Reporting information about individuals in self help groups

by Karuna Krishnaswamy.

The problem

On 14th January 2016, RBI issued a circular for banks to report the financial and non-financial data of the individual members of the Self Help Groups (SHG) that they lend to starting July 2016 to credit bureaus. However, while bank know their groups' credit performance, data on the individual members and internal loans reside, if at all, on paper-based registers maintained by SHGs and cannot be readily accessed. This RBI circular is a reiteration of an earlier one dated 27 June 2014 since banks did not make much progress the first time around. There is a chasm to cross between thinking about one group at a time and thinking about one individual at a time.

In March 2014, there were 74 lakh SHGs, with 42 lakh active borrowers, with loans outstanding of INR 43,000 crore (0.3% of the GDP) and savings of INR 9,900 crore (source). However, there are sector-wide concerns about the quality of the groups. With better data we will know how many groups are active and functioning well, and if loans are being disbursed democratically and not captured by the group leaders. RBI's stated intention is for banks, and MFIs to make an informed decision on lending to SHGs after considering their members' indebtedness and past performance through a credit bureau, monitoring NPAs and for research purposes. NABARD and DFS would like to link SHG members to Aadhaar-based programs like PMJDY to offer a wider range of financial services and for direct benefits transfers. Indeed, this is a first step for banks to graduate SHG members to individual services.

Separately, NABARD's e-Shakti program supports the Digital India program through SHG MIS digitisation as does the National Rural Livelihoods Mission. However, past SHG digitisation efforts have not succeeded because they were expensive and/or operationally difficult to implement. The RBI circular further states that banks may incentivise SHGs to adopt digitisation but that they cannot set any pre-conditions (including opening individual saving accounts) to giving them new loans. This makes it an even more difficult task.

The penalty for not complying with RBI's circular is that the loans to the unreported SHGs will not count as Priority Sector Lending (PSL). Given that the SHG portfolio represents a small fraction of the PSL portfolio of most banks, this penalty may not be high enough.

Gramin Bank of Aryavart (GBA), a large Regional Rural Bank, has successfully piloted a novel agent banking model to, inter alia, service SHGs, that promises to overcome these challenges while offering many other benefits. This article discusses the challenges and how this model that could address them.

Past challenges

Past efforts since 2006 have used a bookkeeping software for SHGs for digitising their information such as a laptop-based MIS by SERP for self-entry, groups paying a computer munshi by Pradan, and NABARD's tablet based software. These efforts have not taken off in a big way due to the following challenges:

1. Funds to finance the hardware and software capital expenditure.
   
2. Hiring and training a computer literate operator who can do the data entry for a large number of groups to make the user fees affordable for the groups without sempiternal support of the SHPI; finding people to do the training, and servicing of the hardware and software.
   
3. Getting SHGs to adopt the MIS and to pay for it.
   
4. Given the low data quality, ensuring data correctness, aggregating, verifying and transmitting (especially in areas without GPRS connectivity) to the bank.
   
5. Securing the cooperation of the banks when it is a low priority for them.

The Bank Sakhi model with dual authentication

Based on a successful demonstration project by GBA, supported by the Rural Financial Institutions Programme (a cooperation program of NABARD and GIZ), NABARD issued a circular on 14 January 2016 encouraging banks to appoint SHG members as their Business Correspondents Agents (BCA) for financial inclusion. An indirect benefit of this approach, that is discussed below, is that individual group member data can be collected at low extra cost.

In this project, SHG members, were trained and appointed as BCAs, popularly called Bank Sakhis, to provide doorstep individual financial services as well as to conduct SHG transactions such as savings deposits, loan disbursements and repayments using a conventional PoS device. They were contracted by Bartronics Pvt. Ltd., the Corporate Business Correspondent (CBC) of the bank. The Bank Sakhis were supported by the SHPI, the Rajiv Gandhi Mahila Vikas Pariyojana and their Block Level Federations who were contracted as a sub-BC to Bartronics.

The key innovation of this model is the use of a dual authentication software that resides on the PoS (NABARD and GIZ have a video explaining this). This software enables BCAs to conduct SHG transactions that require two members' approval to be authenticated. Previously, SHG loan disbursements and loan repayments required two leaders from each group to come to the bank branch, fill in the forms along with two signatures. However, now, SHG transactions are conducted like any other BSBDA transaction by a BCA. An electronic debit to the SHG's group account is authorised using fingerprint-based authentication of the two group leaders by the PoS device. These transactions are recorded in real-time at the bank's core banking system (CBS). The Bank Sakhi does cash-in /cash-out to the group account and its individual members and later squares off at the bank branch. Each Bank Sakhi in UP has a service area of 3.5 villages on average and has five SHGs in her command area due to the low SHG penetrations there.

Banks can now have accurate information in their CBS of the loan disbursal, utilisation and repayment performance of individual members along with their Aadhaar IDs obtained through the KYC process of opening the individual and group accounts. Hence, the information to be collected by banks and reported to RBI can be accomplished through this model at low extra cost.

Assessment of the approach

This approach can mitigate the challenges previously identified with the bookkeeping software approach.

Community adoption : An evaluation of the pilot project revealed that although voluntary, all SHGs leaders reported preferring to transact with the Bank Sakhi. This is because the doorstep service saves two members of each group a (bi) monthly visit to the bank branch which otherwise would cost them a day's time in addition to transportation and incidentals. They do not report unacceptable delays to the group meetings due to the technology.

