About a year ago my supermarket installed its first self-service checkout stations — just a couple at first, and little-used for a time. I hesitated myself, feeling life already featured enough of Mark Twain's "modern inconveniences."

The store has since more than doubled the number of DIY scanners and scales — and now I'm often inconvenienced waiting my turn.

The expanding self-service checkout lane gets me to thinking about ... horses.

Macalester College economist Timothy Taylor has posted a good deal on his excellent "Conversable Economist" blog about automation and how it is or isn't affecting jobs and the economy. Recently, he cited predictions that the revolution in "autonomous vehicles" speeding toward us could soon transform 1 in every 9 American jobs — from truck driving to policing. Some behind-the-wheel professions will simply disappear; others may be much improved.

As the "driverless car" bears down on us, it's interesting to think about how the "horseless carriage" rearranged work in its era. A while back, Taylor quoted an economist of a previous generation named Wassily Leontief on the eclipse of America's horse culture roughly a century ago.

America was home to some 20 million domesticated horses back then. Today there are fewer than half that many, despite the human population more than tripling. Today, of course, horses are kept mainly for recreation, but they used to be, well, America's workhorses. As transportation and agriculture became motorized, many millions of tireless four-legged laborers were put out of a job.

That enormous labor dislocation was politically manageable for an obvious reason. "Horses don't vote," Leontief noted. But what would happen, he wondered, "when you have the same problem with people"?

The fear of technology making human workers obsolete has haunted skeptics of market capitalism for centuries. Economists have mostly dismissed such concerns, and history has mostly proven them right. Jobs that once employed millions vanished — from spinners and weavers to lamplighters to gas-pump jockeys and many more. But they were replaced by new, better-paying jobs (often making, operating or applying new technologies) as overall productivity increased.

It wasn't just the horses, after all, who became superfluous a century ago — but also blacksmiths, stable keepers, harness makers, street sweepers, etc., etc.

Yet manufacturing and accommodating motorized vehicles became a vast source of employment and wealth.

And yet, all this said, it does seem that, more and more, thoughtful economic observers are wondering whether rising wage inequality in our era and falling workforce participation may owe something to the uniquely widespread effects of information technology, and the ways it combines with globalization to transform labor markets.

Taylor remains mostly an optimist about technology and its effects on work. Often-predicted mass unemployment hasn't come, he notes. But he reported on his blog how Leontief observed that, back in the Model T era, "a farmer couldn't ... postpone the change to tractors by feeding [his horses] less oats."

Yet perhaps today's market forces can do something a bit like that, putting downward pressure on the wages of lower-skill workers who perform routine tasks that machines can take over (or make exportable) — while paying premium wages for higher-order skills that help exploit technology's productive potential.

Anyhow, adding to all the uncertainty and complexity is public policy, often designed to help those disadvantaged by changing times. A new study published by the National Bureau of Economic Research — "People Versus Machines: The Impact of Minimum Wages on Automatable Jobs" — probes how minimum-wage hikes affect the pace at which workers in "automatable jobs" are replaced by technology.

Grace Lordan of the London School of Economics and David Neumark at the University of California, Irvine, write that they "focus on automation as it has been one of the dominant forces that has threatened low-skilled jobs in the United States in recent decades … . Minimum wages can exacerbate these changes, when they raise the price of low-skilled labor in automatable jobs, for which machines can be substituted."

Noting that states and cities across the country (including Minnesota and Minneapolis) are boosting minimums to new levels, Lordan and Neumark worry that debate over such policies has overlooked the potential effects on older, lower-skilled workers in jobs beyond service and hospitality industries, including manufacturing.

Their conclusions aren't reassuring. "Based on … data from 1980-2015," they write, "we find that increasing the minimum wage decreases significantly the share of automatable employment held by low-skilled workers, and increases the likelihood that low-skilled workers in automatable jobs become unemployed … including substantive adverse effects for older, low-skilled workers in manufacturing."

As always, new technology will create new jobs while eliminating old ones. "[I]ncreases in minimum wage," Lordan and Neumark add, "will give incentives for firms to adopt new technologies that replace workers earlier. While these adoptions undoubtedly lead to some new jobs, there are workers who will … not have the skills to do the new tasks. Our paper has identified workers whose vulnerability to being replaced by machines has been amplified by minimum wage increases. Such effects may spread to more workers in the future."

All such assessments are subject to dispute — and no doubt a future with fewer drivers and cashiers is coming with or without minimum-wage hikes.

But policymakers should consider carefully how best to help vulnerable workers adjust to the ever more automatable world ahead.

Doug Tice is at Doug.Tice@startribune.com.