PART 2 OF 2 (CH. 3 CONT'D.)
Some facts seem continually to be obscured in the industry's interpretation. Nash provided a belt to hold a reclining passenger in place against the shifting and stopping that would ordinarily be experienced in a moving car. Billboards showed a grandmother sleeping peacefully, held snugly by the belt. It was not constructed, offered or advertised as a belt for collision protection. What are now known as seat belts were not offered by American Motors until the mid-fifties. This reclining-seat "seat belt" was not emphasized in Nash's promotion of the reclining seat option; in fact the belt was completely hidden underneath the seat, and many customers did not even know it was there. There was nothing in the owner's manual about the belt. Nash dropped the feature because it considered it a needless expense. As Ralph Isbrandt, vice president of American Motors, told the Roberts' House subcommittee on Traffic Safety in a 1957 hearing on seat belts, "As we gained experience with the reclining seat, it appeared that this feature actually did not create an increased need for a restraining device."
Gandelot gave further "evidence" of "public apathy" in the small number of letters which General Motors had received from the public about seat belts. He recounted how the seat belts and shoulder harnesses he had tested restricted his ability to reach some of the vehicle controls, rumpled his suit, and gave him aches. He denounced those who were pushing for seat belts as people motivated by "the profit angle."
The arguments General Motors adduced in its opposition to seat belts are less important than the reason for such arguments. The reason is simple: the seat belt is a constant reminder to the motorist of the risk of accident. The seat belt is an emphatic reminder of the second collision, an item that alerts people to expect more safety in the cars they buy. General Motors has never viewed these as desirable expectations to elicit from its customers.
Gandelot and his superior at General Motors, Chayne, watched with skepticism Ford's advertising campaign promoting seat belts as an option for its 1956 models. The public's response to the campaign brought a demand for more seat belts than the company could provide at first. Between September 1955 and January 1956, many Ford purchasers who wanted seat belts could not get them and had to accept delivery of their cars without the belts. Robert McNamara, then vice president of the Ford division, reported in February 1957 that "more than 400,000 seat belts have been sold by Ford since we introduced them," and that no other optional feature "ever caught on so fast."
General Motors was not impressed. About this time, GM's president, Harlow Curtice, had a sharp exchange with Charles Shuman, president of the American Farm Bureau Federation, at a meeting of the President's Committee for Traffic Safety. Shuman wanted to know why the automobile industry as a whole was not offering seat belts as standard equipment. Curtice told him that the idea was impractical and inadvisable.
The Roberts hearings in 1957 brought together expert testimony about the desirability of seat belts as shown in experimental work and accident experience. On the basis of the hearings record, Roberts' special subcommittee on traffic safety concluded that "seat belts, properly manufactured and properly installed, are a valuable safety device, and careful consideration for their use should be given by the motoring public." Charles Chayne appeared at these hearings to repeat the circular argument about the lack of public acceptance or demand for seat belts as a reason for not promoting them.
Gandelot, who was continually called upon to express the General Motors view on the seat belt issue, once told an inquirer, "I delight in living my life each day, realizing that the information I give out is extremely factual." Such a sentiment cannot be faulted; the only difficulty was that GM's chief safety engineer never had any information to give out. While demanding more proof about the value of seat belts, he responded to requests for substantiating his skepticism with answers like this one, made in 1955: "While we certainly have a lot of engineering record films of barrier impact crashes, both normal and high speed, and quite a few simulated impact tests made with a new and very controllable apparatus which we designed and built some time ago, this is all under the classification of engineering data and not for public distribution," He chided his critics in the medical profession by contrasting their lack of knowledge about the seat belt issue with his own "factual view of things," which took into account "only those opinions which have been established on a basis of facts." Yet Gandelot never felt the need to justify the safety of existing vehicle design, however stringent were his standards for those who suggested improvements. In 1954, he offered this astonishing judgment to a physician who was pressing him on the seat belt matter: "Until we have substantially more information I find it difficult to believe that the seat belt can afford the driver any great amount of protection over and above that which is available to him through the medium of the safety-type steering wheel if he has his hands on the wheel and grips the rim sufficiently tight to take advantage of its energy absorption properties and also takes advantage of the shock absorbing action which can be achieved by correct positioning of the feet and legs." A few weeks later he wrote to the same physician, saying that there was very little data available about the effect of seat belts at higher deceleration rates and force values. "This makes me wonder," he wrote, "if, in the public interest, the industry should undertake a fact-finding program. Considering the quantity and type of instrumentation, the anthropomorphic dummies, vehicles and technical personnel required, it would be my guess that such a program would cost upwards of 100 thousand dollars." Gandelot appeared to be turning a long overdue duty of the industry into an act of charity.
General Motors was understandably concerned about the consequences of overt emphasis on safety features as a competitive practice in selling cars. Such an emphasis could only serve to focus public attention on the role of vehicle design in causing injuries during the second collision. Claims by one company that its cars are safer would quicken the interest of federal officials in asking, "How safe is 'safe'?" They might propose that automobiles meet federal safety standards just as trains, ships, and aircraft have been required to do for decades.
It seemed particularly Significant that less than a year after Ford began an unprecedented campaign advertising its "Life Guard Design" ("safety door locks," "safety steering wheels," "safety rear view mirror" as standard equipment, and "crash pads" for instrument panels and seat belts as options) that the Roberts committee opened on July 16, 1956, the first hearings on traffic safety in the history of the United States Congress.
