Zero fiscal multiplier: European version

It’s not just the Fed, check out this article from Business Week:

Short-term rates for borrowing in euros in the forwards market are the cheapest relative to loans in dollars since September. The 50 percent collapse in that spread this month signals investors are betting the European Central Bank will keep its target interest rate at a record low, sacrificing euro strength to prevent deficit cutting by debt-laden economies in the region from stymieing growth.

OK all you Keynesian macro modelers; raise your hand if you included the likely central bank reaction function in your fiscal multiplier estimates.

That’s what I thought.


Tags:

 
 
 

25 Responses to “Zero fiscal multiplier: European version”

  1. Gravatar of StatsGuy StatsGuy
    22. February 2010 at 08:30

    The problem is that the reaction function seems to be asymmetric. When fiscal authorities run deficits, monetary authorities run tight money. When fiscal authorities cut deficits, monetary authorities wait until the economy has shown definitive signs of imploding before they loosen.

    At this very moment, Bernanke is being sworn in for testimony. Someone is going to ask him whether he believes in more monetary stimulus and/or what exactly would trigger the Fed to rapidly change trajectory if more stimulus was needed. He’s going to repeat the Fed’s current path is sustainable and proper, and while the Fed should be flexible they have a strong commitment to clean up their balance sheet, etc.

    Confronted with that, if you were a politician up for re-election in November, and your singular goal was getting re-elected, what would you do if confronted with a backwards-looking monetary authority that was biased against growth whenever confronted with uncertainty?

    [BTW, if we adopted NGDP targeting (with a secure, difficult-to-game mechanism), the fiscal authority could make decisions that took into account a predictable monetary response mechanism.]

  2. Gravatar of Doc Merlin Doc Merlin
    22. February 2010 at 08:44

    Even without monetary policy hampering fiscal policy, I don’t really think fiscal stimulus would work, regardless. Government spending crowds out private spending in the short run, which causes prices to rise. In the long run it damages incentive structures. Supply siders and conservative Keynesians say that high taxes harm the economy. This is true in very the short run, but the real culprit is spending.

  3. Gravatar of ssumner ssumner
    22. February 2010 at 08:48

    Statsguy, You said;

    “[BTW, if we adopted NGDP targeting (with a secure, difficult-to-game mechanism), the fiscal authority could make decisions that took into account a predictable monetary response mechanism.]”

    Yes, they would assume the multiplier is zero, and this would make for more sound decision-making. No more fiscal stimulus, or fiscal anti-stimulus.

    You said;

    “The problem is that the reaction function seems to be asymmetric. When fiscal authorities run deficits, monetary authorities run tight money. When fiscal authorities cut deficits, monetary authorities wait until the economy has shown definitive signs of imploding before they loosen.”

    I find this argument odd, as the article I linked to provides evidence of just the opposite. It argues the ECB is letting the euro fall to offset expected fiscal tightening. In addition, they are also signalling that they will wait longer before raising rates, and of course an expected future loosening of policy is equivilent to a current loosening. Indeed we can see this is the fall of the euro.

    I’m not saying there is no asymmetry at all, but I’d ask you to consider whether there is at least some evidence of a reaction function both ways.

    Regarding Bernanke, I’m not sure monetary policy is too contractionary on average. I think the 2.2% inflation rate of recent decades is reasonable. I am not calling on the Fed to be more dovish, but rather more stable. Keep NGDP growing at a steady rate that will result in them hitting their inflation target on average, just not every year. I really think part of the problem at the Fed is confusion, not just ideology. The Fed certainly wasn’t acting like a bunch of inflation hawks in 2004-07. The median good price was chugging along at a 3% clip, if you saw the Krugman post I linked to recently.

