Tuesday, 29 July 2014
Scott Fullwiler attacks monetary / fiscal coordination.
His article is entitled “Drop It: You Can Call for Helicopter Money but Drop the Call for Coordination”.
The word coordination refers to having having government do $X of fiscal stimulus while the central bank (CB) does $X of QE: the net effect being that the public sector / authorities simply print and spend $X net of any changes to tax.
Basically Scott claims there is no point in the latter printing: that is, getting the CB involved has no effect. This is a fundamental point that needs sorting out because not only were there calls for “coordination” when Scott wrote his article a year ago, as he pointed out, but those calls continue. Examples of those making the latter “call” include Positive Money, David Beckworth and Simon Wren-Lewis. Anyway, I’ll run through Scott’s argument, as I understand it, and try to explain where I think the flaw is.
In the paragraph starting “Let’s think for a minute..” Scott argues that when $Y of coordination takes place, the treasury disposes of $Y of debt in the traditional sense of the word (e.g. “Gilts” in the UK or “Treasuries” in the US), but it still has a debt in that $Y of new base money comes into being, and the CB according to Scott pays interest on that (at 0.25% in the US at the time he wrote the article). But in effect that’s treasury debt because CBs remit their profits to their treasury on a regular basis, and if the CB is on the face of it paying interest on reserves, then the reality is, according to Scott, that the TREASURY is paying the interest, so the treasury is still in debt. Therefor, so he argues, coordination (i.e. having the CB print money and buy debt) has no effect.
In his own words (in blue italics): “Now, what happens if instead the Treasury runs a deficit but there is no “coordination” and instead the Treasury issues T-bills? In that case, the Treasury ends up with new debt that it likewise services at essentially or roughly the Fed’s target rate—it’s well known that rates on T-bills essentially arbitrage with the fed funds rate.”
The no interest on reserves scenario.
Now the first and obvious problem there is that in normal circumstances (i.e. prior to the crunch) CBs did NOT PAY interest on reserves. E.g. the Fed started paying interest on reserves in 2008. So in “normal circumstances” Scott’s argument collapses, unless I’ve missed something. That is, contrary to Scott’s claims, there is a big difference between the authorities, 1, borrowing money from the private sector and spending it, and 2, the authorities simply printing new base money and spending it. To illustrate in 2006, the yield on Treasuries was a bit over 4%. Now there’s a big difference between 4%+ and 0%. Moreover, note that the yield on debt with a few months till maturity was much the same as debt with five years to maturity: reason being that banks were prepared to pay good money to acquire reserves, and ownership of Treasuries with a few months till maturity equals the right to extra reserves in a few months time.
The scenario post crunch was different. After the crunch, the economy was awash with base money, thanks to QE and near zero interest rates. In that scenario, the yield on short term debt is near zero, while the yield on longer term debt will be significant because debt holders run the risk of their investment being diminished by inflation (1.66% for 5 year debt, according to the latter link).
In that scenario, it is true to use Scott’s words that “rates on T-bills essentially arbitrage with the fed funds rate.” But that is only true for SHORT TERM debt.
In that scenario, its true that there is no difference between, 1, the treasury borrowing SHORT TERM and 2, the treasury borrowing short term with the CB buying back the relevant debt. After all, the latter short term debt has a yield of near zero and debt with a zero yield is effectively the same as money.
On the other hand if the treasury borrows by issuing LONG TERM debt, there IS A DIFFERENCE between “1” and “2” just above: to repeat, there is significant yield on long term debt even in the current near zero interst rate scenario.
So I’m claiming that the only scenario where Scott’s “coordination is pointless” idea holds is where there is little effective difference between base money and debt ANYWAY.