Scott Sumner  

The 2008 transcripts: The real issues

PRINT
CBO's Minimum Wage Study: Deal... Media Bias Yet Again...

The Fed releases the detailed minutes of its meetings with a 5 year lag. Bloggers that follow monetary policy closely have been eagerly awaiting the 2008 minutes, expecting all sorts of revealing (and perhaps embarrassing) quotes. Although I haven't yet had time to read the minutes, the excerpts that I have read were about what I would have expected. We already knew what the Fed did wrong, and what they were thinking when they made their mistakes. I'm sure the minutes will add to that knowledge, but I doubt they will alter the general picture.

Dovish critics of the Fed have pounced on many of the embarrassing statements, which look especially appalling in retrospect. More sympathetic observers talk about the benefits of hindsight, significant data lags, etc. Both have good arguments, but both are overlooking an important point.

The Fed does roughly what a consensus of elite academic economists think they should be doing. I don't recall any significant outrage in the academic community, or even among policy pundits in the press and blogosphere, as the Fed made its crucial mistakes in 2008. Matt Yglesias calls Boston Fed President Eric Rosengren a "hero" for his dissent in September 2008. I have no problem with that characterization, but of course that's exactly the problem, isn't it? No massive bureaucratic machine that depends on well-timed "heroics" will ever be reliable. (This is one area where I won't have to try very hard to convince my Austrian critics!)

The real problem with policy in 2008 was not that the Fed wasn't able to forecast the oncoming disaster; I didn't foresee the severity of the recession until the data began to show the crash of late 2008. Rather the real problem was that policy was far too contractionary even given the real time market data that the Fed had available. This can be illustrated with a "tale of two meetings."

To be fair to the Fed, the unfolding economic collapse took a lot of people by surprise. And the Fed did act fairly aggressively when it got around to acting. But these records are reminders that the human beings pulling the strings of the world's largest economy are no better than most other economists at predicting the future.

"We were seriously behind the curve in terms of economic growth and the financial situation," then-Fed Chairman Ben Bernanke said during an emergency conference call on Jan. 21, 2008. It would not be the last time.

The Fed decided on that call to slash its key interest rate by three-quarters of a percentage point, a shockingly bold move after it had decided not to cut rates during another emergency call just 12 days earlier.

Even with the large cut on Jan. 21, the Fed knew it hadn't done enough. Bernanke suggested that it should have cut by a full percentage point or more. Instead, it waited just nine days and slashed another half-percentage point from its target rate on Jan. 30.


In its December 2007 meeting, the Fed discussed whether to cut rates by 1/4% or 1/2%. They opted for 1/4%. Stocks immediately crashed on the news, indeed fed funds futures probabilities, combined with the more that 2% plunge in the market, suggest the decision depressed stock prices by about 5%. Adding in global markets and you are talking about trillions of dollars. All that wealth riding on one quarter point. That's what Bernanke meant by "behind the curve." Market monetarists would say he was "behind the market."

But I'm going to count Bernanke's January make up call as a limited policy success, despite the fact that the recession began in December 2007. With the soaring global oil prices it would have been difficult to prevent a mild recession in 2008. The Fed's vigorous moves in January kept output flat in the first half of 2008. Now let's look at how the Fed reacted to a similar challenge in September 2008, right after Lehman failed. At the meeting of September 16, 2008, the Fed made one of its most revealing errors:

At that time, many Fed officials were far more worried about inflation risks than about the risk of an economic collapse and depression. The word "inflation" occurs 129 times in the Sept. 16 transcript; the word "recession" was uttered just five times. ("Laughter" is noted in the transcript 22 times.)

Even current Fed Chair Janet Yellen -- who was then the president of the San Francisco Fed and had frequently been prescient about the growing risks to the economy -- argued for standing pat on Sept. 16. She did so despite the fact that she still saw signs of growing economic weakness, including a slowdown in demand for plastic surgery in wealthy San Francisco neighborhoods.

Financial markets kept deteriorating in the days after that meeting, prompting an emergency Fed conference call on Sept. 29. Amazingly, the Fed again decided to take no action.

It wasn't until Oct. 7 that the Fed finally got around to cutting interest rates. Even then, some Fed policy makers wanted to quibble about the Fed's outlook for inflation, refusing to believe that the economy had tipped into a deep hole.


