Knowledge is power! The story of debt literacy and credit usage in rural India

Efforts to promote financial inclusion in developing countries have witnessed a massive scale-up in the past decade, thanks to the World Bank’s promotion of “Finance for All” since 2008 and the liberalization of credit markets across the globe. This has resulted in an increase in the share of the banking population from 51 percent in 2011 to around 69 percent in 2017 as recorded by the Global Findex database. Among developing countries, India deserves special mention as a front runner in setting a new institutional context of high levels of financial inclusion by constituting around 55 percent of the global share. This transition in the financial landscape makes it crucial to understand the factors that would determine the usage of financial services by the newly banked population.

Credit is the most common financial service used in the rural regions of a developing country like India. As institutional and market-driven changes like financial innovation and democratization of credit are recent phenomena in rural regions in India, it is important to identify the crucial factors driving the credit behavior in this population. In my job market paper, we examine how credit usage of rural households in this new institutional context is influenced by debt literacy (the ability to make simple decisions regarding debt by applying basic knowledge about interest compounding to financial choices, as defined by the literature).

Background

With the rollout of national-level initiatives to improve financial inclusion in the country (like the Swabhimaan scheme in 2011 and the Pradhan Mantri Jan Dhan Yojana scheme in 2014), around 322.5 million bank accounts were opened for the unbanked population in India over the past decade. Two states (Kerala and Goa) along with three union territories (Chandigarh, Puducherry, and Lakshadweep) became the first in the country to achieve 100 percent financial inclusion (defined as having at least one bank account per household). In our paper, we focus on Kerala, where we zoom in on rural households and exploit a unique setting that allows us to examine the role of debt literacy in credit behavior.

The role of debt literacy in credit usage: How do we study it?

To examine our research question (see figure 1), we depend on primary data that we hand collected from 600 rural households across three districts in Kerala. We measure credit usage by the debt-to-asset ratio of individuals. To measure debt literacy, we use a set of questions that we developed based on the literature. We complement our analysis with a national survey data on rural households, the NABARD All India Rural Financial Inclusion Survey 2016–17 (collected by the National Bank for Agriculture and Rural Development).

We first explore the role of debt literacy in credit usage using ordinary least squares estimation. However, to infer causality, we have to address the endogeneity of debt literacy highlighted by the literature. Endogeneity of debt literacy would mean that debt literacy is influenced by the usage of credit through reverse causality. To address potential endogeneity, we use instrumental variable regressions. We assess the financial exposure of the respondent based on whether they have heard about formal financial instruments such credit cards. We use this as an instrument for debt literacy because hearing about credit cards can influence debt literacy in a financially included rural region as literature gives evidence that innovative formal farm credit options like Kisan Credit Cards (KCCs) are popular options for availing institutional credit in the rural areas. Our second set of instruments are the household head’s education and age which represent family influence on the debt literacy of respondents. Our identification strategy is based on the absence of any direct influence of financial exposure or family influence on current credit usage, other than through debt literacy. The instruments are validated by Hansen J-statistics and F-statistics from the first stage regression.  In the case of women respondents, we also consider “membership in group activities” as an alternative instrument based on the literature as group participation is known to influence their financial capabilities especially in developing countries. We check the robustness of our results via treatment effects estimation based on inverse probability weighting with regression adjustment. This is a doubly robust method which minimizes misspecification errors in regression estimation of treatment effects based on observational data.

What do we find?

Even in a state that has achieved 100 percent financial inclusion, debt literacy levels are quite low in rural areas. Table 1 shows the summary statistics of the debt literacy levels in our sample from Kerala.

Next, we provide evidence that debt literacy has a positive and significant effect on credit usage, particularly for agricultural households and women. This finding is in contrast with available international evidence that shows a negative effect of debt literacy on the level of credit. Our finding that individuals with higher debt literacy tend to hold more debt underscores the importance of debt literacy in their credit usage. We obtain similar results when we repeat this analysis on the national-level data set.

Policy implications

We provide evidence that debt literacy increases credit usage in rural India, where an unprecedented increase in the level of financial inclusion is underway.  Our contention is that debt literacy acts as an empowerment device, especially for agricultural households and women in rural areas. Higher debt literacy may have helped people in our sample seek loans from banks and other financial service providers.  For instance, those who are more debt literate may be more successful in producing the necessary documentation and completing other complex procedures to avail more formal loans, which are the predominant types of loan in our sample.

Our findings suggest that there is scope for policy-based solutions to improve the usage of formal financial services like bank-based credit by vulnerable sections of the population, that is, farmers and women. Policy could focus on improving their debt literacy to overcome the low usage of financial services, hence enabling them to make use of cheaper institutional credit. Our findings also provide lessons for financial institutions as they could use debt literacy as a factor for credit appraisal. Going beyond know-your-customer standards and income, this can help them to do risk-based pricing and provide more financial services like cash credit and individual loans.

Limitations

One limitation of this study could be that although the instrument validity tests show that our instruments are reliable, we understand that our instruments may not be fully exogenous to the extent there could be unobserved confounders (such as psychological factors and social norms) that we are unable to capture. Moreover, the generalizability of our findings would require conducting similar studies in other regions.

Remya Tressa (@RemyaTressa) is a PhD candidate in economics at the Indian Institute of Management Kozhikode (IIMK). More about her research can be found here.