A solution without a problem

From the time of the big bang, up to the middle of 2008, there has never been an example of a central bank that tried to create inflationary expectations, but failed.  But what about last year?  When Brad DeLong asked Bernanke why the Fed did not try to create 3% inflation expectations, Bernanke answered that it was a bad idea.  So I don’t think it’s going out on a limb to assume that the Fed is not trying to create inflation expectations.  So the universe’s perfect record still seems intact, no central bank has ever tried but failed to create inflation expectations.  And Chris Sims seems to agree with me:

A bank that has never in the past announced target paths for the inflation rate or its policy interest rate will have much more difficulty if it tries to begin announcing target paths for inflation for the first time when it hits the ZLB. If (as in the case of the US Federal Reserve) it tries to announce paths of the interest rate without any accompanying target path for inflation, the situation is likely to be even worse. The point of the announcement of a commitment to sustained zero interest rates is to generate expectations of increased inflation. Particularly if there is no history of connecting interest rate paths to target inflation paths in inflation reports, the public is likely to be uncertain how to translate beliefs about future interest rates into beliefs about inflation. Indeed, if the bank (like the US Federal Reserve) is reluctant to accompany its stated commitment to sustained low interest rates with an open discussion of its desired path for inflation and of the risk that inflation will temporarily go above target, the public might rationally perceive that the historical commitment to a ceiling on inflation visible in past policy is likely to trump an announced commitment to sustained low interest rates. In this case, the announced interest rate commitment will of course be ineffective.

I mostly agree with Sims, particularly the last line of this quotation.  An interest rate target might make it harder to create inflation expectations.  Instead, you need an inflation target (or NGDP, as I prefer.)

Is it possible that a central bank’s commitment to inflate might not be believed?  Perhaps, if it used inflation targeting.  If it used a price level target I wouldn’t worry about credibility.   But either way aren’t we spending a lot of time and effort modeling a hypothetical problem that has never occurred and probably never will occur?  Especially given the very real problem that monetary policy is now so tight that we are wasting hundreds of billions in fiscal stimulus in a futile effort to accelerate NGDP growth.  That’s a real problem; in the US, in Japan, and in Europe.

HT:  OGT and Mark Thoma.

BTW, I argued in the previous post that the Fed neutralized the fiscal stimulus.  Matt Yglesias recently linked to that post and made this comment:

I think maybe you need an academic’s confidence in his own theories to accept this as a reason to have avoided stimulus back in early 2009. As either a blogger or a policymaker, I’m more comfortable with the idea of joint fiscal and monetary measures to fight a downturn.

I’ve addressed this issue several times, but newer readers may not have seen my response.  So here it is one more time.

1. The point is not to stabilize inflation (or NGDP, which is what I prefer) rather the point is to stabilize expectations of inflation or NGDP.  And that is much less chancy than it seems.  Take inflation as an example.  If the Fed wants 2.5% inflation expectations, it can keep pumping money into the economy until the 12 month TIPS spread is somewhere between 2% and 3%.  The circularity problem will keep this from working perfectly, but it will allow you to avoid large errors.

2.  Stability in inflation or NGDP growth expectations are actually what we care about.  As long as expectations are anchored, any transitory movements in actual inflation or actual NGDP will have little effect on employment.  That’s why Katrina was such a tiny blip on the macro radar screen, despite disrupting Gulf of Mexico oil production for months.  It had almost no impact on 12 month forward NGDP forecasts, and hence firms had little reason to lay off workers.  What you need to avoid at all costs is a situation like late 2008, when 12, 24, and 36 month forward price level and NGDP expectations plunged sharply.  This dramatically depressed the prices of assets (stocks, commodities, real estate.)  When you combine sharply falling asset prices with sticky wages, you are asking for high unemployment.  And conservatives are wrong if they think wage cuts can easily solve the problem.  Even in an economy like ours, where unions are weak, there is far too much wage stickiness to expect employment to hold up during a severe deflationary shock.  Better to keep expectations anchored.  That will keep you out of a macro environment where wage cuts are needed to restore employment.