The CBC has the operational experience in servicing PoS technology issues in the field, while the Block Level Associations of the SHG federations are well-positioned to train Bank Sakhis.
This service is more likely to be adopted widely than a bookkeeping software since it solves a specific pain-point of SHGs and their leaders and the incentives of the bank, BCA, group leader and members are well-aligned.

In any case, banks can strongly encourage SHG transactions to be routed through Bank Sakhis since dual authentication ensures that the BCA cannot de-fraud the members. Banks simply need to open an individual saving account (as envisaged by PMJDY) for their loans to land in and for repayments, since only transfers from individual accounts to the group account can be recorded by the PoS and not bulk cash deposits for the entire group, into the group account.

Being group members themselves, Bank Sakhis are very familiar with SHG processes whereas BCAs who are not members are less likely to be embraced by other SHGs. In terms of reliability of the service, in our project after two years, only three Bank Sakhis had quit out of 78 (despite low wages).

Costs and benefits to stakeholders: Banks would have access to SHG transactions data at a low cost. Consider a stylised SHG that has 12 members that deposit mandatory savings twice a month, and of which 6 members have a running loan that they repay once a month. At a fee of INR 3 per transaction payable to the CBC, this costs the bank a reasonable INR 96 per group per month, while the INR 7 lakh for developing the dual authentication software is the only additional capital expenditure. The other costs of device, training and servicing are part of the BCA expenses which are incurred anyway by the bank as part of its FI mandates. Furthermore, NABARD's DFIBT department offers subsidies to encourage use of Bank Sakhis; it will reimburse INR 20,000 to the banks for each PoS device and the wages of a block level coordinator to support the Bank Sakhi. Costs could be further reduced by using smart phones instead of PoS devices.

But the larger benefit to banks is moving the SHG transactions out of their branches and to the Bank Sakhi.

GBA found that migrating SHG transactions from its branch to the Bank Sakhi freed up considerable branch staff time, which could be repurposed for originating and monitoring more profitable credit customers. GBA also reports that the delinquency rates of their SHG portfolios have dropped after the Bank Sakhis were introduced. Financial inclusion through CBCs has not been an attractive business proposition for banks. Enabling Bank Sakhis to service the SHGs' credit portfolio will improve the viability of the BC channel per se.

For members, the INR 96 would be lower than the cost of two leaders going to the bank twice a month. Currently there are no user fees but it is likely that the willingness to pay to not go to the bank branch is higher than to pay for entering SHG transactions into a computer. In the past, there have been concerns that SHG federations may not encourage individual accounts fearing that it may lead to disunity and that the group may fall apart. The Bank Sakhi model has been successfully piloted by two RRBs in two states. The federations in these two sites support the model in spirit and in operations.

In our pilot areas, the number of SHGs per village was very low. In other states such as AP, the number of groups per village is much higher, contributing to a much higher fraction of the Bank Sakhi's income and make it worth her while.

Cooperation by banks: The downside to the banks is low. They are required to fulfil FI, PMJDY and PSL mandates in any case; they only need to appoint Bank Sakhis instead of their conventional BCAs. Bank Sakhis have proven to be more effective in servicing FI customers compared to men and are more tied to the village and are less likely to quit. The main challenge for them is to find qualified SHG members, though this restriction could be relaxed in future to scale faster.

Data quality and reliability: Accuracy of the data is expected to be close to perfect since money is changing hands and both parties will ensure that the amounts entered are correct. Challenges to real-time uploads are in those areas where internet reception is poor.

Human resources and training: This approach reduces the need for the eco-system to hire a separate computer munshi in addition to a BCA and hence the entailing hiring and training costs are reduced as well.

Challenges

One limitation of a PoS device is that it is too unwieldy to install an SHG accounting software on it. This may possibly be addressed by using a tablet PC or even a smart phone, instead, with an accounting software in it and the Bank Sakhi could further play the role of computer munshi if she wants to augment her income.

For this model to work at scale, most of a bank's SHGs should regularly use the Bank Sakhi. Hence, the challenges to this are the same as the challenges to the BC model and to technology such as - availability of internet connectivity; the bank branch should not be too far away for the Bank Sakhi to visit for cash management; competent SHG members should be available in each Gram Panchayat; the Block Level Association or an SHPI should be available for hand-holding support in the early days; and a CBC should be operative in the SHG villages. In our early, experience, none of these have been showstoppers.

However, challenges may crop up in future; a scale-up will not have the abundant project management and oversight that the pilot did; identifying and training Bank Sakhis may take longer than the RBI deadlines; Bank Sakhis may get dissatisfied with their low incomes; the required support from the bank branch managers may not be universally available; funds to conduct campaigns to build awareness and trust in the Bank Sakhis may not be available. However, many of these challenges also apply to the accounting software approach.

Conclusion

In sum, the benefits of this approach are too many to be ignored especially if the banks consider their various mandates in toto and in states with large SHG outreach. Stakeholders are better incentivised to support this model than the bookkeeping model. RBI could support this effort by relaxing its instruction that opening a BSBDA cannot be a pre-condition to loan disbursements, since voluntary community participation is a challenge. However, as with large scale-up of any government program, careful design, and ongoing monitoring and evaluation feeding into design improvements, will maximize the likelihood of effectiveness.


Karuna Krishnaswamy is an evaluation specialist with GIZ, Delhi.