Fiscal policy has no ooomph

Here’s Karl Smith quoting Tyler Cowen, and then commenting:

Tyler Cowen asks

“The economist Scott Sumner stated the case against fiscal policy another way on his blog The Money Illusion. Sumner noted that no one believes fiscal policy (unlike monetary policy) could be used to target a price inflation rate of say 4 per cent a year. The implication is that fiscal policy is not very effective in managing overall demand in an economy, so why should we so trust it as a tool of crisis management?”

The simple answer is that knowing exactly what you are doing is far less important in a crisis than near full employment.  Near full employment, not only does the optimal quantity of stimulus vary highly in percentage terms but it actually changes sign. Sometimes you want negative stimulus to keep the economy from overheating.

Far away from full employment much blunter tools can be used. You may not really know what the stimulus effect of fiscal policy is, but you don’t need to know. In a depression for example, optimal quantity of stimulus is big and positive, so you just do something big. You just fling resources in the general vicinity of the problem.

Obviously, I am attracted to this question because I think it is premised on a falsity of “far thinking”, that important situations call for carefully thought-out actions. This is not actually case.

Sometimes the net return-to-action is positive over the entire relevant range. In those cases smart finely tuned action is worthless and getting the most bang for your buck is actively harmful.

You always maximize simply by going full throttle regardless of the cost. This is because marginal benefit always exceeds marginal cost over your range. Note that this will generally not be the point that maximizes the rate-of-return or “bang for the buck.”

.  .  .

This is the perfect realm for something like fiscal policy. Efficiency is not something that central governments are good at. On the other hand, really-freaking-huge, is something that central governments are good at it.

First let me say that I can see lots of possible objections to my argument, but I am pretty sure that Karl misunderstood my claim.  I wasn’t so much arguing that fiscal policy lacked the precision to deliver 4% inflation, I was claiming that it lacked the ooomph, to use Deirdre McCloskey’s term.  I agree that precision is overrated when in an emergency.

Suppose real GDP growth is 2%.  Then if the fiscal authority wants to get 4% inflation, you need 6% NGDP growth.  And if the fiscal authority is getting no help from the monetary authority, then you’d need to deliver 6% velocity growth.

It is true that an expansionary fiscal policy can boost velocity, but it quickly runs up against a huge problem (which is nicely explained in Mishkin’s text.)  You don’t just need big deficits to raise velocity at 6% per year; you need rapidly rising deficits.  Thus a deficit of say 8% of GDP, maintained year after year after year, will only provide a one-time boost to velocity.  In that case you need a deficit of 16% of GDP the second year, then 24% the third.  (Those numbers are illustrative; I don’t know how large the deficit would have to be.) Relatively quickly you’d run out of money, and then you would get assistance from the monetary authority, as Congress would order them to monetize the debt to prevent bankruptcy.  But without help from the monetary authority, fiscal policy is far too weak to target inflation at 4%/year, or indeed even 2%/year.

In contrast, as long as the central bank is independent, and not forced to monetize the debt, it can target many different inflation rates and pretty much ignore what Congress is doing.  Fiscal authorities can’t do it without help from the central bank, but the monetary authority can do it without help from the Congress.  That’s the way our system works in America.  The Fed determines the track for NGDP, and Congress must learn to live with those constraints.  Zimbabwe is different.

Now of course that doesn’t mean there aren’t arguments for fiscal stimulus at the zero bound, but there’s no point in rehashing that debate here.

PS.  David Beckworth and I advocated NGDP targeting in front of about 70 staffers on Capitol Hill earlier today.  Former Dallas Fed President Bob McTeer moderated.  I’m told the crowd was larger than normal—SRO.  Later there’ll be a link put up on the internet.

PPS.  There’s a recent debate about whether money is a bubble.  It can’t be a bubble, as there is no such thing as bubbles.



24 Responses to “Fiscal policy has no ooomph”

  1. Gravatar of Michael Watts Michael Watts
    23. October 2012 at 19:24

    I’ve actually wondered about this, and please forgive my near-total ignorance of the area…

    Using some sort of conventional understanding of what an “investment bubble” might mean, is it possible to conceive of a recession as an investment bubble in currency? Are there parallels? False analogies?

  2. Gravatar of Benjamin cole Benjamin cole
    23. October 2012 at 23:16

    Great blogging. I have commented on Mcteer’s blog. Again the question is raised—if monetary policy is key, should it not be a part pf our democratic processes?
    The Fed belongs in the Treasury Department? I think so.

  3. Gravatar of Saturos Saturos
    24. October 2012 at 01:14

    Go MM! Woo!

    So you disagree with Paul Krugman’s Hayekian definition of bubbles?