The Yellen vote is a clue to the fact that the real problem in 2008 was not an excess of inflation hawks like Fisher and Plosser, nor a lack of "heroes." The real problem was that the Fed was working with a highly flawed New Keynesian policy approach, loosely related to the Taylor rule (albeit not exactly.)

This is why replacing the FOMC with 12 other "highly qualified" economists in 2008 would not have helped. The only way to have prevented a severe recession in 2008 would have been to have 12 supporters of market monetarism, or at least 12 supporters of NGDP level targeting. Back then the two groups would have had a lot of overlap, but since then many elite NKs have endorsed NGDPLT, including Michael Woodford, Jeffrey Frankel, Christina Romer, etc. The profession is moving in the right direction, and I suspect that even Ben Bernanke would do better the second time around.

How do I know that market monetarists would have gotten it right? Note that on the very day of the September 16 meeting, the meeting at which the Fed refused to cut rates due to fear of "high inflation," the TIPS spreads were showing only 1.23% inflation over the next 5 years, well below target. The Fed should have ignored its own worries about inflation, and instead relied on the wisdom of the crowds. The crowd is not always right, but they are more reliable than the Fed, especially when conditions are changing rapidly. Market participants saw the bottom dropping out of the economy using millions of pieces of highly dispersed information, while the clumsy Fed waited for macro data that comes in with long lags. Hayek would understand.

An even better approach would be to set up a NGDP futures market, and use it to guide a level targeting policy of NGDP growth, perhaps at 4% to 5% per year. As long as NGDP growth expectations are fairly stable, it is unlikely that we would have anything more than a very mild recession. Stable NGDP expectations would have also helped to make asset prices less volatile, and recall that the failure of Lehman was partly due to plunging asset prices. On the other hand our financial system is still riddled with moral hazard, so NGDP targeting is not a cure-all. But if we know that NGDP growth expectations would be stabilized, then it would be easier to say no to bailouts.

BTW, all the points in this post were made in earlier posts over at TheMoneyIllusion, well before the transcripts were released. Market monetarists saw the mistakes occurring in real time; I'd guess that other pundits now bashing the Fed for 2008 were mostly silent as the mistakes were actually occurring.

PS. I do know that the recession was already getting severe by September 16, 2008. That's where level targeting comes in. Had the policy been NGDPLT at 5% growth, then market expectations would have been much more bullish in the summer of 2008 and output would have fallen much less in the third quarter. Thus market monetarist policy proposals could have helped the economy even before the fateful September meeting. I worry that readers will get the wrong message from this post---it's wasn't one fateful decision in September, it was a flawed monetary regime that caused the recession.

PS. Marcus Nunes has a very good post, focusing on policy mistakes made in mid-2008. In the comment section of this post you will find more excellent comments by people like Mark Sadowski and Steve.


Comments and Sharing


CATEGORIES: Finance , Monetary Policy



COMMENTS (21 to date)
John Brennan writes:

Observations/Questions,

I have been reading old Milton Friedman columns from Newsweek--in many cases he argued for greater accountability of the Fed to Congress (and indirectly, the public). Are you familiar with his viewpoint on the 5 year lag of minutes? I find this to be inimical to our constitutional system in that it seems to only serve to provide a lag in illustrating random stupidity. Second, you provide speculation on how a fully Market Monetarist board would have handled the situation. What would you speculated how the situation would have resolved itself without the Fed entirely (another policy option recommended by Friedman on several occasions).

Yancey Ward writes:

I think all of this sounds very familiar. MMers sound like they have all the solutions in the future for the problems of 2008, just like Bernanke had all the solutions ready in 2008 for the problems of 1998-2000. I think in the end, all these fancy proposals overlook one key factor- the thing you are trying to control adapts itself to whatever policy regime you are trying to force onto it.

TravisV writes:

Right. The Fed's ad-hoc backward-looking New Keynesian Taylor rule approach is fatally flawed.

The Fed funds rate is a horrible steering mechanism. It locks up right when it's needed most: surprise falls in NGDP.

One key point that wasn't mentioned: unwillingness to tolerate supply-side inflation was a MAJOR contributor to the crash (and NGDPLT would be very effective at addressing that challenge).

