George Selgin on the Fed
Dec 6 2010

George Selgin, of the University of Georgia, talks with EconTalk host Russ Roberts about whether the creation of the Federal Reserve in 1913 has been a boon or a bust for the U.S. economy. Drawing on a recent paper with William Lastrapes and Lawrence White recently released by the Cato Institute, "Has the Fed Been a Failure?" Selgin argues that the Fed has done poorly at two missions often deemed to justify a Central Bank: lender of last resort and smoother of the business cycle. Selgin makes the case that avoiding bank runs and bank panics does not require a central bank and that contrary to received wisdom, it is hard to argue that the Fed has smoothed the business cycle. Additional topics discussed include whether the Fed has the information to do its jobs well, the role of the Fed in moral hazard, and the potential for the gold standard to outperform the Fed.

Allan Meltzer on the Fed, Money, and Gold
Allan Meltzer of Carnegie Mellon University talks with EconTalk host Russ Roberts about what the Fed really does and the political pressures facing the Chair of the Fed. He describes and analyzes some fascinating episodes in U.S. monetary history, discusses...
George Selgin on Free Banking
George Selgin of West Virginia University talks with EconTalk host Russ Roberts about free banking, where government treats banks as no different from other firms in the economy. Rather than rely on government guarantees to protect depositors (coupled with regulation),...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


Dec 6 2010 at 10:56am

Great as always.

Technical note: for some reason my podcatcher uploaded a corrupt version of the file for this week. I don’t know if that’s a problem on your guys’ end or just my podcatcher screwing up or what, but thought I’d bring it to your attention.

[Hi, Adam. It downloaded correctly in my iTunes this morning. Did you delete it and try downloading it again? I did have some technical problems this morning, but it shouldn’t have affected the downloading. All that happened was that the podcast was delayed an hour because the xml file pointing to the podcast was delayed. The mp3 file was fine. Email me at if you continue to have problems.]

Dec 6 2010 at 4:00pm

So bankruptcy is OK as long as it is pre-planned. OK. How does one do this?

Does this mean that you are saying that if the large banks had been forced into bankruptcy in the last crisis that the process would have taken so long that there would have been systemic failure and thus the government had no realistic option but to bail them out?

Dec 6 2010 at 5:54pm

I would like provide some data that I believe calls into question a couple of Selgin’s assumptions.: 1. Bank runs are mostly about insolvency. 2. Banking panics aren’t that bad for the economy (as long as gov’t doesn’t mess it up).

The following is from Gary Gorton’s book, Slapped by the Invisible Hand. The data is panic date, loss per Deposit dollar, and percent change in Pig Iron (used as a rough measure of economic activity).

Year Loss per $ % chg Iron

1873 .021 -51
1884 .008 -14
1890 .001 -34
1893 .017 -29
1896 .012 -4
1907 .001 -46.5
1914 .001 -47.1

These panics caused there depositors to run on many banks and corresponded with large changes in economic activity, yet the overall system was very solvent. These losses per deposit $ can be compared to the Great Depression where 45 cents of every dollar were lost. To say, “people only run on insolvent banks” seems to be very far from the truth.

Dec 6 2010 at 7:51pm

The relationship between bank regulation and runs is likely related to the fact that in places where banks are regulated people actually know whats going on. If no one is regulating the bank how can ANYONE know if its solvent or not? It stands to reason that regulated banks will get more runs. Its just like places with policeman actually punish more crime.

Dec 6 2010 at 11:07pm

It seems to me that the recent Fed audit supports Selgin’s claims. Isn’t the big news that there was no news? It seems that if the Fed’s position had been correct, we should have seen runs or panic when the lending data was released.

Dec 7 2010 at 2:17am

Also, I think it’s been alleged multiple times on this show that the bailing out of Long-term Capital Management was coercive or semi-coercive, while no evidence has been put forth towards that thesis.

Jimmy Cayne and Bear Stearns walked away from the meeting with the important banks and the N.Y. Fed chair, worried about systemic outfall from LTCM. What retribution faced them? Calling the LTCM bailout a gov’t bailout is a big step of credulity, since all the gov’t did was put all the private actors in a room together.

Dec 7 2010 at 9:58am

Great podcast thanks.

1. I think that the central banks being monopolies slowed evolution of banking.

2. I think ideally currency would be backed only by bank assets Then there could be no runs, then you only need a way to set par which is what gold would provide.

IMO the real problem in the monetary system in feedback. One bank fails contracting the money supply that that weakens all the other banks further contracting the money supply. I think that it would be good if the failure of one bank strengthened all the others. Wouldn’t this be true if money where only backed by bank assets? Is it not true that the in the Scottish free banking era things were moving in that direction? Only 2% gold reserves and the option clause and 30% bank capital seems close to money only backed by bank assets?

Dec 7 2010 at 9:59am

In diversity is strength.

Russ Roberts
Dec 7 2010 at 11:47am


I am always careful (as I was on this podcast) to make it clear that the LTCM rescue was government orchestrated not government funded. How voluntary was the participation of the different musicians–the investment banks that the Fed put in the room to agree to rescue LTCM? Somewhat voluntary. As you point out Bear Stearns chose not to participate. If you read House of Cards by William Cohan, you get an idea of the leverage the Fed had and the pressure they brought to bear (pardon the pun). It was significant but not coercive. Cohan also talks about what it eventually cost Bear Stearn not to participate.

The other significance of the orchestration was that even though it wasn’t a government bailout, it did signal that the Fed’s oversight was way beyond its charter, They were not going to let a hedge fund fail. What the Fed would have done had all the players declined is an interesting but unanswerable question.

Daniel Klein
Dec 7 2010 at 12:32pm

Loved it, thanks. I think that George is one of the best research-lecturers around. I always find his podcasts and video lectures highly enjoyable and rewarding. I loved his Good Money video lecture for the Mises Institute.

I thought people might enjoy this from Adam Smith (WN, 437.15):

“Upon every account, therefore, the attention of government never was so unnecessarily employed, as when directed to watch over the preservation or increase of the quantity of money in any country.”

[Quote, in context, is online at –Econlib Ed.]

Dec 7 2010 at 1:33pm

“I am always careful (as I was on this podcast) to make it clear that the LTCM rescue was government orchestrated not government funded.”

I know you are, and I appreciate that.

I’d like to point out also, though, that it’s not really clear what we mean by bailout in this context either. As the people that bailed out LTCM were their major creditors. That is, they are bailing themselves out. It’s a lot different than the 2007 bailouts that were done by an uninvolved 3rd party. The creditors took over a 90% ownership of the fund. Compare this to another alternative, where LTCM goes into bankruptcy. In that situation, the creditors would get 100% of the firm, but the transactions costs could be incredible. Using the legal process to untangle all of the contracts, it could have been years before we knew who lost in LTCM and how much they lost.