  4. Gravatar of Saturos Saturos
    24. October 2012 at 01:15

    Michael, I think Nick Rowe and Ryan Avent have made similar arguments…

  5. Gravatar of J.V. Dubois J.V. Dubois
    24. October 2012 at 01:27

    On the bubble thing – I more or less agree with Williamson. “Bubbliness” of an asset is in my view a measure how much of a value we ascribe to that asset due to the fact that other people view it as important, and it’s “intrinsic” value in terms of utility such asset provides.

    In this view, real assets always have some intrinsic value relative to other goods stemming from the utility they provide. Of course, there may be bubble in real assets. People may purchase gold because they think that everyone wants to purchase gold. But there are also chip producers that want to purchase gold because it is the best conductor out there, so this is probably a price floor for gold – determined as how good it is in conducting electricity compared to copper for instance.

    The point is, that there may be bubbles that will never “fully” pop. Therefore it is really hard to even talk about bubbles.

    Is there a bubble in modern art? Is there a bubble in investment wines? Is there a bubble in veteran cars? Is there a bubble in iPads? Is there a bubble in social media? Only time will tell.

  6. Gravatar of Saturos Saturos
    24. October 2012 at 01:49

    Tyler posted this:

  7. Gravatar of Jim Crow Jim Crow
    24. October 2012 at 01:58

    “It can’t be a bubble, as there is no such thing as bubbles.” Ah, music to my ears. I hate the term bubble. Probably always will. I do believe that prices can be distorted through stupid policies (ah, Nixon, good times) but if something’s freely traded then it’s worth what people will pay for it. No more, no less. That’s your intrinsic value, right there.

  8. Gravatar of Bill Woolsey Bill Woolsey
    24. October 2012 at 02:52

    While this argument about fiscal policy and velocity is true as far as it goes, if the problem is a downward shift in the growth path of nominal GDP, then a one time increase in velocity is exactly what is needed.

    The monetary authority is providing 4% nominal GDP growth, but it is on a 15% lower path. Velocity just needs to rise about 20% one time.

    The monetary authority, being growth rate targeters, I guess, then go foward with 4% nominal GDP growth.

    Excessive focus on growth rates is a terrible mistake.

  9. Gravatar of Saturos Saturos
    24. October 2012 at 04:04

    I assume people have seen this before (but Tyler just linked to it the other day):

    Seem right to you, Scott? I presume Krugman is therefore now arguing for multiple equilibria – or is it the debt overhang, are we supposed to go on running trillion dollar deficits until that clears up?

    Oh, and there’s another thing.

    Shouldn’t that have been posted here first? Instead, I found it on Tyler’s blog.

  10. Gravatar of Saturos Saturos
    24. October 2012 at 04:25

    Interesting, in the Mercatus paper you once again pretend to believe in “bubbles”.

  11. Gravatar of Saturos Saturos
    24. October 2012 at 04:45

    Typo: “we had been
    in a recession since December 2008.” should be “we had been
    in a recession since December 2007.”

  12. Gravatar of ssumner ssumner
    24. October 2012 at 05:17

    Saturos, OK, Krugman says a bubble is when markets are using expectations that can’t be fulfilled. By that definition housing in 2006 was not a bubble. The market believes that prices would be considerably higher 5 years down the road. And as we know from Australia and Canada, that was a belief that could be fulfilled, but was not fulfilled. But it could have been.

    Tyler keeps using total wage data, when he really needs hourly data.

    Yes, I’ve seen that Krugman piece. When Krugman says “How could people be so stupid . . .” he’s not just talking about Republican economists who don’t like fiscal stimulus, he’s talking about himself circa 1999.

    When I talk about “bubbles” in the Mercatus paper, I’m actually talking about “what our society regards as bubbles.” I mean a huge rise and fall of asset prices. But I don’t blame people for being confused.

    Bill, I focused on 4% inflation for a reason, to show why people never talk about fiscal policy targeting inflation. Yes, fiscal policy can do other things, but that wasn’t the point I was trying to make.

  13. Gravatar of Saturos Saturos
    24. October 2012 at 05:57

    When you say that fiscal policy is a tool for controlling AD, you are of course talking about short-run fiscal policy. Perhaps that should be more explicit.

    “fiscal stimulus, aside from being no less capable of contributing to inflation when allowed to do so than monetary stimulus”

    Oversimplifying, obviously.

    I think you should also link NGDP more clearly to MV = PY. It’s really about MV (aggregate spending/money-flow), but we measure it by looking at PY.

    Didn’t Hayek only support NGDP targeting after the 1930’s?

    Is there an established convention for citing blogs like books (“The Market Monetarist) or are you innovating here?

    I think Andy Harless and Britmouse deserved a shout as well, in the final footnote. And Yglesias and Avent are surely at least honorary Market Monetarists.

  14. Gravatar of Ritwik Ritwik
    24. October 2012 at 08:17

    Say the CB imposes no reserves requirements and announces that there will be no further increases in currency.

    Say Treasury wants to target NGDP of 6%.

    Say it can borrow at 4%.

    Say initial debt/GDP is 100%.

