First, the inspiration part.
Organizers say it will almost certainly be the first paper at the prestigious Brookings Papers on Economic Activity that was commissioned based on a blog comment. It is also a rare honor for a graduate student to present a sole-authored paper there; a quick scan of Brookings records shows a similar appearance by the now-renowned economist Jeffrey Sachs when he was a doctoral student in 1979.
“It’s made Matt famous,” said Tyler Cowen, the George Mason University economist who runs the Marginal Revolution blog, and who elevated Rognlie’s comment into a standalone post on his site. “It was brilliantly reasoned and right on target. And very elegant.”
More links here. Even more from Timothy Taylor.
Note that it should inspire high-quality comments, not quantity or snark.
The topic is inequality, which leads to a summary of my talk.
In 1965, the St. Louis Cardinals played their home games in Sportsman’s Park (aka Busch Stadium I). The most expensive seat in the ballpark, a box seat, cost $3.50. A blue-collar worker, who earned about $2 an hour at the time, could treat a family of four to a game in these most expensive seats for less than one day’s pay.
These days, the Cards play at the new stadium, Busch Stadium III. A typical blue-collar worker makes something like $20 an hour The cheapest seat in the stadium still costs less than an hour’s pay. But the most expensive seats cost somewhere north of $800. It would take a month for a blue collar worker to earn enough to treat a family of four to the best seats in the ballpark.
In fact, most seats at the new ballpark are out of reach of blue-collar workers. Why is this? Are the new owners more greedy than Augie Busch, who gave tickets away cheap because he was a nice guy? I think not.
The new owners charge high prices for most seats because nowadays they can. In 1965, the top third and the bottom third of the earnings distribution were not that far apart, so that if you charged prices way above what a blue-collar worker could afford, you would have had mostly empty seats. Today, the top third provide a cadre of highly affluent customers.
In 1965, if you were in the top third and went to a baseball game, chances are that there were people sitting nearby from the bottom third Today, the top third and the bottom third are not sitting in the same part of the ballpark.
I think that the explanation for this comes from the four forces.
1. The New Commanding Heights, which means that over the past 100 years more of the increase in total wealth has been spent on education and health care than on manufactured goods. This trend has become most noticeable in the last thirty years. It means that earnings are no longer split between corporate shareholders and a nearly-homogeneous work force. They are split between high-skilled professionals and low-skilled support staff.
2. Bifurcated marriage patterns. Fifty years ago, one often found a marriage between someone who originated in the top third of the distribution and someone who originated in the bottom third. Since the 1960s, that has become rare. That creates the Coming Apart phenomenon documented by Charles Murray and re-documented by Robert Putnam.
3. Factor-price equalization exacerbates the competitive pressure on low-skilled workers.
4. Moore’s Law means that when computers are able to do a task as well as humans, they soon surpass humans.
Policy interventions to try to stop these four forces or reverse their effects are likely to be futile. The future will be some combination of the Diamond Age scenario (everyone’s basic needs satisfied, with an upper class of Vickys enjoying handmade luxury goods) and a Beyond Therapy scenario, with everyone enhanced by genetic engineering, implants, and drugs.
Fifty years ago, one often found a marriage between someone who originated in the top third of the distribution and someone who originated in the bottom third.
Is this really true? Where is the data? Or was that a marriage between the top third of the distribution was a almost always single (male) earner? So instead of two lawyers marrying, the marriage in 1960 was a lawyer male marrying a potential women lawyer who did not enter the workforce and stayed home. (Or entered for a time period and did not maximize her potential.)
Anyway, even if cross class marrying was more common in 1950 – 1980ish, that could be the historical outlier of the post WW2 period.
Your example may have been very common, but I don’t see how it changes anything in the analysis.
For whatever reason, combining two high income earners and two low income earners increases household income inequality. It may have been an inevitability once women started working their way up to higher income employment, however.
Maybe I’m dumb as to what economists mean by “originate.”
