Even though I’m a fiscal policy economist, I often get hit with questions on other topics. This frequently happens when I’m overseas and I’m put in the position of defending every nuance of free market policy.
I don’t mind pontificating on other issues, but I get frustrated with myself for sometimes not having the specific knowledge to make the best possible case. I was recently asked, for instance, whether the private sector was capable of protecting the health and safety of workers.
I knew enough to discuss the overall cost of regulation, the amount of valuable time diverted to comply with red tape, and some of the research on regulation and job creation.
Moreover, I made the generic argument about how employers have a profit-maximizing incentive to protect the health and safety of their workforce.
But when the person asserted to me that the creation of the Occupational Safety and Health Administration (OSHA) was followed by safer workplaces, all I could do was mumble something about the sun doesn’t rise because roosters crow and that we would need hypotheticals and counterfactuals.
Fortunately, I work at the Cato Institute where there are experts about almost every policy issue, so I was able to track down some analysis about OSHA in the Cato Handbook for Congress.
This then led me to the National Safety Council, where I discovered that the person who was asking me about regulation was correct. As seen in this chart, workplace deaths have fallen significantly since OSHA was created.
There is a discontinuity in the data in 1992 because of a change in methodology, but I’ll stipulate that this doesn’t weaken in any way the argument that the creation of OSHA was followed by lower death rates in the workplace.
But it also turns out that my cop-out response about roosters and sunrises was right on the mark.
Let’s now look at the same chart, but this time we will include data going all the way back to 1933. What we find is that the workplace deaths were falling before OSHA and they continued to fall after OSHA.
Now for some caveats. This chart doesn’t prove OSHA is completely ineffective. Moreover, I”m sure there were state-based workplace regulations in effect in the pre-OSHA era, and I assume the federal government also had some health and safety regulations as well, perhaps through the Labor Department.
My argument is simply the more limited hypothesis that regulations impose considerable costs, which should be taken into account, and that businesses have a profit-maximizing incentive to promote health and safety in the workplace, which is increasingly important as society becomes richer.
So let me put the onus back on the pro-regulation crowd. Given the charts above, shouldn’t there be some sort of obligation to show that regulation has had a positive impact, particularly when costs are added to the equation.
And don’t give me a lazy argument about “even if we save just one life,” because I’ve already shown that a heavy regulatory burden can have a deadly impact.
P.S. Let’s also remember that OSHA generates some bone-headed regulatory choices, such as the crazy example recounted by Dave Barry and this nutty bit of regulatory excess uncovered by my colleague Walter Olson.
[…] That data is accurate, bu what they fail to mention is that workplace deaths were falling at exactly the same rate before OSHA. […]
[…] That data is accurate, bu what they fail to mention is that workplace deaths were falling at exactly the same rate before OSHA. […]
[…] That data is accurate, bu what they fail to mention is that workplace deaths were falling at exactly the same rate before OSHA. […]
[…] Source […]
[…] Intellectually Lazy, there’s a Very Strong Argument for Command-and-Control Regulation URL: https://danieljmitchell.wordpress.com/2012/09/25/assuming-youre-intellectually-lazy-theres-a-very-st… Accessed: […]
[…] in 2012, I shared a chart showing that workplace deaths declined substantially after the creation of the Occupational Safety […]
[…] evidence of societal benefit before government restricts economic freedom (the same argument I’ve used when analyzing […]
[…] some evidence of societal benefit before government restricts economic freedom (the same argument I’ve used when analyzing […]
[…] tape from government are supposedly correlated with good outcomes. One of my favorite examples is the data showing a decline in workplace deaths after the creation of the Occupational Safety and Health Administration. Fans of OSHA think this is […]
[…] better when government employees are lazy (after all, we probably don’t want hard-working OSHA inspectors, Fannie and Freddie regulators, and IRS […]
[…] better when government employees are lazy (after all, we probably don’t want hard-working OSHA inspectors, Fannie and Freddie regulators, and IRS […]
[…] bad thing that bureaucrats are lazy. Do we really want more diligent IRS agents? More hard-charging OSHA inspectors? Do we want Fannie and Freddie regulators burning the midnight oil concocting more affordable […]
…until they’re the ones subject to “command-and-control” tossing around expressed preferences for such nonsense over coffee or beer is cheap – like any idle talk. But let’s look at the actual behavior of those “computer science types” to see their revealed preference. Remember how the computer programming tradespeople reacted to even the barest hint of legislation that would subject them to licensing standards similar to registered professional engineers? And how would those computer science grads like it if OSHA regulated their work product? Or if they had to do all their programming in a so-called straitjacketed programming language – oh, perhaps Ada?
