Data Privacy Concerns Influence Financial Behaviors in India, Kenya

New research from India adds to the growing body of evidence produced by CGAP that customers' concerns about data privacy influence how they interact with financial services to a greater extent than has often been recognized.

Over the past two years, CGAP has been working to understand how much low-income customers care about data privacy and whether their concerns change their behaviors. In 2019, our initial research showed that, when given a hypothetical choice between different loans, low-income customers generally selected higher interest rate loans if they had stronger privacy protections. But these were just hypothetical choices. In our second round of research, we conducted experiments in Kenya and India to see whether privacy concerns influenced real-life financial decisions.

In Kenya, we found that 52 percent of low-income customers selected the costlier of two loan options because of its stronger privacy protections. As we pointed out when we shared the results back in July, the results were striking when considering the financial hardship many people were already enduring in the early months of the COVID-19 pandemic. More striking still was the fact that the average participant said they were willing to pay an interest rate roughly twice as high for better privacy protections.

Our experiments in India corroborate these findings and go a step further by enabling us to compare pre- and post-COVID-19 behaviors.

To carry out the experiments, we worked with research and design firm Pensaar Design and partnered with the money transfer and payments business Eko. Eko caters to low- and moderate-income Indians by providing them with a cash-in point for transfers and payments through a network of over 150,000 small retailers. Our pilot focused on Eko’s internal remittance product, which enables customers to make a cash deposit at a retail location that a recipient can collect in cash from any Eko agent. In exchange for this service, remittance senders pay a transaction fee depending on the amount being remitted.

We offered remittance senders the following choice:

• Option 1: The status quo, where they paid the regular fee for the remittance. This option came with Eko’s standard data protection and privacy features.
• Option 2: An Rs 20 (about $0.25) discount in their transaction fee in exchange for sharing personal data like their phone number and location to be used for marketing purposes.

Although the discount may appear small, it was a significant percentage of the fee. About two-thirds of the customers remitted less than Rs 7,000 ($95) with fees ranging from Rs 30 and Rs 80 ($0.55 to $1.10), meaning that the discount was anywhere from 25 to 66 percent of the fee. The median transfer amount was Rs 4,000 ($55), which carries a fee of roughly Rs 60 ($0.80).

We started this work before the COVID-19 lockdowns began in India. A couple of months into the lockdown, we restarted the experiment. Traffic at stores decreased significantly during the pandemic. Some of the customers who worked in cities and sent money to family members in rural areas had gone back to their hometowns. But we were able to collect a sample of 210 customers before the lockdowns and 632 during the lockdowns.
 

Although more customers accepted the discount during the pandemic, most customers still rejected it. Geographic differences could help explain why more customers accepted the discount. The second sample included customers from North India, who were more likely to take the discount (we do not have pre-COVID-19 data from the region for comparison).

However, customer interviews seem to suggest that, beyond geographic differences, changes in people’s financial situation played a role. The primary reason that customers gave for taking the discount was their need for the extra money. Customers said things like: “I will take the discount as I have no income now”; “I don’t mind. Don’t have money in my account anyway. I live hand to mouth”; and “I can buy 1 KG of vegetables with Rs 20.”

A second group of customers expressed their willingness to share their data because they believed it would give them access to interesting product offers. Customers in this group made comments like: “This has advantages from both sides. I get a discount and I will also get to know about new offers.” We observed similar responses in Kenya, where people believed that selecting a higher privacy loan would reduce their chances of receiving future loan offers.

Finally, a smaller group of customers stated that they were not concerned about data misuse.

We also surveyed the customers that declined the discount. Some said they did not want to receive marketing calls. Others expressed concern about fraud and data misuse. They made statements like: “Just for some discount, I cannot risk losing money from my account” and “No, I don’t mess around with my documents.” Finally, others expressed a general anxiety about sharing their data. Their comments included: “I am scared of all this, don't know when what can happen”; “Something or the other happens with poor people. I don’t want to fall into these traps”; and “No brother, I am a poor man. I don’t know these things.”

The fears expressed by those who felt vulnerable could help explain why people who transferred the smallest amounts were the least likely to take the discount. In a sense, this was a counterintuitive result. Percentage-wise, the discount on fees was more significant for the smallest transactions.
 

Our results from this pilot are consistent with our earlier research: poor people care about their data privacy and place a financial value on it. We hope these results help financial services providers see the importance of having sound data privacy and protection policies and how this could help attract customers in a competitive environment. We also think this evidence is relevant for policy makers. The extent to which people care about their privacy and feel vulnerable to fraud strengthens the case for data privacy regulation.