A world gone mad

Last year I talked about a strange asymmetry in discussions of fiscal and monetary stimulus.  Fiscal stimulus discussions revolved around the question of whether deficit spending could boost real GDP.  Why not NGDP?  I have no idea.  Monetary policy discussions often revolved around the issue of inflation.  Why not NGDP?  I have no idea.  Both types of stimulus are supposed to impact AD, so I don’t see why they are discussed in such different ways.  I recently came across some more examples, which are even more perplexing. 

Last week I did a post on David Frum.  To briefly summarize, he noted that the economy seemed to need stimulus, but he was skeptical of Krugman’s proposal for more fiscal stimulus.  On the other hand, he couldn’t think of any alternative ideas.  Just a week later, he had a sudden brainstorm:

I have a suggestion. I think we should overcome our inhibitions and rediscover the positive side of inflation.

Just so people don’t think I am making fun of Frum, I think he is right.  Three to four percent inflation next year would be good.  (Although I’d prefer slightly lower inflation in the longer run, and also NGDP targeting rather than inflation.) 

But here is what puzzles me.  A week earlier he seemed unable to think of any way to generate more AD, which means, ipso facto, he couldn’t think of any way to generate more inflation.  Then a week later he has a brainstorm; let’s have more inflation!   It would be like BP engineers agonizing for weeks over how to cap the well; and then some pipsqueak piping up suddenly “I have an idea, let’s cap the well!”

My hunch initial hunch was that the average non-economist simply assumes that governments can always generating inflation if they so wish, but boosting AD in a severely depressed economy full of pessimistic people is really hard to do.  They don’t see those two outcomes as being two sides of the same coin. 

But now I’m not sure we can dismiss this problem so easily.  Is it just non-economists who compartmentalize this way?  Is it just non-economists who view fiscal stimulus as affecting RGDP, and monetary stimulus as affecting inflation?  Consider the following from a Reason interview with Anna Schwartz:

Schwartz sounded alarmed, though, at the zealousness with which Bernanke has put “monetary expansion” into practice. She berated the Fed for going too far and predicted that it will have to raise interest rates “in the near future” to arrest inflation. Asked if she sees hyperinflation on the horizon, she exclaimed, “Oh, yes!”

But Schwartz also seems to have undergone a late-life conversion to Keynesian theory. Asked to offer a solution to the crisis, she repeated the ultimate Keynesian maxim: the government should pick up slackening demand in the private sector.

“People are saving, not spending. In order to revive this economy…” she paused, hesitating on the thought, “the government will have to resume spending. By spending, the government will require that the current inventory will be depleted and have to be replenished. And that will bring on additional production and jobs.”

At first I couldn’t make heads or tails of this.  She thinks money is way too easy, which implies AD is way too strong.  But she also thinks we need more fiscal stimulus, because AD is too weak.  But suppose she is approaching this issue in the same way as David Frum.  Money is about inflation and deflation, fiscal stimulus is about real growth and recessions.  They are not two ways of influencing AD, rather they influence two very different parts of the economy.

At this point you might ask whether these views are atypical.  Surely the average economist doesn’t look at things this way.  I’m not so sure about that.  Language is very revealing, and discussions of monetary stimulus focus on inflation, whereas fiscal stimulus discussions focus on real growth—even when it is economists discussing these issues.  This morning a commenter named JimP, gave me an interesting WSJ article which contained this absolutely dismaying survey finding:


This makes me more certain that economists are to blame for the crash of 2008.  Economists apparently support the Fed’s tight money policy.  Even more mind-boggling is the fact that surveys showed most economists favored fiscal stimulus in 2009.  So apparently most economists favored fiscal stimulus and now oppose more monetary stimulus, just like Anna Schwartz.

And remember that most economists are liberals.  They vote 3-1 Democratic.  Admittedly this sample may be slightly different, but I am pretty sure that the group of 44 who oppose more monetary stimulus include at least some economists who vote Democratic.  As I get older I more and more realize that my views on macro are completely different from most of my colleagues.  How could someone favor fiscal stimulus when unemployment was 7.8%, and then turn around and oppose monetary stimulus (which doesn’t increase deficits) when unemployment is 9.5% and inflation is falling below 1%?  What do most economists think the Fed should be trying to maximize?  What is the policy goal?

Obviously the most likely interpretation is that I am the crazy one and the overwhelming majority of economists are right.  Monetary and fiscal policy don’t both affect AD.  Money affects inflation and fiscal stimulus affects real growth.  Perhaps I “misremembered” what I learned in macro 101.  Indeed at this point in my life only Arnold Kling keeps me from doubting my sanity.  Here is what Kling has to say:

Anyway, read Scott Sumner’s whole piece. He keeps trying to stick up for mainstream economics, in a world increasingly gone mad.