Lee Kelly
Dec 7 2010 at 1:36pm


I think the point is that expedited bankruptcy would more quickly establish how related parties are affected. The incentive for government to intervene or for people to run on solvent firms would be diminished. The Fed could provide liquidity through open market operations to prevent “spillover” from adversely affecting solvent firms.


In the days before central banking and government deposit insurance (and especially in countries with relative free banking like Scotland and Canada), banks openly advertised information about their balance sheets, because depositors demanded to know something about the safety of their money. I suspect that today, if not already in the 19th Century, third parties would publish information about competing banks to help depositors to decide where to place their money. After the FDIC was established, banks in the U.S. merely slapped an “FDIC Insured” plaque by tellers’ desks to assuage depositor concerns about solvency. These days, most depositors don’t even know what “FDIC” even stands for, such has the government corrupted the market forces that would otherwise encourage banks to be more prudent. Lamentably, once this transition was completed, it became necessary for regulators to try and correct for the perverse incentives they had created.

Dec 7 2010 at 2:31pm


It’s quite an odd conclusion you draw, since in the recent crisis the banking panic was in non-FDIC insured “deposits” in the repo market (sometimes called shadow-banking). In fact, regular banking institutions have not been failing nor has deposit insurance been invoked for the most part. That is, it isn’t that insured banks took on lots of risk and gambled with it. It was in fact uninsured banking institutions that took on lots of risk, and when it became apparent that some of those institutions would take losses there was a panic in the repo market and a flight to safety.

George Selgin
Dec 7 2010 at 2:54pm

From reading Charlie’s posts, I can only assume that he is knowledgeable enough to be perfectly aware of the answer to his own reply to Lee, to wit: that TBTF and other implicit guarantees did to non-FDIC insured institutions and their creditors what the FDIC tends to do to ordinary banks and bank depositors, which is to cause them to act as if they would not be bearing the downside risk associated with their chosen investments.

C’mon Charlie: admit that you know this.

Russ Roberts
Dec 7 2010 at 2:56pm


It’s true the latest crisis was a panic in the shadow banking system. But almost all the creditors in that system ended up being made whole by the government (just as many had been in crises of the previous two and a half decades) and losing not a penny. Bear Stearns creditors lost nothing. That encouraged continued lending to Lehman and others with similar balance sheets. More here:

Lee Kelly
Dec 7 2010 at 3:14pm


Where did I draw any conclusion about the present crisis? I was primarily talking about events long ago: banks used to advertise information about their balances sheets to prospective depositors, until things like FDIC insurance made doing so pointless. In any case, I agree with your assessment of the run on shadow banks, so far as it goes, but my answer is the same as Russ’s above. The FDIC may have not have insured “deposits” in the repo market, but the government implicitly did and rescued the creditors as expected. Moreover, the collateral used in the repo market was not kept secret from depositors. It was a combination of implicit government guarantees and, perhaps, negligence or corruption of ratings agencies that made “depositors” relatively unconcerned with the collateral.

The suggestion that shadow banking operated in some pure free market alternate money and banking universe is highly misleading.

Dec 7 2010 at 4:06pm

I want to point out that there is a post of mine dangling in comment land that responds to Russ’s comments about LTCM. I don’t want people to be confused whenever that shows up.

This has always been my favorite argument about the crisis. Russ and Lee are both arguing that the bailout that followed the crisis, actually caused the crisis. That is the future caused the past.

Now, the future can and does cause the past, but it does so stochastically, that is, people make projections about the future. Russ wants to look backwards and say everyone new that they would be bailed out and this caused them to behave very recklessly, but we can actually go back and look at what happened and realize it was far from certain that anyone would be bailed out. We can look at the volatility of stocks in the financial sector and note that shot up to very high levels. We have the derivatives underlying these stocks jumping to high levels, and the option implied market volatility jumping to high levels. We can look at the TED spread, which jumped to very high levels. We have Lehman creditors, in fact, not being bailed out. We have the anecdotal stories of Paulson on his knees in front of congressmen trying to put together a bailout. We had numerous different forms of suggested bailout policies. Yet, now Russ wants to go back and said all of these banks knew they were going to be bailed out, so they took on all this risk. Certainly, they might have thought and hoped there would be some bailout, but the market certainly didn’t put a probability at or close to one of it happening.

Secondly, Russ places heavy weight on the precedent of LTCM as telling banks that when they got into trouble, they’d be bailed out by the government, but LTCM was bailed out by its own major creditors. That is, they were bailing themselves out. It is hardly a precedent for what happened in 2007 and 2008.

Finally, the crisis was already well underway when Bear failed. It’s why Bear failed. The lending to Lehman you speak of was really depositing. And the reason they were depositing was because their deposits were collateralized, but depositors were demanding larger and larger haircuts (more collateral for their cash). So, in fact, the bailout of Bear didn’t get people to continue lending to Lehman. They bailed out of Lehman, and rushed into T-bills and Lehman failed.


Any chance we could get Gary Gorton on to talk about the crisis?

George Selgin
Dec 7 2010 at 6:42pm

I encourage Charlie and other readers to supplement their reading of Gorton’s excellent work on the crisis with a perusal of recent papers by John Taylor and Dan Thornton (of the St. Louis Fed) showing evidence to the effect that the scare tactics used to gain support for TARP and TAF helped to turn what had been a temporary credit market seizure into a general market free-fall. The relevant works are in cited in Thornton’s recent paper here:

Russ Roberts
Dec 7 2010 at 7:17pm


I won’t speak for Lee but you are caricaturing my argument. Read the paper of mine that I cited above or listen to the podcast I did on the topic. I argue that the bailouts of the creditors in the past made creditors less prudent. The actual bailouts of the present suggested that they were right to so anticipate.

George Selgin
Dec 7 2010 at 8:58pm

The TED spread jumped up, and then came down again, after Lehman’s failure. It was only after Bernanke and Paulson did their song-and-dance before Congress, warning that another Great Depression was in the offing if Congress refused to give them a gigantic blank check, that it blasted off and stayed there. See, on this, besides the better known studies of John Taylor, Dan Thornton’s more recent paper at:

Lee Kelly
Dec 7 2010 at 8:59pm


I agree that implicit guarantees of government bailouts were not certain. For one, why was there a run on shadow banking if creditors expected to be rescued by the government? The very promise of a bailout should forestall runs just like the FDIC does for ordinary banking. For Russ’s story to make sense the probability of an individual “depositor” being rescued must have been less than 1, because then everyone has an incentive to run even though a government bailout is expected. It’s the worst of both worlds: taxpayers take the fall for losses and there is still an incentive for “depositors” to run.

Russ’s story assumes that some uncertainty existed about whether, by how much, and who would be bailed out. But even a probable bailout would discourage prudence, and that is exactly Russ’s point.

