Syndicate content

Banking the unbanked: Measuring the success of JDY

Sumit Agarwal's picture

There is a big debate about the role of financial markets and products in shaping consumer welfare and real economic activity. In developed economies, there is an increasing discussion that financial sector may have become inefficiently large and products offered to households may have become excessively complex. In contrast, in many developing countries, like India, there has been a significant push to increase the usage of financial products — to “complete” the market.

This article is based on an extensive scientific study that evaluates the Pradhan Mantri Jan Dhan Yojna (“JDY”) launched in India on August 28, 2014. Our study has two modest objectives. First, we document the initial uptake and subsequent usage of banking services — that includes a savings account, overdraft facilities, and insurance benefits — by the unbanked targeted by the program. We compare the usage patterns of banking services of households who got access to banking under JDY with similar households who already had access to banking services before the program. Second, we exploit the regional variation in financial access to explore how expanding access to financial services is related to broader outcomes such as GDP growth, lending, consumption expenditure, retail commodity prices and house prices.

Access to a bank account allows consumers to earn interest on their savings and provides incentives to save more. Savings in the bank account could help circumvent behavioral biases that would otherwise have caused them to spend this money. Allowing access to a bank account reduces transaction costs of transferring money to family for subsistence and saving needs. In the absence of a financial inclusion initiatives like JDY banks may not provide these services to their customers for profitability reasons.

We begin by documenting substantial outreach of the program. In particular, the program led to a large increase in the number of households having access to the formal banking services. The number of accounts steadily increased at a rate of 14% new accounts per month since the start of the program. We also find that the average monthly balance maintained in JDY accounts is INR 482 about 60% of the rural poverty line in India.

Along the usage margin we find that approximately 81% of the new consumers do not deposit any money after account opening in the first six months since the account opening. About 12% of individuals perform one deposit transaction while only 7% perform two or more deposit transactions. The statistics are qualitatively similar for cash withdrawals, with approximately 87% of the sample not withdrawing any cash after opening the account, about 5% withdrawing cash only once and 8% withdrawing cash two or more terms. However, households banked under JDY do perform inward and outward remittances. Approximately 34% (21%) of individuals receive (send) money in their account via inward (outward) remittance during the first six months since the account opening. Examining the frequency of transactions, we find that 17% (15%) of individuals receive inward remittance only once during our sample period, while about 17% (8%) receive (send) remittance two or more times

Overall, our micro-evidence suggests that there was substantial uptake by households under JDY. Moreover, both savings and transactions go up over time for individuals that are banked under the program. This evidence is consistent with learning theories that predict an increase in usage over time as individuals gain familiarity with banking services. The initial usage is infrequent and concentrated among a subset of the consumers with stronger intensity among married account holders.

Next, we exploit spatial (regional) variation in implementation of this program to investigate how access to consumer savings accounts is related to broader economic outcomes such as lending and local GDP growth. To do this, we construct four ex-ante measures of JDY program exposure: (i) number of adults per unit bank branch in a region – this captures the extent of bank branch penetration, (ii) fraction of bank-branches owned by state-owned banks in a region -- since privately owned banks are more likely to open branches in higher income areas with greater financial inclusion, (iii) fraction of unbanked households in a region – this captures the extant level of banking access in each area and (iv) a comprehensive financial inclusion index obtained from CRISIL that uses three parameters as inputs to the index: bank branch penetration, deposit penetration and credit penetration. Higher value of all four measures indicates lower level of financial inclusion. We compare changes in the regional GDP growth, lending growth, consumer expenditure, consumer prices and house prices in regions with greater ex-ante program exposure relative to regions with lower program exposure around program implementation.

We begin our regional analysis by verifying that our ex ante measures of regional JDY exposure in a region before the program indeed correlates with the subsequent intensity of treatment from the program. We observe that there is a strong positive association between our ex-ante exposure measures with both the number of JDY accounts opened and the total amount deposited in these accounts. Next we examine whether JDY is associated with an increase in bank lending. In districts with high ex-ante exposure to JDY, using aggregate data provided by the central bank of India, we observe an increase in aggregate lending in areas with greater ex-ante JDY exposure. We verify these effects are present in our micro data and find an increase in both the number of new loans granted and the amount of loans granted in regions with greater JDY exposure relative to those with lower exposure. We find that the total amount of new deposits brought under JDY is small relative to overall deposits in the banks before the program. In particular, the INR 460 billion deposited in JDY accounts is a mere 0.06% of the pre-JDY deposits in the banking sector. Thus, it is unlikely that the additional lending in more exposed regions reflects relaxation of bank financial constraints due to new deposit inflow. Rather, our findings suggest that JDY may have allowed banks to meet the unmet demand for credit for some households that did not have prior access to formal banking products.

We also examine the impact of JDY on a number of other macroeconomic outcomes at the regional level. First, we do not observe economically significant change in the GDP growth rate in more affected areas. However, given our near term focus it is possible that the overall impact of the program on GDP growth rate will manifest itself over the longer-term as more and more individuals gradually start using these services. Moreover, an increase in lending associated with the program implementation that we document may also require time to affect GDP. Second, we observe an increase in consumption expenditure (proxied by debit card transactions). However, this measure should be interpreted with caution since it does not capture the changes in expenditures financed with cash and may simply reflect relative shift of expenditure activity to JDY accounts from cash transactions. Third, we find some evidence suggesting that the program was associated with an increase in investments. Finally, we do not observe any significant relative change in inflation rate in more exposed areas. This suggests that one of the common concerns that the program may have led to substantially higher price level may be unwarranted at least in the near term.

Our research has shown that financial inclusion program like the JDY can have positive effects both at the micro and macro level. While it is too early to know the long term effects of such programs but in the short term, JDY has been welfare enhancing by providing banking services to the poor.

Comments

Add new comment