Slashdot Mirror
Self-Driving Cars Will Boost the Job Market, Says Marc Andreessen (recode.net)
A future with self-driving cars has induced a lot of anxiety about a resulting loss of jobs, but in fact, they'll create tons more jobs, Silicon Valley investor Marc Andreessen (Wikipedia) said at Recode's annual conference on Tuesday evening. "The jobs crisis we have in the U.S. is that we don't have enough workers," he said. From a report: "It's a fallacy," Andreessen said (specifically citing the lump of labor fallacy and the luddite fallacy). "It's a recurring panic. This happens every 25 or 50 years, people get all amped up about 'machines are going to take all the jobs' and it never happens." Andreessen used the example of the rise of the automobile industry a century ago, which many thought would cost the livelihood of everyone whose jobs were to take care of horses. But "the car then created not only a lot of jobs creating cars" but everything else that happened because of the car: Paved streets, restaurants, motels, movie theaters, apartment complexes, office complexes, the entire buildout of suburban America, etc. "The jobs that were created by the automobile on the second, third, and fourth order effects were 100X, 1000X the number of jobs that blacksmiths had," he said.
Also, everything he described was infrastructure bought and paid for by tax dollars. Folks don't want to pay those taxes anymore and the infrastructure spending has more or less stopped. The build out of suburban America was financed by tax payers. They paid to pave the roads, run electricity, phone & internet, etc ,etc. They're done. They don't want to pay anymore.
As near as I can tell my taxes haven't fallen much. What has changed is that taxes on the 1% have fallen dramatically due to tax cutting, shifting things to capital gains, tax shelters, and keeping money overseas. What has also changed is that a dramatic drop in the middle class has meant that less taxes are paid by many people as they simply make less. If a greater share of the total income was taxed at middle class rates it would be a windfall for government coffers.
Past performance is an indicator of future results. The problem is people interpret results poorly.
I keep repeating this: technical progress increases wealth by reducing costs. Costs are ultimately wage-labor. There's one sustainable way to reduce cost: reduce the wage-hours invested in producing a thing.
Each technical improvement first eliminates some jobs. That gives you transitional unemployment. Lower costs mean lower viable prices, which draws luxury goods down into wider markets: it costs little enough for you to target 100,000,000 middle-class consumers instead of 1,000,000 upper-class consumers, you can price it low enough to target a bigger market. That means either current producers or new competitors will try to take the market and make a bigger profit by lowering prices.
Once prices are sufficiently-low, a good is just a consumer good. Everyone has smart phones now--even poor people--so we compete on price at the bottom and on the spread of prestige across income classes. We have economy cars and luxury cars. The lower-class goods have slimmer margins to try to capture the wider market; as costs come down, we start packing more features into these goods, reducing their price, or both.
So, what happens with those lost jobs?
More features means applying more labor. If you cut costs and then increase features rather than lowering price at a certain market level, then you've invested your displaced labor into producing more stuff--each of those new feature components requires labor, and you shift it from the now-cheaper components to the previously-not-incorporated components.
If you're not boosting features, then you're competing on price to capture those low-end markets. Prices come down in terms of labor-hours--that is to say, prices increase more-slowly than wages for non-changing goods as those goods become cheaper to make. The most extreme form of this is prices decreasing.
Examples?
Cars and phones pack more features into roughly the same price or the same proportion of spending (people tend to expend the same percentage of their income on cars; phones tend to keep at the $350, $500, or $900 price points and pack features, rather than inflating). Hard drives and SSDs tend to fall in price per gigabyte; we see hard drives in particular shipping ~$100 units that keep increasing in capacity (500GB a decade and some ago, several TB today).
Food and clothing increase in dollar-price, but more-slowly than inflation (median household expends 33% of its spending on food in 1950, 12.5% today; 12% on clothing in 1950, 3% today).
New technologies outright fall. Cell phones were available for $4,000 in 1983; small hard drives used to cost hundreds of dollars; and new types of display panels come out at multi-thousand-dollar price ranges for a given size and then fall to a few hundred. SSDs also generally sell by size, and so the 32GB, 64GB, 128GB, 256GB, 512GB, and 1TB models keep falling in price, instead of simply changing the available capacities at a price point as with hard drives.
When the proportion of spending on the same goods falls, consumers have more money. They spend that money on new goods. That requires shipping, retail, and other logistics, all domestic; it also requires manufacture or service provision, which may be domestic or import. This is where new jobs are created.
Caveat: Transitional unemployment means exactly what it says. You eliminate jobs with technology, you need to wait a while for the markets to move around and create new jobs. There aren't new jobs waiting for these people; if there were, we wouldn't have 5% unemployment.
That means, yes, technology eliminates jobs; and, yes, technology creates new jobs. They're in proportion, and there's a lag between them. Both sides are arguing from one absolute, and so both sides are wrong; both sides also typically make ludicrous assertions, like the job-creation assertion that 1 human is replaced with 1 machine
Support my political activism on Patreon.