After 9/11 lots of people (including me) overestimated the risk of terrorism and temporarily lost their bearings.  Did September 2008 do the same thing to the economics profession?

HT:  John and JimP

PS.  John Salvatier sent me this excellent Tim Duy post on the internal dynamics of Fed decision-making:

Salmon believes that Federal Reserve Chairman Ben Bernanke – a Republican – will not break from that party’s consensus that too much has been done already.  Some of Bernanke’s defenders may find that paints a too narrow view of his motivations.  But Salmon also notes that even Democrats are not eager for additional policy action.  If the White House is not willing to push for more, why should Bernanke, especially when it will apparently require him to expend political capital internally?

The fault lies not in the stars . . .



37 Responses to “A world gone mad”

  1. Gravatar of Benjamin Cole Benjamin Cole
    14. July 2010 at 09:25

    Another excellent post. I don’t know how Sumner posts so many shrewdly insightful blogs.

    I think Anna Schwartz is reading newspapers from the Weimar Republic and guzzling absinthe.


    With unit-labor costs going down? With property values going down? With commodities soft? With the CPI sinking? With an overly cautious Ben Bernanke in charge? With the Dow below where it was in 1999?

    I can see why laymen have given up on economists. Some economists say we will do a Japan, and others predict hyperinflation. How long would you go to doctors if one said you were fat and flabby, and the next said you had to eat more and put on some weight?

    Malcolm Gladwell has written some insightful books on the way people view the world. I think most economists are repeating shibboleths and nostrums, dressed up as analysis. The way Supreme Court judges have biases, and then carefully glean through the law to find a principled-sounding platform to support their biases.

    Of course the Fed should be activist now–this should be their hour. Instead, we get the Fed moment of cower.

    BTW, a little tongue in cheek, but I suggest we run a national lottery, that pays out $10 billion for every $1 billion it takes in. Yup, through ex nihilo. My second plan is to secretly goose payouts at racetracks the same way. This avoids moral hazard, and stimulates the economy–and would help the struggling horeracing industry.

  2. Gravatar of Indy Indy
    14. July 2010 at 09:26

    There’s another way of looking at the 9/11 analogy.

    Sept. 2008 was maybe the 1993 WTC car bombing. Since it wasn’t totally catastrophic, somewhat amateurish, and a few perpetrators were caught, the response was minimal – a few cruise missiles. There was no dramatic shift in thinking or feeling of a need to completely reexamine the scale of the threat or the orthodoxy of the minimal tool-kit of potential reactions.

    The threat remained and grew, the real danger increased, but in general policymakers from both parties were pretty self-satisfied that they had done all that was appropriate and necessary. No one was thinking about a successful next attempt.

    But then 9/11 happened and everything gets thrown out the window and people start reasoning from first principles again in an atmosphere of unlimited possibilities – where previously out-of-the-mainstream ideas once thought crazy and morally dubious become reasonable, feasible and attractive. The risk of not doing more is now perceived as greater than the risk of the more going horribly wrong.

    Strange as it seems to say it – this recession and crisis just hasn’t been bad enough yet to spook the Fed and the Economics profession into true upheaval, a crucible where old, false ideas are vaporized and new ones are forged and embraced by the formerly skeptical.

    The Fed currently seems almost satisfied with how things are going, how they saved the world, how much worse things could have been, that 10% unemployment and an absence of deflation “Ain’t that bad really, if you think about it!”.

    And if recovery comes, even slowly and choppily, they will conclude that they should feel perfectly content with their response and the ideas that framed them.

    It will therefore take the trauma of a 9/11 to change their minds and behavior. Maybe a severe global double-dip recession with simultaneous crisis in Europe and China, with another 25% drop in house prices and the stock market, and another 5% spike in unemployment and a Debt-to-GDP ratio already above 100%.

    The Fed and the Economics profession would emerge as a totally different animal. The world would stop being mad, but the price would be enormous.

  3. Gravatar of Benjamin Cole Benjamin Cole
    14. July 2010 at 09:26

    Ahem, that last sentence sould read “the struggling horseracing industry.” Oh, hah ha.

