When Privacy Protects but Excludes: The Hidden Costs of Data Restrictions in Digital Lending

Why It Matters

The study shows that well‑intentioned privacy rules can unintentionally deepen financial exclusion for the very groups digital lending aims to serve, while also eroding fintech profitability.

Key Takeaways

• Android users applied 26% more after call‑record ban.
• Approval rates for Android borrowers fell 16‑18 points.
• Low‑income, young, first‑time, marginalized borrowers faced larger rejections.
• Rejected Android users 2.2 points less likely to secure formal credit.
• Lender profits fell roughly 21‑23% after privacy rule removed CDR enforcement.

Pulse Analysis

Digital lenders have relied on alternative data—call logs, mobile usage, social networks—to extend credit to borrowers lacking traditional credit histories. In emerging markets like India, such data not only inform underwriting but also serve as post‑disbursement enforcement, enabling lenders to pressure delinquent borrowers through their social circles. When Google’s 2019 policy blocked Android apps from accessing call detail records, it created a natural experiment that isolates the impact of privacy regulation on both the demand for and supply of fintech loans.

The immediate reaction was a pronounced demand shift: Android borrowers submitted 26% more applications, reflecting a sizable "privacy tax" they were willing to pay to keep their call data private. Lacking the enforcement lever, the lender responded by tightening screening criteria, slashing approval rates by up to 18 points while keeping interest rates stable. Default rates remained flat, indicating that the lender substituted quantity for quality, excluding borrowers who would have repaid under the previous social‑pressure model. The contraction hit vulnerable segments hardest—low‑income, younger, first‑time, and caste‑disadvantaged applicants saw the steepest declines.

Beyond the short‑run loan flow, the study uncovers a lasting exclusion effect. Rejected Android users were 2.2 percentage points less likely to secure any formal credit over the next four years, a 6% relative drop. This “FinTech ladder” loss underscores that an initial loan often serves as a gateway to broader financial inclusion. Policymakers must therefore weigh privacy gains against regressive credit outcomes, designing data‑protection frameworks that preserve essential enforcement mechanisms or offer alternative safeguards to avoid re‑creating barriers for the underserved.