At the end of a good post on the shift of income shares earned by capital (more) and labor (less) in the US over the past few decades, Krugman writes:
I think we’d better start paying attention to those implications.What implications?
[I]t makes nonsense of just about all the conventional wisdom on reducing inequality. Better education won’t do much to reduce inequality if the big rewards simply go to those with the most assets. Creating an “opportunity society”, or whatever it is the likes of Paul Ryan etc. are selling this week, won’t do much if the most important asset you can have in life is, well, lots of assets inherited from your parents. And so on.
I think our eyes have been averted from the capital/labor dimension of inequality, for several reasons. It didn’t seem crucial back in the 1990s, and not enough people (me included!) have looked up to notice that things have changed. It has echoes of old-fashioned Marxism — which shouldn’t be a reason to ignore facts, but too often is. And it has really uncomfortable implications.As it happens, I've been writing about this for quite some time. It was the focal point of my criticism of Tyler Cowen's "Great Stagnation" hypothesis (e.g. 1, 2, 3, and others), which I said was a "Great Redistribution". Basically the question I'd like to answer is why mean and median incomes have diverged, as pictured in the graph above. A "Great Stagnation" hypothesis seeks only to explain the flattening of median income growth. But we haven't had a Great Stagnation, since mean income growth has continued, at least until the Great Recession.
A Great Redistribution view, on the other hand, says that the structure of the global economy has changed over the past 40 years in ways that benefit (US) capital and hurt most of (US) labor. Specifically, the rise of a low-skill labor force in the former global South has competed away wage gains from low-skill American workers, while the rise of a medium-skill industrialized labor force in places like the NICs has competed away wage gains from medium-skill American workers. Additionally, the rise of mechanized labor (via robotics, which prompted Krugman's post) shifts income from labor to capital. Take a look at this chart:
Wages are converging globally, and since the US had disproportionately high wages this is hurting American labor in relative terms. At the same time, the global market has expanded dramatically. This increases the return to high-skill American labor as well as the owners of capital, who can now sell their production to much larger markets. This is particularly the case for goods and services which are reproducible at essentially zero marginal cost: think intellectual property and entertainment. Since the "high skill labor" and "owners of capital" groups are not mutually exclusive, this shows up in the data as both a) increasing wage inequality, and b) increasing returns to capital.
This is the simplest story in the world... basically just stating comparative advantage, at a mix of sectoral and factoral levels. The fact that it's so novel -- even to someone with a Nobel Prize in international macroeconomics! -- is a point of evidence that our intellectual class is way too focused on explaining everything locally. The Great Redistribution view has plenty of implications for political economy at global and local levels, but it is essentially a rejection of many public choice arguments, which tend to emphasize capture of political institutions by bankers or other oligarchs as the fundamental driving force in recent trends in the American economy.
I'm not sure what Krugman means by "uncomfortable implications". It could mean that the fact that the economy is working the way the way it's supposed to is an inconvenient truth for those who think that our political economy is being wrecked by those who prefer public choice explanations. But I doubt Krugman means that. It could mean that the "Golden Age" of American labor that Krugman loves so much -- the 1950s-1960s -- was a historical anomaly, the result of specific contingent circumstances that are not likely to be replicated ever again (and would be tragic if they were, given that that arose because of two devastating world wars and a Great Depression). But I doubt Krugman means that either. It could mean that the technocratic neoliberal vision is a fraud, and that the politics of distribution is likely to dominate capitalist political economies for the foreseeable future.
In any case, as an example of this Krugman talks about "re-shoring", the process of bringing manufacturing production back to the United States. Krugman suggests that this will have no major effect on employment or the income accruing to labor, because much of this production is done using robots. I think he's right that the direct effects on labor and wages will not be much. The indirect effect could be much higher, however. Why? Because in order to have robots build things, you first have to have factories. Humans have to build those. And you have to have roads to transport the goods. Humans have to build those too. And you have to have shops where the goods can be sold. Humans have to work in those shops. The desire for human labor that is complementary to robot labor can support wage gains for the median worker. That may not be enough to overwhelm the relative redistribution from the median worker to the top 10%, but it can help the absolute numbers.
American labor can benefit in another way: by receiving more non-cash compensation. The trend in the US is to provide more years of subsidized non-work at the beginning and end of life -- longer periods of education, longer retirements as lifespans increase -- and more non-cash benefits -- subsidized health care and education -- in a somewhat egalitarian way. These programs are overwhelmingly funded by the top 10% of wage earners, who are the high-skilled workers and the owners of capital*. To the extent that goods are increasingly created by non-human labor they free up people to do other things, some of which will not be market work. We'll call that "unemployment" or "underemployment" but if we generate sufficient national income to guarantee minimum standards of living at a level that ensures human dignity it will function as quasi-early retirement.
At the same time, quality of life continues to increase rapidly as the marginal cost of entertainment, education, and other goods approaches zero as a result of advances in information technology. This gain is felt by the median member of society as much as the richest person in society, and is more valuable for those with more available time. In terms of maximizing valuable leisure and minimizing alienating labor the typical citizen might be doing better, maybe even much better, than she otherwise would even while the data continue to show that she is doing much worse.
If this is an equilibrium it will have some negative consequences, for sure. Among them will be a reduction in social mobility and an increasingly bitter political economy. But Keynes dreamed of a world in which the gains from capitalism were distributed in a way that allowed people to work less, and some people are still dreaming of it. Marx too: his criticism of capitalism was not just that it generated inequality, but that it created alienation as labor became routinized. Marx didn't care about social mobility... he cared about human dignity. So maybe the left should welcome our new robot overlords (and their capitalist owners) for bringing the vision of Keynes and Marx closer to reality. Instead of slaving away in factories we can all post kittens to Tumblr and write stimulating blog posts. Yeah, maybe it looks like inequality, but it could end up being Utopia.
*The US tax code is already pretty progressive, and is likely to get much more progressive over the coming years, beginning with whatever deal comes out of the fiscal cliff negotiations. At the same time, the US benefit system is one of the least progressive, but I expect this to change over the coming decades for political economy reasons. Ultimately it will be up to the democratic system to manage these structural shifts.