  4. Gravatar of marcus nunes marcus nunes
    14. July 2010 at 09:58

    They haven´t gone mad, just pursuing their self-interest.
    CRomer in 1992:”This paper examines the role of aggregate demand stimulus in ending the Great Depression. A simple calculation indicates that nearly all of the observed recovery of the U.S. economy prior to 1942 was due to monetary expansion. Huge gold inflows in the mid- and late-1930s swelled the U.S. money stock and appear to have stimulated the economy by lowering real interest rates and encouraging investment spending and purchases of durable goods. The finding that monetary developments were crucial to the recovery implies that self-correction played little role in the growth of real output between 1933 and 1942″.
    Full article: http://www.nber.org/papers/w3829

    Now it´s PC to say the opposite:http://blogs.wsj.com/economics/2010/07/14/obama-advisers-says-stimulus-saved-or-created-three-million-jobs/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+wsj%2Feconomics%2Ffeed+%28WSJ.com%3A+Real+Time+Economics+Blog%29

  5. Gravatar of Luis H Arroyo Luis H Arroyo
    14. July 2010 at 11:23

    Surely the anti money expansion don´t take into account that fiscal expansion is more inflationist in the long run than monetary one. And it is much more contractive to revert the fiscal debt than excesive money inflation.

  6. Gravatar of marcus nunes marcus nunes
    14. July 2010 at 12:17

    From the FOMC “circus”:
    [M]embers noted that in addition to continuing to develop and test instruments to exit from the period of unusually accommodative monetary policy (Ha!Ha!Ha!), the Committee would need to consider whether further policy stimulus might become appropriate if the outlook were to worsen appreciably.

  7. Gravatar of Mark A. Sadowski Mark A. Sadowski
    14. July 2010 at 12:29

    You wrote:
    “And remember that most economists are liberals. They vote 3-1 Democratic. Admittedly this sample may be slightly different, but I am pretty sure that the group of 44 who oppose more monetary stimulus include at least some economists who vote Democratic.”

    Very different. The AEA and NBER surveys that show economists are liberal are predominantly of academic economists. The WSJ surveys are predominantly of business “economists” who are not only usually thin on credentials (think John Tamny) but they also usually have a rightward bias (not that I think that one should automatically imply the other).

    As for Anna Schwartz, now that Friedman is dead it is clear who had all the brains in that duo.

  8. Gravatar of Jeff Jeff
    14. July 2010 at 13:12

    Few business economists know much monetary economics. Their job is not to understand economics, it is to help sell the firm’s products. The actual experts are almost all academics or employees of the Federal Reserve System. As Scott keeps pointing out, a lot of monetary economics is counter-intuitive (high interest rates actually go with easy money, etc.). When even Anna Schwartz gets it wrong, you start to realize how rare economic understanding actually is. And some of those who do understand this stuff ignore it when it conflicts with their partisan desires.
    The amazing thing is that, as Scott has said here before, most of what he’s saying about the effects of monetary policy was the conventional wisdom among macro economists only a decade or two ago. Nothing better has come along to replace it, and yet the profession largely ignores it.

  9. Gravatar of Adi Adi
    14. July 2010 at 14:20

    Everyone learns in Econ 101 (or at least I did) that when the Fed lowers interest rates, the aggregate demand curve shifts out, which raises prices and output. I think the reason so many people have thrown that basic Econ 101 approach out the window is that short term nominal rates are up against the zero bound, so expansionary monetary policy is associated with “printing money” rather than lowering rates. And the notion of “printing money” scares the hell out people (or at least throws them off balance), and in their fear (or confusion), they throw the basic logic of freshman macro completely out the window: OMG! printing money sounds irresponsible, so it can’t expand output, it’s going to just lead to hyperinflation!

  10. Gravatar of Bob O’Brien Bob O'Brien
    14. July 2010 at 14:57

    Forgive me for asking an econ 101 question (I am one of those business people and not an economist) but could someone explain to me the difference between GDP, NGDP and RGDP? Also, why does Scott prefer to target NGDP instead of inflation? I know a lot of smart business people but my guess is that not 1 in 100,000 would have a clue about this. Thanks.

  11. Gravatar of TheMoneyIllusion » An odd coincidence TheMoneyIllusion » An odd coincidence
    14. July 2010 at 15:47

    […] the time I was writing this sentence for a post this morning: My hunch initial hunch was that the average non-economist simply assumes that governments can […]

  12. Gravatar of JimP JimP
    14. July 2010 at 16:03


  13. Gravatar of jean jean
    14. July 2010 at 16:45

    I see another solution than the one suggested by Krugman: allow Fed to give dividends to the treasury, even it has negative equity.

