[This is the second and final part of the paper presented in the Evaluation Dialogue on Grameen Replication, held at IRRI, Los Banos, June 12-15, 1997. CARD is one of the largest Grameen replications, in the Philippines.]
The rate of recovery of credit and the demand for repeat loans are indirect indicators of the financial viability at the borrower level. If the member incurs loss in the business enterprise, she would not have capacity to repay the loan. If she had been forced to repay the loan from incomes of other household enterprises, she would not demand a repeat loan and would drop out from the organization, unless, of course, she finds other benefits from the association with the organization. The CARD reports an average drop-out rate of 23%, but this occurs mostly in the first year after the members are recognized. The drop-out rate was unusually high in Marinduque branch, moderate in Masbate, Mindoro and Palanas branches and fairly low in San Pablo, Bay, Dolores and Milagros branches. The drop-out rates drop substantially after the members have taken the second loan.
CARD reports a recovery of credit within due time at over 98%. During the course of the survey the investigators were asked to check the passbook of the borrower and record the number of overdue instalments not repaid at the time of the survey. No overdue instalment was found in any of the cases. Our field investigation thus supports the CARD report on the recovery of loans.
The rate of recovery of interest charged on the loan is another indirect indicator of the financial viability of the enterprise. It will be seen that 89% of the interest due on the first loan and 84% for the repeat loans have already been recovered. For housing loan, however, the repayment rate was 68%.
It will be seen that nearly 97% of the financial loan has been reportedly invested in the enterprise, which generated 163 days of employment during a year on the average for the CARD member, and another 84 days for other members of the households, generating a yearly gross income of P34,550 (P2,879 per month). The contribution of the credit-financed activity to net household income is estimated at P26,884 per annum. The labor productivity is Pl09 per day, about 36% higher than the wage rate prevailing in the mrket. The rate of return on investment is estimated at 117%, which is substantially higher than the effective rate of interest (46%) charged by CARD on the loan outstanding with the member. Thus, the enterprise financed with microcredit is highly financially viable whatever indicator we use.
The rate of return varies significantly across the branches under study. The return is the lowest for the center in the Milagros branch in Masbate. The labor productivity estimated for this branch is lower than the wage rate and the rate of return on capital is negative. But the activity adds P15,803 to the household income. Milagros is a new branch, and as such the amount of financial loan and the capital employed in the enterprise is small. This is a highly poverty stricken area and the CARD members have very little employment opportunities in the labor market in the locality. Inspite of the low return, the members value the CARD credit highly, because it helps them increase household income by reducing underemployment of family workers. During the course of our investigation we found the members of this center were highly motivated and interested in participating in the credit program.
The rate of return is found to be the highest for the center studied for the Masbate branch. The size of financial loan is the highest for the center, and the borrowers have put up large amount of their own fund in the enterprises financed with CARD credit. An average enterprise generated 249 days of employment for the borrower, and another 91 days of employment for the husband and/or other family members, and contributed P46,000 additional income for the household. The labor productivity is almost 68% higher than the wage rate and the rate of return of capital is 210%. Obviously the borrowers in this center should have no difficulty in repaying the loan. The economic performance of the members in the Masbate center is better compared to the respondents under San Pablo and Bay branches, although the latter are older members of CARD and have taken more loans. This is presumably because of the higher incidence of poverty and the lack of alternative employment and income earning opportunities in Masbate, compared to San Pablo and Bay which have well-developed infrastructure facilities and are near to Metro Manila. For San Pablo, the share of equity in total capital employed in the enterprise is very low. The number of days labor used in the enterprise is substantially lower for both San Pablo and Bay, compared to the centers in Masbate and Milagros.
The findings show that in general the financial viability of the enterprises gets stronger with longer association of the members with the credit program. The labor productivity in enterprises run by the new borrowers (P62 per day) is, in fact, lower than the wage rate, and the rate of return on capital is negative when the cost of family labor is imputed by the market wage rate. The members who have received more than two loans have had substantially higher levels of income and employment from the CARD-financed activity. The rate of return on capital is 117% for members who have already taken 3 to 4 loans, and 144% for those with more than 4 loans.
The most common activities financed with the credit are trading agricultural produce, hog raising, retail (sari-sari) store, fish drying and trading, food vending and fishing. Although hog raising is the most popular activity undertaken with the credit, the capital used was typically small and hence the contribution to household income was low, although labor productivity is substantially high. It is often undertaken as a secondary activity using only a part of the loan. The low absorptive capacity of capital in this activity may be due to the problem of marketing of the produce. The trading activities generate very high returns to both labor and capital and contributes substantially to raising household incomes. Fish drying, sari-sari store and trading agricultural produce generate a net return to capital at more than 250%.
The findings presented above amply demonstrate that if microcredit is properly utilized, the financial viability of the enterprise poses no problem. The challenge is how to ensure proper utilization of the loan and recovering the credit from the additional income accruing to the borrowers. The Grameen model of intensive interactions of the bank workers with the borrowers and developing group solidarity and exerting peer pressure through informal organization of the members, are appropriate institutional innovations in this context. The group and the center function as institutions to ensure mutual accountability. The credibility of the group and center as a whole and future benefits in terms of new loans of a larger size are in jeopardy if one member breaks the credit discipline or he/she does not properly utilize the loan and defaults on loan repayments. So, the individual is kept in line by a considerable amount of pressure from other members of the organization. The existence of a well functioning organization thus acts as the collateral for the bank loan.
