Why Robots Won't Cause Mass Unemployment
I made a small note in a previous article about how we shouldn’t worry about technology that displaces human workers:
The lamenters don’t seem to understand that increased productivity in one industry frees up resources and laborers for other industries, and, since increased productivity means increased real wages, demand for goods and services will increase as well. They seem to have a nonsensical apocalyptic view of a fully automated future with piles and piles of valuable goods everywhere, but nobody can enjoy them because nobody has a job. I invite the worriers to check out simple supply and demand analysis and Say’s Law.
Say’s Law of markets is a particularly potent antidote to worries about automation, displaced workers, and the so-called “economic singularity.” Jean-Baptiste Say explained how over-production is never a problem for a market economy. This is because all acts of production result in the producer having an increased ability to purchase other goods. In other words, supplying goods on the market allows you to demand goods on the market.
Say’s Law, Rightly Understood
J.B. Say’s Law is often inappropriately summarized as “supply creates its own demand,” a product of Keynes having “badly vulgarized and distorted the law.”
Professor Bylund has recently set the record straight regarding the various summaries and interpretations of Say’s Law.
Bylund lists the proper definitions:
- Production precedes consumption.
- Demand is constituted by supply.
- One’s demand for products in the market is limited by one’s supply.
- Production is undertaken to facilitate consumption.
- Your supply to satisfy the wants of others makes up your demand for for others’ production.
- There can be no general over-production (glut) in the market.
NOT Say’s Law:
- Production creates its own demand.
- Aggregate supply is (always) equal to aggregate demand.
- The economy is always at full employment.
- Production cannot exceed consumption for any good.
Say’s Law should allay the fears of robots taking everybody’s jobs. Producers will only employ more automated (read: capital-intensive) production techniques if such an arrangement is more productive and profitable than a more labor-intensive technique. As revealed by Say’s Law, this means that the more productive producers have an increased ability to purchase more goods on the market. There will never be “piles and piles of valuable goods” laying around with no one to enjoy them.
Will All the Income Slide to the Top?
The robophobic are also worried about income inequality — all the greedy capitalists will take advantage of the increased productivity of the automated techniques and fire all of their employees. Unemployment will rise as we run out of jobs for humans to do, they say.
This fear is unreasonable for three reasons. First of all, how could these greedy capitalists make all their money without a large mass of consumers to purchase their products? If the majority of people are without incomes because of automation, then the majority of people won’t be able to help line the pockets of the greedy capitalists.
Second, there will always be jobs because there will always be scarcity. Human wants are unlimited, diverse, and ever-changing, yet the resources we need to satisfy our desires are limited. The production of any good requires labor and entrepreneurship, so humans will never become unnecessary.
Finally, Say’s Law implies that the profitability of producing all other goods will increase after a technological advancement in the production of one good. Real wages can increase because the greedy robot-using capitalists now have increased demands for all other goods. I hope the following scenario makes this clear.
The Case of the Robot Fairy
This simple scenario shows why the increased productivity of a new, more capital-intensive technique makes everybody better off in the end.
Consider an island of three people: Joe, Mark, and Patrick. The three of them produce coconuts and berries. They prefer a varied diet, but they have their own comparative advantages and preferences over the two goods.
Patrick prefers a stable supply of coconuts and berries every week, and so he worked out a deal with Joe such that Joe would pay him a certain wage in coconuts and berries every week in exchange for Patrick helping Joe gather coconuts. If they have a productive week, Joe gets to keep the extra coconuts and perhaps trade some of the extra coconuts for berries with Mark. If they have a less than productive week, then Patrick still receives his certain wage and Joe has to suffer.
On average, Joe and Patrick produce 50 coconuts/week. In exchange for his labor, Patrick gets 10 coconuts and 5 quarts of berries every week from Joe.
Mark produces the berries on his own. He produces about 30 quarts of berries every week. Joe and Mark usually trade 20 coconuts for 15 quarts of berries. Joe needs some of those berries to pay Patrick, but some are for himself because he also likes to consume berries.
In sum, and for an average week, Joe and Patrick produce 50 coconuts and Mark produces 30 quarts of berries. Joe ends up with 20 coconuts and 10 quarts of berries, Patrick ends up with 10 coconuts and 5 quarts of berries, and Mark ends up with 20 coconuts and 15 quarts of berries.
50 Coconuts (C)
Give 20C for 15B
20C + 10B
10C + 5B (wage)
30 qts. Berries (B)
Give 15B for 20C
20C + 15B
The Robot Fairy Visits
One night, the robot fairy visits the island and endows Joe with a Patrick 9000, a robot that totally displaces Patrick from his job, plus some. With the robot, Joe can now produce 100 coconuts per week without the human Patrick.
What is Patrick to do? Well, he considers two options: (1) Now that the island has plenty of coconuts, he could go work for Mark and pick berries under a similar arrangement he had with Joe; or (2) Patrick could head to the beach and start catching some fish, hoping that Joe and Mark will trade with him.