Bob Murphy writes:

Scott,

What would make for an awesome future blog post is if you could find 2 or 3 popular pundits who are currently rolling their eyes about the ridiculous Fed officials who wanted to stand pat during these meetings, but then those some analysts back in 2008 didn't object when the interest rate moves were announced.

Daublin writes:

"No massive bureaucratic machine that depends on well-timed 'heroics' will ever be reliable."

Well said, and well worth pondering. The best we should hope for from the Fed is not to be outrageously awful.

For what it's worth, I think many pundits are overly ignoring the other factors of the last decade besides the fed and even the financial sector. The economy has changed, society has changed, and there is a lot of new oversight on economic activity (Sarbanes-Oxley, Frank-Dodd, the employee mandate). Studying the fed, the money supply, and interest rates strikes me as looking under the lamp post. It's like trying to understand a family dealing with drug abuse by analyzing their 401k portfolio.

Don Geddis writes:

@John Brennan: "...without the Fed...". You've left too much unstated, to actually answer your counterfactual. The Great Recession happened because of a sudden fall in NGDP, aka a sudden rise in the value of the Medium of Account (one of the attributes of money) -- combined with sticky wages(/prices) and sticky debt.

At the moment, the Fed controls the money supply (and thus, NGDP). You can "get rid of the Fed", but to figure out what happens then, you need a model of what happens to the Medium of Account after that. What unit of value do banks and homeowners use, when negotiating a mortgage? What unit do workers and employers use, when signing long-term wage contracts?

If the MOA is unstable, then you get recessions. "Get rid of the Fed" doesn't answer whether the value of the MOA will be stable.

Enial Cattesi writes:

@Ward:

just like Bernanke had all the solutions ready in 2008 for the problems of 1998-2000.

More like 1929 ... Or what he thought the problem of 1929 was.

Mr. Econotarian writes:

That said, they did much better this time than 1928-1929! :)

Michael Byrnes writes:

Yancey Ward wrote:

"I think all of this sounds very familiar. MMers sound like they have all the solutions in the future for the problems of 2008, just like Bernanke had all the solutions ready in 2008 for the problems of 1998-2000. I think in the end, all these fancy proposals overlook one key factor- the thing you are trying to control adapts itself to whatever policy regime you are trying to force onto it."

This is not true, for a couple of reasons:

1. NGDPLT, the preferred policy regime of the market monetarists, would have been every bit as useful in 1978 (when the problem was excessive inflation) as it would have been in 2008. In the late 70s, nominal GDP growth was often over 10%.

2. Scott has several posts at his blog where he points out that Chairman Bernanke did not follow Professor Bernanke's advice. Back in 2003, Professor Bernanke knew how to solve the problems that Chairman Bernanke faced in 2008, but for whatever reason the Chairman did not listen to the Professor.

Eric Falkenstein writes:

I think you give them too much perspicacity. I remember working for Jerry Jordan in 1988 and meeting his friend Beryl Sprinkel, and noting that they--Fed insiders at various times--didn't really understand the downward inflationary trend we were in. Reading history, I note that in the 1960s most economists were surprised by the increase in inflation. It's not obvious in real time what velocity is doing.

TallDave writes:

Failure to fire the harpoon of forward guidance will someday be remembered as this generation's real bills doctrine.

Policy wasn't as suboptimal as TGD, but that's damnation with faintest praise -- how much should we have learned by now?

Scott Sumner writes:

John, It is impossible to know how the situation would have been resolved without the Fed, unless there was some assumption about the type of monetary system that would replace the Fed.

I agree the 5 year lag is unneeded.

Yancey, I agree that "the thing you are trying to control" adapts itself to the policy. Indeed I always assume rational expectations, which is what your statement means.

TravisV, Good point about supply side inflation.

Bob, I'm hoping you can do that; I don't have enough time.

Eric, They don't need to know what's going on with inflation that month, just inflation forecasts (which are available in real time.)

The lag in velocity isn't an excuse for the Fed, as they have the option of level targeting NGDP.

TallDave writes:

Yancey: much of the problems of 2008 are still with us, just as Japan still has many problems of the early 1990s -- hence the popular misconception of the "liquidity trap." MM is the only theory that adequately represents the dynamic response you correctly identify -- the Volcker/Greenspan/Bernanke regime created ever falling inflation expectations, but (like BOJ) viewed them as their own ends and consequently has not managed them well at ZLB.