20 Responses to “A solution without a problem”

  1. Gravatar of Doc Merlin Doc Merlin
    3. February 2010 at 20:45

    Maybe Bernanke believes that expectations are less important than money supply?

    Maybe its part of some grand conspiracy?

    Maybe he said what he said to hide the fact that he was illegally using the fed to bail out financial institutions instead of monetary stimulus?

  2. Gravatar of Don the libertarian Democrat Don the libertarian Democrat
    3. February 2010 at 23:22

    If the Govt is going to borrow in a downturn, why not use it to reinforce the Inflation Expectations that Monetary Policy is intended to produce? In fact, the trick it to take the borrowing, which is needed to deal with real fear and real problems introduced by a serious downturn, and use it in such a way as to reinforce the monetary policy.

    Plus, I think Sims has a good point when he says that’s what people believe and expect concerning fiscal policy in a downturn. If that’s so, instead of fighting against those expectations, why not use them?

  3. Gravatar of scott sumner scott sumner
    4. February 2010 at 05:40

    Doc, Maybe, but the reason doesn’t really matter in the end, only what he said matters.

    Don, You said;

    “Plus, I think Sims has a good point when he says that’s what people believe and expect concerning fiscal policy in a downturn. If that’s so, instead of fighting against those expectations, why not use them?”

    It would be nice if people took Sims’ advice, but the Fed doesn’t seem to agree.

    BTW, Sims points out that in the past the size of the budget deficit is not positively correlated with expected inflation. There is little evidence that bigger deficits create higher inflation expectations. To me, the correlation looks reversed. Inflation expectations rose in the late 1960s and the 1970s, when out debt fell to very low levels. Then inflation expectations fell sharply when Reagan started running big deficits. Sims is looking for solutions to non-existent problems. The Fed knows how to create inflation expectations, they simply don’t want to.

  4. Gravatar of OGT OGT
    4. February 2010 at 06:09

    Prof. Sumner- Are there examples of a CB trying and succeeding in creating inflation expectations, particularly in a Zero rate environment? (This is a legit question, not a gotcha)

    I think Sims has some interesting points, in order to change expectations I think the Fed has be to seen to be doing something now. QE may not be strictly necessary as I believe you claim, but it is visible action, as would a bit of monetizing, which judging from the internet reaction the prospect of which caused many people to freak out.

    BTW, have you written anything about the comparative performance of Mervyn King at the BOE? My impression is that he’s been much more aggressive than his counterparts.

  5. Gravatar of ssumner ssumner
    4. February 2010 at 06:24

    OGT. The US did this successfully in 1933, albeit more through dollar devaluation. The WPI rose about 20% in a near zero rate environment. But of course this is the sort of “price” approach I support.

    My main point is that level targeting would almost certainly be credible, as the central bank would have no where to hide if it failed. Each failure would call for even higher inflation the following year.

    You’d expect these cases to be rare, because zero rates occur when the market understands that policy is too contractionary. So you’d need a sudden change of heart at the central bank (that markets didn’t expect), and that doesn’t occur very often at big institutions.

    The focus of my whole approach is not so much how to get out of deep holes (I agree that central banks that are so conservative they get you in the hole, are unlikely to get you out quickly) rather my focus is on forward-looking policies that avoid the hole entirely. And that is something I believe central banks would be interested in. They don’t like high inflation, but they also didn’t particularly enjoy what happened in 2009.

    I do see the point of your question, and it is a good one. My point is that before we worry about what happens when the public doesn’t believe central bank inflation targets, let’s get the central bank to actually set one.

  6. Gravatar of StatsGuy StatsGuy
    4. February 2010 at 06:38

    “To me, the correlation looks reversed. Inflation expectations rose in the late 1960s and the 1970s, when out debt fell to very low levels. Then inflation expectations fell sharply when Reagan started running big deficits.”

