So I spent much of the last couple days working on a research question. I'm catching up on blog reading and I find PGL (Angry Bear) and Arnold Kling (EconLog) discussing Thomas Nugent's column at the NRO. Obviously, I sat up straight in my chair and took notice.
Nugent says,
How does the federal government pay for the damages caused by Katrina? Does anyone asking that question actually know how the government pays for anything? Essentially, the federal government pays for things in just one way — it credits a member bank account. Let’s review the process: The federal government writes a check to a construction company to pay for a bridge. The construction company deposits the check at a bank. When that check clears, the Fed credits the bank’s reserve account at the Fed, and then the bank credits the company’s bank account with “good funds.” Bottom line: Operationally, virtually all of the federal government’s spending per se consists of the Fed crediting an account — that’s all. The federal government doesn’t have any “box of money” that gets “filled” from tax collections and the proceeds from new Treasury securities and then gets “used up” by spending or lending. This is an operational reality. In today’s world of non-convertible currencies, spending is necessarily nothing more than “score keeping.” (If one football team scores a touchdown, and 6 points are added to its score, does anyone ask where the scorekeeper gets the points?)
Oh dear.
Likewise, tax payments simply reduce account balances in the private sector. Nothing “goes” anywhere; the government doesn’t “get anything.” To reinforce this point, if you pay your taxes in actual cash, or buy Treasury securities (government bonds) with actual cash, the Fed shreds the cash. Likewise, if you donate cash to the federal government for Katrina, it shreds it. In fact, if you take a $100 bill and burn it, you’ve donated that $100 to Katrina! Operationally, the entire spending process is not constrained by government “revenue.” Whether or not the government has collected taxes or borrowed is not a factor in the payment process. Any constraints on the process can only be “self imposed.”
Oh dear, again. Mr. Nugent, meet Mr. Ponzi. He has some land he'd like to sell you.
Gotta love Arnold Kling's response:
I'm sorry, but I've read the preceding paragraph several times, and it makes less sense each time.
Whether public or private in nature, spending allocation decisions involve real resources, not just scorekeeping. The Treasury and the Fed are not operationally equivalent. Nugent may be trying to get past the notion of a static budget constraint and asserting the most extreme form of Ricardian equivalence in an intertemporal budget constraint. But even the intertemporal constraint is subject to a transversality condition. At face value, Nugent's statement seems to ignore that. The transversality condition is not "self imposed". Ponzi schemes are simply not allowed. Hyperinflation is not an option.
Oh, I know. It's a rhetorical flourish. I can see that, and you can too. But still.
Continuing with Nugent:
Will private borrowers be crowded out? Impossible. The causation is “loans create deposits,” as taught on day one of every traditional money and banking class. The act of borrowing itself creates exactly that same amount of new liabilities (deposits). The process is “self funding” and circular, as a matter of accounting. The concept of a “pool of savings” that somehow gets “used up” by borrowers is a throwback to the time of fixed exchange rates and gold standards, and has no application in today’s floating-exchange-rate world.
Again, I yield the floor to Arnold Kling:
When I was in grad school, I somehow missed the lecture where they said that government deficits are self-funding in a flexible exchange-rate regime.
The savings curve is more elastic, but not perfectly elastic. There is a difference, and I'd hate to learn that the hard way.
Finally, Nugent says,
The true economic cost of Katrina is the real, physical resources committed to repairing the damage that otherwise could have been used elsewhere to expand productivity or improve overall standards of living. But with today’s excess capacity in everything but energy, there is not going to be much of an opportunity-cost to rebuilding, apart from temporary dislocations of building materials and energy production. In other words, shortages of goods and services due to rebuilding should be temporary and modest.
So the private sector faces real trade-offs (but they aren't bad, so don't worry). But the government faces no real trade-offs; its just scorekeeping.
Sorry, governments face real trade-offs too.
UPDATE: MaxSpeak also took notice.
