Monday, February 20, 2012

Christina Romer: We Need A Regime Change at the Fed

Christina Romer does the Five Books interview and one of her recommended reads is a famous article by Peter Temin and Barry Wigmore titled "The End of One Big Deflation."  This is a great choice since it shows that even in a "balance sheet recession" facing a binding zero percent lower bound, monetary policy can still be very effective by managing expectations.  The key is to radically shift expectations.  Here is Romer discussing the implications of this article for today:
What we learned from the Temin and Wigmore paper is that one way out of a recession at the zero lower bound is by changing expectations. To do that, often what is needed is a very strong change in policy – something economists call a “regime shift”. The most effective way to shake an economy out of a terrible downturn when we’re at the zero lower bound is an aggressive change in policy that makes people wake up, say “this is a new day” and change their expectations. What the Fed has done since early 2009 is much more of an incremental change.
In other words, the Fed has failed to appropriately manage expectations and so we are stuck in a slump.  And I am not convinced that it is now doing any better with its new long-run forecasts of the federal funds rate.  So what in the current environment would rise to the level of a "regime shift"?  What would change expectations enough to catalyze a broad-based recovery in aggregate demand?  Here is Romer's answer: 
I think that what the Fed needs instead is a regime shift. A number of economists have suggested that the Fed adopt a new framework for monetary policy, like targeting a path for nominal GDP. If the Fed adopted such a nominal GDP target, they would start in some normal year before the crisis and say nominal GDP should have grown at a steady rate since then. Compared with that baseline, nominal GDP is dramatically lower today. Pledging to get back to the pre-crisis path for nominal GDP would commit the Fed to much more aggressive policy – perhaps more quantitative easing and deliberate actions to talk down the dollar. Such a strong change in the policy framework could have a dramatic effect on expectations, and hence on the behavior of consumers and businesses.
Such a regime shift would require Bernanke to man up and have his own Volker moment, as previously noted by Romer.  It would be a huge change and that is the point.  A big shock to public expectations, one that would meaningfully change the expected path of future aggregate nominal spending, could be created by a public commitment to a nominal GDP level target.  This is just the medicine the U.S. economy needs right now.  

P.S. No, a nominal GDP level target would not unmoor long-run inflation expectations and it would not depend on a bank lending to work.

P.P.S. Yes, there is evidence for nominal GDP expectations mattering for subsequent NGDP growth.  Here is how I think it would actually unfold. 

12 comments:

  1. Great post. One issue that wont go away though is the argument that catch-up NGDP growth may cause crude prices to explode, canceling or dramatically muting the benefits to RGDP. What do you say to this?

    ReplyDelete
  2. I like the regime change idea--and subsequent regime certainty--but we also have to go to QE for a sustained period, until targets are hit.

    Not (as the Fed did) to QE for a specified period. QE until the target is hit.

    QE takes bonds and gives people money instead. Do it enough, and you monetize debt and boost circulating money. The deleveraging though inflation and actual repayment is good for the economy and taxpayers.

    (Side note: People think of monetizing the debt as helping debtors and thus hurting creditors, and creditors are rich, ergo it is social engineering.

    Actually, since rich people pay income taxes in the USA, they are the ones who will largely have to pay off the national debt (entitlement programs are funded by payroll taxes).

    Ergo, decreasing the national debt through monetization decreases pending tax load on the wealthy. The rig should love the idea, if only they could understand it).

    ReplyDelete
  3. True Anon, but part of the NGDP "mystique" is that it will end the flatline in interest and that will begin to tighten, driving people out of commodities(primary dealers).

    So you may get a spike, then crash of crude oil prices.

    ReplyDelete
  4. It's worth noting that monetary policy today can be effective at the zero lower bound without changing expectations of future monetary policy.

    The counterfactual implies that whatever the central bank purchases today, the recipients of additional cash balances merely wish to hold the money itself. That is, sellers never accept money as an intermediary to realise their demand for real goods and services, but only because they wish to use that money as a store of value.

