The economy is undergoing a range of technological changes whose effects include the displacement of human labor with machines. For example, self-checkouts and online shopping are changes that prevent people from sustaining livelihoods by working as cashiers. ATMs prevent more people from working as bank tellers. Self-driving cars will make it much harder for people to work as cab drivers. And so on. Where will they go? Many pin their hopes for jobs on STEM professions.
According to Bureau of Labor Statistics estimates, the economy had roughly 8.6 million STEM jobs, which represents about 6.2% of the overall labor force. This is a sizable jobs sector. It is bigger than the construction industry (~7 million jobs) and financial services sector (~8 million), and not all that much smaller than the manufacturing sector (~12 million). By many indications, there is reasonably strong demand in these jobs markets as well. The vast majority of STEM occupations earn above-average incomes, and the rate of STEM job growth is expected to be double the general rate of job growth. So the STEM job sector is a fairly big and promising source of work for America’s future.
On its face, STEM jobs seem to hold a promise of absorbing the economically displaced. Perhaps me might hope that people who might have worked in those jobs will instead be programming computers.
It’s possible, but I have doubts. First, STEM jobs are high-skill jobs. They are complicated. As the figure to the right shows (taken from the BLS report), STEM workers are much more educated than the general population. Is it reasonable to expect displaced lower-skill manufacturing or agriculture workers to begin working as programmers, techs, or data scientists?
If we are talking about the highest-skill STEM markets, I harbor some doubts. As someone who teaches applied statistics and data analytics, my view is that a minority of well-educated people have the ability and will to become highly proficient in advanced statistics or difficult programming environments (like R). Heck, a sizable part of the country lacks the basic math skills to understand their own personal finances.
Of course, technology advance to make the world of STEM work more accessible. For example, in data science, there are programs that make it easier to execute complex operations without programming knowledge (e.g., point-and-click Bayesian Analysis, easy to use visualization programs like Tableau). Programs like WordPress make it easy to program web pages. There are a slew of business intelligence dashboard vendors, who deliver ready-to-use statistical analyses to decision-makers with minimal programming or even the need for an in-house analytics expert. I wonder if, as these jobs are simplified and made easier to do, they also become easier to automate. Moreover, simplifying jobs effectively de-skill them, and could conceivably turn better-compensated into worse-compensated jobs.
Hopefully, I’ll be proven wrong. I’m not hopeful that STEM jobs are going to absorb the bulk of those whose occupations will be automated.