The recovery of the loan is facilitated by another institutional innovation of the Grameen, the procedure of collecting the repayments in large numbers of small regular instalments. In a poor household there is always a compulsion of utilizing whatever additional income is generated to satisfy the unmet basic needs. It is difficult for such households to accumulate savings for repaying the loan through large-size instalments. The key to ensuring almost 100% recovery of loans lies in collecting repayments in weekly instalments.
Finally, we should not undermine the importance the Grameen model attaches to appropriate training and orientation to both prospective borrowers and the bank worker for the successful implementation of the microcredit program. The key to the success of the Grameen model is the orientation, approach and human qualities inculcated in the bank workers through a training program based largely on 'learning by doing', that is, through observation of and participation in the on-going activities. This helps them to understand the philosophy and approach of the Grameen model of empowering the poor women through access to credit, to develop qualities required for inspiring trust and confidence in the target group, and to derive satisfaction in serving the cause of the distressed humanity. The bank worker in turn motivates the target group, earns their confidence through the hard work devoted to their service, and convinces them of the need to follow credit discipline. The intensive training before conducting bank business with the borrowers contributes, to achieving this objective. Capital Accumulation The most direct effect of the microcredit is on accumulation of capital, both working and fixed. As the loan is repaid in small instalments every week, it is easy for a borrower to pay the instalment from the income leaving the capital intact. A member is expected to have a larger amount of capital when taking a repeat loan than at the time of becoming a member. CARD provides a repeat loan in substantially larger amounts. Thus, it is possible for the borrower to divert some credit or incremental income for making medium and long-term investments, such as purchase of cattle or acquisition of machinery, tools and equipment. The accumulation of these assets will contribute to increasing productivity of enterprises other than those financed with the CARD loan.
Since the borrower gets a larger amount with every repeat loan, the borrowed capital was higher for longtime borrowers compared to the newer ones. But more significantly, the contribution from own sources was also substantially higher for the older borrowers. The value of livestock holding and the accumulation of capital in machinery, tools and equipment, went up substantially as the number of loans taken from CARD increased. The difference is found highly statistically significant, except for the value of livestock holding. The investment for improvement in housing did not increase much with larger loans for borrowers who contracted up to four loans. However, long time borrowers are inclined to invest substantially higher amounts on housing improvements.
The main objective of the microcredit program is to create employment opportunities for the vast underutilized labor resource by undertaking economic activities on a self-employed basis. Any evaluation of the microcredit program should thus quantify its impact on employment generation. However, as mentioned earlier, the effect of CARD loans on generating new employment is difficult to quantify accurately without conducting a costly and time-consuming regular employment survey throughout the year for the CARD members and a comparable control group. Instead, we asked respondents in the borrower survey to report for each family worker the number of months employed during the year, the number of days employed in a month and the number of hours employed in a day, for both credit financed and other economic activities. From this somewhat imprecise information, it is possible to estimate standard eight-hour days of employment for different members of the household. There has been a large increase in employment for both wife and the husband in the credit financed activity and the difference is found statistically significant. The increase in employment for the wife in other economic activities was only marginal, and statistically insignificant. For the husband the employment effect was in fact negative. It seems that in poor households the husband is forced to overwork in low-productive activities under the pressure to earn a subsistence income. With additional income earned by the wife from the credit financed activity, he can afford to enjoy some leisure. This is the classic example of the backward bending supply curve of labor mentioned in economics literature.
The positive effect of higher employment and capital accumulated would obviously be reflected in higher incomes. The annual income from loan financed activity was 1.9 times higher for households who already contracted three to four loans, and 3.5 times higher for older borrowers compared to new borrowers. There was also a significant increase in income from other economic activities with the increase in the number of loans. The positive impact of credit on income is shown by the statistically significant regression coefficient of borrowed capital. The value of the coefficient suggest that one peso of CARD loan generates a gross income of 3.03 pesos, that is, a rate of return of more than 200%. The results of the econometric analysis confirms the conclusion that the microcredit provided by CARD has had a positive impact on income of the borrowing households.
Total operating expenses of CARD increased from P5.8 million in 1994 to P10.4 million in 1996. As a ratio of total assets, the expenses declined from 39% in 1994 to 32% in 1996. The decline in cost is the result of improvement in operational efficiency over time. The number of active borrowers per bank worker increased from 148 in 1994 to 179 by March 1997, and the amount of loans and savings per bank worker increased from P327,000 to P800,000 during this period.
The personnel cost account for 61% of the total cost and the cost of loan fund is 8.5%. The Grameen model emphasizes heavily the training of the prospective borrowers and the bank workers. CARD was able to keep the cost of this human resource development activity at a relatively low level of about 5.3% of the total cost. The cost on this account is, however, expected to go up with the expansion of the program.
