Revisiting the Impact of Ignition Interlock Devices

Ignition interlock laws (IILs) are designed to combat drunk driving while still allowing offenders to operate their vehicles. Used as both a punishment and a deterrent, ignition interlock devices are an alternative to unequivocal license suspension for driving under the influence (DUI) offenses. An individual with an ignition interlock device is required to blow into the device to start one’s vehicle and periodically required to blow into the device again during usage. Drivers must bear the cost of device installation and upkeep, which averages from about $900 to $1,300 a year. The harshest of penalties require drivers to have a device installed for two years after conviction. These laws have been implemented in all 50 states; however, provisions vary in the requirements for first-time vs. repeat offenders, length of period, and BAC level triggering implementation. After Iowa in 1997, New Mexico was the next state to implement an IIL in 2005. By 2018, 42 states had passed some type of IIL, with 30 states requiring device installation for all offenders.

Research has found evidence that IILs reduce alcohol-related motor vehicle fatalities and DUI recidivism. However, there is no convincing evidence of their impact on DUI arrests, and no research has linked the laws to drunk driving behavior outside of DUI arrests or other enforcement measures.

We study the effect of U.S. state IILs on three sets of outcomes: motor vehicle fatalities, DUI arrests, and self-reported impaired driving. Following previous studies, we classify a state’s law as “strong” if it requires drivers to have a device installed for all offenders, and “weak” if it only applies when offenders have a BAC above some threshold (typically 0.08). Our analysis provides evidence that strong IILs reduce DUI arrests by nearly 4%, with a particularly strong effect of over 5% among women. However, we find no evidence that these effects yield any decline in the number of motor vehicle fatalities. Moreover, we surprisingly estimate that strong IILs significantly increase the likelihood that individuals who report drinking also report having driven while impaired, which might be attributable to the laws raising awareness that even moderate drinking can lead to driving impairment.