A Report Compiled by Palli Karma-Sahayak Foundation (PKSF)

A : Findings of PKSF-MES Impact Assessment Studies by BIDS

1. Impact of MC on Income and Povert

The major objective of MC is to create income among poor households and thereby alleviate poverty. In this respect the question can be asked at two levels, first, whether MC leads to an improvement in income and second whether the increase in income is sufficient to lift them above the poverty level. There is also a third aspect related to the second, that is whether poverty alleviation takes place on a sustained basis. Though most IAS examines the impact of MC on income, the analyses concentrated on mainly the first question. It should be mentioned that most studies consider the total impact of growth income and do not examine whether a redistribution of credit fund has a redistributive impact on poverty and income of households.

Hossain (1985) reported two types of comparison to examine the impact of GB credit on income among borrowing households. On the basis of the borrowers' recall, it was obtained that in January 1980, per capita real income was Tk. 1037 which increased to about Tk 1374 by May-June 1982, i.e. it increased by about 32 per cent over this two and a half year period. Per capita income in Bangladesh increased at a rate of 2.6 per cent over this period which has been, thus, sustantially lower than the growth of income among GB members.

In addition to recall, Hossain (1985) provides a comparison of income of the credit-- participants and non-participants in his sample.' Per capita income of the participants was about 31 per cent higher compared to t:ie control group. Within the project villages, the income of the GB members was about 18 per cent higher compared to the non-members. Both per capita income and ousehold income were found to be positively associated with the amount of credit obtained from GB.

The study had undertaken a regression exercise to control for the other characteristics of the households and to isolate the impact of credit on household income. Ownership of land, education, family composition, non-land asset, etc. were included as explanatory variables along with the amount of credit. Credit was found to have a significant impact even after controlling for the other characteristics.

Hossain (1988) presents a comprehensive picture of the impact of GB on income and expenditure of rural households. On the basis of borrowers perception, 91 percent were found to improve their economic condition after joining GB. Findings from in-depth household survey confirms this positive contribution. The average household income for the GB members was about 43 per cent higher than the income of the target group in the control villages and about 28 per cent higher than the target-group non-participants in the project villages. The study provides a detailed breakdown of the sources of income. The positive effect on income has been due to the large increases in income from processing and manufacturing, trading and transport operations, - activities financed by Grameen Bank loans. Agricultural income was similar across groups. An important finding emerging from his study is that income from wage labor and various self-services was substantially higher for non-participants than for Grameen Bank members within the project village. This indicates that Grameen Bank contributes indirectly to the growth of income of non-participant households through an withdrawal of supply from the labor market on the part of GB members.

Poverty situation has been less serious among GB members. Proportion 'of population living in moderate poverty was 84 per cerit among target group nonparticipants in GB villages, 80 per cent for the target group in control villages and only 61 per cent among GB members.

Extent of increase in income and alleviation of poverty will depend on the length of operation of MC in an area. This has been taken into account in subsequent studies on the assessment of the impact of MC. This is especially important for studies which compare more than one program.

Among the more recent IAS of MC, there has been two studies on the assessment of the impact of PKSF financed MC activities. Alamgir (1994) attempts an assessment of the impact of MC on household income on the basis of perception of the respondent who are borrowers from a number of NGOs financed by PKSF. A direct measure of income has not been provided in this study. In Alamgir's (1994) study, about 43 per cent borrowers reported a substantial increase in household income. 54 per cent reported a fair increase. The study does not provide a cross-sectional comparison of household income among the different categories of households. Income accruing to households from different activities have been compared across programs. Such comparisons cannot, however, provide a meaningful picture because the sample sizes for each activity are small. Method of calculation of income has been different for some of the activities. The link between variation of income and total labor force in an enterprise has not been taken into account.

Rahman (1996) provides two types of assessment of impact on income: one based on the perception of borrowers and the second based on a direct comparison of household expenditure among borrowers and non-borrowers in villages covered by eight NGOs of PKSF. About 97 per cent borrowers achieved an increase in income and only one per cent reported a decline. Household expenditure of those who obtained credit for more than once was found to be 17 per cent higher than those who did not get credit. However, the first time borrowers achieved only a 5 per cent rise in household expenditure over the non-borrowers.

A multivariate analysis by Pitt and Khandkar (1995) on the impact of credit for rural poor shows that the amount of credit taken from GB has a positive impact on household expenditure specially on the expenditure on food. The same data set provides a comparison of the extent of poverty in the areas where the NGO programs operate with areas where they do not. It shows a lower percentage of poverty in the areas covered by NGOs (Khandker and Chowdhury 1995). This study shows that the extent of poverty gradually declines as the landless households get repeat loans. Poverty among households who obtained five or more leans is substantially lower compared to new borrowers or non-members. However, households with 0-1 loans from GB and BRAC have a higher percentage of poor compared to non-member households.

Proshika's IAS (IMEU, PK 1995) also compares the income of participant and non-participant households. In fact, this study conceptualizes income as an indicator of economic empowerment'. Whether or not income within certain low range can be an indicator of empowerment, it is surely an useful indicator of material well being and an important consideration for joining MC programs by rural household. Income of PK members is 27% higher than non-PK households. An important feature of increase in income is that among the female headed households, PK members earn 68 per cent more than their non-PK counterpart.

Mustafa (1995) focuses on the comparative returns from various activities financed by MC as indicators of impact. Trading activities are found to generate largest average annual income. Paddy husking, goat rearing and poultry are the activities with lowest return. These activities use smaller amount of capital. The study also observes that the most profitable businesses are those where members or their husbands have some previous experience and working capital.

Mustafa et. al (1996) measures the impact of Rural Development Program (RDP) inputs (creditand other inputs) of BRAC on material well-being, using the following indicators:

i) Per cent of household with all weather roof

The study compares impact of MC by different sizes of loan and by gender of borrower. The main conclusions are as follows.