Ford terminated its safety campaign in the spring of 1956 after an internal policy struggle won by those who agreed with the General Motors analysis of the probable unsettling consequences of a vehicle safety campaign. The 1956 Ford finished second to Chevrolet in sales, but its failure to be Dumber one had nothing to do with the Ford safety campaign. [2] Even so, it has since been cited to prove that •safety doesn't sell." Working through the Automobile Manufacturers Association and other industry-constituted committees, General Motors found its views accepted by other domestic automobile makers. Vehicle safety became an industry-wide policy matter rather than an individual company matter.
After 1956, industry seat belt policy entered a period where belts were offered as an extra-cost option but were not widely promoted. While saturation advertising and continual repetition of the sales message are deemed necessary to sell automobiles, seat belts were left to win customers without such communication. The manufacturers then seemed mystified because more car buyers did. not demand this option. Chevrolet general manager Edward Cole said in 1959, "One of the startling problems so far as crash injury is concerned is the utter refusal on the part of the American motorists to be strapped into a seat by a safety belt or a shoulder harness. We have made provision in our cars to attach seat belts properly and we have made seat belts and shoulder harnesses available to our dealers. The fact of the matter is that the sale of these safety features is practically nil, indicating a real disinterest on the part of the public in their own safety."
Before Mr. Cole wrote these words, he might have found that Chevrolets, along with other General Motors cars, presented great obstacles to "attaching seat belts properly." In 1961, C. M. Olsen of the American Society of Safety Engineers commented on the unique problems of installing seat belts on General Motors models of the late fifties: "All four-door GM cars are exceedingly difficult in which to make front seat installations. Removing the sharp wire clips deep down in the front seat construction is a strenuous task -- and somewhat like gynecological surgery in the dark -- but has to be done to insure that the belt is not abraded or cut where the user cannot see the damage being done." Mr. Cole had not explained how shoulder harnesses could be installed in the "hardtop" models featuring doors without a pillar to anchor the harness on, and Olsen offered an obvious insight: "I feel that people will otherwise [in cars without pillars] be reluctant to attempt such a difficult do-it-yourself job, or to slit new car upholstery to get the belts through, or pay the price of having it done properly so the belts will not be damaged in use."
Although they had a long record of success 1n creating a public demand for even the most superficial automotive features, the manufacturers lamented the absence of demand for seat belts while they made it difficult for such a demand ever to materialize. Paul Ackerman, engineering vice president of Chrysler Corporation, said to the Roberts subcommittee in a 1959 hearing, "In considering the question as to whether or not we should provide med and permanent attachments for safety belts, 1 intended to explain that many people have very definite objections to the installation of belts in their cars." John Moore, former director of the Cornell project, provided the answer. "No safety device can be used by the public unless it is first made available to the public."
The first step in the drive for availability was to make seat belts standard equipment on all automobiles. The initiative was taken by the New York State joint legislative committee on motor vehicles and traffic safety under the chairmanship of Senator Edward Speno. The committee decided in 1959 that seat belts must come as "standard factory-installed equipment, just as hydraulic brakes and sealed beam headlights." The following year the committee said, "It is the Committee's opinion that the auto manufacturers will not -- now or in the foreseeable future -- install seat belts as standard equipment in all cars unless forced to do so." The Speno committee then gave the automobile makers an opportunity to disprove its prediction. During the 1960 legislative sessions, automobile industry lobbyists defeated a bill requiring seat belts on all new cars sold in New York.
The following year, Senator Speno decided upon a strategy that would show the absurdity of the industry's position. He filed a bill to require new cars to have anchorage units for belts to facilitate and reduce the cost of installation. These anchorage units were merely threaded holes through the car floor, supported by steel plates which could be punched out during fabrication at no added cost to the car buyer. (At that time, a pair of seat belts cost between thirty and thirty-five dollars, plus about fifteen dollars for the mechanic's work in installing them.) The automobile manufacturers resisted. Speno and a group of legislators and administrators went to Detroit to confront company officials directly. The industry must have thought this was a routine Visit by a legislative committee; the Visitors got the routine tour of company plants in a special bus equipped with a loudspeaker and were given a show of crashing a few castoff vehicles with dummies. The usual points were made by the car makers: if New York passed one statute and other states passed conflicting ones, it would make it impossible for the manufacturers to comply; it is sometimes safer to be ejected from a vehicle than to remain inside; it would cost the consumer more; seat belts would hurt automobile sales. General Motors' Charles Chayne told Senator Speno that car safety is best decided by car makers. "A lot of people come here with ideas," he said. "Roberts came here. Ribicoff came here. They went away."
Speno was not impressed. At a dinner for the visiting committee in the Detroit Athletic Club, he told a group of industry vice presidents that the "comfortable delusion of safety the public gets in your cars is in sharp contrast to the broken bodies these cars cause. You've been showing me the ballpark, gentlemen, but you're not talking to me. I hope you will put in the anchorage units. It will cost you almost nothing. But whether you do or not, we're going to legislate it." He asked for a meeting at four P.M. the following day and indicated that he expected a formal reply. The next morning Mark Bauer of the Automobile Manufacturers Association informed Speno that the industry would provide anchorage units in all 1962 models, but they would like to restrict them to the front seat since such a small proportion of people killed are back seat riders. Speno reluctantly made the concession. It was agreed explicitly that following the afternoon meeting there would be a joint announcement. Bauer told Speno that the industry wanted no public release before the meeting. But early that afternoon, four of the automobile companies sent out press releases announcing that they would provide anchorage units in the coming model year. The industry had avoided the joint announcement and preserved the carefully nurtured fiction that all safety advances are made voluntarily.
