The Iowa Caucus was a disaster. The app for tabulating votes, which was developed in secret by a for-profit company named Shadow Inc, failed. It was essentially untested, and when it could not accurately count the votes, the results were delayed for days.
The fiasco in Iowa is not unique; new technology keeps producing such mishaps. Boeing’s 737 Max has been beset with safety issues: Corners were reportedly knowingly cut, pilots were not trained in new systems, and single points of failure were introduced, resulting in two fatal crashes and an extended grounding of the plane. In Michigan, the state developed an algorithm to detect unemployment insurance fraud, and it proceeded to falsely accuse 34,000 individuals of fraud. It’s not only negligence that leads to technological failures. In 2015, for example, it was discovered that Volkswagen had designed and implemented a device to make diesel vehicles appear to meet environmental regulations when being tested. Six executives were criminally charged.
These incidents as well as the mundane workplace struggles against uncooperative corporate software point to a much larger shift. Across industries and institutions, an upper managerial class has expanded rapidly alongside the equally astronomic rise of consultants selling software, services, and advice to companies. In the last few decades, administrative roles and costs in both health care and higher education have grown. In other industries, power and pay have been concentrated at the top. What we are witnessing, in the academic terminology of the anthropologist David Graeber, is an explosion of bullshit jobs. Aside from software and reports, this growing sector of the economy produces nothing, cleans nothing, fixes nothing, maintains nothing—and, at the end of the day, manages nothing.
Yet this cadre of management has not only been granted exorbitant salaries. Insisting on its own importance, it also reproduces itself by calling for more and more management. Just like the old quip that you need a lawyer because the other party’s lawyer will only talk to a lawyer, management has become ubiquitous because managers will only take other managers or expensive consultants seriously.
In order to pay for all this management, costs must be cut, and so companies and institutions are hollowed out: Those who stay and advance become upper management, and those who are directly involved in the production of goods and services are increasingly outsourced, automated, or replaced by contract workers. Management dreams of a capitalism without labor: workers replaced not by robots, but by an endless series of projects, meetings, and brainstorming sessions. This managerial class has deluded itself into thinking it can run corporations, even an entire economy, without workers. So, it overplays its hand, getting rid of the very workers who know how to make things run.
We experience this directly with the increase of self-checkout lanes at the grocery store. The problem is not really the technology—or even the idea of automating various processes—rather it’s that what is often developed is intended to address a static and well-defined problem. In reality, the job of making airplanes, adjudicating fraud, tallying votes, or complying with regulations is complex, dynamic, and best managed by humans who have developed institutional knowledge. This is not to say that software and automation can never work, but rather that for it to work effectively, the software needs to be maintained, fixed, and updated as situations change.
The drive for automation and production at scale has been shaping capitalism since the industrial revolution. Moishe Postone calls this capitalism’s “treadmill dynamic,” where the drive for increased profits pushes companies to innovate, creating relative (but temporary) advantages. But what distinguishes our current epoch is a larger process wherein both capital (i.e., shareholders, investors, owners, etc.) and management’s relationship to businesses is increasingly extractive. The aim is to suck value from businesses’ mere existence rather than from the products produced or the workers employed. The mess in Iowa is not a one-off accident; it heralds a larger shift in institutional structure and ideology. As long as management continues to regard workers and their hard-won knowledge as irrelevant, it will continue to produce such disasters.