Loan Rejection Rate for Women-run Business 2 Times Higher than Men. Formal Finance Bias Needs Fixing

India has made strides when it comes to women’s financial inclusion: The World Bank Findex Report finds that women-owned bank accounts have increased from 43 per cent in 2014 to 77 per cent in 2017; the Reserve Bank of India’s (RBI) annual Financial Inclusion Index (FI-Index) has improved by 10.5 points from 2017 to 2021. But it is quickly becoming clear that this access is not manifesting in greater uptake or usage of financial services by women.

Why are Women Still Unable to Access Formal Finance?

The All India Debt and Investment Survey (AIDIS) in 2019 found that 80.7 per cent women in rural areas had a deposit account in banks (against 88.1 per cent men). This proportion is only slightly higher in the case of women in urban areas, where 81.3 per cent of women had a deposit account. Given the expansion of digital finance, expectations for women’s financial inclusion are also riding high. But it is unclear if and how women would harvest the benefits of smartphone-mediated digital finance.

A 2021 survey finds that 25 per cent women (against 41 per cent men) own smartphones in India. Dvara Research’s upcoming survey of 2,719 low-income individuals indicates that 34 per cent women and 55 per cent men own smartphones. Where women do have phones, they use it differently than men. Earlier work by Dvara Research and partners found that women are wary of using mobile phone due to fear of harassment. Moreover, given the limiting social norms around women’s use of mobile phones, most women prefer using the phone within the confines of their homes and are likelier to use the phone to communicate with family and friends. Together, the lack of ownership and anachronistic patterns of usage imply that women do not generate sufficient data and where they do, it looks different from that of men. Further, it is unclear if loan decisioning algorithms account for the differences between male and female applicants’ data trails. This matters because data has the potential to overcome the barriers that women face in accessing finance such as lack of assets or formal financial history. Yet, credit might be denied to women applicants due to lack of bias correction.

The loan rejection rate for women-owned businesses is two times higher than that for men, despite evidence that women are more disciplined than men in repayments. Even when women are not the primary borrowers of formal finance, studies note that they remain responsible for ensuring timely repayments, and, for maintaining social networks to keep channels of informal borrowing open.

ALSO READ | Women@Work: Working Women Up Globally but Down in India. Is There a Way Out?

How Can We Meaningfully Include Women in Formal Finance?

Enhance cash-in-cash-out networks. A 2021 survey of more than 3,500 retail stores revealed that more than 65 per cent of women would prefer dealing in cash as opposed to other methods. As recommended in a 2019 RBI report, cash-in-cash-out (CICO) networks must be made more robust by the use of more women on-ground banking correspondents (BCs) who are at close proximity to women and are able to further connect them to banks and payment systems.

Improve the quality of grievance redress mechanisms (GRM) to improve the stickiness of formal financial services. Effective GRM can increase the trust in formal financial services and inform the user of beneficial features of the product that may otherwise go unnoticed. For women, GRMs must account for lack of mobile phones by providing on-ground touchpoints that can address queries and register complaints with ease. Emerging research conducted in 2020-21 noted that where women customers are called to discuss grievances with respect to their bank accounts, it is often their male counterparts who answer the call, or hand the phone over to them after answering the call. Therefore, GRMs must be designed while keeping women’s limited access to phones in mind.

Acknowledge and correct algorithmic bias of data-driven decisions. Algorithmic models used for financial decisioning can be biased against women due to lower representation of women in the sample or historical bias encoded while labelling training data. Providers must attempt to identify why bias might be caused and mitigate the same accordingly. For instance, in case of a sampling bias, data must be re-weighted at the training stage, so a lower representation has a smaller effect on the decisioning. Similarly, differentiated use-patterns of women’s data points must be acknowledged as a pre-condition of the algorithm to prevent their treatment in the same manner as that of men.

Even as we celebrate the success of an increase in women’s access to financial services, we must not lose sight of the larger goal, i.e., designing financial services that are meaningful for women. The distance between access and usage can only be traversed on the back of gender-intentional financial services.

Anubhutie Singh is a Policy Analyst with the Future of Finance Initiative at Dvara Research. The views expressed in this article are those of the author and do not represent the stand of this publication.

Read all the Latest Opinion News and Breaking News here