I was under the impression that Gorton’s preferred solution to this problem is to create something like the FDIC but for shadow banking, though please correct me if I am wrong. I admit this would probably be better than the present circumstance, because it would at least forestall runs, but it would also need to be accompanied by regulations of what shadow banks could use as collateral. For me, this is unacceptable: it recreates all the same perverse incentives and government intervention that plagues commercial banking. The best solution is to restore the discipline of the marketplace to money and banking.

Pedro P Romero
Dec 7 2010 at 11:30pm

Great Podcast! George Selgin’s work on money and banking is truly exceptional.


Dec 8 2010 at 12:52am


Thank you for reading and responding to the comments. I respect that you are intelligent and knowledgeable in this area. It’s just that econtalk has been extremely one-sided in presenting a certain view of bailouts, and as one of the show’s biggest fans I’d really like some of the opposing views of other very smart people represented.

If I may, I’d like to challenge some assertions you made about banking panics in the 19th century. First, you said that runs were typically on insolvent banks, but you didn’t mention the banking clearing houses that were formed to counter panics. For those who don’t know, in order to counter banking panics and forestall banks from having to liquidate all of their loans (at fire sale costs), banks subsumed troubled banks and several institutions would come together and form a overarching clearing house bank. They ceased publishing information about individual banks. They acted as central banks. So if bank runs were primarily about insolvency, why would member banks form clearing houses? It only helps the insolvent banks.

My point is not that insolvency doesn’t matter. A small amount of insolvency is needed to cause the panic, but once the panic is under way the cost is very large and the panic itself can cause insolvency (fire sales of loans means writing down loan values means more insolvent banks). Is it not astounding that we had major banking panics where the total depositor losses when all was said and done were a tenth of one percent? Bankers at the time didn’t seem to think that banking panics were nice things that only punished imprudent or unlucky bankers, so much so that they were willing to take on those liabilities in what amounts to a private market bailout.

This seems to suggest that there were two scenarios. One in which a bank was allowed to fail sold its loans at low prices and cause others to fear their banks would fail causing contagion and a run on many banks. Or a second scenario, where the banks saved the failing bank and allowed for an orderly liquidation over time (a bailout). That these two scenarios can exist seems to offer important insights into the present crisis.

Finally, I’m obviously not convinced by Taylor’s work. Here is one response ( But my favorite is offered by Paul Krugman, “you’re looking at some building, and you hear the fire alarm go off, and smoke starts trickling out the windows. Then a lot of fire trucks and firemen arrive — and only after that do flames start shooting out the top of the building. Clearly, the fire department turned a small problem into a crisis.”

Taylor argues that policy uncertainty created by the Fed and treasury’s unconventional intervention caused the crisis to get very bad. It seems so obvious to everyone but Taylor, that perhaps it was the realization of how bad things were getting and going to be that lead Bernanke and Paulson to such extreme measures. Of course, it is also difficult for U.S. policy uncertainty to explain the global nature of the escalation of the crisis.

Thanks again for reading our comments.


Dec 8 2010 at 7:19pm

Great podcast. Much more interesting than I expected.

Russ – Exceptional interviewing skills in this one, at least for me. You seemed to know exactly when I needed more explanation.

Charlie – I don’t know much about the subject, but were the banking clearing houses actually a central bank or were they more distributed and a private solution to the problem of insolvency that have been crowded out by the Fed?

Dec 8 2010 at 9:02pm

“Charlie – I don’t know much about the subject, but were the banking clearing houses actually a central bank or were they more distributed and a private solution to the problem of insolvency that have been crowded out by the Fed?”

They were a private solution to the problem, but the private solution was to form a central bank. All information about the individual banks was cutoff, and the banks shared the liabilities. People were no longer at risk of their bank failing, but they were at risk of the clearinghouse failing. Clearinghouses even issued currency in the form of clearinghouse notes. That isn’t to say they were affiliated with the Federal gov’t though, if that’s what you mean.

George Selgin
Dec 8 2010 at 9:29pm

You’re wrong about those clearinghouses, Charlie. They most certainly weren’t central banks; their certificates were either backed 100% by gold or were issued on terms that made them strictly temporary. They also never commanded legal tender power. Clearinghouse membership was also voluntary.

Moreover, you fail to note why U.S. clearinghouses engaged in special emergency activities. It was because of the infirmities of the state and national banks, the former prohibited altogether from issuing notes, the latter forced to back notes with increasingly expensive U.S. Treasury securities, and all denied the right to branch (with rare exceptions). Canadian clearinghouses, in contrast, never had to issue emergency currency or otherwise come to the last-resort aid of their members.

In fact, as Elmus Wicker, Milton Friedman, and others have noted, the private U.S. clearinghouses appear in retrospect to have been more reliable devices for addressing the infirmities of the U.S. banking system than the Fed turned out to be. But truly free banking along Scottish-Canadian lines would have been better. That means that the Fed was not even a second but a third best solution to the problem of banking crises here.

Much of this is in my paper with White and Lastrapes. So how ’bout y’all read that so’s I don;t have to repeat it here?

Lee Kelly
Dec 9 2010 at 2:53pm

What is wrong with the comments in this thread? Apparently, George Selgin posted the exact same comment two days apart. Other comments seem to have appeared much later than when written.

[Hi, Lee. I’ve removed George’s duplicated comment. It’s true that there were a lot of delays in getting comments posted in this thread, which stemmed from some of the comments’ accidentally including keywords that caused them to get treated like spam. It took me a while to find them, in part because we’ve had some big spam attacks the last few days. The delays were undoubtedly hard on readers and harder still on the participants in the active discussion. I apologize for the confusion and thank everyone for their tolerance.–Econlib Ed.]

Dec 10 2010 at 1:32pm


I’m not entirely sure if you are disagreeing with Gorton on the facts or making a semantic argument (that’s not a central bank, this is a central bank…). I will look at your paper when I have chance, but lest other observers think I’m making this up, I include the following from Gorton (p. 33) quoting some observers closer to the situation.

“In response to a panic, banks would jointly suspend convertibility of deposits into currency. Coincident with this, clearinghouse member banks joined together to form a new entity…As Swanson (1908) put it, “the virtual fusion of the fifty banks of New York into one central bank.” The clearinghouse would also cease the publication of individual bank accounting information…Finally, banks issued loan certificates, which were first used to replace currency in the clearing process, starting with the Panic of 1857. But in the panics of 1893 and 1907, the clearinghouse issued new money, called clearinghouse loan certificates, directly to the public, in small denominations…Other cities typically followed New York and also issued certificates. Sprague (1903): “The position of the more important New York banks is entirely analogous to that of the great central reserve banks of Europe, such as the Bank of England and the Bank of France. Any unusual demand for actual cash…is certain sooner or later to bring about a withdrawal of money from the New York national banks.””

George Selgin
Dec 10 2010 at 6:25pm

Clearinghouses issued loan certificates because legal restrictions prevented national and state banks from handling peaks in currency demand on their own. Again, the arrangements were temporary, emergency ones, that never vested clearinghouses with paper currency monopolies, and nothing of the sort happened in systems with greater freedom in note issue. Indeed, the legal standing of clearinghouse loan certificates was uncertain. Central banks, in contrast, have always enjoyed unique currency issuing privileges, typically amounting to national monopoly rights.

