Friday, March 30, 2012

Monetarists should be thinking about fiscal policy

"Market monetarists" like Scott Sumner and David Beckworth frequently make the case that countercyclical fiscal policy is unimportant - that we shouldn't be thinking about optimal fiscal policy. They make (at least) two basic arguments:

Argument 1: Fiscal policy doesn't matter for aggregate demand because the Fed will just cancel it out.

Argument 2: The Fed is much more technocratic than Congress, and also quicker to act, so stabilization policy is best left to the Fed.

I've addressed Argument 1 before, so now I want to focus on Argument 2. This is not a new argument, but Scott Sumner states it pretty concisely in this reaction to the new DeLong/Summers fiscal policy paper:
The optimal regime relies on monetary policy to steer the nominal economy, and fiscal policy to fix other problems.  So we are going to defend the model how?  A blueprint for failed states?  For banana republics?  Fair enough, but ask yourself the following question:  In a failed state, which is more incompetent branch of government; the central bank or the legislature? 
Yes, the Fed is bad.  But Congress is downright ugly.  Deep down most economists are technocrats.  They see the central bank as being the best and the brightest, the guys who are above politics, who will “do the right thing.”  And how do economists view our Congress?  The terms ’stupid’ and ‘incompetent’ don’t even come close to describing the disdain.  So are we supposed to change our textbooks in such a way that the fiscal multiplier is no longer zero under an inflation targeting regime (as the new Keynesians had taught us for several decades?)  And on what basis?  Because the Fed might be so incompetent that we need Congress to rescue the economy?  In what world does that policy regime actually work?  If you have a culture that has its act together, such as Sweden or Australia, the central bank will do the right thing.  If not, then all hope is lost.
Now, let's think about this for a second. As Scott Sumner says, in an optimal world the Fed is willing and able to costlessly stabilize aggregate demand. We probably do not live in that optimal world. Let's think about the kind of world in which we actually find ourselves. In the real world, there may be the following limits to Fed stabilization of aggregate demand:

Limit 1: Monetary policy may itself be of limited effectiveness (for example, because of model uncertainty and variable lags that mean that the Fed never really knows exactly what it's doing).

Limit 2: There may be political constraints on Fed policy. See Simon Johnson for a brief discussion of some of these.

Limit 3: There may be unavoidable costs to countercyclical monetary policy. The bigger the AD shock that must be canceled out, the larger the costs may be. Robert Lucas, who asserted that costless disinflations can't exist, certainly believed this.

First, let's ignore the first two problems, and just focus on the third. Suppose that whatever the Fed has to do to stabilize aggregate demand - print money and buy stuff, for example, or convince people that it is willing to do so - has some costs. In that case, the less the Fed has to do, the better for everyone. Now suppose that Congress has a procyclical fiscal policy - in booms Congress cuts taxes (as Reagan and Bush did), and in recessions Congress cuts spending ("austerity"). This will tend to amplify whatever AD shocks the world throws our way. It will raise the costs of Fed stabilization policy. In this case, Market Monetarists would naturally want to think about saying to economists like John Taylor, who in practice support this procyclical fiscal policy:


So that gets Market Monetarists halfway to the Brad DeLong position. "Austerity" is bad, and boom-time tax cuts are bad, even if the Fed is the only one doing the stabilizing.

What about proactive fiscal stabilization policy? Here, we have to bring in Limits 1 and 2 from above. If there are technological or political constraints on Fed effectiveness, then it becomes an open question whether or not our best bet is to still stick with "Fed only," or whether to employ a mix of Fed and Congress. It is not immediately obvious that the answer is "Fed and Congress," since it depends on the way in which they interact. For example, if Fed and Congress act in sync, and are simply limited in the size of the actions they can take, then it's probably a good idea to have them complement each other by enacting fiscal and monetary stimulus at the same time. We probably tried to do something like this in 2009, when the ARRA and QE1 happened at roughly the same time. And this seems to be what DeLong and Summers would like. But if Congress is so slow that its actions are just as likely to cancel out the Fed as to help, then we had better stick with only the Fed, despite its limitations.

For a better and more technical discussion of some of these issues, and for a more technical argument against fiscal policy (by Mankiw & Weinzerl), see this post by Brad DeLong.

