BS Jobs and BS Economics

David Graeber’s peculiar article on bullshit jobs (noted earlier by Tyler) does have one redeeming feature, a great example of poor economic reasoning:

…in our society, there seems a general rule that, the more obviously one’s work benefits other people, the less one is likely to be paid for it.  Again, an objective measure is hard to find, but one easy way to get a sense is to ask: what would happen were this entire class of people to simply disappear? Say what you like about nurses, garbage collectors, or mechanics, it’s obvious that were they to vanish in a puff of smoke, the results would be immediate and catastrophic. A world without teachers or dock-workers would soon be in trouble…

This, of course, is just the diamond-water “paradox”–why are diamonds, mere baubles, so expensive while water, a necessity of life, is so cheap?–the paradox was solved over a hundred years ago by…wait for it…can you guess?….the marginal revolution. Water is cheap and its value low because the supply of water is so large that the marginal value of water is driven down close to zero. Diamonds are expensive because the limited market supply keeps the price and marginal value high. Not much of a paradox. Note that, contra Graeber, there is nothing special about labor in this regard or “our society.”

Moreover, it’s good that prices are determined on the margin. We would be very much the poorer, if all useful goods were expensive and only useless goods were cheap.


To put it another way, if there were some catastrophic event and Miguel Cabrera and LeBron James had to haul garbage, they could and at the same level as current sanitation workers. The opposite is not true.

Massive research shows that actively managed hedge funds and mutual funds underperform index funds and ETFs over time. Yet we pay analysts, PMs and traders millions of dollars to actively select portfolios that will likely fail over time. The analysts, PMs, etc are smart and hard working but research shows that over time they will destroy shareholder value. BS jobs are real. Talk to a big 4 auditor and they will spend hours complaining about the joke we call audit. Meanwhile, EMTs and social workers can't pay their bills.

The question is, If people were pressed into service, could they perform these tasks with minimal training? In the case of EMTs, the answer is yes. My proof is that we did so. Prior to 1975, the military produced tens of thousands of medics.

Also, your comparison isn't fair. We have metrics to judge the average fund, but how do we judge the average social worker? Talk to psychiatrists who work in family court and they will tell you what a joke the average social worker visit is.

But, of course, index funds as we know them wouldn't exist without actively-managed funds. If everyone chose to track the market, nothing would trade.

If I may pursue your point...

If everyone chose to track the market, the opportunities associated with active management would be astronomical.

Excellent point. If everyone bought index funds the stock and market would lose its mind.

The paradox is that index funds don't work without active traders. The statistical oddity is that if you added up all of the active trades, you would arrive at...the market, meaning active traders should perform exactly as the index funds in the aggregate.


"The statistical oddity is that if you added up all of the active trades, you would arrive at…the market, meaning active traders should perform exactly as the index funds in the aggregate."

Yes, but due to trading commissions and the substantial fees of running a fund (rent, technology, research, compliance, marketing) even if all the active traders average out to the market the active traders actually underperform the market because of fees and expenses. The longer the horizon the harder it is for the actively managed funds to beat their index.

I understand an analyst or PM can become a bank teller but the inverse doesn't hold true. However, if the bank teller can outperform the average analyst or PM via indexing why should the average analyst or PM earn 100, 200, 300 times the salary of a bank teller? Financial services is filled with these anomalies - I know a programmer who became a high frequency trader. His algorithms earned him great money when they worked, but when they failed he lost his job. Overall, HE LOST MORE MONEY FOR HIS FIRM THAN HE MADE, but since the payoff structure was that success = big bonuses while failure = job loss, he ended up very wealthy while destroying value for his firm!!! And what are the opportunity costs of us herding the best and brightest minds into high frequency trading, financial services and Internet advertising?

I don't know what a social work visit is but I know social workers. Yes, they strike out 75% of the time, but when they have success, the success is oftentimes lifechanging. I know that doesn't mean much to many reading this blog but I think a social worker is a lot more valuable than the 9000th hedge fund or the 500,000th tax accountant. I understand it's Wall Street that pays the taxes that support government services and non profits but on a micro level the typical financial services professional isn't really useful.

"I don’t know what a social work visit is but I know social workers" Really?