Dec 10 2010 at 9:11pm

Ah, now I’m certain your point is semantic. I accurately described clearinghouses above, but you object to them being compared to central banks. That’s fine. I don’t care what we call them.

I wish instead that you’d focus on why I brought them up. I re-listened to the beginning of the podcast, and while I don’t think you said anything outright false–bank runs were typically on insolvent banks or banks that were related to insolvent banks–I worry that it conveys the wrong message. Listening it sounds as if the only people that had to worry about runs were reckless or unlucky bankers that made bad loans, but the clearinghouses give evidence that the mere appearance, the mere fear of being a bank that made bad loans could be catastrophic. That is, the most careful banker could be made insolvent by the contagion of a bank run. That is the only reason a careful bank would take on the liabilities of an insolvent bank. It’s systemic risk all the way back in the 19th century!

Dec 11 2010 at 4:19am

I think the problem clearinghouses addressed wasn’t systemic risk, per se, in the sense that otherwise solvent banks would be in risk of illiquidity from contagion, but rather they addressed the problem that a great deal of New York banks would run the risk of illiquidity every year, when money markets tightened as farmers drew on rural banks’ reserves. Since no banks could issue notes in excess of their security reserves, someone had to step in to meet liquidity demand, and the clearinghouses ended up being that somebody.

Without artificial restrictions on currency issue, as there existed in the National Banking System, the clearinghouse scrip wouldn’t be a necessary part of the financial system.

Regardless, I think it’s worth giving George the benefit of the doubt. When he says only actually insolvent banks tended to experience runs, I can only assume he makes this claim based on having studied records of banks from year to year in the time period in question. I doubt he’s just making it all up.

George Selgin
Dec 11 2010 at 5:39am

MichaelM makes precisely the points I am trying to make. Concerning the nature of pre-Fed bank runs, I draw on work by George Kaufman, Charles Calomiris, and Gary Gorton (see in particular Calomiris and Gorton’s 1991 “The Origin of Banking Panics”). During some major panic runs did afflict solvent banks, forcing them to suspend payments temporarily, though only very rarely causing any to fail. In any event the fundamental problem (as Michael indicates) was the inelastic supply of national bank currency and not the inherent inability of banks to fend for themselves without aid from a centralized LOLR.

Dec 11 2010 at 3:23pm

“When he says only actually insolvent banks tended to experience runs, I can only assume he makes this claim based on having studied records of banks from year to year in the time period in question.”

I know it “feels” like he said that (which was part of my point about the podcast), but his actual statements were much more careful. The banks run on were “in danger of failing” or “on the verge of insolvency” or “heavily involved with other banks in that situation” (which could cover a lot of banks).

Also, I agree that clearinghouses weren’t formed to address systemic risk. It’s just that in times of crisis that’s what they happened to do (most of the time they addressed the illiquidity you described).

Next, you two seem more interested in arguing gov’t bad/free banking good, which I don’t have much of an opinion about. I am more interested in understanding 19th century bank runs as a way of learning about the 2007 crisis, during which time there was also a national bank currency.

Finally, MichaelM said, “Regardless, I think it’s worth giving George the benefit of the doubt…I doubt he’s just making it all up.” While I respect George’s expertise in this are, lots of smart people have examined this period and formed a variety of opinions. If anything, George, like many GMU professors, relishes in holding views far away from other academics. I think it is perfectly reasonable to challenge assertions he makes in a respectful manner. If anything, I would think that he would enjoy the chance to defend his views.

George Selgin
Dec 11 2010 at 5:44pm

Charlie is right in supposing that I don’t mind defending my views at all, though w.r.t. the issues raised here, they aren’t especially unorthodox–as perusal of the sources I referred to will show. But Charlie, I’m not sure what you mean by referring to “national bank currency” in 2007. When I refer to such “currency,” I have in mind the paper circulating notes that national banks were once permitted to issue, and which, along with U.S. Treasury notes, were the only kind of paper currency around prior to the Fed’s establishment. Those notes were phased out in 1935. Before then their supply was constrained by bond-backing requirements. Hence occasional currency shortages that clearinghouses attempted to make up using their certificates, and which the Fed was originally supposed to avoid by means of its unique legal privilege of issuing paper notes backed by commercial assets instead of government bonds.

Dec 13 2010 at 6:07pm

I am not a fan of this podcast. Listening to two people who agree with each other trying to argue against their own positions is not very stimulating.

Dec 16 2010 at 1:51am

FWIW I think Charlie highlights some critical incongruities in the free-banking-over-fed-banking argument. If private banks and/or clearing houses are granted license to issue currency that has no backing at their own discretion (e.g., when the intersection of their own past in/actions and their forward-looking expectations presage a crisis), in what sense could such exercises be regarded as “better” (or different in any material+positive sense) than the operations of the Fed? Moreover, in the absence of certain/well-defined reserve requirements for “free” banks (and/or absolute transparency, a.k.a. forfeiture by free bankers of any/all claims to privacy rights, as well as limited liability protections), what qualifications, if any, should an aspiring new financial intermediary and note issuer be expected to embody? Alternately, if privacy rights and/or liability limits are regarded as equally essential to Freedom, then perhaps everyone would/should open their own bank, thereby ushering in a new golden age of “derivative barter.”

Finally, in a world where individual-level knowledge will never be perfect or complete, and speculation is as a consequence going to be a perennial feature of market behavior, how long would non-clearinghouse member institutions be likely to survive in any (even loosely) interconnected financial system? If the most plausible answer is “not very long at all,” then what real benefits would likely result from such a devolution in monetary affairs — what *net* benefits given the certainty of more frequent (though perhaps less correlated) banking crises under a free banking arrangement?

Russ Wood
Dec 18 2010 at 5:32pm


I actually cheered (I was alone in my car) when you asked Mr. Selgin about the problem of determining which institutions were solvent in the latest crisis, due to the problem of valuing MBS and other complex assets, for which the market had essentially frozen. I did not understand Mr. Selgin’s response about Dutch auctions setting prices. A little more detail there would have helped.

Someday, please do an interview on the “living will” planning for bankruptcy of a financial institution. I don’t see how a self-serving liquidation plan would avoid the delays of traditional bankruptcy, since creditors almost have to be given the opportunity to argue over preferences and priorities. So far as I can tell, the only way to avoid those arguments is to have the administration screw the senior creditors, and get a bankruptcy judge to bless the consummation.

George Selgin
Dec 18 2010 at 9:36pm

Concerning Buiter’s idea for using (reverse) Dutch auctions as a means for valuing troubled assets see:

Concerning living wills for TBTF banks:

Russ Wood
Dec 20 2010 at 6:38pm

Thank you, Prof. Selgin.

