In 2001, there were an estimated 26 million people 70 and older living in the United States.¹ Based on self-reported data from a nationally representative sample, approximately 67 percent of them were drivers.² Older people represented approximately 12 percent of the driving age population and about 10 percent of estimated drivers.

From 1995 to 2001, the number of people ages 70 and older increased 9 percent. However, the number of drivers 70 and older grew even more, increasing 15 percent during the same period.^2,3 In 1995, 64 percent of people 70 and older were estimated to be drivers, compared with 68 percent in 2001.

Lyman et al. reported that in 1995, drivers 65 and older represented 17 percent of the driving age population, 13 percent of drivers in fatal crashes, and 8 percent of drivers in all crashes.⁴ The authors estimated that by 2030, drivers 65 and older will represent 25 percent of the driving age population, 25 percent of drivers in fatal crashes, and 16 percent of drivers in all crashes.

Although the number of drivers 70 and older has been increasing, the rates of fatal crashes among these drivers per population have declined since 1997, from 20 fatal crashes per 100,000 people 70 and older in 1997 to 16 per 100,000 in 2006. The number of fatal crashes among drivers 70 and older has declined since 1997 too, from 4,977 in 1997 to 4,265 in 2006.

Number of fatal crashes among drivers 70 and older per 100,000 people 70 and older

Based on 2001-02 data, drivers 70 and older drove 55 percent fewer annual miles, on average, than drivers ages 35-69.² Many older drivers limit certain types of driving such as driving at night, in the rain, or in rush hour or heavy traffic.^5,6,7,8 Some groups of older drivers are more likely to reduce or restrict their driving. These include women^7,8,9,10 and drivers with physical, visual, or cognitive impairments.^{7,8,9,10,11,12} However, many seniors do not self-regulate or adjust their driving,^6,7 even those with high levels of cognitive impairment.¹²

Older drivers have low rates of police-reported crash involvements per capita, but per mile traveled crash rates start increasing for drivers 70 and older and increase markedly after age 80. Some caution should be used when examining crash rates per mile traveled, especially for older drivers. Older drivers generally travel fewer annual miles than most other age groups and, similar to low-mileage drivers of other ages, they tend to accumulate much of their mileage in city driving conditions. In contrast, drivers who accumulate higher annual miles tend to do so on freeways or divided multilane roads, which generally have much lower crash rates per mile traveled than other types of roads. Hence, the elevated crash rates for older drivers when measured per mile traveled may be somewhat inflated due to the type of driving they do.¹³

Number of crashes per capita by driver age, 2006

Number of crashes per mile traveled by driver age, 2001-02

Insurance claims provide a more comprehensive picture of crashes of all severities than police crash reports because police are not notified of many crashes reported to insurers, and police-reported crashes are more likely to involve injury. Property damage liability claims are filed when an at-fault driver damages someone else's property. Collision coverage insures one's own vehicle against loss caused by a collision. Although not as high as for the youngest drivers, property damage liability claims and collision claims per insured vehicle year (for 2002-04 models) start increasing after about age 65, meaning that seniors more often are involved in crashes.¹⁴

Number of collision and property damage liability claims per 1,000 insured vehicle years by rated driver age, 2002-04 models, January 2001 thru November 2004

A study of older drivers' elevated fatal crash rates per mile traveled between 1993 and 1997 revealed that the main factor was not seniors' overinvolvement in crashes but their fragility, defined as the risk of death in a crash.¹⁵ Fragility increased starting around ages 60-64 and continued to rise with increasing age. Similar results are found with more recent data for 2002-06.

Number of passenger vehicle driver deaths per 1,000 drivers involved in police-reported crashes by driver age, 2002-06

Compared with younger drivers, senior drivers are overinvolved in certain types of collisions — angle crashes, overtaking or merging crashes, and especially intersection crashes. The most common error made in senior-involved crashes is failure to yield the right-of-way. Seniors are cited for this error more often than younger drivers.¹⁶ In a recent Institute study of nonfatal crashes occurring at intersections, drivers 80 and older had fewer rear-end crashes than drivers ages 35-54 and 70-79, and both groups of older drivers had more failure-to-yield crashes and fewer ran-off-road crashes than younger drivers. Reasons for older drivers' failure-to-yield crashes varied with age. Compared with younger and older drivers, drivers 70-79 were more likely to see another vehicle but misjudge whether there was time to proceed. Drivers 80 and older predominantly failed to see the other vehicle.¹⁷

Specific physical, cognitive, and visual abilities may decline with advancing age. However, there are large individual differences in the onset and degree of functional impairments, so age alone is not sufficient information to judge driving ability. Still, functional impairments can interfere with driving and may become particularly evident in stressful or challenging driving situations such as merging or changing lanes. Several studies have shown that higher levels of physical, cognitive, or visual impairment among older drivers are associated with increased risk of crash involvement.^11,18,19,20 Many older drivers also take medications, which can impair driving ability at any age but can be especially impairing for an older person.