Speno went back to Albany and sponsored legislation requiring anchorage units on cars to make sure that there would be no reversal by the automobile manufacturers in the future. The manufacturers opposed the bill, but it was passed. Other states followed New York's example. In 1963 New York, impressed by a Wisconsin law enacted in 1961, passed legislation requiring front. seat belts beginning with all 1965 model cars sold in New York. By this time, the automobile companies, prodded by legislation, were cooperating with the U.S. Public Health Service and voluntary agencies in promoting seat belts. Many government agencies and commercial fleets had installed belts. But the automobile makers were still opposed to standard installation.
The first break in this opposition came from a smaller manufacturer. Early in 1963, Sherwood Egbert, president of Studebaker, announced that his company would install front seat belts on all cars manufactured after February 15, 1963, and contributed this heretical statement: "It is our feeling -- a strong feeling -- that safety measures in motor cars should not come by petition from motorists but that automobile manufacturers should lead in safety equipment."
Under pressure from Speno to begin standard installation before the New York law's effective date of June 30, 1964, the automobile companies finally agreed. In August 1963, they announced that, effective January 1, 1964, they would make front seat belts standard on 1964 passenger cars with list prices adjusted to include the additional cost. Each company alluded to its longstanding interest in safety and seat belts and its gratification for the increasing public acceptance which made such an announcement possible.
Thus the industry rounded out a decade of strenuous opposition before its cars were equipped with a primitive passenger restraint device as standard equipment. The seat belt should have been introduced in the twenties and rendered obsolete by the early fifties, for it is only the first step toward a more rational passenger restraint system which modern technology could develop and perfect for mass production. Such a system ideally would not rely on the active participation of the passenger to take effect; it would be the superior passive safety design which would come into use only when needed, and without active participation of the occupant. It would eliminate the "acceleration overshoot" characteristic of conventional seat belts, which do not prevent the passenger from striking his head or his upper body or both on the corner post, instrument panel, windshield, or header strip. It would also eliminate the "bottoming effect" or the passenger's sliding under, and the backlash or rebound effects.
Protection like this could be achieved by a kind of inflatable air bag restraint which would be actuated to envelop a passenger before a crash. Such a system has been recently experimented with for airplane passenger protection. Both General Motors and Ford did work on a system like this about 1958 but dropped tile inquiry and now refuse even to communicate with outside scientists and engineers interested in this approach to injury prevention. There are a number of general energy-absorption systems that engineering ingenuity could devise to operate either inside or outside tile vehicle.
It has long been recognized that a combination lap belt and shoulder harness -- called the three-point belt -- is more effective than the simple lap belt. It prevents forward jack-knifing and provides lateral restraint against side impacts. Cornell analyzed data from California accident reports and found that simple lap seat belts were quite effective in controlling passenger ejection, reducing dangerous and fatal injury by thirty-five per cent or more. But later data on front seat-belted passengers, released in a 1963 Cornell report, found that in head-on collisions, when passengers stay in the car, there seems to be little difference in injury between those who wore seat belts and unbelted occupants. Cornell added that "the problem is not that the seat belt is a failure but that the front compartment -- the dash panel and steering assembly -- is not providing forward clearance for the head, knees, and torso, so that the body can jack-knife without interference."
The installation of the three-point belt is now being pressed by crash research specialists outside the industry as the second stage in passenger restraint development. This belt presents complications that the automobile makers would like to avoid. Cornell's Robert Wolf told the annual convention of the American Automobile Association in September 1964 what the difficulty is: "Installing a shoulder harness, however, in one's own car is an extremely discouraging project, much like that of trying to fit a homemade seat belt installation ten years ago. The problem is first to find a structurally sound anchor point for the shoulder strap and in a position where the strap doesn't slip off of the shoulder. To make a good anchor point usually requires a good mechanic with a good engineering sense. The chances of early large-scale adaptation to all types of American cars by the simple expedient of the Industry's providing standard shoulder strap anchor points, as was the case for seat belts, seems remote to me because of the difficulty of providing a structurally sound attachment point on hardtops and convertibles, which have no center post to the roof."
Hardtops and convertibles have been gaining rapidly in the percentage of total car sales, reaching almost fifty per cent in 1964. Even the recent sedan models with center posts present formidable difficulties In attaching the upper anchor of the harness and, when installed, give no assurance that they are strong enough to take the pull. Because the manufacturers are on the defensive they take the hard line. Once again, their rationale is based on unspecified tests of only one of the several kinds of possible shoulder harnesses. General Motors president James Roche delivered a statement to the Ribicoff Senate subcommittee in July 1965. He said, "At this time, our plans do not Include the Installation of anchorages for shoulder harnesses. We have con ducted extensive tests and studies of this device. Some of these tests have indicated that In a severe impact situation, shoulder harnesses can do more harm than good. While the harness does restrain the car occupant's forward motion, it also can deflect the impact force into a downward motion, forcing the occupant farther under the seat belt. This downward force can result in highly injurious pressures on the abdominal area. A shoulder harness also can exert dangerous pressure on the occupant's neck, particularly in the case of a relatively high-speed side impact."
It is obvious that poorly designed shoulder harnesses, inappropriately anchored, might result In some Injuries at the same time that others were prevented. But it is just as obvious that good design and Installation at proper anchorage points can avoid these small risks. Crash studies and accident analysis of the effect of these harnesses In England and Sweden, where they are in more widespread use, have shown results highly in their favor.
