgroup pressure and monitoring; as a resultloan defaults are low.

• Working across communities and groupspools resources and diversifies risk.

• Savings mobilization is an integral partof group-based lending.The group-based lending approach has

guided more than 750 nongovernmentalorganizations (NGOs) operating small-scalemicrocredit programs in Bangladesh. Twolarge microcredit programs (BRAC and RD-12) also rely on group-based lending. BRACis an NGO established in 1972, while RD-12 is a public program formed in 1988.BRAC believes that the poor primarily needto develop skills and secure other organi-zational inputs before receiving credit. RD-12 uses a two-tier cooperative structure basedon the credit delivery model of the GrameenBank, though it also emphasizes skill-basedtraining for the poor. Despite their differ-ences, all three organizations have learnedfrom one another.

Microcredit outreachThe advent of microfinance institutions hasbeen a blessing for Bangladesh’s poor—especially women. Although credit is avail-able from informal lenders, the poor findit difficult to generate significant incomefrom such loans because interest rates arehigh. Moreover, group-based lending pro-vides economic security to the poor, whootherwise have to depend on local patron-client relationships for jobs. By forming analliance with a local faction, the poor oftenget secure employment in exchange for their

support in local elections and communityaffairs. But such dependence is coercive anddifficult to break.

Group-based lending provides an alter-native. Microcredit programs mobilize thepoor into groups and provide them with con-sciousness training, offering them a way torealize their potential. In addition, groupmonitoring helps ensure that loans are repaid.

The group mechanism has an addedadvantage for women in Bangladesh. Forsocial and other reasons, women find it dif-ficult to be active in the local labor market.Full-time involvement in nonmarket house-hold production discourages women fromseeking wage employment outside the home.In addition, women are less able than mento influence the allocation of householdresources, including their own time. Group-based credit schemes enable poor ruralwomen to be self-employed at home with-out forgoing their nonmarket householdproduction. Groups also provide a forumfor women to interact and to develop bondsthat provide economic security outside tra-ditional household settings.

Thus microcredit programs are moreattractive to poor women than to poor men.The programs have learned that womenare better credit risks than men. Moreover,reaching out to more women than menpromotes social development and genderequality. As a result Bangladesh’s micro-credit programs target women, and mem-bership among women is high in all threeprograms

By 1994 the Grameen Bank had more than2 million members—94 percent of themwomen (table 1). Nearly 88 percent of BRAC’smembers were women, as were 70 percentof RD-12’s members. Women accounted foran overwhelming majority of active borrow-ers and received more than 80 percent ofannual disbursements in all three programs.Most impressive, women’s loan repaymentrates were well above 90 percent.

Benefits for women and menIdentifying a program’s benefits requires aparticipant-level impact analysis that mea-sures effects on individual and household

The advent

of microfinance

institutions has

been a blessing

for Bangladesh’s

poor—especially

women

Table 1 Female participation in the Grameen Bank, BRAC, and RD-12, 1994

GrameenIndicator Bank BRAC RD-12

Total members (thousands) 2,015.1 1,036.3 452.0Women members (% of total) 93.9 87.7 69.7

Total borrowers (thousands) 1,860.7 739.6 316.0Women borrowers (% of total) 94.2 71.6 77.1

Total annual disbursement (millions of taka) 15,395.3 2,210.0 706.5Disbursement to women (% of total) 91.9 87.0 83.8

Women’s loan recovery rate (%) 99.4 98.0 95.7

Source: Khandker 1998.

welfare, such as per capita consumption,assets, schooling of children, and so on. Whilestudying these effects is important for deter-mining whether a program helps reducepoverty, for several reasons quantifying theeffects of each program is problematic, espe-cially when conditions make it impossibleto conduct an experiment where researchersdecide who receives credit and who does not.

First, programs set their objectives at thestart, and hence are not randomly placed.Thus, without controlling for program place-ment endogeneity, it is difficult to deter-mine a program’s effects. Second,participation in a program is self-selective,so comparing outcomes between partici-pants and nonparticipants does not indi-cate any impact. Finally, when both men andwomen are allowed to participate, house-holds self-select whether a man or a womanjoins the program. Without knowing this,it is difficult to know whether women ormen benefit from a program, or whetherthe gender of a participant matters.

Such problems are endemic to any pro-gram evaluation using cross-sectional data.The study in Bangladesh resolved theseissues using a quasi-experimental surveydesign. Households were chosen at randomfrom both program and control villages,from both target and nontarget groups, andfrom villages with only men’s groups andonly women’s groups. The main findings ofthe research are striking: in many cases theimpact of credit given to women was foundto be much greater than the impact of creditgiven to men.

The household and intrahouseholdeffects of male and female borrowing underthe three programs are shown in table 2.The most important effect of borrowingfrom a microcredit program is its impact onhousehold spending per capita. In all threeprograms a 10 percent increase in women’sborrowing increased household spendingper capita by about 0.40 percent.