In terms of fatalities, older drivers are a danger mostly to themselves and their passengers, who also typically are older and thus more vulnerable to injuries. One study found that per licensed driver, drivers 75 and older kill fewer pedestrians, bicyclists, motorcyclists, and occupants in other vehicles compared with drivers ages 30-59. In the fatal crashes of drivers 75 and older, the drivers and their senior passengers were much more likely to be killed than were occupants in other vehicles. ²¹ However, drivers 70 and older have higher insurance liability claims for damage to other vehicles per insured vehicle year than drivers ages 35-69.¹⁴

The goal of driver screening is to identify people at high risk of crash involvement without falsely identifying other drivers who are not at high risk. So far, there are no tests that accomplish this goal with sufficient accuracy. A few states including California and Maryland are studying multitier screening tests. California developed and pilot-tested a three-tier assessment for driver's license renewal for applicants of all ages.²² The first two tiers consist of nondriving assessment tools (e.g., driving knowledge test, cognitive screening, vision tests, observation of obvious physical limitations). Applicants failing both tiers must pass an on-road driving test to renew their licenses. A pilot study found that lower assessment scores in the first two tiers were related to worse performance on the road test but were not consistently related to crash records during the three years prior to renewal.²² A pilot study in Maryland found that drivers who performed poorly on select cognitive measures were approximately 25 percent more likely than other drivers to have a subsequent at-fault crash after controlling for age, gender, and annual mileage.²³ Although these findings suggest a general relationship between some measures of cognitive ability and crash risk, it is not clear whether these tests are sufficiently sensitive to predict future crash risk with the accuracy that would be required to justify suspending license privileges.

A growing number of states have imposed additional requirements for seniors renewing their driver's licenses. The ages at which special regulations are required vary by state but typically begin at ages 65-75. Approximately 20 states have shorter renewal cycles or require in-person renewals after a specified age. In some states vision tests, driving knowledge tests, and/or on-road driving tests are required for drivers older than a specified age. For example, beginning in 2004, Florida requires that drivers 80 and older pass a visual acuity test when renewing their driver's license. An Institute study of this requirement found that 80 percent of those eligible for license renewal attempted to do so, and only 7 percent of drivers were denied license renewal because they failed the vision test. Of those who did not seek renewal, about half said they thought they would fail the vision test.²⁴

US driver licensing renewal procedures for older drivers

Studies have yielded mixed results. Several studies have shown that vision testing among seniors was associated with lower fatal crash rates per population^25,26 and per licensed driver.²⁷ In one study, states with laws requiring in-person driver's license renewal had a 17 percent lower fatality rate per licensed driver among the oldest drivers (i.e., 85 and older), compared with states without such laws. For drivers 65 and older, fatality rates per licensed driver did not differ for states with and without laws for vision testing, road testing, or shortened renewal periods.²⁸ Similarly, an Australian study found that drivers 80 and older in jurisdictions with age-based mandatory medical and/or road tests did not have lower fatal and serious injury crash involvement rates per capita or per licensed driver compared with a jurisdiction without age-based mandatory testing. Some jurisdictions with mandatory age-based testing had significantly higher fatal and serious injury crash rates than the jurisdiction without age-based testing.²⁹ Thus the effects of special regulations for license renewal among older drivers are not well established.

States may impose license restrictions on a driver of any age, but restrictions are imposed mostly on older drivers. No state restricts drivers based on age alone. The goal of restricted licensing is to allow drivers to continue to drive in safer conditions based on individual abilities. Restrictions may address, for example, vehicle speed, geography, and nighttime driving. The restrictions generally are based on evaluations that may be conducted when drivers apply for license renewal. Drivers may be subject to evaluations by licensing agencies based on referrals from police, physicians, or family. States may establish policies for further testing that include vision screening, road tests, knowledge tests, and/or evaluations by medical advisory boards.

There is little evidence of safety benefits from education courses for older drivers, although several organizations offer such courses. National programs include the Driver Safety Program (American Association of Retired Persons), Safe Driving for Mature Operators (American Automobile Association), and Coaching the Mature Driver (National Safety Council).

