When did GM first offer airbags on its cars? 1988? 1989? Nope — Oldsmobile, Buick, and Cadillac all offered dual front airbags way back in 1974. Known as the GM Air Cushion Restraint System (ACRS), this rare option included the world’s first production driver- and passenger-side airbags, but it was seldom seen and is often misunderstood today. Why was it offered? Did it work? Here’s everything you need to know about these early airbags, in a convenient Q&A format.
Did GM invent the airbag?
No. Concepts for air cushion restraints for automobiles date back at least to the early 1950s, although none were actually offered in production cars until the ’70s.
One of the earliest concepts for an automotive airbag was patented in 1953 by Pennsylvanian inventor John W. Hetrick (as US2649311). Two other early patents were filed in the same period by Robert H. Hodges (US2755125) and Paul M. Maxwell (US2806737), who developed inflatable shields intended to keep occupants from striking the windshield in a crash. Those early designs used compressed air to inflate the cushion, which posed many problems.
The earliest automaker work on airbags may actually have been at Ford Motor Company, which launched its first experimental air cushion restraint study in January 1957. That project was unsuccessful because Ford was unable to get the airbags to inflate quickly enough to be of any value in a collision.
The concept of using a pyrotechnic gas generator rather than relying on stored compressed gas to fill the airbag emerged in the 1960s. Perhaps the earliest iteration of this idea was developed by Yasusaburo Kobori of Japan’s Good Idea Centre Co., who first proposed such a system in March 1963. Kobori patented this system in Japan and the U.S. (as US336045 and US3450414) in the latter 1960s.
Were the GM airbags developed by Oldsmobile?
No. However, beginning in the mid-1960s, GM assigned each of its automotive divisions to coordinate some multi-division engineering programs, and Oldsmobile Division was assigned to coordinate certain safety-related projects, including energy-absorbing steering columns and driver’s-side airbags. The development of the complete Air Cushion Restraint System actually involved several GM divisions, including Fisher Body, Delco Electronics, Electronic Control Systems, Saginaw Steering Gear, and Inland Manufacturing. A lot of the concept work was done by the GM Research Laboratories and corporate Engineering Staff. So, while Oldsmobile did coordinate some aspects of the airbag program, it wasn’t an Oldsmobile project per se.
To the extent that the GM Air Cushion Restraint System could be attributed to any single GM division, it would be Fisher Body.
Why did GM develop airbags back then if they weren’t legally required until the ’90s?
It’s crucially important to understand that the U.S. automatic protection (passive restraint) rule, which became part of federal Motor Vehicle Safety Standard (MVSS) 208, was originally intended to take effect more than 15 years earlier than it actually did. From 1968–1969 into the early ’70s, many automakers — not just GM — did extensive work on airbags, assuming that they might soon be required by U.S. law.
Federal regulators first proposed the passive restraint requirement in July 1969, and the initial “final” rule issued in November 1970 was scheduled to become effective on July 1, 1973. However, most automakers — including GM — strenuously opposed this rule and did everything in their power to obstruct and delay it.
While GM was as opposed as the rest of Detroit to the passive restraint mandate, the GM airbag program had a strong patron in Ed Cole, the former Cadillac and Chevrolet engineer who had become president of General Motors in late 1967. Like many GM engineers of his generation, Cole hated seatbelts. He eventually became convinced that airbags could be a better alternative — which was one of the only things Cole had in common with consumer advocate Ralph Nader. In his 1965 book Unsafe at Any Speed, Nader had argued that seatbelts should have become obsolete in the ’50s in favor of “superior passive safety design which would come into use only when needed, and without active participation of the occupant.”
During Cole’s presidency, the GM air cushion restraint program became one of his pet projects. However, Cole continued to insist that making airbags or any other passive restraint system mandatory would be undesirable and probably infeasible. Throughout the 1970s, GM tried repeatedly to persuade the National Highway Traffic Safety Administration (NHTSA) to defer the passive restraint mandate in favor of smaller-scale, voluntary programs, ostensibly to allow the industry more time to work out the many technical and logistical challenges involved, which were considerable. These proposals were usually accompanied by a promise that GM would eventually offer airbags on a larger scale (and perhaps as standard equipment) at some later date. None of those promises was kept, and in retrospect, it appears the real object was to delay any actual federal passive restraint requirement for as long as possible.
Was GM the first automaker to offer airbags?
Not exactly. Discounting experimental projects, the first automaker to install airbags in production cars was actually Ford Motor Company, which fitted them to a batch of 92 1971 Mercurys, and then to 831 1972 Mercury Monterey sedans.
However, you couldn’t go to a Lincoln-Mercury dealer and buy an airbag-equipped Mercury Monterey. The 1971 batch was retained by Ford Motor Company for use by Ford employees. Some of the 1972 cars were also added to the company fleet, and the rest were supplied to select fleet users, including certain Ford suppliers, several federal agencies, and the Allstate Insurance Company. After 18 months or 40,000 miles, all the airbags were removed from those cars and sent back to Ford for testing and analysis. Ford also brought back 33 of the 1972 cars and fitted them with new airbags for barrier crash testing. Some of the other 1972 cars might have eventually been resold to the public, but only after their airbags had been removed.
At several points in the ’70s, Ford Motor Company considered offering airbags as a production option — initially for the 1974 full-size Ford and Mercury, later for the 1975 Lincoln Continental Mark IV and Ford Thunderbird — but always found reasons not to. Ford eventually concluded that they had “not succeeded in developing a system that, from a technical or economic point of view, we could justify offering to our customers.”
So, GM was the first automaker to actually make airbags available to the general public.
When did GM first start offering airbags?
During the 1973 model year, GM installed its Air Cushion Restraint System in 1,000 specially modified green 1973 Impala four-door sedans as a field trial program.
These 1,000 field trial Impalas had Oldsmobile dashboards with dual airbags, but no seatbelts.
As with the earlier Mercury cars, the field trial Impalas were NOT sold to the general public. About half were retained by GM as part of the company’s own fleet, while the rest were leased to other select fleet operators. In this way, GM was able to keep the performance of the field trial cars “under rather close observation,” which wouldn’t be possible once the ACRS was offered to the public.
The production Air Cushion Restraint System went into production in late November 1973 and became available on some 1974 Oldsmobile, Buick, and Cadillac models beginning in January 1974. The first production car equipped with the system was the 1974 Oldsmobile Toronado pictured above; I unfortunately couldn’t find any color photos of it, but some sources describe it as “copper” (possibly Cinnamon Metallic).