The effective rate of interest to the borrower is the interest charged per unit of outstanding loans, as this is the amount available to supplement the capital employed in the enterprise. The personnel cost accounted for 33% of the amount of outstanding loan, and total cost 54% against the 46% rate of interest earned on this performing asset. It should be noted here that the cost of funds amounts to only 4.6% of the outstanding loan. For judging economic viability we need to estimate the opportunity cost of the loan fund rather than actual cost which includes grants and concessionary funds. If CARD has borrowed the entire loanable fund from the high cost source, this cost would rise to 12%. At this cost of loan fund, the operating expenses would increase to 61.4% of the amount of outstanding loan, and the loss on account of loan operation would increase to 25% of total expenses.
It should, however, be recognized that during the expansion phase of the credit program new branches would be opened every year. These branches will take time to reach full operation, but they have a fixed start up cost. For older branches, the cost of operation may be lower, as these branches should be able to expand their business without adding much to the total cost. Branch level cost function estimated for Grameen Bank branches in Bangladesh suggested existence of economies of scale in the operation of the credit program.
In order to see the extent to which CARD branches reap economies of scale with age, detailed information was obtained on the amount of business and the structure of cost for the eight branches under operation till the end of 1996. Four old branches¾San Pablo, Bay, Dolores and Marinduque¾had reached an average business size of about P 3.03 million of outstanding loans. With this size of business, the personnel cost comes to 18.1% of the amount of outstanding loan, cost of fund 8.6%, and other expenses 6.2%. If all branches operated at this level, the overhead cost on account of the head office would come down to 12.3%. The total cost of loan operation thus would be about 45.2%, almost equal to 46% rate of interest earned on the amount of outstanding loan. The loss that CARD incurs is thus on account of branches which are yet to mature.
CARD has been implementing since 1990, a slightly modified Grameen approach to delivery of microcredit for alleviation of poverty. It adopts the essential features of Grameen, such as targeting poor women as the clientele, organizing borrowers in small homogeneous groups to develop group solidarity and peer pressure to ensure effective utilization and recovery of loans; collecting the principal in small regular weekly instalments so that the repayment does not put pressure on low-income households; developing collective funds with compulsory savings from borrowers for their mutual benefit to cope with financial crisis and saving them from the clutches of usurious moneylenders at times of emergency; and promoting social development of members using credit as an entry point. The modification has been in the area of training of borrowers and bank workers, and in the management and utilization of the collective fund that suit the unique lifestyle and economic conditions of the low-income Filipino.
By March 1997, CARD has mobilized through 13 branches, over 7000 members and disbursed P 82.3 million, of which 76% has already been recovered. The amount of outstanding loans with borrowers has reached P20 million, and savings in the members' collective funds at P l1 million. A survey of 133 borrowers selected from four branches, conducted for this evaluation, shows that CARD has largely succeeded in reaching low-income households with credit. Nearly 70% of its borrowers have no access to land and have very poor housing worth less than P25,000, and they received a share of loans proportional to their numbers. Only 13% of the CARD borrowers have college level education, and 9% had landholding of over one hectare. The average size of a loan taken by a borrower was P9,500, of which P4,940 was financial loan used for running enterprises on a self-employed basis. The most common enterprises are trading agricultural produce, hog-raising, retail stores, fishing, fish drying and trading, and food vending. The average labor productivity in enterprises financed with the loan was P107 per day, 34% higher than the market wage rate of P80 per day. The rate of return on capital was 117%, compared to 46% rate of interest charged by CARD on the amount of outstanding loan. These enterprises add P 2240 per month to household income. Employment, income and labor productivity increased with the number of repeat loans taken from CARD.
lnspite of the high rate of interest charged on the loan, CARD has not yet been able to cover its operating expenses, because of the high cost of operation of this intensively supervised credit program. In 1996, the cost of administration was 29% of the amount of loans disbursed, and 54% of the amount of outstanding loans with the borrowers, which led the institution to incur a financial loss of 17% of total expenses. It is however found that a branch that reaches a business size of P3.0 million of outstanding loan can recover its cost from the interest earned. It takes four to five years for a branch to achieve financial viability.
It is natural for any Grameen replication to incur losses during the period of expansion and consolidation, due to the high start-up cost of opening new branches and it takes four to five years to reach the break even level. CARD has so far covered the loss by mobilizing small amount of grants from sympathetic donors and drawing on available low-cost sources of fund. But it is a key constraint to the expansion of its operation.
Since microcredit has been effective in improving the livelihood of the low-income households, the government should mobilize adequate donor support on behalf of the replicators to reduce the cost of loan funds, so that they could minimize the financial losses. The microcredit institutions should encourage borrowers to undertake small scale production activities through subcontracting arrangements with large-scale business enterprises who could benefit from the low opportunity cost of labor for the borrower households. This would help increase absorptive capacity of capital and reduce the time needed by branches to achieve financial viability. The government should not control the rate of interest charged by replicators in microcredit operations. Without high rate of interest the microcredit operators would not be able to expand the operation on a large enough scale to have a significant impact on poverty alleviation.* Concluded.