  14. Gravatar of Jeff Jeff
    14. July 2010 at 16:55


    RGDP (real gdp) is NGDP (nominal gdp) divided by a price index various called the implicit price deflator or the gdp deflator. Usually, when people refer to GDP growth, they’re talking about the annualized growth rate of RGDP.

    In principle, at least, NGDP is easier to measure than RGDP. You just add up all the dollars earned by anyone in any way, and you’re done. To calculate RGDP, you have to measure prices, correct for changes in quality and in the composition of output, and a million other complications. Corrections for quality and substitution bias in the CPI are already controversial. The GDP deflator is even flakier, in that its construction involves a lot of assumptions that can be questioned. For example, an insurance company takes in 10 percent more revenue this year than last. How to split that 10 percent between output and prices? How do you even measure output for an insurance company?

    Even if you have a good price index to target, nobody thinks prices respond to changes in demand as fast as nominal output does. If monetary policy affects prices only with “long and variable lags”, it makes it hard for the central bank to know whether or not it’s on target. I would expect that price index expectations would similarly contain more noise and less signal than nominal output expectations. But I might be wrong about that.

  15. Gravatar of Bob O’Brien Bob O'Brien
    14. July 2010 at 19:28

    Thanks Jeff.

    Your explaination makes it crystal clear to me.

    Im guessing Scott likes to target NGDP growth instead of inflation because NGDP growth is not only easier to measure than RGDP but is also much easier to measure than inflation.

  16. Gravatar of Bob O’Brien Bob O'Brien
    14. July 2010 at 19:31


    One more question for you. How can the fed make NGDP go up?

  17. Gravatar of Doc Merlin Doc Merlin
    14. July 2010 at 20:12

    @Bob O’Brien:
    “One more question for you. How can the fed make NGDP go up?”

    In its monetary authority role:
    1.Print more money and do things with it,
    2.Change the inflation target
    3.Start buying lots of foreign currency (or gold) with its dollars

    In its banking regulatory role:
    1.Reduce real costs associated with banking by relaxing costly regulations
    This would also affect RGDP.

  18. Gravatar of JeffreyY JeffreyY
    14. July 2010 at 21:57

    I think it’s not obvious to most people that inflation and RGDP are connected. The empirical graph that connected them in the 60s was the Phillips curve, but people have heard that stagflation debunked it (when, AFAIU, stagflation only changed the y axis from “inflation” to “dInflation/dt”). When I try to explain how the connection works, I have to fall back on sticky prices, which people are skeptical of.

    I also have to blink when you say AD==NGDP. I see that that identification makes sense, but intuitively “aggregate demand” matches better with “demand for real stuff” aka RGDP in my head.

  19. Gravatar of Lorenzo from Oz Lorenzo from Oz
    15. July 2010 at 02:46

    Speaking as a former national account statistics adviser for the Australian Parliament, NGDP is a statistical construct, but a rather more solid one than RGDP because of the deflator issues that Jeff identifies.

    As for the CPI, all sorts of fun and games involved.

    One of the reasons I like Scott’s take is that NGDP is total productive transactions, and we want more transactions. The most obvious difficulty that strikes me is that money is used for goods, services and assets and NGDP only covers expenditure on asset creation (i.e. investment) rather than total turnover of assets. One cannot even say that assets are intertemporal when goods and services are not, since expenditure on goods and services are connected to expectations of future income (see Friedman’s permanent income hypothesis).

    As for the more general point of the post, it shows in part how much macroeconomics is the poor relation within economics. It simply does not have the robustness of adherence that conventional microeconomics does: for good reasons, alas.

    As Scott points out, there is a drastic lack of a common analytical language and, as Brad Delong has pointed out, is still going around and around a framing that date backs to 1839.

  20. Gravatar of DD DD
    15. July 2010 at 05:30

    From Friedman’s Noble lecture.


    “Professional controversy about the relation between inflation and unemployment has been intertwined with controversy about the relative role of monetary, fiscal, and other factors in influencing aggregate demand. One issue
    deals with how a change in aggregate nominal demand, however produced, works itself out through changes in employment and price levels; the other, with the factors accounting for the changes in aggregate nominal demand.
    The two issues are closely related. The effects of a change in aggregate nominal demand on employment and price levels may not be independent of the source of the change, and conversely the effect of monetary, fiscal, or other
    forces on aggregate nominal demand may depend on how employment and price levels react. A full analysis will clearly have to treat the two issues jointly.
    Yet there is a considerable measure of independence between them. To a first approximation, the effects on employment and price levels may depend only on the magnitude of the change in aggregate nominal demand, not on its source.”

