Bullshit Jobs

by David Graeber

  So Simon’s job was basically to be the all-seeing eye that determined how different parts of a bank’s many moving parts fit together and iron out any incoherences, vulnerabilities, or redundancies he might find. In other words, he’s about as qualified to answer the question as anyone could be. His conclusions?

  Simon: In my conservative estimation, eighty percent of the bank’s sixty thousand staff were not needed. Their jobs could either completely be performed by a program or were not needed at all because the programs were designed to enable or replicate some bullshit process to begin with.

  In other words, forty-eight thousand of the bank’s sixty thousand employees did nothing useful—or nothing that couldn’t easily be done by a machine. These were, Simon believed, de facto bullshit jobs, even if the bank workers themselves were deprived of the means to assess or collectively analyze the situation, and expected to keep any suspicions to themselves. But why didn’t the bank’s higher-ups figure this out and do something about it? Well, the easiest way to answer that question is to observe what happened when Simon did suggest reforms:

  Simon: In one instance, I created a program that solved a critical security problem. I went to present it to an executive, who included all his consultants in the meeting. There were twenty-five of them in the boardroom. The hostility I faced during and after the meeting was severe, as I slowly realized that my program automated everything they were currently being paid to do by hand. It’s not as if they enjoyed it; it was tedious work, monotonous and boring. The cost of my program was five percent of what they were paying those twenty-five people. But they were adamant.

  I found many similar problems and came up with solutions. But in all my time, not one of my recommendations was ever actioned. Because in every case, fixing these problems would have resulted in people losing their jobs, as those jobs served no purpose other than giving the executive they reported to a sense of power.

  So even if these jobs didn’t originate as flunky jobs, which presumably most didn’t, they ended up being maintained as such. The threat of automation, of course, is an ongoing concern in any large enterprise—I’ve heard of companies where programmers will show up to work wearing T-shirts that say “Go Away or I Will Replace You with a Very Small Shell Script”—but in this case, and many like it, the concern went to the very top: to the very executives who (if, for instance, they are involved in private equity in any way) pride themselves on the ruthlessness with which they acquired other corporations and saddled them with enormous debts in the name of downsizing and efficiency. These very same executives prided themselves on their own bloated staffs. In fact, if Simon is also correct, they did so because that’s what a large bank really was: it was made up of a series of feudal retinues, each answerable to some lordly executive.24

  on some ways in which the current form of managerial feudalism resembles classical feudalism, and other ways in which it does not

  The upper quintile is growing in size and income because all the value created by actual productive workers in the lower quintiles gets extracted by those at the top. When the top classes rob everybody else, they need a lot more guard labor to keep their stolen loot secure.

  —Kevin Carson

  If we return to the example of the feudal overlord in chapter 2, this actually makes perfect sense. I was using feudal overlords and retainers as a metaphor at that point. But in the case of banks, at least, it’s not clear how much is metaphor and how much is literal truth. As I pointed out, feudalism is essentially a redistributive system. Peasants and craftsmen produce things, to a large extent autonomously; lords siphon off a share of what they produce, usually by dint of some complex set of legal rights and traditions (“direct juro-political extraction” is the technical phrase I learned in college),25 and then go about portioning out shares of the loot to their own staff, flunkies, warriors, retainers—and to a lesser extent, by sponsoring feasts and festivals and by occasional gifts and favors, giving some of it back to the craftsmen and peasants once again. In such an arrangement, it makes little sense to speak of separate spheres of “politics” and “the economy” because the goods are extracted through political means and distributed for political purposes. In fact, it was only with the first stirrings of industrial capitalism that anyone started talking about “the economy” as an autonomous sphere of human activity in the first place.

  Under capitalism, in the classic sense of the term, profits derive from the management of production: capitalists hire people to make or build or fix or maintain things, and they cannot take home a profit unless their total overhead—including the money they pay their workers and contractors—comes out less than the value of the income they receive from their clients or customers. Under classic capitalist conditions of this sort it does indeed make no sense to hire unnecessary workers. Maximizing profits means paying the least number of workers the least amount of money possible; in a very competitive market, those who hire unnecessary workers are not likely to survive. Of course, this is why doctrinaire libertarians, or, for that matter, orthodox Marxists, will always insist that our economy can’t really be riddled with bullshit jobs; that all this must be some sort of illusion. But by a feudal logic, where economic and political considerations overlap, the same behavior makes perfect sense. As with the PPI distributors, the whole point is to grab a pot of loot, either by stealing it from one’s enemies or extracting it from commoners by means of fees, tolls, rents, and levies, and then redistributing it. In the process, one creates an entourage of followers that is both the visible measure of one’s pomp and magnificence, and at the same time, a means of distributing political favor: for instance, by buying off potential malcontents, rewarding faithful allies (goons), or creating an elaborate hierarchy of honors and titles for lower-ranking nobles to squabble over.