Anyway, my point is that it doesn't seem very fruitful for Market Monetarists to ignore fiscal policy issues and simply keep repeating "We should leave it all to the Fed." Fiscal policy is important even if you believe the Fed is best. Market Monetarists should be speaking out against "austerity" and boom-time tax cuts. And the argument that "Fed only" is the optimal stabilization policy is itself highly non-obvious and needs more empirical support, not just intuition or theoretical postulation.


  1. Argument 1: Fiscal policy doesn't matter for aggregate demand because the Fed will just cancel it out.

    Let's ignore questions about whether it would make any sense for them to do so now and instead imagine that congress had acted to boost aggregate demand during the last boom (or that they act to do so during the next boom). With the economy at near full potential, any increase in aggregate demand will only bring inflation, so the Fed would be wise to act to decrease aggregate demand and thereby "cancel out" the action of congress. But it's not exactly cancellation; aggregate demand is unaffected but other things change.

    What would the Fed do, exactly? Contract the money supply so that interest rates go up. This would be a good thing. This would mean next time we're in a recession, we wouldn't be at near zero interest rates, so the Fed will actually have room to act to stabilize the economy. Credit will be harder to come by. Do you think our current problems have anything to do with credit being too hard to come by? Maybe if the flow of credit is a little tighter, Wall Street will put money in the places most likely to create a return, rather than throwing money everywhere they think they can make some amount of money.

    Congress needs to act to increase aggregate demand, not for the short term but for the long term. Monetarists should demand this, because it is required if we want to get rates back to a range where monetary policy is effective at managing the economy. We need to use fiscal policy to discourage savings and investment and encourage spending, so that the Fed can use monetary policy to encourage savings and investment and discourage spending. That will make it possible for them to encourage more spending in a slowdown.

  2. "Suppose that whatever the Fed has to do to stabilize aggregate demand - print money and buy stuff, for example, or convince people that it is willing to do so - has some costs. In that case, the less the Fed has to do, the better for everyone. Now suppose that Congress has a procyclical fiscal policy - in booms Congress cuts taxes (as Reagan and Bush did), and in recessions Congress cuts spending ("austerity")."

    What do you have in mind as that cost? And why wouldn't fiscal policy incur that cost? The cost I think of is inflation, but if inflation is increased by increased AD, then fiscal policy incurs that cost as well. I can't think of any cost extra fed action would have that extra fiscal policy wouldn't have.

    There is, however, a cost that additional fiscal stimulus incurs that additional monetary policy doesn't: adding to government debt. I think if government debt was at like 20% I'd probably agree with you, but with the US's fiscal position the way it is, I think monetary policy has a large advantage.

    People aren't doing procyclical policy just for the hell of it. They're doing because our government debt is crazy big. If we can get closer to budget surplus without pushing the economy into a recession, we should do that. The way to do that is substitute fiscal policy for monetary policy.

    1. I'm not sure if I agree that our debt is too big; interest rates on US Treasuries are still insanely low. The market is saying, "borrow, borrow, borrow."

      As far as higher inflation goes, I'd like to see higher inflation, or at least higher inflation expectations. It would mean less of this overabundance of liquidity. So in this situation, I'm not sure that's a drawback for fiscal policy.

  3. Hellestal7:14 AM

    "the less the Fed has to do..." Right.

    Nearly every single person who discusses Fed action defines doing "little" or doing "nothing" differently, and none of the people who suggest that the Fed do little or do nothing ever seem to consider seriously what precisely "doing little" is supposed to mean. They seem to forget that a change in expectations is, all by its lonesome, a change in monetary policy even if no immediate action is visible to the naked eye. People could say the Fed is doing little to nothing if they hold the monetary base steady, but that's not actually nothing, because it would mean letting interest rates go where they will. An announcement to hold interbank interest rates fixed isn't doing nothing, because it would mean accommodating whatever changes in the monetary base necessary in order to maintain stable interest rates. The Fed could keep its exchange rate to the euro (or a basket of currencies) stable, and define that as doing little to nothing, and then interest rates and the base would vary accordingly in order to hit the target, and so on and so forth. They're always doing something.