The study attempted to identify a critical mass of NGO input as a combination of length of membership (more than 2.5 years) and the amount of loan (more than Tk 2500) and tested which indicators of impact respond to such critical level. Statistical tests demonstrated that the value of housing structure and total assets are significantly higher for male members who are above the thresholds. However, other indicators of material well-being, namely, cash earning, food stock and expenditure did not vary significantly for the two categories, those above and below the threshold. For female members of the groups, the difference in the values of most of the variables were found significant. The results also indicated that the improvements have not been'induced by better initial endowment of the members. In addition, it was observed that for the female members, positive impact takes place at a slightly lower level of NGO inputs, compared to male members.

The study adopted the method of project control comparison for assessment of impact of BRAC inputs and the sample was designed accordingly. But a project-control comparison of household income has not been provided which would help to get further insights into project impact. Moreover household income as an indicator of impact did not receive adequate emphasis. Income per household rrember gives a better indicator of material well being of a household, which has not been provided. Appropriate sector/occupation wise breakdown of income has not been provided.

BIDS (1990) Provides a comparative Picture of the impact of various poverty alleviation Programs. The study focused mainly on economic impact. The indicators used in the study include: household income and expenditure, type of expenditure, food intake, employment, land transactions and housing.

The findings of the above study indicate a Positive impact of all types of programs (govt. and NGOS) on households' economic conditions. Some indicators like household income show a greater response compared to consumption expenditure or housing situation. Statistical tests for the difference in the mean are significant between members and non-members. But among the members of various programs studied, the difference in mean household (or per person ) income is -lot statistically significant. Some of the programs has effected changes in the primary occupation of members and a few Programs affected only the secondary occupations. Nutritional inadequacy have been found to be lower among programme members, however the magnitude of difference in rice consumption is not significant between members and non members and among the various programs.

The study examined how far the program participants come from the target group- It was revealed that a significant Proportion of non-target households have been included in most programs. NGOs perform better and largely reach the target group whereas in govt. programs, inclusion of non-target group members have been significant. Therefore the impact of such program on income would be- somewhat exaggerated, because the control group has been chosen from landless target group only.

Research Division of BRAC has taken up a number Of specialized studies on various aspects of poor households. For example, Hadi, A., Nath, S.R. and Chowdhury, A.M.R. (1996) examines the impact of membership of BRAC and other NGOs on per capita expenditure and other social and health related standards enjoyed by households. The study reported a small improvement in both food and non-food expenditure among BRAC members compared to non-program target group. The same study shows a greater improvement among members of other NGOS.

Sen (1996) compares the characteristics of the participants and non-participant target group households (owns less than 0.5 acres of land) and traced the changes between 1989-90 and 1994 through repeat surveys. The targeted programs in which the households participated include the national level agencies such as GB, BRAC etc. as well as the small local NGOS.

During this period, the participants increased their labor force participation rate, and increased the accumulation of land. Comparison of the MC participants and nonparticipants on the basis of panel data shows that the participants started with lower per capita income (PCI) and reached a slightly higher level after five years. They achieved a 3.2% rate of growth of PCI compared to 0.2% by the non-participants (Sen 1996). It did not attempt an analysis of the correlates of credit which contribute to the enhancement of income. The impact of the amount of credit, the organization involved, other inputs etc. should also be examined through the use of panel data.

2. Impact of MC on Employment and Labor Market

Studies on the assessment of economic impact of MC look at a number of other issues in addition to the impact on household income and expenditure. To understand the process of increase in income, some of the studies analyze the changes in employment and occupation. In addition, a number of studies examined the impact of credit on savings and capital formation.

All case history based impact assessment studies emphasize the increase in female employment and in the labor force participation among women. Case studies on female borrowers of GB and other credit programs narrate how the hard working women have been engaged in income eaming activities and thereby increased the household income. Yunus et. al (1982), Ray (1987), Rahman (1996) and many other studies describe cases where women borrowers use a small amount of money and gets involved in an economic activity and work for long hours. In addition, the case studies show that, for male workers, credit has contributed to the improvement of labor productivity through the addition to capital in the existing enterprises. In some cases, credit enabled them to have ownership of the equipment used instead of having to rent it and thus enhanced his income.

Studies on the impact of GB by Hossain (1985, 1988) show that MC encourages the involvement of household members in non-agricultural occupation. This is confirmed by other studies (Rahman 1996, Mustafa et. Al 1996). The objective of MC is to generate self-employment which can enhance household income. Studies have shown that MC raises the probability of being engaged in self employment (Rahman and Khandker 1994).

Many of the studies did not examine the impact of credit on the employment situation. This is partly due to the difficulties of collection data on employment on the basis of recall method. To obtain year round data on employment and to capture the seasonal variation of employment, multiple surveys over a year is the ideal approach. Most surveys involved one shot data collection and therefore employment data has been presented either in the form of labor force participation rate, based on usual status or the amount of employment during the month preceding the survey.

Hossain (1985) identifies the generation of self-employment as the main mechanism through which MC has been effective in accelerating the growth of income. This has been reflected in the higher labor force participation ratio among GB borrowers (30.1 %) compared to non-participant target groups in project and control villages (26.4% and 24.3% respectively).

Hossain (1988) and Rahman et. al (1 994) provide gender disaggregated data on labor force participation. Positive changes in labor force participation among both men and women from households borrowing from GB has been revealed by the data in Hossain (1 988). The increase is much higher among women. Similar findings have been reported by Rahman and Khandker (1994). However, the latter study found an increase in labor force participation rate only among women. This study found a positive impact of GB on the hours of employment among both male and female workers. It has been found that employment impact extends to the non-participants as well. Employment among both participants and non-participants in the project villages were found to be larger than the similar group in the control village. The same study revealed that BRAC and BRDB programs did not have such positive impact on employment.

An indicator of economic empowerment used by the !AS on Proshika (PK) is the wage settlement process and the PK members’ ability to influence the wage rate. In this respect. the PK members were observed to be slightly better off than others. However, the wage setting process in the rural labor market may be influenced by a host of economic and social factors and may not be simply determined by the employer or employees as asked in the question. The other type of market power, assessed in terms of where and to whom product is sold shows an improvement for the PK members. They can sell at a distant market and directly to wholesalers rather than to the mohajan.