Which GM cars were offered with airbags?
The GM Air Cushion Restraint System option, RPO AR3, was available on most 1974 to 1976 full-size (B-body, C-body, and E-body) Oldsmobile, Buick, and Cadillac models:
- Oldsmobile: Delta 88 (except convertible), Ninety-Eight, and Toronado; not available on Custom Cruiser wagon
- Buick: LeSabre (except convertible), Electra 225, and Riviera; not available on Estate Wagon
- Cadillac: Calais, De Ville, Fleetwood Brougham, and Eldorado coupe; not available on Eldorado convertible or limousines
The ACRS was not available on any Pontiac models, or on any Chevrolets other than the 1,000 1973 Impala field trial program cars. The airbags were never offered on any intermediate or smaller cars. The ACRS option was discontinued after the 1976 model year and never became standard equipment on any GM models in the ’70s.
How many GM cars were built with airbags in the 1970s?
A 1983 study conducted for the NHTSA by the National Analysts division of Booz-Allen Hamilton presented the following production figures:
| Make | 1973 | 1974 | 1975–1976 | Total |
|---|---|---|---|---|
| Chevrolet | 1,000 | 0 | 0 | 1,000 |
| Oldsmobile | 0 | 2,433 | 3,050 | 5,483 |
| Buick | 0 | 2,177 | 276 | 2,453 |
| Cadillac | 0 | 908 | 1,399 | 2,307 |
| Total | 1,000 | 5,518 | 4,725 | 11,234 |
The totals for the regular production option apparently came from contemporary reports in Ward’s Auto World and from GM. A November 1976 item in the Wall Street Journal also reported that sales of the ACRS option had totaled 10,243, not including the 1,000 Impala field trial program cars. However, a 1988 SAE paper about the Air Cushion Restraint System by GM Safety & Crashworthiness Systems engineer Harold J. Mertz reported total production (including the 1973 Chevrolet field trial cars) as 11,321 — 78 more than previously reported. That figure might include cars retained by GM rather than being sold.
Unfortunately, I have no breakdowns by model.
ACRS cars were rare, and not many survive today. The easiest way to spot one is to look for the distinctive four-spoke steering wheel, which had a horn button on each spoke. All Air Cushion Restraint System cars used this same wheel, with the name of the division in the small label in the lower right-hand corner of the airbag cover. Airbag-equipped cars also had a unique “AIR CUSHION” warning lamp on the instrument panel. ACRS-equipped cars were not available with tilt or tilt/telescope steering wheels because of their unique steering column. Saginaw did some work on adapting the ACRS column for the tilt/telescope option, but that combination was never offered in the ’70s.
How much did the GM airbag system cost?
The suggested retail price of the ACRS option was $225 in 1974, $300 in 1975, and $315 in 1976.
Many customers and some dealers felt these prices were too high. Two contemporary GM marketing studies showed that buyers were potentially interested in airbags if the price was between $100 and $200, but interest dropped off rapidly if suggested prices rose above $200 (which was the equivalent of almost $1,500 in 2026 dollars). However, GM claimed in March 1975 that they were actually losing about $250 on each airbag-equipped car even at the suggested retail price.
Did the GM airbag cars still have seatbelts?
The 1973 Chevrolet Impala field trial program cars had no seatbelts. All regular production 1974–1976 Air Cushion Restraint System cars came with front lap belts, but not shoulder belts.
GM did recommend using the lap belts, but the ACRS was intended to comply with the original MVSS 208 standard, which required frontal impact protection “by means that require no action by any vehicle occupant.” In addition to the airbags, the driver’s side of the ACRS had a special type of impact-absorbing steering column and a padded metal knee brace to help keep an unbelted driver upright in a crash.
Cars with the airbags did not have the seatbelt interlock system that was fitted to the ignition switches of most 1974 and some 1975 U.S. cars.
Aren’t airbags dangerous if you’re not wearing a seatbelt?
Yes, and that danger is greater with older “first-generation” airbags like the GM Air Cushion Restraint system, which were intended as passive restraints (rather than supplemental restraints for belted occupants).
The original federal MVSS 208 passive restraint standard required automatic protection systems that could restrain an unbelted crash test dummy in a 30 mph barrier impact while keeping the forces on the dummy’s head, chest, and legs below specified limits. To meet those requirements, airbags had to deploy very rapidly — within about 40 ms — and with considerable force. GM estimated that in a 30 mph impact, an unbelted driver would strike the inflated airbag with a force of about 2,300 lb (roughly 10,200 N), which would be absorbed by the controlled deflation of the airbag and by the energy-absorbing steering column.
In human volunteer tests with average-size adult male drivers, airbag deployment didn’t usually cause injuries more severe than minor bruises and abrasions. However, real-world experience later demonstrated that smaller, shorter occupants could suffer serious injury from the airbag deployment (sometimes even if they were belted). Unbelted occupants, especially front passengers, were also more likely to be “out of position” (e.g., too close to the airbag, or at an unusual angle), which increased the risk of injury. A particular problem was young children standing or lying unrestrained on the front seat.
Then as now, wearing a seatbelt reduced the risk of being injured by the airbag deployment — an occupant properly restrained by a seatbelt would hit the airbag with less force, and was less likely to be out of position relative to the airbags when they deployed.
However, until the late 1980s, very few Americans — on average, no more than 20 to 25 percent — wore seatbelts. No U.S. state had mandatory seatbelt use laws until 1984. Before that, many legislators and most state governors believed supporting such laws would be political suicide. Seatbelt education campaigns seemed to have little impact on Americans’ seatbelt use rates. Even ardent safety advocates were convinced that there was nothing anyone could do to make Americans buckle up voluntarily. This left federal regulators looking for ways to either mechanically force people to wear seatbelts (with automatic belts or ignition interlocks) or else provide equivalent crash protection without seatbelts.
The possibility of unbelted occupants being injured or killed by airbags was not unknown in the ’70s, and in fact GM made some design changes to the ACRS prior to production to try to reduce that risk. However, airbag advocates — including Ralph Nader and the Center for Auto Safety — generally dismissed the danger, or insisted that it could be alleviated with incremental design changes.
Did the early GM airbags work the same way as on modern cars?
Broadly speaking, yes, but there were many minor differences.