  If all of this very much resembles the inner workings of a large corporation, I would suggest that this is no coincidence: such corporations are less and less about making, building, fixing, or maintaining things and more and more about political processes of appropriating, distributing, and allocating money and resources. This means that, once again, it’s increasingly difficult to distinguish politics and economics, as we have seen with the advent of “too-big-to-fail” banks, whose lobbyists typically write the very laws by which government supposedly regulates them, but even more, by the fact that financial profits themselves are gathered largely through direct juro-political means. JPMorgan Chase & Co., for example, the largest bank in America, reported in 2006 that roughly two-thirds of its profits were derived from “fees and penalties,” and “finance” in general really refers to trading in other people’s debts—debts which, of course, are enforceable in courts of law.26

  It’s almost impossible to get accurate figures about exactly what proportion of a typical family’s income in, say, America, or Denmark, or Japan, is extracted each month by the FIRE sector, but there is every reason to believe it is not only a very substantial chunk but also is now a distinctly greater chunk of total profits than those the corporate sector derives directly from making or selling goods and services in those same countries. Even those firms we see as the very heart of the old industrial order—General Motors and General Electric in America, for example—now derive all, or almost all, of their profits from their own financial divisions. GM, for example, makes its money not from selling cars but rather from interest collected on auto loans.

  Still, there is one crucial difference between medieval feudalism and the current, financialized version. We’ve already mentioned it earlier in the chapter. Medieval feudalism was based on a principle of self-governance in the domain of production. Anyone whose work was based on some kind of specialized knowledge, whether lace makers, wheelwrights, merchants, legal scholars, was expected to collectively regulate their own affairs, or including who would be allowed to enter the profession and how they would be trained, with minimal supervision from anybody else. Guilds and similar organizations typically had elaborate hierarc
hies within (though not always so much as they do today: in many medieval universities, for instance, students elected their professors), but at the very least, a medieval sword smith or soap maker could go about his work in the confidence that he would never have anyone who was not himself a sword smith or a soap maker telling him he was not going about it correctly. Industrial capitalism obviously changed all that, and the rise of managerialism in the twentieth century drove the process even further; but rather than this in any sense reversing under financialized capitalism, the situation has actually worsened. “Efficiency” has come to mean vesting more and more power to managers, supervisors, and other presumed “efficiency experts,” so that actual producers have almost zero autonomy.27 At the same time, the ranks and orders of managers seem to reproduce themselves endlessly.

  • • •

  If one wants a parable for what seems to have happened to capitalism over the last forty-odd years, perhaps the best example I know is the Elephant Tea factory outside Marseille, France, currently occupied by its employees. I visited the plant a few years ago, and one of the occupiers—who took me and some friends on a tour of the grounds—told us the story of what happened. Originally, it was a local enterprise, but during the age of mergers and acquisitions, the company was bought up by Unilever, owner of Lipton, the world’s largest tea producer. At first, the company left the organization of the plant more or less alone. The workers, however, were in the habit of tinkering with the machinery, and by the nineties, they had introduced a series of improvements that sped up production by more than 50 percent, thus markedly increasing profits.

  Now, in the fifties, sixties, and seventies, there was a tacit understanding in much of the industrialized world that if productivity in a certain enterprise improved, a certain share of the increased profits would be redistributed to the workers in the form of improved wages and benefits. Since the eighties, this is no longer the case. So here.

  “Did they give any of that money to us?” our guide asked. “No. Did they use it to hire more workers, or new machinery, to expand operations? No. They didn’t do that, either. So what did they do? They started hiring more and more white-collar workers. Originally, when I started working here, there were just two of them: the boss and the HR guy. It had been like that for years. Now suddenly there were three, four, five, seven guys in suits wandering around. The company made up different fancy titles for them, but basically all of them spent their time trying to think of something to do. They’d be walking up and down the catwalks every day, staring at us, scribbling notes while we worked. Then they’d have meetings and discuss it and write reports. But they still couldn’t figure out any real excuse for their existence. Then finally, one of them hit on a solution: ‘Why don’t we just shut down the whole plant, fire the workers, and move operations to Poland?’ ”

  Generally speaking, extra managers are hired with the ostensible purpose of improving efficiency. But in this case, there was little to be improved; the workers themselves had boosted efficiency about as much as it was possible to do. But the managers were hired anyway. What this suggests is that what we are really dealing with here has nothing to do with efficiency but everything to do with changing understandings of the moral responsibilities of corporations. From roughly 1945 to 1975, there was what is sometimes referred to as a “Keynesian bargain” between workers, employers, and government—and part of the tacit understanding was that increases in worker productivity would indeed be matched by increases in worker compensation. A glance at the diagram on the next page confirms that this was exactly what happened. In the 1970s, the two began to part ways, with compensation remaining largely flat, and productivity taking off like a rocket (see figure 7).