3. Savings, Credit Market, Asset Accumulation and Microcredit

So far we have focused on the impact of MC on the final impact variables: income and expenditure. However, the increase in current income and consumption expenditure describes only a part of the benefits of MC. Equally important components of impact include savings, asset accumulation, and expenditure on human capital leading to human capital formation. Changes in savings and asset ownership will have an impact on the operation of the rural informal credit market. Such changes deserve special attention because these are likely to be instrumental in generating a second round of growth of income.

While we discuss the impact of MC on saving and asset accumulation, it should be mentioned that the voluntary savings require more attention than the compulsory ‘group fund' savings. Almost all the large MC institutions impose the requirement of compulsory illiquid group fund savings. But the borrower-savers do not have ready access to this saving. Therefore the household savings not deposited with the NGOs are crucial for meeting emergency expenditure needs, not only for consumption but also for investment. To assess the impact on savings one should recognize that in addition to cash savings, rural households' saving take various forms of direct investment which should be accounted.

Credit from MCIs is expected to reduce the dependence of poor households on the informal credit market. One of the mechanisms through which poverty is perpetuated is the borrowing from money lenders at exorbitant rate of interest. Once a poor household borrows to meet an emergency need or for social functions, it can hardly free itself from the grips of money lenders. Institutional borrowing can help to free themselves from the clutches of these informal lenders. A few studies provide information on this aspect.

Mustafa et. al. shows that households with access to credit from BRAC, improved their position in this respect. This is indicated by a process of withdrawal from the informal credit market. The average value of informal borrowing is larger for the middle of the scale input levels (from BRAC) and declines for the top end of the scale for both male and female categories (for respective categories from Tk. 5445 and Tk. 4233 to Tk. 2333 and Tk. 1244 per borrower household). The harmful process of dependence on informal credit market declines not only in terms of amount borrowed but also in terms of the declining share of consumption as a purpose of informal borrowing. For the female members the decline is steady and gradual (with increase in inputs ) whereas for the male members it fluctuates and, nevertheless, is lowest in the largest input category. Informal borrowing for consumption purposes declines from 55 per cent and 67 per cent of the total borrowing, respectively to 32 per cent and 41 per cent respectively for the male and female borrower in the largest combined input level category. Similar pattern is observed with respect to the use of institutional credit. The use of BRAC credit to meet consumption and other hardship needs increases initially at medium level of input (1 1 and 14 per cent respectively for medium categories) and then it declines sharply in the largest input level (4.5 and 5 per cent respectively).

Hossain (1988) examines the impact of MC provided by GB on the rural credit market. One of the basic premises of the Grameen Bank,, is that the poor non-farm operators, particularly those employed in cottage industries and petty trade are exploited by village moneylenders. The effective rate of interest involved in -the process is often implicit and the exploitation takes place not only through high rate of interest but also through a control over the input and product trade. By exercising a monopolistic control in the interlinked input, product and capital markets, the 'Mahajon' can extract a large proportion of the current income of the borrower. Grameen Bank credit is expected to enable the poor non-farm entrepreneurs to become independent of the financial market intermediaries.

His study shows that the dependence of GB participants on non-institutional sources was lower in comparison to similar groups of non-member households in the project and control villages. Only 6 per cent GB members received loans from noninstitutional sources, compared to about 18 per cent of the bank's target group in the control villages and 20 per cent among non-members within the target group in the project villages. Among the target group in the control village only 9 per cent households received institutional credit. With GB intervention, where the Proportion of households receiving institutional credit was 30 per cent. The share of institutional loans in total was about 78 per cent in the GB village.

MC is expected to contribute to accumulation of both working and fixed capital. Investment of MC on fixed capital can be viewed as net saving if the loan is repaid without having to sell other fixed asset, i.e., from current income. This is possible because the loan is repaid in weekly installments, leaving the capital intact. Hossains’ (1988) study shows that about 80 per cent of households reported accumulation of nonagricultural capital after joining the Bank and the average amount of investment was higher for long time borrowers than for newer ones. His findings show that the share of equity in total investment increased from 11 per cent to 28 per cent for members borrowing four times or more. An important form of capital accumulation among borrower of GB is the ownership of livestock. The average number of cattle owned increased by about 67 per cent during this period. Among the borrowers, 37 percent owned livestock before they became members which increased to 65 per cent during their membership with GB.

Khandker and Chowdhury (1996) shows that MC has a positive impact on savings and asset accumulation by households. Current savings of those who obtained MC is many times higher than the savings of those who did not get credit. Total asset ownership showed similar difference. The amount of saving increases sharply with the number of loans obtained by a household. Among the types of assets, greatest increase occurred in house and transportation assets and livestock. Increase in landed assets among members was not significant.

Mustafa et al (1996)'s study also observed a positive impact of credit on savings and capital accumulation. A comparison of those with large and small (cumulative) amount of loan from BRAC-RDP shows that the former group has achieved a larger amount of current savings and possess more assets in the form of house and livestock. An increase in ownership of house is the most significant change among the borrowers who obtained credit above the threshold level. The increase in the ratio of productive asset to total asset is not, very large. However, it should be mentioned that the estimates of productive assets does not include the value of homestead area or living quarters. But in the rural areas where household and farm/firm overlaps, the house and the homestead area contributes substantially to productivity levels and low income rural households. Such households usually spend their savings initially on the improvement of housing condition.

Rahman (1996) does not provide a direct estimate of savings but presents a picture of investment expenditure out of current income. It is observed that the first time borrowers did not spend much on capital assets, but those who borrowed more than once, spent a much larger amount on this head.

Increase in saving and investment among MC program participants, vis-a-vis the non-participants is also revealed by the IAS of Proshika (IMEC, PK 1995). It shows an astounding increase in savings by poor households. PK households' saving is 2.4 times of the amount saved by non-PK households. However, the proportion of income saved is not very high among the PK households. The two groups save only 7 and 3 per cent of their respective household income. Such low average savings rate may be due to low range of income among the poor households. In addition, it is di ' fficult to obtain accurate estimates of rural household savings because of the informal nature of such savings. A large part of the saving may be directly reinvested which often goes unnoticed.

