Paul Krugman has generated quite a bit of buzz with two recent blog posts (I, II). In the posts, Krugman notes that labor’s share of income has recently been falling, and that this throws cold water on the usual education story. If the gains of productivity increasingly go to capital, then what matters is not education and skills, but who owns the capital. He sees this rise in capital’s share of income as being the result of automation, which has shifted gains to those who own the robots instead of those who work the lines.

I like where Krugman is going with this, but there is reason to be cautious, at least when it comes to trusting the BLS estimation of labor’s share of income. In 2004, the Cleveland Fed released a paper cautioning people against an exclusive reliance on the BLS in response to growing discussion about labor’s recent dip:

First, the “historic lows” in labor’s share are only observed in some series produced by the Bureau of Labor Statistics. Other measures of labor’s share — for example, for the nonfinancial corporate business sector, or the macroeconomy more broadly — are currently near their historic average.

The other metrics do not appear to be regularly maintained. The authors of the paper put them together themselves, and I do not know if anyone has updated numbers. But if there was divergence then between different metrics, there certainly could be divergence now.

The main takeaway for Krugman still holds even if labor’s declining income share is not to blame. As I have pounded over and over again, education is not the solution to inequality. If it was, the rapid expansion of educational attainment over the last four decades would have caused substantial reductions in inequality, but as we know the reverse is true. Krugman’s tentative theory looks to be that capital’s share is rising, which undercuts the ability to reduce inequality by increasing labor’s productivity. My dominant thought is that education is much more a positional good than people imagine, and that ramping up education — whatever its other merits — is unscaleable as a solution to inequality. These views are compatible of course.

Outside of the education-is-not-the-solution thesis, the other issue raised by Krugman is how exactly do we handle this going forward. It seems like this will cause a problem because our existing institutions put so much weight on income from jobs, but income from jobs will be declining, as will perhaps the amount of work to do. Peter Frase has made a whole craft industry poking fun at these kinds of remarks. The mocking goes something like this: In the coming future, robots will be so productive that we will have masses of goods at very little cost and very little human work to go around?! How terrible!

Frase goes on to point out that the only reason this seems terrible is because of a cultural obsession with income coming from jobs. The left-liberal in particular is utterly obsessed with what Jacob Hacker termed “predistribution.” That is, the left-liberal really wants to put into place institutions that create pre-tax outcomes that are more equal. For some reason, there is something particularly magical and special about institutional forms that redistribute income before the transfer system has a chance to do its part.

As Frase points out, if our future is one of automation-fueled abundance, this obsession with predistribution has to go. Otherwise, we will craft institutions of artificial scarcity and create work that people do not even need to do, all so people can have jobs and capture their income through labor. The problem is that in our society, it’s not having good pay, but having a good-paying job that marks success, dignity, and independence. There is a deeply moral character to that understanding as well, and even most liberals believe in it. That is why you will almost never see a Democrat advocate for redistribution: it’s always about policies they think will change the predistribution. And those things which are redistributive are smothered in rhetoric that tries to obscure their true nature (e.g. how you are just getting what you “paid into” Social Security and Medicare).

Anyhow, this will be a really tough cultural shift to pull off. I hope it happens, but I am not optimistic.