  21. Gravatar of scott sumner scott sumner
    15. July 2010 at 06:28

    Benjamin, Yes, the fear of hyperinflation seems nuts. BTW, there is an amusing typo in your final sentence. That would get me down to the track!

    Indy, Good point. The reason it wasn’t worse is because although we are somewhat incompetent, we aren’t THAT incompetent. Bernanke wouldn’t let NGDP fall in half.

    Benjamin, I see you caught it too.

    marcus, Some of it is self-interest, but a lot of non-government officials are also saying nutty things.

    Luis, Good points.

    Mark, Good point, but I can’t imagine they differ all that much. Why did so many business economists support fiscal stimulus?

    Jeff, Very well put.

    Adi, Yes, I’ve also been thinking along the same lines. By the way, this “compartmentalization” of thinking also occurred during the Great Depression. People said monetary stimulus wouldn’t help, then the same people said a lot of monetary stimulus would give us hyperinflation. (And in Japan as well.)

    Bob, Any attempt to target RGDP will leave the price level indeterminate in the long run (hyperinflation or hyperdeflation.) That’s because we don’t know the natural rate of RGDP. NGDP is safer. If it turns out the natural rate of growth is 2.3%, then 5% NGDP growth gives you 2.7% inflation over time. If the natural rate of growth is 3.1%, then 5% NGDP growth gives you 1.9% inflation over time. NGDP is better if you care about both inflation and stable growth.

    JimP, That’s actually good, as we need to focus on practical solutions–the helicopter drop is a distraction. The Fed has all the tools it needs.

    Jeff, That’s also a good response to Bob.

    Doc Merlin, Thanks, those are good answers to Bob.

    JeffreyY, The better correlation is between NGDP and RGDP. But even that doesn’t hold very well during supply shocks like 1974. But right now it does hold.

    No, AD isn’t RGDP. Consider the economy at full employment (vertical AS curve) You can still boost AD a lot, but RGDP can’t grow at all.

    Lorenzo, I’ve thought about asset turnover, but in the end I just don’t buy it. Ultimately we are interested in the business cycle, and unemployment. The forex market may have a turnover that is ten times bigger than GDP, but it doesn’t create many jobs (except for traders). Does this make sense? I think we want something correlated with employment–construction of services, nondurables and assets.

  22. Gravatar of scott sumner scott sumner
    15. July 2010 at 06:49

    DD, Thanks, That is also my view.

  23. Gravatar of Doc Merlin Doc Merlin
    15. July 2010 at 07:51

    “Im guessing Scott likes to target NGDP growth instead of inflation because NGDP growth is not only easier to measure than RGDP but is also much easier to measure than inflation.”

    NGDP is also a hell of a lot easier to predict, short term inflation estimates tend to be absolutely horrible.

  24. Gravatar of Bob O’Brien Bob O'Brien
    15. July 2010 at 08:04

    Thanks Doc Merlin,

    Say the Fed does #1 and prints money to buy an aircraft carrier and build a few new highways. Does this mean, in a time like now with very tame inflation, we get an increase in employment, NGDP etc. without adding to the debt or deficit? If yes, this seems too good to be true, where is the down side?

  25. Gravatar of Jeff Jeff
    15. July 2010 at 10:20


    Absolutely, but this is actually a trick answer. The trick is that Federal Reserve Notes (dollar bills) are not counted as part of the national debt. So the Fed printing money does not add to the debt or deficit, but the increased taxes from people who build the carrier reduce the deficit.

    Now you might argue that the debt should include (Fed assets – Fed liabilities), but notice that when the Fed prints money, it puts it into circulation by buying something. [Usually, that something is a Treasury security, but recently it could have been an MBS (mortgage-backed security)]. So the result is that Fed assets and Fed liabilities both increase by the same amount.

    The Treasuries bought by the Fed do still count as part of the national debt subject to the debt limit. But “debt held by the public” decreases because of the Fed purchase, and Federal Reserve Notes aren’t counted as part of that. “Monetizing the debt” doesn’t make it go away, it just moves it into the Fed vaults in exchange for currency or reserves.

    Hope this helps.

  26. Gravatar of Bob O’Brien Bob O'Brien
    15. July 2010 at 12:47

    Thanks Jeff,

    I think I understand. The government prints and spends money and makes a bookkeeping entry such that it has more assets and owes more money to itself and as long as this does not drive inflation too high all is well.

    This seems to imply to me that printing and spending money to increase NGDP is similar, maybe even the same, as fiscal deficit spending. For example the Fed could print money and the government could use it to pay off the $800 Billion stimulus bill debt instead of buying the aircraft carrier.