It is difficult to interpret the investment data presented by this (IMEC, PK 1995) study. Average amount invested is Tk 5525 and Tk. 3933 respectively among PK member and non-member households which are much higher than the savings. For the PK members, the amount invested is possibly supplied by PK loan. A summary of the findings of some impact assessment studies is presented in table 1

                  Table – 1
A summary of the findings of IAS

4 Impact of MC on Women

MCIs have opened up new opportunities for rural women, which were quite unthinkable for them. Uneducated and assetless poor women becomes recipient of institutional credit by virtue of MCI activities. This has enabled them to achieve what may be considered as a fundamental right. Among 4 million landless members who obtain credit from GB and other MCls, 90 per cent or more are women. Therefore, it will be useful to provide a picture of the progress in the percentage of female borrowers among the micro credit organizations.

4.1 Progress of NGO Credit among Women

During the initial years, GB worked mostly among men. In 1983, women members were 45 per cent of total. The percentage of female borrowers increased rapidly during the late eighties and early nineties. From 1993, more than 90 per cent borrowers are women (Table 4.1). BRAC maintained a more balanced distribution of male and female borrowers for a long time. During the late eighties, female borrowers were around 65 per cent which increased to 74 per cent in 1993 (Table 4.2).

                                           Table 4.1
Gender Distribution of Borrowers of Grameen Bank 1983-1993

Source : Grameen Bank, various reports, Khandker et. al. 1995

                                     Table 4.2
Gender distribution of borrowers of BRAC 1989-1993

Source : Khandker and Khalily 1995

                                               Table 4.3
Gender distribution of borrowers of NGOs covered by PKSF

Source : Annual Report, PKSF, various years.

ASA had both male and female groups at the beginning. But as soon as they shifted their emphasis to the provision of credit, all the groups were turned into women's groups, the wives female household members being the substitute for men. This gives an impression that women can play a useful role in channeling credit.

Most of the smaller NGOs also work mostly among women. An idea about the gender distribution of the recipients of loan from small NGOs can derived from the distribution of PKSF funds which is disbursed mostly through small NGOS. In 1994-95 and 1995-96 women constituted 87 per cent of all borrowers.

These information suggest that women in Bangladesh have captured the micro credit market. This is an indicator of success for women, but one also needs to go deeper into the explanation of this trend and review the impact of MC on women's empowerment.

4.2 Micro Credit and Women's Empowerment

A major objective behind taking up ME activity organized by micro-credit is to utilize family's female labor force who would be otherwise underemployed. Social reasons and lack of wage employment opportunities for women impose constraints on the way to fuller utilization of women's labor even if the poor families are in desperate need to raise family income through intensification of family labor use (Rahman 1996, 1990). A number of studies have shown that a large part of the labor input in family enterprises is provided by women.

A number of studies on impact assessment of MC looks at the impact on female employment. It has been observed that the magnitude of female labor is larger in enterprises financed by loans provided to women. Rahman (1986), Rahman and Khandker (1995) and Pitt and Khandker (1995) have concluded that female labour supply shows a positive response to micro-credit supply. Self-employment of women, and thereby total household employment is higher in villages with NGO credit programs (Rahman and Khandker 1995).

Whether women's employment can be considered as a step towards empowerment will depend on the interlinked changes that take place with the changes in employment. A control over income and asset, and the improvement of consumption, increase in contraceptive use etc. can be considered as components of empowerment of women which could be achieved through women's greater involvement in income earning activities.

Several Studies Provide definite evidence Of improvement in women's situation resulting from access to NGO credit. Whether the causality works through the receipt of credit or through income enhancement based on the investment of credit is not always clear. Schuler and Hashemi (1 995) provide evidence on women's empowerment through GB and BRAC credit. Using logit regression, they have shown that the membership of GB and BRAC has positive impact on women's status both directly and indirectly (through their income earning capacity),

Another study (Mustafa et. al. 1995) on the impact of BRAC has concluded that BRAC has brought about gradual changes in the lives of women. Members involved in BRAC programs have gained a greater control over their income and has been able to gain control on expenditure.

An early study (Rahman 1996) on the impact of GB credit provides evidence that the women who borrowed from GB enjoyed a better standard of living for themselves (in terms of food intake and possession of clothing) and a greater control over family decision making. Though the amount of food consumed (in terms of number of meals) by women was not higher than men among the borrowers, the variety of food items consumed was better. Moreover, the amount spent on clothing by borrower women was higher than women from non-borrower households of target group.

A number of studies on impact assessment of MC looks at the impact on female employment. It has been observed that the magnitude of female labor is larger in enterprises financed by loans provided to women. Rahman (1986), Rahman and Khandker (1995) and Pitt and Khandker (1995) have (concluded that female labour supply shows a positive response to micro-credit supply. Self-employment of women, and thereby total household employment is higher in villages with NGO credit programs (Rahman and Khandker 1995).

Whether women's employment can be considered as a step towards empowerment will depend on the interlinked changes that take place with the changes in employment. A control over income and asset, and the improvement of consumption, increase in contraceptive use etc. can be considered as components of empowerment of women which could be achieved through women's greater involvement in income earning activities.

Several studies provide definite evidence of improvement in women's situation resulting from access to NGO credit. Whether the causality works through the receipt of credit or through income enhancement based on the investment of credit is not always clear. Schuler and Hashemi (1 995) provide evidence on women's empowerment through GB and BRAC credit. Using logit regression, they have shown that the membership of GB and BRAC has positive impact on women's status both directly and indirectly (through their income earning capacity).

Another study (Mustafa et. al. 1995) on the impact of BRAC has concluded that BRAC has brought about gradual changes in the lives of women. Members involved in BRAC programs have gained a greater control over their income and has been able to gain control on expenditure.

An early study (Rahman 1986) on the impact of GB credit provides evidence that the women who borrowed from GB enjoyed a better standard of living for themselves (in terms of food intake and possession of clothing) and a greater control over family decision making. Though the amount of food consumed (in terms of number of meals) by women was not higher than men among the borrowers, the variety of food items consumed was better. Moreover, the amount spent on clothing by borrower women was higher than women from non-borrower households of target group.

