1. INTRODUCTION AND MOTIVATION

Lack of access to the formal credit facility through formal banking and financial institutions has been one of the major hurdles faced by the poor people of the developing countries. There are two main reasons widely discussed. Firstly, the commercial banks and institutions think that the loan demanded by the poor people are in very small amount and it is not economical for the banks to grant these loans. Secondly, these poor people often failed to provide any collateral, this makes their loan requirements risky, and conventional banks and institutions avoid entertaining these types of loans (Secondi, 2008). The solution to these problems was provided by microfinance programs. The primary objective of microfinance programs around the world is to reduce the poverty by providing small loans to the neglected poor people without the condition of collateral.

Microfinance has gained a considerable appreciation over the past three decades due to its mechanism of providing the credit access to the neglected poor people. In pursuit of the fight against poverty, small loans were provided by the microfinance institutions (MFIs) to the poor people so that they may utilize them to establish small businesses or expand their existing business and achieve self sufficiency. However, there has been considerable differences in the rate of interest charged on these loans by the various MFIs, repayment rates and level of self sufficiency achieved by the borrowers around the different regions of the developing countries (Ahlin et al 2011).

MFIs normally use group lending methodology to expand the credit access to the poors. This methodology helps the microfinance clients to keep a check on the proper utilization of the loan amount and also allowing only those individuals to be the part of the group who possess a good moral and financial reputation in their society. Eventually, much of the lender’s responsibility of the supervision is shifted on to the group members. This helps the group members to effectively control the problems arising from incomplete information on the financial health of the individuals intending to join the group (Armendariz and Morduch, 2005). It has been viewed that the group loan given under joint liability lending contracts play an effective role in enforcing the contract conditions of loan repayments through peer monitoring by the group members. Under the joint/ group liability contracts, each group member is responsible for the repayment of the loan in case of any member reports default. Therefore, each group member act as guarantor for the others. It had been argued by various researchers that in a joint liability contracts the group members are well informed about their fellow group members compared to the MFIs or banks. Therefore, this peer monitoring by the group members is much more effective and a cheaper than the MFI’s (or bank’s) monitoring (Varian, 1990: Stiglitz, 1990 and Banerjee et al 1994).

There are a number of studies[1] suggesting that the group liability programs work better in addressing the problem incomplete information faced by the lenders and banks, as much of the responsibility has been shifted to the group members, resultantly lowering the cost faced by the MFIs. Some studies such as Tassel, 199 and Ghatak, 1999 argued that the in group liability loans, the self selection of the group members helps in reducing the problem of incomplete information and can generate higher repayment rates. There are various laboratory experimental studies tried to investigate the group liability loans and effects of peer monitoring on repayment rates. Such as, Cason et al (2009) in their study found that even the monitoring was a costly activity and subjects were required to pay the cost of monitoring, but still the group partners preferred to monitor each other. They also observed that when the cost of monitoring paid by the group members was less than the cost paid by the lender, the group liability programs performed better than the individual liability programs. They also pointed out that in the absence of difference in monitoring cost (among group monitoring cost and lender’s cost), the repayment behavior of the subjects remained almost similar in both group and individual lending cases. In another study conducted by Kono (2006) in Vietman, reported that even in the presence of monitoring and imposition of penalties to the defaulters, the default rates were higher in group liability cases compared to the individual liability cases.

There are few studies explaining the importance of social ties between the group members towards effective monitoring and loan repayment rates. There are a number of ways the social ties can be effective, one of the most important factors is the full information about the group members. This complete information makes it is much easier for the group members to monitor their fellow partners, and if required they can enforce the contract terms (Besley and Coate, 1995, Abbink et al 2006).

However, we were interested in investigating the effect of peer monitoring and choice of microfinance contracts on the loan repayments. The peer monitoring has been discussed extensively in the literature, however, the evidence on the choice of microfinance contracts (i.e. The choice among the interest and non-interest based microfinance contracts) is non existent. We do observe studies addressing the choice among the individual and group lending contracts by the microfinance lenders (Gine and Karlan, 2008, Bhole and Ogden, 2010, Madajewicz, 2011). We conducted laboratory experiments with the students of the University of Sydney, Australia during October, 2013. In our study, there were four treatments and the subjects were offered the loan under the group liability scheme. A pair of two subjects constituted a group and there were 74 groups in total (148 subjects). In the first treatment, subjects can choose to monitor their partner’s repayment actions, then they were required to choose a contract from the given two contracts (PLS and IB). After choosing the contracts, they were recording their repayment actions. In the second treatment, subjects were allowed to go for monitoring, but they were not allowed to choose the contracts rather the experimenter had allocated the both PLS and IB contracts randomly and evenly. With the third treatment, there was no monitoring choice available to the subjects, but they were allowed to choose from the PLS and IB contracts. In the fourth and last treatment, subjects were neither allowed to monitor, nor allowed to choose from the PLS and IB contracts, the contacts were allocated by the experimenter in a randomly and evenly way. There were five rounds for each treatment and after every round the subjects were swapped randomly and it was ensured that the same subject should not be the matched with the other more than once. The subjects were able to identify their group partner through their roll numbers, but they cannot see and talk to him. There was a monitoring fee[2] for all those subjects who were opting to monitor their group partner. Yet, while monitoring they can impose penalty on to their defaulted partner. The decision of monitoring or otherwise was a common knowledge for both the group partners.

This experimental study was conducted to address the following research questions.

1. Whether peer monitoring and peer pressure (imposition of penalty) affect the repayment rates positively?
2. Are there any choice preferences for PLS and IB contracts? How choice of the contracts affects the repayment rates?
3. Are there any gender specific differences in repayment rates?
4. Whether the degree of religiosity affects the repayment behavior of the subjects?

