DRI-270 for week of 10-6-1: How is Job Safety Produced?

An Access Advertising EconBrief:

How is Job Safety Produced?

The best-selling book on economics in the 20th century was probably Free to Choose, the 1980 defense of free markets by Milton and Rose Friedman. It contained a chapter entitled, “Who Protects the Worker?” In it, the authors highlighted the tremendous improvement in the working conditions and living standards of workers from the Industrial Revolution onward. What, they inquired rhetorically, accounted for this? The Friedmans suggested “labor unions” and “government” as the likely top two answers to any poll taken on this subject.

One of the nation’s leading experts on the subject of risk and safety is W. Kip Viscusi, long an economics professor at Harvard, Duke and Vanderbilt universities and now affiliated with the Independent Institute. In an essay on “Job Safety” for the Fortune Encyclopedia of Economics, Viscusi wrote: “Many people believe that employers do not care whether their workplace conditions are safe. If the government were not regulating job safety, they contend, workplaces would be unsafe.”

The Friedmans and Viscusi knew something that the general public doesn’t know about job safety; namely, that free markets and competition are what keep workers safe. The notion of a “market” for risk or safety seems hopelessly abstruse to most people. The general attitude toward competition can best be described as ambivalent. Still, it is the job of economists to make the complex understandable. Herewith an explanation of how job safety is really produced.

Compensating Differentials

Most people seem comfortable with the fact that wage differentials exist between jobs. Moreover, the direction of difference is not random. Different types of manual labor may differ radically in the element of physical risk to which workers are subjected – coal mining, for example, presents a much higher probability of death or severe injury than does loading-dock work. The greater the risk associated with an employment, the higher the wage its workers will command.

In free markets, wages result from the interaction of supply and demand. Does the “risk differential” reflect variations in the supply of labor or the demand for it? Either or both. The toll of current and future mortality in coal mining – from accidents and “black-lung” disease, respectively – tends to restrict the supply of labor to the profession, driving up miners’ wages on that account alone. Coal’s high-BTU content makes a miner’s output from a 40-hour workweek much more valuable than that of the dock worker, so the demand for coal miners exceeds that for dock workers – a factor also tending to drive miners’ wages above those of dock workers.

This logic extends to other characteristics of employment, beyond those ordinarily associated with risk. Library work is viewed as pleasant because of its low-key, low-stress, peaceful character and attractive environment. This attracts a plentiful supply of applicants for low-rung library jobs (pages and assistants) and the continuous pursuit of graduate degrees in library science (required for librarians). This bountiful supply of labor tends to depress wages within libraries below those of comparable other jobs, such as clerks, cashiers, tellers and such. The particular attractions of library work influence people to accept lower wages than would otherwise be acceptable – in effect, library employees receive part of their payment in kind rather than in cash.

Economists use the term compensating differentials as shorthand to denote and explain differences in work-related remuneration that compensate for differences in how different jobs are perceived or experienced. The phenomenon was first observed and categorized by the great Adam Smith in 1776 in his magnum opus, An Enquiry into the Nature and Causes of the Wealth of Nations. Smith observed that positive wage differences would exist for occupations that were dirty or unsafe, such as coal mining or butchering, those that were odious, such as performing executions, and those that were difficult to learn.

Compensating differentials play a key role in job safety. Opponents of markets – who tend to be the same people promoting government regulation of job safety – insist that employers are too parsimonious to spend money on job safety. And why should they? From the employer’s standpoint, the anti-market man maintains, expenditures on job safety are a waste because they are a cost that does not contribute to the employer’s revenue. The compensating differentials argument supplies a potential motivation for the employer’s investment in safety. A safer workplace will increase the worker’s willingness to accept lower wages, thus allowing the employer to recoup his investment over time, just as if the investment allowed him to earn more revenue.

