. . . who made a legitimate point about how jobs are changing.
The hunt is on for offensive clips of Michael Bloomberg talking off the cuff, and over the weekend a new one circulated. In the version passed around, Bloomberg nonchalantly tells his Oxford audience that he could teach anyone how to farm — “even people in this room”: “It’s a process; you dig a hole, you put a seed in, you put dirt on top, add water, up comes the corn.” He adds that modern “information economy” jobs require more “gray matter.”
The denunciations came quickly and furiously, many of them pointing out that farming these days is a pretty high-tech endeavor. There are three things to know about this episode.
First, the version of the clip that went around left out important context. Bloomberg was very explicitly not talking about modern farmers, but rather about “the agrarian society” that “lasted 3,000 years” before the industrial era, which lasted 300. His point was that the modern economy does not create good, reliable jobs for low-skilled workers the way that past economies did. You can see the full discussion starting around 41:30 here (though the actual query that prompted Bloomberg’s rambling answer starts at 37:30):
Second, while the added context renders countless responses to the clip irrelevant, it doesn’t change the fact that Bloomberg comes across as a condescending jerk. His entire comment feels like an in-joke between him and the elite Oxford audience. At one point he talks of the need to provide the “dignity of a job” to the masses so they don’t “set[] up the guillotines someday.” And of course, even primitive farming hundreds of years ago was not just a matter of dig hole, plant seed, water, up comes corn.
But third, the broad outline of his economic analysis is correct — unremarkable, even — in terms of the basic facts rather than the tone with which they are spoken. So even as Bloomberg proved he was almost comically out of touch, he raised important issues that policymakers really should be thinking about.
It is true that, after living most of their history as hunter-gatherers, humans switched to agriculture and later to industrial production. It’s also true, by definition, that in a society where the overwhelming majority of people actually work as farmers — as was the case even into America’s early years — farming must be something that the overwhelming majority of people can learn to do. And most important, it’s true that work is changing today in ways that give a big advantage to high-IQ people with college degrees, while making it difficult for folks at the bottom as those agricultural and manufacturing jobs disappear.
When Bloomberg emphasized that old-school work involved “processes,” he evoked the concept of the “routine task,” meaning something accomplished by following and repeating a specific series of steps. It has been extremely well documented for more than a decade and a half that jobs focusing on these types of tasks are in marked decline, largely because routine tasks are the easiest type of work to automate.
Back in 2003, one of the earlier papers on this subject found that about 60 percent of the “demand shift” toward college-educated workers from 1970 to 1998 could be explained by computerization. More recent work confirms the decline of routine jobs and adds that robots can displace competing human workers and reduce wages. A famous book about a decade ago argued that, in a “race between education and technology,” the demand for highly skilled workers had outpaced the supply of people with the needed education and training, leading to high wages for a select few and widening inequality.
Unlike farming hundreds of years ago or manufacturing work after World War II, many of these new jobs are not the kind of thing that just about any healthy person with a good work ethic can learn to do. They might require extensive training and technical knowledge, a four-year degree or more, specific personality traits, etc. Some people are of the belief that everyone in the country, no matter their natural intelligence or disposition, is capable of acquiring these things with enough effort, but even if that’s true, we have not come anywhere near making it happen. It’s a legitimate matter of concern.
As I spelled out in a print piece last year, I don’t think it’s time to panic over automation right now, though the combination of automation and trade has certainly clobbered manufacturing work in the areas that relied on it most. Nationwide, unemployment is low and productivity growth is anemic, which is the opposite of what you’d expect to see if robots were taking jobs at a pace that should worry us.
But I don’t think that’s going to be true forever, as robots become able to do a higher and higher percentage of the things humans can. While it’s far in the future, the ultimate fear is that robots will become capable of doing just about anything productive that most people could do — and do it cheaply — paving the way to a world where there’s no need for a lot of people to work at all.
At that point, maybe government handouts like a “universal basic income” will be enough to pacify the folks left behind. Or, as Bloomberg speculated, maybe the pitchforks really will come out for the arrogant technocrats who run everything.