Bullshit Jobs

by David Graeber

  —obscure nineteenth-century joke

  A bourgeois paradise will supervene, in which everyone will be free to exploit—but there will be no one to exploit. On the whole, one must suppose that the type of it would be that town that I have heard of, whose inhabitants lived by taking in each other’s washing.

  —William Morris, 1887

  If the preceding chapters merely described forms of pointless employment that have always been with us in one way or another—or even that have always been with us since the dawn of capitalism—then matters would be distressing enough. But the situation is more dire still. There is every reason to believe that the overall number of bullshit jobs, and, even more, the overall percentage of jobs considered bullshit by those who hold them, has been increasing rapidly in recent years—alongside the ever-increasing bullshitization of useful forms of employment. In other words, this is not just a book about a hitherto neglected aspect of the world of work. It’s a book about a real social problem. Economies around the world have, increasingly, become vast engines for producing nonsense.

  Figure 2 Distribution of the labor force by sector, 1840–2010

  How did this happen? And why has it received so little public attention? One reason it has been so little acknowledged, I think, is that under our current economic system, this is precisely what is not supposed to happen: in the same way as the fact that so many people feel so unhappy being paid to do nothing defies our common assumptions about human nature, the fact that so many people are being paid to do nothing in the first place defies all our assumptions about how market economies are supposed to work. For much of the twentieth century, state Socialist regimes dedicated to full employment created bogus jobs as a matter of public policy, and their social democratic rivals in Europe and elsewhere at least colluded in featherbedding and overstaffing in the public sector or with government contractors, when they weren’t establishing self-conscious make-work programs like the Works Progress Administration (WPA), as the United States did at the height of the Great Depression. All of this was supposed to have ended with the collapse of the Soviet bloc and worldwide market reforms in the nineties. If the joke under the Soviet Union was “We pretend to work; they pretend to pay us,” the new neoliberal age was supposed to be all about efficiency. But if patterns of employment are anything to go by, this seems to be exactly the opposite of what actually happened after the Berlin Wall came down in 1989.

  So part of the reason no one has noticed is that people simply refused to believe that capitalism could produce such results—even if that meant writing off their own experiences or those of their friends and family as somehow anomalous.

  Another reason the phenomenon has been able to sail past people’s heads is that we have developed a way of talking about changes in the nature of employment that seems to explain a lot of what we see and hear happening around us in this regard, but is, in fact, profoundly deceptive. I’m referring to the rise of what’s called the “service economy.” Since the 1980s, all conversations on changes in the structure of employment have had to begin with an acknowledgment that the overall global trend, especially in rich countries, has been for a steady decline in farming and manufacturing, and a steady increase in something called “services.” Here, for instance, is a typical long-term analysis of the US labor force by sector (see figure 2).1

  Often it’s assumed that the decline of manufacturing—which, incidentally, hasn’t declined that much in terms of employment in the United States, by 2010 only returning to about what it was at the outbreak of the Civil War—simply meant that factories were relocated to poorer countries. This is obviously true to an extent, but it’s interesting to observe that the same overall trends in the composition of employment can be observed even in the countries to which the factory jobs were exported. Here, for instance, is India (see figure 3, below).

  Figure 3 Sector contribution to GDP (%), India

  The number of industrial jobs has remained constant or increased slightly, but otherwise the picture is not so very different.

  The real problem here is with the concept of a “service economy” itself. There is a reason I just put the term in quotation marks. Describing a country’s economy as dominated by the service sector leaves one with the impression that people in that country are supporting themselves principally by serving each other iced lattes or pressing one another’s shorts. Obviously, this isn’t really true. So what else might they be doing? When economists speak of a fourth, or quaternary, sector (coming after farming, manufacturing, and service provision), they usually define it as the FIRE sector (finance, insurance, real estate). But back in 1992, Robert Taylor, a library scientist, suggested it would be more useful to define it as information work. The results were telling (see figure 4).

  Figure 4 Information as a Component of the Economy

  As we can see, even in 1990, the proportion of the workforce made up of actual waiters, barbers, salesclerks, and the like was really quite small. It also remained remarkably steady over time, holding for more than a century at roughly 20 percent. The vast majority of those others included in the service sector were really administrators, consultants, clerical and accounting staff, IT professionals, and the like. This was also the part of the service sector that was actually increasing—and increasing quite dramatically from the 1950s onward. And while no one, to my knowledge, has pursued this particular breakdown through to the present, the percentage of information jobs was already rapidly on the increase even in the latter half of the twentieth century. It seems reasonable to conclude this trend continued, and that the bulk of the new service jobs added to the economy were really of this same sort.

  This, of course, is precisely the zone where bullshit jobs proliferate. Obviously, not all information workers feel they are engaged in bullshit (Taylor’s category includes scientists, teachers, and librarians), and by no means all those who felt they are engaged in bullshit are information workers; but if our surveys are to be trusted, it seems evident that a majority of those classed as information workers do feel that if their jobs were to vanish, it would make very little difference to the world.

