Perhaps eventually John Maynard Keynes will finally be right about the four-day workweek. A pilot program in Britain among 61 companies turned out to be popular with workers—hardly a surprise—but was also popular with management:
Companies that participated could adopt different methods to “meaningfully” shorten their employees’ workweeks — from giving them one day a week off to reducing their working days in a year to average out to 32 hours per week — but had to ensure the employees still received 100 percent of their pay....Companies’ revenue “stayed broadly the same” during the six-month trial, but rose 35 percent on average when compared with a similar period from previous years. Resignations decreased.
....Those who took part were less likely to report that they felt they did not have enough time in the week to take care of their children, grandchildren or older people in their lives. The time men spent looking after children increased by more than double that of women, pointing to positive effects of a shorter workweek on gender equality — though there was no change in the share of housework men and women reported taking on.
I'm still trying to cut my workweek to five days, but my brain won't let me. Stupid brain.
Jokes aside, this is a good idea—though for a reason that isn't immediately obvious: robots are on track to eventually start taking away our jobs permanently. That is, there won't be "other" jobs for us to do that robots still can't perform. They'll be able to do everything.
When that starts to happen, maybe in a decade or so, one way to deal with it gracefully would be steady cuts in weekly work hours as human work becomes less necessary—but with the same pay.¹ This could happen with direct government subsidies or simply by government mandate. But something like it is going to happen.
A nationwide change to a four-day workweek would be sort of a pilot run for this. Corporations would learn how best to implement shorter workweeks and researchers would have some time to figure out which types of workweek flexibility work best for different kinds of companies. This will make the transformation to a robot economy smoother, less scary, and more efficient when it finally happens.
¹Or higher pay, since mass robotification will certainly increase productivity massively.
Except that in America, with our (technically) unofficial religion of Calvinist Capitalism, all our thought/political leaders will continue to insist that work is the only legitimate path to affording life's necessities. People driven out of work by automation will be told to get a new job (regardless of whether there are jobs available) or to hie themselves to the workhouse/prison/graveyard.
Je suis d'accord. Our theocracy needs a violent shock and even 9/11 (its subsequent military misadventures) or the Sixth of January couldn't shake loose the parishioners from the Golden Peso.
Keeping people paid will be the hard part. I’m sure the oligarchs are licking their chops at the thought of AI bots raking in as much cash as possible while immiserating the great unwashed masses.
they then run into the problem of needing customers.
As the likely untrue since its too perfect story of the auto exec touring automated factory and saying these robots are great but do they buy cars?
After many years of predicting doom from robots taking all the jobs it appears that Kevin may be catching on to how automation has actually worked over the last 200+ years. It has meant more stuff for everyone, and until around 1940, fewer working hours for those in non-intellectual jobs. Robot weaving machines took the jobs of hand weavers in the early 19th century, spawning the previous Luddite movement, but other jobs arose, as they will when what Kevin is now calling robots take the jobs of white-collar workers. Actually automation has already taken the jobs of many white-collar workers, such as clerks who had to manually copy out every document. But despite a huge increase in automation since the start of the industrial revolution, the unemployment rate has not increased.
The reason that real wages have been stagnant and working hours have not decreased for many decades is not robots, it is a combination of politics and international competition. Since the big switch of the parties on racism, Republicans have used culture wars to get the votes to curtail the political power of wage-earners.
How best to implement a four-day work week:
Literally just do it. That's all it takes. Just set expectations/schedules for four days instead of five. It's literally that easy.
Aside: of course there's something in here about automation and robots. That's not how automation works. What happens is humans just start doing different tasks, not fewer tasks. That's part of why productivity has increased for decades - because automation and other improving technology frees up people-hours from one task, allowing them add another task. The four-day week imperative is entirely separate from this.
The short term pain this will cause to Labor will be exquisite. Without unions, Capital will twist and wring every centime possible until - horrors - government and the electorate assert themselves.*
*This obviously refers to our government as rational government having already tried this made plans ahead of time.
"robots are on track to eventually start taking away our jobs permanently"
Try telling that to pro-immigration types who say we need more people because there are lots of job openings.
There are lots of job openings - which makes it easier for people to leave bad jobs for better ones. A lot of the bad jobs are not yet targets for automation - although I was impressed at the rate berry farms in NW WA managed to automate planting, tending, and harvesting those crops. It went from: "What's THAT weird ass thing?" to: "Hey, there are a whole lot of weird ass things trundling through all the fields around here!"
Brother-in-law was a manager in a paper plant that instituted 36 hour work week a couple of decades ago. Three days of 12 hour shifts, four days off. Initially, all the workers complained. After a week or two, they were very happy.
> ¹Or higher pay, since mass robotification will certainly increase productivity massively.
