The Car Allowance Rebate System (CARS) was a fiscal stimulus program intended to boost auto sales by providing consumers with incentives to replace old, low-mileage vehicles with new, fuel-efficient models.  Commonly referred to as “Cash for Clunkers,” the 2009 program allowed individuals to trade in vehicles that were less than 25 years old, in drivable condition, and had a combined city and highway fuel-economy rating of 18 mpg or less.  A government voucher (up to $4,500) could then be applied to the purchase price of a new vehicle with a fuel-economy rating of at least 22 mpg.  The nearly 677,000 “clunker” trade-ins were scrapped, preventing them from entering the used car market.

Previous studies have examined the impact of the CARS program on the economy (Mian and Sufi; QJE 2012) and the environment (Li, Linn, & Spiller; JEEM 2013).  These studies find that the CARS program led to large increases in automobile sales during the two months of the program.  Most of the induced purchases may have been “borrowed” from purchases that would have occurred in the near future, however, and estimated reductions in CO₂ emissions are modest.  While CARS does not seem to have had an overwhelming effect on economic growth or environmental quality, researchers have overlooked the potential public safety benefits generated by the program.

Over three-quarters of CARS trade-in vehicles were manufactured prior to 1999.  Newer automobiles are considerably safer than the aged vehicles the CARS program removed from roadways.  Cars sold in the U.S. have been required to have airbags on both the driver and passenger sides since model year 1999.  Second generation airbags, which are safer for children, became standard in 2001, and a three-point safety belt (both shoulder and lap belt) in the center rear seat became standard in 2007.  

We expect that there are two mechanisms by which the removal of clunkers from roadways may have led to reductions in traffic fatalities.  First, as described above, the CARS program removed aged vehicles which may not have been subject to the same safety protocols required of newer automobiles.  Second, older cars are subject to more natural wear.  Vehicle age is positively associated with vehicle-failure-related crashes, and this risk may be exacerbated in states without vehicle safety inspection programs.

To determine the effect of the CARS program on traffic fatalities, we pursue a difference-in-difference strategy.  A baseline DD model compares across two dimensions: 1) rates of traffic fatalities pre- and post-CARS and 2) rates of traffic fatalities for vehicles eligible for CARS vouchers (“clunkers”) and those not.  A DDD model then introduces a third comparison: rates of traffic fatalities in states with required vehicle safety inspections and those with none.

Initial DD/DDD models use detailed information on U.S. traffic fatalities from the NHTSA’s Fatality Analysis Report System.  These national analyses are supplemented with a state-specific DD model.  Complete accident report data from the state of Wisconsin, allows us to consider accident-related injuries as an additional traffic safety outcome.