Like modern airbags, the ACRS airbags were inflated using pyrotechnic gas-producing generators. The passenger-side inflator assembly also used some stored gas. Unlike later airbag systems, the ACRS used argon rather than nitrogen to fill the airbags.
The passenger side of the ACRS actually had TWO air cushions, one inside the other. The smaller inner cushion acted like a knee brace, inflating to fill the space between occupants’ legs and the dashboard. The larger outer cushion inflated around the knee bag to restrain occupants’ torsos and heads. These passenger-side airbags were large enough to protect both right front and front center passengers.
Airbag deployment was triggered electrically by three inertia switches activated by two crash sensors: a bumper impact sensor, located in the nose, and a passenger compartment G-sensor, located in the plenum behind the firewall.
Compared to modern systems, the sensors were enormous, and rather primitive. GM set the activation threshold for the bumper sensor to the equivalent of an 11 mph barrier impact (representing a deceleration of about 18g), while the passenger compartment sensor threshold was the equivalent of an 18 mph impact (a deceleration of about 30g). GM set these thresholds fairly high to keep the airbags from deploying inadvertently in fender-benders or on bumpy roads. Nonetheless, there were at least two incidents of the airbags deploying when the car’s frame bottomed out on a railroad crossing.
GM was concerned about the danger of front passengers being injured by airbag deployment in low-speed impacts, so on production ACRS cars, the passenger-side airbags were installed under the passenger side of the dash (rather than above the glovebox) and had a novel two-stage inflation system. The airbags were activated like this:
- “Low-level” deployment (>11 mph/18g impact): The bumper sensor would trip two of the three inertia switches, fully inflating the driver’s side airbag, triggering the first passenger-side generator, and releasing the passenger-side stored gas to partially inflate the passenger-side airbags.
- “High-level” deployment (>18 mph/30g impact): Like a low-level deployment, but the passenger compartment G-sensor would also trip the third inertia switch, triggering the second passenger-side gas generator to fully inflate both passenger-side airbags.
This two-stage system has been compared to modern “smart airbags,” which can change their deployment force based on factors like collision speed and occupant weight. The GM ACRS approach had a similar goal, but unlike modern airbags, the system had no means of detecting occupants’ weight, position, or seatbelt use — it could only respond to different levels of impact force.
Even back then, federal regulations required the Air Cushion Restraint System to have a self-diagnostic system with a dashboard warning lamp that would light up and stay lit if there was a system fault. The passenger compartment sensor also incorporated a crude type of crash recording system. It used a series of fusible links that would blow or not blow depending on whether the airbags received an activation signal and whether they actually deployed. This was to help crash investigators assess how the system performed in a collision, including determining whether there had been a high-level or low-level deployment of the passenger-side airbags.
Did the airbags actually work in crashes?
Yes, and GM studied every accident they knew about that involved an ACRS-equipped car. The NHTSA also closely examined accident data from cars with airbags.
There was a sticker in the glove compartment of cars with the ACRS option, urging owners to call a toll-free number to report airbag deployments and “tow-away” or occupant injury accidents. GM also encouraged dealers and insurance agents to report accidents involving airbag-equipped cars. In all, GM recorded and studied a total of 216 accidents involving ACRS cars through December 1980, although most were fairly minor.
The ACRS had some notable success stories. For example, in a November 1974 accident near the St. Lawrence River in Quebec, Canada, an airbag-equipped 1975 Oldsmobile Ninety-Eight Regency collided head-on with a 43-ton tractor-trailer rig at a closing speed of about 44 mph. Although the Oldsmobile was totaled (as you can see from the above photo), the driver and four passengers — none of whom were wearing seatbelts — all survived with mostly minor injuries. (One of the unbelted kids in the back seat was knocked unconscious upon hitting the back of the front seat.)
There were also five recorded fatalities in airbag-equipped cars. In three of those cases, GM concluded that the “violence of impact” was so severe that the occupants would probably not have survived in any case. For example, in March 1976, an airbag-equipped 1974 Oldsmobile Delta 88 (above) collided head-on with a 1972 Chrysler New Yorker at a closing speed of about 100 mph, killing all occupants of both cars. The Oldsmobile’s airbags inflated, but the passenger compartment was so badly crushed that it didn’t make any difference.
GM used a “matching case” approach to compare accidents involving airbag-equipped cars to past accidents of similar “injury potential” involving cars without airbags.
They eventually concluded that the ACRS was 21 percent effective in mitigating serious injuries to the driver and 16 percent effective in mitigating serious injuries to front seat passengers.
However, GM estimated that front passengers were actually 34 percent MORE likely to suffer mild to moderate injuries (particularly leg injuries) with the ACRS than without it. This was especially true if the passengers were small children — the accident case histories include a couple of harrowing examples of why child seats really should have become mandatory long before they actually did.
The fifth of the five recorded ACRS fatalities was a one-month-old infant who was lying unrestrained on the passenger seat at the time of the collision and was apparently killed by the passenger-side airbag deployment.
Did the airbags ever go off accidentally?
In a number of cases, yes. There were at least six inadvertent deployments, resulting in at least five lawsuits against GM. Three of those incidents occurred during servicing, one due to a vehicle fire, and two due to sensor or wiring problems. (In May 1975, Cadillac recalled 2,000 ACRS-equipped cars to correct wiring defects.) I’m not sure if the railroad track incidents were deemed inadvertent deployments.
Why did GM drop the airbags after 1976?
GM maintained that the ACRS option was discontinued after the 1976 model year due to slow sales and a lack of consumer interest. However, a likely contributing factor was the retirement of Ed Cole on October 1, 1974.
Although Cole had been enthusiastic about airbags, division managers apparently were not, nor was Cole’s successor, E.M. (Pete) Estes. Within six months of Cole’s retirement, GM was already saying that it would no longer offer airbags on any cars in the 1977 or 1978 model years. In mid-1976, Estes briefly suggested that GM might continue to offer airbags on the Cadillac Eldorado and Oldsmobile Toronado for 1977 and 1978, but that didn’t actually happen.
Why didn’t the airbag option sell better?
This has been a controversial question since the 1970s. GM insisted at the time and for years afterwards that the ACRS option simply “was not successful in the marketplace.” Airbag proponents contended that GM had deliberately sabotaged the option’s commercial prospects in order to demonstrate that there was no demand for airbags, despite surveys and marketing studies showing that people WERE interested.
