Crash tests yield data, safer autos.

The Insurance Institute for Highway Safety, based in Arlington, Va., broad backing from the automobile insurance industry. No question our support has been very, very strong," said Brian O'Neill, the institute's president. "We have most of the major insurers supporting us, and they do so with enthusiasm. CEOs of those companies have served on my board."

It is little wonder that the institute has such strong industry support, given the wide publicity its work receives and the undeniable effect the work has had on highway safety. The most visible and well-known benefit has been the improvement in vehicles as a result of the institute's high-speed frontal-offset crash program--the one in which it crashes vehicles into specially designed barriers at 40 mph and films and measures the effects on test dummies inside. "When we started the program in 1995, there were very few good performers," said O'Neill of the vehicles tested. "In the last two years, we have not had a bad performer."

The test results helped push manufacturers to build stronger occupant compartments, O'Neill said, which has resulted in significantly lower mortality rates on highways and fewer injuries to the legs, which are in the part of the compartment that is first and most affected by a frontal crash. Leg fractures have cost insurers about $3 billion a year, but leg injuries are on the way down, O'Neill said.

To protect its objectivity, the institute buys vehicles for its crash tests right off dealers' lots. Of the institute's $12.4 million budget, it spends nearly $1 million a year for the purchase of these vehicles, O'Neill said. Opened in 1995, the facility and property cost the institute $8 million, but it has improved the institute's credibility and clout.

The crash tests also led to the wider use of air bags. "We and some insurers fought the fight for a long time," said O'Neill. "Without our efforts and Allstate's and State Farm's, air bags would have come much later."

The institute's affiliate organization, the Highway Loss Data Institute, gathers, processes and publishes data on insurance losses for each kind of vehicle.

Some of the institute's crash-test work has influenced lawmakers to take action. The institute helped years ago to develop a seat-belt enforcement campaign in North Carolina," Click it or ticket." Since then, 35 states have adopted the seat-belt program with the help of federal funding, and national seat belt use rose to 79% in 2003 from 75% in 2002.

Similarly, the institute vividly documented what happens to unrestrained children in crashes, leading to nationwide child-restraint laws. O'Neill said about 90% of children now are strapped in, as opposed to only 10% when the institute first became involved.

The institute's work has led to graduated licensing programs for teenagers, which typically restrict such new licensees to driving in daylight and with no more than two teen passengers. More recently, the institute has touted programs that install cameras at intersections to deter drivers from running red lights. O'Neill said these programs have reduced right-angle crashes by 60% to 70% where they are in use. They are only beginning to spread around the country, he added.

In work dating back to the 1980s, the institute demonstrated that raising the legal drinking age would reduce drunken-driving rates among teens. All states have since adopted 21 as the drinking age; the institute estimates the higher age has saved 21,000 teen-age lives and prevented significant numbers of serious injuries.

In the next few years, the institute hopes to persuade automakers to offer greater side-impact protection through air bags as standard equipment. With whiplash claims from rear-end crashes still near $8 billion a year, it would like to see continued improvement in head-restraint designs. And it would like to see better bumpers. Not only are most current bumpers ineffective at preventing damage in low-speed crashes, but they don't line up between high-tiding sport utility vehicles and lower-riding cars. "We're working with international research groups supported by insurers to address the underride/override problem," said O'Neill.

The crash-test facility has led to better relationships with automakers, according to O'Neill. "Relationships have been very positive," he said. "They work very closely with us, and we share all of our information with them." Manufacturers have asked the institute to play a role in solving the underride/override crash problem, for example, by joining the Alliance of Automobile Manufacturers in developing voluntary standards. "They ask us to play those roles because they know our credibility is so high," O'Neill said.

Better bumpers appear to be more a political than an engineering problem. O'Neill said style is a higher priority for automakers than effective bumpers. He also said that automakers earn significant amounts from the sale of parts to replace those damaged in crashes.