Summary: My post A guaranteed minimum income: faux solution for the new industrial revolution sparked discussions in the comments and on Twitter, with some powerful insights and incisive questions. Here are answers to a few of the questions raised.
A guaranteed minimum income (GMI) is a powerful tool for fighting poverty. As such it has been implemented in many nations (in different forms) and proven to work. My previous post said that it would not work if expanded to offset the job destruction coming from the new industrial revolution. That is, it could provide subsistence level income to the unemployed (preventing mass poverty) — but little more, with ill effects.
- Automation might lead to a massive increase in the number of people collecting subsistence-level GMI. While preventing suffering, it would help create a two-tier society of rich and proles — as the productivity gains from the new industrial revolution went to the owners of the machines. We could end up living in the world of 1800, a high-tech Pride & Prejudice.
- Above subsistence-level GMI would reduce people’s willingness to work at many of the boring, dangerous, or unpleasant jobs that pay low wages — either destroying them or boosting their wages (and so accelerate their automation) — either way boosting the program’s cost.
- It would be politically difficult to implement an above subsistence level GMI. There is support for fighting poverty, although the decades-long rollback of welfare in America shows the limits are quickly reached. But there is today little cultural basis for large redistributive income transfers.
Objections were raised to this analysis. A brief note follows about each of these complex issues, just sketches of an opinion.
- Can we afford a high GMI?
- The GMI is a transitional step, followed by more radical redistributive measures.
- Is a high GMI is politically feasible?
- How would a large class of idle people affect society?
(1) Can we afford a high GMI?
Let’s run the math. The Federal government spends $2.2 trillion on social security, Medicare, and welfare benefits. State and local governments spend $0.688 trillion ($489B on Medicaid and $199B on welfare). Total: $2.9 trillion. That’s paid for by taxes which are on the low end of the post-WWII range (lower than those of most of our peers), with a small deficit (~3-5%) — which also fund a monstrous military/intel budget ($800 – $1T, depending on definition — almost half of total world spending).
Assume the GMI replaces the existing welfare and social security system (probably more cheaply). Assume a high GMI with an average cost per household of $40 thousand/year. That assumes a higher benefit since a GMI would supplement, not replace, the wages for many (most?) people on it. On the other hand, the program would probably have to provide some form of medical care coverage. For comparison, the 40th percentile of household income is $41 thousand; a $15/hour wage at 40 hours/week yields $31 thousand/year.
Under these complex assumptions each trillion dollars spent brings very roughly another 25 million people on the GMI “dole”, equivalent to 16% of the current working population (151 million), with a loss of their income taxes. We probably can afford to spend an added trillion dollars per year if military spending is reduced and taxes are increased (current GDP is $18 trillion). That would take the percent of Americans employed very approximately down to near the 1960 low (note: the graph double counts the part-time reservists with jobs, roughly 800 thousand).
A cautionary note: we can’t predict effects of massive automation
We reach the limits of reliable analysis when looking into the future of the industrial revolution. Automation on a larger scale — eliminating more than a sixth of all jobs in a short period of years — would probably require massive changes in our social, economic, and political systems. I doubt this can be usefully modeled, as we would be deep into the unknown.
The difference between a low and high GMI disappear in the uncertainties if too many jobs disappear too quickly. A 2013 study by two Oxford professors found that 47% of all US jobs are at risk from “computerization”. The rapid advance of technology will increase that number.
Such automation might happen fast. To see how quickly jobs can vanish, look at the number of horses in America — down 90% in two generations (of people).
- 1915: 20,500,000
- 1949: 6,000,000
- 1955: 2,000,000
(2) The GMI is a first step, followed by radical redistributive measures.
This is a logical plan, but I suspect unlikely to work. The 1% (a somewhat arbitrary label for the plutocracy) has been gaining strength since the 1970s, both in income, wealth, and political power. If we get a jump in productivity along with large-scale job losses (neither are certain) and put the unemployed people on the dole (a subsistence-level GMI) — would the 1% become stronger or weaker?
We can only guess about the result of so many hypotheticals, but I suspect the answer is weaker. Perhaps people on the GMI will unite and vote ever greater incomes from the federal treasury. Or perhaps they will form a massive new underclasss, content to receive crumbs for nothing as the fruits of the new industrial revolution flow to the 1% — making them ever stronger.
(3) Is a high GMI politically feasible?
A guaranteed basic income does little to change the distribution of income, unless it is set quite high. A high GMI seems politically problematic in today’s America because it would be seen as generous welfare.
To a large fraction of Americans taxes are theft and welfare is for free-loaders. Make it generous, equal to what people can earn by working, and the 1% will find a large number of supporters eager to kill it. After all, we already see the power of this framing as the existing welfare system was rolled back during the past 2 decades.
This can be easily field tested. Buy some beers in a blue collar bar and start a discussion about a GMI. My guess is that most of the folks there will agree with me.
(4) How would a large class of idle people affect society?
The long-term unemployed on a GMI might become wonderful citizens, wise and educated. Or they might become aimless mobs, consuming entertainment, booze, and drugs. Either way, they’ll be dependents — creating an America unlike the citizenry of independent craftsmen, merchants, and farmers imagined by Jefferson and Jackson (although both past and future rest on the work of non-citizens — slaves then, machines tomorrow).
It would be an experiment. I suspect it would be tried only under duress, after all other solutions had been tied. Chaos is more likely if we fail to plan for the likely effects of automation, and instead react to events.
Other posts in this series
- 50 years of warnings about the new industrial revolution. It’s here. Ignore the naysayers.
- The coming big inequality. Was Marx just early?
- The coming Great Extinction – of jobs.
- Steps to make the tech revolution boost America, not just the 1%.
- The robots are coming, bringing hope of a better future.
- Our future will be Jupiter Ascending, unless we make it Star Trek.
- Well-meant minimum wage increases will accelerate automation.
- The battle of institutions vs. technology = rising wage inequality.
For More Information
For an introduction to the problem see “Will Humans Go the Way of Horses? Labor in the Second Machine Age” by Erik Brynjolfsson and Andrew McAfee in Foreign Affairs, July 2015.
If you liked this post, like us on Facebook and follow us on Twitter. See these books to learn about the new industrial revolution now in progress…
- The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies by Erik Brynjolfsson and Andrew McAfee (2014).
- Rise of the Robots: Technology and the Threat of a Jobless Future by Martin Ford (2015).
- The Future of the Professions: How Technology Will Transform the Work of Human Experts by Richard and Daniel Susskind (2016).