At the eighth Stapp Car Crash Conference, held In Detroit in October 1964, all the automobile companies had representatives present. None denied the superiority of shoulder harnesses over lap belts. Several, especially Chrysler's Roy Haeusler, actively advocated the use of harnesses. Dr. Paul Joliet, chief of the U.S. Public Health Service's division of accident prevention, has urged that shoulder harnesses be made standard equipment on new cars.
But the lack of an adequate center post, or any center post at all, on most models remains a problem. The search for making the seat belt more effective leads, as General Motors accurately foresaw years ago, to probes of other design inadequacies. In this case, the focus is on the seat Structure. Seats that tear away from their moorings and add unbearable "g" forces to a passenger already hurtling forward are one of the most common design failures recorded by crash investigators. The General Motors Engineering Journal May-June 1955 reported that for a GM seat to be considered satisfactory, it had to withstand a load of one thousand pounds. This means that two 150-pound persons sitting in the rear seats and striking the back of the front seat at only a 3-1/2 "g" force (or any combination thereof) would dislodge the seat from its moorings. In recent years, seats have been a little more firmly anchored, but the problem remains. Medical investigators reported a case in which a 195-pound football player, seated in the back seat of a car involved in an accident, was thrown against the back of the front seat, pushing it forward and crushing to death a front- seat passenger.
In March 1965, Product Engineering reported the development of an integrated seat by an automobile company supplier: "Current seat belt anchoring hooks the belt to bolts in the car floor; the new system anchors the seat, then attaches the belts to the seat. And that's the safety feature; positive seat anchoring should prevent the seat from being tom loose during a crash. The seat is designed to accommodate a retractable harness system and headrests (to prevent the head from snapping back on rear end collisions). It can be added to existing cars or incorporated as original equipment at little or no extra cost, according to the manufacturer, which presented prototypes to all the domestic car producers."
What is important in this example, as in other examples of automobile safety features, is not the particular design, but the performance function which is ignored by contemporary automobiles. UCLA's Derwyn Severy has pointedly criticized the industry at technical meetings for not designing a seat that will prevent the neck or spinal injury of the common rear- end collision. "It is the one most easily corrected by design and the one given least attention after perhaps the steering wheel and shaft," he said in 1964. Yet university crash injury researchers have not succeeded in getting industry specialists to discuss this problem in open forum on a high technical level. It is the most neglected aspect of passenger restraint.
Seat belts are now standard equipment and their installed cost to the car buyer is about one third of what they cost five years ago. Nearly thirty per cent of all automobiles on the road are equipped with seat belts, and the number of motorists using them is steadily increasing. The growth of habitual seat belt usage will accelerate now that the seat belt bas been removed from its place as the ugly duckling of the automobile world's vast array of optional equipment and gingerbread.
The passenger compartmentIn a collision, an automobile passenger can be adequately restrained and still be injured or killed if another vehicle, a tree, an abutment, or any other striking object invades the passenger compartment. Nearly a third of all injury-producing accidents involve either roof impact caused by a car rolling over or penetration of the side wall of the vehicle cabin.
The two elements of the car's structure most directly involved in sum accidents are the chassis frame and the body frame. The purpose of the chassis frame is to give proper support for the body and chassis components. The body frame, which has been welded or joined with bolts to the chassis frame, is the other load-bearing structure in the car.
It is also a function of the car's body structure and frame to absorb collision energy and maintain what collision specialists call the "structural integrity of the outer shell whim surrounds the restrained passenger." But when it comes to design and manufacture for such performance in collisions, the automobile industry has either ignored the statistical evidence of the problem or is deliberately withholding knowledge about it. Despite the reports of Cornell's Automotive Crash Injury Research project and other crash injury research groups on the significant role of car frames and bodies in side-impact crashes, there is not a single discussion of the subject to be found in the technical literature produced by the industry's engineers and stylists. There is neither published evidence nor claims by the companies to any proving-ground tests of direct side-impact crashes involving the passenger compartment. Nor is there in the technical literature any attempt to establish load criteria, to evaluate existing frame types, or to study the relative adequacy of proposed alternatives. In this critical area of automotive engineering there is instead almost total confusion -- leaving the consumer helpless to make any meaningful distinctions about the relative safety of the various types of body structures and frames employed.
A case in point is the "X" or "cruciform" type chassis frame. This frame was introduced in '957, primarily to reduce the problem of restricted headroom and difficult entry into the "low-profile" automobiles that were becoming popular after the mid-fifties. The X frame construction does not have side rails along the passenger compartment, as did most previous conventional frame designs. From the time the cruciform type frame was introduced, it was widely used by General Motors on Chevrolet, Buick, and Cadillac. The Ford Motor Company continued to use frames with side rails, and it was evident that the two companies held strongly different opinions about the two designs.
In the fall of 1959, a photograph of a Chevrolet Impala that was broken in half after striking a tree broadside was widely circulated in newspapers throughout the country. The frame had severed at the intersection of the X. The report of the General Motors investigators who rushed to the scene attributed the severance of the frame to the semi-airborne position of the car as it struck the tree. This had apparently allowed the engine mass to act as the head of a sledge hammer. At the General Motors engineering center in Michigan the conclusion was that "automobiles are not designed to withstand sum tremendous lateral forces -this would be extremely uneconomical."