    The Fed is simply unable to "do little". There is literally no such thing. The idea doesn't manage to be coherent. Any target they choose comes with a corresponding set of circumstances necessary to achieve that target. No exceptions. If they choose a target, and it so happens that reaching that target means no change in interest rates for three years (or no change in the base, or whatever), that is still the farthest thing in the world from doing little. It's just a different way of doing something, because the choice of the target is itself the actual course of action.

    The best thing the Fed can do, the clearest signal they can send, is to choose a proper target and stick with it. Sumner makes this point again and again, his choice being an NGDP path level target. If they've got a proper target, they can be very busy indeed, even if to the world at large it seems like they're doing nothing but sitting on their duffs. Doing little is never an option.

    Once you start to think of the process in those terms, the proper terms, then the limitations of fiscal policy immediately become more apparent.

  4. costs, what costs? show me a paper quantifying the "costs and benefits." wheres the threshold? how slippery is the slope?? great lets have a debate about the costs.

    The real costs are that "inflation expectations become unmoored." Its a phenomenal testament to the credibility of the Fed that they can get 2% inflation in the face of high UE and an output gap.

    Now as for fiscal policy: I live in MD. DC/VA/MD feeds from the fiscal trough and constantly fights over the funds (for example recently to move 12000 FBI jobs from DC to a new headquarters in MD). There is no such thing as temporary fiscal stimulus because politicians hate to "cut jobs." even a road needs maintenance after its built. sure lets have a road AND a new purple line, why not. How many times have we renewed "temporary" investment tax credits and the AMT... so many times even now for the one that expired in Dec companies largely it will be renewed, eventually.

    And yes, many people think fiscal stimulus will make the Fed retract its "till 2014" guidance (which would effectively tighten). If Bernanke is signalling he can hold it down, he should use a different color smoke.

    the fiscal trough has created a culture of dependency in MD/DC/VA. they would rather pursue federal money and casinos than export-freindly investments in rail, ports, and so on. What are all those contracters in DC/VA/MD going to do when we actually wind down the wars in the middle east? Eisenhower warned about the military industrial complex. bureaucacies do not like to shut themselves down.

    So frankly, even if there is a mutliplier, i dont care. There is no excuse for the Fed not to be doing its job. pointing the finger is not a defense.

    1. Anonymous9:13 AM

      "There is no such thing as temporary fiscal stimulus because politicians hate to "cut jobs."

      Umm, was the ARRA permanent? I was not a huge fan of the way ARRA was put together, but I wasn't confused about it becoming permanent.

      If the bill is written such that the stimulus spending is temporary, it will be temporary unless Congress acts to make it permanent. If it's not temporary, it won't be.

      Lately, it seems that tax cuts are more likely to become permanent dispite any risks to the long-term fiscal position and no matter how weak the stimulus effect of such tax cuts. My example would be the Bush Tax Cuts which were set to expire and which Congressman Ryan seeks to make permanent.

      Seriously, this argument is shopword and not reflective of any deep thinking about the role, characteristics, and limits of fiscal stimulus in a situation where monetary policy is contstrained.

    2. If the bill is written such that the stimulus spending is temporary, it will be temporary unless Congress acts to make it permanent. If it's not temporary, it won't be.

      well, fair enough if things were different they would not be the same. Sure in an ideal world, would fiscal stimulus actually be temporary? Sure, but in an ideal world i am a billionaire sitting on a beach in Aruba with... you get the idea.

      As a big believer in incentives, the INCENTIVE for every congressman is to bring home the bacon (yes: republicans are already talking about bringing bak earmarks in time for the election. APPLAUSE!!!).

      nope just temporary pork here!

      P.S. its not "pork" if its spend on things I want.

    3. ... also just so you understand I am an equal opportunity pork-basher, "tax expenditures" in the name of stimulus are pork too.

