Michael Hicks: Automation and today’s labor market challenges


Today, many businesses struggle to hire the workers they need.

Whatever the causes, this current challenge will surely prompt widespread changes by employers. This type of adjustment isn’t a new phenomenon, but economists allocate very little time explaining the mechanics of change. Nor do we explain that these types of corrections are normal and generally, if not always, make society better off. This is true across many types of labor market changes. But, even as the world improves, there are some winners and losers, or rather each of us experience some benefits and costs. That too is worth explaining, along with some examples.

When businesses cannot hire enough workers at the wage they think is appropriate, they call it a labor shortage. Of course, workers get a voice in work as well, and a business thinking a wage is fair doesn’t matter if a worker doesn’t agree. This process of workers matching with employers is messy and slow, and government cannot do much about it. We try of course, and states are all funded by the federal government to create an online help-wanted database. It’s even possible that in a few years, with a few million dollars more, some states will have just as good a system as monster.com had back in 2004.

The big challenge in matching workers isn’t information about job availability, but rather information about job quality and wages. High worker turnover is a sign of mismatched information about job and worker. In times when there is more demand for than supply of workers, wages should rise. This increases the relative cost of workers and makes automation more cost effective.

In the decade and a half prior to the Great Recession, the U.S. experienced considerable labor market disruption due to inexpensive technology. This technology ranged from digitization of many machines, to inventory control and tracking to robotics. Whenever the cost of labor-saving technology is less than the cost of labor, firms automate some part of work. That experience should inform what we now see in labor markets. As it becomes expensive to hire new workers, business will automate some tasks.

As technology improves and more tasks can be automated, workers tend to fall into two categories—those who are mostly complements to the automation, and those who are primarily substitutes of automation. Complementary workers keep their jobs and are often paid better. Workers who are mostly substitutable become unemployed.

Of course, most of us fall into a continuum between the two. Technology helps us do some things better and replaces other tasks. Here, formal education appears to play a huge role. More formal schooling enables workers to more easily learn new tasks in the wake of new technology. Education, particularly higher education, serves a critical role in making a worker more ‘automation proof.’ Let me offer an example.

The field of economics is almost unrecognizable from the days I was an undergraduate student. Today, the internet holds abundant data, and desktop computers permit us to perform astonishingly complex analysis. This new technology permits one economist to do in a day what a whole team of researchers might have needed several months or years to perform back in 1980. This made economics less expensive, while improving quality. It also added a stunning number of research questions for us to ask and answer. The result was a large increase in the demand for economic research and economists.

This is a largely successful example, but it is worth noting that economists who couldn’t adapt to the new techniques and technology largely disappeared. There are plenty examples of workers not adapting well.

This weekend, my family ordered groceries through a well-known delivery service. We’d purchased two deliveries during COVID and the second was about to expire. The experience was horrible. The scanned products, like cereal were fine, but the fruit purchase was a disaster. For Granny Smith apples, we received Gala apples. Instead of plums we received nectarines, and instead of bananas we were delivered plantains.

I don’t mean to be cruel, for there are people I know and respect who perhaps eat plantains. But, it must be said that plantains are to bananas, what firewood is to cantaloupe.

Oddly, this wasn’t even the worst part of the delivery. Nearly all the wrong fruit we received was either bruised and spoiling or so unripe as to be inedible for a week or so. This experience highlights precisely the type of human skill that machines are very unlikely to replace. Almost all the fruit was labelled, so the personal shopper should not have made such errors. But, if you haven’t actually seen a plantain, you’d be easily fooled into thinking it was a gloriously large banana. If you’ve ever eaten one, you’d never make that mistake again. It is the culinary equivalent of petting a skunk you think is a cat.

Technology may be able to distinguish between bananas and plantains, but we are a long way from low-cost methods of choosing the perfectly ripe cantaloupe or watermelon. Humans are uniquely suited to make such distinctions. Clearly 200,000 years of evolution and a tightly integrated sense of smell, vision and touch confer us with unique capacity to judge such matters. I find it unlikely that anytime soon a robotic personal shopper will replace fruit chosen by a human. This personal shopper didn’t have a robotics problem, but instead a problem of carelessness.

These two skills vary across the entire domain of formal education. Formulating and testing a scientific hypothesis usually requires close to eight years of college education. Distinguishing ripe from unripe fruit requires a few human senses and some experience shopping with a discerning parent. In both cases, technology can combine to make these productive jobs, the adjustment period can be long and require substantial changes to consumers and businesses.

Despite the many skills we are gifted with as humans, the best evidence is that the humans who are exposed through more formal education will be more productive. That higher productivity will mean higher wages. There is a natural limit to selecting fruit by hand, but there are no such limits to many other tasks that combine with automation. There will be some opportunities for these peculiarly human tasks. But, it has always been the case that more formal education is a key to insulating individual workers and the places they live from the risk of automation.

Michael J. Hicks is the director of the Center for Business and Economic Research and an associate professor of economics in the Miller College of Business at Ball State University. Send comments to [email protected].

No posts to display