General Motors spokesmen continued to defend the cruciform type frame as offering substantial resistance to side impacts because of the rocker panel and floor pan underbracing members -- even though by 1965 all General Motors models except the Buick Riviera had abandoned the design in favor of the perimeter type. In 1960 the General Motors technical center offered proof that a unitized structure with side rails can also split into two pieces. A picture of a Ford Thunderbird, torn in half after slamming against a telephone pole and tree, was offered as evidence to critics of the X type frame.
This comparison enraged Ford engineers. Fletcher N. Platt, a highly talented research engineer at Ford, retorted that the Thunderbird case involved a telephone guy-wire that had "acted as a knife on the entire body structure." In contrast, he said, "the Chevrolet that broke in half failed at the center of the X frame after hitting a tree." Platt said, "The X frame has no advantages from the standpoint of passenger protection. It requires less material to support the four comers of the car, but it is obviously less rigid and provides little lateral [side] protection to the passenger compartment." He suggests consulting any "'unbiased' structural engineer regarding these two designs." Mr. Platt might not consider Mr. Harry Barr, vice president for engineering of General Motors, qualified for the designation 'unbiased,' but Mr. Barr did admit grudgingly, under questioning, that the Oldsmobile perimeter type frame had some advantages over the Chevrolet X type frame in side-impact crashes at speeds of about fifteen miles per hour. Further proof that some General Motors engineers agreed with Ford's Platt came in the form of an internal memorandum prepared by the Oldsmobile division in 1963 in whim the Oldsmobile "guard-beam" frame was described as offering an "extra margin of protection" over the X type frames of Chevrolet, Buick, and Cadillac.
The manufacturers may disagree about the relative effectiveness of different kinds of body frames, but they say little or nothing about the comparative safety of the conventional sedan and the so-called hard-top models.
In the hard-top models there is no center door pillar from the window sill upward and no upper half of the door frame. The same is true of convertible models, but at least the customer is on obvious notice when he buys a convertible, while the hard-top resembles the conventional sedan in the apparent security of the enclosure.
One danger in the hard-top model was cited by Robert Wolf of ACIR, who said, "It is quite common in a side impact of a four-door hardtop car for the center post to tear out at the floor attachment joint, where the post is loaded severely in bending. These posts are probably not designed to withstand a severe crash load -- they are there for other purposes."
Still another hazard is the dangerous consequence of a "rollover" in a hard-top model. Without the upper center post to support the roof structure, the hard-top offers less protection to its occupants than does the conventional sedan. In many accidents involving roll-overs, the hard-top has been described as having "crumpled like a Japanese lantern." One official of the Fisher Body Corporation said of General Motors bard-tops that they were "on the borderline." But who knows what the borderline is?
If the companies insist on pillarless construction, there are various engineering approaches that can strengthen the crash resistance of their cars. The manufacturers themselves have patented practical kinds of latch and reinforcing member arrangements which lock the side of the vehicle into an integral unity by the use of multiple latch locations. Nothing has been done to apply these patents to current automobile production.
Likewise, the provision of roll bars to protect against the impact of a roll-over is another possible improvement that is viewed negatively by the industry. Many physicians with an interest in automobile races have been impressed with the protection given drivers by roll bars or equivalent reinforcement when their vehicles go through spectacular accidents, sometimes flipping over and over for hundreds of feet. Dr. John States, president of the American Association for Automotive Medicine, has urged the automobile makers to incorporate roll bars in their designs. But such a safety feature would apparently inconvenience the designers of hard-tops and, even in sedans with upper center posts, would involve changes which are abhorrent to the cost analysts and stylists.
In the whole area of reinforced and strengthened body and chassis structures, the industry has steadfastly avoided testing, research, and change for safety. While gearing its public relations to stories of vehicles crashing at proving grounds, it continues to ignore the work of the men who have done the necessary studies. One such man, James J. Ryan, a recently retired professor of engineering at the University of Minnesota, has done extensive car collision experiments. Just one of his findings suggests the direction the car makers could follow. Mr. Ryan said recently, "From our tests we have determined means of strengthening the structure of the vehicle to prevent displacement of the walls, the door, and the posts and the penetration of the driver's compartment. The forces of impact could be reduced four times by the proper construction of any vehicle without increasing its cost or weight."
***
It has become evident that the Cornell data playa central role in any discussion of the second collision. After half a century of automobile usage, a staff of only nine people began, with federal support, the first statistical reporting system on how interior car designs injure and kill motorists. The time for analyzing the design of automobiles had come, and the crucial distinction between the causes of accidents and the causes of injury was shown with unmistakable clarity. The driver could no longer be the scapegoat for industry negligence in the design of their vehicles. From the day De Havens group began work in 1952, segments of the automobile industry suspected that things might never be the same again if they remained aloof from Cornell's probings.
Two events in 1955 moved the industry to act. The U. S. Public Health Service joined the Department of the Army in support of Cornell's Automotive Crash Injury Research (ACIR), thus assuring continuity and growth to the project. Early in the year ACIR released a comparative study of automobiles manufactured from 1940 to 1949 and those manufactured from 1950 to 1954 on the question of whether the newer group produced more or less injury than the older group in similar accidents. The study concluded that "on the most conservative basis, 'new' (1950-54) car designs have not demonstrated any improvements in the injury effects produced by accidents. When injury-producing accidents occur, occupants of 1950-54 ears are injured more often than occupants of 1940-49 cars. Further, there is a statistically significant increase in the frequency of fatality among the occupants of 'newer' cars. The contention that present day automobiles are 'safer' in injury-producing accidents is not borne out by the facts."