  5. I disagree with all three limitations that you assume.

    1. If the Fed faces some informational constraint such as the one you mention, the Government faces the exact same constraint. If monetary policy faces the limit of not knowing exactly why it's going on, fiscal policy faces exactly the same limit. Additionally, Sumner's 95 paper shows that even with information lags, a CB targeting a futures price performs better than if it followed a money supply rule, so you would also need to explain why 1$ of additional money supply is a better if it's financed by gov't borrowing than a $1 of additional money supply which is created automatically by the fed when inflation expectations are bellow target. A gov't following a counter cyclical spending rule in essence follows a tax financed money supply rule, which, crucially, fails to take into account the information that can be inferred from market expectations and can thus outperform futures-prices CB spending only if market expectations are systematically wrong.

    2. There ARE political constrains to the CB, this is why it's better if central bankers operate with no discretion. A CB that is OBLIGED to transact in some futures market if NGDP or CPI are falling bellow target is much less likely to to yield to political pressure than a government or CB with discretion. Sumner's paper covers this too.

    3. There may be such unavoidable costs indeed; but this sounds a lot like the confidence fairy (lots of people believe in it but nobody has ever seen it). I am not aware of any model that predicts any such costs; quite the opposite since monetary policy allows the gov't to earn seigniorage. I am not aware of any historical incident where the CB intervened to stabilize NGDP path and such "unavoidable costs" occurred.

    I think that, presently, the cost-benefit analysis dictates that fiscal policy should be used only in the liquidity trap in addition to monetary policy (so that banks will swap excess reserves for gov't bonds and the government can spend the cash instead). In other instances monetary policy is best for stabilizing the economy. Also, forgive me to say so but I don't think you've even read Sumner's paper; which I did even though I am not an academic, because I thought it might be fun to find out what people's opinions actually are BEFORE attacking them :)

    I thought you were a saltwater guy, one of those that Paul Krugman boasts that understand the models of the other side before criticising them. Well, maybe fresh is the new salt...
    I also

    1. In order, I suppose.

      1) I'm not sure I understand why you think government expansionary fiscal policy would necessarily have the same constraints as monetary policy. In Noah's piece, he only gives informational problems as an example. It's highly possible, in my opinion, that things like infrastructure spending and even transfer payments would not have the same limitations, for example, because of risk-averse human behavior in the financial sector, that the Federal Reserve's monetary policy has.

      2)As I understand it, Noah was making an argument for what "is," not an argument for what "should be." Yes, it's all well and good if the Federal Reserve somehow had a system that shielded it from all political pressures but it's pretty clear that Bernanke has been under pressure during certain times by inflation hawks, and that's just the way it is, that limits the reality of monetary policy in the current Federal Reserve's banking methods.

      3) I can mostly agree with you here; there's no reason to believe in unavoidable costs if we haven't seen evidence for them, just like there's no reason to believe confidence in austerity will lead to growth.

      However, to give Noah the benefit of the doubt on this one, it may be true that the type of real devaluation that the Bernanke Doctrine has created may have a very real long-term negative effect on long-run supply, or wages, or interest rates, (I'm just throwing things out there) in the United States. I'm not very experienced with monetary economics as of this point though.

      - Thomas Cole

    2. 1) You are right in the sense that infrastructure payments could be more productive than monpol but this is irrelevant to stabilization policy and even Cochrane would agree that if, say, building a highway has some kind of value that exceeds costs, we should go ahead and build it regardless of where we are on the business cycle. Noah said something completely different, i.e. that fiscal policy could be used for stabilizing the economy. In that sense, the gov't would be facing exactly the same informational constrains; and Sumner has a pretty good case that a futures targeting CB would be better than a money supply rule when dealing with such problems. And a government that follows a rule that says "spend when U goes up and P goes down" is in essence a following a very discretionary money supply rule, so Noah's case holds only if you assume that market inefficiency is the norm, not the exception. I am not 100% sure on this, but if Noah wants to convince MM guys like Sumner he should explain where they got it wrong.

      2) MM guys also make the case of what "is". Sumner's blog is full of post on how the fed's policy IS too tight and that a CB that follows a futures targeting rule would simply not have the discretion to keep money so tight.

      3) Hysteresis is a real concern regardless of fiscal and monetary policy, I think we can all agree there. As I said, fiscal stabilization should be used in the ZLB, but i can't think of any reason why it would be expansionary in normal times. And in fact, this is also a prediction of IS/LM, at least in its Romer interpretation, or any other kind of dynamic or static model which incorporates a CB that reacts to inflation shocks. I think even Krugman would agree with this, at least that's what his papers on Japan's liquidity trap suggest.

