Britain needs more AD, no wait . . . less AD

In a number of posts I’ve suggested that there is a basic confusion over stabilization policy.  Instead of viewing fiscal and monetary stimulus as two ways of influencing AD, most people talk as if fiscal policy affects real GDP, and monetary policy affects prices.  Today I’d like to argue that this logical error is deeply embedded in our profession, and indeed is a major cause of the crash of 2008.  Here’s The Economist:

The surprising weakness was caused in part by unusually cold weather but it is not fully explained by it. A frail economy ought at least to be free of price pressures but Britons have had no such luck. Inflation rose to 3.7% in December.  . . .

Such a run of bad news looks like an argument for a policy rethink. If the economy stays weak, it may not be robust enough to withstand further deficit-cutting measures, including a planned rise in national-insurance contributions this April. The persistence of high inflation (it has been well above the 2% target for most of the last three years) calls into question the idea that the Bank of England could counter the effects of fiscal tightening by easing monetary policy. Its benchmark interest rate is already as low as it can feasibly go, and a further round of “quantitative easing” would stretch too far the gap between the bank’s objective of low inflation and its actions. A concern that businesses and wage earners might think policymakers were going soft on inflation led two of the bank’s nine-strong monetary-policy committee to vote for an increase in interest rates this month (see article).

I don’t have a major problem with any individual sentence in this article.  And they end up arguing against a change in the policy mix, which I think is a defensible view.  But I do have a problem with the way the issues are discussed.  There’s very clear implication that it’s the duty of monetary policy to deal with (high) inflation, and the responsibility of fiscal policy to address growth.  If you don’t believe me, if you think in both cases The Economist was focused on AD, and randomly chose to mention the inflationary effect of AD when talking about monetary policy, and the growth impact of AD when discussing fiscal policy, then I have a challenge for you.  Please send me an article from the British press that suggests that Britain needs to address the 3.7% inflation with further fiscal austerity, and needs to address the slow recovery with more monetary stimulus.  That’s the policy I favor, but I doubt you will be able to produce a single article that favors that policy mix, and justifies both changes in the way I describe.

In talking to many ordinary people, in reading the financial press, and in talking to many economists, I have gradually become convinced that almost everyone compartmentalizes these issues in a way that isn’t justified by current macro theory.  I will argue that this has been very harmful.  Not just recently, but also back in the Great Depression.  Here’s what Keynes said in his famous letter to the NYT on December 31, 1933:

The other set of fallacies, of which I fear the influence, arises out of a crude economic doctrine commonly known as the quantity theory of money.  Rising output and rising incomes will suffer a setback sooner or later if the quantity of money is rigidly fixed.  Some people seem to infer from this that output and income can be raised by increasing the quantity of money.  But this is like trying to get fat by buying a larger belt.  In the United States today your belt is plenty big enough for your belly.

People often claim that Keynes didn’t believe in liquidity traps.  If a liquidity trap is a situation where increases in the quantity of money have no impact on AD, then Keynes certainly did believe in liquidity traps.  In the preceding paragraph he is clearly making the “pushing on a string” argument.  But notice that he makes it for output, not prices.  In the very next paragraph Keynes switches gears:

It is an even more foolish application of the same ideas to believe that there is a mathematical relation between the price of gold and the price of other things.  It is true that the value of the dollar in terms of foreign currencies will affect the prices of those goods which enter into international trade.  In so far as an overvaluation of the dollar was impeding the freedom of domestic price-raising policies or disturbing the balance of payments with foreign countries, it was advisable to depreciate it.  But exchange depreciation should follow the success of your domestic price raising policy as its natural consequence, and should not be allowed to disturb the whole world by preceding its justification at an entirely arbitrary pace.

First a little context.  The first sentence of this paragraph is a jab at George Warren, an FDR advisor who claimed a mathematical relationship between the price of gold and the price of goods.  Irving Fisher had similar views, although was perhaps a bit less dogmatic than Warren.  Note that Keynes calls this a “foolish application of the same idea.”  In other words, saying 20% more money means 20% higher incomes and the idea that 20% higher gold prices means 20% higher goods prices are basically two sides of the same fallacy.  Except that they aren’t at all.  Indeed Keynes knows this, he knows that Warren’s dollar depreciation program did raise the US price level, and raised it dramatically.  The WPI rose strongly throughout 1933, and the weekly increases were highly correlated with changes in the dollar price of gold.

More importantly, everyone knew this.   If Keynes had attempted to apply the liquidity trap concept to prices, his reputation would have been torn to shreds.  He almost always applied the monetary policy ineffectiveness concept to real output, almost never to prices.  Especially if the monetary shock was large, and would obviously be inflationary.  Keynes was no fool.  He was careful to suggest there was no “mathematical relation” with prices, a very different argument for no effect at all.  But there is a problem here.  The Keynesian model says inflation and real growth go hand in hand.  When the economy is at less than full employment (as it certainly was in 1933); one cannot have rising prices without rising output.

Here’s an important difference between Keynes and Paul Krugman.  Krugman does understand the implication of Keynesian theory.  He does understand that if monetary policy is ineffective in raising output, then ipso facto it is ineffective in raising prices.  In 2010 the Fed was accused of putting the cart ahead of the horse—artificially pushing prices higher with monetary policy (QE2, which depreciated the dollar) rather than letting prices rise “naturally” as a consequence of economic recovery driven by more domestic spending.  Keynes would have opposed QE2; he would have called it “unnatural” and “arbitrary.”  Krugman, who understands Keynesian theory far better than Keynes ever did, supported the policy.  (BTW, Keynes got his way;  FDR gave up on monetary stimulus two weeks later and never tried it again.)

Allan Meltzer once wrote that it was odd that Keynes never suggested that central banks target inflation at something like 4%, so that they could avoid the zero rate bound, and thus the need for fiscal stabilization policy.  I think Keynes would have been horrified by that idea.  For Keynes, the “natural” price level is a stable one, unless you need a bit of reflation to make up for a previous plunge in prices.  The “natural” way to boost AD is with more government spending, or more investment.  Money should not be a “belt” that prevents economic recovery, but it also shouldn’t be used to “artificially” push the economy higher, by boosting prices.

You’re probably thinking that we have advanced far beyond the primitive analysis of the 1933 Keynes (the low point of his career, BTW.)  Unfortunately not.  Consider the 100 most prominent macroeconomists.  When AD started plunging sharply in the second half of 2008, how many vigorously and loudly insisted on the need for a much more aggressive monetary policy stance?  Here’s the list:

1. . . .

