Is Automation Really to Blame for Lost Manufacturing Jobs?
foreignaffairs.com
foreignaffairs.com
Her real point is that the electronics industry is doing great, while some other parts of of US manufacturing are not. Fig. 2 is useful. Textiles, apparel, furniture, and paper products are the big losers. The apparel industry was totally crushed by imports from low-wage areas. That took out the US textile industry, since exporting textiles to be made into apparel and re-imported was pointless.
US motor vehicle output is up, not down. That hasn't been crushed by imports.
Primary metals (which is mostly steel and aluminum) are up.
More breakdown of the computers and electronics sector would help. We'd probably see semiconductors through the roof and consumer electronics down. That data is available.
This is useful. Trade is an industry-specific problem. The American apparel industry is mostly gone, but we have cheap clothing. The metals industry complains but isn't in real trouble. Motor vehicles are doing OK. A breakdown of consumer goods by category would help. That would show where to apply tariffs.
Paper products are down because newspapers are obsolete, and they used vast amounts of paper, and because offices use far less paper. Not import-related.
[1] https://upjohn.org/sites/default/files/pdf/state-of-american...
https://www.cnbc.com/2017/10/26/china-is-the-largest-market-...
Maybe when the local factory shut down in years past it transferred 500 jobs straight to China. That doesn't necessarily mean that it kept employing 500 people for the same output in the years since. Re-shoring the same volume of production may mean significantly fewer jobs today, in e.g. textiles:
"U.S. Textile Plants Return, With Floors Largely Empty of People", 2013
https://www.nytimes.com/2013/09/20/business/us-textile-facto...
... Take Parkdale: The mill here produces 2.5 million pounds of yarn a week with about 140 workers. In 1980, that production level would have required more than 2,000 people. ...
Or in the case of fields that still are very little automated, like garment sewing, it would require pretty much the same number of workers now as 20 years ago. As you say, it's very product-specific.
If we broaden the category a bit from "manufacturing jobs" to "industrial jobs" (including things like coal miners and workers at coal fired power plants), you also see strong effects from process-and-input changes. It takes far fewer people to produce a terawatt hour of electricity from a modern combined-cycle natural gas generating plant than from a modern (worse, built-in-1970) coal plant. It also takes fewer workers for a TWh from solar PV or wind power. The rapid growth of jobs in renewable energy is a temporary excursion, mostly an artifact of current rapid expansion. At steady-state it will take fewer people to supply electricity from a renewable-heavy electricity system than a coal-heavy system. That's one reason that I am bullish on the long term price-competitiveness of renewable electricity but skeptical that "green jobs" are a panacea for declining industrial employment in other sectors.
The music industry is the best example of what happens.
Basically, things that used to require all sorts of manufacturing and distribution layers are now more or less compressed into a single digital device.
I also think theirs a good Stephen Hawking quote I like to think of when people discuss how in a post scarcity world with full automation life will be a utopia and its something along the lines of "If machines produce everything we need, the outcome will depend on how things are distributed,"
Guess what automated robotic things you now see that weren't there before ?
Could be that people are not buying as much stuff. I have read that credit card debt is back to 2008 levels. Maybe everyone is maxed out; not good.
I am of the opinion outside metropolis areas the entire nation is running on Empty as far as finance goes. The people I know who live here and work normal blue collar jobs run just under the max on their credit cards and keep alive by paying minimum. Opening new lines of credit to pay for emergencies
Granted this is only my personal experiences but its common enough its disturbing.
When people run out of money to buy things what happens ?
If they keep trying to “sell food” or other illegal activities out of their home, they will be put in prison.
Only if. How about trillions in debt instead.
Yes, it is. Do we miss those jobs? No, we don't, because we're so much wealthier when we can replace human labor with machine labor. That frees up humans to do higher-order tasks, improving our standard of living.
Yes, there are winners and losers with any change, but long-term we are far better off than we could have imagined at the dawn of the Industrial Revolution.
