Review: Bullshit Jobs

The rise of pointless work and what we can do about it

David Graeber (2018)

Bullshit Jobs explores the phenomenon of pointless work, seeking to provide a basic taxonomy of its kinds, document its effects, suggest potential causes, and hint at some potential solutions. The most important function performed by the book is to provide a useful label for the phenomenon of pointless work: bullshit. Actually defining what it is, however, takes some doing.

Most people will be familiar either with anecdotes or experience jobs where the main requirement seems to be to look busy rather than accomplish any particular task, and most people will have at least some experience of having to fill out forms or prepare for meetings that seem to vanish into the ether never to be heard from again. If I’m being paid to surf the web, what difference would it make if I didn’t show up? If no one reads these stupid forms, what difference would it make if I filed them in the recycling bin? These questions get to the heart of the bullshit work phenomenon, and characterise its two key facets: bullshit work and bullshit jobs. The forms are bullshit work - parts of a worthwhile job which do not improve the performance of the job or the outcomes for those it serves (a favourite example is overloading teachers and nurses with paperwork); while the websurfing employees are doing bullshit jobs - jobs where there is apparently no end product or service to which they are meaningfully contributing.

Graeber doesn’t put up much of a fight against nitpicking in the definition:1 the taxonomy is mostly taken from a synthesis of a couple of hundred testimonials and is more thematic than concrete. Nevertheless, he does defend an important distinction between bullshit jobs, which are pointless or make life worse for everyone,1 and shit jobs, which produce extremely valuable outcomes but are difficult, unpleasant, and/or poorly remunerated.2 In most cases Graeber is happy to defer to an individual in their assessment of whether their job is bullshit, in part because it is the feeling of working a bullshit job which interests him.

Bullshit jobs rot the soul. Graeber relates a fusillade of accounts from people working bullshit jobs which portray the frustration, anxiety, and misery they engender. There are key parallels between the symptoms of bullshit jobs and those of existential angst: both are caused by what should be desirable situations, and both are a response to apparent meaninglessness. Sinecures, after all, are easy - browsing the web in an office building for eight hours a day is substantially more desirable than scrubbing the same office building for eight hours a night for a quarter of the pay. It is quite possible that Graeber’s correspondents, as a self-selecting sample, represent a small minority of people who can’t be satisfied with such arrangements, but perhaps there is something more universally human in the desire for meaningful work. Graeber locates this universal in the milk of human kindness: people want to feel like their lives are of benefit to others, and frustrating this desire by having them do pointless or harmful work produces serious negative responses.

From my perspective, Bullshit Jobs successfully makes the case that some jobs are not worthwhile contributions to society, and that these jobs can distress those who work them. I particularly like the characterisation of employment as having moved from an outcome-oriented proposition (paying people for what they do) to a presence-oriented proposition (paying people for being available).3 This is largely harmless in itself, but produces pathological outcomes when combined with ever-increasing rounds of efficiency monitoring. It’s fine to keep a accountant on staff rather than using them as a consultant - they build up a knowledge of your business, they’re there to ask stuff to anytime and won’t but otherwise engaged, and it stops them working for the competition - but you won’t have accountancy work enough to keep them occupied 40 hours a week. Then comes the efficiency monitoring, to which an accountant twiddling their thumbs for 35 hours a week outside of the end-of-year rush looks like a massive waste of resources. The defense against piles of additional work is simple: look busy even when you’re not. The 80’s and 90’s were a whirlwind of employees perfecting the looking busy defense, and now it is so established that many companies simply assume the employees are busy, even when they state that they aren’t.

Where Bullshit Jobs is weaker is in its argument that most jobs are being increasingly bullshitised. Nursing, for example, includes far more paperwork (and consequently less actual nursing) now than it did 40 years ago. This phenomenon is present in almost all jobs nowadays. I am, however, unconvinced by many of the examples provided in the book: I don’t doubt that there are forms somewhere which people fill in which are of no practical use to anyone at any point, but, as a scientist used to working with data, I can certainly see the benefits to acquiring data and using it to gain understandings of how systems function at higher levels. It might be a pain in the arse for a doctor to have to log each of their patient’s side effects online, but it will allow us to get a quicker and clearer picture of the notoriously complicated interactions drugs have. Likewise, checklist culture looks an awful lot like bullshitisation, but it saves lives. Similarly, many standards are labyrinthine and wholly arbitrary,4 but they allow systems to work with one another and we could not have the modern world we do without them.5 Keeping records of what has been done and when is tedious and takes up time that could be used to actually do the things being recorded, but it can be vital in discovering what happened if things go wrong.

I certainly accept that there are likely to be massive inefficiencies in the paperwork attendant on many jobs - genuinely pointless data being collected and necessary data being ignored - but I’m not convinced that the growing focus on ensuring things are done consistently and properly is bad. I do acknowledge, perhaps more pertinently for Graeber’s thesis, that this focus on record-keeping etc. can produce frustrations similar to those arising from bullshit jobs. I experience similar issues sometimes when going through the various steps required to ensure the science I produce is open and reproducible: it’d be so much quicker and more fun to just run experiments, but it’s essential that they are done in a way which allows the results to be checked and mistakes corrected.


  1. The actual definition given is “a form of paid employment that is so completely pointless, unnecessary, or pernicious that even the employee cannot justify its existence even though, as part of the conditions of employment, the employee feels obliged to pretend that this is not the case” (pp. 9-10)  2

  2. Graeber allows for parasitic jobs to be included in the bullshit category. Examples he provides include airbrushing cosmetics advertising material to make it look like products work, telemarketing, and most of the finance sector. Part of Graeber’s thesis is that the products of shit jobs are increasingly captured by those doing bullshit jobs, and thus bullshit jobs are a both a product and a cause of shit jobs. 

  3. One of the most fascinating parts of the book is in the earlier chapters where the history of work is discussed, and it is persuasively argued that the rhythm of work natural to most people is intermittent bursts of activity interspersed with extended calmer periods on a variety of timescales. The idea of constantly working at one pace (particularly a backbreaking pace) is decried as a catastrophic error and deep injustice. 

  4. To pick an example with which I am incidentally connected: the standard for a recent version of JavaScript, which ensures that websites work the same way in all modern web browsers. 

  5. Many houses probably still have a drawer full of old mobile phone chargers from the days when each manufacturer (and sometimes each model!) of phone had a different power socket:battery interface. Nowadays almost everything uses one of the USB standards.