Well that didn’t take long.  But if I had to list the number who forcefully advocated fiscal stimulus, it certainly would be a quite long list.  The general view was; “Well, rates are already pretty low, and the base has been rising.  So monetary policy can’t be holding back the recovery.  The belt is not too tight.  Time to engage in fiscal stimulus—which is more certain to boost spending.  After all, G is part of C+I+G, but I don’t see “M” in that formula.”

This was a tragic error.  The easy fiscal policy and tight money of 2009 have put us where we are today.  By 2010 the Keynesians realized the fiscal stimulus wouldn’t do the job (and to their credit, a few Keynesians like Krugman argued this from the beginning.)  So by 2010 you had famous Keynesians like Alan Blinder desperately seeking monetary policy options.  Since they can’t easily break away from the “interest rates are all that matters” view of monetary policy, Blinder ended up gravitating to the idea of negative rates on bank reserves, a wacky scheme first mentioned in a couple publications by yours truly in early 2009.  Others said printing money was worth a shot, after all, what do we have to lose?

I still think our language is holding us back.  We should never discuss AD issues by talking about prices and output; we should always refer to NGDP.  If we did so, we’d be much less likely to discuss policy in the confused way The Economist does in the excerpt above.  It would sound bizarre to say “The UK government should refrain from fiscal austerity, lest NGDP falls, and they should refrain from monetary stimulus, or else NGDP might rise.

That doesn’t mean the P/Y split is completely inconsequential.  In terms of human welfare we’d prefer to have more real growth and less inflation.  But that’s a supply-side issue, which can’t be addressed through demand-side stimulus.  Some commenters have made clever attempts to argue that fiscal stimulus would be better for aggregate supply, but I guess I’m just too libertarian to buy that argument.  I’d rather have monetary policy target NGDP along a stable 5% growth trajectory, and have all fiscal decisions subject to rigorous cost/benefit considerations.  Because a “targeting the forecast” approach means expected future NGDP would always be on target (even though current NGDP may be below target) there would be no “depression economics.”  Fiscal policy makers should operate in a classical world with real opportunity costs.

How will we know if the quasi-monetarists have won?  When everyone else starts using their language.  The more I see bloggers like Matt Yglesias and Brad DeLong talk about NGDP, the happier I am.  It doesn’t matter if they stay Keynesian—NGDP talk will slowly and insidiously lead to quasi-monetarist thinking.  That’s why Krugman slapped down David Beckworth last year; Krugman argued that talk about NGDP would lead to dangerous monetarist ideas.  At the time I argued that he was being illogical.  Now I see that the field of macro can’t be understood in strictly logical terms.


Tags:

 
 
 

37 Responses to “Britain needs more AD, no wait . . . less AD”

  1. Gravatar of Mark A. Sadowski Mark A. Sadowski
    5. February 2011 at 10:08

    Scott,
    I’ve been reflecting on the “stagflation” business and Menzie’s, Andy’s and your statements concerning currency depreciation in a small open country (although my mother would have been taken aback by the UK being described as “small”). Menzie mentioned the impact of rising energy prices, similarly Andy mentioned a negative AS shock and you pointed out the positive effect on AD by currency depreciation would far outweigh any negative effects on AS (I hope that’s a fair summary).

    I looked into the energy situation and as expected thanks to North Sea oil production the UK is far less dependent on energy imports than the Eurozone. Also it seems to me that the effects of the pound’s depreciation are far outweighed by the general runup in energy prices since their collapse in late 2008. If energy were the culprit we should also be seeing higher inflation in the Eurozone and we clearly are not. It seems to me the “depreciation induced negative AS shock through energy prices thesis” doesn’t really ring true (as applied to the UK).

    I know you’ve brought up the issue of higher tax rates and increased government expenditures as having negative effects on British AS. But these seem too modest to me to explain their rise in inflation in the midst of such a pathetic recovery. There clearly is some kind of supply side drag at work but I still don’t see where it’s coming from.

  2. Gravatar of nanute nanute
    5. February 2011 at 10:08

    And here’s Bill Vickrey talking about inflation and unemployment in 1996: The main difficulty with inflation, indeed, is not with the effects of inflation itself, but the unemployment produced by inappropriate attempts to control the inflation. Actually, unanticipated acceleration of inflation can reduce the real deficit relative to the nominal deficit by reducing the real value of the outstanding long-term debt. If a policy of limiting the nominal budget deficit is persisted in, this is likely to result in continued excessive unemployment due to reduction in effective demand. The answer is not to decrease the nominal deficit to check inflation by increased unemployment, but rather to increase the nominal deficit to maintain the real deficit, controlling inflation, if necessary, by direct means that do not involve increased unemployment. http://www.columbia.edu/dlc/wp/econ/vickrey.html

  3. Gravatar of W. Peden W. Peden
    5. February 2011 at 10:39

    Prof. Sumner,

    “When the economy is at less than full employment (as it certainly was in 1933); one cannot have rising prices without rising output.”

    Just to see if I understand what you are saying correctly: you’re saying that an increase in NGDP can only result in an increase in overall prices insofar as it raises demand for goods & services above the level at which suppliers of goods & services can respond?

    Or another (!) way of putting it: if prices rise as a result of NGDP, then that implies that there are no resources that are laying idle as a result of a nominal shock?

    There’s a lot of talk here in the UK about how the UK can get sustainable growth. You’ve convinced me that there is no single policy that would do us so much good as adopting a NGDP target of 5.6% (the trend rate in the UK’s Great Moderation between 1993 and 2007).

  4. Gravatar of Cameron Cameron
    5. February 2011 at 10:40

    Great post.

    We can complain about conservative opposition to monetary stimulus all we want, but at the end of the day I think it is the ambivalence (at least until recently) of influential keynesians like Krugman and Blinder that has prevented easier policy.

    And it’s bad enough that fiscal policy is much more expensive to carry out, but it seems clearly less effective as well!(based on both market reactions and economic data – how do those pessimistic about monetary policy explain the recent and significant improvements in the economy?)

    If this is true, fiscal stimulus opposition has probably on net helped the economy by leaving keynesians (and the fed) with no other option than easier monetary policy.

  5. Gravatar of bill woolsey bill woolsey
    5. February 2011 at 11:10

    I usually use some version of the GDP deflator to measure the price level.

