David Graeber wrote a piece On the Phenomenon of Bullshit Jobs which niggled at me for a long time. On one hand, anyone who has worked has seen people doing work that didn’t really need to exist. On the other hand, the way he catalogued bullshit jobs rang false to me. He singles out administration, financial services, academia, and public relations, among others, as entirely bullshit fields. But this seems cartoonish – of course there are scammers in every line of work, but a civilization that hanged all the administrators would quickly rediscover the importance of coordination.
I think a better way of approaching the problem is to start with this axiom: a bullshit job is a job whose demand is artificially created. And in practice, there is a sliding scale of bullshit. On the pure-bullshit side, you have “rubber room” jobs, where incompetent employees are placed with nothing to do except stare at the walls, in the hope that they will quit in shame and in the meantime encourage the others. Another example would be the sort of jobs you’d imagine giving the boss’s idiot son. In both cases, the work creates no value to the company, and you couldn’t imagine a company waking up one morning and deciding to hire someone off the street to do this work. The demand is entirely driven by the political considerations within the company.
For the next level, imagine the economics textbook example of two women who start off as stay at home moms, taking care of their children, and decide instead to babysit for one another instead. Suddenly they become employees, and GDP goes up, even though they’re doing exactly the same work as before. Unlike the rubber room workers, these women are doing real work. If one of them were to jet off to Tahiti, diapers would start to pile up and people would notice. Still, their “job” is in a real sense bullshit – they are no more productive than they were before.
The next level is even more subtle. Let’s take the example of the Joint Strike Fighter, the bloated project aiming to produce the next generation of fighter planes. With an estimated program cost of $1 trillion and an estimated unit cost almost 5 times that of the F-15, the program has been plagued by absurd cost overruns. Now, if we look at the engineers working on the project, they seem to be doing serious, non-bullshit work. These are smart guys working on real technical problems, and at the end of the day fighter planes are in fact rolling out of the factories. And yet, if we had a government that was competent at planning a project like this, four out of five of them would be out of a job. Contra Graeber, bullshit jobs aren’t a matter of how technical or sexy a job appears. Despite being STEM-based, working on cool projects, and ultimately producing results, these engineers’ jobs only exist because of demand driven by political incompetence – they are to a large extent bullshit.
And this leads us to a whole swath of government jobs, which aren’t characterized by hilarious cost overruns but are still largely created by artificial demand. Tyler Cowen in The Great Stagnation points out the real difficulty in figuring out how to value the activities of government. $100 worth of work by a Health and Human Services bureaucrat, while perhaps not devoid of value, seems less certain to yield objective value than $100 of jeans. But if we’re going to be skeptical that bureaucrat, we need to turn skepticism much farther afield. Many of the people who read this essay have jobs that are paid for by the government. Academics who are dependent on government grants, doctors who are dependent on payments from Medicare and ACA-mandated private insurance, civil and environmental engineers whose companies live and die by government contracts – the list goes on.
The tricky thing is that these jobs are only partially bullshit. Doctors, for example, do a lot of valuable work, and even in your preferred utopia, doctors would still be needed. But the way the medicine is practiced is inevitably distorted by its status as an outsourced government job. Medicine has to be done in a way that is legible to a Medicare bureaucrat. Incredible amounts of effort are spent on licensing and credentialing. And hospital billing departments wage endless trench warfare with private insurance companies, whose own bureaucrats are fighting to find reasons to deny payment to hospitals – a perverse set of incentives that exists only because of the way the government has set up healthcare.
There’s a real sense in which every one of these jobs is a vital one, at least in the Rube Goldberg way things are set up now. A hospital that fired half its billing department would soon cease to remain solvent. A doctor who practiced medicine without regard to Medicare regulations would soon find himself without a license. And somehow, in between all this nonsense, doctors and nurses find time to actually take care of patients. But it’s hard to look at all of this and say that this is really the best way we could be doing things. In a well-managed healthcare system, things would be a lot more streamlined – and lots of those billing department cubicles would be vacant.
Now, putting on one’s libertarian hat, the obvious counterargument to the notion of bullshit jobs is “if these jobs are so useless, why not start a company without them, and corner the market?” The fact is, many institutions that provide both the jobs are in fact quite efficient – efficient at meeting the artificial demand of government contracts. The engineers at Lockheed Martin are, to all appearances, working hard at the problems assigned to them. The problem is not that they are slacking off, the problem is that their assignments are asinine. And there’s no way to profit from that dysfunction, short of trying to start a competing government.
Elon Musk is at once a challenge to and an exemplar of this phenomenon. By forming SpaceX, he demonstrated that there was tremendous slack in the way that Boeing and Lockheed Martin had been fulfilling their launch contracts for NASA. Because of his superior ability to organize a launch company, he has been able to outcompete the existing contractors, stealing a sizable chunk of their contracts. Nevertheless, it’s worth noting that SpaceX is still in the business of landing government contracts – chasing the artificial demand driven by NASA. (Private satellite launch contracts are a growing but still small portion of their revenues.) And even if you just want to fight existing contractors head to head, it seems like the table stakes are on the order of “founding two successful tech companies and being Elon Musk.”
This picture is more complex than the one painted by Graeber. There aren’t, for the most part, a cabal of obvious villains or laughably pointless jobs. With the exception of the rubber room, most of these jobs have a logic of their own – there is good reason for each of them to exist in the complex ecology that is the modern economy. There is very little you, as a private citizen, can do, either to tear down bullshit jobs or outcompete them as an entrepreneur. But this picture does leave us with some takeaways.
For one, our economy is driven by artificial demand to a much greater extent than most people realize. The government itself only employs a limited number of people. But huge swathes of the economy, including the modern growth industries of education and healthcare, only exist in their present form because of government spending and government mandates. Western countries may have an aesthetic preference for channeling this artificial demand through nominally private entities, such as aerospace companies, hospitals, and universities. But in a real sense these engineers, doctors, and professors have government jobs, the same sort of government job as your stereotypical paper-pushing bureaucrat. They may still do valuable work, just as some bureaucrats do, but much of their work has a large admixture of bullshit – and in a sane world wouldn’t exist.
A corollary: since the demand here is artificial, the jobs themselves are political artifacts. We’re used of thinking of jobs as roles that emerge as people figure out how to fulfil existing demands. But to a certain frame of mind, “jobs” are commodities to be doled out. We understand this perspective when politicians scheme to place army bases and VA hospitals in key congressional districts, but it extends further than that. Debates over issues as far-flung as the appropriate level of education spending, the role of nurse practitioners, and the value of massage therapy are thinly laundered debates about how much artificial demand to create for bullshit jobs.
A final lesson here is that qualitative economics has an important role in helping us understand how the world works. The problems with GDP as a metric of well-being are well known. But even a statistician who is willing to look at, say, “GDP minus government” will miss many of the ripple effects that these distortions have on the economy. He might note the dollar value of Medicare payments, but include billing managers as purely private-sector workers, not noticing that the demand for their work is in large part driven by government regulations. Quantitative economics can get us part of the way towards understanding the world, but if we want to truly understand what sorts of work are socially valuable, we have to be comfortable thinking qualitatively about how individual jobs work, why they exist, and what social value they create.