I think Gordon disagrees: https://www.brookings.edu/wp-content/uploads/2016/07/tt20-united-states-economic-growth-gordon.pdf
Gordon is much more pessimistic about technological advances than Kevin.
I have just skimmed Gordon's article, so far, but here are some issues with his 'headwinds': 1) He argues that high school- and college-completion rates are plateauing or down, holding down productivity growth, but notes the "inability of 40 percent of college graduates to find jobs requiring a college education". How would growth be greater if we graduated more students, such that 50 or 60% were underemployed? 2) The fourth 'headwind' is growing public debt (he also worries about private debt). He doesn't provide any explanation of how a higher debt/GDP ratio will retard growth. I would argue that it depends what the debt finances. The US is expanding infrastructure spending and adopting industrial policies that will increase productivity in multiple ways.
Is it that we aren't creating more new jobs that require a college education because we don't enforce anti-trust laws?
We aren't investing enough in a) research and development, and b) advanced manufacturing, using automation to reduce labor-cost content instead of labor arbitrage, i.e. exporting jobs to low-wage countries.
Because there isn't enough competition?
Gordan may be right, but few enough economists seem to be able to predict a year into the future that I'm not too concerned about their predictions. Gordan's concern with income and wealth disparities does seem to be involved in various ways with slow productivity growth. One that nobody talks about is really terrible management.
Of course most senior management was never that great. However, most of them had started near the bottom of the company and worked their way up. Their advancement wasn't based on total BS. Their peers had twenty or thirty or more years to see what they were made of. The earnings of the top people at the biggest corporations were large, but not so large as to make one enormously rich in a few years. They had been in it for the long haul and that was their mindset.
Now the possibility of getting rich almost overnight exists. Annual incomes of $20 million or more are possible. The new VP or President or CEO may come in with no experience of the industry, much less the particular business. His aims are to maximize his income for the next three or so years, puff up a resume, and find the next job. What happens to the business after that is of no consequence. In reality he can manipulate the balance sheet, BS the investors, and artificially bump up the stock price. Naturally this sort of super short term thinking and total disdain for the company's future is not going to result in the sort of investments that pay off in the long term. And for these guys five years is an unimaginably long time.
Strongly agree, and I would add that the focus on short-term financial performance over long-term sustainability attracts the wrong people into top management. Boards hire finance grads who have little understanding of the advanced technology needed to continuously improve productivity, consequently under-investing in it. Or investing in the most-hyped generic tech, instead of what is appropriate to their business.
"robots are on track to eventually start taking away our jobs permanently. That is, there won't be "other" jobs for us to do that robots still can't perform. They'll be able to do everything."
Unlikely. There is literally no way any robot could possibly do any of the jobs at the small company I work for. About 50% of nearly everyone's job is about building relationships with other people. Robots are never going to do that.
As I was writing this paragraph, someone posted in Slack that a happy customer toke them 'shoutout to @Susan who is "the best as always!"'. Will customers ever say this about some "AI"?
And as for writing code (which everyone now thinks AIs can do)--I can barely write requirements that my offshore team can implement without copious questions and much checking and rechecking of what was done. Sure an AI can write a bubble sort algorithm or the like. But you know what coders almost never need to do? Write bubble sort algorithms or anything like that. It's all complex business logic. Inserting new logic and new features into vast and complicated applications that can't break anything that already works. So no, I am not at all worried or excited that "robots" will start taking jobs away.
(sarcasm, yes) "Meanwhile, many CEOs were beginning plans to cut benefits for employees who were now no longer considered 'full time' as they worked less than 40 hrs. a week. 'I've been looking at upgrading my yacht game', one was quoted, who asked to remain anonymous. 'I've been searching for more ways to easily cut wages and they're just handing me this on a silver platter, thank you!.'"
Don't get me wrong - great idea for us as humans, but unless it's iron-clad, we'll be screwed by it.
"That is, there won't be "other" jobs for us to do that robots still can't perform. They'll be able to do everything." Will they be able to hose me off and change the diaper on my demented ass?
It matters how you implement this. I worked for a computer company that changed to a 4-day 10 hr day where workers could choose whether to have Monday or Friday off. That made it impossible to schedule meetings for Monday or Friday and difficult to reach people in their offices without keeping track of who was in or out on which day. It messed up carpooling arrangements on those days. For women with young children, finding day care for an extra two hours per day was difficult and it made a longer school+childcare day for kids. For managers and supervisors, we had to work 5 10-hr days per week because we had to be present whenever any workers there, BOTH Mondays and Fridays. The company went back to the old schedule ultimately. It might make more sense for the country to adapt a standard work week of the same 4 days, but it doesn't work out well to let everyone do what they want.
AI researchers think the 10 year timeline is bullshit. I'm going to side with the experts.