Default effect analysis

Description

Measures how much a pre-selected option drives choice, and why, by comparing outcomes under different defaults or active choice.

Defaults are among the most powerful levers in choice architecture: Madrian and Shea found 401(k) participation rose from about 49% to 86% under automatic enrolment, and organ donor consent differs dramatically between opt-in and opt-out countries. Default effect analysis treats every pre-selected setting as an intervention and measures its effect, using randomised comparison of defaults or active choice where possible, or regression discontinuity and before-after designs where not. It also decomposes the mechanism: inertia and effort, implied endorsement by the designer, and reference-point shifts. Rigour comes from randomisation, measuring long-run outcomes rather than just initial selection, and checking whether people stuck with a default that harms them, which raises an ethical as well as a statistical question.

Attributes

Loss type
Siloed
Effort
Weeks

Steps

  1. Inventory all pre-selected options, pre-filled fields and pre-ticked choices in the flow, and record each one's current uptake rate from logs.
  2. Select defaults with material consequences for users (money, privacy, safety, workload) and state for each the hypothesised mechanism: effort, endorsement or reference point.
  3. Design arms: current default, alternative default and enforced active choice; calculate sample size for the expected difference (default effects are often 20 to 40 percentage points, so a few hundred users per arm frequently suffices).
  4. Randomise users to arms and measure initial selection, later switching, and downstream outcomes over a pre-registered window of at least 30 to 90 days.
  5. Probe mechanism with a short follow-up survey or 6 to 10 interviews asking whether users noticed the default and whether they read it as a recommendation.
  6. Evaluate welfare: check whether outcomes for defaulted users are good for them, choose the default that best serves the majority, and preserve easy opt-out.

Sources

  • Paradox of Choice (Schwartz)
  • Think Bigger (Iyengar)
  • Happiness by Design (Dolan)
  • Methods Atlas (Gopinath)
  • The Inquiry Series, Ch 22 — Choice Architecture, Default Effects & Option Set Design

Also known as

  • Default option testing
  • Opt-in versus opt-out analysis
  • Default nudge evaluation

Categories

  • Analytics & Behaviour
  • Behavioral Science & Economics