I live a strange life - due to a funny life situation, I travel regularly between India and the US. One of the starkest differences is observing how many jobs machines have taken in the wealthy US, leaving the US with a dire surplus of labor.
A day in the life of Prashant from Gurgaon
Introducing Prashant, a software architect. Prashant wakes up early, heads to a nearby gym, and works out with his personal trainer. After he gets back from the gym his wife and family are awake. They all eat a light breakfast prepared by their cook. Prashant lives in Gurgaon, but works in Pitampura - 1.5 hours away. He doesn't want to lose productivity so he tethers his laptop to his phone and updates JIRA tickets while his driver worries about the road.
Prashant's wife Aishwarya is also a working professional, so she drops her children off at the creche (Hinglish word for day care) where she stores her children while not in use. After getting home from work, Prashant and Aishwarya are tired and don't want to cook. So as usual they eat the food their cook prepared - tonight it's a spicy paneer chili. After dinner they take a walk. Prashant develops a craving for paan - a delightful contrast to the oily and spicy paneer chili he ate before. So he stops by one of the local paan-walas working by the side of the road. Aishwarya also stops to buy a single cigarette.
Prashant and Aishwarya then walk home, enjoying their minor indulgences.
A day in the life of Mike from California
Introducing Mike from California, a software salesman extrordinaire. Like Prashant, Mike also wakes up early, chugs some brotein and then heads to his a crossfit box where starts a set of deadlifts. Suddenly he hears a disembodied voice - "Don't bend your back on the deadlift, Mike!" The AI critiques his form and guides him through a personalized workout designed to maximize his gainz. After his workout the AI gives him tips on taking a good swelfie.
After his workout he goes back home to his kitchen. "Tea, Earl Grey, Hot" he says, and with some hissing his kitchen-bot prepares him tea. Mike then takes out his phone, orders an egg white omellete, and a few minutes later the bot has produced it. After eating, Mike heads for the shower while the cleaning bot rapidly scurries out of his way.
Mike then takes out his phone and orders an Uber. The self driving Uber shows up and takes him to his office. When he gets to the office he notices his coworker Katrina watching her children on the nannycam. He sees a baby sleeping in the arms of a large, cuddly furry humanoid shape - Katrina's friendly NannyBot. It looks like a stuffed toy but the AI inside keeps her children safe.
After heading home, Mike goes out for a run - he likes to separate his cardio and strength training. As he is coming back from his run, he stops by one of the many food stalls on the way back. "Give me 2 fish tacos, less oil" he asks. He then taps his phone to the terminal to pay. A slightly distorted voice replies to him - "Si senor, tres minutos." Apparently the speakers on the TacoTron need a little adjustment, but three minutes later the bot hands him delicious fish tacos.
Mike then heads home, ready for bed.
Wait what?
My Indian readers probably see quite a bit of themselves in the story of Prashant. But my American readers probably recognize that the story of Mike from California is pure science fiction. Readers with very sharp eyes might notice that the photo of TacoTron is actually a forgery!
In the US, self driving cars are currently research projects owned by Google and Tesla. Mike can't afford an Uber every day so he needs to drive himself to work. There is no kitchen bot, cleaning bot or TacoTron. Katrina, like many professionals was unable to afford child care - in real life she quit work to care for her children herself.
Relative to India, the US has significantly fewer jobs. But machines didn't steal these jobs - these jobs just don't get done.
This is not really what you'd expect if there is a surplus of labor. A surplus is a situation where we have so much supply of something that we are simply unable to consume it all. Yet in the real world we find ourselves in a situation where we desperately want more labor to be performed.
Where did the jobs go?
Lets think about some examples of jobs that get done in India, but don't get done in the US. It's easy to imagine that a paan-wala might not get much business in the US - lots of Americans just don't like paan. But what about the paan-wala's other line of business, selling single cigarettes?