Dec 27 2010 at 2:20pm

Very interesting theories presented in this podcast.

It leaves me with a few questions:

1) Is Prof. Selgin saying Prof. Friedman was incorrect in his judgement about the fed?

2) What evidence is there that there was not systematic risk in the Continental Illinois (1984) bailout? Prof. Selgin said that only two institutions had as much as half of their assets invested in Continental Illinois. But I would need to know more to know if this tells me something. Is 50% the right number? Is it too high or too low? Was that number at the fingertips of the decision makers at the time, or is this just another Monday morning quarterback situation?

Comments are closed.


About this week's guest:

About ideas and people mentioned in this podcast:Books:


Podcasts, Blogs, and Video:



Podcast Episode Highlights
0:36Intro. [Recording date: December 2, 2010.] Reminder: On Twitter as EconTalker. Topic today: Has the Fed Been a Failure? Could also put an exclamation point instead of question mark--not a traditional way to title a paper. Has the Federal Reserve improved economic performance since its establishment in 1913? Mainstream view at least until recently--lender of last resort; it's eliminated the bank runs and panics of the 19th century. Early years marred by the Great Depression, but once understood, we got the Great Moderation, great prosperity. You and your co-authors disagree. Let's start with the role of the Fed as the lender of last resort. Hasn't that been a good thing? Wasn't the 19th century this tumultuous set of bank panics and bank runs that we don't have to worry about any more? What's the evidence on that issue. The 19th century was marred by bank panics and runs but it's important to realize that the few studies that try consistently to account for panics find that until the Bank Holiday of 1933, the panics were not less frequent or less severe than they had been before the Fed. If anything you find an increase in panics and an increase in severe panics, although it's true that the majority of these occurred during those last three years of the period in question, banking panics of the 1930s. There were about 5 of them in all, including the most severe ever. Afterwards you don't have panics to the same degree as before, but as we argue in our paper, what really brought panics to a halt for a while was first, the Reconstruction Finance Corporation (RFC) which bought tremendous amount of bank stocks in 1933 and the Federal Deposit Insurance Corporation (FDIC) and Federal Savings and Loan Insurance Corporation (FSLIC) by providing deposit insurance eliminate what had become one of the more important proximate causes of the panics before the 1930s. The RFC was a Hoover New Deal agency. Unfortunately, not well-appreciated; far more instrumental than the Fed was in allowing the banking system to be reopened after the National Bank Holiday without renewed withdrawals and panics. Another thing at least as influential--the withdrawal of gold from the banking system was suspended by executive order during the Bank Holiday. The big run that started at the end of February 1933 and continued till the National Holiday was mainly a run on gold because people feared devaluation. Once they couldn't get gold any more that took that motive away.
5:13Bank runs generally. I'm a depositor in a bank; I put my money in the bank, with the promise the bank will do two things: it will pay me some interest along the way and it will give me my money when I ask for it; and when I ask for it, it will be there. The bank in turn, in order to pay that interest, keeps not 100% of the money in the vault. They lend it--they are an intermediary. The risk, though, is if everybody showed up on the same day to get their money out, it wouldn't be there, as George Bailey explains in It's a Wonderful Life. There's no reason to think everybody will show up on the same day unless people fear they won't be able to get their money out. That's where the idea of a bank run or bank panic occurs. Not just a bad piece of luck that everybody at the same time wants their money out; it's some kind of uncertainty or fear that the bank is not solvent. Talk about what happened before FDIC insurance and before the Fed, what in the 19th century led to these runs and how did the market respond? First, important to note that even though a lot of economists are fond of theories that people run on banks just because of fear, the historical record overwhelmingly shows that bank runs, in the absence of insurance, have tended to be runs on banks that were in danger of failing, and not because of the runs but because of bad loans. It wasn't animal spirits. Animal spirits are much overrated as a factor in historical bank runs. Usually people run on banks that were on the verge of insolvency at the time of the run or heavily involved with other banks that were in that situation, indirect basis for fearing they were in trouble. True also in U.S. experience; in the United States runs and panics have been much more common than in other countries, and what's even more revealing is if you look at different banking systems in the 19th century, especially when they tended to be different--today all more or less the same as far as the role of central banks and insurance--you find that runs and panics were a much bigger problem in the more regulated system. And the United States is at the top of that list among industrialized nations at the time. When you go to systems that were relatively less regulated, you find that panics were either unknown or infrequent. Why would regulation encourage panics? Let's take the United States, since regulations there did the most harm. The kinds of regulations I'm talking about included restrictions on branch banking. Most chartered state banks couldn't branch; there were some exceptions. National banks couldn't branch at all until the McFadden Act of 1927 allowed them to branch only in those state jurisdictions where the state banks could branch; so that didn't make much of a difference. The lack of branch banking alone made for a system of very weak banks and lots of them. The argument there would be if I'm a standalone bank and can't branch out geographically, one place, then I am at the mercy of economic forces in that one geographical area. So if there's a factory layoff that significantly affects homeowners in that area, they won't be able to pay their mortgages back; as a result, my mortgages will go bad and I may have trouble honoring my promises. Really comes from the first principle of finance: if you want to have a safe portfolio and avoid insolvency you need to be able to diversify. In those days, banks could not buy and sell loans as they can do today. Weren't many marketable securities out there besides the Federal ones. Banks were heavily dependent on the success on the firms and farms in the local economy. When you had local shocks, they easily succumbed to those shocks and runs were often in response to the perception that banks were in danger of failing or that some local banks had already failed, which was a good predictor that others were in trouble.
10:35What was the argument--issue of branch banking, also called unit banking--what was the public choice or non-public choice argument there? What explains why we passed those restrictions? What was the public argument and the underlying public choice reality? Has to do with the manner in which banking was treated as a business. Unlike other businesses, historically you couldn't start a bank with a general incorporation procedure, at least that was the case universally until 1937, so you had to get a specific act of the legislature in order to get permission to start a bank. Government privilege that was jealously guarded. Every banker who succeeded would line up and oppose the entry of other banks, particularly in their town. Created a tradition of restricting bank charters to unit banks in individual towns, tradition that prevailed until the 1990s. From an established unit banker's point of view it was a no-brainer to want to keep other banks from being able to compete in their territory. They particularly feared invasion by the New York banks because a bank that was in New York had access to that central money market, so had a competitive advantage; could perhaps profitably branch into the rest of the country. Conversely, unit bankers in the country didn't expect they could get a foothold in New York, so opening up wouldn't give them a symmetrical advantage. Are you suggesting that if I lived in Cincinnati, OH, in 1850, I only had one bank to choose from? In major cities over time the number of banks grew. In Ohio you might not have had much of a choice because back then they had a state monopoly. But in some cities you would have had a choice; depended on the attitude of the state legislators. You had a choice, but within those choices they were all where they happened to be physically. All unit banks. So a successful bank that wanted to open a branch on the other side of town to compete with another existing bank wouldn't be able to. Hard time. There