    Scott, that’s because sovereign default risk (which was low in the 80s) was dominated by interest rate changes and currency flows. In the early 80s, interest rates spiked, dollars poured back into the US, foreign countries gladly financed the Reagan Recovery with huge loans, the dollar revalued, and the currency flows dramatically lowered the “market” inflation prediction. This crashed several other economies that had taken out dollar denominated debt to fund consumption and insecure investments (e.g. latin america).

    The story might be different if default risk begins to dominate interest rate risk – but if that happens, we’re in bad bad bad shape. And it’s looking more likely.

    With regard to the current “recovery”, I think people are underestimating the sheer amount of fiscal spending that would need to occur to offset the deleveraging – we still haven’t taken a good look a municipal and state cuts, which are just starting to really hit this year. I think some conservatives are excited about this, because they think it might help cut local govt costs. It’s really just cutting services without cutting wages.

    Meanwhile, we have massive evidence that wage cuts/unemployment, high nominal debt and real interest rates, and lower asset prices are encouraging more homeowners to walk away…


    A fifty fold increase – and those strategic defaults are increasingly being absorbed by FHA. All in the name of keeping imports cheap… This is who we’re fighting against, btw:


    Having said that, I agree with Yglesias – my issue, however, is how money enters the economy. From a political viewpoint, one of the chief challenges to monetary policy is that it benefits the speculators first before the benefits bleed into the broader economy (and it means higher gas prices), and the public has many reasons right now to hate both of these.

  7. Gravatar of 123 – TheMoneyDemand 123 - TheMoneyDemand
    4. February 2010 at 08:39

    Luckilly Bernanke perfectly understands the importance of the desired path of inflation, because of that, he took the initiative of publishing the medium term forecasts of FOMC members.

  8. Gravatar of Marcus Nunes Marcus Nunes
    4. February 2010 at 08:54

    The desired path of inflation is well known (1.5% – 2%). When the Fed publishes the medium forecast of inflation it is releasing info about the likely path of i. If MTIE are greater than 2% everyone will come to expect a rise in i.

  9. Gravatar of vimothy vimothy
    5. February 2010 at 07:09


    How does the Fed expand the money supply?

  10. Gravatar of vimothy vimothy
    5. February 2010 at 07:31

    A further thought: obviously the money supply expanding only creates inflation if it is being spent. So any expansion will only stimulate AD if some other thing causes AD (spending) to rise. The money supply appears to be orthogonal.

    Yet another thought: will more AD even cause inflation if there is a large output gap and idle resources?

  11. Gravatar of ssumner ssumner
    6. February 2010 at 06:48

    Statsguy, Deficits weren’t inflationary in the 1980s. They weren’t in Japan in similar circumstances to ours. The markets don’t expect high inflation now. Could it happen? Sure. But it’s not likely.

    Defaults are driven by low NGDP. Wage cuts aren’t the cause, they are just one more symptom of low NGDP.

    You said;
    “Having said that, I agree with Yglesias – my issue, however, is how money enters the economy. From a political viewpoint, one of the chief challenges to monetary policy is that it benefits the speculators first before the benefits bleed into the broader economy (and it means higher gas prices), and the public has many reasons right now to hate both of these.”

    I don’t like this kind of statement. If read literally, it means you favor a barter economy. I suppose you’d say you weren’t refering to “monetary policy” but rather “expansionary monetary policy.” But the policy I favor, more stable NGDP growth, is a speculators nightmare. They thrive in an economy where NGDP moves around erratically, driving all sorts of asset bubbles.

    123, Do you have a link for the recent forcasts? I saw them quite a while ago, and wondered “why I they forcasting failure?” (Especially one and two year forecasts.)

    vimothy, This confuses nominal and real variables. NGDP can rise from prices increases, not just from more purchases of goods and services. If people like to hold cash balance equal to 10% of income, and you increase cash by 8%, then nominal income will rise by 8% even if people don’t make any more trips to the store.

  12. Gravatar of vimothy vimothy
    6. February 2010 at 09:24

    I agree that NGDP can rise from price increases, but I don’t see how, even if the Fed controls the money supply (I don’t think that it does), a price increase is given from the bare fact of an increase in the money supply alone, absent any change in behaviour. That is, if there is an increase in the money supply but no increase in aggregate demand, there is no inflation. If people simply choose to hold greater cash balances, there is still no upward pressure on prices.

