Editing Seat belt (section)

===Risk compensation===
Some have proposed that the number of deaths was influenced by the development of [[risk compensation]], which says that drivers adjust their behavior in response to the increased sense of personal safety wearing a seat belt provides.

In one trial subjects were asked to drive [[go-kart]]s around a track under various conditions. It was found that subjects who started driving unbelted drove consistently faster when subsequently belted.<ref>{{Cite journal |vauthors=Streff FM, Geller ES |title=An experimental test of risk compensation: between-subject versus within-subject analyses |journal=Accident Analysis & Prevention |volume=20 |issue=4 |pages=277–87 |year=1988 |pmid=3415759 |doi=10.1016/0001-4575(88)90055-3 |url= https://deepblue.lib.umich.edu/bitstream/2027.42/27199/1/0000202.pdf |hdl=2027.42/27199 |hdl-access=free }}</ref> Similarly, a study of habitual non-seat belt wearers driving in freeway conditions found evidence that they had adapted to  use by adopting higher driving speeds and closer following distances.<ref>{{cite journal |vauthors=Janssen W |title=Seat-belt wearing and driving behavior: an instrumented-vehicle study |journal=Accident Analysis & Prevention |volume=26 |issue=2 |pages=249–61 |year=1994 |pmid=8198694 |doi = 10.1016/0001-4575(94)90095-7}}</ref>

A 2001 analysis of U.S. crash data aimed to establish the effects of  legislation on driving fatalities<ref name=CohenLeinav/> and found that previous estimates of seat belt effectiveness had been significantly overstated.

According to the analysis, seat belts decreased fatalities by 1.35% for each 10% increase in seat belt use. The study controlled for endogenous motivations of seat belt use, because that creates an artificial correlation between seat belt use and fatalities, leading to the conclusion that seat belts cause fatalities. For example, drivers in high-risk areas are more likely to use seat belts and are more likely to be in accidents, creating a non-causal correlation between seat belt use and mortality. After accounting for the endogeneity of seat belt usage, Cohen and Einav found no evidence that the risk compensation effect makes seat belt-wearing drivers more dangerous, a finding at variance with other research.