      Oh and by the way, I am not some kind of right wing lunatic, I'm very happy to pay half my income to the Swedish government so everyone can have access to descent healthcare. But I don't think the Swedish government should adjust its spending relative to the business cycle, save for the special case of the ZLB. I generally agree with Noah on politics but this post is simply too lazy to be taken seriously.

    3. 1) My point here is that these information constraints may not result in the same actual limitations for both monetary and fiscal policy, because financial markets behave differently from consumer ones. Even though fiscal policy has the same informational constraints, an equal magnitude of expansionary policy doesn't necessarily have the same effect for both policy tools, even when we're not in the ZLB.

      Using your example from your first post; "you would also need to explain why 1$ of additional money supply is a better if it's financed by gov't borrowing than a $1 of additional money supply which is created automatically by the fed when inflation expectations are bellow target." To me, it seems obvious that the government borrowing that money and just giving it to a random person to expand demand is different from the CB expanding the money supply through the bond market; because the resulting money may have different velocities, for example.

      2) I see your point here and I agree. However, for the purposes of argument; let's assume that the CB has political pressures and there's nothing we can do about it. Isn't that on its own, an argument for fiscal policy?

      3) When I said "it may be true that the type of real devaluation that the Bernanke Doctrine has created may have a very real long-term negative effect," I was referring to the potential negative effects of active monetary policy, not the potential hysteresis effects of inaction, which I think deserves a distinction. If it turns out there's no such thing, I have no reason to disagree with you here.

      I didn't mean to accuse you of being right-wing or anything.

  6. Anonymous2:34 PM

    Automatic Stabilizers

    UI is fiscal policy that is legislated in advance and is countercyclical.
    There is no reason why automatic stabilizers cannot be expanded along the lines of an "Infrastructure Bank" proposed by Adam Samwick. Build and repair infrastructure when labor is cheap and in surplus. Cut back when the labor market is tight. Wiser use of tax dollars. More infrastructure bang for the buck.

    Automatic stabilizers address the timeliness issue.

    Given the impotence of monetary policy at the zero bound, it might be wise to visit expanded automatic stabilizers as an additional tool to combat the high negative social costs of extended unemployment and jobless recoveries.

    -jonny bakho

  7. "Market Monetarists should be speaking out against "austerity" and boom-time tax cuts."

    1. MM call tax cuts fiscal stimulus, they speak out against austerity all the time.

    2. The real crisis value of NGDPLT is that it pisses on booms, and keeps crisis from happening in first place.

    3. The real growth value of NGDPLT is that it shrinks govt. and keeps public employees from getting raises in good times. Any growtn on public side increases artificially and immediately NGDP and crowds out private sector.

    It will make more sense if you view it as a visible cap on annual growth. Under a transparent growth cap, with immediate pissing on booms effects (raising private borrowing rates) - govt shrinks.

  8. Noah,

    As Hellestal above implies, you need to be more precise about what you mean by the Fed's "doing" something and why that would have costs. Scott Sumner argues that expected NGDP is the only coherent criterion for the tightness/ease of monetary policy, so from his point of view, the Fed never has to "do" anything: the optimal policy is always to do nothing (although doing nothing might involve buying or selling assets in an attempt to avoid doing something).

    One possibility is to argue that there is an optimal size for the Fed's balance sheet and there are costs to deviating from that optimal size. (Maybe the optimal size is zero, but the cost of deviating is far offset by the benefit of having some dollars in circulation.) I think that's sort of implicitly Bernanke's view: he would like more fiscal stimulus because he would prefer to have a smaller balance sheet. Particularly when the Fed is buying risky assets this makes sense: there is a risk of taking losses on those assets, which could result in the Fed becoming technically insolvent, which would in turn reduce the Fed's flexibility in responding to future inflation because it would set a minimum on the amount of base money that the Fed would have to leave in circulation.

