AI and Income Inequality

Does automation always lead to lower costs and greater productivity? Well, obviously, yes. A machine that can assemble 395 paper cups per minute is going to save a lot of money if we install it in our paper cup factory, no question. Even if we keep all out artisanal handcrafts of paper cups and put them to work on the line, we’re still going to pull in more profit.

Or no. Now that we’ve had automatic check out stations and robot cleaners long enough, we’ve learned that the machines don’t always save money or get more work done than the humans. Often the humans end up spending a lot of time serving the machines, but they’re still working and they’re still being paid. Sometimes, in fact, the robots increase the human workload.

In fact, robots making your burger and fries can lead to greater income inequality.”

Automation affects workers who keep their jobs

Justie makes the point we’ve often made here, that human beings are still having to do a lot of the work when robots supposedly take over their jobs. Robots like Flippy can flip burgers, but they need human coworkers to put lettuce and tomato on that burgers, to wrap them up, and to deliver them to the human customers. But he also points out that the fact of having robots in the restaurant can give employers an excuse to cut staff.

The narrative about fast food automation is that the pandemic labor shortages in hospitality and agitation for a higher minimum wage, combined with lower prices on robots, made it worthwhile for fast food owners to invest in automation. Justie suggests that this narrative is causing fewer workers to work harder in fast food establishments.

He makes another strong point: automation requires standardization, which can also be harmful to humans. “In fast food,” he says, “this means small menus with minimal customization.” It probably also means a requirement for the humans to work more precisely and consistently, which is not natural for humans. It makes their jobs less creative, more tedious, and at the same time more difficult.

Income inequality?

Justie’s story has bad guys, for sure. Automation cuts labor costs but only by making the human workers who are left work harder for the bosses who make higher profits. But our old friends Daron Acemoglu and Pascual Restrepo have taken a more academic approach and demonstrated that “between 50% and 70% of changes in the US wage structure over the last four decades are accounted for by the relative wage declines of worker groups specialized in routine tasks in industries experiencing rapid automation.” In fact, “Automation technologies expand the set of tasks performed by capital, displacing certain worker groups from employment opportunities for which they have comparative advantage.” They have the numbers to back up these claims.

When machines take over the dull, dirty, dangerous jobs — as they should, we’d say — the people who do those jobs don’t usually get to step up to more creative and fulfilling work. They usually end up in the service sector, earning much less.

Harry Holzer of Brookings agrees. “In general,” he says in a discussion of how AI will increase the rate of worker displacement even in professional fields,  “automation also shifts compensation from workers to business owners, who enjoy higher profits with less need for labor.”

Justie’s solution is stronger protections for workers. Holzer looks to education. “Our most important challenge is to improve the breadth and quality of education and training,” he says. “To become complementary to AI, more workers will need what researchers call 21st century skills. These include communication, complex analytical skills that often require careful judgements of multiple factors, and creativity. The onus is on K-12 and postsecondary schools to adapt and provide greater emphasis on teaching such skills.”

As is so often the case, the largest measure of success will probably depend on government, eduction, and industry to work together. We in the industrial space can develop strategies for retraining or upskilling displaced workers to minimize job losses. We may be running out of time to chat about this, though. It may be time to take action if we want to avoid the catastrophes being predicted.