  27. Gravatar of Lorenzo from Oz Lorenzo from Oz
    15. July 2010 at 13:05

    If you want further grounds for pessimism about economic prospects try the Baltic Dry Index:
    The BDI is one of the purest leading indicators of economic activity. It measures the demand to move raw materials and precursors to production, as well as the supply of ships available to move this cargo. Consumer spending and other economic indicators are backward looking, meaning they examine what has already occurred. The BDI offers a real time glimpse at global raw material and infrastructure demand. Unlike stock and commodities markets, the Baltic Dry Index is totally devoid of speculative players. The trading is limited only to the member companies, and the only relevant parties securing contracts are those who have actual cargo to move and those who have the ships to move it. … Tuesday’s close lower on the Baltic Dry Index marked the thirty-third consecutive decline of the index, an indication of the daily rate for a ship carrying dry bulk goods such as grain, coal and iron ore.

  28. Gravatar of scott sumner scott sumner
    16. July 2010 at 08:06

    Jeff, Thanks for the answer to Bob, Just to be clear, even if the base did count as part of the debt, money supply increases would not raise the debt (nor would they decrease it.)

    Lorenzo, Oh oh. Last time the BDI fell sharply was late 2008.

  29. Gravatar of Andy Harless Andy Harless
    19. July 2010 at 10:02

    Obviously I’m going to have to start reading your blog more regularly. A few thoughts:

    1. Under today’s circumstances, fiscal policy is more precise than monetary policy. (It’s easy to get a vaguely reasonable estimate of the effect of a fiscal policy move, e.g. extending unemployment benefits, on AD. Much harder to estimate the effect of, say, a $200 billion purchase of 5-year T-notes by the Fed.) I think this is somehow related to the conception that most people seem to have that fiscal policy affects real output and monetary policy affects the price level. I’m not sure exactly how.

    2. The Fed moves last, as they say. Given that AD tends to affect inflation over a longer horizon than it affects real growth, it makes a little bit of sense to attribute inflation to the last mover. We’re contemplating short-run fiscal policy moves that can be estimated to affect real growth, but in the long run, the Fed will determine whether they affect inflation.

    3. I don’t think it’s quite true that “monetary stimulus…doesn’t increase deficits.” True that it doesn’t show up in today’s reported deficit, but the Fed’s Treasury purchases do affect the long-run fiscal picture in a way that involves the same kind of risk as having the Treasury borrow more. Basically, the Fed will, if things work out well, either have to sell those assets at a loss or to increase the interest rate it pays on reserves, either of which will reduce the Fed’s profit and thereby increase future deficits. It’s largely a matter of accounting whether you take the deficit now (new Treasury borrowing) or later (Fed purchases of outstanding Treasuries to be hopefully liquidated at a loss).

  30. Gravatar of scott sumner scott sumner
    20. July 2010 at 05:40

    Andy, Good questions:

    1. I have argued the exact opposite. Because monetary policy is the last mover, there is no stable fiscal multiplier. If the Fed is inflation targeting then the fiscal multiplier is always precisely zero, as fiscal stimulus does not boost inflation.

    I advocate a forward-looking monetary policy tied to NGDP futures. We don’t have that regime, but we do have TIPS spreads which allow us to measure the impact of monetary policy on inflation expectations in real time.

    2. Good point about long run inflation–I agree. But my problem with all this is that there is no reason to talk about fiscal policy as a stablization tool, but to ignore the role of money as a stabilization tool. Even worse, when there is discussion of using monetary policy in terms of stabilization, it is couched in the language of inflation. People will say “why not have the Fed increase inflation to boost growth?” But the Fed doesn’t directly control inflation, they control NGDP growth which then gets partitioned into prices and output according to the slope of the SRAS curve. But NGDP increases caused by fiscal stimulus also get partitioned into prices and output according to the slope of the SRAS curve. Yet people talk as if the SRAS curve is flatter for fiscal led NGDP expansion than monetary led NGDP expansion.

    3. I have several comments here. If you assume the EMH is true, then the expected loss on Fed Treasury security purchases is zero. And even if the EMH doesn’t hold, they can buy relatively safe Treasury notes for which the losses would be small. On the other hand the expected budget cost of the $787 billion stimulus is in the hundreds of billions (even assuming a growth dividend.) But my more important argument is as follows. Monetary stimulus does not require any additional purchases of Treasury securities. The Fed can lower the demand for base money, and that has the same effect as an increase in supply. They have already more than doubled the monetary base, which is more than enough if all that extra cash went out into circulation. Indeed it would be too much. And it is easy to get all that extra cash out into circulation. Just charge a negative interest rate on ERs (including vault cash), and then set a modestly higher inflation target to discourage the private sector from hoarding cash in safes. There is a reason it is called high-powered money. If done right a little goes a long way.