    Another possibility is that there is some kind of optimal interest rate (not the natural interest rate, which varies over time, but something sort of unconditionally optimal) and that deviations from it are costly. I actually think this is reasonable. When the interest rate is too low, you get asset price instability, because the time discounted value of future cash (or service) flows is highly sensitive to changes in the anticipated growth rate. When the interest rate is too high, well, I don't have a good reason why that's bad, but it pretty much seems to suck. So fiscal policy should aim at keeping the natural interest rate close to the unconditionally optimal interest rate.

  9. "Fiscal policy doesn't matter for aggregate demand because the Fed will just cancel it out."

    Inn the case of massive stimulus in a booming economy, this is what the Fed should do. In the case of where we are right now, this assumes that the Fed is evil and/or incompetant.

    So Sumner assumes evil/incompetancy when he needs it, and not otherwise.

    There are no honest right-wing economists.

  10. I guess the way I look at monetarism is, the Fed basically has one dimension of control on the economy, and the economy is a multidimensional thing. They can print money and buy stuff, or sell stuff and burn money. So if they want to boost aggregate demand, they have a way to do that. If they want to boost interest rates to prevent debt driven bubbles from getting out of hand, they have a way to do that. But they cannot possibly do both.

    Fiscal expansion can do both, just as austerity can bring both interest rates and aggregate demand down. You can also use fiscal policy to increase aggregate demand without raising the deficit if you want -- after all, the rich and the poor to not spend their money at the same rate, and there are plenty of ways to change the tax code to favor spending over saving, like getting rid of the special rate on capital gains. Monetarism seems to me to be a belief that the economy only has one dimension that matters and cannot benefit from moving in the directions monetary policy cannot move it.

    The other thing that bugs me about monetarists is that, in spite of the fact that we've just gone through a major crisis resulting from expansion of credit, judging by this comment thread they not only believe that the best way to boost aggregate demand is to search the mountains and valleys for anyone who could be taking on debt and isn't, and make sure it is possible for them to do so, but they believe that, of all the tricks to have in our bag, should forever remain the only trick in our bag.

  11. Limit 2 has clearly been present during this downturn where we have Republicans going as far as suggesting Bernanke should be tried for treasoni , etc.

    Appreciate you pushing back against the anti-fiscalist line a little. As much as Sumner wants NGDP I sometimes think he'd rather have no NGDP and no fiscal stimulus than NGDP with fiscal stimulus. In other words his strongest passion, seems to be opposition to fiscal stimulus.

  12. I can't understand the real difference between "steering the nominal economy" and "fixing other problems". One can distinguish analytically between nominal and real measures of economic activity. But there is no such thing as the "nominal economy" that can be separated out for the "other things" and managed independently. It's all one integrated job.

    Nor is there a real-world way of institutionally separating the stabilization function of government from public investment and redistributive functions of government. It's all part of the same job.

    The Fed can print money and buy stuff, but it simply can't buy enough of the kinds of things that we really need to buy, and in as productive and effective a way, as can the political branches of US government, where the real spending authority of the public lies. Ultimately, the Fed just engages in the buying and selling of limited classes of financial assets.

    So Sumner thinks Congress does a bad job. Then why doesn't he put his intellectual energies to work showing them how to get active and put our public treasury to work accomplishing real good. We have massive problems of social decay, public underinvestment, incoherent national purpose and unemployment of resources that can't be fixed by Fed money management and expectations setting, and is manifestly beyond the scope of private sector entrepreneurial. It's going to take an activist national government directing serious resources. We're way beyond the point where the only macroeconomic policy we need from the national government is the limited kind of "stabilization" that can be provided by the Fed.

  13. The other thing that bugs me about monetarists is that, in spite of the fact that we've just gone through a major crisis resulting from expansion of credit, judging by this comment thread they not only believe that the best way to boost aggregate demand is to search the mountains and valleys for anyone who could be taking on debt and isn't, and make sure it is possible for them to do so, but they believe that, of all the tricks to have in our bag, should forever remain the only trick in our bag.

    Excellent point Eric L.

    1. Hellestal12:08 PM

      No one in this comment thread has mentioned debt.

    2. Dan Kervick, monetary policy does not need more borrowing for it to work. We have not been making calls for more debt. I thought we had this conversation before.