For its part, General Motors shrugged off the findings. Some Ford and Chrysler officials, however, were more sensitive to the possible consequences of this kind of information. An independent project, solidly financed, was acquiring the statistical capability to evaluate on a comparative basis the safety of automobiles based on their actual accident injury experience. The officials realized that it would be to the industry's advantage to establish their presence in ACIR's work. Before the end of 1955 Ford and Chrysler each announced a two-year grant to ACIR of $100,000 per year. In 1957 General Motors finally joined them in providing financial support through the Automobile Manufacturers Association. During the past several years, ACIR bas relied on annual grants of $175,000 from the Automobile Manufacturers Association and $300,000 from the U. S. Public Health Service.
From the standpoint of protecting its interests, the industry has never received so much for so little. The result has been an impressive perpetuation of the status quo in vehicle safety design, in spite of the potentially devastating impact of the collected data. Right from the beginning a close liaison was established between ACIR and the automobile industry. ACIR's director, Robert Wolf, said recently that interim studies and preliminary findings are often reviewed with the Automobile Manufacturers Association. The AMA is consistently asked for guidance and usually reviews drafts of reports before they are released to the public. Prior to a major announcement, such as the one made in November 1964, called "Automobile Crash Injury in Relation to Car Size," it has been common practice for ACIR to meet with industry representatives and go over the wording in the release.
Why Cornell finds it necessary to seek the advice and approval of the AMA concerning statistical analysis and reporting of data dealing with past accidents is not explained. Certainly ACIR has an adequate statistical staff and all the necessary data-processing equipment. The answer, in large part, lies in the AMA's desire to exercise a reviewing function which assures that ACIR does not name makes and models. To say, for instance, that the steering assembly is a major instrument of injury is a finding that can be tolerated by the automobile companies, but to have ACIR reports say that Make A's steering column is twice as likely to injure the driver as those in Make B, C, and D, would be damaging; it would tell consumers, insurance companies, and interested public agencies that some cars are not as safe as other cars.
The manufacturers have been almost entirely successful in making ACIR see matters their way. On only two occasions has Cornell named the brands of cars involved in ACIR reports. In 1964 ACIR's B. J. Campbell reported that an analysis of door latch effectiveness on very late model cars showed little difference between General Motors, Ford and Chrysler. Three years earlier, when a Cornell report found significant differences In door latch failure among the "Big Three," it deleted the car names and replaced them with Brand X designations. Another instance came in November of 1964. The Cornell report, called "The Safety Performance of 1962-1963 Automobile Door Latches and Comparison with Earlier Latch Designs," was based on data from 24,342 cars in which at least one occupant was injured during an accident. Among its more interesting conclusions was: "The doors of General Motors cars were tom off more frequently than those of Ford or Chrysler and the type of hinge damage appeared to be different, too: the General Motors hinge appeared to snap off cleanly with little or none of the deformation or twisting observed for other cars." ACIR was specific with its figures:
PERCENTAGE OF CARS WITH DOORS TORN OFF In the past two years there have been indications that ACIR is not entirely satisfied with the constraints placed upon it as a result of its "understanding" with the Automobile Manufacturers Association, but the chafing has not yet resulted in any blossoming of scientific independence.
However cautious ACIR has been in seeing that its internal workings and projected studies be kept from the public view, it made a mistake with the formally announced and suddenly suppressed Shoemaker and Narragon report. This was an analysis of steering column penetration scheduled for release in November 1963. ACIR director Robert Wolf gave a preview of the findings in an address he delivered that month at a Liberty Mutual Life Insurance Company conference on the automobile and public health. Wolf said, "This study, which examines accidents involving standard American cars, compacts and European cars, shows clearly that injury to drivers is strongly increased when column penetration occurs." He noted that in accidents of similar severity, the column on some makes of cars held up much less effectively than on others. Wolf then cited the report as "Narragon, Eugene A., and Shoemaker, Norris E., Steering Column Penetration in Automobile Accidents. Automotive Crash Injury Research, Cornell Aeronautical Laboratory, Inc., Report No. VJ-1823- 4, November 1963." Several months earlier in a Laboratory pamphlet entitled "Transportation Research," the same reference appeared. The month of November ended, and there was no report. There has still been no such report released. In ACIR's annual report for 1964, it was disclosed that the Shoemaker study was released to the AMA for the "purpose of securing technical guidance for use in a final report." The report also said, "The ACIR staff is still not satisfied that the best approach to the study has been formulated." Since the study pertained to what Robert Wolf called "important comparisons between car makes," caution indeed had been the order of the day.
The general explanation about statistical difficulties given by ACIR is not persuasive for two reasons. First, the ACIR seven-man statistical staff, headed by Dr. Jaakko Kihlberg, is acknowledged to have a high order of technical skill. Second, statistical difficulties of such seriousness would seem to have been discoverable well before the announcement that the report would be issued on a specific date.
As the Cornell data has accumulated to levels permitting more relined analysis of makes and models, private criticism by certain crash injury research specialists and observers of ACIR's tabu against naming manufacturers and models has mounted as well. Yet future plans for topics and studies to be undertaken by ACIR give no indication that analyses by manufacturer or make will be published.