There is no question that the Air Cushion Restraint System option was clumsily marketed. Although officials of the three production divisions obviously knew about the option ahead of time, the actual launch was a mess: The ACRS didn’t go into production until November 29, 1973, two months after the start of the 1974 model year. While ACRS-equipped cars were theoretically available for delivery by January 1974, the divisions didn’t add the option to their dealer data books until February. So, even dealers and salespeople who were interested — and there were some — were sometimes unsure if or when it would actually be possible to order airbags. The availability of the option wasn’t officially announced to zone sales managers until March 1, and GM’s first and only corporate print ad (above) didn’t appear until a few days after that. Within just two weeks of this ad, GM officials were already telling reporters that sales of the airbag option were disappointing!
In addition to the March 1974 print ad, there was a 10-minute “Mini-Theatre” film about the airbag option, which was provided to dealers for sales training and to show to prospective customers.
For the most part, the divisions, which were principally responsible for advertising and marketing, didn’t do much to promote this option. The airbag system was omitted entirely from some early 1974 brochures, although there were ACRS-specific brochures later in 1974. The illustration above is from the Oldsmobile version.
Later model-specific and full-line brochures and ads sometimes mentioned the airbag option, but usually only in passing, and often without explaining what the Air Cushion Restraint System was or why a buyer might want it. I think the problem was that the divisions didn’t know how to market the ACRS. It wasn’t exactly a convenience option (except insofar as it didn’t require the ignition interlock), but describing it as a safety feature might suggest that cars without it were unsafe, which wasn’t a message GM wanted to send.
Marketing studies confirmed that consumer awareness of the airbag option was poor. Customers who did know about the airbags often had questions about the system that dealers were ill-equipped to answer, like how much it would cost to replace the airbags after deployment. Dealers wanted more tools to help them explain the system, and felt that the divisions should have offered sales incentives to encourage salespeople to promote the option, as Cadillac later did with the diesel V-8 in the Seville.
Availability of the ACRS option was spotty. In later years, GM insisted that it had established enough production capacity to sell 100,000 systems per year, or about 300,000 over three years. However, GM’s statements to the NHTSA at the time said otherwise. Ed Cole told the NHTSA in August 1973 that GM would not be able to install more than 50,000 systems for the 1974 model year; GM director of automotive safety engineering David Martin told NHTSA officials in May 1975 that GM could produce no more than 10,000 additional systems during the rest of the 1975 model year and all of 1976. So, actual capacity in the 1974–1976 period was no more than about half the 300,000 units GM later claimed. Availability eventually became a further problem for dealers — by 1976, ACRS-equipped cars had to be special-ordered, and could take up to three months to arrive.
GM conducted several marketing studies of consumer response to the airbag option, so they had some idea of the problems. However, none of those studies was used to adjust the marketing strategy. One was conducted in May 1971, well before the ACRS became available. The others were held in May 1975, shortly after GM decided to discontinue the ACRS option for 1977, and in December 1976, after the option had been dropped. None of these studies was used to adjust the marketing approach, and the latter two were essentially postmortems.
On September 27, 1979, three years after the ACRS was dropped, Pete Estes appeared on NBC’s The Today Show and admitted to host Phil Donahue that GM “didn’t really push airbags.” Nonetheless, in April 1981, Betsy Ancker-Johnson, VP of GM’s Environmental Activities Staff, testified before Congress that airbags were “economically impracticable” and that the limited sales of the 1974–1976 Air Cushion Restraint System and a similarly unpopular automatic seat belt option offered on the Chevrolet Chevette “must be viewed as reliable and persuasive measures of the absence of any significant consumer demand for passive restraints.”
In a February 1983 statement to a U.S. Senate subcommittee, Insurance Institute for Highway Safety (IIHS) senior vice president Ben Kelley lamented, “Had GM … marketed the crash-protection benefits of its air bags even half as aggressively as it markets stereo cassette decks, vinyl roofs, and ‘convenience packages’ — it is hard to think it could not have created a demand for air bags on the part of many car buyers.”
How did GM get back into offering airbags?
That is a long and messy story, which is really too complicated to fully recount here.
In January 1977, GM signed an agreement with Gerald Ford’s secretary of transportation, William T. Coleman Jr., to produce up to 300,000 more ACRS-equipped cars during the 1980 and 1981 model years. The incoming Carter administration was dubious about that agreement — new NHTSA administrator Joan Claybrook didn’t think it was legally enforceable, and new DOT secretary Brock Adams was concerned that it would actually delay wider installation of airbags — but they asked GM to adhere to the deal anyway. GM instead dragged its feet; announced a dramatic increase in the estimated retail price of the airbag option (to a hefty $581); and then repeatedly pushed back its schedule for the planned airbag reintroduction. For a few months in 1980, they continue to report that airbags would be available on some cars for the 1983 model year, but in April 1981, GM instead announced that its air cushion restraint program had been canceled entirely.
Around the same time, the Reagan administration moved to completely rescind the MVSS 208 automatic protection rule, which still hadn’t been implemented. This resulted in a federal lawsuit (State Farm v. DOT) that went all the way to the U.S. Supreme Court. SCOTUS ruled in June 1983 that rescinding the passive restraint rule had been unlawful and ordered the NHTSA to revise its decision. In July 1984, secretary of transportation Elizabeth Dole issued a new final rule that restored the passive restraint requirement and set a new phase-in schedule, beginning in the 1987 model year. By September 1, 1989, all new U.S. passenger cars would be required to have either passive seatbelts or a driver’s side airbag.
In 1988, after years of excuses and empty promises, GM once again began offering an airbag — driver’s side only — as an $850 option on the 1988 Oldsmobile Delta 88 Royale Brougham. It wasn’t until the 1989 model year, when a driver’s side airbag also became optional on most Cadillac models, that GM airbag installations actually topped the 100,000-unit mark. Well into the ’90s, many GM cars still had cheaper passive seatbelts instead, which later became a commercial liability as buyers demonstrated that they preferred airbags to automatic seatbelts. Any technological lead GM might have had thanks to the ’70s ACRS program had long since been squandered.
So, what was the point of offering the airbags at all?
Back in October 1973, GM automotive safety engineering director David Martin urged the NHTSA “to take advantage of the GM air cushion program for 1974-75 and use this experience as a basis for determining the future course of MVSS 208.” However, what GM really wanted for the “future course of MVSS 208” was for it to go away.