    3. But you have no plausible transmission mechanism. It's all based on bad theory about rational expectations and the quantity theory that assumes the the spending behavior of the real economy responds to expectations about the monetary base. It seems to me that to the extent that is even remotely true, it's purely a matter of inflation expectations building inflation premiums into asset prices.

      The whole thing is based on a view of real economy activity being the result of technocratic trickle down from the relatively small circle jerk of asset market hustlers and CNBC watchers - the people who actually pay attention to what Ben Bernanke says and even know who he is. It's a completely upside-down view of the drivers of the real economy.

    4. Hellestal, of course no one mentions debt, because the MMers have this weird view of the central bank as having a magical ability to sprinkle money all over the real economy, and refuse to recognize that the CB only does its real work through the credit markets.

    5. Hellestal11:29 PM

      If we prejudicially define "real work" as work through the credit markets, then the central bank only does its real work through the credit markets, by definition. Unfortunately, we could, perhaps, miss something by doing that. What might we miss?

      Well... the velocity of money is not constant. That's simply a fact, one of the facts that sunk old monetarism. Even if the total stock of money were roughly constant, the velocity of money would not at all be stable. Sometimes the average dollar circulates more quickly through the economy, and sometimes less quickly, having nothing to do with the total number of dollars that are out there. It would behoove us to do a couple things: 1) to try to understand what might cause the velocity of money to increase or decrease, independent of any changes in the money supply, and 2) to think about how central bank policy might influence this process.

      This is not at all to deny the importance of credit. But it's only after we've thought about this process in just a bit more depth, which involves money movement without debt, that we should, in my estimation, dismiss the process as magical if it doesn't work.

      1) What might change the velocity of money, independent of the stock of money? (Again, we all realize that changes in the stock of money through expansion of credit is important, but we want to focus here right now on just that one process of money movement, to understand this one process by itself.) It's pretty simple, actually. It's going to relate to money demand, how much money people want to hold at any given time. That, in turn, is going to relate to inflation expectations, as you refer to above. (None of this is even MMism, by the way. It still relates to just plain old Keynesianism. Krugman would say something very similar, just google his name and "credibly promise to be irresponsible")

      Given high inflation expectations (regardless of the stock of money), people are going to want to hold less cash. But as it turns out, the economy as a whole is incapable of holding less cash. The amount of money here is fixed. If I buy something with cash, someone else receives that cash. If they want to get rid of the cash, then someone else receives that cash. Each individual micro actor wants to decrease their cash balances -- because holding cash is a smidge more costly when we have a smidge higher inflation expectations -- but it's utterly impossible for the economy as a whole, macro, to hold less cash, because someone new always ends up with the money. That leads to what people call the "hot potato". Cash is too hot to hold, inflation expectations are a bit higher than before, but the cash isn't just going to disappear out of circulation, so it would absolutely lead to a higher velocity of money as people pass it back and forth.

      And people would pass it back and forth by buying things.

      We have thus a higher velocity of money. You pointed out higher asset prices from higher expected inflation, but that's not the only process going on here. Higher inflation expectations will lead people to buy more things as they try to decrease their cash holdings. That's the first question I asked about. Now it's time for the second.

      2) Can the central bank influence inflation expectations? Answer: Yes. Of course.

      And that's it. That's the whole process. I see nothing magical about it. Every single step of the process is based on entirely uncontroversial economic theory, fully compatible with Keynesian thought. MMism would come in, for example, with the specific mechanism recommended (path level targeting) to influence inflation (or NGDP) expectations, but that is not even necessary to tell the story so far. It is, therefore, "real work" for the central bank to influence inflation expectations.

    6. It seems magical to me. It's all based on the idea that a very significant portion of economic agents in the country have inflation expectations based somehow on what the Fed says or does, and that they respond to the inflation expectations that they do have in the manner you describe - by spending more readily.

      Is there empirical evidence to back these claims up? Or just theoretical models?

    7. Hellestal12:55 AM

      "It seems magical to me."

      Every step on the path is the sturdiest brick to these eyes.

