Does Bernanke deserve the Bagehot Badge of Honor?

EconTalk Extra
by Amy Willis
George Selgin on Monetary Poli... Philip Tetlock on Superforecas...

How much credit (or blame) should the Fed receive for its response to the 2008 financial crisis? What about former Fed chief Ben Bernanke? This week, EconTalk host Russ Roberts sat down with George Selgin of Cato's Center for Monetary and Financial Alternatives to discuss just that, and we think some of their answers might surprise you...

So please let us know what you thought and learned this week, and let's keep learning from one another. You're our inspiration!


1. Selgin and Roberts both seem to think that the Fed policy of paying interest on reserve held at the Fed neutralized any stimulative effects of monetary policy. What empirical evidence would you need before accepting this claim? How might Bernanke respond?

2. In discussing the failure of Lehman Brothers, Selgin says that the failure was in itself a good thing, "but in the context it was quite harmful." What does he mean by that, and what does his analysis suggest the Fed should have done?

3. Why is Selgin, who identifies as a market monetarist "fellow traveler," ultimately against the Fed's Quantitative Easing? How does his stance compare to that of EconLog's Scott Sumner? (For more on Sumner's thoughts, check his EconLog archive as well as this EconTalk episode from 2013.)

4. Roberts mentioned the possibility of having Bernanke as a guest on EconTalk? If that happened, what would you want Russ to ask him?

Comments and Sharing

TWITTER: Follow Russ Roberts @EconTalker

COMMENTS (14 to date)
Daniel K. Robin writes:

Of course paying interest on reserves inhibits funds from entering the economy. The question is how much. Was the impact of stimulus greater?

Daniel K. Robin writes:

Of course paying interest on reserves inhibits funds from entering the economy. The question is how much. Was the impact of stimulus greater?

Greg G writes:

I don't believe that anyone has a reliable way to quantify how many good loans were prevented from being made by the Fed's policy of paying interest on reserves. I sure don't.

Nevertheless, I believe the effect is much smaller than most people think. There is a big shortage of creditworthy borrowers. There is a big surplus of readily loanable funds. Net interest margin on loans is at an all time low. The bottleneck is too few borrowers, not too few lenders. All interest on reserves can do is reduce the number of lenders.

Our most innovative companies used to be the biggest borrowers. Today they are cash machines more likely to be looking to lend than borrow. Sovereign wealth funds are piling up investable cash. Income inequality is growing with a larger share of income flowing to those inclined to invest it. There is a worldwide glut of liquid hot money looking for creditworthy investments.

At no time in history has more investable money been chasing fewer credit worthy borrowers.

By the way, it's my understanding that the recent rise in the Fed Funds target did come with a quarter point jump in interest paid on reserves.

George Selgin writes:

#1: See

Shawn writes:

George is awesome. Invite him more often!

Scott Sumner writes:

1. I completely agree about the contractionary effect of IOR. As far as the importance of the policy, it is hard to determine for two reasons:

a. It would be hard to determine even if we knew the counterfactual path of all the other policy instruments in the case of no IOR.

b. We don't even know the counterfactual path of other policy instruments.

2. I agree about Lehman. It was the right move not to bail it out, but it harmed the economy in the short run. Had we not done all the earlier bailouts, Lehman would have behaved differently, and been less of a problem. One needs to think in terms of long term policy regimes that are transparent and predictable. That's what we haven't been doing in banking, we've created a highly unpredictable policy regime, which creates both moral hazard and banking instability.

3. I partly disagree on QE. I say partly, because if I had been dictator of the Fed we probably would not have done QE. I'd have switched to NGDP targeting, level targeting. In that case QE probably would not have been needed---or at least much less would have been needed.

In addition, I would have paid no IOR, or perhaps done negative IOR. Again, that might have eliminated the need for QE.

But we didn't do my preferred policies, and in that case I still think QE was better than not doing QE. The bad side effects that some worried about (high inflation or perhaps capital losses to the Fed) seem increasing unlikely to occur. I have to hold my nose when saying this, but I still think QE was better than non-QE.

jw writes:

Scott, I again applaud you for jumping into the fray in a comments section.

