By Matthew Yglesias. SlowBoring.com.
April 25, 2021.
Economic progress has slowed down, not sped up
Billionaires’ Row in New York City (Photo by Gary Hershorn/Getty Images)
Paul Graham, the influential venture capitalist, wrote a good essay recently looking at the difference between the richest people in America today versus the richest people in the America of 1982.
It’s a good essay and you should read it, but the spoiler is that today a much larger share of the richest people are founders especially of “tech” companies.
And he goes on to lay out what is, I think, a widely-shared Silicon Valley View of the past 40 years of the American economy, namely that the rise of today’s ultra-rich founders in contrast to the heirs and heiresses of 1982 reflects the rise of a more dynamic economy:
So it's not 2020 that's the anomaly here, but 1982. The real question is why so few people had gotten rich from starting companies in 1982. And the answer is that even as the Herald Tribune's list was being compiled, a wave of consolidation was sweeping through the American economy. In the late 19th and early 20th centuries, financiers like J. P. Morgan combined thousands of smaller companies into a few hundred giant ones with commanding economies of scale. By the end of World War II, as Michael Lind writes, “the major sectors of the economy were either organized as government-backed cartels or dominated by a few oligopolistic corporations.”
In 1960, most of the people who start startups today would have gone to work for one of them. You could get rich from starting your own company in 1890 and in 2020, but in 1960 it was not really a viable option. You couldn't break through the oligopolies to get at the markets. So the prestigious route in 1960 was not to start your own company, but to work your way up the corporate ladder at an existing one.
This Silicon Valley View gets two important things right:
Progress and change are good and important, and whether or not you find yourself annoyed by annoying people who use innovation as an annoying buzzword, it is true that innovation is central to human progress.
The rich tech founders who dominate the billionaire league tables are not bad guys who got rich by exploiting the masses; they got rich primarily through a mix of lucky breaks and good ideas that have made things better.
But there’s a big problem with the Silicon Valley View. If you look at the tippy top of the economy, you see a lot fewer heirs in 2021 than you had in 1981 and a lot more tech founders — it seems like the economy has grown more dynamic. But if you look at the aggregate productivity statistics, you see a sharp slowdown in the mid-1970s that we never really recover from. So the change at the top of the Forbes rankings can’t be a cause or a consequence of the greater dynamism of the American economy, because the American economy has become less dynamic, not more.
The great productivity slowdown
There are different ways to measure and think about productivity, but one of the most important is what the economists call total factor productivity (TFP). Growth in TFP is growth in output that isn’t accounted for by increases in the amount of capital goods or labor used. And the most dramatic illustration of the slowdown is this chart that Eli Dourado made where he compares utilization-adjusted TFP to where we’d be if we had continued with the old pattern of steady 2% annual growth.
A lot of theory and math go into creating that measurement, so you don’t need to take it all the way to the bank with you.
The point is that basically all measures of productivity show something similar. A Bureau of Labor Statistics report that came out this month looks just at the productivity slowdown since 2005 and notes it amounts to about $95,000 less in output per worker than we would have had if we’d sustained those higher growth rates. The number gets so big because the recent TFP slowdown was paired with a slower rate of growth in human and physical capital, amounting to an all-around collapse of labor productivity.
Many people have seen the charts showing that median wages have grown much slower than mean output per worker, and drawn the conclusion that productivity growth doesn’t matter anymore, and it’s all about inequality.
This is not right — the Obama administration’s 2015 Economic Report of the President did the math and showed that the productivity slowdown has been a bigger deal for middle-class income stagnation than inequality.
One upshot of this is that the leftists who just want to complain about billionaires and inequality and forget about innovation and technology are wrong.
But it’s also a huge problem for the Silicon Valley View, which is that the rise in inequality is okay because it’s blessed us with all this innovation. It’s true that, in theory, accepting more inequality as the price to pay for more innovation would be a good deal. So if it were actually the case that the economy had grown more dynamic since the mid-70s, that would be great. But what actually happened is that even as the oil shocks of the 1970s ended, productivity growth remained sluggish all throughout the 1980s and early 1990s. Then there was this roughly 10-year burst from 1996-2005 associated with big box stores and trade with China. But then, instead of charging into a bold new economy, things flattened out again.
Computers played a big role in the productivity burst associated with the rise of Walmart and other large, efficient, modern retailers, but the consumer-facing technology boom of smartphones and the internet has been something of a dud in terms of productivity.
Myth and mismeasurement
The standard Silicon Valley View of this is that it’s all a form of mismeasurement. It’s obvious that computers and the internet have transformed our lives, and if the aggregate statistics don’t reflect that, then so much worse for them.
