Choice architecture is the design of the decision environment, including its defaults, feedback, and framing. In Choice Architecture, Thaler, Sunstein and Balz (2010) found that new lane markings cut crashes 36 percent in the six months after installation. The comparison ran from September 2006 to March 2007, against the same six-month period the year before. A related example shows Illinois's mandated-choice organ-donation policy produced a 60 percent sign-up rate, versus 38 percent nationally.
What the Study Found
The new lane markings reduced crashes 36 percent in the six months after installation, compared with the same period a year earlier. Adding beacons and signs to the lines cut crashes a further 47 percent over the next six-month period. Illinois's mandated-choice organ-donation question produced a 60 percent sign-up rate, versus a 38 percent national average. A Johns Hopkins checklist cut the ten-day line-infection rate from eleven percent to zero. Over the next 15 months it prevented 43 infections and 8 deaths, saving 2 million dollars.
"With an eye on these nudges, choice architects can improve the outcomes for their Human users."
Thaler, Sunstein and Balz (2010), Choice Architecture.
Methodology
This chapter is a conceptual essay, not a single empirical study. It draws on secondary reports and case examples to illustrate six choice-architecture principles. Reported figures come from outside sources cited in the text, such as city transportation records and hospital studies. No formal dataset, sample size, or statistical model is presented in the chapter.
Key Statistics
| Metric | Finding | Context |
|---|---|---|
| Lake Shore Drive crash reduction (lines only) | 36% fewer crashes | Sept 2006-Mar 2007 vs. Sept 2005-Mar 2006 |
| Lake Shore Drive crash reduction (lines + signs) | 47% fewer crashes | Mar 2007-Aug 2007 vs. Mar 2006-Aug 2006 |
| Illinois organ-donor sign-up rate | 60% vs. 38% national average | Mandated-choice question at license renewal |
| Johns Hopkins line-infection rate | Fell from eleven percent to zero | Ten-day rate after checklist adoption |
| Line infections and deaths prevented, savings | 43 infections and 8 deaths prevented; $2 million saved | 15 months following checklist adoption |
| Anesthesia critical incidents caused by human error | 82% | Vs. equipment failure |
| Tulsa double-sided printing savings | More than $41,000 a year | Default-setting policy change |
Lines Only vs Lines and Signs
| Measure | Lines Only | Lines and Signs |
|---|---|---|
| Crash reduction | 36% | 47% |
| Comparison period | Sept 2006-Mar 2007 vs. Sept 2005-Mar 2006 | Mar 2007-Aug 2007 vs. Mar 2006-Aug 2006 |
Why This Matters
Small, low-cost design changes can measurably shift behavior even when a decision's substance stays the same. Architects who set effective defaults, give timely feedback, or design against predictable errors can shift outcomes without limiting choice. Libertarian paternalism, Thaler and Sunstein's term for this approach, aims to nudge choices without restricting them. For anyone designing forms or policy, some default is unavoidable, and its choice deserves deliberate thought.