History of the CPSC's Requirements in its
Bicycle Safety Standard

home                    lights


This is in response to Robert Lockard's inquiry for more information about the history of the federal government's bicycle safety standard that requires, among other things, the all-reflector system of nighttime safety equipment that physically cannot work. The following account is long and complex, both because the story is complicated and because of my intimate knowledge of the events. If the subject of governmental regulatory and scientific mistakes interests you, read on. Others may pass to the next letter.

In 1972 I managed an engineering department and was a prominent veteran-aged bicycle racer. In that year, four events affected my life. Three different government bodies started making mistakes about bicycles and cycling. One body prosecuted me for not riding as dangerously as they required, which, naturally, I had refused to do. After being convicted, I tried to obey their law, and in half an hour it damned near killed me - I survived only by considerable skill aided by lots of luck. The fourth event was a disagreement with my employer about what my engineers should be looking at; the bosses wanted quick cost savings while I thought about a better-operating manufacturing operation in a longer term. I therefore decided to fix the three governmental problems about cycling before going on to my next job. I thought that it would take two years, but I am still doing it. The law that damned near killed me was rapidly repealed; the other two are still items of controversy.

The Consumer Product Safety Commission of the United States was preparing to issue its bicycle safety standard. The authority it used was the Hazardous Substances Act, which allowed regulation of "toys or other articles intended for use by children." Its starting point was the Bicycle Manufacturers' Association specification BMA/6. BMA/6 described a crummy bicycle of the type bought by non-cycling parents for their children, because that was the only type made by the firms who belonged to BMA. Some of its requirements pertained to safety, including the one for the all-reflector system which was far more dangerous than the state-required headlamp and rear lamp or rear reflector. Others pertained to durability, the ability to stand up to neglect and abuse. Others pertained to manufacturing convenience. The CPSC's engineers thought that all of these requirements were correct safety requirements and wrote them into its proposed standard. The proposed standard did not limit itself to bicycles intended for children.

As Robert Lockard correctly stated, proposed regulations of this type must be published for public comment for 90 days before final issuance. What organization should respond to a proposed safety standard for bicycles? The obvious organization was the League of American Wheelmen, the only national organization of cyclists with such concerns. (The national racing organization paid attention only to racing.) LAW was an amateur organization with little money, and at this period it had hired its first executive director upon payment of a subsidy by the BMA for this purpose. That XD made damn sure that the LAW's directors did not hear about the proposed bicycle standard until after the allowed comment period had expired. I had been alerted, read the proposed standard, was apalled, and within the allowed time wrote extensive comments that detailed its engineering errors. After that comment period, the only parties who had standing to take further action were those who had already made comments within the allowed period, those who made or sold bicycles, and those who purchased bicycles. LAW, the national rganization of bicycle consumers, which did none of those things, was forever excluded.

I described the standard in the cycling press, and the CPSC received a torrent of angry mail. Those who rode good bicycles were angered at the prospect of having to ride crummy toy bicycles, and a significant proportion of the letters were literally obscene. The CPSC's chairman quieted the uproar by making a public announcement that those who wrote these letters need not worry because the standard applied only to "toys or other articles intended for use by children." Once the uproar had died down, the CPSC quietly decided that since some children had grown as large as some adults, every bicycle sold in the United States was intended for use by children and must therefore comply with the standard. The exceptions are very few, and include some bicycles that are specifically built for use by children that cannot be used by any normally sized adult! The idea that a bicycle that was built for me (custom made bicycles are a significant part of the good bicycle business), to my dimensions, with the equipment that I specified, for long distance touring or for road racing, or one built for Greg LeMond for use in the Tour de France, could be intended for use by children, is a legal fiction that is beyond comprehension. Yet that is the way that the law stands today.

It became obvious that the standard could not stand unless it permitted good bicycles. I regret to say that the most esteemed cyclist cum engineer of the period (from the Great Depression until just before this time, very few cyclists were of the professional or technical classes, except for poorly paid professors) volunteered to help the CPSC revise the standard so that it did not prohibit good bicycles. In that form, the standard was issued.

