While there are many contradictions, inconsistencies and subjective interpretations in the CoC assessments which seem to draw the attention away from the real issues, we hope the sponsors and developers of the SIDBI-World Bank COCA tool recognize the fact that it far from being a reliable and valid psychometric measure of code of conduct assessments

While previous articles looked at the award of free points in the code of conduct assessments (COCA) reports and the peculiar findings arising therein, this article looks at issues of subjectivity and inconsistency in CoC assessments with tangible examples.

1.    It has been observed that Bandhan has only recently incorporated a cash flow analysis of clients’ households in its loan forms. Rightfully Bandhan did not receive a high score on O (Observance).

The case of Equitas is contrary to this. One month before the COC assessment took place, Equitas is said to have rigorously changed its policies (i.e., one month after the AP crisis started). 

 This COCA judgement appears to be based on soft factors: the assessment is relying on the interpretation of the management and the management 'feels' that is has been able to overcome the problems. Even though agents exist in areas where Equitas operates (MFIN-NCAER study: Here's the proof that microfinance agents are thriving in Tamil Nadu) and problems due to agents have not been not been ruled out by M2i COCA report itself, Equitas still receives a high score (7.6 out of 9) on O (observance) as shown in Table above Thus, we have the cases of Bandhan and Equitas, both of which had changed their policies (Bandhan on Loan Appriasal and Equitas on CoT) after the crisis and stated that they were implementing the new policies now. On one hand, Bandhan was rightfully given a low score whereas Equitas got a high score-even though (as per the COCA reports) both the MFIs had not fully overcome the observed problems. There seems to be a great deal of subjectivity and inconsistency here with regard to the COCA tool

2. With regard to pricing, I realise that transparency in pricing is the major concern in this COCA tool. Given so, I find it contradicting that SKDRDP, which charges by far the lowest interest rates, gets the lowest score because "it accounts for and communicates interest on a flat rate basis". (http://www.sidbi.com/micro/COCA%20SKDRDP.pdf)  This is the main reason for providing exceptionally low grading whereas the above report states that SKDRDP's "Annualised Percentage Rate is among the lowest for the MFIs in India." This needs to be noted carefully. 

The implication of this is that MFIs that mention their interest rates on a declining balance but charge higher (APR) rates of interest are rewarded with higher (and better) scores. And MFIs who state their interest rates on a flat basis but charge lower (APR) rates are penalised with lower scores. I am not sure that this is fair at all!

That transparency on pricing is not always understood by the clients becomes evident from the Equitas COCA report. Equitas is known for its transparency in pricing and it communicates these on a declining balance basis. The Equitas COCA report however states that "the level of awareness of the clients about the effective interest rates and method of application was found to be low." (http://www.sidbi.in/Micro/COCA%20Equitas.pdf). This being the case, what then is the great benefit from communicating interest rates to clients on a declining basis?

Without question, clients are surely better off with lower (annualised percentage) interest rates (stated on a flat basis) as compared to higher rates of interest presented to them on a declining balance.

3. Bandhan received a high score on Staff Conduct (97%) (please look at Section 1: Scores and facts in the report page no 2 given at http://www.sidbi.com/micro/Bandhan.pdf).

The COCA report says: "None of the clients reported any misconduct by the staff. However, the internal audits were not found covering staff conduct issues adequately and explicitly."  (Page no 13)

The key question here is how can M2i come to an overall score of 97% on staff conduct when internal audits do not cover staff conduct? Moreover, M2i allotted the maximum score of 11 points out of 11 points on Observance (please look at Annex 1, page no 15 in the above report) even though a serious concern (internal audits were not found covering staff conduct issues adequately and explicitly) was observed at Bandhan. When staff conduct is not under the purview of the internal audit team, what is the guarantee that staff conduct is indeed good and as per policy? And given such a situation, how can Bandhan be given the maximum score for observance?

4. Also related to staff conduct is the case of Equitas. Equitas received a 100% score (35 points) on staff conduct (please look at Section 1: Scores and facts in the report page no 2 given at http://www.sidbi.in/Micro/COCA%20Equitas.pdf). It may be good to realise that staff conduct is the indicator that accounts for about 28% of the total COCA score (please look at the above report, page no 22) and it is a very important indicator.

The report says that in the past,

"Equitas has faced problems pertaining to involvement of unauthorised agents in the client origination process, particularly in some of the branches of Chennai, primarily on account of high sales targets of the SOs (sales officers) and weaker controls". (page no 7)

Thus the SOs were dealing with the clients through the unauthorised agents. So maybe the relationship of the SOs with the agents may have been good, but with whom did the ultimate clients deal and how good was that relationship?

Further, the report says that Equitas has done away with targets and related incentives for enrolment of new clients from "November 2010 onwards in order to reduce the likelihood of the SOs getting involved with the unauthorised agents. In M2i's opinion, despite all sincere efforts, some of these agents may still be in existence." (page no 7) 

Given that, in M2i's own assessment, agents may still be (and perhaps are) in existence, surely, it cannot be expected that they (these agents) will follow any policies on staff conduct. So how can M2i justify a 100% score for Equitas on this aspect? It seems that not-so-good practices here are rewarded with the maximum score!

5. Ujjivan has experienced problems with unauthorised agents and the COCA report says that:

 "there are instances of presence of unauthorised agents and influential group leaders as has been pointed out in many Internal Audit reports of Ujjivan. The assessment team also observed many instances where the same center leader has been the leader for many years and that the same person is leader of more than one MFI center/group." (Page no 8) (http://www.sidbi.com/micro/COCAUjjivan.pdf)

How is it possible that Ujjivan receives a 92% score on client origination (please look at Section 1: Scores and facts in the above report, page no 2) while these serious concerns are found to be in existence? It is also inconsistent that these serious problems have not been addressed as an area of improvement. M2i considers enhancing client awareness of the declining balance interest rates and improving the dissemination of grievance redressal mechanisms to be more important for Ujjivan.

6. Last but not the least, SKDRDP received a low score on client origination amongst others because "the organization does not have a policy on avoidance of unauthorized agents." (http://www.sidbi.com/micro/COCA%20SKDRDP.pdf)

The same report also says that "at present, SKDRDP mostly operates in areas where other MFIs are not operating. In the current assessment no evidence could be found to suggest that unauthorized agents are affecting the operations of the organization in any significant manner." (Page no 7)

In other assessments like those of Basix, Equitas and Ujjivan, the scores for client origination were high because the MFIs have a policy not to deal with unauthorised agents. However in practice, and especially in the recent past, these three NBFC-MFIs have been actively using agents. The COCA reports also do not rule out the use of agents by some of the MFIs (at the time of the assessments). That being the case, why is their score for Client Origination and Targeting (COT) higher than SKDRDP, where no serious evidence with regard to use of agents was found by the M2i team. 

7. And I could go on but it is time to stop...

From the analysis given above, it becomes very clear that there are many contradictions, inconsistencies and subjective interpretations in the CoC assessments and these seem to draw the attention away from the real issues. I hope that the sponsors and developers of the SIDBI-World Bank COCA tool recognize the fact that it far from being a reliable and valid psychometric measure of code of conduct assessments...