A challenge in evaluating these courses is to design studies that separate the effects of the course from the effects due to differences in the types of drivers who participate. Drivers who choose to take these courses typically have lower crash rates prior to the course than those who do not participate. None of the evaluations of older driver improvement programs or educational initiatives have found a reduction in subsequent crash risk among participants relative to comparison groups.^{30,31,32,33,34}

Technologies to enhance occupant protection in crashes include seat belt force limiters that limit the force of restraints during a crash and advanced airbags that reduce the likelihood of injuries resulting from airbags. Improved head restraint systems help protect occupants against whiplash and other neck injuries. Some vehicles also provide ergonomic features that may assist older drivers, including adjustable pedals and seat adjusters as well as bigger and brighter displays and controls.

Vehicle technologies intended to prevent crashes may help drivers of all ages. Electronic stability control (ESC), a technology intended to increase a vehicle's stability, has been found to be highly effective in reducing single-vehicle fatal crash involvement risk.³⁵ Other crash avoidance systems include lane departure warning systems and adaptive cruise control, which maintains a set distance from the vehicle ahead. Despite the potential benefits of these technologies for helping to prevent crashes, there is some concern that systems requiring attention or responses from drivers may lead to cognitive overload or distraction from the driving task itself, especially for older drivers.

Much can be done to improve roadway safety. Improving the visibility of road signs and pavement markings through lettering, size, or color can be particularly important for older drivers who may have visual impairments due to macular degeneration, glaucoma, cataracts, or other health factors. Intersections are a particular problem for older drivers, and countermeasures may include adding left-turn lanes and left-turn traffic signals. One study found that low-cost modifications to intersections (e.g., making traffic signals more visible, adding a dedicated left-turn lane) resulted in a 13 percent greater reduction in injury crashes per licensed driver for drivers 65 and older compared with drivers ages 25-64.³⁶

Another approach is to reconfigure existing or new intersections as roundabouts, which reduce vehicle speeds and eliminate some of the most complicated aspects of traditional intersections. In a study of intersections that were converted from stop signs or traffic signals to roundabouts, injury crashes were reduced by 76 percent.³⁷ However, older drivers favor roundabouts somewhat less than younger drivers. In surveys taken at least one year after the construction of new roundabouts in six communities, 65 percent of drivers ages 65 and older favored the roundabouts, compared with 70 percent of drivers 35-64, and 74 percent of drivers 18-34.³⁸ No studies have focused on the effects of roundabouts on crashes among older drivers who may find them difficult to navigate. Adding features to roundabouts such as advanced warning signs and directional signs may encourage older drivers to choose routes with roundabouts as opposed to conventional intersections.³⁹

Better and more widely available alternatives to driving for older people, particularly in rural communities, are needed. Walking may not be feasible in some communities, and friends and family may be needed to provide rides. Other alternatives, although not available in many areas, include public transportation and community-based systems such as van programs and volunteer drivers. In a national telephone survey, public transportation was the usual mode of transportation for 5 percent of all adults 75 and older and for 14 percent of adults 75 and older who did not drive.⁴⁰

In 2006 there were about twice as many pedestrian deaths per population among those 70 and older than among those younger than 70. However, pedestrian nonfatal injury rates per population were about twice as high for people ages 10-20 than for people 70 and older,⁴¹ suggesting that the increased risk of fatality among older pedestrians is due largely to an increased likelihood of death from their injuries.

Countermeasures to reduce injuries and fatalities among pedestrians of all ages may particularly benefit seniors. An Institute review of traffic engineering measures to reduce pedestrian-motor vehicle crashes identified several effective approaches.⁴² One is to reduce traffic speeds by using roundabouts and speed bumps. In European studies, roundabouts were shown to reduce pedestrian crashes by about 75 percent.^43,44

Traffic signals that stop all vehicle traffic for part or all of the pedestrian crossing signal also have been shown to reduce pedestrian crashes. One study found a 50 percent reduction in pedestrian crashes at intersections with such signals, compared with intersections without them.⁴⁵ Other measures include making pedestrians more visible (e.g., increasing intensity of roadway lighting, especially at night) and separating pedestrians from vehicles (e.g., providing sidewalks or refuge islands in the medians of two-way streets).

Another improvement is to extend the time available for pedestrians to cross with a green light. This can be especially beneficial for older pedestrians. A 2000 Institute study found that providing pedestrians a 3-second head start through a leading pedestrian interval (a signal that allows pedestrians to begin crossing before the release of turning vehicles) reduced conflicts between pedestrians and turning vehicles as well as the incidence of pedestrians yielding the right-of-way to turning vehicles.⁴⁶