I immediately sued the CPSC for review and rescission of the standard, citing 16 engineering errors and the improper issuance authority. One other cyclist, a federal employee who knew governmental procedures, also sued about the improper issuance authority. After that, 13 bicycle companies sued. Once the bicycle companies were given sufficient time to sell their present stock of non-complying, and therefore presumably dangerous, bicycles, they dropped out. Think of this. The only persons who carried through objections to the bicycle safety standard of the Consumer Product Safety Commission of the United States were pre-eminent consumers of the best and safest bicycles and one was on the verge of being recognized as the pre-eminent worldwide authority on bicycle safety! The objectors did not want a standard with safer requirements; they objected that the standard should be repealed because it was useless, improper, and in some respects dangerous.

By this time I had spent my savings. I was living a hand-to-mouth existence, surviving on a minute amount of engineering consulting. I ran the case entirely on my own, without assistance from anybody, and with no previous training in the law. In the end, the DC Court of Appeals (I had sued in San Francisco, whose court might have ruled more in my favor, but the national and international bicycle companies required that the case be moved to DC) ruled that in 4 of my 16 points the government had no rational case whatever (which is the legal standard for rescissing a governmental regulation), but not in the most important one, which concerned the danger of the all-reflector system. When I tell attorneys of my achievement against the full forensic power of the federal government, they are amazed that anyone could have achieved that. Their amazement shows just how bad the CPSC's regulation was, and essentially how bad it remains.

Consider the type of arguments made. One requirement was that the front fork withstand running into a curb or a wall. The requirement was a durability requirement that the BMA intended to show parents that if they bought a BMA bicycle they could expect that it would withstand such treatment without needing repairs. The CPSC put it into the safety standard under the impression that it was a safety requirement. The fact of the matter is that if a cyclist hits some object hard enough to get injured, the strength of the front forks is immaterial. The bicycle stops and the cyclist flies through the air, to hit either the object or the ground. Even if you strapped the cyclist to the bicycle with a safety belt, because he is 4/5 of the total mass, he would merely pull the bicycle along behind him. In the course of the initial comments, the legal brief, the response brief, and my rebuttal brief, as I remember, the CPSC produced 3 different engineering arguments as to why it had put that requirement in its regulation, all of them wrong. This isn't rocket science. It is covered in the first-year physics course taken by students of physics (like me) or engineering. However, because the CPSC had produced an argument that the court couldn't see was engineeringly incorrect, the court allowed the requirement.

In another requirement, the standard required that wheel rims be sufficiently strong so that the spoke nipple heads would not pull through the rim when subjected to a large overload. The test had been in BMA/6, so the CPSC adopted it. The trouble was that the requirement resulted from some failures during wheel assembly. Confronted with wheels that failed during assembly, the manufacturers required their rim suppliers to provide sufficiently strong rims, and developed a test to determine this. Sounds great, doesn't it? The trouble was that if the rim survived the rather crude assembly process that was used for the wheels of cheap bicycles, it didn't fail in service. The CPSC argued that when a bicycle went over bumps the tension of some spokes increased, so that some spokes might fail, leading to successive failure of a large number of spokes producing a catastrophic wheel failure that threw the cyclist to the ground. So said its engineers, giving their official opinion. There are three troubles with this opinion. The first is that in the hundred years since the invention of the tension-spoked wheel nobody had recorded such an accident. The second trouble is that when wheels fail they don't fail like this. True, some spokes fail, but spoke failure is random and rarely involves more than one or two spokes. And when wheels fail from going over bumps, the rim is indented, thus loosening the spokes, as can be easily seen. The third trouble, well that was not solved until later. However, the CPSC argued that it was intolerable that the law should be construed to require that the CPSC had to wait until people had been killed in an accident (naming this 'the body count') that its engineering science had predicted before it was authorized to issue a regulation. The court accepted this argument, with its corollary that the CPSC's engineering science was accurate.

Consider the scientific aspects of the government's argument. Engineering science predicted that accidents of type A occurred. The equipment involved in these accidents had been in widespread, worldwide, use, starting more than 100 years before. However, such accidents had never been reported. The government never presented any theory about how the bicycle wheel carries its load, and therefore about how it might respond to overload. However, the evidence from the type of overload that was assumed, going over bumps, had been frequently observed to produce only the opposite type of failure, indentation of the rim that loosened spokes. The government never presented any calculations concerning what it asserted. Since it had no theory, it could hardly do that. Considered as engineering science, the government's argument was hardly convincing. The predicted accidents had not been observed, while the conditions that supposedly caused these predicted accidents had been observed to produce the opposite result. At the very best, this was junk science. Yet the court accepted it. Not merely that. The court held (argued is a better word) that it was proper legal practice to accept this accept junk science.

