Do Defaults Save Lives?¶
Johnson, E. J., & Goldstein, D. (2003). Do Defaults Save Lives?. Science, 302(5649), 1338-1339.
Cited by¶
3 citations across 3 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Mandatory vs. Default Norms
- Madrian and Shea (2001) documented this empirically in their landmark "power of suggestion" study, showing that switching a 401(k) plan from opt-in to opt-out enrollment raised participation rates from roughly 49% to over 86% while concentrating contributors at the default contribution rate and asset allocation; Johnson and Goldstein (2003) found analogous effects for organ-donation consent, where presumed-consent (opt-out) countries exhibited effective consent rates above 85% while opt-in countries clustered below 30%.
This sourceCross-country comparison demonstrating that presumed-consent (opt-out) organ-donation regimes produce effective consent rates above 85% while opt-in regimes cluster below 30%, despite formal availability of opt-out in both cases.
- Madrian and Shea (2001) documented this empirically in their landmark "power of suggestion" study, showing that switching a 401(k) plan from opt-in to opt-out enrollment raised participation rates from roughly 49% to over 86% while concentrating contributors at the default contribution rate and asset allocation; Johnson and Goldstein (2003) found analogous effects for organ-donation consent, where presumed-consent (opt-out) countries exhibited effective consent rates above 85% while opt-in countries clustered below 30%.
Domain-specific¶
- Default Effect
- Johnson and Goldstein's (2003) cross-national analysis of organ-donation consent rates found that countries using opt-out defaults (donation is the default; you must actively decline) had effective consent rates of roughly 85–99%, while opt-in countries had rates of roughly 4–27%, the two distributions not overlapping — a gap the paper attributes to default assignment on the strength of its accompanying randomized experiment rather than to underlying preference differences
This sourceJohnson and Goldstein's comparison of effective organ-donation consent rates across eleven European countries, in which the opt-in and opt-out distributions do not overlap, with the causal reading resting on the paper's accompanying randomised experiment rather than on the country comparison alone. The paper itself — Johnson and Goldstein's Science study pairing a cross-national comparison of organ-donation consent rates with a randomised online default manipulation. The effective-consent-rate ranges in the paper's Figure 1: 4.25–27.5% across the four opt-in countries and 85.9–99.98% across the seven opt-out countries, two distributions with no overlap. The Germany/Austria pair from the paper's Figure 1 — ~12% against ~99.98% effective consent in neighbouring populations differing only in the default — with the causal reading resting on the paper's accompanying randomised experiment rather than on the country comparison alone.
- Johnson and Goldstein's (2003) cross-national analysis of organ-donation consent rates found that countries using opt-out defaults (donation is the default; you must actively decline) had effective consent rates of roughly 85–99%, while opt-in countries had rates of roughly 4–27%, the two distributions not overlapping — a gap the paper attributes to default assignment on the strength of its accompanying randomized experiment rather than to underlying preference differences
Mechanisms¶
- Default Label Review
- Its strength is that it attacks the strongest form of hidden norm — the default effect, where people disproportionately keep whatever is preselected
This sourceShows that changing the preselected no-action default sharply changes organ-donor consent choices.
- Its strength is that it attacks the strongest form of hidden norm — the default effect, where people disproportionately keep whatever is preselected
Verification¶
This reference passed the adversarial substantiation pipeline: it was checked to exist and to support the claim it is attached to. See how references were verified.
Registry ID ref:b32c3b937f55 · see in the full table