Brad DeLong also took notice of Arnold's excellent post. Note - Nugent has been playing this bastardized version of Keynesian macroeconomics for a long time.
I didn't see DeLong commenting on the Nugent piece or on Kling's response. I did just come across MaxSpeak, however.
You are right. It was Max and not Brad. So many liberal economist bloggers, so little time!
Nugent is taking a lot of criticism for this in the econoblogosphere, but I think he is making a perfectly legitimate Keynesian argument, and the reason he is wrong is a much more subtle one than his critics are suggesting. He states explicitly that there is slack in the economy, and he assumes implicitly that monetary policy is unresponsive to changes in the amount of slack. While I don’t agree with the latter assumption, I do think it may be defensible. The Fed claims to react to changes in economic conditions, but for the past 15 months their behavior has been amazingly uniform. Perhaps rather than, say, a Taylor rule, they are following a policy like, “Raise the funds rate by 25 bps every month-and-a-half until it reaches 5%, and then leave it there for a year.” If so, and if the Katrina reconstruction is finished by mid-2007, then Nugent’s argument may be valid.
I can't overlook the other stuff just because he could possibly be right about the amount of slack. Sure, I can see a Keynesian argument that because of the slack in the economy the opportunity costs associated with this ramp-up of spending might be lower than they would be at full employment. I wouldn't fully agree with it, but time will tell who is right. He could even argue that monetary policy is unresponsive, or (why not go all the way?) that it's ineffective. But that's just the last part of a long essay. The rest of it is a bunch of nonsense about how fiat money and flexible exchange rates make government finance just about scorekeeping. If you take that part at face value, he has found a perpetual motion machine. I can't overlook that, even if he says something at the end that might have some validity.
Governments face real trade-offs.
knzn - I get your point (even if Arnold Kling wants to bash it). But don't be so quick to defend Nugent and William Mosler (Nugent's mentor). They have a long history of misrepresenting what Keynes said to excuse LONG TERM fiscal stimulus. You understand what Keynes wrote. They don't.
There is a perpetual motion machine. Roosevelt used it in 1933: he devalued the dollar, got the Fed to ease, passed the New Deal, and suddenly all kinds of new stuff was being produced out of thin air. Of course the machine loses efficiency as you move up the Phillips curve. Nugent acknowledges this when says, “Budget deficits are only too large if they usurp the private economy’s need for physical capital and labor, thereby precipitating an inflationary surge.”
Hyperinflation is an option, and Nugent acknowledges it’s a bad one, but you don’t get hyperinflation unless you start to stress the resources of the real economy. As long as you measure everything in dollars, Nugent’s Ponzi scheme works, and if you measure things in real terms, you’re back in Nugent’s last paragraph. It’s not “just” the last part of a long essay; it’s an integral part that discusses the conditions in which the rest of the essay doesn’t apply and asserts that those conditions do not hold. (Only when the real economy is stressed does it become inappropriate to measure things in dollar terms.)
You can make a lot of subtle points about why Nugent’s argument (or mine) isn’t quite right. The Phillips curve has a slope to it; it’s not just a backward L-shape. Inflation (especially hyperinflation) does have real effects, so my point about measuring in dollars isn’t quite right. The fractional reserve nature of the banking system might enable the Fed to follow an unresponsive policy and still require the government to take on real obligations. But these are all minor points. Nugent is quite right to criticize those who believe that purely financial variables are economically meaningful apart from their real effects. That fallacy is quite common and much more consequential than the minor problems with Nugent’s argument.
pgl: I calls 'em like I sees 'em. If you quote me something where they misrepresent Keynes, maybe I’ll agree with you. But it seems to me Keynes would (and in fact did) support long-term fiscal stimulus for a country (like Japan recently, perhaps until a few weeks ago) that has a long-term problem with resource underutilization. It’s not entirely absurd to argue that the US (and more pointedly, the world) is in that condition today. At Angry Bear, I did point out specifically the big substantive problem with Nugent’s argument, but the Fed is not helping my case by doing their imitation of a trained monkey. They ought to at least put stuff in their statements about needing to counteract the effects of fiscal policy.