Let’s say that in the not too distant future I buy a 10 unit apartment building in NYC. Rents are high and the market price is $2M a unit, so $20M and my credit is good and I have to put $1M down at a zero interest rate in a zero inflation rate environment to make the math simple. Interestingly, the exact same building is for sale in Yonkers, but there the prices are only $100K/unit.

Then a shock occurs. All of a sudden my tenants can only pay enough rent to justify $1M/unit or $10M asset price, and obviously my micro-NGDP spending just went down by 50%. So can I rely on the Fed to reinflate my asset back to $20M so that my micro-NGDP spending gets back to where it was? Or even 5% more if it is a year later?

How is this not anther Greenspan put? How does it not significantly increase moral hazard? If the Fed doesn’t save me, who is on the hook for the $10M risk that just became a loss? Sure, I may lose $1M, but the lender lost $9M and if the Fed bails him out the public has just spent $9M absorbing a socialized risk.

So getting to the title of this thread, let’s compare:

Bagehot: The Fed is the lender of last resort. It does this by:

1. Lending freely in a crisis
2. Self limiting that lending via a high interest rate
3. Lending only on good collateral

Bernanke, Yellen and to a lesser degree Sumner: The Fed is to centrally manage the economy. It does this by:

1. Targeting a single metric, a 2% inflation rate or a 5% NGDP
2. Lending freely in a crisis, not just to provide liquidity, but to rebalance almost all bank balance sheets, leading to a huge increase in excess reserves
3. Lending at zero or even negative rates
4. Lending on newly created or even toxic collateral

I am afraid that if Bagehot makes any sense at all, then all three fail to receive the badge of honor.

The economy is complex and chaotic. Business cycles will never be tamed. There are a huge number of economic variables which are in conflict most of the time. The Fed can only act at a significant lag in data and then only on incomplete data. Their published economic forecasts even six months ahead are horribly inaccurate. They are also hamstrung by unrestrained fiscal policies and deficits that also increase risk. There is tremendous political pressure to bail out favored constituents.

Now, I do not envy the job, nor do I have all of the answers. However, a little humility in what a few people in power can know or, more importantly, control might be in order. The Fed cannot create wealth.

Michael Byrnes writes:

jw, I think you are misplacing Sumner in your apartment example.

Your example seems to fit sort of right in line with the Fed's actions as described (and criticized) by Selgin. Had you been, in 2008, "The Big Bank of jw", the Fed might have said:

Holy Cow! If the BBjw goes under that would be disastrous for the economy as a whole, so we've got to do some kind of intervention to keep BBjw afloat. But we don't want to let this affect the stance of monetary policy and risk inflation, so every dime of emergency liquidity that goes to BBjw will be offset by more contractionary policy elsewhere.

That's exactly the opposite of the type of policy that Selgin and Sumner would prefer to see from the Fed:

If we ease or tighten only as necessary to maintain a stable level of spending economy wide, then BBjw can be allowed to fail without creating any major adverse shocks to the broader economy, inflation will remain stable, and spending by toxic institutions will, by market processes, be reallocated to stronger ones - the exact opposite of what happened under the Fed's sterlized bailouts.

If BBjw was an strong institution suffering from lack of liquidity due to a economy-wide nominal shock, then Fed actions aimed at maintaining stable spending would likely have prevented it from going under in the first place. If BBjw was a weak institution suffering from insolvency due to its own poor decisions, then Fed actions aimed at maintaining stable spending wouldn't save it but might make it easier for BBjw to go through the bankruptcy process and sell off its better assets.

jw writes:


My argument was trying to go beyond just the financial crisis and show that assets and spending are closely linked, besides the theoretical wealth effect, and that inflating NGDP and spending also would inflate assets, increasing risk.

In any case, I still believe that neither monetary nor fiscal stimulus can create wealth. Sumner doesn't have the counterfactual evidence that his NGDP targeting would've prevented the crisis and I don't have the counterfactual evidence that a Bagehotian approach would have produced a sharper, shorter recession and much greater real GDP growth afterwards.