This misses the actual utility of having an equities market in the first place. The *individual* motivation for trading is to attempt to beat the aggregate performance of the market, but as has been noted that's a logical impossibility when taken across the entire market; not everybody can outperform, unless the servers and trading floors are moved to Lake Woebegone. In fact, trading itself must be zero-sum from a performance-relative-to-an-index standpoint. Every buyer at price X faces a seller at X, putting aside spreads and other transaction costs. The general utilities involve providing liquidity for entrepreneurs, information discovery and the facilitation of deploying capital. The potential rewards must be outsized because, as was also noted, if they were pretty mundane why would anybody do anything other than purchase an index? And if there are no differences of opinion, there is very little reason to engage in the costly task of information discovery and analysis. (And then the deployment of capital would be, hard though it might be to imagine after 2001 or 2008, significantly crappier.)

As to your point on relative value, you again miss the real issue. You "think a social worker is a lot more valuable than the 9000th hedge fund or the 500,000th tax accountant." Swell. I, OTOH, am trying to setup a business, and so I have very little interest in social workers and a great deal of interest in whether I can find the 500,001st tax accountant or have to bid up the price of the existing 500K. There is no "we", there are your interests, my interests and some variable level of overlap.

It's important to note that they only underperform net of fees. Cochrane covers some related topics here:

Looks great - I will defnitely check it out - thanks

A small number of active fund managers have actually beaten the market over time; some by a very significant margin. For example, Peter Lynch and the Fidelity Magellan fund and Bruce Berkowitz and the Fairholme fund. So perhaps there's a risk premium associated with the cost of paying a fund manager. If you choose a good one, you can make more (and so can they).

"A small number of active fund managers have actually beaten the market over time; some by a very significant margin."

Yes, but it may be luck (or steady insider trading ala SAC Capital). If there are 10,000 actively managed funds you would expect a handful to outperform over time - just by chance. The hard part is identifying that handful and finding them early on. By the time a pattern of success emerges it is often too late. For example, I believe I read research where they determined that the top funds over various time periods are just as likely to be in the bottom half of performance over the next time period as they are to be in the top half.

You would expect them to outperform during certain years "just by chance", certainly. Outperforming the market consistently, and making 15% to 20% more than market return over 10 to 20 years is more than simple chance. If it's insider trading, and they can get away with it for over a decade or two and still never get caught, nor investigated, then that's a skill worth paying for.

We is too many people. You may pay fund managers but I do not. Now is sometimes seems that we are taxed to pay for politicians to scam us.

Athletes indeed have real abilities that would be difficult for most others to replicate. What are the abilities that CEOs have that are the equivalent of homering or dunking? Things that other individuals are literally unable to do. To continue the sports analogy the CEO should be compared to the coach of the team not the players correct? The coach cannot "do" anything other than direct the players that do the actual work. And certainly direction is useful but the best coach is not going to beat Lebron with scrubs. In addition what are the things that a Fortune 500 CEO does that the company is not already paying other people to do as well (it would seem the smaller the company the more important the CEO not the reverse)? It appears very hard to measure CEO performance and separate CEO performance from sector performance. Let's just say I think it is a lot easier to evaluate how much Lebron should be paid compared to how much a CEO should be paid.

In football coaches seem more important than all but a few great players. Basketball not so much.

This reminds me of the BBC's The Public Philosopher episode from April 2012, "Should a Banker be paid more than a Nurse?

We can give the poor their own Dream Job, and it still increases consumption and reduces price levels.

It takes years of education to become that ignorant.

Totally disagree. This strikes me as one of those "superficially plausible until you reason through it" things, like deficit spending.

Graeber is overrated and became too self-confident after his book Debt, the first 5000 years which I enjoyed (the first 100 pages). Writing like Montaigne requires much more reading than Graeber has done.

There is also an aspect of leverage here. Most of our high paying jobs are highly leveraged. A good manager increases the productivity of their entire team. A software developer writes software that is used again and again and against by many many many people. The people doing obvious work are often very poorly leveraged.

A teacher has a lot of leverage - thousands of students over a career. I manage 6 people - and none of them even need me. Mgmt is usually a BS job.