This positive incentive may have a negative counterpart as well. The ability of workers to sue for tort injury provides an incentive to improve worker safety. (In this regard, Viscusi makes a vital distinction: Firms must correctly understand and anticipate liability in order to feel this incentive. The most famous tort liability case was the massive asbestos liability case, in which longtime principles of tort liability were overturned in order to find large companies like Johns Manville liable for worker illnesses contracted many years before the link between asbestos and mesothelioma was uncovered.)

The Market for Job Safety

The most common way of assessing risk is to calculate the approximate rate of death or injury per annum. For example, a job requiring physical labor entailing moderate risk might result in one death per 10,000 workers per year. Workers in this employment should expect to earn a modest premium – perhaps $500 – $700 per year – over workers doing labor involving essentially no risk of death. Another way to view this premium would be to call it the amount that workers would willingly give up to avoid the risk they bear.  And this amount also sets a ceiling on what employers would pay to improve jobsite safety, since any amount below this will save the employer money, while any amount above it will cost more in safety expenditures than the amount the employer could save in avoided wage premia.

The market for safety is one in which workers assess the risk characteristics of jobs they contemplate. Their assessment determines their willingness to work at that job and the wage at which they will work. It is obvious that the successful functioning of this market demands that workers correctly assess a job’s risk/safety profile.

“How well does the safety market work?” Viscusi asks rhetorically. “For it to work well, workers must have some knowledge of the risks they face. And they do.“[emphasis added] He cites one study showing that 496 workers correctly paired a higher risk of injury with a higher level of danger in their industry. Only 24% of workers in women’s outerwear manufacturing and communications equipment characterized their industry as “dangerous.” But 100% of workers in logging and meat products described their industry as dangerous – correctly, as it turned out.

Are workers ever wrong about the risks they face? Well, they sometimes mis-estimate the level of risk they face, not by assuming it to be zero but by wrongly assuming to be higher or lower than it is. But the evidence strongly suggests that the market does work.

Another datum supporting this conclusion is the general reduction in job risk throughout the 20th century. As real income rose throughout the century, we would expect that workers would take some of their gains in the form of risk reduction; that is, they would deliberately seek out less job risk because the increase in real wages allows them this luxury. In effect, this implies that safety (or risk-reduction) is a normal good, something workers choose to “purchase” more of when their real incomes rise. In fact, that is exactly what did happen over time. Real wages roughly tripled from 1933 to 1970 and average death rates on the job fell from about 37 per 1,000 workers to about 18.

Still another aspect of the market for safety is the incentive it provides to learn. This applies to both employee and employer. Do workers keep track of new information developed about job-related safety hazards? Yes; the evidence of this is the high quit-rate (about 33%) for workers to learn that job risk has risen since their initial hire. Since the hiring and training process is expensive for employers, this represents an incentive for them to hold down those risks.

Government Regulation as a Way to Improve Job Safety

To Americans under forty years of age, it must seem as though the federal government has always been omnipresent in economic life. Actually, the bulk of federal-government regulation is the legacy of two historical periods – the New Deal administration of President Franklin Roosevelt from 1932-1945 and the Great Society regime of President Lyndon Johnson from 1963-1968. Most of the non-financial regulatory apparatus, including the agencies dealing with health and safety, were created in the late 60s and early 70s. The publicity created by the muckraking exposes of consumer activist Ralph Nader played a key role in stimulating the implementing legislation for these agencies. (Viscusi’s career began with the two years he spent as an apprentice in the Nader organization prior to his academic training.)

In 1970, the federal Occupational Safety and Health Act created the agency called OSHA (Occupational Safety and Health Administration). The agency is an attempt to engineer a theoretically safe workplace and implement it by government fiat. This de-glamorized mission statement highlights the agency’s glaring flaw: the substitution of technological criteria for purposes of solving economic problems. OSHA’s attempt to ban formaldehyde from the workplace in 1987 resulted in rulemakings that were estimated to cost $72 billion for each life they purported to save. To add insult to this grievous injury to economic logic, the U.S. Supreme Court ruled that OSHA regulations could not be subjected to any cost-benefit test, thus enshrining the agency’s right to commit acts of fiscal and economic lunacy with apparent impunity.

