» Paradox of Thrift, meet Paradox of Toil

Paradox of Thrift, meet Paradox of Toil

Keynes has been rediscovered in the past 2 years – not that he was ever lost, nor ever ignored – with Krugman, Reich, the Obama spend-to-save-jobs team, and the myriad pundits exploiting his theories to pass various spend-to-save policies. The model, of course, was FDR’s New Deal platform, that won him the election, but probably kept us in the Great Depression for a few years too long, and the effects of which we are still reeling from.

One of the famous New Keynesian points centers on the “Paradox of Thrift”. From Paul Krugman’s blog:

The story behind the paradox of thrift goes like this. Suppose a large group of people decides to save more. You might think that this would necessarily mean a rise in national savings. But if falling consumption causes the economy to fall into a recession, incomes will fall, and so will savings, other things equal. This induced fall in savings can largely or completely offset the initial rise.

For the full explanation, click here.

In an interesting twist, Gauti B. Eggertsson, an economist at the New York Fed, just published a paper that shows that under certain conditions, if everyone decides they want to work more, the economy will actually achieve lower employment in equilibrium. Here’s a quick summary from the paper:

This paper is about a paradox: Under particular assumptions, and in a particular environment, if everyone tries to work more, this in fact, reduces aggregate employment in equilibrium. This is the paradox of toil. It corresponds to a classic fallacy of composition. Just looking at what one person does, holding everything constant, can be misleading once you aggregate everything up. This is true even in a representative agent economy like the one I study in this paper, where all workers look and act exactly the same. The reason for this is general equilibrium: When everybody does something nothing is constant. In models of general equilibrium, it can be highly misleading to build intuition and draw inferences looking at only one individual in partial equilibrium.

…If everybody tries to work more in response to the intertemporal disturbance, this puts downward pressure on current and future wages. What happens? Firms cut their prices today and in the future and stand ready to supply whatever is demanded at those prices. Then what? This leads to expectations of deflation, which increases the real interest rate — the difference
between the nominal interest rate and expected inflation — and the central bank can’t offset this by cutting the nominal rate due to the zero bound. Higher real interest rates lead to lower demand because people prefer to spend in the future rather than today, since prices are expected to be lower in the future (and thus the return on savings higher). Because of lower spending today, firms demand less labor. Thus more labor supply, leads to lower wages, more deflation, higher real rates, leading to less spending, leading to less hiring of workers. Therein lays the paradox.

For the full paper, click here.

Why is this interesting to us now? Well, for starters, those “certain conditions” are a short-term nominal interest rate at the zero bound and declining output and prices. The Economist summarizes the implications in the following way:

An implication of this is that policies designed to enhance aggregate supply (of which an increased supply of labour, perhaps through cuts in marginal tax rates, is an example) will not work in the short run if the problem is deficient aggregate demand. Essentially, at the zero bound with falling prices and output, labour demand increases with real wages. But Eggertson shows that while policies aimed at stimulating supply suffer from some version of this problem, this is not true for those—like government spending, temporary sales tax cuts, or investment credits—that aim to bolster demand.

One thing that prevents this model’s conclusions from being literally true is that the strength of its results depend on labour markets being perfectly flexible. If wages are not perfectly flexible downwards, as most New Keynesian models assume, then the increase in labour supply does not exert as strong a downward pressure on prices (which is part of the feedback loop that makes the model work). More generally, though, he makes the point that the effectiveness of fiscal measures at the zero bound may differ dramatically from their effectiveness at positive interest rates, which is the environment most empirical studies are forced to use. No wonder there is so much disagreement about the fiscal multiplier!

http://www.economist.com/blogs/freeexchange/2010/02/paradox_toil

For me the implications are tied to everything else we’ve been discussing – always, but especially at 0% interest rates, the government is limited and often policies will have the opposite of their intended effects.

Last 5 posts by Yaron Sadan

GS catching up on our conversations on demographics - August 12th, 2010
Europe revisited - August 12th, 2010
Starting the day off - August 11th, 2010
Fed day: Who cares? - August 10th, 2010
From the Fed: Future Recession Risks - August 9th, 2010