In summary, our results indicated that the monitoring had a significantly positive effect on the loan repayment rates in both with and without choice of contract cases. There were significantly higher proportion of subjects, choosing the PLS contact against the IB contract. Although, there was a high preference for PLS contracts, but at the same time we did not notice any change in repayment rates for both PLS and IB contracts. Moreover, we noticed lower repayment rates in no monitoring with no selection of contract treatment. We did not observe any gender related differences. Also the results revealed that the degree of religiosity was not affecting the repayment behavior of the subject. For the empirical analysis of this experimental work, we used the percentages to analyze the repayment rates. Paired t-test and McNemar’s test were also applied to examine the differences in behavioral responses of subjects across treatments. Finally, we applied panel logit regression methodology to check the significance of monitoring choices against the no monitoring cases and PLS contract against the IB contract in the presence of demographic and other control variables.

The rest of the chapter continues as follows. Section 2 gives a literature overview. Section 3 describes the experimental contracts, design and execution. Results are discussed in Section 4 and in Section 5, we analyze the results and concluded the study.

2. RELATED LITERATURE

Microfinance gained popularity and appreciation over the past three decades for its mechanism of offering small loan to those poor people who were generally excluded from the provision of formal credit facilities. The microfinance use its innovative lending methodology in the presence of incomplete information of the clients and non availability of physical collateral. Generally, lending techniques adopted by the MFIs were individual or group lending. The group lending can be used as screening tools whereas, the joint liability could work as the monitoring device within the group. Nevertheless, the group lending and joint liability programs of micrfinance are considered as efficient instruments in the expansion of credit facility to the poor (Morduch, 1999: Armendariz and Morduch, 2005).

Under the group lending strategy, all the members of the group were provided with the micro loans in their individual capacity. All the group members were then made jointly responsible for the repayment of the each other’s loan in case of default, also future they were denied from any loan facility by the MFIs (Morduch, 1999 : Ghatak and Guinnane, 1999). It had been argued that the success of the group lending strategy has been the outcome of various reasons. Firstly, the screening of the clients by the group members, this helps in selecting the credible and trustworthy persons in the group. Secondly, the group members keep a proper check on the utilization of the loan and also they keep on observing the efforts of each of their partners have exerted in order to make his investment project successful. Lastly, each group member faces a peer pressure from his partner which forces them to comply with the repayment conditions of the loan, along with peer pressure the group members have to face social pressure from their society they were living which enforces the contract conditions of loan repayment. In this way it became possible for the lenders to shift their burden on to the borrowers and the problems occurred due to incomplete information can be handled at the borrowers end (Microcredit Summit Compaign, 2005).

In the existing literature, there are a number of works explaining the positive impacts of microfinance programs for the small line and investment activities. It has also been pointed out that the implications of microfinance programs kept on varying from borrower to borrower and repayment behavior was not same across the borrowers (Cr´epon et al. (2011); Banerjee et al. (2010)). On the other hand, there are studies focusing on the merits and demerits of group/joint-liability loans and stressed that the joint liability loans were successful in increasing the repayment rates (Banerjee et al. (1994); Van Tassel (1999); Wydick (2001)). Entirely the same, it had also been mooted by a number of researchers that the espousal of a group or joint liability loans may lead to increase the risk loving behavior among the borrowers. They may put in a much riskier projects because the repayment responsibilities had been portioned out among the group member. The building of this attitude potentially invites the free-riders (Gine et al. (2010); Fischer (2010); Barboni et al. (2012)). There have been a number of factors addressed in the literature which could involve the repayment behavior of the individuals positively. Such as monitoring either by the lender or by the group members, peer pressure and peer sanctioning, social affiliations among the group members, opportunity for future loans, group size and lower tier of interest rates (Floro and Yotopoulos, 1991, Wydick, 1996, Wydick, 1999).

Gin and Karlan (2010) conducted randomized field experiments in the Philippines over a point of three years found that group liability affected the development and advancement of the microfinance program exerting extreme social pressure on the group members, resulting in discouraging the good willing clients to adopt. They likewise did not notice any difference in repayment rates for both individual liability and group liability cases.

Ghattak and Guinnane (1999) has given a comprehensive analysis on the effect of screening, monitoring and enforcement in group lending. They also observed that the lender’s cost of group screening, monitoring and enforcement can be reduced if they follow the group liability strategy. It has also been argued that in a group or joint liability cases where the default or failure to repay the loan will affect all the group members, including the defaulting member (as everyone in the group has to pay back the loan) and the whole group will stand disqualified for the future borrowing will encourage the group members to monitor each other. In this way, the group lending has the potential to increase the repayment rates. However, it has been viewed that even in the group lending case the monitoring and enforcement through imposition of penalties are costly, but at the same time effective in reducing the lender’s risk, due to shifting of responsibility on to the group members (Stiglitz, 1990: Varian, 1990).

The researchers have consistently argued that the controls or restrictions like peer monitoring by the group members, restriction on further loan payment to the defaulter, social pressure from community through close knit social ties and effective monitoring by the MFIs on the individual as well as group borrowers could potentially increase the chances of loan repayment (Abbink et al., 2006a; Gine and Karlan, 2010; Cassar et al., 2007; Karlan, 2007). In a recent study by Al-Azzam et al., (2012), inferred that peer pressure on the fellow counterpart within a group could result in a higher rate of payment.

[1] For example, Armendariz , 1999 and Rai and Sjostrom, 2004.

[2] See Section 3.2 for details.