    I am not using the prices of imported goods and I am counting the prices of capital goods and government goods.

    If you think that Fed policy is supposed to be controlling the “cost of living” rather than trying to avoid using the changes in the price level (including wages) to adjust the real quantity of money to the demand to hold it, then perhaps the CPI is the way to go.

    Anyway, if the dollar falls in value, and foreign goods get more expensive, this will raise the CPI. Only to the degree that this causes people to buy more domestic products or else foreigners choose to buy more U.S. goods will this raise the demand for U.S. output and so tend to raise the GDP deflator.

    In other words, controlling CPI inflation translates into using monetary policy to control the exchange rate. Oh, and raising interest rates will tend to create a net capital flow, raise the dollar, lower import prices, and lower the CPI.

    Now, can fiscal policy control demand for domestic product without causing changes in the exchange rate? Well, the U.S. Treasury sells bonds to foreigners and uses the proceeds to pay firms to build roads in the U.S. Interest rates rise in the U.S., attracting the net capital inflow, the dollar rises, imported goods are cheaper, and the CPI falls. This tends to reduce the demand for U.S. import competing products and exports (that damn trade balance leakage.)

    The immediate effect of the added demand for government goods doesn’t raise the CPI. Of course, the employed workers may purchase consumer goods and services.

    Anyway, a slight loosening of monetary policy (lowering interest rates) can prevent the Fiscal policy from raising the value of dollar, lowering import prices, and lowering the CPI.

    So, monetary policy (interest rates) controls the CPI and fiscal policy (funded by capital inflows and outflows) controls domestic demand.

    Interest rate targeting and targeting the CPI (core or otherwise) is a big mistake!

    Keep final sales of domestic product growing at a slow, steady rate by making sure the quantity of money remains equal to the demand to hold money. Let interest rates and the exchange rate, and the prices of imported goods adjusted based on market forces.

  6. Gravatar of W. Peden W. Peden
    5. February 2011 at 11:29

    Bill Woosley,

    Good points. Also, focusing on CPI means that monetary policy can ignore major overheating e.g. there could be excessively high NGDP growth, but the price increases all go into housing because supply factors are reducing the cost of consumer goods but not houses.

    One consequence of this blog has been to make me see the limitations of inflation targeting in general.

  7. Gravatar of anon anon
    5. February 2011 at 11:46

    “There’s very clear implication that it’s the duty of monetary policy to deal with (high) inflation, and the responsibility of fiscal policy to address growth.”

    It seems that you want to bring back pre-1980s monetary policy. The Great Stagflation of the 1970s happened because monetary authorities maintained the existing path of NGDP growth in the face of a supply-side squeeze in the real economy, so inflation expectations became unanchored. (Yes, there was an attempt to control inflation via fiscal policy: the notorious “Whip Inflation Now” buttons.) I thought all monetary economists had learned something from that particular mistake.

  8. Gravatar of Morgan Warstler Morgan Warstler
    5. February 2011 at 11:54

    “Please send me an article from the British press that suggests that Britain needs to address the 3.7% inflation with further fiscal austerity, and needs to address the slow recovery with more monetary stimulus.”

    “I’d rather have monetary policy target NGDP along a stable 5% growth trajectory, and have all fiscal decisions subject to rigorous cost/benefit considerations.”

    Whoa! Hold the phone!

    Excuse me Mr. Sumner, sir… but, dude – I have been yelling at the top of my voice for months on end that you should tie those two things together explicitly in your own daily diatribes.

    I have said, Scott, the more you demand we cut government spending (public employee salaries):

    1. the more conservatives will listen to you.
    2. the greater the productivity gains we’ll get out of the public sector. This means privatization, this means repealing Davis-Bacon, this means giving the state weaponry to bargain on pensions.

    If you believe making cuts to Fiscal is how you stabilize prices / grow the economy SAY WHY.

    Say, look here:

    http://www.gpo.gov/davisbacon/davbacsearch.html

    If the average worker on a government project earns 50% less, then we can hire 2x as many, or pay less in taxes (put more in taxpayer pockets). Or say less debt to roll over, as interest rates go up… say something.

    Look, I’m not an economist, but can’t you figure something out some simple narrative like this:

    Since we have been printing money, prices any not gone up, because even as input costs have risen, companies have been loathe to raise prices, so they have fired employees – becoming more productive. They are at their wits end, in the current situation, they can’t fire more people, and if commodity prices go up, they are going to have to raise prices. Inflation looms!

    BUT, within the public sector we have seen 20+ years where productivity gains have been negligible by comparison, and while it is only 20% of our economy – a 25% gain in productivity over the next 5 years (targeting 5% per year), means we are gaining what? 1% a year in GDP?

    Doesn’t an increase in real productivity offset any price increases? You pay less taxes, you can afford higher prices, time to hire private sector workers into SMBs.

    And if there is no inflation coming, well doesn’t mean even greater effect of the productivity gains?

  9. Gravatar of justanothereconomist justanothereconomist
    5. February 2011 at 11:58

    Scott-

    “This was a tragic error. The easy fiscal policy and tight money of 2009 have put us where we are today.”

    Tight money policy of 2009? Money base [from FRED] almost doubles from 2008-09-10 874.796 to 2008-12-31 1690.796. That’s before 2009 even begins. Money base didn’t even near double in the 1970s when NGDP was soaring, and yet NGDP hasn’t doubled in the current situation.

    This is unprecedented monetary stimulus. While you may not like the liquidity trap, you have to admit that something is different now than then- Zero lower bound, whatever you want to call it, but something is very different. And it’s not just interest on reserves, there were massive excess reserves in Japan and the USA in the 1930s with no interest on reserves.

    How much is enough to not have tight money? It’s easy to keep criticizing, as we will never see a money base on 10 trillion, so your ideas can never be tested out.

    Despite all the monetary stimulus, the USA is not even hitting the implicit 2% target that Bernanke admitted to on 60 minutes. Weird right?

    Cameron- Keynesians like me are not opposed to monetary stimulus. But how much is enough? I support QE1 QE2, as many QEs as you want, as its worth a shot, but its impact is minimal.

    “(based on both market reactions and economic data – how do those pessimistic about monetary policy explain the recent and significant improvements in the economy?)”