Aversion to naming the manufacturer or make of car is not the only way that ACIR pays interest on the funds supplied by the Automobile Manufacturers Association. For almost a decade, ACIR has been providing each sponsoring automobile company with microfilm copies of accident photographs and police and medical reports of cases involving that company's products. For example, General Motors receives case reports relating to GM automobiles. These cases are provided only to the manufacturers. Furthermore, when an unusual occurrence of structural collapse or an injury relating to a particular make is observed, even if it is just one clinical investigation, notification is given to the producer of that automobile.
The exclusive funneling of specific case materials to the automobile makers by ACIR raises serious questions of public policy. ACIR's work is largely financed and supported by public agencies and funds. Over sixty per cent of its annual funds come from the U. S. Public Health Service, but the public contribution is much greater than indicated by that percentage. ACIR receives data for only a small fraction of its true cost since police and public health personnel freely contribute their time in preparing the specially designed report form that ACIR supplies them. This information should be considered a national data bank to be used for the benefit of the public generally.
In the present situation an injured person cannot obtain even the reports pertaining to the accident in which he was involved. Yet victims of marine or air disasters or their legal representatives have the explicit legal right to the detailed accident investigation data gathered by the Coast Guard or Civil Aeronautics Board. The Cornell data should be freely available to the public. In his final report on four years of investigation of fatal automobile accidents in the Boston area under a U. S. Public Health Service grant similar to that given Cornell, Alfred L. Moseley urged, "The findings should be public records so that justice and fair play in criminal and liability proceedings would be assured."
ACIR has rebuffed requests from public agencies to release to them even a small portion of the data which ACIR has given to the manufacturers. The New York State Joint Legislative Committee on Motor Vehicles and Traffic Safety (the Speno Committee) got in touch with Mr. Wolf in May 1963, taking note of an ACIR study released in 1961 that showed significant variations in door opening frequency among cars made by the "Big Three" manufacturers. The committee requested identification of the manufacturers and photographs of door latch and hinge failures in order to give it a basis on which to determine what design differences in the various door latches and hinges were associated with a higher frequency of door openings. The committee was in the middle of its pioneering investigation into vehicle safety and the need for safety design standards. ACIR turned down the committee's request, but a year and a half later decided it was wiser to publish the fact, with accompanying photographs, that General Motors had the worst door-opening record, followed in order by Ford and Chrysler.
On March 27, 1965, the Speno Committee wrote to Dr. Paul Joliet, chief of the Division of Accident Prevention in the U.S. Public Health Service, the organization that administers the federal grant to ACIR. The Speno Committee said that since the basic case data that ACIR supplies to the manufacturers is available to the Public Health Service, the committee would like to review these records in order to make its own analysis. The committee further pointed out that it considered the Cornell data to be publicly owned and therefore accessible to public agencies -- local, state, or federal.
Dr. Joliet called in the principals of the ACIR project to review the Public Health Service's policy concerning the issues raised by the Speno Committee. The result was a blanket endorsement of the status quo. Dr. Joliet stated flatly that ACIR was "free to determine with whom they wish to discuss the nature of any preliminary analyses they have performed, to whom they wish to make available any of their raw data material," and also to determine whether they wish to consult with their sponsors regarding publication of particular preliminary or final analyses. Further, the release of case data material to other investigators or other parties at interest is at the discretion of the principal investigator and the institution."
Not only has Dr. Joliet's division endorsed ACIR's policy of sharing its data only with the car manufacturers, but also it has denied itself the use of the data. Though the Division of Accident Prevention has the right to receive the same case material given to the manufacturers, it has deliberately chosen not to do so. When asked the reason for this policy, one employee of the division answered, "Who wants hot potatoes?"
Dr. Joliet and his associates seem to believe that their responsibility ends after they determine the value of the research proposal they are financing. In view of the Public Health Service's legal mandate, this is a remarkably limited role. The Division of Accident Prevention's key purpose is to plan and conduct "a nationwide accident prevention program aimed at encouraging and assisting state and local health and other agencies in the development, operation and improvement of local accident prevention programs." Dr. Joliet has told many Congressional committees that his division's interest is in preventing deaths and injuries. Presumably empirical data would help in this work.
To permit public funds to he mixed with industry money in such a project as ACIR and to give researchers full discretion to give data to manufacturers while denying it to all others is nothing short of an abdication of the public trust. By this action, the Division of Accident Prevention of the U. S. Public Health Service is sanctioning what amounts to a subsidy of the automobile industry, since the industry is the exclusive recipient of data that is paid for mainly through taxpayer contributions. This is a real bargain for the automobile manufacturers, whose contribution to ACIR amounts to the equivalent of only 2¢ for every car they sell.
There is no evidence that the industry has improved the safety of its vehicles as a result of the case reports it obtains from ACIR. In an article generally sympathetic to the automobile makers, Automotive News in May, 1965 commented, "Regrettably, the companies are making little use of these reports."
In addition to disseminating its case data exclusively to the automobile industry, ACIR appears to have an unreal vision of how its studies would find ultimate application to the design of safer vehicles. According to the Cornell scientists, their dreamed-of progress would proceed this way: 1) "Statistical studies discover a problem, define the problem area, and point toward a solution; 2) laboratory and engineering work result in a solution which is then incorporated into the vehicle; and 3) statistical studies evaluate effectiveness of the solution and indicate the need for further refinement."