Rahman (1986) used a number of indicators to examine the impact of GB membership on the awareness among women and their decision making power. It was observed that the female borrowers had a greater decision making Power compared to the wives of male borrowers. Moreover, GB member women demonstrated greater awareness level in terms of most indicators (the indicators used are: whether castedvote, knows how to prepare oral saline, aspirations for education of sons and daughters, preferred age of marriage for daughter).

A discernible Positive impact Of women's involvement in MC financed selfemployment is the increase in contraceptive use (Schular and Hashemi 1995). GB is not actively engaged in promoting contraception. Yet, a greater use Of contraceptive in villages covered by GB has occurred through the strengthening of women's economic roles which give them more autonomy and control over decisions affecting them and their family. In addition, the involvement in income generating activities contributes to their self-confidence and ability to have access to family planning materials.

Micro-credit programs not only improve the contraceptive Prevalence rate among the borrowers, but improve the rate for the community as a whole (Schuler and Hashemi 1995, Khandker and Latif 1995). The effect on contraceptive use among non-members is a diffusion effect resulting from the changing fertility norms among GB members. As the number of contraceptive users in a community increases, it becomes easier for all women to adopt family planning, as their fear of side effects and social restrictions is reduced

Positive impact of credit and related inputs provided by BRAC on women's status has been reported by Mustafa et. al's study (1996). Though the credit obtained by women has been often used in the enterprises managed by the male members of the household, the access to such finance improved women's decision making role in the family. Better treatment by husbands for the wives have been reported to be evident.

During the initial years of the program, male members were opposed to formation of groups by women which has recently been reversed. Women who are trained in certain skills, such as, poultry vaccination, village health worker skill, other occupation skill etc. are accorded particularly enhanced status by the villagers. Women's decision making power has been observed to be positively associated with the length of membership with BRAC and the amount c-f loan. Role of men as the sole decision maker in the household was observed to be declining and in most cases decision was taken jointly. In the early stages, when small amount of loan has been involved, women were often found to be taking independent decisions.

Rahman (1996)'s study on the impact of PKSF credit examined the relationship between the length of NGO credit and gender differential in the access to basic needs (food and clothing). Such differential did not change in favour of women as the households obtained a larger number of loans. Larger number of loans increased the access to food and clothing for both men and women.

Mizan (1994) studied the impact of GB membership on the decision making power of women. To test the assumption that a greater participation of women in GB leads to a higher decision making power, the author developed a composite unidimensional variable representing decision making power. The value of this score was compared for borrowers and non-borrowers and the difference was significant. The decision making score increased with the cumulative amount of loan. In addition to years of loan from GB, the amount of loan may also positively influence household decision making because a larger loan increases the financial resources of women. In the household context, the larger loan is expected to give higher financial power and, in turn, enhance their decision making power. The findings confirm this expectation and such power increased with the years of loan until they reached the sixth year after which it remained at the same level.

A multivariate analysis was also conducted, which shows that GB participation has a significant impact on women's decision making power even after controlling for demographic factors and other household characteristics. Though the inclusion of some of the participation variables may introduce the problem of endogeneity, the general point that such participation has an important impact is reflected in both bivariate and multivariate analysis.

Gibbons and Todd (1993) observes that GB loans help the Poor assetless women to increase their earning and as a result their social position within the family and community has improved remarkably as has their self confidence and self esteem, There is a strong relationship between how economically active the loanee is and her status.

Mahmud et. al's (1966) study uses a number of indicators to reveal the Positive impact of BRAC membership on women's status. These indicators include both economic and social parameters. It has been demonstrated that member's contribution to household income is significantly higher than the contribution by non member women in target group as well as in non-target group. Women from BRAC member households were found to have higher rate of savings. They were less dependent on noninstitutional credit. They utilized a larger percentage of their non-institutional loan in income earning activity (60%) in contrast to target group non-members, who used most of their loan for consumption. However, it is a matter of concern that a large percentage (40%) of loans from BRAC has been used for purposes other than income generation. Moreover, only 20 per cent of women among the members were employed in income earning activities which implies that many of the loans have been used by male working members in the family. Yet, the activity rate among women from BRAC member families was higher than other groups of households.

Zohir's (1990) study observed that SB has Positively influenced women's status in the society and the social influences pertaining to women's security. The former arises due to the loan disbursement to female members who can now have greater say on family matters.

Zohir's study (1990) also reports that even though a large Percentage of SB credit goes to women, the loans are used by husband and/or the father of the female borrower and therefore the involvement of women in credit programme is not reflected in their participation in the workforce.

Zohir (1990) obtained Positive impact of 'SWANIR\JAR' on social aspects like Family Planning and literacy; program villages in this study were in a better position In these respects, compared to the control villages.

To examine social status Of women, Mahmud et. al. used the following indicators: women's mobility, decision making regarding small purchase and large purchase, violence against women, Political awareness, legal knowledge, incidents of injustice and women's participation in protest. BRAC membership was found to improve women's mobility and awareness. Some improvement in decision making was observed. But the incidents of injustice and violence were no less among the members. The explanations offered by the authors is that members are more vocal about these and others may not have disclosed such incidents.

The above discussion on the impact of women's involvement in MC provide ample evidence of social progress among women and in the community as a whole. Such Progress has been due to two interlinked forces: the income generation through ME and the involvement in the innovative loan program which involves processes and practices which are outside the existing social norms.

One should however, be cautious in generalizing the Positive role of micro- credit for women. It has been observed in a number of studies on the use Of micro-credit (Rahman 1996, 1993, 1986) that women cannot always use the credit in enterprises which they run independently by themselves. They usually need assistance from male members of the household and as a result they cannot exercise full control over their loan and credit based resources.