I dont know about that. Our team has been operating without a manager for about a month and productivity has dropped significantly. Maybe your team doesnt need you, but my guess is things would start to fall apart if they just didnt have a manager. Or maybe you just suck.

A good manager is all about greasing the politics of your organization. You select and focus the output of the team.

There are select hierarchical environments where you can work more or less by consensus - usually a function of organization size - but even then leaders emerge. I'd say that if you feel redundant as a manager there's a good chance you operate in a relatively frinctionless organization where your reports' interest are aligned.

A teacher has thousands of students over a career, but has little or no effect on most of them (says this former teacher).

Horrible analogy. Students are customers, not employees.

A bad manager decreases productivity of the whole team - and gets paid hefty sums anyway. In addition - the way he ruined the team will be a "valuable experience" elevating his market value.

Snarky this morning aren't we? I manage 6 people because I work in a large organization that requires a management structure. However, the structure is flawed - although I technically manage 6 people, they each work on projects with people outside my chain of command. Meanwhile, I spend 95% of my time on my own projects (i.e. not performing management work) and those projects are independent of the people I supervise. The people I supervise are strong - which is why I don't spend much time managing them. Shouldn't managers know when to give smart people the ball and let them run?

As I said - the structure is flawed but that is out of my hands. My point was I get extra pay because I'm a "manager" but I don't really do much managing because it isn't required. Fortunately I have my own projects but I'm afraid in many organizations the managers act as roadblocks and don't do much actual work.

Reminds me of a manager I once had. As we had logs of what we achieved, it was pretty trivial to calculate that whilst managing a 2 man team, that manager managed to contribute just over -1 man day of productivity when in the office. Days when all three of us were in were less productive as days when one of us (and not the manager) were in the office.

But the price of diamonds, like the salaries of CEOs, are kept high, not because of the scarcity of the resource, but through manipulation of the market. Just as the price of water could one day rise much higher, not so much because of a dwindling resource (though the resource is dwindling), but because of manipulation (privatization) of the market. The salaries of teachers are not kept low due to a surplus of teachers, but because we, as a society, fail to value them at their true level, and they, historically, have not had the political power to demand more. God knows what we would be paying teachers today if they hadn't had unions - probably barely above minimum wage, and there are plenty of people who would love to see them reduced to that.

Of course the supply of teachers is kept low. Any job that requires certification has a control on supply.

On the other hand, the supply is kept artificially high. Why else would high school students be required a year of gym and a semester of health.

Did you mean the _demand_ is kept artificially high?

Both. The supply is kept low because of certification and the demand is kept high by graduation requirements.

There are also a big supply of dreamers who want to be teachers.

If your work is "obvious" it's probably because lots of people can do it. Someone who makes a thousand other people 0.5% more efficient is probably doing something incredibly un-obvious and most people could not do it.

Sounds hopelessly conspiratorial to me. "All the wrong people get paid too much and all the right people get paid too little because of some insidious manipulation"?

Maybe not all CEOs, but the most famous ones got paid a lot because they are worth a lot. Bill Gates and Steve Jobs generated a simply staggering amount of wealth for their respective companies. HP would pay a gigantic sum of money for someone who could do that for them.

If Teachers arent valued at their 'true level' its because there is no competitive market for their skills. You dont get paid your 'true value' via political pressure, you get it by competing in a free market with others.

To rephrase your argument:
"Maybe not all CEOs, but the most famous ones got paid a lot because they are worth a lot. Stephen Elop and Steve Ballmer generated a simply staggering amount of wealth for their respective companies."

Re: Ballmer:

The stock went up 5% when he announced his resignation. And "Microsoft's market capitalization fell about $330 million during his tenure as CEO."

So an unknown, future CEO is worth more than Ballmer. And the stock fell while he ran the company. Yet he made many millions. This is what "average" people (and myself) don't understand. Maybe one can argue that Microsoft would have been down more than $330mil under a non-Ballmer CEO?

Or he did the best thing he could for his shareholders by retiring. They made billions!

Remember, you cant guarantee ahead of time who is going to be the Steve Jobs and who is going to be every CEO of HP for the last 20 years.