      It's been thoroughly, tirelessly documented over many cases that economic agents respond to expected inflation by attempting to rid themselves of cash in favor of other things. Inflation is a penalty on cash, and people's attempts to reduce their holdings in the shadow of that penalty is a hallmark of countless historical episodes. Just as inevitable is the inability of people as a whole to succeed in reducing their cash balances and avoiding the penalty, because somebody has to end up with the money.

      This is why you can't directly measure any change in people's desire to hold real cash balances. Even as people wish to hold less, they are unable to do so -- the money always goes to someone else's hands.

      We can, however, see the broader effects, and we have seen these broader effects again and again. To repeat, this isn't an observation that exclusively comes from MMers. Keynes himself noticed this process. For an extreme example, there is the joke from Keynes in his "Tract on Monetary Reform" that makes the point comedically talking about the Austrian experience:

      A prudent man at a cafe ordering a bock of beer should order a second bock at the same time, even at the expense of drinking it tepid, lest the price should rise meanwhile.

      You can pull real-world examples from any other period you'd like, and you'll get the same response, as people attempt to move out of cash for safety to avoid taking an inflationary hit. Currently, the best-told example might be the This American Life program on "The Invention of Money", which talks about how Brazil controlled its own inflation. (The transcript is available, too, for quicker reading than listening.)

      I talked to one man who used to have nightmares about money sitting on top of his dresser. Sitting still, just losing value.

      And on and on and on. The shoeleather costs of the 1970s in the US were notable, though less severe than the most memorable historical examples.

      The next step in the process is just as solid. You have already acknowledged the potential effect of higher inflation expectations on asset prices. There is a great chart you can google, "Is it really as simple as don't fight the Fed", which shows asset price responses (S&P 500) to both Fed action and announcement. Yes, even in a liquidity trap, the Fed can influence inflation expectations -- and therefore clearly influence asset prices (as you previously noted, and the chart shows) and also spending in the economy.

      The theory itself is tame uncontroversial stuff, and even better, it is indeed backed by lots of evidence.

  14. Noah, I have been thinking about fiscal policy but in a different way:

  15. "But if Congress is so slow that its actions are just as likely to cancel out the Fed as to help"

    Yeah, but finance crisis and debt recessions last for years, maybe even lost decades. There's a lot more time to act and get projects online than with the old style recessions.

    And in general, with regard to the argument that government is horrible and you can't rely on it to do anything well:

    1) The same people who always say this (Republicans) go out of their way to make it so. It doesn't have to be this way. During the progressive era of most of the 20th century, government became much more professional, ethical, and competent. You can find many governments in the world that are very professional and competent. Not to mention the fact that so much of what government does is so high social NPV and so poorly fitted to the free market (externalities, monopoly, etc.), that even with substantial waste it's still well worth it. Look at government Medicare and compare its efficiency to the private health insurance industry. Medicare beats it hands down in efficiency, and government health care systems in many other advanced countries make our free market system look ridiculously inefficient by comparison.

    2) Even with the shortcomings that government does have, it's very helpful for economists to say what is optimal for government to do, even if there are serious problems getting it to do it. Knowing what's ideal tells us what to push for, and what to support when it is offered, rather than just knee-jerk saying if its government it must be bad and opposed -- yea anarchy!

    1. Your find point is the "all recessions aren't the same" argument. It is a fundamental point here that has just been ignored.

  16. On a general point
    "that whatever the Fed has to do to stabilize aggregate demand - print money and buy stuff, for example, or convince people that it is willing to do so - has some costs."

    (Um WHAT stuff surely matters - and isn't this just another name for fiscal policy financed by printing money). I think the WHAT stuff is meant to be financial assets. But notice something, financial assets are in the hands of the already well to do, and the fed is BRIBING them to swap them for cash (which it is promising to devalue). This has distributional consequences. Then they say - but monetary policy is easily reversable - yeah by BRIBING the same people again. No wonder the rich are in favour of it rather than building hospitals and bridges and such - giving the money directly to people without jobs.

    Just in case people think I'm an old fashioned Keynesian - I actually think what we need is much bigger automatic stabilisers - and international financial reform.

  17. The behavior of politicians reminds me of a personal experience with a troublesome bear in Minnesota. The big bear appeared at my cabin porch one day drawn by aromas of some steaks cooked the night before.