Inequality: Krugman vs. Stiglitz

by Benjamin Studebaker

There’s an interesting debate going on within Keynesianism at the moment about whether or not the present economic malaise in much of the western world can be accredited to the persistent rise in inequality that has transpired over the last thirty years or so. Arguing in favour of the inequality connection is Joseph Stiglitz; arguing against is Paul Krugman. I’d like to examine what both economists have to say on the topic and deduce as best I can my own view on the subject.

To start off, let’s substantiate the underlying assumption of the debate, that inequality is, in fact, increasing. Using data from the OECD, we can see that if you live in a developed country that is not Turkey, Greece, France, Hungary, or Belgium, inequality has risen where you live:

Gini coefficient is a measure of inequality. Low inequality nations like Sweden usually score around 0.25 while high inequality nations like the United States often approach or exceed 0.4. As you can see in the chart, the red cubes for the 2008 measurement are nearly uniformly higher than the blue bars representing the 1985 measurement (and in some of these countries (the USA and UK in particular), 1985 is considerably more equal than say, 1975). With the central premise of higher inequality established, we can then discuss whether this inequality has negative impacts connected to economic malaise.

Let’s start with Stiglitz. Stiglitz proposes a series of connections between rising inequality and the recent economic malaise:

Lower wages as a percentage of GDP translate into a middle class having to spend in excess of its earnings to support growth, reducing the economy’s ability to sustainably maintain high growth rates without generating higher household debt.
Investment in the next generation at the familial level is crowded out by household debt used to fuel growth (or compounded further into student loan debt).
The rich often pay effectively lower tax rates than the middle class through tax avoidance, so weaker middle class incomes relative to GDP depresses tax income relative to GDP, reducing the perceived budgetary scope for stimulus.
There is a statistical relationship between higher inequality and economic instability in general for these reasons.

Krugman takes on board points 2 and 4, but claims that they are not specific to this crisis and are ongoing problems we have had for some time. In response to point 3, Krugman points out that the tax system in the United States is, despite avoidance, still mildly progressive. I am broadly inclined to agree with Krugman here–while higher earners do try to avoid paying quite a bit of tax, they still, broadly speaking, end up paying more than lower earners:

However, it should be noted that Krugman himself has claimed that inequality has created some specific tax problems, namely with regard to the payroll tax cap. Payroll taxes fund Social Security, but, as currently structured, income beyond the approximately $106,000 cap is not subject to payroll tax. Greater inequality does have the effect of making the financial foundation of Social Security consequently appear much less secure than it is–if the cap were lifted, it is estimated that Social Security would remain solvent until at least 2060.

The crux of the debate centres on Stiglitz’s first point, which amounts to an “underconsumption” hypothesis. Krugman is suspicious of the underconsumption hypothesis because it unites broadly left wing egalitarian goals with Keynesian economic policy and argues that the former favours the latter. Krugman believes that this may be a form of biased, wishful thinking. In answer to it, he offers up data on the savings rate. Krugman’s argument goes like this:

For inequality to be producing underconsumption, there would need to be evidence that the rich, who are receiving a larger portion of the income, save substantially more of that income without consuming it than the poor do.
Consequently, the hallmark of underconsumption would be a high savings rate, representing more of the income diverted to the wealthy being saved rather than spent and corresponding producing an economy that chronically suffers from inadequate demand.
Since the savings rate has in reality been falling, underconsumption is not possible.

The savings rate is indeed doing as Krugman suggests, if we look at the Federal Reserve data:

However, I think Krugman has the wrong conception of the means by which inequality produces under-consumption It is not so much that the rich consume or invest so much less of their income than the poor do, it is that when inequality rises wages are replaced with borrowing. When inequality is rising, instead of raising wages to keep up with economic growth, wages remain stagnant or grow more slowly, actually falling as a share of GDP. In order for demand to support this growth, the additional purchasing must come from somewhere other than wages. Where does it come from? Borrowing, with the result being increased household debt. Who lends the money? The rich, the very people whose incomes have increased. So the rich supply the additional demand, but they do it not by consuming themselves, but by loaning the money to their own workers for their consumption. No one saves more money in this scenario–the rich lends all of their money out, and the poor spend everything they earn in wages and more. Now, it’s true that you don’t have under-consumption before the crisis in this story–the borrowing supplies the money needed to sustain demand, at least for a while. However, eventually a Minsky moment occurs in which the market realises that the aggregate debt level produced by that borrowing is unsustainable, and the result is a private debt crisis (which can evolve into a public debt crisis if states nationalise private debts to protect creditors, as happened in most western countries, though notably not Iceland). The under-consumption is the result, once the borrowing stops, and it is what we are seeing now as, year after year, consumers fail to recover their pre-recession spending habits.

So, to summarise this alternative story:

Wages as a share of GDP stagnate, but the economy continues to grow, with the proceeds of that growth going to the rich.
To sustain growth, the rich lend their money to the poor rather than pay them higher wages. The borrowed money substitutes for higher wages and produces growth.
However, eventually the debts accumulated throughout the economy become unsustainable, and the economy collapses.
As an added bonus, states take on this debt and consequently become debt-burdened themselves and less able to provide for fiscal stimulus.

The hallmarks of my alternative story would be:

Falling wages as a percentage of GDP
Soaring household debt in the run-up to the crisis which is only clearing slowly in the aftermath and consequently continuing to retard growth.
A low savings rate as everyone is either a debtor or a creditor.

As we saw above, the savings rate did indeed fall in the run up to the crisis. What about wage share and household debt? Let’s look:

Sure enough, wage share has been falling, as per the Federal Reserve: