The U.S. Experience with No-Fault Automobile Insurance: A Retrospective


cal costs) from his or her own insurance company. In the United States,
add-on no-fault is an important variation in which an injured party
can recover from his or her own automobile insurance without any
restriction on also filing a tort claim against another driver. Another
important variation, choice, allows individual drivers to choose whether
to accept, in exchange for lower premiums, restrictions on their right
to sue other drivers.
No-fault approaches to automobile insurance were first proposed in the 1920s, modeled after the workers’ compensation no-fault
approach to workplace accidents. For the next 40 years, numerous academic studies decried the use of the tort system to compensate injured
victims of automobile accidents. Commentators focused on the following failings of the tort system:
(1) As a result of the fault standard, many victims were either not
compensated at all or undercompensated. (2) There was a long
delay in providing compensation to injured persons. (3) The seriously injured were often undercompensated while victims with
minor injuries were often overcompensated. (4) The process of
establishing fault created high administrative costs. (5) Victims
and injurers had large incentives to be dishonest in their efforts to
improve their cases. (R. Keeton and O’Connell, 1965, pp. 2–6)
Because it minimized litigation and administrative costs associated with determining who was at fault for an accident, supporters of
no-fault supposed it to be less expensive than the tort system.
Massachusetts passed the nation’s first no-fault automobileinsurance law in 1970, and many other states soon followed. A number
of insurers and consumer groups supported no-fault over the opposition of the trial lawyers, and, for a while, it appeared as though it was
a genuinely superior policy innovation.
Over time, however, dissatisfaction with no-fault grew, primarily because the hoped-for premium-cost reductions never materialized.
Several states repealed no-fault laws and realized premium-cost reductions. Political debate about no-fault increasingly focused solely on the
issue of consumer premium costs, and the other justifications for the
no-fault approach on which its original proponents relied lost political

Summary xv
salience. Many insurers and consumer groups that once supported no-fault as a means of reducing rate increases no longer support it.
We demonstrate that the perception that no-fault auto insurance
claim costs were higher than other auto-insurance systems was largely
accurate. Total injury costs per insured vehicle gradually began to
diverge across systems in the late 1980s, with no-fault becoming substantially more expensive than tort. Whereas injury costs under no-fault were only 12 percent higher in 1987 than those under tort, this
difference had ballooned to 73 percent by 2004. Surprisingly, we also
found that states that restricted lawsuits against other drivers—in an
attempt to reduce costs—actually exhibited higher claim costs than
states that permitted these lawsuits.
Why were no-fault regimes unexpectedly more expensive? We
identify medical costs as a primary contributing factor. Medical treatment in no-fault states was vastly more expensive than in other states.
Controlling for a broad range of personal and accident characteristics,
we demonstrate that claimants in no-fault states are more likely to claim
the use of virtually every type of medical provider, from the emergency
room to chiropractor, and visit each type of provider more frequently
than claimants in other states. We also show that the same medical
care costs more to the auto-insurance system in no-fault states than
in tort states and that most of this cost divergence occurred during
the 1990s. In particular, prior to 1987, medical charges to the auto insurance system for individuals in no-fault states were only slightly
higher (5.7 percent) than for comparably injured individuals in tort
states. However, by 1997, the disparity had grown beyond 40 percent. While we discuss plausible explanations for these trends, further
research is necessary to determine exactly why medical costs in no-fault states grew so dramatically during this period. One possibility is
that no-fault insurance shifts medical costs associated with auto accidents from the first-party health-insurance system to the automobile insurance system.
We also demonstrate that, while no-fault states had lower levels
of litigation activity and devoted a smaller share of payments to noneconomic damages in the 1980s than did tort states, by 2007, the two
systems had largely converged on these characteristics. No-fault has