As for Ballmer, Microsoft shareholders will have to live 215% increase Net income an the nearly triple annual revenue gained under Ballmer's reign.

I think we can assume this guy is a teacher. That said, there are few pure markets free of asymmetry and/or external manipulation. Government is often there putting its thumb on the scale. Everywhere market participants have some advantage they use, which results in price anomalies. The fact that the cost of basic education of children has gone up over time indicates just such a manipulation. In this case it is a form of self-dealing by politicians and unions.

You know what happens when you assume. Sorry. I'm not a teacher. Just a guy who values what teachers do, even the PE and Health teachers.

My health teacher taught us that there is such a thing as "tetesterone", and that drugs are bad, mmkay.

I think we can assume this guy has never Googled "ad hominem."

In American high schools, most of what is supposed to be learned is instead "memorized and forgotten" soon afterward. To a large extent, high school is high class daycare. That sounds harsh but ask yourself, "How much do you remember of what you supposedly learned in high school? How many high school final exams could you actually pass today?" And realize that as a reader of this blog, you are probably in the top few percent of people who took school seriously.

In terms of social usefulness, 50K a year is overpaying for most high school teachers.

"The salaries of teachers are not kept low due to a surplus of teachers, but because we, as a society, fail to value them at their true level"

There is no such thing as a "true value". Their value is relative to the needs and desires of all the people observing them, and each one of those is a different value influenced by differences in how the observer values their characteristics and by individual differences between teachers. You are seduced by the simpler picture of homogeneous widgets sold to a very similar set of consumers (though even in that case there are obvious differences in how the consumers value things, hence the title of the blog.)

Now, you might argue that the mass of taxpayers are individually undervaluing teachers in light of their own true interests (a mass of miscalculation, as it were), but then I think you'd have to make a case that you know their business better than they do, and do so in light of the fact that in practice education is both rivalrous (see class size as an issue) and excludable (see the bidding up of property values in neighborhoods with good schools). What I think you'll ultimately find is that taxpayers place a high value the schools, and the teachers that staff said schools, where *their own children* go. Everybody else's, not quite so much.

I.E. selfish a-holes? Yes. Precisely. A selfish a-hole is unable to value the contributions of a teacher and therefore should not be put in charge of determining the teacher's salary.

I agree with this part: "The salaries of teachers are not kept low"

It's worth noting that setting wages equal to marginal productivity optimizes total productivity by encouraging each person to perform the job where he can make the greatest marginal contribution. It's a greedy algorithm in both senses of the word.

It's also not even theoretically possible in most cases to pay wages equal to a person's total (non-marginal) productivity. Say a production process requires inputs from ten different people doing radically different types of work with different qualifications. How do you decide how much of the overall productivity to attribute to each worker (or, for that matter, to capital)? Without taking marginal productivity into account, there's no right way to divide it. And even if there were, the overall productivity is a function of the price of the good being produced, which is determined by its marginal value.

"An abstract for this item is not available."

Also "the more obviously one’s work benefits other people" is not a good measure of the marginal value of such work. Markets are really good at turning very diffused efforts into very concentrated benefits. If someone invented a process that shaved off .0001% of the cost of refining a gallon of gas, the market would reward him with wealth beyond anyone's dreams. Benefits like that are very real but not very obvious (until you ask the guy where he got the money for his yacht). Markets help us capture such benefits.

> If someone invented a process that shaved off .0001% of the cost of refining a gallon of gas, the market would reward him with wealth beyond anyone’s dreams.

That seems unlikely.

Likeliest scenario is for such a person to be the employee of a refinery company. In that case, they will get the standard $70k-$130k per year that such a job would entail and all the marginal benefit would go to their employer.

Somewhat likely would be an employee of a university. Their invention would either be public domain or commercialized by their educational institution, not them.

Least likely would be an independent researcher with commercial acumen. They would have to spend $2k on a patent plus $18k on a patent lawyer, and then shop their patent around the relevant oil companies. Even then, it's likelier that oil companies will just wait for the patent to expire rather than pay anything.

Entirely impossible would be an independent inventor founding their own refinery. That would require an absurd amount of capital unlikely in a petroleum refining specialist.