  I think this is important to emphasize because despite the lack of statistics, there has been a great deal of discussion since the 1990s about the rise of information-oriented jobs and their larger effect on society. Some, like former US Labor Secretary Robert Reich, spoke of the rise of a new tech-savvy middle class of “symbolic analysts” who threatened to gain all the benefits of growth and leave the old-fashioned laboring classes languishing in poverty; others spoke of “knowledge workers” and “information society”; some Marxists even became convinced that new forms of what they called “immaterial labor”—founded in marketing, entertainment, and the digital economy but spilling outside as well into our increasingly brand-saturated, iPhone-happy daily lives—had become the new locus of value creation—leading to prophecies of the eventual rebellion of the digital proletariat.2 Almost everyone assumed that the rise of such jobs had something to do with the rise of finance capital—even if there was no consensus as to how. It just seemed to make sense that, just as Wall Street profits were derived less and less from firms involved in commerce or manufacturing, and more and more from debt, speculation, and the creation of complex financial instruments, so did an ever-increasing proportion of workers come to make their living from manipulating similar abstractions.

  These days, it’s hard to recall the almost mystical aura with which the financial sector had surrounded itself in the years leading up to 2008. Financiers had managed to convince the public—and not just the public, but social theorists, too (I well remember this)—that with instruments such as collateralized debt obligations and high-speed trading algorithms so complex they could be understood only by astrophysicists, they had, like modern alchemists, learned ways to whisk value out of nothing by means that others dared not even try to understand. Then, of course, came the crash, and it turned out that most of the instruments were sc
ams. Many weren’t even particularly sophisticated scams.

  In a way, one could argue that the whole financial sector is a scam of sorts, since it represents itself as largely about directing investments toward profitable opportunities in commerce and industry, when, in fact, it does very little of that. The overwhelming bulk of its profits comes from colluding with government to create, and then to trade and manipulate, various forms of debt. All I am really arguing in this book is that just as much of what the financial sector does is basically smoke and mirrors, so are most of the information-sector jobs that accompanied its rise as well.

  But here we return to the question already raised in the last chapter: If these are scams, who, exactly, is scamming whom?

  a brief excursus on causality and the nature of sociological explanation

  In this chapter, then, I want to address the rise of bullshit jobs and to suggest some reasons this may be happening.

  Of course, in earlier chapters, particularly chapter 2, we looked at some of the more immediate causes for the creation of useless employment: managers whose prestige is caught up in the total number of their administrative assistants or underlings; weird corporate bureaucratic dynamics; bad management; poor information flow. These are important in understanding the overall phenomenon, but they don’t really explain it. We still have to ask, Why were such bad organizational dynamics more likely to occur in 2015 than they were in, say, 1915, or 1955? Has there been a change in organization culture, or is it something deeper: a change, perhaps, in our very conceptions of work?

  We are faced here with a classic problem in social theory: the problem of levels of causality. In the case of any given real-world event, there are any number of different reasons why one can say it happened. These, in turn, can be sorted into different kinds of reason. If I fall into an open manhole, one might attribute this to absentmindedness. But if we discover there has been a sudden statistical increase in the number of people falling into manholes in a given city, one must seek a different sort of explanation—either one must understand why overall rates of absentmindedness are going up there, or, more likely, why more manholes are being left open. This is an intentionally whimsical example. Let’s consider a more serious one.

  At the end of the last chapter, Meena noted that while many people who end up homeless have a history of addiction to alcohol or other drugs, or other personal foibles, many others are teenagers abandoned by their parents, veterans with PTSD, and women fleeing domestic violence. No doubt if you were to pick a random person sleeping on the streets or in a shelter and examine his or her life history, you would find a confluence of several such factors, usually combined with a great deal of just plain bad luck.

  No one individual, then, could be said to be sleeping on the streets simply because he or she was morally reprobate; but even if everyone sleeping on the streets really was morally reprobate in some way, it would be unlikely to do much to explain the rise and fall of levels of homelessness in different decades, or why rates of homelessness vary from country to country at any given time. This is a crucial point. After all, consider the matter in reverse. There have been moralists throughout the ages who have argued that the poor are poor because of their moral turpitude: after all, we are often reminded, it’s easy to find examples of people born poor who became wealthy owing to sheer grit, determination, and entrepreneurial spirit. Clearly, then, the poor remain poor because they didn’t make an effort they could have made. This sounds convincing if you look just at individuals; it becomes much less so when one examines comparative statistics and realizes that rates of upward class mobility fluctuate dramatically over time. Did poor Americans just have less get-up-and-go during the 1930s than during previous decades? Or might it have had something to do with the Great Depression? It becomes harder still to hold to a purely moral approach when one also considers the fact that rates of mobility also vary sharply from country to country. A child born to parents of modest means in Sweden is much more likely to become wealthy than a similar child is in the United States. Must one conclude that Swedes overall have more grit and entrepreneurial spirit than Americans?

