08/08/2014 09:18 am ET Updated Oct 08, 2014

Inequality and Growth: What I Think We Know and Don't Know

This new S&P study on inequality and growth is creating a bit of buzz, mostly because it's from a firm we associate with financial markets, not inequality analysis.

I found the study to be a worthy collection of the arguments that have been brought to bear on this interesting question of growth and inequality. Though in an earlier study I raised some questions about the claim that high inequality dampens growth, I clearly think there's something there, especially as regards mobility barriers (e.g., underinvestment in low-income children), over-leverage leading to inevitable busts, and the interaction of wealth concentration and money-in-politics.

And for the record, it actually makes sense that financial firms would worry about this. If inequality above a certain point leads to slower growth or for that matter, greater social upheaval, that could "queer the deal" for them too. Also, inequality-driven populism can come from the right as well, as in the drive to shutter the Export-Import bank -- if you don't see market linkages there, recall the sharp decline in Boeing's stock the day after Rep. Cantor, a Republican ally of the bank, lost his primary.

What follows is my brief overview of the basic channels by which economists think inequality hurts growth, with a few thoughts on the evidence in support of each factor:

-Ours is a 70 percent consumer spending economy, and the broad middle class has a higher propensity to spend their income than the rich. So if the rich claim most of the growth, GDP will growth more slowly.

Evidence: This is a solid theory but the evidence is weaker than you'd expect. At least in advanced economies, correlations between rising inequality and weaker consumer spending don't jump out at you.

In part, this relates to something I discuss below: bubble-driven wealth effects boosting consumer spending of middle-class people such that even with stagnant incomes, their spending is up. Also, while wealthy people surely save more than the less well off, they can be pretty robust spenders. Finally, establishing causality is extremely tricky.

Take the current recovery, which has been found to be uniquely imbalanced as the vast majority of the growth in income has flowed to the top of the scale. Yet, while consumer spending grew more slowly in this recovery compared to earlier ones, so has GDP, such that spending as a share of GDP hasn't changed much. But is slower GDP growthdue to slower spending or are they both the victim of austere fiscal policy, the "zero lower bound" problem at the Fed (the interest rate they control is already zero and they can't lower it further), weak investment, the sharp reversal in the wealth effect when the housing bubble burst?

I'd argue the latter, but either way, this simple, logical claim is not easy to substantiate.

-Underinvestment in future productive resources. Our historically high levels of inequality threaten to undermine both parents' and society's ability to invest optimally in the productive capacity of the future workforce. As the S&P report puts it, "the impact of income inequality on future generations of qualified workers is particularly disconcerting."

Evidence: Again, solid theory but not a lot of proof. The S&P report points to much lower college graduation rates among the poor but of course that's always been the case. The question is whether this differential has expanded as inequality has gone up. Moreover, as I show in figure 3 my analysis of this question, there's no obvious deceleration in quality of the labor force over the decades-long period in which inequality has grown.

On the other hand, one can imagine that this effect would be very slow moving, operating with decade-long lags. So to my thinking, especially when you combine the private underinvestment problem with the public one I note below, this is a surely a potentially worrisome development.

-Regressive tax and transfer policies. S&P notes that while our tax and transfer system is progressive at any point in time, meaning post-tax inequality is always lower than pre-tax, it has become less so: "the equalizing effect of transfers and taxes on household income was smaller in 2007 than it had been in 1979."

Evidence: This is a well-known development and it makes sense when you think about both some regressive changes in the tax code (the reduced tax burden at the top of the scale) and the interaction of our aging demographics and the transfer system. Re the latter, Social Security and Medicare represent an increasing share of federal transfers, and while they're progressive, they're less so than most other parts of the transfer system.

But how is this a drag on macroeconomic growth? S&P doesn't say and it's not obvious. That doesn't mean it's not problematic, of course, but growth linkages here have yet to be made.

-Leverage, bubbles, and busts: OTEers know that I think this one is big. The chain of events is simple and empirically plausible: the wedge of inequality funnels growth narrowly to the top, leaving the broad middle to face stagnant incomes. However, in part because so much income has accrued at the top among those with high savings rates, credit is cheap and--more on this important link in the chain below--under-regulated.

This chain has numerous relevant impacts to the question of inequality and growth. As noted, it helps explain the lack of correlation between weak income growth for the majority of households and aggregate consumption, in essence, substituting credit and wealth effects for income growth. It helps to explain recoveries that look OK on the surface, even while most families are treading water. And it explains the shampoo cycle economics that have been so problematic in recent decades ("bubble, bust, repeat"), thus creating a potent link between inequality and growth.

Evidence: I found this paper by Cynamon and Fazzari to be the most systematic and compelling analysis of this chain. Still, while I think they're onto something, it is true that we've had also lots of bubbles and busts when inequality was low. So more work on this potential process is needed.

-Money in politics: Finally, my other top candidate for how inequality hurts growth: the toxic interaction of wealth concentration and our uniquely money-driven politics and policy.

Evidence: The linkages here include dysfunctional politics that hurt growth in obvious and visible ways from government shutdowns, under-regulation of financial markets, threats to default, and underinvestment in public goods, from medical research to transportation to education.

Convincing political science research has demonstrated the influence of money in policy preferences, and unless you're willing to make a case that the political interests of the top few percent happen to be coincident with policies that foster faster growth, this dynamic also explains how inequality can hurt growth. Of course, supply-side tax cutters offer precisely that prescription: enrich the top and the benefits will trickle down. But decades of evidence have found this to be fairly dust. To the contrary, a central point of the inequality debate is that periods of high inequality are defined by the fact that a rising tide can no longer be counted to lift all boats.

So that's the skinny on what I think we know about inequality and growth, though there's of course much more at the Center for Equitable Growth, a fascinating new'ish institute built around these very questions. I'm quite certain there's something to this important connection but I don't think you see it through some of the most expected channels. Which makes it that much more interesting.

This post originally appeared at Jared Bernstein's On The Economy blog.