was another thing contributing to the support for unit banking, best summarized by Bray Hammond in his book about banking before the Civil War: attitude was banks are dangerous and monopolistic, therefore we should have as few of them as possible. This was the prevailing attitude of the public, not of the bankers themselves. What you've got was a persistent unit banking structure that was only finally completely broken down by the reform of the 1990s. Back to the Fed: as a lender of last resort, the fact that a banker has a window he can go to, to get liquidity, was not an important ender of the bank-run phenomenon. Ended by the provision of deposit insurance, at least as a panic-stopper. What role did the Fed play, what role has it played as a lender of last resort, and what is the justification, benefits or costs? The Fed has acted as a lender of last resort, as an emergency lender, but not as a classical lender of last resort necessarily on those occasions. It did do some last-resort lending during the crises of the early 1930s, in February 1932 for example, but only to member banks. Did not see itself as responsible for other banks, and that left a lot of banks in the cold. Afterwards it tended to play an emergency lending role increasingly starting in the 1970s and massive interventions during the recent crisis. According to the classical lender of last resort doctrine, usually attributed to Walter Bagehot in his book Lombard Street. Bagehot was the second and most famous editor of The Economist magazine back in the 1870s, and there is still a Bagehot column in the magazine. What he said in response to crises in the English banking system was that the Bank of England had a public responsibility to do something it hadn't been doing up to that point, which was to see to it that external drains of specie did not result in its clamping down on credit, but rather that it kept the flow of credit going to the private sector while raising its interest rates to stem the outflow of gold. Specie is gold or silver. Bagehot was making a recommendation for the English banking system, but it's important to note that in doing so he explicitly said the English banking system was deeply structurally flawed because it had a central bank. If it hadn't been for this central bank which was capable of mismanaging money in a way that could ultimately trigger external drains of gold or specie, there wouldn't have been any need for advising it to act in a more publically responsible manner. Bagehot expressly points to the relatively free Scottish system of the time, not as free as it had been before Peel's Act of 1845, as an ideal system and makes it clear that in that kind of decentralized monetary system you don't need a lender of last resort. A lender of last resort according to Bagehot if you are stuck with a central bank is a central bank that lends freely at high rates of interest, but only to solvent firms, during times of crisis.
19:13Idea would be that as a bank, if I made a lot of horrible loans, and as a result when people showed up to get their money I didn't have enough to pay them, and I go to the central bank and say: I just need a little bit of liquidity here because I'm a little short this month. If you are short this month because you made a lot of bad decisions, Bagehot would say: Don't lend them money. If it was just randomly a lot of people showed up this month for their money, but I'm solvent, meaning the expected flow over time--the income--is large enough to cover the outflow, then a loan is okay. Crucial microeconomic distinction. Bagehot had sound microeconomics as a basis for what he was recommending. It was not that the central bank should try to save or bail out insolvent institutions--institutions whose assets before they experienced runs were more than their liabilities. The job of the lender of last resort was to prevent spillovers from the failure of insolvent institutions or international specie outflows from bringing down otherwise solvent institutions by making sure those solvent institutions don't fail for lack of liquid reserves. Lombard Street available on Econlib. If you read the last page you'll see where what he is doing is what we would call today a second best solution to the crises in England. The first best solution would be some form of free banking. The Fed's record has been one of persistently violating Bagehot's rule to lend only to solvent institutions and let the insolvent one's fail so that we have an efficient market situation. A situation where people are responsible for their actions and pay a price if they act recklessly. The reason it's crucial is not only to not throw good money after bad, but you also create a moral hazard problem. That moral hazard problem itself can become one of the more potent causes of irresponsible lending. Look what the Fed has done. It bailed out Franklin National--in the 1970s--and it was insolvent when the Fed bailed them out. Did they know that? They had good reason for deeply suspecting it. Every time the Fed does that, it raises the ante, because of moral hazard; there will inevitably be an even bigger failure from insolvency when this happens. So along comes Continental Illinois, 1984, which was unquestionably insolvent--few could have doubted it at the time. Monolog podcast, by Russ, on this issue. Soften question: challenged you: did they know they were insolvent? The key thing is not--they didn't delve into every bad loan--but they knew they'd made a bunch of bad loans, knew they were in trouble because those loans were not paying as they were supposed to; but in addition, the justification for these rescues--and they were rescues of the creditors of these institutions--was not: Oh, they just need a little bit of time. It was: If they fail, the systemic problem is going to be too horrible. They didn't pretend that they were solvent. They didn't care. They appealed, for the first time in the case of Continental Illinois to what's come to be known as the Too Big to Fail doctrine. This doctrine was supposedly a reason for ignoring Bagehot's microeconomically sound advice because the spillovers would be too big. On the face of it it makes no sense. If big spillovers are what you are worried about, then big support for the still-solvent firms is what's called for. There is no reason to bail out any firm because of big spillover effects it might cause if those spillover effects can be avoided by Bagehot's means of making sure there is ample liquidity for all the still-solvent firms in the market. Moreover, Too Big to Fail has always been a lie in every instance. In the case of Continental Illinois, there were the usual gloom and doom statements by the authorities about what would happen to the rest of the economy if it failed, but eventually it was determined--by George Kaufman--that only two banks would have lost more than half their capital in the aftermath of the failure of Continental Illinois if there had been no support for its creditors. Same thing happened with Long Term Capital Management (LTCM), 1998. There, there was also a private rescue alternative that the Fed authorities short-circuited with a sweeter deal. Although in that case, the Fed did not rescue the creditors of LTCM. They orchestrated a rescue by private firms. Semi-coercive and instead of another rescue that would have taken place. The case of LTCM, Continental Illinois, Franklin National--none of those justified any sort of departure from Bagehot's principles, and I do not believe that the departures in the recent crisis, even more dramatic, have been any more justified.
26:54Challenge you systemic risk argument. You argued that if you are worried about systemic risk, you should offer the liquidity to the solvent banks. But isn't the argument the following? Had Continental Illinois failed and not been bailed out to lenders to Continental Illinois, two other banks would have gone bankrupt. They would have lost half their capital. So the question is: were they solvent or not? You are saying: those two weren't solvent. But isn't it possible that if enough banks had lent money, if they were so interconnected--and this is the argument, I hate it by the way, about the current crisis--so many banks were interconnected, so many had lent or were holding overnight loans, repo loans of Bear Stearns or Lehman Brothers, that had Bear Stearns been allowed to fail, so many creditors would have gone bankrupt, and in turn other guys expecting them to pay wouldn't be able to receive, so the whole system is kind of insolvent. No one bank, but a domino effect. What is your response to that argument? It is an assertion without evidence. We are expected to take it as a matter of faith; as with the boy who cried wolf, we know from past assertions of that type that they have been untrue. It wasn't true of Continental Illinois, of LTCM. Of course, Lehman Brothers was allowed to fail, and the spillover effects from that turned out to be surprisingly minor, as Peter Wallison has shown. Not so sure about that. Hard to know. Unclear. If Bear Stearns had been allowed to go under, the effects of a