Bagehot was prescient in this 140 year old excerpt from "Lombard Street" ( where he is arguing against using market rates in a crisis and for higher rates:

"The more numerous the demands on the Bank for bullion, and the more variable their magnitude, the more dangerous is the rule that the Bank rate of discount should conform to the market rate. In former quiet times the influence, or the partial influence, of that rule has often produced grave disasters. In the present difficult times an adherence to it is a recipe for making a large number of panics."

Regardless of whether it's Bernanke, Yellen, Sumner or Krugman, and regardless of the metric targeted, when they are advocating for managing a complex economy they are violating Hayek's knowledge problem.

Michael Byrnes writes:

jw wrote:

I still believe that neither monetary nor fiscal stimulus can create wealth.

I think that's the wrong way to look at it. Exchange plays a central role in the creation of wealth, and the quality of the monetary system can facilitate or hinder exchange. I don't think that means that monetary policy creates wealth, but it can certainly affect the creation of wealth.

If there is going to be a Fed with a monopoly on the creation of base money (as is the case in the US), then it has to do something.

Regardless of whether it's Bernanke, Yellen, Sumner or Krugman, and regardless of the metric targeted, when they are advocating for managing a complex economy they are violating Hayek's knowledge problem.

Monetary policy isn't "managing a complex economy", it is managing one product (base money) over which it has a monopoly. Is Apple violating Hayek's knowledge problem when it decides how many iPhones to produce in a give year?

Now, Bernanke's Fed departed from monetary policy to engage isomething like fiscal policy by choosing to bail out Bear Stearns, to let Lehman fail, to bail out AIG, etc., and to, in effect, tighten policy across the rest of the economy in order to offset the largesse it showered on its chosen winners.

That's a vastly different kind of policy than anything Sumner would have advocated.

jw writes:


Of course we agree that a monetary exchange system is required to facilitate wealth creation. Getting that job done right is hard enough.

And as discussed before, the Fed has been assigned the unenviable task of the keeping unemployment low as well.

So we have had low inflation and a declining and now very low unemployment rate for almost a decade. Welcome to Uptopia!

Does anyone think that what it took to get us here is sustainable or even correct?

Bernanke and Yellen have openly stated that their goals were to try and manage the economy. I don't see how targeting an X% NGDP is different. You can't know what X is and you can't know which knobs to turn and when to create X. Focusing on N/GDP is like a company who only looks at their income statement and not their balance sheet, especially the debt side of it.

If Apple builds too many iPhones and loses enough money to bankrupt themselves, they cannot create money to make up the difference.

Kbl writes:

great episode. Very interesting

This was a great episode and George had a remarkable knack for distilling the complex into understandable components. I found his subtle inversion of the Bernanke book title, "The Courage to Act" to the actually much more difficult but more wise in the long-run decision, "The Courage Not to Act" one of the more memorable one-liners of the conversation.

Karol Partyka writes:

I think episode was far to critic about the FED policy.

I think that the interpretation of the interest on reserves as a strengthening of the monetary policy is correct. This policy makes sense if we assume that the ultimate policy goal is to smooth the natural changes in the interest rates.

After the initial crisis the market interest rates plumed down sharply due to decrease of profitable investment opportunities. To keep this process in check the FED improvised and used interest on reserves to make sure no one is borrowing below the price given by the FED.

In this light this policy does not have anything to do with stimulating the economy. It is rather a statement that FED will not accept an interest rate below their target.

Moreover this policy takes the loanable funds from the markets onto the FED balance sheet and uses them to fund the asset purchases. So the FED acts as an Intermediary between Money Market and Capital Market. (look at the liabilities of the FED. the increase of spending is funded by the increased reserves held by the banks)

So FED is basically saying to the market: I will give you r% on your funds and invests the funds into the portfolio of assets.

I am looking forward for the next episode on Monetary policy. This one was surely enjoyable.

I would recommend Perry Mehrling(Columbia University) for the interview on Monetary Policy. He has some deep insights about monetary policy and Bagehot Rule.

Comments for this podcast episode have been closed
Return to top