David R. Henderson  

Krugman's Evidence for Ricardian Equivalence

A SOPA Analogy... Economists and Influence, A De...

Paul Krugman has been an outspoken critic of Ricardian Equivalence, David Ricardo's idea, revived by Robert Barro in 1974, that, holding government spending constant, when the government cuts taxes and increases deficits, people will save more to pay the inevitably-higher taxes down the road.

But Krugman's latest blog post gives some crude indirect evidence for this. Referencing a McKinsey Global Institute Report, he shows that as U.S. governments have increased debt as a % of GDP (by 19 percentage points), the private sector (households and non-financial corporations) have reduced their debt by virtually the same percent of GDP (18 percentage points). He argues plausibly that financial sector debt should be ignored.

Of course, over that time period, government spending has not been held constant but has spiked up. Still, it seems that a crude test for Ricardian equivalence is whether increases in government debt are offset by declines in private debt. I'm not Mr. Macro, and so I could be missing something subtle or even obvious, but this evidence seems quite striking.


Comments and Sharing

CATEGORIES: Fiscal Policy

COMMENTS (14 to date)
John David Galt writes:

Private sector borrowing does decrease when government borrows more, but the fact has a much more direct explanation than Krugman's: The banks have only so much money available to lend at any given time, and the more of that capital the government grabs for itself, the less is left for the rest of us to use.

Soon there won't be any left over, and we can forget about a recovery then.

david writes:

Well, taking the argument as given: scroll down and consider Spain. Would Barro-Ricardo as a fundamental assumption regarding forward-looking behavior hold for Americans but not Spaniards?

But that aside. First, the crude test as described seems like a test of crowding-out rather than Barro-Ricardo equivalence per se; of course one can derive crowding out from Barro-Ricardo plus some assumptions on capital markets but crowding out is neither necessary nor sufficient.

Second, what Barro-Ricardo (and crowding-out) are about is net debt positions rather than debt positions alone. The graphs are on the latter; that is why Krugman can discount financial debt (because those generate assets for other financial organizations, and presumably financial organizations are good at juggling assets and liabilities whereas non-financial institutions may not be). So you are misreading the graph.

The intuition is a New Keynesian one about debt being "sticky" and troublesome even if the individual is still net solvent, not about real net savings/investment or who owns what share of the national capital stock. You can't see that on the graph.

Daniel Kuehn writes:

There seem to be so many other reasons to deleverage right now, that I think it's probably a little much to call this evidence of Ricardian Equivalence.

Daniel Kuehn writes:

I think the real issue with Ricardian Equivalence isn't the rational response - it's the "equivalence" idea.

Clearly private actors respond to public debt because they have some expectation of taxation in the future. That, I think, is completely uncontroversial. What's controversial is whether that implies that financing methods are irrelevant.

What should also be uncontroversial (and what makes a test like this tricky) is that not only do private actors respond to public decisions - public actors respond to private decisions. We see increasing deficits precisely when the private sector is deleveraging or expected to deleverage.

So what is Krugman showing us? The first or the second uncontroversial behavioral response? Or both? And if the first exists (which, again, I think is uncontroversial), does that mean we should believe Ricardian Equivalence?

Pablodius writes:

Much of judgement and decision making show that individuals don't make sound, rational decisions with perfect information over long time horizons but instead make decisions based upon their most recent exposure to numbers, events and experiences.

Individuals are myopic and hyperbolically discount the future. How would Ricardian Equivalence hold under one or the other or both conditions? If an individual's time horizon is less than when the tax increase will occur, then that individual would not be making a decision to save now to offset the future tax increase. They are not even considering the future tax increase. Alternatively, an individual discounts the future tax increase so much that the NPV will be offset by future wage growth or by future savings that can begin at a later date.

Mike Hammock writes:

My first reaction to graph 6 here was "Ricardian Equivalence". I was surprised that Richard Koo's reaction was total bafflement.

Les Cargill writes:

John David Galt:

"...The banks have only so much money available to
lend at any given time, and the more of that
capital the government grabs for itself, the less
is left for the rest of us to use."

This isn't really true, at least not how the fed is structured. Treasury and Fed can trade without any of that money seeing the light of day.
You have to work harder than this to show Ricardian Equivalence.

Jason writes:

It seems like things went the other way around for this one. Private borrowing decreased and government borrowed increased as a result. If the causality ran in the direction of your hypothesis, private borrowing would fall in response to increased government borrowing.

William Occam writes:

Consumer debt did not go down because people saved more. It went down because they defaulted


Evan writes:

I think that the big thing is that government spending has increased. That money has to go somewhere, and it makes sense that it would decrease private sector debt as well. Where else would it be going?

Second, Ricardian equivalence is meant to apply to consumers as well. Perhaps savvy financial companies are compensating for the impending tax increases (still, I doubt it) but what about average people? I find that unlikely. Savings rates always go up during recessions. I think that the 2003-2006 years show why Mr Ricardo was wrong the best, even though, again, government spending increased.

Hope this helps, and remember to take what I say with a gram of salt. I'm just a high school student.

I'm pretty sure the public-private debt relation is basic accounting identity rather than anything theoretical.

Also, I have to say that the idea people are seriously considering RE is sad indictment on the state of economics. The average person doesn't even know the level of or understand government spending and debt. Even Ricardo didn't believe RE.


david writes:

It's not a graph of net debt, guys (which would have a total equal to the current account surplus). Stop kneejerking to the term "Ricardian equivalence" or "Krugman".

Tom Maguire writes:

William Occam beat me to it - the household debt reduction is due to defaults. From the McKinsey report:

"US household debt has fallen by $584 billion, or a 15 percentage-point reduction relative to disposable income.

Two-thirds of household debt reduction is due to defaults on home loans and consumer debt. With $254 billion of mortgages still in the foreclosure pipeline, the United States could see several more percentage points of household deleveraging in the months and years ahead as the foreclosure process

Jim Glass writes:

Ricardian Equivalence, David Ricardo's idea...

Actually, Ricardo argued against it. It's funny how some things get their names.

Comments for this entry have been closed
Return to top