Ed Cole may have loved the idea of banishing seatbelts in favor of airbags, and he was happy to spend $60 million of the corporation’s money developing them as a technical exercise, but GM was adamantly opposed to any federal requirement that they install or even offer airbags or other passive restraints. In January 1977, about two and a half years after his retirement from General Motors, Cole conceded in a letter to IIHS president William Haddon Jr. that “the only way passive restraints are going to get to first base is to make them mandatory,” but he certainly wasn’t willing to admit that while he was still answerable to GM shareholders.
The GM Air Cushion Restraint System saved a number of lives — that much is clear from the accident case studies. However, it was sold in such small numbers that its overall impact on auto safety was minuscule. Even the accident data GM gathered was really too scanty to be statistically useful, and GM muddied the waters by making confusing, contradictory statements about that data. (Harold Mertz’s 1988 paper provided a much more coherent summary.) GM didn’t get much out of the program except another club to swing at the NHTSA and Congress in insisting that airbags would never be accepted by the public.
Contrary to what GM and other automakers insisted, airbags did not turn out to be “economically impracticable,” and buyers eventually embraced them despite their cost. Federal law has explicitly required dual front airbags and three-point seatbelts for all new U.S. passenger cars since the 1998 model year, and for all new U.S. light trucks and multipurpose vehicles since 1999. The safety value of widespread airbag adoption proved to be considerable: The NHTSA estimated that from 1987 to the end of 2008, airbags saved 28,244 lives in the United States.
However, designing airbags to act as passive restraints (rather than as a supplement to seatbelts) came at a price: Between 1990 and 2009, the NHTSA recorded 296 confirmed airbag-related fatalities, 191 of them children. Starting with the 1998 model year, the NHTSA reluctantly allowed “depowered” airbags, which dramatically reduced the number of airbag-related deaths at some cost to the airbags’ value as passive restraints. It wasn’t until the arrival of more sophisticated “smart airbag” systems in the ’00s that this basic contradiction was finally resolved.
Related Reading
Vintage Road & Track Report: The Facts Behind Airbags – Do We Need Them? *Updated (by yohai71)
Air Bags play a particularly important role for Americans, who for the most part were late in taking up the wearing / using safety belts.
GM fitted front seat safety belts as standard equipment from as early as 1964, but for unknown reasons, American buyers rarely used these safety belts and in some cases even removed them. Even in 2026, large passenger buses in the USA often provide passengers with excellent safety belts but few, very passengers use them.
The Safety Belt remains the single most important safety feature, but sadly even in 2026, there are still people who don’t appreciate how important it is to use safety belts.
Australia was the first country in the world the introduce and sticky enforce laws mandating the use of safety belts. Australia mandated to the use of safety belts in 1969! These laws together sticky enforced speed limits. You cannot exceed 110 Ks (about 70 MPH) anywhere in Australia and many major freeways have 80K (45 MPH) limits.
But Australia has a very small road toll when compared with other developed countries. The USA kills an average of 45,000 people everything year and seriously injures over 300,000 each year. This must cost America a lot of money but is not surprising as across the USA many cars and trucks regularly drive at 80 to 90 MPH speeds. Put simply the Americans simply drive too fast.
Mobile phone use has only added to the above problems. Mobile use needs to be banded totally from cars. As good as airbags are they pose a serious risk to short adults (under 5 feet tall), these short people need to sit to close to the airbag.
Great article nevertheless.
The first Australian seatbelt use law was in Victoria, which was announced in November 1970 and took effect on 22 December of that year, not 1969; it initially didn’t apply to children under 8. Seatbelts were not compulsory in all Australian states until January 1972. (The 1969 Australian Design Rules had required manufacturers to INSTALL belts, but mandating that occupants use them was a separate matter.) However, Australia was definitely a leader in seatbelt law, and the aggressive enforcement had a huge impact. I have a 1984 paper that reports that by November 1982, Victorian driver seatbelt use was up to more than 95 percent and outboard front passenger use was over 91 percent, which cut the vehicle occupant fatality rate by more than half.
The U.S. was substantially more resistant to seatbelt laws, and to seatbelts in general. Even after states here started passing mandatory use laws in the ’80s, overall seatbelt use didn’t top 50 percent until the early 1990s. Once it did, the fatality rate plummeted dramatically.
U.S. traffic fatality rates today are actually quite low. The U.S. Department of Transportation measures both total fatalities and the fatality rate per 100 million vehicle miles traveled, which reflects the number of vehicles on the road and how much they drive. The NUMBER of fatalities in recent years has been as low or lower than the U.S. in the ’80s, but the fatality RATE is roughly half, or less than half.
It’s worth pointing out that the vast majority of US fatalities are due to drunk or otherwise impaired (drowsy) driving. If you’re alert and sober you’re far less likely to be involved in a serious crash, even if you’re recklessly hurtling along at a terrifying 46mph.
It’s a lot, but it’s not the vast majority. NHTSA data puts the current figure at about 30 percent of traffic fatalities.
This must cost America a lot of money but is not surprising as across the USA many cars and trucks regularly drive at 80 to 90 MPH speeds. Put simply the Americans simply drive too fast.
The top speed limit for New Zealand is 68 mph, and the nation has a higher fatality rate (7.2 per 100 billion vehicle kilometers traveled) to that of U.S. (6.9 per 100 billion vehicle kilometers traveled).
The German rate is even lower than the U.S. rate – 4.2 per 100 million vehicle kilometers traveled, even though there are large portions of the Autobahn with no posted speed limit.
Fascinating. My only concern with airbags is that once they are old enough, they are as compromised as any other aged component in a car but far more of a hazard if it deploys due to deterioration.
That said, it is a brilliant life-saving invention. I drove some Valiants without seatbelts in my 20s, including interstate, and revelled in the feeling of freedom. Now I feel naked without a seatbelt, particularly on the more barge-like conveyances I’ve enjoyed, as it holds me to the seat whilst conveyance wallows. Great piece Aaron.
I share your concern about them as they age.
And yes, I drove my Beetles all over the country without belts without ever a thought of that windshield so close to my face. And I was ejected from one of them in a crash, which very luckily I survived with almost no injuries except a concussion. Beetle doors had a bad habit of popping open in even low speed crashes.
The frequency of doors popping open on ’50s and early ’60s cars was really alarming. Cornell ACIR data in the ’50s showed that the doors popped open in about 45 percent of occupant-injury-causing accidents. Safety latches reduced that to around 28 percent, which was better, but still not confidence-inspiring.