The impact on household consumptionof borrowing by women was about twice asgreat as for borrowing by men. Women’s bor-rowing also seems to improve children’s wel-fare much more than men’s borrowing.Except for BRAC’s impact on girls’ schoolenrollment, women’s borrowing improvedboys’ and girls’ enrollments more than men’sborrowing. And in contrast to men’s credit,women’s credit had a large and statisticallysignificant impact on the nutritional well-being of boys and girls. Borrowing from amicrocredit program also increases house-hold net worth. However, the impact of bor-rowing on household net worth is higher forcredit to men than for credit to women. Thatis, men invest more than women in physicalcapital.

Microcredit also influences householdreproductive behavior. Credit to womenincreases fertility (except under theGrameen Bank program) and reduces con-traceptive use. In contrast, credit to menreduces fertility (except under the BRACprogram) and increases contraceptive use.This finding suggests that borrowing mayencourage female borrowers to have morechildren, since women’s self-employment

In many cases

credit given to

women has a

greater impact than

credit given to men

Table 2 Impacts of microcredit borrowing on household outcomes(percentage change for a 10 percent increase in borrowing)

Grameen Bank BRAC RD-12Male Female Male Female Male Female

Household indicator borrowing borrowing borrowing borrowing borrowing borrowing

Per capita spending 0.18 0.43 0.19 0.39 0.23 0.40Net worth 0.15 0.14 0.20 0.09 0.22 0.02Boys’ school enrollment 0.07 0.61 –0.08 –0.03 0.29 0.79Girls’ school enrollment 0.30 0.47 0.24 0.12 0.07 0.23Boys’ height for age –2.98 14.19 –2.98 14.19 –2.98 14.19Girls’ height for age –4.92 11.63 –4.92 11.63 –4.92 11.63Recent fertility –0.74 –0.35 0.54 0.79 –0.74 0.50Contraceptive use 4.25 –0.91 0.40 –0.74 0.84 –1.16

Source: Khandker 1998.

and nonmarket household production maybe jointly produced. Still, family planningprograms should not be targeted exclusivelyto women; men also need to be targeted foreffective population control.

Finally, microcredit smoothes householdconsumption and labor supply. Householdsthat experience more seasonality in con-sumption are more likely to participate inthese programs, indicating that the pro-grams are well-targeted. Interestingly, con-sumption smoothing is higher for women’scredit than for men’s. On the other hand,women’s access to microcredit seems toreduce men’s heavy workload in the peakseason.

While the household impact of women’sborrowing is important, it is equally impor-tant to explore how women benefit fromtheir participation in microcredit pro-grams. Women benefit from microcreditprograms through the cash income gen-erated by self-employment and the assetsthey acquire in the process. A 10 percentincrease in credit to women increases theirparticipation in the labor force by 1.0 per-cent for the Grameen Bank, 0.7 percentfor BRAC, and 0.8 percent for RD-12 (table3). The greatest impact, however, is onwomen’s nonland assets, which—with a 10percent increase in borrowing—increaseby 2.0 percent for the Grameen Bank, 1.2percent for BRAC, and 2.2 percent for RD-12. In contrast, credit to men does not have

any significant impact on women’s laborsupply or women’s holding of nonlandassets.

ConclusionAn increase in assets and the generation ofcash income may empower women withinthe household, increasing consumption forthem and their children and contributingto other measures of welfare (such as school-ing). One measure of this increased empow-erment is the fact that women’s borrowinghas an independent effect on the allocationof household resources, with men investingmore in physical capital and women invest-ing more in human capital.

These results represent only the short-run effects of microcredit programs.Whether these effects are sustainable overtime remains to be seen. But even if theyare sustainable, microfinance should not beconsidered a panacea for reducing povertyor reaching poor women. Only people whohave the ability to be self-employed can bor-row, and only about 40 percent of the eli-gible poor households in Bangladeshparticipate in microcredit programs.Entrepreneurial ability is unevenly appor-tioned among any population—includingwomen and poor people. Other programs,such as wage employment schemes, arerequired for those who cannot make pro-ductive use of microcredit.

Further readingKhandker, Shahidur R. 1998. Fighting Poverty

with Microcredit: Experience in Bangladesh.New York: Oxford University Press forthe World Bank..

This note was written by Shahidur R. Khandker(Senior Economist, PRMGE/DECRG).

Microfinance

should not

be considered

a panacea

for reducing

poverty or reaching

poor women

This note series is intended to summarize good practice and key policy findings onPREM-related topics. PREMnotes are distributed widely to Bank staff and are alsoavailable on the PREM website (http://prem). If you are interested in writing aPREMnote, email your idea to Asieh Kehyari. For additional copies of this PREMnoteplease contact the PREM Advisory Service at 87736.

Prepared for World Bank staff

Table 3 Impacts of women’s borrowing on women(percentage change for a 10 percent increase in borrowing)

Grameen Indicator Bank BRAC RD-12

Women’s labor hours 1.04 0.72 0.77Women’s nonland assets 1.99 1.15 2.17

Source: Khandker 1998.