The success of credit in empowering women may be understood in a wider perspective if one makes a comparison of women's self-employment with women's wage employment in large scale industries in the urban areas of Bangladesh. Women's wage employment has been observed to bring about even a greater social change (Majumdar and Zohir 1991, Kabeer 1995) and a promise of regular source of income which is higher than earnings from casual wage employment or from self-employment in rural micro enterprise (Rahman 1994).

It can be argued that investment in ME will by no means prejudice against large scale industrialization and women's employment in the formal sector. not created fixed . capital that will need to be scrapped if large scale industrialization takes place. More importantly such industrialization is still not in sight even in the distant horizon of the country's rural development program

4.3 Other Social Impact

In this section a number Of social impact indicators are examined and how various mc programs influence these varia@les will be discussed. These indicators include health and nutritional status, vulnerability and crisis coping mechanism etc.

Gibbons and Todd (1993) presents some findings on the impact of GB participation on the health of the children. Increased income and women's control over it has generally resulted in an improvement in the health of the children of the GB long term loanee households. Anthropometric measures were used to test the impact on health of 2-10 year old children, who were bom when their mothers have been members of GB as compared to such children from the control group. 'Grameen' children were found to have higher value of all the anthropometric indicators. These averages were much higher than the national averages. However, it was observed that there was still some malnourishmnt among the children of GB members and therefore it was concluded that they should be encouraged to translate the nutritional slogans into actual practice.

Chowdhury and Khandker (1 995) examined the nutrition status among recipients of loan from GB, BRAC and BRDB. However, this study did not provide a comparison with a control group and therefore the impact of program cannot be identified. GB borrowers were in a better situation compared to other programs.

Fuglesang et. al. (1986) support the view that the awareness about nutrition is better among the GB members and the 'sixteen decisions' have been an effective tool for raising awareness and to create aspirations for better living.

IMEC, PK (1995) examined the impact of PK membership on the social empowerment. Among social empowerment, they have chosen some conventional indicators like literacy rate, school enrollment rate, sanitation practices etc. In addition, social empowerment includes the decision making role of women. In general, most of the indicators reveal a positive impact of PK-membership.

A study on BRAC (Mustafa et. al. 1995) revealed that households' vulnerability to seasonality and the situation of food security, are responsive to the length and amount of credit. Food security is more sensitive to levels of credit than the pure length of membership. The food security levels show close relationship with the amount of credit and the phenomenon is more prominent for the female borrowers. Data on the distribution of food security shows that the proportion of households in the 'severe food deficit' category declined from 39 per cent for the 'no-loan' category to 12 per cent for the greater than Tk 10,000 credit category. The proportion of food surplus households increases from 1 1 per cent to 28 per cent for the respective loan categories. Improvement in seasonal vulnerability is observed to be more systematically related to the amount of credit among female loanees.

Seasonal fluctuation of a household's consumption level and material well being is an important dimension of poverty in Bangladesh. Such fluctuation causes hardship for the Poor which is manifested by a decline in consumption levels in slack season compared with the better season. The access to micro-credit is expected to result in a reduction in the differences between the lean season and the peak season with respect to per capita weekly rice consumption, per capita weekly expenditure on food, per capita total expenditure, seasonal food stock, and per capita weekly cash earning. The study observed that the declining seasonal vulnerability among borrowers is revealed by their enhanced coping capacity. Recent studies on poverty have put forward various indicators to assess the coping capacity of households. One of the most important indicator is the mechanism used to cope with seasonal downturn in the economy. The case studies in Mustafa et. al. (1996) reveal that the coping capacity of the members in older VOs improved as they obtained credit for a longer period. Their coping mechanisms are of less damaging nature for the household resource base Diversification of employment and the use of kin and friendship-networks, instead of the harmful ones such as informal credit with high rate of interest and asset depletion were reported to be used as mechanisms to cope with seasonality. The credit from BRAC helps the diversification of earning sources in the lean season specifically.

Another indicator of security is the capacity to ensure a stable earning in the long term. This in turn will depend on a households ownership of productive assets. The long term security of a household is indicated by the increasing value of real capital in the asset profiles. The average value of the revenue earning assets shows a steady increase with an increase in the length of participation and size of credit (increasing from Tk 3,419 to Tk 6,452 for the females borrowers with largest loan. The share of the house structure in total asset value is large for all RDP input levels. Higher value of a house suggest the use of high value, more durable building materials. Investment on housing is quite common among rural families because a house adds to their productive capacity, which Is a rationale for such investment even though its depreciation rate may be high.

Atiur Rahman's (1989) study also shows a positive impact of GB credit on the ability to withstand disasters. Following the 1987 and 1988 floods, a household survey of two flood affected localities, was conducted covering both GB members and control groups. He sought to explore the capacity of the rural poor to survive and re-establish themselves following natural disasters It was observed that GB members depended less on informal credit and local money lenders and in most cases used working capital from their business or their own savings to survive disasters.

Chowdhury (1989) examined the impact of a number of NGO credit program and GB program. For all programs positive social benefits in the form of awareness and economic strength was obtained. Economic benefits were found to have empowered them. Some of the programs increased the participants access to public resources.

5 Case Studies on MC

5.1 Examples of Case Studies

Ray (1987) describes stories of a large number of GB members, to derive insights into their plights. For example, a number of case studies describe the fear and trauma faced by women who became GB members in the early days. Here we quote some portion of the story of Kamala Begum:

Case study on 'overcoming fear'(Ray 1987)

Kamala is a member of Chamaree village. Her husband was a day laborer without any farmland. Kamala got taka 2500 as the first GB loan on the basis of her proposal to purchase a cow which could produce milk to enable her to pay weekly installments to the GB. But she failed to buy a suitable cow with that amount. She got nervous. She was not sure of her ability to manage the accounts of a loan of Tk. 2500 if she used it for other purposes. At the same time she considered that her default of the loan would place her group in jeopardy. She hated to enter into such trouble. After considering all these possibilities, she returned Tk. 2000 and obtained only Tk. 500 as loan. She used this as capital for being engaged in 'paddy husking' activity, She earned sufficient income and successfully repaid the loan. She had overcome her nervousness and uncertainties. She obtained a second loan of taka 2000.

