Easy Credit, Uneven Evidence: A Scoping Review of Digital Consumer Credit, Repayment, and Household Financial Wellbeing in Southeast Asia

Abstract

Purpose: Digital consumer credit, chiefly buy-now-pay-later (BNPL or PayLater) services and app-based online loans, has spread quickly across Southeast Asia, and concern about borrower harm has grown with it. This scoping review maps the empirical evidence on how consumer digital credit relates to borrowers' repayment, debt burden, financial wellbeing, and debt-related stress in the region; identifies which borrower characteristics and consumer-protection factors that evidence examines; and locates its principal gaps.

Design: Fifteen SciSpace and Google Scholar exports generated on 18 September 2026 (1,114 records; 547 after deduplication) were reconstructed and re-audited, including 47 records an earlier ranking cap had omitted, and a supplementary semantic search on 21 September 2026 added 57 new records. Reports were assessed against explicit population, concept, and context criteria, and data were charted into five outcome domains without statistical pooling.

Findings: Thirty-two reports were included: 27 from Indonesia, three from Malaysia, and two from the Philippines. No eligible report from Vietnam, Thailand, Singapore, or the other ASEAN members was retrieved. Twenty-seven of the 31 dated reports appeared in 2025 or 2026. Twenty reports were cross-sectional surveys of modest, usually location-specific samples (median n = 207), and eight were qualitative or purposive-sample studies. No report used administrative or credit-bureau repayment records, a longitudinal panel, or a quasi-experimental design, and none evaluated the effect of a consumer-protection measure. Much of the quantitative evidence measures perceived risk, risky-borrowing propensity, or subjective wellbeing rather than observed arrears or default. Qualitative studies consistently describe repeated borrowing, the normalisation of debt, family strain, and psychological distress. Financial literacy was the borrower characteristic examined most often, yet its association with borrowing risk was not consistently protective.

Originality and limitations: The review reports its AI-assisted search and its audit transparently, including 18 reports whose full text could not be verified and 426 earlier title-and-abstract exclusions that were not re-audited. The evidence supports a descriptive map rather than conclusions about causal effects, and it identifies the data infrastructure regulators in the region would need to assess digital-credit harm with confidence.