Yeah, those are only “sociological” measures but to me they’re pretty conclusive evidence proving that “computer science types” really don’t trust “command-and-control” over people.
Still I like the idea of some sort of mathematical demonstration of AI’s conjecture that “government is NP-hard”. Also a rigorous mathematical demonstration of the communication problem that is one of the major criticisms of socialism and all other forms of economic centralism in the State made by Austrian economists probably already exists. “Computer science types” should see the analogies between a command economy and forms of parallel and distributed processing.
” I’m equally skeptical of the competency of unregulated free markets which are at least as vulnerable to those same anarchists. Would you trust a military to the free market?” — Zinc
There are two significant problems with this statement. By definition free market actors *are* anarchists; at least within the economic sphere. The same is just as true for you when you head to the produce isle and choose to buy spinach instead of broccoli based on your whims.
The second issue, within the context of your post, is the idea that regulation that imposes civil fines is not an economic cost of business like any other. For example, theft is a normal course of running a brick-and-mortar storefront and a business simply folds the cost of that loss into its pricing structures. When speaking of regulatory issues, of any kind or worth, if the punishment is no more than a fine then the regulation is utterly meaningless. It becomes no more than a simple accounting cost and will be treated as exactly such.
While it is not proper to state that such regulations are ‘free market’ it is not proper to state that they are ‘not purely economic’ like every other cost in a ‘free market’. Such solutions carry all the flaws of regulation and all the flaws of free market solutions at one and the same time. So the answer to your question as to whether it can be fixed is rather trivial: All breeches of regulation need to be criminal in nature. As well this pierces the corporate veil and places the burden on each actor in the hierarchy that is just ‘following orders’.
Regulations dictated by government fiat are worse than bad. Central planning and regulation is just another arm of the dictatorship that is creeping upon us.
Individual humans have the right to decide the best tradeoffs for themselves in human and economic terms. If you need to consult statistics and spend time analyzing all the seen consequences of said regulations – so be it.
Don’t forget the unseen as Bastiat reminded us. Better yet – look at the last 5000 years of human history. Even better – investigate agorism or Ludwig von Mises – that’ll get you started. My conclusion? Human beings are much better off now than they were in almost all cases, in spite of the state – certainly not because of it, in any sphere.
Humans have a bright future, if we can rid ourselves of the state – if we don’t? All bets are off.
Forgive me if I am asking the obvious here:
– OSHA is Federal. So I am guessing that it regularized a patchwork of state regs? Might this be beneficial and increase efficiency? (Especially with suppliers.)
– BUT: won’t a Federal agency cause increased distortions b/c of BOTH Knowledge Problems and Agency Problems?
” I’m equally skeptical of the competency of unregulated free markets which are at least as vulnerable to those same anarchists. ” – Zinc
So all things equal you’d rather have costly regulations damaging our economy?
Sorry, the regulations have to do good, have to do provable good AND have to do enough good to justify their cost. Otherwise they’re a net loss in one form or another.
IMO if you’re not benefiting anyone significantly and you’re taking a lot of money out of the economy you’re doing harm, not good.
Sure we might have fewer people who die from certain workplace accidents, and that’s great. But when we also have fewer people who can afford to go to a doctor to catch a life threatening illness early because we’ve spent that moeny on OSHA to gain a marginal benefit we lose that bonus.
If the OSHA regulations result in a net benefit; great. Keep them. But the economic cost from them does result in a negative that has to be considered before determining if they’re net positive or negative in outcomes..
Without running the numbers, determining the costs, and specifying the exact benefits you don’t have a program that we know actually benefits us overall.
While I obviously haven’t studied every possible example, it seems to me that this tendency for regulations to be “closing the barn door after the horse has fled” is true in most cases, particularly for safety-related regulations. Seat belt use was rising before use was made compulsory. Same for motorcycle helmet use. Drunk driving was declining before penalties were stiffened. Regulations probably helped change attitudes about these things, but it is a very heavy-handed way to merely change attidudes.
Zinc: the real problem isn’t so much that “government is usually incompetent”, rather that government becomes increasingly incompetent over time. Check out Parkinson’s Law and the Peter Principle. You could also read Ibn Khaldun to find out that this problem has been with us for centuries.
WRT a free-market military, there is lots of literature on the subject, but that’s an entirely different story.
Today OSHA is more interested in maximizing fines assessed than maximizing worker safety. And that’s from a relative that worked for OSHA.