  31. Gravatar of Andy Harless Andy Harless
    20. July 2010 at 15:09

    How can you charge a negative interest rate on vault cash? That sounds like an invitation to subterfuge. Any kind of low-maturity highly liquid instrument (including T-bills) would be a close substitute for reserves, so you would have to restrict banks from holding those. You’d have to somehow distinguish between real short-term lending and bogus short-term lending that really consists of banks parking their cash in a place where it won’t be subject to the penalty. A regulatory nightmare.

  32. Gravatar of ssumner ssumner
    21. July 2010 at 07:46

    Andy. You’d simply allow a minimum amount of vault cash interest free, and then charge banks who held vault cash balances above that normal ratio. Vault cash has traditionally been a part of reserves, usable for reserve requirements (don’t know if that’s still true) so it should be possible to count it accurately enough to prevent banks from parking $100s of billion in vault cash. Even if you assume bankers are criminals and would risk going to jail to cheat the Fed out of a bit of interest, then you’d also have to consider the problem of pulling this off. Banks get their cash from the Fed. What if the Bank of America suddenly asked the Fed for $10,000,000,000 in hundred dollar bills? Clearly the Fed would investigate. I’m not saying it would work perfectly, but it would work well enough to dislodge the vast majority of the ERs currently being held.

    As far as swapping reserves for T-bills, that is precisely the point. Whatever banks choose to hold instead of ERs, it still has the same effect of pushing the ERs out into circulation, where it becomes part of the money stock. Cash is a much less close substitute for T-bills than ERs. Cash can be stolen, for instance, unlike T-bills and ERs.

    Again, this isn’t my prefered policy, they really need NGDP or price level targeting, in which case none of this would be necessary. But it is a way to ease money without bloating the Fed’s balance sheet any further.

    BTW, Robert Hall liked my idea.

  33. Gravatar of Andy Harless Andy Harless
    21. July 2010 at 10:38

    I don’t see how it would do any good to push ERs out into circulation by having banks buy T-bills. True, currency and T-bills aren’t perfect substitutes, but they’re pretty close. When the interest rate on T-bills goes more than a few basis points below zero (enough to cover the insurance on cash held in a safe deposit box), most of the non-bank investors who own them will choose to invest in currency instead, because the return will be higher. Banks will effectively still be holding excess reserves, but instead of holding them in the vault and recording them as an asset of the bank, they’ll hold them in a safe deposit box and record them as an asset of the customer. Banks own balance sheets will have T-bills on the asset side, and whenever they need to use the “reserves”, they’ll simply sell a T-bill to the marginal customer, who will then take the “reserves” out of the safe deposit box. It’s only slightly less convenient than having the cash in their own vault.

    I suppose you could prohibit banks from buying negative-yield assets (except reserves), though. That wouldn’t be too hard to enforce, and I can’t think of any obvious problems.

    On a separate point, regarding the question of relative precision of fiscal and monetary policy, it’s true that, if you view them separately, subject to each other’s reaction function, then the monetary policy reaction function certainly does throw a wrench into the impact of fiscal policy. But if you view them ceteris paribus as tools between which we are choosing (and if you consider the tools that are actually being widely contemplated), fiscal policy is a lot more precise, at least in the short run. For this reason, it’s easier to imagine monetary policy overshooting to an extent that would produce inflation, and it’s also easier to imagine a monetary policy action undershooting to an extent that it has no significant impact on real output. If you hold monetary policy constant (in the sense of holding the interest rate constant and not doing additional QE — which is surely the sense most people would think of it), then fiscal policy can be expected (if it’s effective) to affect real output but have little effect on inflation (assuming we face a highly convex Phillips curve — or if we face a linear Phillips curve but aren’t concerned about inflation rates that are below, say, 2%). We may be legitimately concerned, however, that monetary policy actions (holding fiscal policy constant), pursued on a sufficiently large scale, create a serious risk of inflation. It makes sense to me, for example, that Anna Schwartz may be thinking, “Use fiscal policy because it will probably help, and we can estimate the impact well enough to be confident that it won’t cause inflation, but avoid using a lot of monetary policy, because we don’t know what the effect will be, and it could well cause inflation.”

