One of the oddities of the Keynesian model is that although it predicts that saving will fall during recessions, many people assume the model predicts that saving will rise during recessions. If you don’t see this, start with the easiest version, a closed economy Keynesian cross model. The famous Keynesian “paradox of thrift” predicts that an attempt by the public to save more will not cause the equilibrium quantity of saving to rise. Rather income will fall, and then at the lower income level saving will again equal investment. But it doesn’t stop there. More sophisticated Keynesian models show that falling income will tend to reduce investment. Indeed all business cycle models predict investment (and hence saving) will fall during recessions, even as a share of GDP. So when you are in recession both saving and investment tend to be a relatively low share of national income.
The preceding implies that merely looking at saving as a share of GDP is not very useful. Yes, it is lower during recessions, but not because reduced saving causes recessions—indeed the Keynesians would claim something closer to the opposite.
The next question is whether saving by sector is a useful variable to look at. In particular, if one sector has a high saving rate could that sector be causing a recession? That’s not at all clear. If aggregate saving tells us very little, it’s not clear why the components of saving (household, business, government, etc.) would be particularly informative. They might be, but you’d need a model.
Paul Krugman has a suggestive new post:
And another notes to myself post. Below are corporate profits (after tax and inventory valuation adjustment) and nonresidential fixed investment (roughly speaking, business investment), both measured as shares of GDP. These aren’t exactly matched figures, because not all business investment comes from corporations. Still, I think they illustrate an important point. Business investment isn’t actually all that low; you expect it to be relatively weak in a weak economy with excess capacity, but in fact it’s about as high a share of GDP as in the middle Bush years. What’s really out of line with previous experience is the level of corporate profits, which is arguably serving as a kind of sinkhole for purchasing power.
Just when you think Krugman is going to draw some policy implications, he steps back, and lets the facts speak for themselves. And I think that’s a wise move, as it’s not at all clear what the high level of corporate saving actually means. Krugman points out that investment is about normal for this stage of the business cycle, which suggests that “animal spirits” are not the big problem here.
Are there policy implications for corporate saving rates? Maybe, but I doubt there are stabilization policy implications. I doubt that the corporate “sinkhole” is making the recession worse. It seems to me that to draw policy implications you’d have to construct a bizarre Rube Goldberg-like model of stabilization policy:
1. Monetary policy should target NGDP or inflation. There’s your model of AD.
2. If monetary policy fails to target NGDP or inflation, then fiscal policy should try to cause either M or V to move in such a way as to stabilize the economy. For instance, fiscal stimulus might raise interest rates and thus raise V. Or if the Fed is targeting interest rates, it might cause the Fed to raise M.
3. If monetary policymakers fail to do their job, and fiscal policymakers fail to do their job, perhaps some sort of policy directed at corporate saving might affect V or M.
What might that policy be? An excess profits tax? No, that’s contractionary. Relaxing intellectual property laws? That might redistribute wealth from capitalists to workers, and perhaps might decrease the private sector marginal propensity to save. However it might also cause a stock market crash, reducing the Tobin q, and hence corporate investment. More propensity to save, but less animal spirits.
I’d like to see intellectual property rights weakened (although it would reduce stock prices and thus hurt my personal finances.) But not as a stabilization policy.
To conclude, never reason from a sector of GDP, at least if you are trying to explain changes in GDP. First analyze monetary policy. If that policy is dysfunctional, then try to figure out how changes in fiscal policy might change the precise way in which monetary policy is dysfunctional.
And then? There is no “and then.” Even the second part of the preceding suggestion is probably too difficult for economists to model in the real world; trying to go even further (into various types of private saving) would be completely pointless.
Tyler Cowen responds to Krugman’s “sinkhole” claim:
“So corporations are taking a much bigger slice of total income — and are showing little inclination either to redistribute that slice back to investors or to invest it in new equipment, software, etc.. Instead, they’re accumulating piles of cash.”
I am confused by this argument. I would understand it (though not quite accept it) if corporations were stashing currency in the cupboard. Instead, it seems that large corporations invest the money as quickly as possible. It can be put in the bank and then lent out. It can purchase commercial paper, which boosts investment.
Maybe you are less impressed if say Apple buys T-Bills, but still the funds are recirculated quickly to other investors.
I don’t much like either post. Tyler seems to believe that high levels of corporate saving do not cause depressed levels of AD. I have no complaints with that broader claim. But you can’t really make that argument by pointing to the fact that funds saved get recycled into investments. Yes, increases in realized saving lead to increases in realized investment, at the aggregate level. But if and when Krugman does draw some policy implications from the “sinkhole” of corporate saving, he’s way too savvy to ignore the saving/investment identity. He’ll talk about income distribution, propensities to save, and dysfunctional monetary policy. Or at least imply those assumptions.
BTW, The Krugman quotation provided by Cowen does have one peculiarity. It actually doesn’t matter very much whether corporations give money back to investors, except second order effects relating to agency problems. But it was preceded by this paragraph:
So, I’ve had a mild-mannered dispute with Joe Stiglitz over whether individual income inequality is retarding recovery right now; let me say, however, that I think there’s a very good case that the redistribution of income away from labor to corporate profits is very likely a big factor. Here’s corporate profits as a share of GDP:
Krugman shouldn’t use the term “redistribute” in completely different ways in back to back paragraphs. The argument that higher earnings by capital will lower the MPC is far more powerful than the argument that redistribution within the corporate sector has an effect on the MPC. After all, capital gains are (effectively) taxed more lightly than dividends, so investors might actually be better off if corporations waited a while before paying out profits to shareholders.
If Krugman wants to make a case that inequality matters for AD (and please God don’t let him become another Stiglitz) then he’d want to make it on the basis of individual income distribution, not redistribution among capitalists.
I’m not happy with my post–but perhaps after getting comments I’ll see where the real issue here is–it’s all too vague for me to grapple with right now.