However, the case studies of Ray are more useful to describe these qualitative aspects. Description of cases on education and income generation could be more useful if supplemented by quantitative data.

Case Study (Rahman 1997): Sharif's Way of Managing Access to Land

This story of a Grameen Bank member shows how a small loan can help a household to come out of poverty and distress and remove a crucial bottleneck in their endeavor to utilize an existing skill. Sharif is an energetic person in his early thirties, married to Nilufar who is about 22. Sharif worked in their family enterprise of cotton yarn spinning. His father managed the business and he himself along with his three brothers worked in the actual yarn processing. Income from this business was Sufficient to maintain the family of five members.

But two years after Sharifs marriage, his father died. Sharif, started to manage the business as he was the eldest among the brothers. With his experience and hard work, he felt he could carry on the business successfully. His younger brothers, however, did not take into account the hard realities of life and engaged in extravagant spending without asking Sharif's permission . As a result, the business failed and they were left in debt. Even this downturn in business failed to alert the younger brothers to their responsibilities as partners. At this point, Sharif and Nilufar moved to a separate household and the joint family broke down.

But Sharif did not have any capital to run his business. All his savings and even the proceeds from the sale of a sewing machine owned by Nilufar went to the repayment of debt. They were saved from plunging into further debt when Nilufar became a member of GB and borrowed Taka 3000 to buy a sewing machine. Nilufar's neighbours were members of GB and they provided her with information on 'his when she was looking for a source of funding. She bought a sewing machine for Taka 2000 and purchased tools for yarn spinning with the rest of the money. Nilufar earned around Taka 100 per week from sewing, and Sharif started his own business of yarn spinning. Unfortunately, due to the small size of capital it was not very profitable. The open space they used for spinning was owned by another person who charged Taka 1 00 per month as rent.

Nilufar obtained another loan of Taka 4000 in the next year. They used the money to buy a small piece of homestead land and constructed a small hut which became their home. They chose their homestead land in a location where there was some government owned (khas) land nearby, which they could use for their spinning so that they did not need to pay any rent for the land.

At present Sharif brings raw material from traders on credit, processes the yarn and sell it for Taka 10 per kilogram. Their weekly income from this is around Taka 200. The return from this is low because a substantial part of the profit is extracted by the traders. Nilufar plans to take another loan next year so that they can purchase raw materials with that cash and sell the processed yarn directly to the manufacturers. That will more than double their weekly earnings.

The couple have accumulated personal savings of Taka 200 and group-fund saving to Taka 300 with the GB. Such savings are important for self-employed families whose earnings are crucially related to their labor input. It means that they can draw money from their savings in times of emergency.

Case Study (Rahman 1996): Romila: ttie Hard Working Entrepreneur Financed by

a Small Local NGO

Romila's husband's grocery shop hardly earned enough to maintain the family of five members. Capital in the shop was small. Romila gave her husband a part of the fund she obtained as loan from a PO of PKSF. This not only increased the return from the shop, but also ensured the cooperation from her husband in her efforts of organizing her own self-employment activities. She invests the rest of the loan in her business of making 'Pitha', and other snacks which she sells at the nearby market centre.

Time use pattern of Romila

She starts her day early so that she can do the chores and provide breakfast to the family by 7:30 am and go to the street side to sell her 'Pitha'. She cooked those last night and went to bed quite late. By 9:00 am it is sold and Romila comes home. She starts preparation for the second round of 'Pitha' making which she sells in the afternoon. She employs one woman to pound rice. Rice flour is used in most of the Ipitha'. She says that, while she makes all these snacks, it is of not much extra effort to cook the midday meal for the family. By 4:00 pm she is again ready with her snacks and sells them till sunset. She has much more to do after she comes home. This is the time for cooking and serving dinner for the family. After dinner she cooks the snacks and Pitha for selling in the morning. Paying back the loan in weekly installment is no problem for her. After meeting all costs and repaying loan, she gets a return of about 100 taka per day. Such a return had been possible because of 9-10 hours of hard work.

She had demonstrated judicious entrepreneurial decision by hiring a woman to do the rice-pounding, which enables her to make snacks at a larger scale. She has no problem from the demand side and can sell as much as she can produce. She had cleverly chosen the place for selling her product at the spot in the corner of roads along which long lines of workers pass (most of whom work in the garments and other industries in the area).

With the additional income, Romila purchased household goods and furniture. She had arranged to take electricity connection for her house. 'After all this hard work by the side of the heat of the 'Pitha'frying, don't I deserve a good sleep with an electric fan in my room' - Romila justified her endeavour for improving the standard of living. This strive for better living is of course the driving force behind the hard work by entrepreneurs in all societies and at all times.

B. Finding of PKSF ME Study by BIDS

Palli Karma-Sahayak Foundation (PKSF) has commissioned a longitudinal Monitoring and Evaluation Study (MES) which is being conducted by the Bangladesh Institute of Development Studies (BIDS). Started in 1997, the studies will be completed by October 2000. The study covers 3026 sample households spread over 91 villages located in 13 of PKSF POs operational areas. The first and the second round surveys under the study were administered in 1998 and 1999 and the third and final round survey is being conducted presently. A comprehensive report addressing all the objectives and issues of the study will be written after the final round survey. Meanwhile BIDS has prepared two preliminary reports incorporating partial analyses of some analyses based on the first and the second round surveys. Besides these two reports, BIDS has also prepared other reports – (a) Census Report; (b) Impact Assessment Studies on Microcredit; (c) ME Manual (draft); (d) three case studies on three POs and e) a number of workshops papers and published two newsletters based on the findings of the study. Some findings on the impacts of microcredit as presented here are gleaned from the reports and papers prepared by BIDS.

Section 1

Economic Impacts of Microfinance

1.1 Income

The 1998 survey found the average annual income of land-poor participant households to be higher than that of the non-participants. Self-employment activities had higher contribution to total income for the participants for both eligible and non-eligible groups. Amongst eligible households, such activities contributed more than 50 percent of total income, while the corresponding figure was only 43 percent in case of the eligible non-participants. Estimates from the second BIDS survey suggest that nominal household income increased by 19 percent in program villages and by only 13.5 percent in control villages. The participant households were better able to sustain their self-employment activities than the non-participants. It was generally observed that the 1998 flood had more adverse effects on land-poor households, leading to deterioration in the rural income distribution. However, access to microfinance and other assistance from the MFIs enabled the participant households to better cope with floods.