    But money base is lower today than at the beginning of 2010. While the recent round of QE is a good idea, it’s still hasn’t increased the money base much. We’re still below early 2010 levels of money base. So you believe, like Scott, that an increase of onver a trillion dollars of the money base wasn’t enough, but 600 billion more will be? I guess we’ll wait and see.

    http://research.stlouisfed.org/fred2/graph/?chart_type=line&s%5B1%5D%5Bid%5D=BASE&s%5B1%5D%5Brange%5D=5yrs

    “If this is true, fiscal stimulus opposition has probably on net helped the economy by leaving keynesians (and the fed) with no other option than easier monetary policy.”

    Maybe you didn’t notice, but we had massive fiscal stimulus through the tax cuts. And the economy is picking up because of this. It’s weaker than the equivalent spending increase, but it still is stimulative, especially the unemployment benefits and other increases in disposable income for the liquidity constrained (lower class). You really think the economy is picking up because of more excess reserves?

    Look at how closely excess reserves follow money base… it’s uncanny:

    http://research.stlouisfed.org/fred2/graph/?chart_type=line&s%5B1%5D%5Bid%5D=AMBNS&s%5B1%5D%5Brange%5D=5yrs

    http://research.stlouisfed.org/fred2/graph/?chart_type=line&s%5B1%5D%5Bid%5D=EXCRESNS&s%5B1%5D%5Brange%5D=5yrs

    That’s a Hicksian liquidity trap- purchases of assets by the central bank are just held as cash balances i.e. excess reserves, and don’t impact the economy as they stay in the bank vaults.

  10. Gravatar of Luis H Arroyo Luis H Arroyo
    5. February 2011 at 12:00

    Good post, I agree; perhaps because I´m of the most skeptical about the fiscal policy and its effects. The article of The Economist is absurd, horrible.I thought that only in Spain it was possible to read this sort of things. A rise in fiscal debt & a monetary contraction do increase interest rate…
    In any case, I´m not capable to understand as UK is a country with so high inflation (higher than US or Germany). I suppose that its famous “free market reforms” has some fugue in some part… Perhaps price intervention in public utilities. When some years ago i work on UK´s economy, I observed that its consumer prices take much more time that US ones to fall during recession.

  11. Gravatar of Nick Rowe Nick Rowe
    5. February 2011 at 13:03

    If my childhood memory of 1960’s Britain is correct: they used to argue for tighter fiscal policy to control inflation, and looser monetary policy to lower interest rates and promote investment and long run growth. (Not the same though, of course, because that’s a supply-side channel for long run growth.

    Part of the problem is that the Quantity Theory and Neutrality of Money is intuitive. Double M, the result is you double P and leave Y unchanged. But Y=C+I+G.

  12. Gravatar of W. Peden W. Peden
    5. February 2011 at 14:01

    Anon,

    US average NGDP growth was over 10% in the 1970s, whereas the post-war average NGDP growth from 1946 to 1969 had been about 6.5%. So, no, monetary authorities in the 1970s weren’t trying to maintain the existing path of NGDP growth. In fact, monetary policy in the 1970s wasn’t based on NGDP growth at all, but on the silly “twin mandate” of inflation + unemployment back when the Fed really saw it as their job to control unemployment.

    If NGDP had stayed at about 6.5% in the 1970s and the double digit inflation was caused by an average annual RGDP contraction of 5%, then your explanation of the Great Stagflation may hold some water. However, that just didn’t happen: NGDP nearly doubled, while annual US RGDP growth in the 1970s was about 3.5%, only about 1% less than the 1960s average. So the rise in inflation in the 1970s can be understood as the result of a rise in NGDP without a corresponding rise in RGDP, as central bankers drove the Phillips Curve-era monetary policy of trying to control unemployment using increases in the quantity of money into total absurdity.

    In general, the facts don’t match what you’re saying at all, and your comparison of Prof. Sumner’s prescription for monetary policy with pre-Volcker monetary policy makes absolutely no sense at all.

  13. Gravatar of W. Peden W. Peden
    5. February 2011 at 14:09

    Nick Rowe,

    The classic policy in the 1960s was monetary tightening to control inflation, fiscal expansion to boost growth. That’s a great way to gradually pound the private sector into submission.

    There were also the folk at the Department of Applied Economics at Cambridge (and their less extreme versions in wider academia/policy making circles) who argued that monetary policy barely mattered at all. This, combined with the theory (beloved of Joan Robinson et al) that wages were determined exogenously by power and bargaining* meant that the proposed solution in the 1960s and 1970s from many Keynesians was permanent incomes policies to fight inflation + fiscal expansionism to boost growth.

    Economic planning was also very, very popular in the 1960s and 1970s. The full apparatus of planning (the National Enterprise Council, the Neddies, quantitative exchange controls, price controls and so on) weren’t dismantled until well into the 1980s.

    While Americans had worked out that money matters by the 1960s, “Money doesn’t matter” was still a perfectly respectable claim in British economics well into the 1970s. In fact, even in the 1980s and early 1990s many people still claimed that there was no way of combining an acceptable level of unemployment and inflation without incomes policies.

    * Books II-IV of the GT were used to back up this claim!

  14. Gravatar of W. Peden W. Peden
    5. February 2011 at 14:20

    Luis H Arroyo,

    Most of UK consumer price inflation (2.1% of 3.7% according to some estimates) is made up of the results of fiscal policy e.g. the 2.5% increase in VAT and increased taxes on fuel.

    On the other hand, high UK RPI cannot be explained by such factors.

    Our problem seems to be that our economy has suffered a lot from the financial crisis and apart from manufacturing it is still suffering. As NGDP has slowed recentely, this has resulted in a fall in RGDP rather than inflation.

    There is a very good case for more quantitative easing in the UK, especially if our economy remains in recession in Q1 2011, but the Bank’s inflation targeting regime will probably stop this from happening. The UK needs a NGDP target and monetary stimulus.

    The plan in June 2010 was to cut spending and boost monetary easing to compensate, as in 1981 which brought us out of our most severe post-war recession. Rather to my surprise, cutting spending is going very well (the government has exceeded its deficit reduction target this year by about £2.9 billion) but boosting monetary easing is being complicated by price index factors, when we need to be looking at NGDP.

  15. Gravatar of Lorenzo from Oz Lorenzo from Oz
    5. February 2011 at 14:28

    Scott, to follow on W. Peden’s summary, does the following capture what you are saying?

    NGDP = total monetary value of productive (as in ‘counts to GDP’) transactions.