Lehman collapse would have been different. Once Bear Stearns creditors were rescued, that encouraged people to think that Lehman would be rescued as well. I'm agnostic about what we learned from Lehman's collapse. But you have to concede the fact that there was a great deal of unease when Lehman went under because of, say, money market funds, which had lent Lehman money. Which they shouldn't have! I think they lent them money because Bear Stearns had been rescued, in search of higher yield. In the aftermath of Lehman's collapse, the Fed backstopped the money market funds. So we don't really know the full impact of the Lehman collapse, but I would argue that it doesn't tell us a lot. We have to be very careful here. There are two questions here. One is the Too Big to Fail question of whether the interconnections were such that the insolvency of Lehman meant that other firms were bound to become insolvent if it failed. On that score, I don't think the spillover effects were substantial. The other kind of spillover effect is the kind Bagehot was arguing central banks should prevent, and that is where panic generated by one firm's failure, one insolvency, leads to a run on or shortage of liquidity for otherwise solvent institutions. That's what seems to have happened to all of the money market funds with the single exception of Reserve Primary. Reserve Primary really was insolvent; and Bagehot would say to let them fail, or not fail in this case--wasn't a question of failure--it was a question of breaking the buck. Which means not being able to pay 100 cents on the dollar to the investors. So people who had bought money market funds on the presumption--strange presumption but it's what we have because of habits and the regulatory environment--that your capital is totally safe. Turns out it wasn't. Not a matter of failure because the fund doesn't have any liabilities that are fixed. Just a matter of paying what the assets are worth. No evidence that any other money market fund was going to break the buck, let alone go broke. Therefore, to the extent that you had a liquidity crisis affecting the funds, that's not a reason for not letting Lehman itself fail. How would you categorize the famous meeting where Hank Paulson called in something like 20 of the nation's largest banks, the CEOs, and said: I'm giving you a lot of money right now; you have to take it; here's a little slip of paper to sign for it. He argued at the time that that was a way of reducing any liquidity constraints on these banks. They were all forced to take it on the grounds that that way we wouldn't know if any of them were insolvent or illiquid. The claimed worry was, if it was voluntary, the banks that took it would then be viewed as candidates for failure and there would then be a run on them. What's your take on that episode? I think that's probably correct. What Bernanke and Paulson were worried about was a repeat of what happened back in the 1930s when the RFC was forced to disclose who it was lending to; this supposedly contributed to runs on the banks receiving funds from it. Though actually the evidence for that has been questioned. Appears not to have been so severe as sometimes been claimed. What would Bagehot say about this? I would say same as him. If we have a lot of banks that won't fail and therefore don't want this support, what we have here is an arrangement designed to protect insolvent institutions, partially by making sure no one can tell they are insolvent even as they are being bailed out. I don't see the rationale for that. Wouldn't the claim be there could be an illiquid but solvent institution? An illiquid institution need not fear that there would be any stigma from borrowing from the Fed. Such an institution normally, with adequate liquidity provision to the open market, would also have been able to borrow from those banks that did not suffer from a lack of liquidity. The open market situation after Lehman's failure and after Reserve Primary's breaking of the buck was very bad, and the Fed needed to provide liquidity at that time because interbank liquidity seemed to be seizing up. But it's important to realize that it didn't do that. It was too busy making sterilized loans to insolvent institutions to attend to the illiquidity of the general market.
35:27What should it have done? I think it should have aggressively engaged in open market operations without any sterilization instead of lending to particular insolvent institutions. Should have considered making open market purchases of a broader range than usual. Nothing in Bagehot that says central banks should only buy Treasuries. Bagehot wrote at a time when they made loans instead of engaging in open market operations, but a relaxation of the Treasuries-only policy would have allowed the Fed to purchase other kinds of assets in the open market, private assets, but not for more than they were actually worth, would have been beneficial at that time. Contemporary example: I think the argument the Fed would respond with to your criticism is we were at a point where many of the assets these banks were holding were mortgage-backed securities; their prices were either in free-fall or hard to assess, and therefore it was hard to know who was liquid, illiquid, or who was insolvent. Buiter has an answer to that question as well, proposal he made some time ago in the course of an argument that there is never any need for the Fed to do other than engage in open market operations and then let the market determine where the liquidity went. If you design the open market with a Dutch auction or something like that, you can pretty much allow the auction to determine the value of the securities and do so in a way that will avoid the moral hazard problem. So there is a way that the Fed could have opened up open market purchases even to the very questionable assets without exposing itself or ultimately the taxpayers to risk of loss by buying them at artificially inflated prices. One last comment: my view is that even if all these arguments by those who wanted more intervention--even if they are right, even if there is all this interlocking systemic risk, once you accept that, the evidence for it is weak--it's just asserted because these guys all interact with each other. But suppose it's true. Suppose when the first domino falls you knock them all down. Once you say that, you basically create a welfare program for Wall Street that allows them to perpetually enlarge themselves with borrowed money, inflate the size of their institutions, make riskier and riskier bets and pay themselves well. Until it collapses. But they get rescued and then move on and do it again. It's a vicious cycle, and there's no way it's going to do anything but get worse; or you have to resort to such draconian regulation of the banking system to offset the effects of moral hazard that you end up without a financial system that can cater to the financial needs of an advanced capitalist economy. You've got a public utility. There's another reform that's relevant here. Allowing firms to fail--the spillover effects that that has depend on the way the failed firms are resolved. Bankruptcy in principle, if it's done quickly enough, can actually be a perfectly good way of minimizing the spillover effects. It has been noted that bankruptcy procedures can be too slow-moving to meet the needs of avoiding spillovers in the case of financial failures, but here the English proposal, legislation that requires larger financial firms, the ones that might be considered too big to fail, although nobody from the Fed or elsewhere has told us where the cutoff is, to create so-called Living Wills. These are specific plans that the institutions themselves come up with that would amount to procedures for expediting bankruptcy by specifying exactly what should happen in the event that they are insolvent. Sensible reform and that's what should be done in the United States. Who would think that's a good idea? You and me; maybe some of our listeners. The delay in bankruptcy makes it hard to let firms fail, so why don't we speed that process up or even create some regulations? The alternative would be let's keep the current system that continually bails out people who have made bad loans, and large, rich, giant institutions. How would that play out politically? There's a tendency to think about: what is the best policy. That's not what actually happens. Political sausage factory. In that world, the banks get a lot more say than you and me. There are a lot of economists who benefit from the size and power of the Fed, and they too think it's a good idea to avoid systemic risk and not go through this messy bankruptcy procedure. Political incentives for which policy emerges and which would be the best one; might not be the same thing.