    Now the Treasury runs a primary deficit of 2% every year and pays interest of 100% * 4% = 4% of GDP. So it runs a total deficit of 6% every year.

    In this world, the Treasury delivers 6% NGDP growth at a constant fiscal deficit of 6% and stable debt/GDP of 100%. It’s as simple as that. No currency or reserve increases required. Velocity will increase in same ratio as NGDP.

    The kind of relation between velocity growth and budget deficit that you posit is logically incoherent.

  15. Gravatar of Saturos Saturos
    24. October 2012 at 09:02

    Ritwik, constant fiscal deficit implies constant (but greater) reduction in money demand (income and interest-rate effects), so velocity jumps, but doesn’t keep rising.

  16. Gravatar of Saturos Saturos
    24. October 2012 at 09:30

    Nick Rowe (once again) explains it all correctly:

    Seriously, I don’t care if he isn’t that good at math, the man is probably in many ways the best theoretical macroeconomist alive. (Well, at least in “conventional fields”.)

  17. Gravatar of Ritwik Ritwik
    24. October 2012 at 09:42


    Nothing of the sort. Bank deposits grow in the world I’ve posited. Perhaps by 6% p.a. Perhaps lesser. Perhaps more. Doesn’t matter.

    The observed money velocity, where money is defined in the Scott Sumner way – zero interest bearing reserves + currency – grows in this world. The ‘money’ demand falls, and keeps falling as long as the NGDP is expected to grow. Because money, and its velocity, are both artifacts.

    If David Hume had written down the original equation as CV=PY where C was credit and V was the velocity of credit, the equation would be just as true as it is now. And just as unhelpful in understanding anything.

  18. Gravatar of Saturos Saturos
    24. October 2012 at 09:56

    Ritwik, how will your deficits cause bank deposits to start growing at 6%? And if it isn’t exactly 6% growth, how do you guarantee 6% NGDP growth?

  19. Gravatar of Saturos Saturos
    24. October 2012 at 09:59

    The money multiplier doesn’t rise whenever people want to borrow more money, you know.

  20. Gravatar of Saturos Saturos
    24. October 2012 at 10:01

    A bank that lends my deposits will have to take into account my propensity to spend them, even without reserve requirements. And that requires having reserves on hand. The backing of deposits can’t converge to zero.

  21. Gravatar of Saturos Saturos
    24. October 2012 at 10:02

    CV = PY? The price level depends on the unit of account. Credit is an activity that can occur even without such a thing.

  22. Gravatar of Ritwik Ritwik
    24. October 2012 at 10:37

    1. Deposits are backed by government debt.

    2. Credit is denominated in the unit of account. In fact, the unit of account evolves precisely to keep track of credit.

    3. Bank deposits will rise and fall for the same reason they fall and rise now.

    Anyway, relax. I don’t claim to have solved the entire macroeconomic model. I just showed an equilibrium condition where the debt/GDP is stable, fiscal deficits are a constant % of the GDP and the observed velocity falls and keeps falling.

    Scott’s conception of deficits having to rise and keep on rising is trivially false, based on not thinking through the government budget constraint in a growing economy.

    Neither currency nor reserves is critical to anything, except in that they’re legally required or behavioural legacies. Currency is a historical legacy that lets govt+CB make seigniorage revenues. Required reserves are a bank tax.

  23. Gravatar of Saturos Saturos
    24. October 2012 at 11:29

    Not to turn this into another MF-y saga, but

    1. Money is not really a liability:

    2. Just because credit is denominated in the unit of account, doesn’t mean its circulation accounts for the price level. Apples are denominated in the unit of account, yet the apple stock times the velocity of apples does not equal NGDP.

    3. Sure, but how does that answer my objection

    I thought you “showed” that velocity keeps rising, not falling?

    What happens in an economy without a government?

    I think the main problem with these criticisms from you and others is that you are attached to the notion that macro has to be deep and complex, and that no “trite” explanation could fully account for the sacred mystery of why business cycles really really. There has been controversy for decades, after all. The answer can’t be simple. These “market monetarists” must be kidding themselves.

    I must confess that I on the other hand am pleased at the success of a group of economists whose fundamental ideas square with the intuitions I have had about the economy since high school, and have had such predictive success.

    How would we test your theory of what money is?
    Ours, of course, based on very concrete notions of a well-defined class of entity (base money) moving from agent-holding to agent-holding, is eminently so. Sure, we could be wrong somehow – how would you demonstrate it?

  24. Gravatar of Economics Student Views » ECON430-Topic #4: Fiscal and Monetary Policy Economics Student Views » ECON430-Topic #4: Fiscal and Monetary Policy
    4. April 2015 at 12:45

    […] or not monetary policy even works. Scott Sumner, noted advocate of NGDP targeting, talks of how fiscal policy does not have the capacity to truly affect the economy in the necessary ways, and that the interest on reserve rate can be used in a variety of […]

Leave a Reply