    Money is the fundamental economic signalling device (“I am willing to spend x on this”, “I am willing to sell this for y”).

    The more smoothly supply can respond, the more any rising signal of money generates more output. The less smoothly supply can respond, the more any rising signal of money generates higher prices. This is true in any particular market but also in the aggregation across all markets.

    Inflation is the degree to which the money signal increases systematically and persistently beyond the output response across all markets.

    But if folk think their money will speak loudly later, they will hold on to it more, leading to a lowering of the aggregate money signal. If folk think their money will speak less loudly later, they will tend to spend it more, leading to an increase in the aggregate money signal.

  16. Gravatar of Lorenzo from Oz Lorenzo from Oz
    5. February 2011 at 14:38

    And, to follow on:

    Thinking terms of RGDP goes astray, because it misses out how and what economic agents are signaling to each other. Thinking in terms of money alone goes astray, because it misses how money signals about goods and services.

  17. Gravatar of Indy Indy
    5. February 2011 at 14:43

    Trying to move a whole profession is hard. I’m not sure it can be done with the same old personalities – these things usually require a generational shift. Eventually you’ll be vindicated, but I hope it’s sooner rather than later. I don’t know if we can wait a whole generation.

  18. Gravatar of Mark A. Sadowski Mark A. Sadowski
    5. February 2011 at 15:08

    W. Peden,
    You wrote:
    “Most of UK consumer price inflation (2.1% of 3.7% according to some estimates) is made up of the results of fiscal policy e.g. the 2.5% increase in VAT and increased taxes on fuel.”

    If that’s true it explains a lot.

    However, how can that be? The tax increases didn’t go into effect until January and the 3.7% figure is for December yoy CPI.

  19. Gravatar of scott sumner scott sumner
    5. February 2011 at 15:10

    Mark, What about the VAT increase? BTW, I wasn’t suggesting that all of the upswing in inflation is due to statist policies, I think that would be unlikely, I doubt even half is.

    Nanute, That sounds like price controls.

    W. Peden, You said;

    “Or another (!) way of putting it: if prices rise as a result of NGDP, then that implies that there are no resources that are laying idle as a result of a nominal shock?”

    First of all this isn’t about my views, it’s about Keynesian views. The Keynesians say that if more AD pushes up prices, it will also push up output, unless at capacity. And the economy was not at capacity in 1933. Some extreme Keynesians might say inflation implies output at capacity, but I don’t think that view is widespread.

    Cameron, Those are all good points.

    Bill, I agree with your policy views. I’d add that a program like FDR’s gold price targeting did raise both the CPI and the GDP deflator, so it wasn’t only import prices.

    anon, You said;

    “It seems that you want to bring back pre-1980s monetary policy. The Great Stagflation of the 1970s happened because monetary authorities maintained the existing path of NGDP growth in the face of a supply-side squeeze in the real economy, so inflation expectations became unanchored.”

    You’ve got it exactly backward. In the 1970s the Fed let NGDP growth rates soar, to about 11% between 1971 and 1980. Not controlling NGDP was the major mistake.

    Morgan, You said;

    “Excuse me Mr. Sumner, sir… but, dude – I have been yelling at the top of my voice for months on end that you should tie those two things together explicitly in your own daily diatribes.”

    Have you ever considered the possibility that after all that yelling, I’ve now heard your message to gut public employees?

    Justanothereconomist, You said;

    “Tight money policy of 2009? Money base [from FRED] almost doubles from 2008-09-10 874.796 to 2008-12-31 1690.796. That’s before 2009 even begins. Money base didn’t even near double in the 1970s when NGDP was soaring, and yet NGDP hasn’t doubled in the current situation.”

    The monetary base also soared from 1930-33, and yet most economists seem to think tight money greatly aggrevated the Great Depression. That suggests that most economists don’t agree with your view that changes in the monetary base are a reliable indicator of the stance of monetary policy. I use NGDP growth expectations, which fell sharply during 2008-09. You are free to use whatever indicator you wish, the main point is that monetary policy was too tight for the needs of the economy.

    But there is another problem with your argument. As the monetary base was doubled in late 2008, the Fed began paying interest on reserves, at a rate higher than the next best alternative (T-bills). The Fed indicated the purpose of IOR was contractionary. I worked. Even without IOR the base is not a reliable indicator (see Japan), with IOR it’s completely useless.

    You said;

    “How much is enough to not have tight money? It’s easy to keep criticizing, as we will never see a money base on 10 trillion, so your ideas can never be tested out.”

    We could get all the stimulus we need with a dramatically smaller base. Just set a higher inflation target, or do negative IOR. You don’t have to print lots of money.

    How much evidence do we need that there is no such thing as a liquidity trap before people accept that fact?:

    1. FDR’s highly successful program of ending deflation, which raised the WPI by 20% in 1933-34.
    2. The Fed stating that it has plenty of ammo, but it sees no need right now to use all sorts of policy options such as lower IOR, higher inflation targets, level targeting, etc. But will pull them out if needed.
    3. The fact that all sorts of asset markets responded strongly to rumors of QE2, despite the fact that the liquidity trap model predicts no effect on those asset markets.
    4. The fact that Japan tightened monetary policy three times during 2000-06, despite Krugman claiming the BOJ was valiantly trying to create inflation, but just couldn’t get the hang of it.
    5. The fact that just about every highly respected macroeconomist (including Krugman and Bernanke) has published papers on how central banks can boost AD when at the zero bound. It’s a pity Bernanke won’t try the ideas he published–like level targeting.

    There is no doubt that the zero bound has been a difficult issue for real world central bankers. But that’s mostly because they insist on targeting interest rate, which isn’t the smartest thing to do when rates are at zero. Or when rates are above zero.

    You said;

    “Maybe you didn’t notice, but we had massive fiscal stimulus through the tax cuts. And the economy is picking up because of this.”

    You’ve got the timing wrong. The drop in unemployment from 9.8% in November to 9.0% in January was not caused by the tax cut, unless you want to argue for an expectations channel. Of course Keynesians shot themselves in the foot when they argued that Obama couldn’t be blamed for things in early 2009 because the stimulus money hadn’t been spent yet. That implies no expectations channel. And the recent “tax cut” isn’t that big, it was mostly keeping the status quo–only the payroll tax was cut 2%. Market indicators suggest that QE2 strongly raised inflation expectations, even before the tax deal. I admit the tax deal probably helped somewhat, but it wasn’t decisive.