This is a deeply ignorant statement that renders invisible over 2 billion completely disenfranchised people in the third world. Just because you don't experience the consequences - while you gorge on the bounty of human progress somewhere amongst the top decile of our species - doesn't mean they aren't there.
In the past 150 years, the human race has made extraordinary progress, lifting billions out of poverty.
There are still many globally who are left out, and I am aware of that. But you can't argue against the widespread gains to humanity through improved technology.
Western nations, in general do a piss poor job with people in poverty, as we think ourselves beyond the problems of the third world, at least in the united states, its clearly not true - and dont have the right systems in place to allow folks to either fully participate in the low end of the market, or give them meaningful government assistance.
Business right now is frightfully capital intensive - and has a high barrier to entry (paperwork mostly), and in many places there are no jobs available - what is someone to do?
This is a terrible way to view it. If we pack twice as much on a chip vs last year, did output go up? But price went down by about the same amount. Should we measure output in transistors? Printed die area? Dollar price of chips? Number of man-hours required to produce things? Labor cost?
If Intel spent a bunch of cash on a new fab, made circuits with twice as many transistors, reduced prices on the old model and sold the new ones at the old price, then sold the same number as last year, did "output" really change? If yo u think it did, then I ask if it changed in any way that is meaningful to economic measures - it didn't change GDP, it didn't change cost or amount of labor, it didn't even change the companies profits.
Probably not. The ones who were constrained by computing power are now doing higher quality work - FEA comes to mind as something that seems to produce more accurate results with increased computing capability. Games get "better" but their market doesn't get bigger - just more demanding. Hollywood does more CGI, but that doesn't actually increase their market. These are all things that don't affect GDP.
Manufacturing is inherently about mass producing, which means it's always been about scale, which means doing more without adding proportionately to labor. Automation is just one of the many things that help scale, it's just one of the many things that requires capital investment.
The bigger problem for manufacturing jobs has been globalization. When you can take your capital to a market with cheaper workers, then labor becomes a scaling tool (ironically?) in that you can throw labor at a problem to get your labor costs down. What's interesting about this, though, is that at some point you can't do this anymore. Textiles, which is considered level 0 for manufacturing, has nowhere left to go for cheaper labor than Bangladesh (without significant political instability issues).
https://qz.com/1269172/the-epic-mistake-about-manufacturing-...
I mean it's definitely a contributing factor, but only because of the axioms that drive our definition of a good company.
The ideal company is seen as an investment vehicle that creates shareholder value. Full stop.
We don't measure well performance of a company in terms of hiring people. Nor do we even measure it in how well they do what they are there to do. It always getsmeasured in value returned to investors. Investors are generally not laborers in this day and age. And any company that can make more long term by automating and dumping the need for laborers will.
You can blame Automation all you want, but it's something far more basic that makes automation a worthy goal in the first place.
And no, neither automation nor the incentives that make it desirable are "easy" problems to fix.
Even if that were true, that's not a factor that just suddenly started to have force in the last couple of decades.
While it is certainly true that nothing legally changed (edit: this isn't true: there is a full on attack on collective bargaining ongoing) in the more recent 30-40 years to allow for the massive accumulation of wealth and corporate practices that make the 'FULL STOP' in the original comment more accurate than before, the modification of tax law certainly has had a major impact.
> On August 12, 1981, Jack Welch made a speech at The Pierre in New York City called ‘Growing fast in a slow-growth economy’. This is often acknowledged as the "dawn" of the obsession with shareholder value. [0]
Milton Friedman's 1970 op ed in the NYT [1] arguing that profit is the primary social responsibility of a corporation has also been influential on executive thinking in this regard.
[0] https://en.wikipedia.org/wiki/Shareholder_value
[1] https://www.nytimes.com/1970/09/13/archives/a-friedman-doctr...
I feel like you're implying we should. But people having to work is a cost, both to the company and to society as a whole. It's like suggesting we measure the effectiveness of an engine by how much smoke it emits. The measure of an engine is the power it's generating, and the measure of a company is whether it can produce something that's more highly valued than the sum of costs, an attribute that is pretty tightly knit to shareholder value (certain modern era tech companies notwithstanding...).