  4. Gravatar of ssumner ssumner
    22. February 2010 at 08:54

    Doc merlin, I agree that fiscal stimulus would be pretty weak in any case, but the other problem is that I don’t know how to define the stance of monetary policy. What would a monetary policy look like that did not try to offset fiscal stimulus? Would it be a steady MB? Or steady short term rates? Neither of those are good indicators of monetary policy, as money can get much tighter even if rates are steady (as we saw in 2008.) Thus the counterfactual issue is surprsingly hard to even define. A steady monetary base is a vastly different policy from a steady short term rate. So the multiplier from one assumption will be very different than from the other. But any assumption is arbitrary. In my view all that matters is what the central bank will actually do—that is the reaction function that must go into any estimate of multipliers.

  5. Gravatar of StatsGuy StatsGuy
    22. February 2010 at 10:54

    ssumner:

    “I find this argument odd, as the article I linked to provides evidence of just the opposite. It argues the ECB is letting the euro fall to offset expected fiscal tightening.”

    Yes, but this is AFTER the economy has shown definitive signs of badness. Pan-EU growth was 0.4% last quarter (if you believe that). Latvia and neighbors are in full blown depression. German exports and imports each dropped ~18% in 2009, only partly recovering from these anemic levels in December. Moreover, we have yet to see how credible the ECB promise to ease in response to fiscal cuts will be, given that we’re witnessing a brutal fight over who will replace Trichet…

    And right now, consider the leader:

    http://www.bloomberg.com/apps/news?pid=20601109&sid=aY0eQB4FPBaI&pos=14

  6. Gravatar of Doc Merlin Doc Merlin
    22. February 2010 at 14:19

    @Scott’s response to me

    Good point.

  7. Gravatar of D. Watson D. Watson
    22. February 2010 at 14:50

    “That’s what I thought.”

    Classy.

  8. Gravatar of JimP JimP
    22. February 2010 at 17:24

    A WSJ article on price level targeting – with no mention of Scott. But at least its a start.

    http://blogs.wsj.com/economics/2010/02/21/an-alternative-to-the-2-inflation-goal/

  9. Gravatar of Kevin Donoghue Kevin Donoghue
    23. February 2010 at 02:32

    “OK all you Keynesian macro modelers; raise your hand if you included the likely central bank reaction function in your fiscal multiplier estimates.”

    Presumably you won’t settle for an LM curve, so what’s “the” likely reaction function supposed to be? A Taylor rule of some sort? Or Galbraith’s version, which takes into account which party is in power?

  10. Gravatar of Scott Sumner Scott Sumner
    23. February 2010 at 05:46

    Statsguy, OK, Those are certainly reasonable views, but would you agree that the article asserts the opposite? The article claims the euro is weakening precisely because of expected ECB easing. It may be wrong, but its argument is identical to mine.

    Does the ECB’s head have a lot of power, or is it simple majority vote?

    D. Martin. I meant that as a (lame) joke. (Not sure if you are serious or sarcastic.)

    JimP, And more importantly no mention of the fact that Bernanke himself recommended the policy for countries in a liquidity trap.

    Kevin, I don’t use IS-LM, so forgive me if this is wrong, but I presume the LM curve is equivalent to no reaction in the fed funds target, but the base moves as necessary to keep fed funds on target. If so, that’s not the one I’d pick, at least on a Woodfordian long run expected sense.

    I’d be happy with assuming a forward-looking Taylor Rule reaction function. In any forward-looking nominal targeting regime, fiscal stimulus will have no effect as the Fed would adjust policy to keep expected growth in the nominal variable on target. Sure, the central bank doesn’t actually do exactly what I say, but it is also true that the central bank isn’t passive, but reacts to nominal changes in the economy, and even expected changes. I don’t think there is much doubt that the weakness in the economy back in March, which showed up in forward-looking indicators like stock prices, commodity prices, and LT T-bond yields, scared the Fed and led to that asset purchase program.

    I’m not trying to say the multiplier is precisely zero in all cases. I am saying:

    1. It’s zero if the Fed is doing its job.
    2. It’s actually closer to zero than many suspect, once you take likely central bank responses into account.