In technical terms, the observation here is that GDP counts what is bought and sold and thus doesn’t capture the consumer surplus inherent in the provision of free, ad-supported goods.
And it’s unquestionably true that the internet has brought us a lot of that stuff. But for mismeasurement to do the work the Silicon Valley View needs, you can’t just say that the internet features consumer surplus. You have to say that the amount of consumer surplus has accelerated relative to what people gained in past decades. That does not seem remotely plausible to me. The rise of free, ad-supported broadcast radio and television was a really big deal that transformed the world more than the rise of on-demand streaming video.
Chad Syverson looks internationally and finds there’s no correlation between the IT intensity of a country’s economy and the degree to which it’s been hit by the productivity slowdown.
You can also try to measure this in different ways. Erik Brynjolfsson, Avinash Collins, and Felix Eggers find that the median participant in their experiment would require $48 per month to be willing to abandon Facebook, which suggests the company is raising human welfare in unmeasured ways.
How much would a person in 1982 have demanded to give up access to CBS forever? Unfortunately, nobody did that experiment. A different experiment actually did get people to abandon Facebook for four weeks, and found that doing so “increased subjective well-being … and caused a large persistent reduction in post-experiment Facebook use.” After all, it would be a mistake to travel back in time to 1982, ask smokers how much money they would require to never smoke again, and then conclude that tobacco companies were generating massive unmeasured welfare gains.
Separate from the specifics of Facebook, I think these technologies haven’t improved welfare as much as the Silicon Valley View wants them to because they don’t aggregate up in the right kind of way.
Picking whatever show I want to watch off one of the various streaming services I subscribe to is obviously better than watching “The Single Guy”1 because it happens to have the timeslot between “Friends” and “Seinfeld.” But in terms of the overall impact on quality of life, everyone watching the same stuff had certain kinds of offsetting benefits. Any given Friday, you could rely on chats about the previous evening’s Must-See TV. The point here isn’t to go full Luddite and say “actually, it was better when we had no choices.” Just that we’ve been having a lot of fun staring at various screens the whole time. Smartphones are more genuinely transformative because they let you extend screen-fun into times and places where it wouldn’t go in the 1980s. But George Jetson only worked nine hours a week — that’s what a productivity revolution looks like. You get a huge increase in actual leisure time, not the ability to squeeze a little more phone time into the dead parts of your day.
I don’t want to veer too far off into the realm of hot takes here — just to say that we should have a relatively high bar before we proclaim the GDP figures useless, and I don’t think mismeasurement theorists have cleared it.
More innovations, fewer rich founders?
Back during the 2012 presidential campaign, Mitt Romney slammed Barack Obama’s green jobs initiatives as having wasted a boatload of cash on failed companies like Solyndra and Tesla.
But don’t forget, you put $90 billion, like 50 years’ worth of breaks, into—into solar and wind, to Solyndra and Fisker and Tesla and Ener1. I mean, I had a friend who said you don’t just pick the winners and losers, you pick the losers, all right? So this—this is not—this is not the kind of policy you want to have if you want to get America energy secure.
Yep. The lame-brain government that can’t do anything right gave Tesla a low-interest loan during the worst days of the financial crisis which allowed the company to survive, repay the loan early, and now Elon Musk duels with Jeff Bezos for the title of world’s wealthiest man.
That $90 billion figure is wrong, but it is true that the American Recovery and Reinvestment Act gave subsidized loans to greentech companies to try to keep the industry alive and thriving. It’s also true that ARRA was, in retrospect, much too small. Suppose the government had spent four times as much on the program that helped save Tesla and that meant we had two high-growth electric car startups by the mid-teens? The world would be a much better place, since as much as many people love their Teslas, the nature of consumer goods is that nothing is perfect for everyone. But competition is often bad for individual businesses. The combined market caps of Tesla and Tesla 2 might be smaller than that of Tesla alone today. At a minimum, in a world with more innovation in the electric car space, it’s likely that Musk personally would be a lot less rich.
Or think about Bezos. Back in February 2000, he made a timely sale of convertible bonds2 in Europe right before the NASDAQ crashed in March.
We tend to remember the dot-com era as full of charlatans with doomed business ideas. But at the time, a lot of people thought they were visionaries. And a lot of other people thought Bezos was a charlatan. If one other 1990s-vintage e-commerce player3 had been as savvy as Bezos with his market timing, maybe we’d have had two big players in that space in the early 21st century. Again, good for the world, but bad for the richest guy on the planet.