Consumption of cereals generally correspond with the income of different sample groups. Thus, in program villages, program participants among the land-poor households were found to consume higher quantities of rice, wheat and pulses compared to the non-participants. An average eligible participant household consumed 7.21 kg. of rice, 0.99 kg. of wheat flour and 0.12 kg. of pulses over a period of three days (during 1999 survey). In contrast, an eligible non-participant consumed 6.4 kg. of rice, 0.41 kg. of wheat flour and 0.11 kg. of pulses during the same period.

1.2 Food Security

Household-level food security is no more perceived narrowly; and such security may arise from enhanced ability to procure food from the market with earnings from non-farm activities. The BIDS study finds the program participants to have greater access to the land rental market, and thereby better ensure employment (on rented-in farm) and food security from own production. The study estimates that about 26 percent of rice consumption by the land-poor households is met out of own production (after sale). Because of greater access to cultivable land through the rental market, the participant households had marginally higher amount of rice from own production than the non-participants.

1.3 Wage

Wage earning contributed about 23 percent of total annual income for the land-poor households. In absolute terms, wage earning of the participant households was about 8 percent higher than that of the non-participants. Such difference arose primarily due to higher wage earning by the participant households from transport and other non-agricultural activities supported by microcredit.

1.4 Employment

The land-poor participant households are better able to ensure employment on own farms due to their better access to the land rental market. As noted before, wage employment in non-agricultural sector is higher for the participant households. Moreover, higher net income from self-employment activities suggests better employment opportunity for them in such activities. All these generally reflect positive impact of program participation on employment.

1.5 Assets – land and non-land

Average size of land owned by participant households is lower than the non-participants; 91 decimals compared to 149 decimals. The BIDS study however suggests that the eligible participants mortgage or rent-in more land than the non-participants, and therefore, have larger operational holding.

In the case of income-earning assets, such as, poultry and livestock, positive impact of microfiannce is quite evident. The findings clearly reveal that a higher percentage of program participants among the eligible households own poultry, goat/sheep and cows, when compared with the non-participants.

Ownership of such assets as bicycles, hand tubewell and rickshaw/van is found to be more prevalent in program villages, compared to the control villages. Within the program villages, compared to the non-participants, a higher percentage of the land-poor participants own bicycles (12.5% compared to 8%), general boat (3.7% compared to 1.6%), irrigation equipment (1.23% compared to 0.23%), radio (17.9% compared to 12.6%) and rickshaw/van (8% compared to 2%).

The 1999 survey by BIDS finds no significant difference across sample groups in the incidence of sales and purchases of non-land assets. Only in the case of rickshaw/van, a larger percentage of the land-poor participants reported of its purchase, compared to the non-participants. One however needs to account for the instability in such ownership, since about 1.6 percent of the eligible participant households had sold such asset during 1998-99.

Section 2

Social and Other Development Impacts

2.1 Fertility and Contraceptive use

2.2 Health and Nutrition

There is positive program placement effect on nutrition status.

2.3 Sanitation and Drinking water

Upon comparing program and control villages, the BIDS study finds small positive influence of participation on waste disposal, but only among the land-poor households. There is also a positive effect of program participation on the use of sanitary toilets. With regards to the method of hand-washing, there is no clear evidence on program impact. The use of pure drinking water is universal because access to tubewell water is widespread.

2.4 Literacy and School enrollment of children

Adult literacy rate, whether defined in terms of ability to sign or in terms of ability to read and write, is significantly higher among the eligible participants compared to eligible non-participants. The BIDS study also found that among the eligible or land-poor households, program participation increases the chance of both boys and girls (aged 5-9 and 10-14 years) to be enrolled in schools.

2.5 Social mobility

Mobility of women, captured in terms of commuting to a number of selected places, do not vary significantly in most cases across participant and non-participant households. However, women of participating households visit health centers and NGO offices more frequently than women of non-participating households.

2.6 Women Participation and Household Welfare

Women membership in microcredit program has a significant influence on women’s time allocation decision in poor households to the extent that this is possible by reducing women’s labour time in non-market production, household maintenance and childcare, none of which is independently affected by women’s membership status. Hence, participation in a microcredit program has a beneficial effect on household welfare by increasing women’s income contribution to the household. Participation also has a beneficial effect on women’s welfare since it increases their total employment time by reducing their other work responsibilities. Participation of women in microcredit program has also increased their mobility.

C. Findings of IDA Study

A rapid impact assessment study conducted by the World Bank / IDA in early 1999 also shows that the poor have benefited from the microcredit program of PKSF in several ways. It finds that

(a) income has increased for 97.93% of borrowers;
(b) quality and quantity of food intake have improved for 88.59% of borrowers;
(c) clothing improvement has been reported by 87.85% of borrowers;
(d) housing conditions have been improved for 75.26% of borrowers;
(e) children education has been improved for 75.41% of borrowers;
(f) sanitation condition has been improved for 68.74% of borrowers; and
(g) overall quality of life improvement has been reported by 94.96% of borrowers.

The World Bank / IDA study also finds that land ownership of 28.5% borrowers has increased and they on average acquired 30.6 decimals of land, significant for a poor family. Ownership of houses with standard wall and CI roof has increased from 25.6% to 44.3% and houses having walls and roof with CI materials has increased from 7.3% to 36%. About 87% of borrowers perceived that they have accumulated asset from the income of microcredit. Their net worth is about Tk. 26,794 which is significant for a poor family. The self evaluation of the borrowers on their family condition has been very positive. The number of families having food deficit (and other incidental expenses) has been reduced from 82.2% to 9.6% over a period of 5 to 6 years, and also 77.3% of the families have either ‘adequate’ or better economic condition compared with 4.7% before they joined the microcredit program of POs.