You're well aware of the flaws of the statist solution, but the right risks losing that argument absent a better riposte than warnings of imminent hyperinflation that never materializes.

TallDave writes:

Yancey: much of the problems of 2008 are still with us, just as Japan still has many problems of the early 1990s -- hence the popular misconception of the "liquidity trap." MM is the only theory that adequately represents the dynamic response you correctly identify -- the Volcker/Greenspan/Bernanke regime created ever falling inflation expectations, but (like BOJ) understood them as their own ends rather than as part of the dynamic feedback cycle identified by Fisher and Friedman, and consequently has not managed them well at ZLB.

You're well aware of the flaws of the statist solution, but the right risks losing that fight absent a better riposte than warnings of imminent hyperinflation that never materializes.

Yancey Ward writes:

Dave,

My problem is that MMers are just assuming their policy would be "better". There is no basis for this belief since there is no actual evidence. Note how Sumner just posits that you set NGDP at a level rate of increase and let the market take care of the rest. Really, how can any one know how the economy would adapt to a regime where NGDP grows at 5%/yr and everyone knows it is going to grow at 5%/yr. Scott just points to periods in the past where it was growing at 5%/yr and assumes those results are the same as his proposed policy for making it a permanent expectation. I think one cannot know what perversions this regime will lead to, but one can be fairly certain that perversions of it will arise and destroy it just as surely as other regimes were destroyed in the past, and when it happens, there will another economist writing about a new policy tool that he promises will work. Rinse and repeat.

Scott Sumner writes:

Yancey, You seem to assume that prior to being adopted, we had no way of knowing how other monetary reforms might fail. But that's not true. We knew what their flaws were, even before they were adopted. And we also knew what problems would be addressed.

Which specific problem can we expect with NGDP level targeting? Just saying that "unforeseen" problems might occur is not enough, anyone can do that with any government policy reform, or indeed any business strategy. Something always might go wrong. But why? What is the flaw in this specific proposal? We want people to begin expecting NGDP targeting, and to change their behavior on that basis. On the other hand with a gold standard you do not want people to recognize the system and act accordingly. It could be disastrous. That's a big difference.

The other flaw with your argument is that we you can't beat something with nothing. Either we never do monetary reform, in which case we'd still have the monetary regime that caused the Great Depression, or we reform the best we can. Are you suggesting we never deviate from the monetary regime that caused this crisis?

Brett writes:

5% nominal GDP growth seems about right, since that would put you at the nominal rate for 1999 (a good year for GDP growth, job growth, and wage growth across the all income quintiles). Could you push it with a 6% target, or would that be risking accelerating inflation?

TallDave writes:

Yancey -- it's not an assumption so much as an epiphany: NGPLT should work better than inflation targeting because it's so clearly a more accurate model at ZLB.

The difference is like Einsteinian physics supplanting Newtonian physics because the former works better at speeds approaching c -- MM just explains our reality better than the alternatives.

What would 2008-2014 look like with NGDPLT? We can't say with certainty, but without the largest fall in NGDP since TGD it's very likely unemployment would be lower, RGDP would be higher, gov't deficits would be lower -- and to boot we would have little if any QE or fiscal stimulus. This was basically the world before the Fed entered ZLB and ran into relativistic effects that confuse its poor Newtonian model of the economy.

CdnExpat writes:

[Comment removed for supplying false email address. Email the webmaster@econlib.org to request restoring your comment. We'd be happy to publish your comment. A valid email address is nevertheless required to post comments on EconLog and EconTalk.--Econlib Ed.]

C. Carafa writes:

One problem with NGDP targeting is that for most of 2007 and 2008 it would have suggested monetary tightening. Inflation was running high, and right until September 2008 future inflation expectations were elevated as well. From the viewpoint of 2007/8, targeting NGDP at, say, 4% would have meant holding rates, or indeed, raising them.

c8to writes:

"With the soaring global oil prices it would have been difficult to prevent a mild recession in 2008." - ahem?

i'm guessing here you mean politically difficult - it would not have been difficult at all to avoid a drop in NGDP...right?

Comments for this entry have been closed
Return to top