    I’m no Republican, so I won’t argue against Galbraith on that one. It’s possible he is right. I see some cases that seem to support his view. But I’d point out the Fed did a pretty good job under Clinton. And under Carter they were too expansionary early, so their mistake was more complicated than merely tight money in late 1979.

  11. Gravatar of StatsGuy StatsGuy
    23. February 2010 at 08:51

    Yes, I agree with you on what the article says… The fact that the article lags reality by 6 months limits its usefulness, however.

    Central bankers love fiscal austerity, because it makes their job easier – that being keeping inflation low. With sufficient fiscal austerity, they can keep low _nominal_ rates (thereby deflecting criticism that they are using tight policy), and keep inflation low. The economy might implode, but they can claim they are “doing all they can”.

    (Today, btw, German business confidence for February is sucking wind.)

    ECB monetary policy is primarily run by the Exec Board (not the governing council, which just covers transitional stuff). Six members – one is currently German and is expected to step down for Weber.

    However, Weber (Alex, not Max) is loud and a forceful personality. Having Weber take the Presidency is like Plosser taking the Chair in the US, but worse – there are 8 other members in the US, only 5 other members in the ECB.

    Remember, the ECB has ONLY ONE policy objective, and that is monetary stability – keeping inflation at OR BELOW 2%. That is its only goal. Unlike the US Fed, it doesn’t even (notionally) have a second objective. From the EU national govt perspective, then, depending on the ECB to prop up the economy is a lousy bet.

    In their rush to “insulate” agencies like the Fed/ECB from politics, the politicians did a pretty good job at handicapping themselves. (Although, in the US, the Fed still looks out for the banks.)

    Meanwhile we have Ben Bernanke complaining that the fiscal deficits are making it hard for the Fed to hit its 2% inflation target (after Bernanke helped crash the economy and destroy tax revenues for 5-10 years).

    Well cry me a river…

  12. Gravatar of Kevin Donoghue Kevin Donoghue
    23. February 2010 at 12:59

    Scott,

    AFAICR, Brad DeLong tells us that every macro model has an IS-LM representation. That’s a strong claim but maybe it’s true in some sense. However I was merely thinking of the simple textbook case where the stock of money, M, is held constant and the demand for money function, L, is increasing in Y and decreasing in r. So the curve L(Y,r)=M is upward-sloping in (r,Y) space. That is to say, fiscal policy isn’t ineffective in that model unless we’re already at “full” employment (in which case fiscal stimulus is just plain stupid, not that that deterred LBJ or Harold Wilson).

    Very often when you criticise such simple Keynesian models you seem to me to be ignoring the fact that they are designed to analyse situations of involuntary unemployment, with full employment as a limiting case. You’re not the worst of course. John Cochrane has spiralled out of earth orbit into some region where involuntary unemployment is simply inconceivable.

    Anyway, that’s a bit beside the point I wanted to make, which is this: even very elementary Keynesian models take into account the fact that the effects of stimulus can be offset or even nullified by monetary policy. No self-respecting Keynesian macro-modeller ignores that. So your jibe misses its target.

    But New Keynesian models (for which I have no special affection) also highlight another point, which the Old Keynesian models neglect: when the lower bound on nominal interest rates kicks in, fiscal stimulus is a very effective way of easing borrowers’ fears regarding future monetary policy. What better way to assure the market that deflation is not on the cards than by increasing government debt?

    I’m wandering a bit off-topic here, but while I’m in the mood I want to make one more point about fiscal policy. The idea that “public works” can help to pull an economy out of a slump is much, much older than Keynesian economics. It was commonplace in Victorian times. Economists were familiar with the idea and their objection, as the old joke has it, was: “that’s all very well in practice, but it doesn’t work in theory.” With few exceptions, Lionel Robbins being the most notable, even the “classical” economists of the 1930s supported public works as a practical matter. (And Robbins recanted eventually.) What was new in Keynes was that he presented a theory to justify the practical prescription. You may not like his theory – you’re not alone – but the challenge is to create a better one which matches the facts, which stubbornly refuse to conform to classical models, old or new.