  34. Gravatar of ssumner ssumner
    22. July 2010 at 05:11

    Andy, I think you slightly misunderstood my argument for pushing reserves out into cash, but it’s probably my fault. These issues are complex and I don’t always mention all my background assumption. It is always true that temporary cuurrency injections have little impact, and permanent currency injections have an expansionary impact. This is true when rates are at the zero bound, and it is almost equally true if interest rates are not at the zero bound. The argument for negative IOR is exactly the same as the argument for more QE. In either case the extra money flowing into the economy might be hoarded by banks or the public. I does not solve the problem of currency injections that are only expected to be temporary, unless it changes expectations about the future money supply.
    That is why inflation/NGDP targets are really the key to monetary policy. AS long as inflation expectations stay below target, monetary injections won’t help. My point about negative IOR was two-fold:

    1. It might be interpreted by the markets as a signal that the Fed wants more inflation.
    2. At worst it allows the Fed to do a trillion in QE, without increasing the size of the Fed’s balance sheet. Lowering the demand for base money is identical to increasing the supply. Each may or may not work for the same reason.

    You said;

    “If you hold monetary policy constant (in the sense of holding the interest rate constant and not doing additional QE “” which is surely the sense most people would think of it)”

    This in my view is precisely the problem with the view of “most people”. By far the Fed’s most powerful tool is signals about the future stance of monetary policy. Most people don’t seem to realize this, and falsely view a Fed with a stable M or a stable i (two very different assumptions during most periods, BTW) is “doing nothing.” The most powerful monetary policy shock in American history was the dollar devaluation of 1933, and yet there was little change in M or i. But future expected M rose sharply. One of the reasons why fiscal stimulus failed was because all the Fed chatter early in the year about “exit strategies” had the effect of holding down inflation expectations. If fiscal stimulus does not raise inflation expectations, then it has no chance of working. (Yes, the increase may be small when there is slack, but there must be some increase. And slack also makes an inflation overshoot from monetary stimulus less likely.)
    In the real world you can never take monetary policy out of the equation. That doesn’t mean fiscal stimulus can’t work, I think it can sometimes raise inflation expectations, but it does require some cooperation from the Fed.

    My work on the Great Depression convinces me that there is little chance of overshooting toward high inflation from a liquidity trap. When FDR stopped devaluing the dollar in early 1934 many conventional economists wanred him that it was an inflationary time bomb. Indeed that is one reason he stopped. They talked about long policy lags. George Warren warned him that the inflation would end once FDR stopped devaluing the dollar, as wholesale price indices at the time were heavily weighted twoard commodity prices, which respond instantly to monetary shocks. Warren pointed out that commodity prices remained far below pre-depression levels (FDR’s target was 1926 prices), and hence further dollar devaluation was required. The WPI had already increased 20% from March 1933 to February 1934, when the dolalr was again fixed to gold. In the next 6 years it only increased another 5%, less than 1% per year. Warren was right and the economic establishment was wrong. Lags are not a problem.

    Look at market sensitive prices and you can see whether you have done enough monetary stimulus. Monetary policy is far more precise that fiscal stimulus. Monetary policy can and should peg the price of NGDP or price level futures contracts, fiscal stimulus depends completely on the future expected policy of the Fed, Even if current M and current I are held fixed, the Fed can and does powerfully shape expectations with signals about future policy. Bernanke probably lowered inflation expectations a bit yesterday.

    I agree, BTW, that it would be silly to forbid banks from holding negative rate T-bills.

    Thanks again for all the great comments.

  35. Gravatar of Lorenzo from Oz Lorenzo from Oz
    24. July 2010 at 00:12

    Scott: The forex market may have a turnover that is ten times bigger than GDP, but it doesn’t create many jobs (except for traders). Does this make sense?
    Yes, because I was thinking of assets more in a rather vague and unhelpful way.

  36. Gravatar of TheMoneyIllusion » Look! The economy’s sinking! Someone should do something! TheMoneyIllusion » Look! The economy’s sinking! Someone should do something!
    24. July 2010 at 08:05

    […] has never made any sense to me.  In the comment section of a recent post, Andy Harless provided the most plausible explanation that I have seen thus far: Under today’s […]

  37. Gravatar of Look! The Economy’s Sinking! Someone Should Do Something! Look! The Economy’s Sinking! Someone Should Do Something!
    26. July 2010 at 05:24

    […] has never made any sense to me.  In the comment section of a recent post, Andy Harless provided the most plausible explanation that I have seen thus far: Under today’s […]

Leave a Reply