One of the most pernicious arguments Americans made against seatbelts was the notion that it was better to be “thrown clear” of a bad wreck. Even GM engineers promoted this idea in the ’50s, although they knew it was false — ejection was the leading cause of occupant deaths.
Being thrown clear was the default safety position of racing drivers in open wheelers until the 1960s. While it sort of made sense for them, perhaps the notion transferred to civilians in tintops from this legacy.
I think there has been some effort to make modern airbags fail-safe (meaning that, in theory, they’d be more likely to NOT deploy because of a fault than the reverse), but the phrase “aged pyrotechnics” is definitely always somewhat nerve-wracking.
According to a post I saw on Facebook the other day, Kramer’s “Assman” car was airbag-equipped. It was a green Chevy and looked like a 1973 model, so it’s at least possible. Interesting article BTW!
Indeed it was. Close inspection reveals the Olds 88 dash and unique steering wheel.
I’d sure like to know the back story on how the producers got ahold of that car.
You totally missed the first standard ACR equipt US cars, the 1989 Chrysler Lebaron coupe and convertible. I had a 1989 Lebaron parts car I bought for $50 with no title for the front end parts to repair a 1989 Lebaron convertible (also an ACR equipt car) for my then 15 yo granddaughter.
Yes; this article was about the history of GM’s air bags.
How did GM get back into offering airbags? Because eventually the federal government said you couldn’t use those stupid door mounted seatbelts as a substitute, they had to offer airbags with conventional seatbelts.
This was a great read.
Well into the ’90s, many GM cars still had cheaper passive seatbelts instead, which later became a commercial liability as buyers demonstrated that they preferred airbags to automatic seatbelts.
Like Matt said above, these were just stupid – and potentially unsafe in 4-doors – door-mounted seatbelts that met the letter of the law but I don’t think there’s one owner who actually used them as a “passive restraint system”.
It’s incredible that the first U.S. automaker to make air bags available thumbed its nose at the law until they had no longer had a choice.
GM did resume offering airbags on some models well before that, including on Cadillacs. However, the knowledge (by 1991) that the feds were going to require airbags by 1997–1998 did probably accelerate their adoption on the remaining holdouts.
Wasn’t Chrysler Corporation the first American company to make air bags standard equipment in the late 1980’s and early 1990’s?
Yes, in 1989. This article is specifically about the history of GM’s air bags.
Chrysler made a drivers side airbag standard on some 1989 K-cars and the M-bodies. However the 1989 Lincoln Continental had standard dual front airbags, a first for US cars.
Great article on a subject I knew little about other than they existed. There is a reason every race car on this planet has seat belts. Buckle up! The life you save could be your own!
Good information here .
Not wearing seat belts is stupid . period .
Pops added seat belts to our cars in the late 1950’s .
-Nate
A very interesting article. I had always associated early airbags with Oldsmobile and didn’t know you could get a Buick or Cadillac equipped with them. When I was starting out as an apprentice mechanic I clearly recall being warned never to work on the steering column or under dash wiring of a car so equipped unless I had the service manual. Many years later when I was a shop foreman we had an accidental airbag discharge while a mechanic was wiring up a 2 way radio in a ’96 or so Ford.
He was working under the dash and not injured, and he either didn’t know or wouldn’t admit how it had happened. The Ford dealer suggested that 12 volts had been applied to the wrong place as he handed me a jumper wire found in the passenger footwell along with a hefty bill.
I can also think of 3 workers who survived collisions relatively unscathed that likely would have resulted in at least serious injuries if not for the airbag. We take this stuff for granted today, and it took far too long to become a standard.
Here in BC seat belt use became mandatory around the time I started driving in the late ’70s, and it was quite rigidly enforced for awhile. I was recently reminded of this by a very young RCMP officer who stopped me for not wearing my seatbelt, saying he knew I didn’t have it on because he didn’t see the shoulder belt. He seemed amazed that my ’65 Corvair didn’t have them. Or airbags. Or a collapsible steering column. Or even 4 way flashers. He didn’t write me a ticket. He did comment that I might as well be riding a motorcycle as far as protection goes.
We’ve come a long way, even if GM and others had to be dragged kicking and screaming the whole way.
My parents paid to have seatbelts installed in our 1958 Plymouth in the early 1960’s. There were four belts installed, so whoever sat in the middle had to share a belt with the occupant sitting by the window. Which meant I, as the youngest kid, had to share a belt with my brother. I absolutely HATED doing that. My parents would try to make a game of it, announcing “Fasten your seatbelts, prepare for takeoff!” like we were in a plane. To ten year old me that was, as the kids say now, totally cringe.
The Bonhams auction promoting ‘possibly the last example remaining’ seems to have been in 2009, at the Washington, D.C., auto show in 2012 a similar vehicle was seen in the safercar.gov area. If it’s the same vehicle, the government agency (NHTSA) had made a few customizations since acquiring it.
I remember these, and still probably own a copy of GMs airbag system brochure from 1974 or 75. I remember thinking that this was the future, but the future took longer to arrive than I expected!
By the time the new law took effect in the 80s, I had my qualms about air bags. But the alternative was clunky, uncomfortable, obtrusive “automatic” belts. From then on I preferred air bags.
Great article!
Aaron–this was a very interesting article; in 1974 I was living in Kansas City and remember reading an article in the Kansas City Star about a 1974 Oldsmobile equipped with the air bag. The driver was involved in an accident and the air bag deployed; the driver was uninjured. I remember a photograph in the newspaper of the vehicle showing the air bag after deployment. Apparently GM was doing little or no advertising of their availability as I previously knew nothing about this until I read your interesting article.
Was that the incident of the man whose Oldsmobile hit a bus? The accident is described in the case histories; those don’t have names, but the story was written up in the press at the time and said the driver’s name was Arnold Arms, a physician. He apparently lost consciousness behind the wheel of his 1975 Olds Ninety-Eight, crossed the center line, and hit a (fortunately very slow-moving) city bus in Overland Park.
Aaron-thank you for filling in the details, the accident was a long time ago, and I had forgotten about some of the details. Thank you for filling in the details!
And why would Ford test airbags with a 1958 Chevrolet?
Probably because they already had one lying around from another project or engineering evaluation.