  These figures are for the United States, but similar trends can be observed in virtually all industrialized countries.

  Where did the profits from this increased productivity go? Well, much of it, as we are often reminded, ended up swelling the fortunes of the wealthiest 1 percent: investors, executives, and the upper echelons of the professional-managerial classes. But if we take the Elephant Tea factory as a microcosm for the corporate world as a whole, it becomes obvious that wasn’t all that happened. Another considerable chunk of the benefits of increased productivity went to creating entirely new and basically pointless professional-managerial positions, usually—as we’ve seen in the case of universities—accompanied by small armies of equally pointless administrative staff. As we have seen so often, first the staff is allocated and then someone has to figure out what, if anything, they will actually do.

  Figure 7

  In other words, the feudal analogy is not even really an analogy. Managerialism has become the pretext for creating a new covert form of feudalism, where wealth and position are allocated not on economic but political grounds—or rather, where every day it’s more difficult to tell the difference between what can be considered “economic” and what is “political.”

  Another classic feature of medieval feudalism is the creation of hierarchies of ranked nobles or officials: a European king might grant land to a baron in exchange for providing a certain number of knights to his army; the baron, in turn, would grant most of that land to some local vassal on the same basis, and so on. Such devolution would proceed, through a process of “sub-infeudation,” down to local lords of the manor. This was the process by which the elaborate ranks of dukes, earls, viscounts, and so forth that still exist in places like England originally came into being. In India and China, matters were typically more indirect; the usual practice was to simply allocate the income from a certain territory or province to officials who were likely to actually live in the nearest city, but for our purposes here, the result is not so very different.28

  As a general principle, I would propose the following: in any political-economic system based on appropriation and distribution of goods, rather than on actually making, moving, or maintaining them, and therefore, where a substantial portion of the population is engaged in funneling resources up and down the system, that portion of the population will tend to organize itself into an elaborately ranked hierarchy of multiple tiers (at least three, and sometimes ten, twelve, or even more). As a corollary, I would add that within those hierarchies, the line between retainers and subordinates will often become blurred, since obeisance to superiors is often a key part of the job description. Most of the important players are lords and vassals at the same time.

  how managerial feudalism manifests itself in the creative industries through an endless multiplication of intermediary executive ranks

  Every dean needs his vice-dean and sub-dean, and each of them needs a management team, secretaries, admin staff; all of them only there to make it harder for us to teach, to research, to carry out the most basic functions of our jobs.

  —anonymous British academic29

  The rise of managerial feudalism has produced a similar infatuation with hierarchy for its own sake. We have already seen the phenomenon of managers whose job it is to manage other managers, or the elaborate mechanisms Irene described whereby banks set up a hierarchy of offices to endlessly rarify what’s ultimately an arbitrary and meaningless set of data. Often, this kind of managerial sub-infeudation is a direct result of the unleashing of “market forces.” Recall here Kurt, with whom we began chapter 1, who was working for a subcontractor to a subcontractor to a subcontractor to the German military. His position was the direct outcome of market reforms supposedly designed to make government more efficient.

  The same phenomenon can be observed in a dozen different fields. For instance, the multiplication of levels of managers whose basic job is to sell things to one another has come to dominate almost all “creative industries”—from books, where editors at academic presses in many cases don’t even read half the books they are supposed to have edited, because they are expected to spend most of their time marketing things to other editors; to the visual arts, where recent decades have seen the rise of a whole new stratum of manager
ial intermediaries called curators, whose work assembling the work of artists is now often considered of equal value and importance to the art itself; to even journalism, where the relationship between editors and reporters has been complicated by an additional level of “producers.”30 Film and television have fared particularly badly. At least, so it seems from testimonials within the industry. Where once the Hollywood studio system relied on a relatively simple relation between producers, directors, and writers, recent decades have witnessed an apparently endless process of managerial sub-infeudation, resulting in a daunting array of producers, subproducers, executive producers, consultants, and the like, all in constant search for something, anything, to actually do.31

  I received several testimonies from workers in TV “development”—that is, small companies in the business of coming up with programming ideas to pitch to larger ones. Here’s an example that illustrates just how much the introduction of market elements within the process has changed things:

  Owen: I work in development. This part of the television industry has expanded exponentially in the last twenty years. TV used to be commissioned by one channel controller who would ask producers he liked to make whatever shows they wanted. There was no “development.” There was just making the show.

  Now every company in TV (and film, too) has its own development team, staffed by three to ten people, and there are more and more commissioners whose job it is to listen to their pitches. None of these people make TV shows.

  I have not gotten a show sold for four years. Not because we are particularly bad but because of nepotism and politics. That’s four years that have amounted to precisely nothing. I could have sat with my thumb up my arse for four years, and nothing would be any different. Or I could have been making films.