    Luis, Thanks, I can’t answer your question about inflation.

    Nick, Yes, that’s very different. I don’t think monetary stimulus has any long run affect.

  20. Gravatar of W. Peden W. Peden
    5. February 2011 at 15:49

    Mark A. Sadowski,

    I can’t claim to have the figures, but perhaps there’s a degree of “smoothing out” scheduled price increases. Also, while the VAT was delayed until 2011 to allow retailers time to adjust, the UK’s emergency budget itself was back in June 2010.

  21. Gravatar of OGT OGT
    5. February 2011 at 15:49

    One question that your exchange with JustAnotherEcon brings to mind on Britain is what tools the BOE is using in terms of the monetary base and bank reserves and how does that differ from the Fed? It has obviously resulted in more effective monetary growth, if not yet real growth.

    On Britain again, wrote before I think the Cameron gov’t is creating structural friction in the midst of a recession. Most recessions are some mix of real sectoral adjustment (recalc) and monetary disequilibrium brought about inter-temporal demand shifts (AD). I wouldn’t necessarily think the budget shifts are solely responsible for Britain’s step backwards, though, especially since most of the cuts seem to be back loaded into 2014-2015.

    As a thought experiment, if we encountered a country with significant enough supply constraints that real growth was stuck below 1%, how would a Sumner led Central Bank recalibrate NGDP growth? What metric would you use to evaluate a change in trend growth potential Or would you? I would assume that there would be some political pressure to do so (Though 4% inflation is not quite the end of the world is has come to be regarded as).

  22. Gravatar of scott sumner scott sumner
    5. February 2011 at 15:52

    Lorenzo, Most of what you say is true, but I’d put it this way. RGDP is misleading because stimulus doesn’t directly affect RGDP. It affects NGDP, and RGDP responds according to conditions on the supply side.

    Indy, I’m not going to be able to move the profession, others will have to say the same things.

  23. Gravatar of scott sumner scott sumner
    5. February 2011 at 15:57

    Mark, Wasn’t the VAT increased in two steps? The press is blaming the VAT, surely they must have some justification?
    (No sarcasm please.)

    OGT, There is no need to even keep data on inflation and RGDP. The central bank needs to decide what NGDP growth rate it thinks is best, and stick with it regardless of RGDP growth. The problems usually attributed to inflation are actually caused by NGDP growth.

  24. Gravatar of Mark Phariss Mark Phariss
    5. February 2011 at 16:00

    Why 5% NGDP growth? Why not the long-term 6.5%? More generally, how should an NGDP growth rate target be determined?

  25. Gravatar of Mark A. Sadowski Mark A. Sadowski
    5. February 2011 at 16:12

    Scott wrote:
    “Mark, Wasn’t the VAT increased in two steps? The press is blaming the VAT, surely they must have some justification?
    (No sarcasm please.)”

    No, the VAT was increased in one stage from 17.5% to 20% on January 4th. And the UK press articles I’ve reading only talk about it’s effects on inflation going foreward not on the already high inflation rates. That’s why I’m somewhat confused by W. Peden’s statements.

    P.S. The quantity (and quality) of my sarcasm is directly proportional to the amount of sleep I’ve foresaken. I’m not quite there yet.

  26. Gravatar of StatsGuy StatsGuy
    5. February 2011 at 17:56

    Good post.

    I would note that the policy recos Bill W. puts forward seem to be a lot like Joe Gagnon’s. Since I’m a pragmatist, tho, I see more opportunities for govt action that passes the cost/benefits tests. Start with energy taxes and investment credits.

    If you haven’t yet, watch Mervyn King’s speech.

    Also, a thought – the current Muni debt crisis, how much of this is because of the deflection in the NGDP growth rate?

  27. Gravatar of Richard W Richard W
    5. February 2011 at 18:40

    Some of the VAT base effects people may be referring to are related to VAT being cut to 15% Dec. 2008 until January 2009 when it went back up to 17.5%. The problem with VAT explanations is they assume full pass through to consumers. The Office of National Statistics estimate that the temporary cut had only a pass through of one-third.

    What I would like to know is if inflation is such a problem why have we not seen that reflected in the real effective sterling exchange rate? Such a level of negative real rates should lead to a depreciation in the REXR, and it has moved sideways for the last year.

  28. Gravatar of Richard W Richard W
    5. February 2011 at 18:45

    That should be Jan. 2010

  29. Gravatar of Mark A. Sadowski Mark A. Sadowski
    5. February 2011 at 19:04

    Richard W. wrote:
    “Some of the VAT base effects people may be referring to are related to VAT being cut to 15% Dec. 2008 until January [2010] when it went back up to 17.5%. The problem with VAT explanations is they assume full pass through to consumers. The Office of National Statistics [ONS] estimate that the temporary cut had only a pass through of one-third.”

    Much Thanks. Now things are more clear. Still I’d like to see a proper accounting of the overall inflation effects.

  30. Gravatar of justanothereconomist justanothereconomist
    6. February 2011 at 03:27

    Scott-

    “The monetary base also soared from 1930-33, and yet most economists seem to think tight money greatly aggrevated the Great Depression. That suggests that most economists don’t agree with your view that changes in the monetary base are a reliable indicator of the stance of monetary policy.”

    That’s fine, that doesn’t mean it’s right. This is dangerously close to an argument from authority. Monetary tightening caused a recession starting in mid1929, but monetary policy can’t explain the severity of the GD. Monetary policy was much more contractionary by any (non circular) measure in 1920-1921.

    “I use NGDP growth expectations, which fell sharply during 2008-09. You are free to use whatever indicator you wish, the main point is that monetary policy was too tight for the needs of the economy.”

    This is 100% circular. Of course the monetary increase didn’t cause an increase in NGDP expectations, that’s why it’s a liquidity trap. Monetary loosening had little to no effect on NGDP. If you define monetary policy ineffectiveness as tight money, then yes, there is only tight money and no liquidity traps.

    “But there is another problem with your argument. As the monetary base was doubled in late 2008, the Fed began paying interest on reserves, at a rate higher than the next best alternative (T-bills).”
    Yes, but monetary expansions in Japan and the US in the 1930s had weak effects, lots of excess reserves, and no IOR. IOR was a bad move, but there would still be excess reserves without IOR.