The late ’50s Ford program was an experimental study, not any kind of production program. It was VERY common for advanced engineering or research projects to use other automakers’ cars. For instance, some of Ford’s FWD engineering work for what became the Cardinal project was done using second-hand Saab cars. The phrase is “test mule.”
What an interesting article. My memory, now 63, tells me that my dad talked about an ACRS equipped GM car at one point until he found out that they didn’t offer the option of tilt wheel at that time (he always appreciated that option as he was tall and it made his entry and exit easier)
At first, I was not a fan of air bags especially after the wife’s aunt had one deploy in a newer Maxima ( i belive the first year they were offered) She was cut up very badly from that air bag. But then I realized how important they were and began to embrace the idea.
Yes, the ACRS was not compatible with the tilt/telescope wheel, so your dad was right about that. That actually had less to do with the airbag itself than with the special energy-absorbing steering column that went with it, which was different than the regular one — it had to be able to take a higher vertical load than the standard column. Saginaw said they were working on applying the same changes to the tilt and tilt/telescope wheels “for a later introduction date,” but it didn’t happen. Of course, a lot of newer cars have airbags with tilt and telescope wheels.
And today everything tilt/telescope wheel has to be unlocked, moved and locked. No more pull the lever and the spring pulls the wheel up and out of your way.
My sister broke the turn signal lever off of his 76 Cadillac and he had a tough time finding a shop other than the dealer who could fix it. He finally asked one mechanic and it was explained that they had to keep the telescope portion compressed.
The 1990-1993 Ford Mustang didn’t offer a tilt wheel option but all the other airbag equipped Fords did. For some reason I guess there was an issue with the Mustang’s steering column or the dashboard to where they couldn’t offer it.
I learned of airbags after reading an insurance-company advert, I think Allstate, that showed a wrecked Olds 98 that the driver survived without major injury, pushing for airbags to become mandatory. That’s the only reference I saw about airbags relating to GM. I also recall a Consumer Guide test of the Marquis praising the huge glovebox, which they noted was originally designed to house an folded airbag. I would have thought the likes of Consumer Reports would have chastised GM for discontinuing airbags in 1977, but no. It was all praise everywhere for the new downsized B/C bodies, and no mention there or anywhere else that you could no longer get the dual front airbags (the discontinuation of early ABS cars went similarly unnoted.
> In human volunteer tests with average-size adult male drivers, airbag deployment didn’t usually cause injuries more severe than minor bruises and abrasions.
Who would volunteer for doing something that had a decent chance of physically hurting them? This also illustrates a problem the persisted until recent years – crash test dummies and safety systems themselves were designed for male bodies, ignoring about half the population that was female (and usually kids of either gender).
The 10-minute “mini-theatre” was likely not a videotape – 1974 predates either Betamax or VHS. It was likely a 8mm or 16mm film cartridge or reel, played in a machine I occasionally saw in the office room area adjacent to the main showroom, away from where customers usually were. These machines were used primarily to train sales staff; I doubt many potential buyers saw this.
At what point was it decided that airbags weren’t really suited as passive restraints (no fastening seatbelts required) and instead sold as “supplementary restraint systems” (SRS) to be used along with conventional lap and shoulder belts?
Yes, Allstate took out a number of pro-airbag ads — they actually did more to promote the ACRS option than GM did. They also had an airbag fleet with about 600 cars, which at times included airbag-equipped Mercury, Chevrolet, and Volvo test program vehicles as well as some production ACRS models.
That isn’t as difficult as you might think, generally. The specific answer in this case is “Air Force guys”: In late 1970 and the first half of 1971, GM, the NHTSA, and the USAF did a cooperative human volunteer program at Holloman AFB, with 35 impact sled tests on 16 human guinea pigs from the Aerospace Medical Research Laboratories. There were various safety precautions, so it wasn’t just Spam in a can, and since all the volunteers were between 19 and 28, I suspect they were more dismayed by having to wear white leotards than the possibility of getting injured.
Yes, that was a big problem that the NHTSA didn’t begin to address until the late 1990s.
You’re probably right. The retrospective marketing study (which was compiled in the early ’80s) described it as a “videotape.” However, now that you mention it, I found a 1982 article on GM dealership training programs that says that “prior to 1979” GM used Super 8 film cartridges, so that’s probably more likely than tape. Was the machine you recall one of these contraptions? https://www.mecum.com/lots/492674/1970s-chevrolet-dealership-mini-theatre-display-cabinet/
The ACRS Mini-Theatre film was shown to at least some customers — in retrospective interviews, several customers recalled having seen it at the dealer. Dealers seemed to show it to interested customers because they didn’t really have any other tools to illustrate how it worked. Some wished for some kind of demonstrator model, which I think would have been difficult, but goes to show the difficulty in selling an option where there was nothing for customers to see or touch.
I think most automakers of the late ’80s and early ’90s described their airbags as SRS, probably as a hedge against liability. For instance, on the 1988 Oldsmobile Delta 88 Royale Brougham, the airbag was called the Supplemental Inflatable Restraint. However, the recognition that airbags alone were a problem was really in the mid-’90s, when the press picked up on the cases of airbag-related fatalities and serious injuries. The NHTSA was very conflicted about this, but they finally had to start taking serious action around 1996.
I don’t recall the exact appearance of the Mini-Theatre, but I only recall the smaller blue part shown at the right side of those photos, although they may have been placed atop the larger brown part. The little specks and dots that flicker on the film shown here are indicative of film prints, not usually seen on videotape (unless it was converted from film). I know I watched a few of them circa 1976 at dealerships, but the ones I saw were mostly aimed at sales staff showing how their cars were better than their competitions’. I don’t think potential buyers were the main audience, although some salespeople showed them to prospective buyers. Many of these 10-minute-or-thereabouts films are now on YouTube or elsewhere online.
Also worth noting is that the car buff books like C/D and R&T were very against airbags in the 1970s.
Fascinating article, Aaron! When I joined the Insurance Institute for Highway Safety in 1979, several senior staff including my first boss had the experimental airbag equipped 1973 Chevys and 1975 Volvo 240s. By that time, all were fitted with lap belts up front but no shoulder belts. Later, my boss acquired one of the production Oldsmobile 98s. It was originally black but he had it repainted white (not a good idea because any dings in the paint showed the black underneath).
As you mention, IIHS was front and center in pushing for both airbags and mandatory belt use laws (the latter after political will on the part of state governments made them feasible).