It seems difficult to believe that the judges could not envision the possibility that the $72 billion committed to saving that life had alternative uses that included saving multiple other lives. Yet the idea that the Constitution should codify any kind of respect for economic logic remains outside the legal mainstream to this day, despite the efforts of scholarly judges like Richard Posner and Frank Easterbrook to bring their substantial economic learning to bear.

Viscusi notes that “increases in safety from OSHA’s activities have fallen short of expectations. According to some economists’ estimates, OSHA’s regulations have reduced workplace injuries by at most 2 to 4%.” He compares the fines OSHA collects in the average year (about $10 million at the time Viscusi wrote) to the size of the aggregate risk premium embedded in U.S. wages (about $120 billion at that point). Obviously, the market for safety was disciplining employers and employees alike much more powerfully than OSHA.

As the Friedmans pointed out, though, “government does protect one class of workers very well; namely, those employed by government.” Government employees have job security and incomes linked to the cost of living. Their civil-service retirement pensions are also indexed to inflation and superior to anything available from the Social Security system most Americans are tied to by law. Those government employees who retire early enough to log enough quarters of private-sector employment to qualify for Social Security benefits can “double-dip” from the government pension trough. Needless to say, this is not exactly the concept that OSHA, et al, were designed to further.

Labor Union Bargaining

The role of labor unions in securing improvements in job safety is limited to whatever provisions the union might succeed in embedding into negotiated labor contracts. Unions cannot add to the market incentives to improve safety – incentives that would exist whether unions existed or not. Indeed, if anything, the opposite is true.

Unions can succeed in raising the wage received by their members. They do this either by restricting the supply of labor by limiting the legal supply of workers to union members, or by legally bargaining for a wage higher than the one that would otherwise prevail in a free market. Either way, the result of this above-market wage is unemployment of labor. To the extent that workers leave the unionized industry for employment elsewhere, the higher unionized wages are counterbalanced by lower wages elsewhere.

The wage premium for risk will represent a lower fraction or percentage of the higher, unionized wage than of the market-level wage. Thus, labor unions dilute or lessen the impact of wage premia in creating job safety for workers.

Risk Compensation

The most powerful development in the economics of risk and safety over the last four decades has been the recognition of risk compensation behavior as an offset to rulemaking by government. In the early 1960s, University of Chicago economist Sam Peltzman began to investigate federal-government automotive safety laws designed to force automobile companies to add safety equipment to cars.

The laws made no sense to him. He could see that car companies had incentives to add safety improvements to cars, provided customers wanted them – and he didn’t doubt that many consumers did. But he didn’t see why the companies had to be, or should be, forced to do something that that might well be in their own interest anyway or, alternatively, might not make sense at all.

Peltzman’s research, summarized in a now-classic 1975 article in the Journal of Political Economy, found that the safety regulations did not improve safety on net balance. That is, they either failed to improve actual safety or the lives saved or injuries avoided were offset by other lives lost and injuries incurred because of the laws and safety measures taken.

The key overall principle at work was risk compensation. Safety measures like air bags, seat belts and anti-lock brakes made people feel safer. Consequently, the most risk-loving individuals drove faster and incurred more driving risk to offset the death-and-injury risk that had been reduced by the new safety measures and equipment.

Peltzman’s results were initially greeted with massive skepticism. But forty years of research have vindicated them resoundingly. The “Peltzman Effect” is now recognized worldwide by social and physical scientists. It has been verified empirically in research involving motorcycle and bicycle accidents as well as automobile crashes, and in such diverse fields as athletics, children’s play, recreational pursuits like skydiving and fields like insurance and finance.

Really, risk compensation is not nearly as counterintuitive as it seems upon first exposure. The logic of command-and-control government rules is that most people are mindless robots who are incapable of perceiving incentives, let alone acting in their own interest – but who are capable of following rules laid down by government.  Alternatively, they are docile enough to pay fines ad infinitum after racking up violations. The glaring exceptions are government rule makers, who are well-informed and well-intentioned enough to make the rules that the robots are supposed to follow.