    “The Fed indicated the purpose of IOR was contractionary. I worked. Even without IOR the base is not a reliable indicator (see Japan), with IOR it’s completely useless.”

    I’d love to see any model or empirical evidence where an interest rate of 0.25% causes the kind of monetary contraction you imply to it. I can’t think of any that would come even close. It’s a quarter of a point. The fed funds is not actually at zero, it’s more like an eighth of a point, so the gap is really an eighth of a point.

    “FDR’s highly successful program of ending deflation, which raised the WPI by 20% in 1933-34.”

    Svensson optimal escape from a liquidity trap involves a devaluation and repeg such that you commit to a higher price level. This is exactly what FDR did. What I can’t figure out is why the price level didn’t exceed 1929 levels until the War- this is odd in any monetary school I think.

    In any case, a liquidity trap (for me) is that open market operations have a very weak effect on output. Naturally, you don’t need to debunk every conception of liquidity traps out there, but I don’t see this as inconsistent. Now should the USA do this (depreciate and then peg)? This is an interesting question that I haven’t seen discussed much.

    “3. The fact that all sorts of asset markets responded strongly to rumors of QE2, despite the fact that the liquidity trap model predicts no effect on those asset markets.”
    Sure, all those excess reserves strengthen balance sheets.

    Maybe I should stop beating around the bush. If monetary stimulus works, then why is all the money base going into excess reserves? Regardless of theoretical debates about liquidity traps, IOR, etc. how can this affect the economy if the reserves are essentially returned to the Fed? Let’s assume monetary policy is very effective now for a second- how does that happen if monetary base increases simply show up as excess reserves? I mean just in an accounting sense, don’t more reserves have to be loaned out at some point? Even if the effect works through non-banks actors like say industrial corporations. If say GE expects higher inflation, then real interest rates fall and they borrow more. Won’t this necessarily reduce excess reserves if the policy is working? Even with a Brunner style portfolio reallocation, don’t the reserves stop being excess and get translated into extra demand for longer term assets?

    Another way to say it is: how does QE-> money base-> NGDP if not through a fall in excess reserves? I am willing to reexamine by beliefs about your take on macro if QE was not just expanding excess reserves pari passu with money base, but I can’t get over that.

    I have no real right to demand that you address this, but you’re one of the best bloggers about responding to commenters, even with confrontational commenters like me.

  31. Gravatar of bill woolsey bill woolsey
    6. February 2011 at 11:30

    The anonymous comment about how the seventies involved pegging NGDP growth in the face of a supply side squeeze so inflation expectations became unhinged shows a problem. Apparently, some people confuse “NGDP” with “RGDP.” If the Fed tries lowers interes rates to maintain real output growth in the face of a supply shock then inflation expecations might become unhinged and interest rate targeting can lead to disaster.

    But, who favors a target for real output growth? The target is for money expenditures on output. And interest rate targeting is always dangerous.

    Justanothereconomist:

    You really think a liquidity effect exists when open market operations have no effect output? Do you mean money expenditures on output, or the real volume of output.

    Anyway, open market operations in what? BAA corporate bonds? 30 year treasuries?

    Anyway, Scott is defining monetary “tightness” as the quantity of money relative to the demand for money. There is a slight twist because it is in the future (say one year out) and translated into expecations of money expenditure.

    You want to define it by historic standards. Base money is really high compared to what it was in the past. Quasimonetarists say that it is low compared to the demand to hold base money. (and more importantly, it is expected to be low compared to what the demand for base money is expected to be.)

  32. Gravatar of bill woolsey bill woolsey
    6. February 2011 at 11:37

    Scott:

    I think an expansionary monetary policy will raise the CPI, including one that specifically targets a lower exchange rate.

    Can an expansionary fiscal policy raise real output without raising the CPI? Maybe. But if the foreigners lose confidence, then the CPI will skyrocket.

    Wow.. I think I am on to something. This _is_ they way they think. Maybe there were on to something when the claimed that all of us quasimonetarists are are too focused on closed economy models. While I think our approach takes an open economy into account in many explicit ways, it does appear difficult to make sense of what they are trying to do.

    Well, maybe.

  33. Gravatar of Scott Sumner Scott Sumner
    6. February 2011 at 21:32

    Mark Phariss, 5% has been the recent trend rate, it implies about 2.5% inflation, close to the BOE goal.

    Mark, Look at Richard’s comment below, he says rates rose earlier from 15% to 17.5%.

    Statsguy, You said;

    “Start with energy taxes and investment credits.”

    I presume you mean carbon taxes–I agree. Investment credits are inefficient; eliminate taxes on capital–that’s the best way to encourage investment.

    There’s no doubt the the muni debt crisis like all the other debt crises (including Greece) is heavily related to the drop in NGDP. Some are almost totally caused by that, others (like the subprime crisis and the Icelandic/Irish/Greek crises), are partly exogenous.

    Thanks Richard.

    Justanothereconomist, You said;

    “That’s fine, that doesn’t mean it’s right. This is dangerously close to an argument from authority.”

    Not at all, you implied that my “tight money” argument was wacky, because the base soared. I showed it’s not wacky. I almost never argue from authority. My definition of tight money is different from almost everyone else’s definition.

    You said;

    “Monetary tightening caused a recession starting in mid1929, but monetary policy can’t explain the severity of the GD. Monetary policy was much more contractionary by any (non circular) measure in 1920-1921.”

    I strongly disagree. If you are asserting that the monetary base fell more sharply in 1920-21, I agree. As I said, I don’t find that a useful indicator of monetary policy. You may not like my measure, but it is no more circular than those who use M2 or nominal interest rates (two of the most common indicators.)

    You said;

    “This is 100% circular. Of course the monetary increase didn’t cause an increase in NGDP expectations, that’s why it’s a liquidity trap. Monetary loosening had little to no effect on NGDP. If you define monetary policy ineffectiveness as tight money, then yes, there is only tight money and no liquidity traps.”

    Not quite, NGDP expectations aren’t actual NGDP. Of course Krugman made the same argument against Friedman and Schwartz’s M2. That more MB would not boost M2. So I assume you consider M2 a circular monetary indicator?

    Here’s what I am claiming. The government could peg the price of a NGDP futures contract at a price that rose 5% per year. Then make the market by offering to buy and sell unlimited numbers of the contracts at the target price. Is that circular? How’s that different from pegging the price of gold, or foreign exchange, or the fed funds rate?