    To some degree, this may even be true for the financial institution that directly sells to the central bank, e.g. banks selling bonds to the central bank and just holding the excess reserves instead. However, to the extent that such banks had purchased additional bonds in order to sell them to the central bank, the previous holders of the bonds now have larger cash balances as well--there is a chain of transactions where one party receives additional cash balances that they otherwise would not have. For the liquidity trap to literally be true, all of the sellers in this chain must want to hold onto all of the money.

    Furthermore, even if investment demand is so weak that banks cannot find people to lend unwanted reserves to, they can ultimately just spend those reserves to pay employees or other goods and services--people don't refuse the general medium of exchange because interest rates are near zero.

    ReplyDelete
  5. Voodoo.

    I'm not saying voodoo can't work. Maybe it will and maybe it won't. If it works, it will work in the same way similar forms of faith healing, mesmerism and priestcraft can work.

    But it's a stretch to call this "monetary policy." It's just behavioral performance art that happens to use the Fed board of governors as a prop.

    You could possibly get the same result, whatever they might be, if Obama fired his treasury secretary, his economic advisers, and the head of the IRS, and brought in a new team supposedly dedicated to "employment level targeting," even if they did nothing else different.

    Would that be a change in "fiscal policy"?

    ReplyDelete
  6. Anonymous 1:

    What Anonymous 2 said. Also, the voracious appetite for commodities coming out Asia and other emerging markets is important too.

    ReplyDelete
  7. Lee, I agree with your points. Now if the market participants understand this and anticipate it to happen then once again we back to expectations mattering.

    ReplyDelete
  8. Dan, "Vodoo" worked rather well when FDR tried it 1933. It also seemed to work rather well when Sweden tried it during this crisis. And the early evidence from Japan's new inflation target in managing expectations seems to be making some difference. And, as I provided in the link above, there is systematic evidence for expectations mattering. There is a reason why expectations management is a huge part of modern macro--it works.

    ReplyDelete
  9. I'm not doubting that expectations matter David. I'm just suggesting that expectations management has left the realm of monetary policy. The fact that the central banker might be able to play such a role at any given time is a historical contingency based on the central banker's role as "Important Guy Who Exerts Mysterious Power We Don't Understand Over the Business Cycle."

    ReplyDelete
  10. David,

    I suppose my implicit point was that Keynesians and market monetarists mean something slightly different when talking about expectations at the zero lower bound. In short, Keynesians don't think the central bank could increase aggregate demand unexpectedly, but only by promising to pursue more aggressive policy at some time in the future when the zero lower bound is not a concern. Market monetarists, however, also see channels through which monetary policy could work right now, regardless of expectations.

    ReplyDelete
  11. the Fed should stick to inflation targeting; they have done a relatively good job at it. The Fed cant target GDP growth since Money is Neutral in the Long run. Also our employment problem was one of AD but now has become a Supply problem ( since we have seen a shift in the beveridge curve leading to more structural unemployment caused by imobility of labor markets). this imobolity can be fixed by supply-side policies of deregulation, right to work, and persuing policies that would decrease production costs. after we shift LRAS further to the right, we will experience a temporary recessionary gap. At that moment the fed should look to loose monetary policy to increase AD to optimal output and full employment at equalibrium with LRAS.

    ReplyDelete
  12. the Fed should stick to inflation targeting; they have done a relatively good job at it. The Fed cant target GDP growth since Money is Neutral in the Long run. Also our employment problem was one of AD but now has become a Supply problem ( since we have seen a shift in the beveridge curve leading to more structural unemployment caused by imobility of labor markets). this imobolity can be fixed by supply-side policies of deregulation, right to work, and persuing policies that would decrease production costs. after we shift LRAS further to the right, we will experience a temporary recessionary gap. At that moment the fed should look to loose monetary policy to increase AD to optimal output and full employment at equalibrium with LRAS.

    ReplyDelete