While watching Bill Maher last night, I heard a common economic myth that I thought I should clear up.
“We Are Losing All The Jobs To Robots”
So, lets take a look today at the data to see if this thesis makes any sense.
I have asked this question for many years. Where on this chart did we lose all the jobs to robots?
Job openings per sector:
Yes this is manufacturing job openings!
Other sectors look the same too.
Get my drift!
Total Job Openings
From Doug Short:
One of the reasons I had a 2% handle with my GDP predictions over the years is that prime age labor force growth peaked in 2007. Its only slowly growing again now, but we should have better demographics in the next decade.
Household formation is a big key for solid consumption. Young people need to buy stuff because they haven’t before, while older Americans tend to save more in their mid 50’s until death. Ages 17-29 and ages 49-65 are very heavy in this cycle.
Census population map very useful.
We can see here prime age labor force growth peaked in 2007, unlike the 1980’s and 1990’s
From Calculated Risk:
Another myth, we don’t make anything anymore. Yes, technology has displaced some workers in the manufacturing sector but we are still the 2nd biggest manufacturing country in the world .
Have a wonderful Happy 4th of July weekend!
Just remember, when (They) say, we have lost all the jobs to robots, we have over 154 million working Americans, 81.5% full time working profile and highest job openings in the 21st century!
Logan Mohtashami is a senior loan officer at AMC Lending Group, which has been providing mortgage services for California residents since 1987.
Why do you only show the FRED data for the value or output of the manufacturing sector, but not show the level of employment in manufacturing, it undercuts your whole thesis.
I haven’t seen anyone claiming robots are taking jobs in restaurants, retail stores, or construction. Service sector jobs that require social interaction require far more complex generalized AI, and something like construction jobs, where the environments are very diverse, and each building construction is usually uniquely architected, meaning that if a robot did exist that could build houses in arbitrary environments, reconfiguring it for each new job would probably be prohibitively expensive at this point to make the cost savings a wash.
However, in manufacturing, where the benefits of scale and high volume mean millions of identical products, and standardized parts, with a standardized manufacturing line, robots can excel, and indeed, US manufacturing output has risen dramtically over the last 20 years while the labor participation in manufacturing and # of factories has declined.
That in and of itself, is a slam dunk that automation (robots), have replaced labor supply needs.
I think you will need to revisit your thesis when you factor in supply chain logistics jobs. Warehouse distribution centers are now being heavily automated, and if self driving vehicles in the next 20 years take over shipping and delivery, another 10 million jobs will quickly disappear, and many many more if car ownership declines in favor of hail-on-demand self-driving vehicles.
I also think same day delivery, e-commerce, and automation in fast casual restaurants, is also reducing retail sector needs. Whether brick-and-mortar retail continues to decline will depend heavily on social/human factors and not economic efficiency.
If you break down manufacturing data itself, you can make a case that 670,000 were lost over the decades due to Robots and we still have over 12 million manufacturing workers as of today.
In regard to manufacturing, even trade surplus Germany has manufacturing jobs fall as a % of the workforce as well as Japan.
For that data and thesis I wrote this
I agree, across all OECD advanced economies, manufacturing has been declining, it’s basically deindustrialization of labor. But I think the pace of automation is moving faster than most people realize, we’re in the early stages. Not a day goes by that I don’t read some new breakthrough paper in machine learning. Those have not translated into actual products yet, it’s like the internet in the 1990s, or the PC revolution in the mid 80s, most people could not have imagined those resulting in WeChat or SnapChat.
The actual outcomes will be unpredictable at this point in 20 years.
The U.S. has over 150,000,000 people working with over 5,700,000 job openings and mass scale of baby boomers leaving the workforce and dying off in the next 30 years.
Replacement workers are the issue for the world, not robots.
The U.S. does have a prime-age labor force growth cycle left in it to replace the millions and millions of boomers retiring and dying of the next 3-4 decades.
Japan, China, Europe doesn’t
More robot data here
Comments are closed.