As for when airbags came to be regarded as supplemental restraints (to be used with lap/shoulder belts), Mercedes-Benz deserves credit. The company started equipping some of its 1984-85 models with driver airbags and made them standard in all US models (still driver-only) in 1986. IIRC, this was the first use of the abbreviation “SRS” (supplemental restraint system) stamped into the airbag covers.
Many of the pics here came from my best friend Phil in Quebec who owns the green ’75 Electra, which I’ve ridden in many times, he restored it himself, it’s a very nice car. Another shot of it:
pic 2
I just saw this thread after Roger told me about it. Finally saw some production numbers! All I knew is that 123 cars like mine (Electra Limited 4 door hardtop for 1975) had it. I have seen a few more including another of those 123 with the Park Avenue package that I wanted to buy as a teenager. I hate shoulder belts and I’m not a big convertible fan so I always had an interest for the last of the cars without shoulder belts!
Here’s a picture from the driver module that I had to remove to fix the horn.
I’m wondering about the value and effectiveness of an airbag that’s never deployed, for example, on these early ’73 Chevys. Would you trust it? …or have the components deteriorated?
In theory, the airbag warning light should illuminate (and stay on) to warn of malfunctioning components.
I can say as a retired staff member of the Insurance Institute for Highway Safety that we crash tested two different 1973 Impalas when they were older cars. The first was in 1985 when we removed from outdoor storage a nonrunning car and simply replaced the 12-volt battery. The airbags fired normally when tested.
The second was for the dedication ceremony of our new crash testing facility in December 1992. The car we used was a running example and as in the first case, the airbags fired normally in the crash test.
I have a 1998 Nissan Frontier purchased new, and I’m not worried about the airbags not working properly. The warning lights (for both the system and the passenger airbag specifically) illuminate for a few seconds when I activate the ignition switch and there’s no visible deterioration of the airbag covers in the steering wheel and instrument panel.
I see a number of potential concerns, which unfortunately don’t have easy answers. The first is that when any car hits age 50, many components may have deteriorated: wiring, plastic connectors, the nylon airbags themselves. (The fault detection system would probably indicate the former, but not the latter.) The second is that Fisher Body developed special tools for servicing the ACRS, including an electronic analyzer for tracking down electrical faults. Finding these tools in serviceable condition now would not be easy. The third problem is that replacement parts are now nonexistent, so if you identify corroded wiring or some other fault, fixing it would be daunting.
These kinds of problems are not really specific to airbags so much as any rare option on an old car that’s well beyond any window of manufacturer support, but they are concerning.
Very interesting, and a good overview of the first airbags. I was aware of GM’s tentative offering, but wasn’t aware that Ford had done some experimental work on them much earlier. It’s not too surprising that Ford would do such a thing in that era – a time when they had their own out-of-the-blue safety initiate. It’s also telling that both were under-marketed and not embraced in the market. I could be wrong, but the first make to really embrace air bags was probably Mercedes: as I recall, they were an option (driver’s side at least) early in the 80s, and most Mercedes had them going into the 90s.
Anyway, I’m a little surprised at the naïveté of Ed Cole at least (less so at Ralph Nader) to think of air bags as a replacement for seat belts. The reality is, an unrestrained occupant is a huge physics problem that even highly sophisticated modern systems struggle with. A well-engineered, pre-airbag car (like a Mercedes or Volvo) will still protect a belted occupant far better than the safest modern car will an unbelted occupant in most situations. Air bags do some good, but are still a secondary safety feature to seat belts. You don’t see air bags in NASCAR or F1 because they don’t need them. On the other hand, if you think compliance was bad 50 years ago, just imagine asking people to put on a restraining 5-point harness and helmet – incidentally, figuring out head restraints was one of the biggest pieces of the puzzle in preventing injury and death in racing accidents. A good case study was Dale Earnheart’s death, which I guess you’d say I watched live on TV.
I have two other quibbles with air bags (as well as many other safety features). One is cost and complexity: the “this will save lives” argument is valid, but everything in life is a trade-off. A safer car doesn’t do you much good if you can’t afford it, or when the systems inevitably break down and fail. The emphasis should be on completely passive safety measures, lest you end up with a car that’s golden in crash tests, but a death trap ten years later when the electronics start to fail. And the suggestion that we mandate away old cars “for our safety” is completely unacceptable to me. It has to come down to good, educated choices by consumers more than regulation. The other problem is implicit trust in technology – Titanic thinking. The only accident you’re guaranteed to survive is the one you avoid. I’d much rather see more stringent requirements for driver’s licenses than on the cars themselves. Maybe we’re still behind the curve on that, but regulations on cars themselves have become onerous – only this time the big automakers are all for it, because you either have to buy all the technology required from them, or have incredibly deep pockets (like deeper than all but a handful of corporations) to build your own engineering and supply chain. The only successful major automotive startup of my lifetime is Tesla, and that would never have happened without massive EV subsidies that the big makes were too slow and inflexible to take advantage of.
Stricter driver licensing and more comprehensive driver training should in theory produce safer drivers. But in reality (and we’ve known this since the 1960s), this will not be the case. People already know that they shouldn’t speed, run red lights, drive while impaired, and text while driving, but they choose to do these things anyway. Improved knowledge of safe driving habits will not lead to correspondingly better behavior, except of course during the road test when a state trooper is sitting beside them.
Also, cars and drivers exist in the real world, where there are innumerable factors that are beyond any one person’s control. Even an expert driver may be distracted by a screaming child and look away at the wrong instant; there may be an unexpected wet or oil-slick patch of road, or a piece of debris that flattens a tire; a car sideswiped by another vehicle that was trying to avoid a child who ran out into the road may be turned into a 1.5-ton billiard ball. I knew someone who was badly traumatized by a car accident in which they were rear-ended at a stoplight and pushed into the car in front of them hard enough to trigger airbag deployment — there was absolutely nothing they could have done differently to avoid or mitigate it, which in some ways was worse, at least psychologically, than if they had made some avoidable mistake.
I used to rant about our country needing better drivers, not idiot proof cars, way back when. Eventually I realized that cars have become appliances, and that public transit is not an option for most people in the US, including the many millions who would rather not fool with owning or driving a car if they didn’t have to. Then, out of my twenties, I realized I was also no longer infallible nor immortal, and never had been. Now, as a forgetful old retired guy, I am happy that our daily drivers all have airbags, and several other improvements that make survival, when somebody blows it, much more likely.
Excellent work, Aaron, as usual. Thanks.