That the Phillips Curve has a slope to it is not a subtle point. That IS the point.
This is the same go-round that we had once before. We are in agreement on the big picture. In Keynsian (or at least Hicksian) terms, when the LM curve is flat, fiscal stimulus makes sense. When resource utilization is slack, opportunity costs are low and budget deficits are not as inflationary.
But you recognize that when resource utilization is not slack, there are real consequences to all that extra spending.
I have read Nugent's piece several times now, and I get the same impression every time. He is implicitly assuming that resource utilization has enough slack that these considerations (like the slope of the Phillips Curve) are nothing at all to worry about. He says crowding out is "impossible" (his word) because of fiat money and fractional reserve banking. He says the process is "self-funding" and "circular". There's your perpetual motion machine. Far from being a minor point, his whole policy recommendation hinges on arguments like that. He has unhinged the real interest rate from fiscal policy. And rhetorically, he's justifying that statement with recourse to the circular nature of fractional reserve banking rather than properly bringing in the slackness argument at this point. That's bold rhetoric! It's that cavalier approach to the issue that I find most frustrating. I mean he literally says that any constraints on the process are "self-imposed". But you and I know that it's not "self-imposed". That's the problem. If you take this essay at face value, you could be get the wrong idea. Whereas, if you were to write a similar essay, I think you would probably not use words like "self-imposed" or claim that crowding out is impossible because of fractional reserve banking.
What can I say? I'm picky about rhetoric.
You like to mention 1933. Yep. Lots of slack there--a lot more than today. And yet, as I always point out in discussions like this, business investment as a percentage of GNP was not at its lowest in 1933, but in 1943. Somewhere in that decade, we crossed a line and that perpetual motion machine that FDR discovered started to dissipate some heat. Crowding out is not "impossible". Even FDR found that out.
(Of course, the cost/benefit calculation is another matter. The point is that it wasn't a free lunch.)
I guess the substantive question turns on just how much convexity there is in the Phillips curve (and how persistent is inflation, and of course, where we are now on the curve). If the true curve has the traditional econometrician’s form – perfectly linear with inflation coefficients that sum to 1 – then Nugent’s point is not substantively valid, because the Fed’s only real choices are in the timing of growth and the target for inflation. At the other extreme is the infinitely convex Phillips curve, in which case the perpetual motion machine works perfectly as long as you don’t push it too far, and Nugent is right on the money, provided you accept his premise that there is slack in the economy.
The reality, I believe, is somewhere in between: there is no free lunch, but lunch can get very cheap very quickly as you move down the Phillips curve. In that case, it may be reasonable to talk about budget policy and interest rate policy as if they were completely separate decisions. Nugent takes this a little too far by treating the decisions as not just separate but unrelated. (“Arguments that interest rates are going higher because of deficits are misplaced; interest rates are going higher because of Fed policy.” Hello? Didn’t the deficits stimulate the economy, and didn’t the Fed react to the strengthened economy?)
After rereading Nugent’s article in its entirety, I am rather less impressed than I was at first. I do think he had some valid points to make, but they are too diluted with things that cloud the issues. For example, I’ve always thought of crowding out as being ultimately a real phenomenon. He is able to deny its existence by casting it as purely financial. And the reference to fixed exchange rates is a red herring. The only difference under Bretton Woods is that there would be a small gold outflow (not in my beloved Mundell-Fleming model, but probably in real life), but to me a gold outflow is just another bookkeeping entry. It’s not as if the French were going to cart off the gold and make jewelry out of it. Of course, if exchange rates were permanently fixed, the gold outflow would be a real effect, because eventually you start to run out of gold and have to raise interest rates, but in practice exchange rates have never been permanently fixed.
Yep. Now we're on the same page!