SBBLost propertyReport a loss

What we hold ourselves to

Every number on this site has a source, or it is labelled an assumption. That is not a posture — it is a test that turns the build red.

The widely repeated claim that “reporting within 30 minutes gives over 70% recovery” could not be sourced anywhere. It therefore does not appear on this site.

Established

~130 000

Items handed in to the operator’s lost-property service in a year

watson (2024) — Fundgegenstände im Zug: SBB erhöht Preise für verlorene Dinge

Established

~60%

Share of handed-in items reunited with their owner

What this does not support: This is the share of items that REACH the service. It says nothing about items never handed in.

SWI swissinfo.ch (2024) — Der Fundus der SBB-Fundsachen gleicht einer Wundertüte

Established

~180 000 / ~120 000

Loss reports filed each year, against items actually handed in

What this does not support: The two figures are counted differently and are not a clean subtraction. They bound the gap; they do not measure it.

RUBICON — Lost property service — Nova Find

Established

40%

Strangers who contacted the owner of a wallet containing no money

Science (2019) — Cohn, Maréchal, Tannenbaum & Zünd — Civic honesty around the globe

Established

51%

Strangers who contacted the owner when the wallet contained money

What this does not support: Return rates RISE with the value inside — the opposite of what self-interest predicts. Honesty is not the scarce resource.

Science (2019) — Cohn, Maréchal, Tannenbaum & Zünd — Civic honesty around the globe

Established

72%

Strangers who contacted the owner when the wallet held a large sum

Science (2019) — Cohn, Maréchal, Tannenbaum & Zünd — Civic honesty around the globe

Established

Every wallet in the experiment carried a business card with the owner’s email — the finding measures what people do when a channel to the owner exists

What this does not support: An item left on a seat carries no such card. This is the most load-bearing detail of the study for this product, and the one most often dropped when it is cited.

Science (2019) — Cohn, Maréchal, Tannenbaum & Zünd — Civic honesty around the globe

Established

The authors attribute returning to altruistic concern plus an aversion to seeing oneself as a thief, both of which strengthen as the loss to the owner grows

Science (2019) — Cohn, Maréchal, Tannenbaum & Zünd — Civic honesty around the globe

Established

Paying for prosocial behaviour can reduce it — money crowds out the motive it tries to buy

What this does not support: Independently, a Swiss transport operator may not claim a finder’s reward (VPB Art. 77 Abs. 2, SR 745.11). The legal bar comes first; this explains why removing it would still be a mistake.

American Economic Review (1997) — Frey & Oberholzer-Gee — The Cost of Price Incentives

Measured here

6 704 / 9 034

Real journeys and calls for one operating day, distilled from the published timetable feed

What this does not support: One operating day, not a full year. Enough to make the trip picker real, not enough to model demand.

Open Data Platform Mobility Switzerland (2024) — GTFS timetable feed (opentransportdata.swiss)

Assumption — not measured

Assumption — not measured

Reporting while still aboard should raise the chance an item is recovered, because the vehicle reaches its terminus within 10–30 minutes, turnaround cleaning follows within 5–15, and another passenger may take the item at any point

What this does not support: Reasoned from mechanism, NOT measured. No published figure links time-to-report to recovery rate; the widely repeated “report within 30 minutes” claim has no traceable source. The product instruments time-to-report so this can eventually be answered rather than asserted.

Sources