But missed in even Alex's rebuke is the additional insight that there is a fundamental, even fundamentally just, reason why simple supply differences should determine worth -- they tend to select/filter for personality and cognive traints that any society neing built from the ground up would want to have. How many of those holding the higher paying, less supplied, jobs could not, in a pinch, be trained to fill the gap left by the vanishing worthies? I would wager anything that, with training, a pool of engineers, financiers, surgeons, professional managers, etc. would produce more adequate teachers, garbage collectors, and nurses, than the other way around.

Two thoughts:

I don't think that Graeber is making an economic argument. While this might be hard for an economists to understand, it is possible to make proscriptive arguments about how the world ought to be, without reliance on economic thinking. Thus, while Graeber may or may not understand the concept of marginal utility, it is not at all clear that he needs the concept in order to successfully make his argument. This should be obvious to anyone who view the labor market in context. it is clear that the policy framework and the marginal value of a job given that framework play a role in remuneration for work. Just one example should make this clear, lawyers have successfully created a legal monopoly that ensures that they are paid more for their services. This makes pay for the average lawyer greater than it should be and it decreases consumer surplus. It is also a policy that contributes to the problem Mr. Graeber is discussing. Doctors do the same (with greater success) by limiting the number of medical schools and by preventing nurse practitioners from delivering unassisted medical services. Different policies lead to different labor markets, that is why, for example, France and the United State pay workers differently for different work.


I think that average firm size also contributes to the distribution of jobs towards professions that deal with risk. As a rank amateur I can't show this relationship mathematically or with resort to technical lingo but my own observations suggest that larger firms devote a larger percent of their resources to accounting for legal and financial risk. Thus as average firm size increases the number of dollars chasing legal and financial risk also increases. In both professions (but law especially) utility is measured by covering your ass. In the same way that ultra wealthy individuals devote larger sums of money to dealing with tail risks (they consume more insurance than individuals) large and wealthy firms devote more money to what graeber somewhat crassly refers to as paper pushing.

Alex, maybe you can help me sort out a condundrum that's plagued me for years:

1. The diamond-water paradox suggests that there is no relationship between utility and price.

2. The doctrine of revealed preferences (and really, price theory in general) is based on the idea that prices are the *only* practicable *measure* of utility.

These seem irreconcilably contradictory.

Can you help?

I'm not Alex, but I'll help.

When you write "utility" and note the paradox, you are conflating absolute utility with marginal utility. If I rewrite your conundrum appropriately, it'll cease to be a conundrum.

1. The diamond-water paradox suggests that there is no relationship between *absolute* utility and price.
There is however, a strong relationship between *marginal* utility and price.

2. The doctrine of revealed preferences is based on the idea that prices are the only practicable measure of *marginal* utility.
Prices do not, however, reveal absolute utility.

If that does not help, then I suggest David Friedman...

Well you did such a good job explaining that, perhaps you can explain why the fundamental advances in civilization, such as the kind that won Nobel Prizes and otherwise helped many people, were done for free or next to nothing. Is the fame associated with inventing something useful--at the margin--really so powerful? If so, then I posit that we need to strengthen IP laws to give further incentive to those people not motivated by mere fame, to invent something useful. By contrast, if you argue that "people invent regardless of incentives" (fixed supply, vertical line), then marginal utility does not explain anything.

I call this the "Inventor Paradox"--fundamental advances in society, of the kind that shifts curves outwards (not marginally but wholesale shifts), are done essentially by chance, by people motivated by non-monetary compensation. If you can figure out whether more people would invent if motivated by money you would win some sort of a prize--maybe even the Nobel Prize--if that's what motivates you.

Thanks Jody. Very well explained.

And here I find the crux. There is no way to measure a good's absolute utility in terms of money (IOW, in terms of all other goods).

And the mistaken, illogical conclusion that has been arrived at from that: because absolute utility cannot be measured, there is no such thing as absolute utility.

Which is why Alex is able to make the statement he does: "there is nothing special about labor in this regard".

The very same logic applies to any other "good": there is nothing special about life in this regard, or years of life, or a happy, healthy, and productive life.

Alex Tabarrok seems blissfully unconcerned that his "innovation renaissance" is needed in precisely the same post-1973 economy of low productivity growth which has seen CEO compensation reach escape velocity.