The fallacy of this reasoning is illustrated by the history of just one item. So far, the door latch is the only vehicle feature that has gone through this sequence. First, Cornell found that the risk of serious injury or death was markedly greater when occupants were ejected than when they remained in the car. In the pre-1956 cars, ACIR data revealed that at least one door opened in nearly half of the injury-producing accidents. Then, in its 1956 models, the industry introduced so- called safety door latches, which involved a simple design change that was at least thirty years overdue. Finally, in 1961, Cornell released a study showing that door opening frequency in the 1956-1959 models, compared with the pre-1956 models, was reduced by about thirty per cent The next door latch improvements came in 1962 from Ford, in 1963 from General Motors, and in 1964 from Chrysler. In other words, for ten years motorists were used as guinea pigs while the car makers were awaiting statistics on how many of them were being thrown from cars during collisions before deciding to inch forward with the next improvement.
Statistical evidence is, after all, only one basis on which to decide the need for safer design. Clinical studies of a single, or a small number of cases can define a safety problem that demands a design change. Even before waiting for blood to be shed or a mangled vehicle to be investigated, as in the General Motors door hinge failure, advance design analysis and testing under collision conditions could detect a large majority of hazards before the final mass-production specifications are completed.
To move a manufacturer to action should not -- as it did -- require statistical confirmation by Cornell that the rear-view mirror is one of the ten top instruments of injury in automobile collisions. It is enough to know, as the Ford Motor Company knew in 1964, according to their consultant Dr. Donald Huelke, of twenty fatal cases which occurred when the victims struck the rear-view mirror in Ford Falcons. If the possibility of this particular hazard did not occur to the automobile designers before the vehicle was built, these fatal cases are proof of the need for redesigning the rear-view mirror.
***
ACIR has been subjected to some unfounded criticism. Certain foreign-car manufacturers, for example, intimated that the Cornell group made its "Big Car-Little Car" study -- which found, under similar accident conditions, a considerably higher incidence of serious injuries and deaths occurring in small-car accidents -- under pressure from American car manufacturers. In fact, the study was done on Cornell's initiative. But some fundamental criticisms of ACIR are justified. ACIR scientists have not displayed much commitment to giving a broader significance to their work. Like their colleagues at the Harvard School of Public Health, UCLA, and Wayne State University (all working with federal funds and industry assistance), they have been in possession of information that is relevant to the elimination of millions of casualties, and the expertise to utilize that information. Like their colleagues, they have shown only a slight appreciation that their special roles should require them to state forcefully in public forums the issues for discussion and resolution. As nuclear physicists and medical scientists learned years ago, public discussion is of great importance to their research undertakings. Ultimately, the successful implementation of research findings provides the public support for additional research. The absence of scientific statesmanship among these independent accident-injury researchers working under federal grants explains to a great degree why their funds have not increased noticeably for a decade. These scientists who do not make known to the public the importance of their work and the practical possibility of a vastly safer vehicle cannot, of course, enjoy public support.
The ACIR staff might well refer back to the testimony before the Roberts subcommittee in 1959 of Dr. T. P. Wright, vice president for research at Cornell University. As an engineer with wide experience in problems of transportation safety, Dr. Wright addressed himself to the question of whether engineered safety design of vehicles can result in dramatic reductions in the annual highway injury and fatality toll. His answer was, "Most decidedly, yes; with these provisos: "If a concerted effort is made toward fuller utilization of the information which scientific research has already provided; if appropriate support for present and future investigation and research can be assured; if present and future findings can be channelized to individuals and organizations willing and able to act on their implications by applying them at the practical level; if appropriate public educational measures are assured and maintained." Then, in words which should have weighed heavily on the minds of the university accident-injury researchers, Dr. Wright added, "Furthermore, as a matter of personal ethics, I should consider myself guilty of a crime against humanity if, for whatever reason, I were responsible for prolonging the ravages of a disease which is the unnecessary and shameful byproduct of the greatest transportation system the world has ever known.... Delay will be measured in inexorable terms of human life, suffering, and permanent disability."
John Moore, the director of ACIR between 1955 and 1960, did lend his assistance to the Roberts Committee in its attempt to establish a public record on vehicle safety problems. The present director, Robert Wolf, delivered two addresses in 1963 and 1964 when he recommended corrective measures that were available and effective for improving automobile crashworthiness. Small as these efforts have been, they are improvements on the timidity of other university scientists and engineers working in the area of vehicle safety. Perhaps with the accumulated record of industry intransigence and the opening up of diversified sources of financial support from public agencies (recent contracts from General Services Administration and the Department of Commerce fund for special projects are the first indications), ACIR will rise to its public responsibility.
ACIR should make public its general and specific findings on design hazards. It should explain the deeper issues of why such hazards persist year after year, and the engineering feasibility of producing much safer automobiles. The sooner ACIR performs these missions of scientific statesmanship, the sooner the scientific-engineering community and the major institutions which form public policy will be awakened to their long-neglected responsibilities to save lives.
_______________
NOTES:1. There are many ways to design steering columns to prevent what engineers call their "rearward displacement relative to the firewall and instrument panel." The design alternatives are cheap, practical and long known to the manufacturers. (See Fig. 4)
2. That was not the only year that Ford failed to exceed Chevrolet in sales. Moreover, the 1956 Ford, in contrast to the Chevrolet and the Plymouth, was barely changed from the previous year. Ford's Robert McNamara released publicly in early 1957 detailed figures on safety option sales and market surveys showing the marked success of the safety features in attracting purchasers. But to the delight of the industry the saying that in 1956 "Ford sold safety and Chevy sold cars" caught hold and became a standard response to critics of the automobile companies. It is interesting to note that Ford officials never went out of their way to deny this erroneous impression unless they were specifically requested to do so.