  I doubt most contemporary conservative moralists would wish to argue this.

  One must, then, seek a different sort of explanation: access to education, for example, or the fact that the poorest Swedish children aren’t nearly as poor as the poorest American ones.3 This doesn’t mean that personal qualities do not help explain why some poor Swedish children succeed and others do not. But these are different kinds of questions and different levels of analysis. The question of why one player won a game rather than another is different from the question of how hard the game is to play.

  • • •

  Or why people are playing the game to begin with. That’s a third question. Similarly, in cases like these, where one is looking at a broad pattern of social change, such as the rise of bullshit jobs, I would propose we really need to look not at two but at three different levels of explanation: (1) the particular reasons any given individual ends up homeless; (2) the larger social and economic forces that lead to increased levels of homelessness (say, a rise in rents, or changes in the family structure); and, finally (3), the reasons why no one intervened. We might refer to this last as the political and cultural level. It’s also the easiest to overlook, since it often deals specifically with things people are not doing. I well remember the first time I discussed the phenomenon of homelessness in America with friends in Madagascar. They were flabbergasted to discover that in the wealthiest and most powerful country in the world, there were people sleeping on the streets. “But aren’t Americans ashamed?” one friend asked me. “They’re so rich! Doesn’t it bother them to know everyone else in the world will see it as a national embarrassment?”

  I had to concede it was a good question. Why didn’t Americans see people sleeping on the streets as a national embarrassment? In certain periods of US history, they certainly would have. If large numbers of people were living on the streets in major cities in the 1820s, or even the 1940s, there would have been an outcry and some kind of action would have been taken. It might not have been very nice action. At some points, it would probably have meant rounding up vagrants and placing them in workhouses; at other times, it might have involved building public housing; but whatever it might have been, they would not have been left to languish in cardboard boxes on public thoroughfares. Since the 1980s, the same American was more likely to react not with outrage at how social conditions could have come to this pass, but by appeal to explanations of the first level—and conclude that homelessness was nothing more than the inevitable result of human weakness. Humans are fickle beings. They always have been. There’s nothing anyone can do to change this fact.4

  This is why I emphasize that the third level is simultaneously political and cultural—it bears on basic assumptions about what people are, what can be expected of them, and what they can justifiably demand of one another. Those assumptions, in turn, have an enormous influence in determining what is considered to be a political issue and what is not. I don’t want to suggest that popular attitudes are the only factor here. Political authorities often ignore the popular will. Polls regularly find roughly two-thirds of Americans favor a national health care system but no major political party there has ever supported this. Polls also show most Britons favor reinstating the death penalty, but no major political party has taken this up either.5 Still, the larger cultural climate is clearly a factor.

  • • •

  In the case of bullshit jobs, this means we can ask three questions:

  1. On the individual level, why do people agree to do and put up with their own bullshit jobs?

  2. On social and economic levels, what are the larger forces that have led to the proliferation of bullshit jobs?

  3. On the cultural and political levels, why is the bullshitization of the economy not seen as a social problem, and why has no one done anything about it?6

  Much of the confus
ion that surrounds debate about social issues in general can be traced back to the fact that people will regularly take these different explanations as alternatives rather than seeing them as factors that all operate at the same time. For example, people sometimes tell me that any attempt to explain bullshit jobs in political terms is wrongheaded; such jobs, they insist, exist because people need the money—as if this consideration had somehow never occurred to me before. Looking at the subjective motives of those who take such jobs is then treated as an alternative to asking why so many people find themselves in a position where the only way they can get money is by taking such jobs to begin with.

  It’s even worse on the cultural-political level. There has come to be a tacit understanding in polite circles that you can ascribe motives to people only when speaking about the individual level. Therefore, any suggestion that powerful people ever do anything they don’t say they’re doing, or even do what they can be publicly observed to be doing for reasons other than what they say, is immediately denounced as a “paranoid conspiracy theory” to be rejected instantly. Thus, to suggest that some “law and order” politicians or social service providers might not feel it’s in their best interest to do much about the underlying causes of homelessness, is treated as equivalent to saying homelessness itself exists only because of the machinations of a secret cabal. Or that the banking system is run by lizards.

  sundry notes on the role of government in creating and maintaining bullshit jobs

  This is relevant because when, in the original 2013 essay about bullshit jobs, I suggested that while our current work regime was never designed consciously, one reason it might have been allowed to remain in place was because the effects are actually quite convenient politically to those in power; this was widely denounced as crazy talk. So another thing this chapter can do is clarify a few things in that regard.