In the US, we actually do have a few such people. In 2014, an American (alleged) cigarette wala became quite famous for a little while. His name was Eric Garner. Unfortunately Eric Garner didn't become famous for doing a great job of selling cigarettes. He became famous because the police choked him to death while he told them 11 times that he couldn't breath. (Warning: that's a link to a video of police killing a guy for no good reason. You probably don't want to watch it.)
Perhaps people in the US can't find jobs as cigarette walas because they don't want to be killed.
As an example of another job, consider food vendors. In India there are a huge number of pani puri walas like this one:
He starts the night as a guy with food but not money; he ends the night as a guy with money but not food. In some of the more corrupt and westernized areas (e.g. Bandra) the police will actually carefully police the vendors and make sure only the approved ones get to work. But that's a very special case - for my western readers, Bandra is roughly the Bombay equivalent of Beverly Hills.
In the US, the situation is far more fraught. Consider the NYT article Outlaws Make Better Lunches. Apparently all food walas are technically illegal, but if a food wala maintains good relations with assorted politically connected people (police, nearby brick&mortar competition) then law might not be enforced. As one might expect, the number of people who are good at cooking is far smaller than the number of people who are good at cooking and crony-style politics. From the article:
...nor can it fully legalize street vending, because opening up hundreds of rent-free spots would damage the real estate economy.
There was a similar motivation behind the lunch truck crackdown in Jersey City back in 2010 and 2011 - lunch trucks were providing quality meals to people and outcompeting the local Au Bon Pain and Burger King. Unsurprisingly the city government didn't support free markets and competition and instead made lunch worse for all the local workers.
Perhaps this is where the jobs went? Destroyed by protectionism demanded by politically powerful actors, namely the "real estate economy"?
When one starts looking for jobs destroyed by protectionist regulations, it's not very hard to find them. Until last year, to become a hair braider in Texas one would need to engage in 2250 hours of training, pass four exams, and spend thousands of dollars. As one would expect, this results in fewer jobs for people, higher prices, and women with less carefully maintained hair. The situation in Texas has gotten better since Isis Brantley was willing to get arrested repeatedly while standing up for her economic rights; most other sectors of the economy won't have someone with similar courage.
Quite a bit of research has been done and the scientific consensus is quite clear: occupational licensing harms consumers and costs jobs.
It's also worth thinking about the supply side. In India, there is a lot of domestic labor performed which could, in principle, also occur in the US. Think maids, butlers, drivers, and the like. Why doesn't it? Lets consider the incentives of a person to take such a job. In the US, we have a strong economic "safety net" which provides tremendous disincentives for work.
In the "poverty" regime of approximately \(0-\)15k/year (I put "poverty" in scare quotes because in India this would be considered "rich" ), there is virtually no incentive to increase one's earnings. There are in fact sharp cliffs where earning $1 more can actually cause your disposable income to go down. In terms of consumption you see even starker effects. See also this CBO report. In short, if a person could earn $15k-20k/year as a maid, they have no incentive whatsoever to actually take this job; they can also have $15-20k/year of consumption by taking advantage of the safety net and they don't even have to work.
India also has a safety net - the Rural Employment Guarantee Act. But India's safety net is structured quite differently - it's along the lines of the Basic Job Guarantee I proposed in an earlier post, or the Civilian Conservation Corp that the notoriously right wing FDR enacted. It's structured roughly as follows; if someone can't find work, the government will provide them low wage labor (typically 120-200 rupees/day). So in India your choices are either work for the government for a very low wage or work for an individual at a moderately higher wage. As simple incentives would predict, when increased labor increases your income, more people are willing to offer labor in the market.
The idea that robots have somehow taken our jobs and we have a surplus of labor is nonsense. A surplus is when you have too much of something. In the US, we have a situation of too little labor; houses are dirty because maids are expensive, professional women quit the workforce because they can't afford child care, and high skill individuals spend lots of time on household labor they don't enjoy. Robots didn't take our jobs.
As a corollary of this argument, the Chinese and Mexicans (or whoever protectionists like Trump and Bernie are complaining about today) also didn't take our jobs.