Alternative theory of bullshit jobs: Technology and economies of scale mean that only a few people can produce all the real value in the economy. Transferring value then becomes easier than creating real value at the margin. This is what the bullshit jobs are. Lawyers, marketing, sales, finance workers, government workers produce less value than they get paid (myself included). They transfer value from competitors/consumers/targets to their bosses or even themselves.

We're basically economic mercenaries. We call our jobs bullshit because it doesn't matter much if we do our jobs or not. Sure, it matters to our employer because it helps them best their competitor / target. But it doesn't make the world any better off and we know it.

This is excellent--thanks for writing this, and but for the government shutdown arguably you'd not have the opportunity to write this, though gov't workers are in fact a huge surfing presence online during working hours.

Now consider this: b.s. jobs means that if you did not do your job, somebody else would, since it's so easy to do (take money from the producers). But also I would say that in science there are 'b.s. jobs' or 'b.s. research' or 'b.s. inventions' in that if you did not do the research, or invent the product, it would not matter since somebody else would inevitably discover or invent the same thing, since the conditions for doing so are "in the air". Sorry if you don't get what I mean, but I think most people in tech fields know what I'm talking about (it's related to TC's "low hanging fruit" as well--the easy advances in science are done first).

Most people want to do things that help people. They often do it for free, its called volunteering. Some even pay to do real work, such as when you pay to pick fruit at a farm.

Jobs that are useless or hurt people aren't things people want to do. However, market failures often create opportunities for such jobs to exist. Since most people would rather do a job that isn't evil you have to pay them more to do a job they know is evil or useless. Kind of like "moral hazard pay".

Now *that's* an interesting theory.

The value of the low wage worker as consumer is under valued.

This discussion assumes that an economy would function just fine is it was one CEO/capitalist/worker making $1,000,000 and a million workers making $1 for a total GDP of $2,000,000.

Except, individual consumption would have a mean of $1 and the best the CEO could consume might be $100 or even maybe $1000, but it defies logic for the CEO to consume a million times more than the mean.

What could he possibly consume? Maybe robots? So each year he devotes $999,990 to consuming robots, well except that will increase production say 10%, so he will consume $999,990 the first year and $9,999,990 in robots the eighth year.

The trouble for the past decade is basically just that. At least while building all the shopping malls and all the houses that so many could not afford. Now the high earners can't consume what they earn so they send cash chasing stocks and bonds because they know not to loan to low wage workers. And governments are borrowing and giving that to the low wage workers so they can consume what is being produced.

At least they will do that for T-bills as long as the first and second enumerated powers of Congress are honored and exercised respecting the wealthy who can't think of a better thing to do with their money than T-bills.

What happens in 15 days if T-bills seem as valuable as a $100,000 mortgage on a condemned Detroit house or a Detroit city bond with a governor determined to pay 5 cents on the dollar based on the idea that workers are worthless.

A really bad answer. That marginal utility explains everything in economic life is one of the dumbest economic statements of all time. Go back to study economics properly.

Tabarrok's response glibly explains nothing: a statement that everything is determined marginally does nothing to explain the determinants of supplies and demand, producing some particular marginal price.

The real answer would involve institutional economics, which is much more work than denouncing an ideological opponent with a red herring. When you examine the structure of the markets and the cultural assumptions they are embedded in, abstract statements like Tabarrok's become obviously vapid.

Take diamonds, for example. The DeBeers monopoly has controlled the supply for decades by stockpiling excess production. The DeBeers monopoly has promoted diamonds as a positional good for the middle and lower classes for decades, creating wasteful competition. If diamonds were valued primarily for their industrial uses, their prices could be much lower.

Wages for various jobs vary widely between times depending on unionization, corporation and employment laws. Wages vary widely based on cultural assumptions of "women's work", lower class job classifications, etc. Wages vary widely depending on bargaining power, self-serving, etc. Wages vary widely based on whether there is a winner-take-all structure to the market. Etc.

To put it another way: imagine that labor was as powerful as the DeBeers cartel. What would the value of labor then be?

Wow, in that case your response is an even poorer form of economic reasoning.

Comments for this post are closed