  13. Gravatar of Doc Merlin Doc Merlin
    23. February 2010 at 20:44

    @ Kevin Donoghue
    “Anyway, that’s a bit beside the point I wanted to make, which is this: even very elementary Keynesian models take into account the fact that the effects of stimulus can be offset or even nullified by monetary policy. No self-respecting Keynesian macro-modeller ignores that. So your jibe misses its target.”

    But keynesian models treat fiscal and monetary stimulus as exogenous; reality it goes one step further. If the money is very expansionary, government will be drawn to borrow a lot because rates are low. And because the GDP to debt ratio is falling due to the long term debt being inflated away. This leads to government expansion.

    If the government uses fiscal stimulus, it often will result in monetary stimulus being less aggressive.

    My argument is in your model you can’t treat them both as exogenous.

  14. Gravatar of scott sumner scott sumner
    24. February 2010 at 06:15

    Statsguy, Those are all very good points. The ECB target has two great weakness; one is the focus on P rather than NGDP. The other is a lack of symmetry, as you note.

    Kevin, You say,

    “Very often when you criticise such simple Keynesian models you seem to me to be ignoring the fact that they are designed to analyse situations of involuntary unemployment, with full employment as a limiting case. You’re not the worst of course. John Cochrane has spiralled out of earth orbit into some region where involuntary unemployment is simply inconceivable.”

    Then you’ve misread me. I absolutely never ignore the reality of involuntary unemployment. I am very new Keynesian in terms of my view of the upward sloping SRAS and sticky wages and prices. And wages aren’t sticky at the individual level, they are sticky at the aggregate level. So that makes unemployment involuntary at the individual level.

    Cochrane needs to better explain his position. I don’t agree with his views on monetary policy and Keynesian economics, so it is hard for me to understand where he is coming from. His long paper last year was actually much better than Krugman indicated, if you read the entire paper (which Krugman obviously didn’t.) But his short sound bites are confusing, he doesn’t indicate why velocity would be unaffected.

    You said;

    “Anyway, that’s a bit beside the point I wanted to make, which is this: even very elementary Keynesian models take into account the fact that the effects of stimulus can be offset or even nullified by monetary policy. No self-respecting Keynesian macro-modeller ignores that. So your jibe misses its target.”

    I don’t follow this. If you are saying they understand the problem at a theoretical level, then of course I agree. But if you are saying that the multiplier estimates that they publish actually take into account expected Fed reaction, then I disagree. Do these models consider Fed statements about “exit stategies” to be tightening? I doubt it.

    You said;

    “But New Keynesian models (for which I have no special affection) also highlight another point, which the Old Keynesian models neglect: when the lower bound on nominal interest rates kicks in, fiscal stimulus is a very effective way of easing borrowers’ fears regarding future monetary policy. What better way to assure the market that deflation is not on the cards than by increasing government debt?”

    You had me on your side until you said “what better way”. I’ll tell you a cheaper way; rather than running up $800 billion in more debt that must be financed with distortionary taxes or spending cuts in the future, why not just have the Fed announce a higher target for NGDP? A target consistent with what the fiscal policymakers are trying to achieve. And if they won’t, why not appoint a new Fed chairman and fill the two empty seats with Krugman and DeLong? (Yes, I’m half kidding about DeLong, but you get my point.)

    Regarding you last point, I have always argued that Keynesian economics is absolutely the common sense way of looking at the world. If consumers get pssimistic and sit on their wallets, isn’t it obvious the economy will slow down? That’s the paradox of thrift. When I teach it to my students I always discuss things like the fiscal multiplier as a very commonsense way of looking at things. On the other hand the whole excess cash balance/fallacy of composition way of looking at the monetary transmission mechanism is far more counterintuitive than looking at the interest rate mechanism. You are right that all this predates Keynes, as it’s just common sense.

