Wednesday, January 18, 2012

Once More on Ricardian Equivalence

I can’t let this rest.  Why should anyone bother with this idea?  RE depends on the specification of an intertemporal government budget constraint: any increase in the fiscal deficit in the current period must imply a corresponding decrease in the future, so that the relationship between the present value of the tax and spending streams remains unchanged.  But why?

Here are three arguments against the assumption of such a constraint.

1. The government’s debt/GDP ratio can and does change permanently.  Governments can borrow a lot, run up this ratio and then continue to live at their new level of indebtedness.  There is no reason why the US government cannot continue at the current debt/GDP ratio rather than the one we had on the eve of the financial crisis.  There is also no reason why this ratio cannot be increased to some higher level and maintained in perpetuity.  No doubt there is some level of indebtedness that is not sustainable, but we don’t know what it is, and in particular there is no compelling argument that says we are up against that constraint today.  (Credit markets, for what it’s worth, think US debt sustainability is a non-issue, given that they do not require a risk premium.)

2. The debt/GDP ratio is predictably altered by changes in nominal GDP.  Both real and price level effects matter.  Robust real economic growth can reduce the debt burden with no contribution from fiscal stringency, and if deficit spending increases subsequent growth (both through capacity utilization and hysterisis effects on potential income) a portion of the debt burden is directly offset.  (Models that assume away such effects simply beg the question.)  Just as relevant is the price level impact: inflation can erode the debt burden, and the question of whether such erosion is “optimal” from a representative household point of view (whatever that means—probably nothing) is irrelevant to whether such inflation will actually occur.  Note, incidentally, that the net effect of inflation on private sector wealth depends on whether the country in question runs a persistent current account surplus or deficit.

3. Finally, what we should now know from history is that it is far more likely that governments will default on their debts than pay them off.  Show me an RE model that accounts for even the possibility of default.....I’m waiting.

What does it say about the state of economics that a theory with no apparent connection to reality can spawn untold dissertations and journal articles and even crowd out rational thinking about current policy alternatives?

4 comments:

Ralph Musgrave said...

Agreed. Bill Mitchell had a go at debunking Ricardianism here last year:

http://bilbo.economicoutlook.net/blog/?p=13236

Plus I had a go here:

http://ralphanomics.blogspot.com/2010/10/ricardian-equivalence-is-nonsense.html

Plus I like this blog post on professional economists’ tendency to prefer theory to reality:

http://unlearningeconomics.wordpress.com/2012/01/10/thats-ok-in-practice-but-does-it-work-in-theory/#comment-451

Bruce Webb said...

Also agreed. But you could make your argument even stronger by swapping debt service/GDP ratio for debt/GDP tout court. It surely makes a big macro difference if you are faced with financing debt at 50% of GDP at 10% interest as opposed to say 150% of GDP at 1%. The former position is roughly where some southern members of the EuroZone were a couple of years ago and increases on either side of the equation or worse both puts them in a world of hurt. On the other hand the latter position is roughly where Japan is sitting, with the added advantage that most of that debt is issued in their own currency and held domestically.

In this continuum the U.S. and France are in different ways closer to Japan than say Portugal no matter what S&P might say.And today's successful bond sale by France and continuing ones by the U.S. at 1% rates makes the nominal debt/GDP ratios almost irrelevant. At a minimum secondary to debt service public and private in their respective countries.

And all this BEFORE converting nominal rates on long bonds to real rates. I haven't seen anyone take a look at what this is doing to the top line number for say debt service on US debt held by the public but somewhere behind the scenes 10% of existing ten years are being rolled over from something around 7% to 1% each year this current unnatural state of free money persists.

I only wish I was smart enough to suss out the long term implications of this. But it only takes ten fingers and the back of an envelope to figure out that simple debt/GDP is not capturing the whole story here

Peter Dorman said...

Good point, Bruce. I was writing in response to the standard defense of an intertemporal budget constraint, that debt/GDP sustainability requires a ceiling, and (unwarranted assumption), if sustainability is to be at the current level, the ceiling bites now. This argument seems utterly indefensible to me, which is why I posted.

In the real world, of course, you have to look at interest rates to calculate a (dubiously) sustainable fiscal deficit. I agree that such a deficit fluctuates with interest expense, but putting this into an intertemporal framework requires some assumptions about the probability distribution of future interest rates -- which I would be loath to make.

As for the current period income effects of negative real interest rates on long term treasuries, you would have to consider also the impact of low rates on privately issued debt, whose net ownership structure is more complex. Go for it.

Meanwhile, my question is, why does nobody defend the assumptions on which RE is based against criticisms like mine, while econ heavyweights continue to debate the implications of RE for policy?

Bruce Webb said...

Well I think I still agree. Then again my formal training is in Medieval History and Celtic Mythology, so discount this for that.

It just seems to me that conventional theories about the burden represented by high public debt to GDP ratios unconsciously build in 70s and 80s level nominal and real rates. That is when Bush I was talking "deficits as far as the eye can see" the unspoken assumption was that debt service would eat us up, everyone understanding 22% rates on your credit card making a mockery of minimum payments. Seems to me that neither economists or people at large have not internalized the difference that negative real rates has on sustainability. Now if you don't use that short term free money to grow long term GDP things will go south. Which is why you don't use it to pay out dividends via tax cuts. But the world is willing to pay us to use their money to invest in future productivity and putting on simple debt/GDP blinders on is insane.