Mark Zuckerberg is rich because lots of people like to use Facebook. But he’s also a smart investor. When he paid $1 billion for Instagram, lots of people thought he was nuts and it was a sure sign of a huge bubble in tech stocks. I got out of the second-guessing Mark Zuckerberg game a long time ago, so my contrarian take was that he was making a savvy play to head off a potential future competitor. Today I think the conventional wisdom is that regulators should not have allowed that deal to go through.4 I don’t think there’s any conceivable universe in which the FTC would have blocked that acquisition — I was there and at the time, everyone thought Facebook was overpaying — but suppose they had blocked it. Or suppose Instagram’s founders and investors had insisted on way more money.
Again, I think that’s a better world with more innovation and more competition but probably poorer tech founders. You could say the same about Google’s purchases of YouTube and DoubleClick.
There’s often a kind of culture war between media and tech, where the people raising these competition concerns are “anti-tech” or calling the rich founders names. But I’m really trying to say the opposite here. It’s not that these guys are so rich because they’re bad. But it is true that one reason they’re so rich is that we haven’t had as much innovation and high-growth startups as we ideally should have had. The huge explosion of innovation in household appliances that came after the Great Depression and World War II didn’t create the same kind of singular fortunes, in part because taxes were really high, but also in part because the innovation explosion was genuinely huge, so it was hard for any one guy to be the dishwasher billionaire. We just haven’t had that much tech, and the tech we have had has been limited.
The importance of the real world
The other big factor here is just that software has had a limited ability to influence the real world.
Housing is a huge part of the household budget, and we have important housing scarcity issues in America and larger ones in important parts of Europe. Technology is a key part of the answer to housing scarcity — we need a way to fit more dwellings onto a quasi-fixed pool of high-quality parcels of land. The good news is that the technology to accomplish this — apartment buildings, with elevators if necessary — already exists. The bad news is that it’s generally illegal to use it. You can’t just knock down the brownstones near the B and C lines on the Upper West Side and throw up huge apartment buildings. Nor can you build tall apartment towers near most of the Caltrain stations in the Bay Area or the stations of the expensive and ambitious LA Metro.
That’s a tragedy, and we’d have much higher productivity if we used the best technology available for one of our most important commodities.
But we’d have even higher productivity if we had smart inventors and savvy investors trying harder to innovate in the housebuilding space. The fact that the existing best technology is often illegal to use, however, is a huge disincentive to focus on housing. It’s easy to imagine a world in which American dwellings are larger and cheaper, and therefore more filled with manufactured goods, and therefore a larger share of the population is involved in relatively high-productivity work in construction and manufacturing rather than in low-productivity food service work.
What would we do with fewer food service workers?
Well, one restaurant near my house has responded to hiring challenges by putting QR codes on all the tables. You scan the code, order from an online menu, and the server brings your food. I think most restaurant owners would be reluctant to annoy their customers by abandoning the conventional method of ordering. But if the United States experiences a nice long run of full employment, then more and more business owners will find it worth their while to innovate. Right now, a lot of fast-casual places let you order on your app to pick up, or you can order at the counter. The apps are nice, but to achieve real productivity gains you need to go app-only which, again, I think will only happen when companies face objective economic pressure from an extended run of full employment.
We’ve had generally sluggish labor markets for most of the past 20 years, so a lot of technical work has gone into solving the problem of “how can I utilize an oversupply of cheap labor?” That’s Uber, that’s DoorDash, that’s “the gig economy.” And while it’s better to have Uber than to not have Uber, it necessarily disappoints in productivity terms. With better policy, those technical skills could be put to use solving the problem “how do I cope with a paucity of cheap labor?” and we’d be moving into the utopian future.
A new era of innovation
Long story short, we’re not going to get anywhere by scapegoating successful business founders or by telling ourselves that distributive problems are the only ones that matter.
But the observation that we have a lot of founders and relatively few heirs in the contemporary Rich Guy League Table doesn’t do the kind of work that the proponents of the Silicon Valley View want it to do. The pace of productivity growth and economic dynamism has gone down, not up. The extreme riches of the richest Americans reflect the value of their innovations, but also the paucity of alternative innovators. Some of that is just bad luck, but some is bad policy.
Then a whole other set of bad policies has unduly limited the scope of activities where we’ve allowed the best technology to be applied. And catastrophic failures of monetary policy have created a situation where improving the productivity of big, mass-scale enterprises hasn’t been especially worthwhile.
Out of all these failures, only one — too much willingness to let tech incumbents acquire new startups — really has anything to do with the political agenda that rich Silicon Valley people have pushed. But it’s still a sea of failures for America as a whole. We can do better and we should do better. And we should also be open to the possibility that a more dynamic, more innovative, more competitive economy might birth fewer giant fortunes and actually do quite a lot to lift the net worth of random heirs who joint-own shares in lots of stuff.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.