In a seemingly benign report [PDF] presented earlier this month about middle class incomes and innovation, the chairman of the White House Council of Economic Advisers, Jason Furman, claimed that jobs with a median wage paying less than $20 an hour have an 83% chance of being automated. Those with less than a high school education had a 44% chance of having their jobs automated.
The report was presented to the National Academies of Sciences, Engineering, and Medicine on December 15. It tried to find solutions for the slowdown in middle class income and labor productivity growth.
Furman noted that median family income had grown by 3% between 1948 and 1973, but only 0.4% between 1973 and 2015. He explained that middle class income growth was determined by the middle class’s share of national income, labor participation rate, and labor productivity growth.
While the labor participation rate stayed roughly the same but for more women versus men working, the middle class’s share of the national income declined significantly (from 68% to 52%) as did the average annual labor productivity growth rate (2.8% to 1.8%) in approximately the last 30 years.
These trends—increasing inequality and stagnant wages—have been a major focus of American politics recently, especially after the financial crisis and Great Recession. While the report leaves out the primary culprits—corporate concentration and financialization of the economy by Wall Street—its analysis does offer some insight into why the economy has stalled out for many Americans and how elites think about the problem.
Furman offers four explanations for the growth slowdown, both in labor productivity and generally (measure as total factor productivity) based on the theory of innovation driving those factors: older populations such as America’s are less innovative, the country is between innovation waves, all the easy innovations have been taken, and the economy is not dynamic enough.
The solutions Furman offers are to increase federal support for research and development, improve and increase business investments in innovation through tax reform (such as increasing the Research and Experimentation tax credit), reform the patent system to prevent excessive litigation, and increasing competition generally.
Furman notes that fewer companies and subsequently jobs have been created in the last 35 years, than the decades preceding them and that corporate capital has continued to gain increasing returns in a very safe investment environment.
To put it bluntly, we have an economy mostly run by Wall Street financiers and corporate monopolists. Furman, in true Obama Administration fashion, laments the problem but fails to substantially provide a way to attack it. Not surprising when you consider those same financiers and monopolies underwrite both political parties and the consulting and think tank set Furman travels in.
In short, the rent-seekers have taken over.
But what is easily the most striking part of the report are Furman’s projections on the future of low-skilled work and workers. Using data from the Bureau of Labor Statistics and related studies on automation [PDF 1, 2, 3], Furman claims those working in jobs where the median hourly wage is less than $20 are going to be completely screwed by the coming wave of automation, with 83% of jobs currently done by humans being done by machines instead.
Furman also calculates that 31% of jobs with a median hourly wage of $20 to $40 will also be automated. Jobs with a median hourly wage of over $40 only have a 4% chance of being automated.
Furman claims there is also a correlation between educational attainment and likelihood that a job will be automated, with shares of jobs most likely to be automated highest with workers with less than a high school education and lowest with those with a graduate degree.
In other words, the lower your education level and hourly wage is, the more likely you are to be replaced by a machine in the coming years. So, just in case you thought the working class hasn’t taken enough kicks to the teeth the last decade, here come some more.
While any projection needs to be mitigated by an understanding that the information used for the projections is imperfect, the trend seems pretty clear here. The only hope rests in the notion that automation and technology ultimately don’t kill jobs but change them.