Nothing about actual human behavior suggests that human beings conform to this model. Evidence clearly reveals human beings as rational subject to the informational constrains under which they all labor. The idea that we react to the presence of rules that run counter to our predilections is fully consistent with this picture. It is perfectly clear why OSHA’s rules have “fallen short of expectations” – because OSHA failed to realize that when they force people to obey rules against their will, they take happiness away that people will strive to regain. That is true by definition; that is what “against their will” means.

The Common Sense of the Free-Market Approach to Job Safety

In free markets, workers demand a “wage premium” to reflect the degree of danger or “unsafety” they perceive in a job. They don’t “demand” it by walking into an employer’s office and banging on the desk – they don’t have to. They just work only when and where wages rise sufficiently to compensate them for the risk they bear. This voluntary approach allows the amount of work supplied to equal the amount employers seek at the equilibrium market wage. This contrasts with the approach of labor unions, which creates involuntary unemployment by insisting on a bargained, above-market wage and/or working conditions that employers would not voluntarily provide.

The common sense of the wage premium can be expressed in figurative terms: “In our (workers’) opinion, this wage premium reflects the degree of danger – above the norm or average – that we associate with this job. You (the employer) are free to make any safety modifications in the job or working environment that will cost less than this amount (in the aggregate), but beware of spending more than this. Meanwhile, we have freely chosen to accept the currently-existing risks – as we perceive them.”

This provides a framework for efficient improvements in job safety. Without it, we are left with vague, grandiose rhetoric about how “nothing less than absolute safety is tolerable for our workers” or “how would you like your son or daughter to work in such an environment.” That kind of nebulous talk is complete rubbish. Every human being willingly takes risks every day of their lives, consciously or not. One of the most important parts of growing up is learning the risks of everyday life – which ones are necessary, which ones are reasonable and which ones are foolish. It makes no sense whatever to whine that people “shouldn’t have to risk their lives mining coal” so “big corporations can make profits.” Does it make sense to say that people can willingly risk their lives climbing mountains, fighting bulls, racing automobiles, jumping out of planes, fighting fires and so on – but they can’t risk their lives to keep people warm in the winter? Free markets allow the individuals directly concerned – workers, employers and consumers – to gauge the risks and calculate which improvements in safety are worth making and which aren’t. It recruits the people most willing and able to bear risk by offering them a premium for their efforts. It warns the timid by differentiating jobs according to risk – if all jobs paid the same a tedious and dangerous process of trial and error would be required to learn which jobs they should avoid.

Contrast this reasoned, rational approach with that of government regulatory agencies. They substitute their own technological, engineering view of safety for the free-market approach and impose it on the public in the form of command-and-control, one-size-fits-all, take-it-or-leave-the-country rules and regulations. One might reply that engineers have a more informed view of safety than do workers and employers. Yet research by leading experts like W. Kip Viscusi shows that market wage premia closely track technological and ex post statisticalmeasures of risk. And the government approach runs the risk of being skewed by politics; the regulatory agency’s objective studies may be overridden by a determination to please their bosses in the administration or Congressional patrons upon whom their funding depends.

The final word belongs to formal logic, which declares that there is no such thing as a pure engineering optimum in resource allocation. An engineer can determine (say) the optimum output from given inputs into a particular machine, but he or she can never determine the value to place on the inputs or output. Only producers and consumers can do that; that is why we need markets to solve economic problems. The formaldehyde case, noted above, shows the ghastly extremes to which engineers and bureaucracy can go when given free rein.

How is Job Safety Produced?

Our investigation reveals that job safety is produced primarily by free markets through wage premia and voluntary improvements in safety enacted by employers. It is produced much less efficiently, less productively and more haphazardly by government and even less proficiently by the actions of labor unions.