    I still agree with Hicks and Friedman that what is really new about the GT is he made it all true “in theory” as you put it, with the assumption of absolute liquidity preference as the limiting case, and partly true when the LM curve sloped upward. In a sense the exception in the classical model becomes the norm in the Keynesian model, and vice versa. I know a lot of Keynesians point to unemployment rather than absolute liquidity preference, but I still think the classicals had an adequate theory of sticky wages causing unemployment. But they lacked a theory of the liquidity trap.

  15. Gravatar of scott sumner scott sumner
    24. February 2010 at 06:18

    Doc Merlin, That’s a good point about the problem with treating them both as exogenous. Imagine a car with two steering wheels, and where the two drivers don’t agree on where we should be going. Also imagine the steering wheel called “M” is much more powerful than the steering wheel called “G”. Not try to model how changes in G affect the direction of the car. It isn’t easy, and that’s why I have so little confidence in fiscal stimulus.

  16. Gravatar of StatsGuy StatsGuy
    24. February 2010 at 14:00

    Doc:

    “If the money is very expansionary, government will be drawn to borrow a lot because rates are low.”

    It’s not merely long term debt being inflated away, it’s increased tax revenues due to higher employment and economic activity. But it’s not true that this draws govt to borrow a lot – indeed, historical data suggest govt. borrows LESS because revenues are higher and the gap is thus lower. (During the 1994 to 2000 expansion, govt debt declined massively…)

    You might be arguing that continued long periods of expansion do not give govt any incentive to cut waste – and perhaps periodical crises help break logjams and trim embedded programs. This is a long term structural issue, however, and even then is questionable – expenses may increase during downturns even as revenue is dropping.

  17. Gravatar of Doc Merlin Doc Merlin
    24. February 2010 at 17:33

    @statsguy

    “During the 1994 to 2000 expansion, govt debt declined massively”

    94 to 00 didn’t have high inflation.

    I think you misread when I said “money expansion” as “economic expansion”, I meant inflationary rates of money expansion.

  18. Gravatar of Doc Merlin Doc Merlin
    24. February 2010 at 17:34

    @statsguy continued:

    A better example is the seventies where inflation was high.

  19. Gravatar of TGGP TGGP
    24. February 2010 at 23:08

    Via Jerry O’Driscoll, even Keynes didn’t believe in countercyclical fiscal policy.

  20. Gravatar of scott sumner scott sumner
    25. February 2010 at 05:57

    Doc and Statsguy, The government borrows most heavily when money is extremely tight and NGDP is falling.

    TGGP, I agree. But he did favor more public investment in the Great Depression, however he wanted to keep the investment going even after the economy recovered.

  21. Gravatar of Doc Merlin Doc Merlin
    26. February 2010 at 15:06

    “Doc and Statsguy, The government borrows most heavily when money is extremely tight and NGDP is falling.”

    So government borrows the most money when its the absolute hardest time for people trying to borrowing money? That seems extremely pro-cyclical.

  22. Gravatar of scott sumner scott sumner
    27. February 2010 at 06:26

    Doc Merlin, Just because money is tight doesn’t mean it’s hard to borrow. Tight money is usually accompanied by very low interest rates. So tight money often means easy credit.

  23. Gravatar of Doc Merlin Doc Merlin
    2. March 2010 at 20:20

    Hrm, in the finance world, the only definition of “tight money” that I have heard used is “its hard to get loans.” So, I have to ask, what exactly do you mean by “tight money.” Is for you “tight money” a synonym for “NGDP not rising as fast as it should,” or what?

  24. Gravatar of scott sumner scott sumner
    3. March 2010 at 06:24

    Doc, Yes, I mean low NGDP, or lower than desired inflation.

    What if interest rates are 15%, but it is easy to get loans (as in the 1970s). Is money tight?

  25. Gravatar of The semantics of stimulus, and how to win friends « Shewing the fly The semantics of stimulus, and how to win friends « Shewing the fly
    12. November 2011 at 12:15

    [...] The way I personally travelled this road was basically starting from intuitive Keynesianism, to ‘the central bank moves last’  to monetary disequilibrium theory. Alternatively, you could travel the (shorter) road from [...]

Leave a Reply