    I understand that you seem to think changes in the monetary base tell us whether money is easy or tight, but very few people look at the world that way, and for good reason.

    You said;

    “Yes, but monetary expansions in Japan and the US in the 1930s had weak effects, lots of excess reserves, and no IOR.”

    Temporary monetary expansions are not effective, regardless of whether rates are zero or not. Permanent monetary expansions are effective, regardless of whether rates are zero or not. The BOJ was not even trying to inflate, indeed they were specifically trying not to inflate. They reduced the monetary base by 20% in 2006 when there was a threat of inflation rising above zero. The OMOs of 1932 were completely ineffective, because the gold standard made them de facto temporary. As soon as FDR left the gold standard, he had no trouble raising prices and NGDP.

    It’s not about the monetary base, it’s about POLICY. What is the central banks inflation target? Is it doing level targeting? What is it’s goal? These are the key questions. Zero rate bounds reflect cautious central banks.

    Yo uasked why a 1/4 point could be so important. As long as you think that interest rates determine the stance of monetary policy, you’ll be agreeing with Joan Robinson that money couldn’t have been easy in 1923 Germany, after all, nominal rates were high! Interest rates tell us NOTHING about the stance of monetary policy. A quarter point is higher than the rate on T-bills, so banks prefer to hold reserves to T-bills. It might well cause real demand for base money to double. Yet, it’s possible that even at a zero rate hoarding would be an issue, in which case you go negative, or institute a higher inflation target. But a quarter point is certainly enough to make reserves preferable to T-bills. By the way, the IOR rate was about 1% in 2008, when the real damage was done.

    You asked:

    “What I can’t figure out is why the price level didn’t exceed 1929 levels until the War- this is odd in any monetary school I think.”

    What’s the mystery here? The US massively increased its demand for monetary gold during the 1930s. That’s deflationary. The WPI did rise around 30%, but would have risen much more if the Fed hadn’t hoarded a large fraction of all the gold that had ever been mined since about 5000BC.

    You said;

    “In any case, a liquidity trap (for me) is that open market operations have a very weak effect on output.”

    You are mistaken, the criteria is nominal output, not real. Monetary policy has no effect on real output when we are at full employment, but nobody would say we are in a liquidity trap.

    I don’t favor the Svensson approach, level targeting is the best way of avoiding the zero bound problem.

    You said;

    “Sure, all those excess reserves strengthen balance sheets.”

    What does that have to do with asset markets rallying on news of QE2?

    You said:

    “Maybe I should stop beating around the bush. If monetary stimulus works, then why is all the money base going into excess reserves? Regardless of theoretical debates about liquidity traps, IOR, etc. how can this affect the economy if the reserves are essentially returned to the Fed? Let’s assume monetary policy is very effective now for a second- how does that happen if monetary base increases simply show up as excess reserves? I mean just in an accounting sense, don’t more reserves have to be loaned out at some point? Even if the effect works through non-banks actors like say industrial corporations. If say GE expects higher inflation, then real interest rates fall and they borrow more. Won’t this necessarily reduce excess reserves if the policy is working? Even with a Brunner style portfolio reallocation, don’t the reserves stop being excess and get translated into extra demand for longer term assets?”

    I want to get economists to stop thinking about short run changes in the monetary base as the causal factor. In the short run the base should be endogenous, and reflect the demand for base money at a given inflation or NGDP target. That makes me sound post-Keynesian, but the difference is that in the long run the supply and demand for base money determine the nominal aggregates. It’s silly for the Fed to try to boost NGDP by printing money–there’s plenty of money in the economy. They should set a higher inflation target. So why did I favor QE2? Because politically, they can’t do that. So they do QE2 as a disguised inflation target. The Fed is signaling a determination to persist with easier money even when the economy exits the zero bound. The markets pick up that signal and interpret QE2 as a de facto slightly higher inflation target. That’s why the TIPS spread rose on the rumors. I know it sounds very convoluted, and I don’t blame people for being puzzled with my argument. But money is almost infinitely complicated, and getting more so everyday as they add nonsense like IOR.

    If you want to call what I just described a liquidity trap, because OMOs don’t work unless the raise inflation expectations, that’s fine. But just don’t argue the Fed is out of ammo.

    BTW, You are confrontational, but also ask very well informed questions. I no longer have the time to give the sort of full answers I could in a one on one talk for an hour. But there are also my 800 posts if you are really interested in my take on monetary economics. 🙂

    Bill , On the open economy question I think people are mixing up two important but distinct events.

    1. QE2 did slightly depreciate the dollar, and that did cause a little bit of inflation.
    2. World commodity prices are rising mostly on developing country demand, as David Beckworth’s new post shows very clearly.

  34. Gravatar of Mark Phariss Mark Phariss
    7. February 2011 at 09:31

    So is 5% just for the UK; would the US target then be set ~6-6.5%?

    Once the target had been set based on past trend, how would either CB know if/when/how/how much to adjust the target (as trend from that point forward would simply be a result of the target)?

  35. Gravatar of ssumner ssumner
    7. February 2011 at 17:04

    Mark, You initially set it based on past trends, then gradually shift it if you think that trend is not optimal. The trend in the US is also about 5%, at least since the early 1990s.

  36. Gravatar of 英国は総需要を増やすべき、いや待て…減らすべき by Scott – 道草 英国は総需要を増やすべき、いや待て…減らすべき by Scott – 道草
    9. February 2011 at 22:30

    […] Scott // スコット・サムナーのブログから”Britain needs more AD, no wait . . . less AD“ (February 5th, […]

  37. Gravatar of Britmouse Britmouse
    16. February 2011 at 07:11

    Scott, your influence is growing over here.

    Merv King was forced to deny he targets NGDP in the Bank’s inflation conference today, then spent the rest of the time arguing that low rates are necessary for growth and we should ignore inflation. He is becoming a hate figure in the left-wing (“inflation: bad for workers”) and right-wing press (“inflation: bad for savers”), so he must be doing something right.

    http://blogs.wsj.com/source/2011/02/16/it-looks-like-the-bank-of-england-is-targeting-nominal-gdp/

    And better:

    http://blogs.wsj.com/simonnixon/2011/02/16/mervyn-king-stands-up-to-his-critics/

    “Indeed, many of those attacking the BOE today for not doing more to cool the economy are those who normally fill their days attacking the government for leading the country to almost certain ruin. Odd.”

Leave a Reply