You may believe that I failed to convince the judges in a fairly simple case in which there could be only one rational answer. I admit that I was under great pressure, having to meet tight schedules, running a case with thousands of pages of record, against the federal government but entirely alone and without money, concerning a technology whose science I was developing at the time, and in a legal process in which I had to research and feel every step of the way because I was entirely untrained in its use. I admit that I did not produce the best possible initial and rebuttal briefs that are possible, but I think that I did damned well.

There are two reasons why I failed, and those reasons concern deficiencies in the governmental and legal processes.

The first is the legal standard that there must be a rational connection between the purpose of a law and the means by which it proposes to achieve it. Sounds great, doesn't it? Indeed, it is necessary. The problem is what the legal and judicial professions consider to be a rational connection. Any flummery of words will suffice to justify the government's decision. Even if the action is supposed to be based on scientific or engineering grounds, it doesn't have to meet scientific or engineering standards, merely lawyers' standards of wordplay. I recognize that most governmental actions are not based on scientific or engineering standards; for example, what income tax rate to charge persons with high earnings. That decision is based on other grounds entirely. Yet in the case of safety regulations, which supposedly are based, and are intended to be based, on scientific or engineering grounds, those standards don't apply. In those cases, the court is required, by precedent, to accept arguments that merely sound rational to the legal mind, regardless of their scientific irrationality or implausability.

The second is that in regulatory cases such as this no trial of fact is permitted. The trial of fact is presumed to have taken place in the regulatory proceedings themselves. Therefore, by the normal standards of law, only legalistic appeals are permitted, and the only materials that can be submitted are arguments and rebuttals to them (in effect only written arguments are effective because of the very limited time allowed for oral argument). I took my bicycle into the courtroom to serve to illustrate some of my points, but the CPSC's lawyers objected to it as evidence, which is not permitted. When the judges told me that excluding my bicycle didn't matter because they would understand me because all of them had had bicycles in their youthful years, I knew that my case was lost. If my controversy had been against bicycle manufacturers who wished to market dangerous bicycles, then it is possible that a reasonable facsimile of a trial of fact could have taken place during the proceedings before the CPSC. However, my controversy was with the CPSC itself, the organization that issued the supposed bicycle safety standard that was, in parts, useless; in other parts, detrimental to cyclists; and in a few parts was outright dangerous. It is as though one had to sue the trial judge who is also acting as his own expert witness, and unchallengeable at that. Had the legal procedures permitted me to have brought the CPSC's engineers to the witness stand I would have destroyed both their professional standing and their credibility before any court.

So what about the third problem with the strength of the wheel requirement? The tension-spoked bicycle wheel had been used for 100 years without sufficient theory of how it worked. (Yes, an engineer for the Royal Air Force, in the time when airplanes still landed on motorcycle wheels, had worked out the theory, but nobody remembered that.) There had been many different arguments, but no solution. Several years after this case, several university groups, and at least one individual who had access to Boeing Co's mainframe computer, were working on the problem. In casual conversation with that individual, beside the road one evening, I suddenly realized the likely solution and how to measure its operation. I went home, spent the next morning building a bit of test equipment, and by noon I had demonstrated the solution. Once I published it, the university groups came through with confirmation that it agreed with what they had been working on by more complicated processes. The bicycle wheel must have pre-tensioned spokes, or it cannot work, which has been amply demonstrated by experience. As it rolls along the ground, the portion of the rim adjacent to where the tire touches the ground is pressed inward, thus reducing the length of the few spokes (2 to 4) that are connected to that portion of the rim. I measured the change in length of a spoke as the wheel rolled and calculated the reduction in tension over the low region. The reduction of tension in those few spokes equals the load carried by the wheel. My first experiment got it within 5%, and that was good enough for me. Greater load, as from going over a bump, would result in greater reduction in tension than in level riding. That demonstration proved that the CPSC's engineering science must have been junk, but the demonstration came four years too late for the case.