If companies had to maximize employment they would need to automate fewer things, which would lower productivity and result in less wealth being produced. The entire society would end up being poorer.
This happened in the Communist bloc within living memory and we don't need to make the same mistake again. It's not a coincidence that the countries with the best quality of life are the ones that allow corporations to pursue wealth creation over other goals and then redistribute the wealth through taxation and social benefits. If you want to redistribute wealth it's good to actually have some wealth.
The problem is that we haven't come up with a way to properly pay people in a system where productivity has increased past the point that "full employment" - in the traditional sense - is no longer needed, or possible.
Where you land on how that conflict should resolve is going to depend on your views.
I disagree, at least if democracy is also a goal. Full employment gives individual citizens economic power to go with their political power. Merely "ensur[ing] people have access to goods and services" turns them into economically disempowered dependents.
The older group were workers. They earned income. They invested it, and now are attempting to maintain control of their investments with the last power they have, voting.
Those same young people that have jobs now, that put it in retirement, will be the old people investing in the future.
I believe the same cost also applies to companies. There can be (and almost always is) an environmental, societal, etc cost.
The grimmer reality is that companies compete with one another, and customers - be them consumers, businesses, non-profits, or administrations. They all mind things like cost and quality to a degree, so you need to constantly improve your processes, and indeed automate, to stay in business.
Unfortunately, this is where the perverse incentives start coming into play. Eventually, you hit that point of diminishing returns as an actor in the market.
One of a few things then has to happen.
A) Downsizing, hyper-optimization, and automation to create the illusion of growth and ROI to keep the investment bucks rolling in. This leads to decreased employment, and if unchecked creates value deserts at a certain point of optimization. You end up setting the bar so high to break into an area, that smaller endeavors without some massively disruptive edge being managed by people not interested in being bought out stand a chance at being able to survive.
B) You fail due to investors selling off your stock to reinvest elsewhere. Given enough sell off, you'll likely start seeing stock price go down. No one wants money in a company that can't grow anymore.
The situation in B) can lead to
C) Vertical/Horizontal integration. This tends to be risky, because you start bumping heads with regulatory agencies if you aren't careful, and the word 'monopoly' or 'oligopoly' starts getting thrown around. You CAN make incredible growth numbers by merging with the right things, but the network effects of an emphasis on shareholder returns and a minimization in financial payouts to laborers means a steady positive feedback loop is created, siphoning economic power to the people with shares, which is increasingly NOT the people working jobs. Eventually, there may be wealth transfer via inheritance from shareholding parents, but wages have been minimized for discouraging a large portion of the working population from buying stock.
This can be exacerbated as new goods price starts to take into account that there are generally at least enough well off enough people to the not+well-off people where prices will stabilize on the higher end of the scale over time. Leaving people looking for cheaper lower quality goods they can afford now.
This leads to a manufacturing race to the bottom. See current status quo with manufacturing moving to Asia. This can hold out only until some other metric starts to significantly raise the bar in terms of minimum cost to produce or transport goods. Which coincidentally, environmental concerns are starting to do. Once things start getting more expensive to produce, and the cost gets offloaded to the customer...
Well... You'll have a bunch of pissed off people who can't afford stuff, hired or not.
It'll take a long while. Heck, it's been what, 100, 200ish years, for our population to start getting to the scale technologically where we can even get a non-trivial number of people to even CONTEMPLATE theatre planet's capability to absorb our economic externalities is limited to within a few human lifespans.
It isn't an easy problem. A lotta moving pieces.
Oh, wait...
However, I think 'automation' is really just a bad proxy, technology is -augmenting- workers and re-architecting supply chains. Consider how transformative the standardized shipping container is for an import/export business - regardless of what countries they operate in.
I mean I think you're engaging in selective reasoning.
go dig a trench by hand and get back to us.
If cheap labor is available you use that. If you can't expert the work to a developing nation automation might be the next best option. There's no single boogey man to blame.