Change it to top 50% and bottom 50%. Assume only men had jobs. Wouldn’t that mean that all people in the top 50% end up with someone in the bottom 50%.
It can’t be this easy, can it?
Something related to 1, 3, and 4 is what Megan McArdle calls “The Great Inversion” of middle class people moving back into gentrifying cities from the suburbs, but that observation is the effect, not the casual force.
The question is why are people moving back into the cities, when real estate in the cities is so much more expensive, (i.e. despite the fact that the “rent is too damn high”), the appeal of city life is not noticeably greater than it was a decade or two ago, and the cost-differential is so high that it practically erases the compensation and lifestyle gains?
The answer seems to be the explosion in manufacturing labor-productivity, that mirrors the similar and earlier explosion in the agricultural sector. In the US, the manufacturing share of jobs has gone from 1/3 after WWII to 8% today. Yet US manufacturing output is at a peak because output per worker has gone up 300% in the last 40 years. (see, e.g.).
All this must be combined with the collapse in shipping costs (you can move an entire container across an ocean for a few hundred bucks these days). If it’s cheaper to move manufacturing 10 miles away from the customers, you might as well move it 10,000 miles away (which is related to factor price equalization (FPE) and international free trade arrangements).
So, what that means is that “tangible production activities” can now only ever occupy only a small percent of the labor force, and everyone else is doing services.
If those services are outsourceable then those jobs will disappear from a high-labor-cost country (FPE again).
So that means most, and an increasing fraction, of the labor force must be engaged in “non-outsourceable services.”
I think you can split “non-sourceable services” into two key groups. There are ones that require mere physical proximity by their very nature: say, surgeons on the one hand, or ‘servant’ jobs that haven’t been fully automated yet (food preparation, maid service, landscaping, chauffeurs and deliverymen, etc.) on the other.
And there are ones that derive a significant benefit from face-to-face coordinations, interactions, and the kind of practice, training, and development that builds organizational capital.
And this last category of activity naturally benefits from network effects, which will tend to encourage people to increasingly centralize and cram together in a decreasing number of increasingly large and dense super-hubs.
And because those places will be the only possible source of the vast majority of lucrative employment, people will have no choice but to cram into them, no matter how high the cost of living (to include taxes and other urban frustrations), which will certainly guarantee that the rent will be “too damn high”.
And that’s what is leading to the Great Inversion, which is really driven by the forces of The Great Centralization.
NB: I can’t take credit for the foregoing analysis, or the name “The Great Centralization”, which is the insight of another online commentator who goes by the alias ‘Vladimir’ or sometimes ‘Vladimir M.”
That’s very interesting, and seems to fit my own experience.
It would also explain why the most extreme levels of wealth inequality are maintained in large, urban, democratic centers.
“Why are people moving back into the cities.”
They aren’t. Megan McArdle is quoting from Alan Ehrenhalt’s “The Great Inversion,” but in his own book he points out that what’s happening isn’t middle-class people moving back into the cities. It’s that the poor are now moving out, too. The result is that central cities are gentrifying, but it’s because of who is left, not who is moving in.
That’s not consistent with what we observe.
If it was all about poor people leaving and ‘survivor composition bias’, raising certain averages, then hub-city populations should be declining, and rents should more stable with all that newly-abandoned supply. Also, you would not observe much new construction or renovation.
But take Washington DC as an example. You can say it’s ‘special’, but ‘government’ is perhaps the ultimate example of non-outsourcable service jobs which benefit from face-to-face coordination and physical proximity to the highest echelons of power. And of course, any service industry (e.g. lawyers) which engages in successful rent-seeking and which would have a lot of non-local competition were it not for laws prohibiting ‘outsourcing’ and setting up other barriers to entry falls in a similar category.
Population within the District boundaries is up 16% from its low about 15 years ago, and real estate prices in all but the crummiest two wards are now back at their bubble peaks. The median rent for a two bedroom is now around $2,800. New building is happening everywhere the city allows it, and at a frantic pace. Not just new condos and apartment buildings, but people turning old townhomes into garish ‘pop-ups’ all over.