[…] If you’re intellectually lazy. […]
Could the drop in workplace deaths also be associated with the fact that a lot of heavy industry has moved off shore? Therefore workplaces are inherently less dangerous.
Another anecdote to add to the list of OSHA idiocies. Soon after the agency started, an inspector gigged a company because the office typewriter didn’t have a grounded plug. It didn’t have a grounded plug because it was a manual typewriter.
Government is NP-Hard revolving around how much it is attempting to ‘command-and-control’. Quantified perhaps by total ‘lines-of-code’ involved in the the written laws and regulations or total population.
It’s a pet theory, but everything seems to point that way for me.
Just as a for instance: The postal service is presented with -almost- the Traveling Salesman problem, with additional complicating caveats.
But note that exactly the same underlying problem is trivial if it is split into the individual tasks, planned and performed individually. “You, deliver (or cause to be delivered) this one letter for the lowest effort possible.” The discrepancy is large enough that FedEx, UPS, and a slew of other companies sprang up to do a slice of the job. For less. With a profit. (And if you add -every- messenger, courier, and letter-and-package delivery service, a -lot- of profit.)
I’d -really- like someone like Cato to work up a formal proof though – computer science types really seem enamored of ‘command-and-control’, a little solid evidence that arises more from mathematical proof than sociological studies could be quite significant.
I suggest that you plot this data on a semi-log graph. A process that is producing change at a constant rate will be a straight line; one that is producing a changing rate will show a curvature. By my eye from the plot shown, the rate of decline is getting smaller with time (this would show on a semi-log plot as a significant ‘upward’ curvature), implying that the benefits of future regulation may be much less than one is likely to be led to believe by the proposers of the regulation. (This isn’t surprising – in a ‘perfect’ world the curve has to asymptote to zero.)
Child labor existed throughout history until the 20th Century. What reduced child mortality and finally ended child labor was not legislation, but the improved standard of living brought by near-laissez-faire capitalism.
Last paragraph: http://eh.net/encyclopedia/article/whaples.childlabor
Related to workplace safety laws: the historical statistics I have seen show no effect of child labor legislation on child mortality:
a) Among boys, ages 10-14 in Sweden, the mortality rate was 35/1000 in 1775, and 15/1000 in 1880. Child labor legislation was passed in 1881, and the curve of the graph was not affected. There was no sudden drop in child mortality, but continued smoothly downward.
b) In England and Wales, child mortality, ages 10-14, both sexes, was 9/1000 in 1840 and had decreased to 3/1000 by 1910.
c) Belgium’s child mortality rate (10-14) was decreasing steadily for 30 years before any child labor legislation was passed in 1886, and did not decrease more rapidly afterwards.
These statistics are based on graphs that Stuart Hayashi of Hawaii has in his Facebook pictures. They are for his upcoming book, currently titled: The Nature of Liberty: The Biological Case for Individual Rights.
The sources are:
a) R. Steckel, L. Sandberg: “Was Industrialization Hazardous to Your Health? Not in Sweden!”, Health and Welfare During Industrialization
b) England and Wales: Total Population – Death Rates by Year of Birth, Human Mortality Database
c) Belgium – Death Rates, Human Mortality Database
Random reward and the chickens will believe that whatever they are doing is repsonsible for the reward. Works with lib-progs too.
The continuity of the chart makes sense to me. At the time that OSHA came into effect, it essentially institutionalized the safety rules of the industry I was in (Electric utilities). The conventional wisdom was that by the time you had a proficient lineman trained, the companies investment was so high that it was far more economic to protect him than replace him. Economic self-interest took care of the safety and health issues.
I have no data however for low skill or sweated industries which could be quite different.
While I could certainly believe that a regulatory agency could easily make things worse instead of better, I’m not inclined to presume that no regulations are always better. While you would seem to be open minded regarding regulations backed by evidence, I do detect a general disdain for them in your tone. Even if OSHA were shown here to be ineffective, the question then is, can it be fixed? Unfortunately, a widespread belief that government is usually incompetent becomes a self-fulfilling prophecy, successful government is a team effort, and when many team members are ‘just tear it down’ anarchists, government will likely remain incompetent. I’m equally skeptical of the competency of unregulated free markets which are at least as vulnerable to those same anarchists. Would you trust a military to the free market? Why should it be an exception, if it should?
Dan,
I believe the OSHA thing has a near-exact parallel with the Clean Air Act, i.e. air quality was improving long before the law, but supporters just start the chart when the law was passed.
RGK