Point-of-sale installment credit isn't a payment method — it measurably raises total spending, stacks onto expensive debt, and its costs land hardest on the least able to carry them.
Five studies of buy-now-pay-later / pay-in-4 credit: transaction-data causal estimates (Di Maggio; Maesen; Bian), the debt-on-debt channel (Guttman-Kenney), and the consumer-side record (Akana). Peer-reviewed backbone: Maesen & Ang (J. Marketing) and Guttman-Kenney et al. (JBEF); the working papers are clearly-labeled supplements. Working collection — the synthesis post compares effect sizes, disagreements, and open questions.
NBER Working Paper 30508 · ≈65 cites
BNPL access lifts spending +$130 on impact and ≈$60/week permanently — retail liquidity “sticks where it hits.”
Journal of Marketing 89(3) 13–35 · accepted manuscript (White Rose) · ≈50 cites
When a big US retailer switched on pay-in-4, purchase incidence rose ≈9pp and amounts ≈10% — the button lifts the whole checkout, not just the split purchases.
Journal of Behavioral and Experimental Finance 37, 100788 · ≈90 cites
19.5% of UK credit cards carry BNPL instalments — 0% debt refinanced onto ≈20% APR balances, mostly by younger consumers and deprived areas.
NBER Working Paper 31202 · ≈50 cites
Inside a billion-user wallet, BNPL access lifts spending ≈5.6% yet default stays <1% — when instalments are the only credit in town, most users refuse to revolve.
Federal Reserve Bank of Philadelphia, CFI Discussion Paper 22-02 · ≈19 cites
Users call BNPL convenience, not credit (52.6%) — but heavy users hit payment trouble at 14.7%, and one missed instalment costs $7–10.
Peer-reviewed backbone: Stijn Maesen et al. 2024, Benedict Guttman-Kenney et al. 2023. Core claims in the synthesis stand on these; working papers and reports are kept as clearly-labeled supplements.
Candidates that belong here but have no legally fetchable open-access full text yet. Listed honestly rather than forced in — hand me a legal copy and they get the full treatment.
Field experiment: installment options raise purchase intent and willingness-to-pay by shrinking the pain of paying. The cleanest psychology complement to the transaction-data papers.
Blocked: No fetchable open-access copy — SSRN and the SMU repository both bot-blocked automated access; INFORMS version paywalled.
Documents debt stacking across five providers — the multi-app blind spot every transaction study names but can't see.
Blocked: files.consumerfinance.gov blocks automated retrieval; grab manually and it slots straight in.
Fed analysis of who uses BNPL and how constraints interact with overspending, debt accumulation, and default risk.
Blocked: kansascityfed.org unreachable from our network (connection timeout); no repository mirror indexed.
Nationally representative US survey: lack of emergency savings and income volatility predict BNPL adoption — the distress leg of the story.
Blocked: ScienceDirect (hybrid OA) returns 403 to automated fetch; no repository copy indexed.
Merchant-side transaction DiD — same design family as Maesen & Ang, so it reads as a replication datum rather than new territory. Held here under the no-rehash rule: it only becomes a member if the collection ever grows a dedicated merchant-effect section.
Blocked: Publisher PDF bot-walled; the indexed repository mirror (Tampere) is proof-of-work gated.