  13. Gravatar of ssumner ssumner
    6. February 2010 at 10:09

    vimothy. The Fed injects money by buying bonds. If people hold the extra cash that would be a change in behavior. It would raise their ratio of cash to income.

  14. Gravatar of vimothy vimothy
    6. February 2010 at 10:27

    Yes, but if they don’t spend it then the too much money isn’t *chasing* the too few goods. It just sits there. Only when actors choose to stop increasing their cash balances does AD / NGDP recover.

    Maybe it’s also worth reiterating Keynes’ insight that money’s elasticity of production means that labour is not diverted into its production when demand for it increases. This demand for money is a black hole for AD. And since total spending is total income, a certain amount of deficit spending to stabilise the system, allowing the private sector to achieve its desired level of spending relative to income, is necessary. Otherwise the self-reinforcing preference for money will continue to drag the economy down.

  15. Gravatar of ssumner ssumner
    7. February 2010 at 10:35

    vimothy, Keynes’ liquidity trap was discredited decades ago. If the central bank targets P or NGDP, the money suppy becomes endogenous.

    If banks hoard reserves the Fed can simply impose an interest penalty on ERs.

  16. Gravatar of vimothy vimothy
    7. February 2010 at 12:09

    How do you discredit a tautology? You’ll have to explain, I’m afraid. Everywhere I look I can see the effects of this problem! It’s an excellent insight.

    The money supply is already endogenous. The CB targets a base rate; liquidity preference of market determines differential; supply of loans infinitely elastic at a given interest rate, and collateral and regulatory requirements. There is no reserve constraint. For instance, see here:

    “… the typical strong emphasis on the role of the expansion of bank reserves in discussions of unconventional monetary policies is misplaced. In our view, the effectiveness of such policies is not much affected by the extent to which they rely on bank reserves as opposed to alternative close substitutes, such as central bank short-term debt. In particular, changes in reserves associated with unconventional monetary policies do not in and of themselves loosen significantly the constraint on bank lending or act as a catalyst for inflation.”


  17. Gravatar of scott sumner scott sumner
    8. February 2010 at 07:27

    Vimothy, What tautology are you referring to?

    I agree the money supply is also endogenous if interest rates are targeted. I think we can now all see the madness of targeting interest rates; once they fall to zero you can no longer use your policy lever. That’s why I oppose interest rate targeting.

    My preference is for NGDP futures targeting. Set thet MB at a level where NGDP is expected to grow along a gradually rising trajectory, say 5% a year.

  18. Gravatar of vimothy vimothy
    8. February 2010 at 08:58

    Keynes’ tautology–the one we’re discussing!

    Demand for money contributes nothing to AD. Money that is not spent does not contribute to total income, because total spending *is* total income, by definition. If it goes somewhere, it must have come from somewhere. How is this disputable–can you explain?

  19. Gravatar of scott sumner scott sumner
    9. February 2010 at 06:28

    Vimothy, I’ve never heard of something called “Keynes’ tautology” I really don’t know what tautology you refer to. Is it C+I+G+NX=Y?

    Don’t confuse stocks and flows. Money is a stock, income is a flow. I agree that in a closed model income equals expenditure equal output, but that tautology doesn’t explain why they change over time.

  20. Gravatar of 123-TheMoneyDemand 123-TheMoneyDemand
    10. February 2010 at 07:05

    The latest FOMC forceasts are here:

    They are from November, and they were approved again in December:
    “Although some of the recent data had been better than anticipated, most participants saw the incoming information as broadly in line with the projections for moderate growth and subdued inflation in 2010 that they had submitted just before the Committee’s November 3-4 meeting; accordingly, their views on the economic outlook had not changed appreciably.”
    FOMC minutes from January are not yet available.

    Current long run target for inflation is 1,7-2. What is important is inflation forecast for 2010-12 which is approximately only 1.0-1.6.

Leave a Reply