While these options weren’t Patrick’s top choices before the robot fairy visited, now they are great options precisely because Joe’s productivity has increased. Joe’s increased productivity doesn’t just mean that he is richer in terms of coconuts, but his demands for berries and new goods like fish increase as well (Say’s Law), meaning the profitability of producing all other goods that Joe likes also increases!
If Patrick chooses option 1 and goes to work for Mark, then both berry and coconut production totals will increase. Assuming berry production doesn’t increase as much as coconut production, the price of a coconut in terms of berries will decrease (Joe’s marginal utility for coconuts will also be very low), meaning Mark can purchase many more coconuts than before.
Suppose Patrick adds 15 quarts of berries per week to Mark’s production. Joe and Mark could agree to trade 40 coconuts for 20 quarts of berries, so Joe ends up with 60 coconuts and 20 quarts of berries. Mark can pay Patrick up to 19 coconuts and 9 quarts of berries and still be better off compared to before Joe got his Patrick 9000 (though Patrick’s marginal productivity would warrant something like 12 coconuts and 9 quarts of berries or 18 coconuts and 6 quarts of berries or some combination between those — no matter what, everybody is better off).
Give 40C for 20B
60C + 20B
16C + 7B (wage)
Give 20B for 40C
24C + 18B
If Mark decides to reject Patrick’s offer to work for him, then Patrick can choose option 2, catching fish. It involves more uncertainty than what Patrick is used to, but he anticipates that the extra food will be worth it.
Suppose that Patrick can produce just 5 fish per week. Joe, who is practically swimming in coconuts pays Patrick 20 coconuts for 1 fish. Mark, who is excited about more diversity in his diet and even prefers fish to his own berries, pays Patrick 10 quarts of berries for 2 fish. Joe and Mark also trade some coconuts and berries.
In the end, Patrick gets 20 coconuts, 10 quarts of berries, and 2 fish per week. Joe gets 50 coconuts, 15 quarts of berries, and 1 fish per week. Mark gets 30 coconuts, 5 quarts of berries, and 2 fish per week. Everybody prefers their new diet.
Give 50C for 15B + 1F
50C + 15B + 1F
5 fish (F)
Give 2F for 20C + 10B
20C + 10B + 2F
Give 25B for 30C + 1F
30C + 5B + 2F
The new technology forced Patrick to find a new way to sustain himself. These new jobs were necessarily second-best (at most) to working for Joe in the pre-robot days, or else Patrick would have pursued them earlier. But just because they were suboptimal pre-robot does not mean that they are suboptimal post-robot. The island’s economy was dramatically changed by the robot, such that total production (and therefore consumption) could increase for everybody. Joe’s increased productivity translated into better deals for everybody.
Of course, one extremely unrealistic aspect of this robot fairy story is the robot fairy. Robot fairies do not exist, unfortunately. New technologies must be wrangled into existence by human labor and natural resources, with the help of capital goods, which also must be produced using labor and natural resources. Also, new machines have to be maintained, replaced, refueled, and rejiggered, all of which require human labor. Thus, we have made this scenario difficult for ourselves by assuming away all of the labor that would be required to produce and maintain the Patrick 9000. Even so, we see that the whole economy, including the human Patrick, benefits as a result of the new robot.
This scenario highlights three important points:
(1) Production must precede consumption, even for goods you don’t produce (Say’s Law). For Mark to consume coconuts or fish, he has to supply berries on the market. For Joe to consume berries or fish, he has to supply coconuts on the market. Patrick produced fish so that he could also enjoy coconuts and berries.
(2) Isolation wasn’t an option for Patrick. Because of the Law of Association (a topic not discussed here, but important nonetheless), there is always a way for Patrick to participate in a division of labor and benefit as a result, even after being displaced by the robot.
(3) Jobs will never run out because human wants will never run out. Even if our three island inhabitants had all of the coconuts and berries they could eat before the robot fairy visited, Patrick was able to supply additional want satisfaction with a brand new good, the fish. In the real world, new technologies often pave the way for brand new, totally unrelated goods to emerge and for whole economies to flourish. Hans Rosling famously made the case that the advent of the washing machine allowed women and their families to emerge from poverty:
And what’s the magic with them? My mother explained the magic with this machine the very, very first day. She said, “Now Hans, we have loaded the laundry. The machine will make the work. And now we can go to the library.” Because this is the magic: you load the laundry, and what do you get out of the machine? You get books out of the machines, children's books. And mother got time to read for me. She loved this. I got the “ABC’s” — this is where I started my career as a professor, when my mother had time to read for me. And she also got books for herself. She managed to study English and learn that as a foreign language. And she read so many novels, so many different novels here. And we really, we really loved this machine.
And what we said, my mother and me, “Thank you industrialization. Thank you steel mill. Thank you power station. And thank you chemical processing industry that gave us time to read books.”
Similarly, the Patrick 9000, a coconut-producing robot, made fish production profitable. Indeed, when we look at the industrial revolution and the computer revolution, we do not just see an increase in the production of existing goods. We see existing goods increasing in quantity and quality; we see brand new consumption goods and totally new industries emerging, providing huge opportunities for employment and future advances in everybody’s standard of living.