That’s not descriptive of a city that’s merely hollowing out of poor people. Those are the signs of new demand. Those are affluent professionals moving in and bidding things up.
And that is what you see in every major hub city these days. To the extent other cities (or even countries) are flailing, as in the rust belt, it is mostly the case that they are sending all their talented young people to the shrinking number of super-hubs, with the demographics and real-estate price trends to prove it.
And here’s another data point (HT: Megan McArdle):
This same phenomenon is playing out in growing and prosperous hubs all over the developed world. Middle-to-Upper-Middle class people are moving into the hubs because those are the only places the majority of them can find lucrative, “non-outsourceable services” employment, which requires that they all get closer to each other to coordinate, build organization capital, and do all the other things that must be done by legally protected locals and/or in close proximity.
The rest of the countryside is for whatever minimal economic ecosystem is required for commodity production, some mass-manufacturing, tourism, but those only employ a small and shrinking portion of the labor force. There is also a lot of what Dr. Kling calls ‘leisure’ (retirement, or idled-dependency) out there, but that’s not relevant to the labor market situation.
The bottom line is that there is no place for most people in developed countries to get middle-class jobs except within commuting distance of an urban hub, and that the centralizing phenomenon is continuing to accelerate in its intensity.
Because real estate is the classic scarce resource for which there is zero-sum competition, these workers have little choice but to allocate all the gains from their employment towards bidding up the price of land close to the center, which, yes, provides a windfall capital gain to the prior owners of that real estate, but erases whatever standard-of-living gains would have been gained by their high-productivity labor, because they’ve got to use so much of their money to pay for shelter, and they can’t escape the rent rat race.
And because these urbanites have to allocate such a large (and growing) fraction of their wages to real estate costs, and for completely different reasons the costs of government, higher education, and health care continue to rise, then they have much less left over for expensive manufactured goods, which is why we see growth in that area of personal consumption expenditures declining.
You use the interesting example of the widening of the gap in best-seat to standard-seat costs as an illustration of (the effects of) widening income inequality. I also thin there is room to analyze it as a way to begin illustrating how this nominal gap may end up being much smaller and relevant than the the polemicists make it out. First, much of the gap goes away when recognizing that a good chunk of the goods and services which define consumption (“wants”) at the 1% level have experienced a massive relative price increase — deflating income trends at the same inflation rate is not realistic. Second, how much of the gap goes away when we you add to each consumer’s (the top and the bottom) basket value that was not available in 1965 — being able to watch an enormous amount of baseball on either free or affordable TV.
Arnold forgot to add, “Have a nice day!”
I’m not sure about the durability of the “Commanding Heights.” Education and healthcare are surely the most sclerotic sectors in our economy, and are most prone to thorough-going disruption. They survive in their current form solely because of government over-regulation.
The unpopularity of the ACA demonstrates the thin ice on which the medical industry rests. And that a leading presidential candidate is running against the academy does not auger well for that institution.
So Mr. Kling is right about the past. But not about the future.
He has also described it essentially as a receding of everything else.
I will also take issue with Mr. Kling’s claim for genetic engineering. It is easy to engineer an organism for a specific trait, e.g., ability to survive weedkillers. But humans are complex mixtures of lots and lots of traits, and engineering that (to any significant degree) will prove impossible. We’re just waaaay too complicated.
Compared to successfully engineering a human, predicting the stockmarket is easy.
I suspect one thing we will do first will be to find populations that resist diseases based on single proteins or gene variants and provide gene therapy to those deficient. So it will start out modestly. He did say it was the future.
We’ll never get beyond that modest start.
Also, genes usually have multiple consequences. They cause disease–but they also prevent mental retardation. It’s best to think of the genome as an ecosystem, with all the parts tightly interconnected.
But ecosystems aren’t tightly connected. You can add in lots of alien species, and lose lots of local species, and hardly anyone would notice a difference. The United States is a good example. It is vastly different than it was in 1492.