How about the all-reflector system, which is the requirement that I really wanted to be removed? That requirement was obviously dangerous because the all-reflector system could not work for some very obvious car-bike collision situations, but at that time there were no detailed accident statistics on which to prove their importance. I had worked on all the other engineering errors because I hoped that demonstrating them all would persuade the court to return the whole regulation for rewriting, which would impress the CPSC's engineers with the need for careful engineering analysis of everything that they had done. The CPSC stated, in its reply brief, that it had considered and researched the nighttime accident situation and it was convinced of the correctness of its original official finding in the Federal Record that "the all-reflector system provided adequate visibility to motorists under lowlight conditions." To support this claim, the CPSC quoted a range of 800 pages of the record. I felt that they hadn't done this, but I had insufficient time when preparing my rebuttal brief to read those 800 pages to discover what they had done.

Therefore, I could not prove to the court that they had lied. The CPSC also misstated the issue by bringing in arguments that I wanted to ride dangerously, or wanted to double the price of bicycles, or that children, who were supposed to be incapable of using lights, would be protected by reflectors. The combination of complexity and limited time overwhelmed me. Years later, when I took the time to research that claim, I discovered one sentence in those 800 pages that concerned this subject. It stated that better illumination, or better reflectorization, might reduce the nighttime accident problem. That's all that it said. It failed to distinguish between street lighting, headlamps, and reflectors, and it referred to no study of nighttime car-bike collisions at all. Yet the court accepted that claim. Two years later, when detailed national statistics of car-bike collisions both in daylight and at night became available, those statistics demonstrated that 75% of the car-bike collisions that were probably caused by darkness were of the type and from the direction in which a headlamp would help but a reflector would not.

Why did the Consumer Product Safety Commission of the United States go to such lengths and to such lies to produce and justify a supposed safety standard that had very little possibility of reducing accidents and greatly endangered consumers? The first reason was that the CPSC was first advocated and then staffed by zealots. They obviously believed that a great many injuries to consumers were produced by products whose manufacturers didn't care about the number of people they injured. The bicycle regulation was the first big task that the CPSC undertook, and if that was demonstrated to be a failure many heads would roll and the life of the organization might be at stake. Bicycles are involved in a great many injuries to consumers, but half of those are caused simply by falling off, a problem that cannot be avoided as long as the bicycle is a bicycle. Defective mechanical design, which is all that the CPSC can regulate, is responsible, except for the all-reflector system, for only about 1% or less of injuries to cyclists. However, the CPSC claimed 17% in an early study, but that was after the accident population had been winnowed of all accidents that clearly were not related to mechanical failure, and included those caused by defective maintenance, which were the great majority of those atributed to mechanical causes, as well as those caused by defective design.

The CPSC was claimed to be a very necessary part of government that would produce a great improvement in the quality of life. However, the first big project that it undertook was one that it was physically impossible to fulfill, because very few of the accidents to users of that product were of types that the CPSC could do anything about. Then it foolishly adopted, without understanding them, some industrial requirements and tests that had nothing to do with safety, plus a dangerous requirement which the industry being regulated had intense interest in getting into law. Afraid that all of this would be exposed if any part became questioned, the CPSC did everything that it could to keep the whole mess concealed, including lying. It should have considered its actual instruction from Congress, which was to protect consumers. The reasonable (I don't like to use the word 'rational' in this sense, because it conceals too much verbal irrationality.) action would have been to rethink the problem and decide what it had authority to do, and to what extent it could use persuasion beyond the scope of its direct authority, that would make cycling safer, and get on with it. Instead, it chose to protect itself.

After its triumph with the CPSC and my case, the BMA tried to get the states to adopt the all-reflector system in their traffic laws. I was the prime opponent arguing against that proposal, and the National Committee for Uniform Traffic Laws and Ordinances overwhelmingly rejected the BMA's proposal. (I think that the proposal received no yes votes at all.)

Now, twenty years after the initial chaotic foolishness, the CPSC has started to reconsider the problem of nighttime protective equipment. It has held two meetings on the subject. However, those meetings have demonstrated both that the CPSC's engineers don't understand the problem of nighttime car-bike collisions and that they are determined to protect the existing requirement for the all-reflector system by, at most, improving the reflectors. Furthermore, the BMA has continued to present its arguments for that all-reflector system, backed up by the makers of reflective materials, most prominently 3M in advocating reflective tires. As I have written above, it doesn't matter how good the reflective materials are, or where positioned, because the motorist's headlamp beams don't shine on the reflectors at the time when the motorist must take the action that prevents the collision.

As promised, or maybe threatened, this has been long and detailed. I hope that those of you who are interested in some aspects of the several subjects discussed herein have found it worthwhile to have perused it.

Return to: John Forester's Home Page                             Up: Lights