42:03Other aspect of the Fed: the smoother of the business cycle. The claim is the 19th century was full of recessions and depressions, whereas the 20th century--had that little thing in the 1930s, awful, and until recently the business cycle had been tamed. We had the Great Moderation and if you just look at the data, it's clear that the Fed seems to have smoothed things. When I interviewed Milton Friedman in 2006 he argued that yes, the general smoothing of monetary growth that the Federal Reserve had learned to do led to better economic performance. What's your verdict on that? The present state of research is such that we no longer can view things in that more optimistic way. Already when you interviewed Milton Friedman, revisionist work had been done, mostly by Christina Romer and some by other economists, and some since 2006 that has drastically changed our understanding of the volatility of output and the length and severity of recessions before the Fed's establishment in 1913. Bottom line is new research suggests that the post-war business cycle has not been less severe overall than the pre-Fed business cycle. That is, contractions have not been less frequent, shorter, less severe, and output has not been substantially less volatile. That's going to hold post-war period. It doesn't deny the existence of the Great Moderation. Concerning the Great Moderation, however--roughly 1982-2007, lasting through the whole Greenspan era--nobody denies that, but two facts: First, most recent data suggests that the Great Moderation has ended. So, whatever was going on there if it was an improvement in monetary policy, well apparently the moderation is over. That shouldn't surprise anyone. It's just an outlier! One data point! Next time it won't happen because we'll learn from it! The other thing is that since the Great Moderation there has been a large body of research devoted to asking what is the cause of this Great Moderation, or what was the cause? That research has steadily chipped away at the early presumption that improved monetary policy was behind it. That research has either attributed it to a decline in the number of external shocks the economy was exposed to--the Good Luck hypothesis, this was just a period when we were lucky because we hadn't been exposed to as many fundamental disturbances as had been the case in the past. Or the research has said there have been structural changes, but it wasn't improved monetary policy; other changes have been behind the Great Moderation. More dynamism. Globalization, financial innovations, changes in demography, in the management of inventories. Host of factors. Same studies that have identified these factors have also tested the alternative hypothesis of improved monetary policy and reject that as a likely cause. Tough question: if you are out there right now you might think: That's right, the Fed stinks. Sympathetic to that viewpoint, but going to challenge my own biases here. Isn't it true that any of these exercises trying to assess the quality of the Fed's performance--let's admit the fact that the macro economy is a very complex beast, alive, emergent phenomenon, we don't measure it very accurately. Trying to compare the Fed's performance from 1913 to the present, to the pre-1913 period is kind of silly because the data we have available for earlier period is horrible, messy. And what you do is going to confirm your biases. If you like the Fed--I'll play John Taylor here--you could argue that if the Fed had followed the behavior it had followed previously, we wouldn't have had the current crisis; Great Moderation would have continued. Aren't we just indulging our prejudices when we have these kinds of arguments? Sure, prejudice is always a factor in driving the kind of research people do. Don't think you can ever get away from that. People start out with priors, and they pursue those priors. Your question about whether we can make useful comparisons between the system after and before 1913: it's an absolutely essential question. What our research in the paper is trying to do, motivated by the following observation: if you were to ask almost any economist today whether the Fed is worth keeping, whether it's been successful or not, you would have gotten a positive answer, and that answer would have been based on conventional wisdom about the empirical record. Most people are happy to defend the Fed without performing a fancy, elaborate and perhaps even impossible counterfactual exercise to determine whether their optimism or faith in the Fed is justified. What our paper is trying to do is say: The evidence out there showing an initial case showing that the Fed clearly has made things better. Give us a little of that evidence. Evidence based on the most recent statistical studies concerning what has happened to the variability of output, number of banking panics, the length and frequency of recessions, the lender-of-last-resort conduct of the Fed, the price level behavior, the predictability of inflation in the price level--all of those statistics don't show a clear improvement. That is not saying that therefore we know that another system would have worked better. Our limited claim is simply that based on the available studies today, there is no reason for thinking the Fed has been successful or that no system could possibly be better; and we need to do more systematic research about alternatives. We need to stop being complacent about the Fed. It's a burden of proof argument: most people's priors are that the 19th century was hell; since 1913 it's been great except for the Great Depression and the current financial crisis. But the 19th century had some really bad depressions: Depression of 1894, not as good a level of data but really bad depression. What you are suggesting that when you actually look at the data we have, it's the burden of the Fed defenders to make the case. Is that a good summary? Yes, it is.
52:32Want to turn to the alternative arrangements, but want to ask intermediate question first. As economists we are trained to think of the Fed as kind of like the handles that run the taps in a bathtub. Some cold water, some hot water; want tub to go up, you turn the water on; might have a couple of handles to determine if it's hot or cold water, but you have those tools. You want it to go down, you pull the plug to drain it; you can even control the rate at which it drains. A technocratic view of the Fed as a tool. If we step back for a minute--and this is a little bit unpleasant but I think important--if we weren't so tied to the Fed, because we are all interested in monetary policy as economists, why would we ever think it would work that way? I'm not going to suggest that Ben Bernanke or Alan Greenspan are bad men--we are all human beings, with flaws and high IQs, but other than that, normal human beings. Would the model for the Fed be a public choice model? I would never think they are going to run the taps or their other choices like hand out the towels like they do in a textbook that says how it could work. That would be silly. I would expect them to be a political institution, responding to economic actors that have a big incentive to influence their decisions--that would be the banks. Why would I come to the conclusion that what the Fed does most of the time is help the banks rather than the rest of us? How do you think that model, which is very alien to the way most economists think about the Fed--they think of it as this running the bathroom tub metaphor--but if you took a slightly less attractive, almost sinister view, not a conspiracy: not like the banks call up Bernanke or Greenspan or any previous Chairs and say the bag of money will be on the stairs, make sure you are good to us tomorrow. Not that but just that the general forces encourage keeping Wall Street happy, keep the large banks happy because they influence how the President behaves. They make donations to their parties; and as a result the President wants to make the Fed chair happy, and the Congress--etc. Cynical? I think that is a realistic view of the Fed. I tend in my own work to give it the benefit of the doubt and to emphasize more the informational challenges that make it highly difficult, practically impossible to operate the tasks in the ideal manner that textbooks suggest. But I think in fact the reality is mostly the public choice reality you have in mind. Simply not ever going to behave the way the textbooks say it and other banks behave. My focus is on economists. What appalls me is the way economists idealize the Fed and other central banks, writing in textbooks about the way they performed as if it was actually the ideal way they could perform and the only way they could perform--in the absence of information that's not actually available to them. What I really would like to do is change the way economists talk about central banking and monetary policies. I don't expect them to go whole hog for a public choice explanation for these institutions of the sort you find plausible and I also find plausible, but I would at least like to see them handle them according to the way they have performed in reality to some extent--that is, to stop portraying them in textbooks as if they behave ideally when we have all the evidence we need to see they are after all constantly making often very severe mistakes. Brief story about the informational challenge, which is a huge challenge. When I was in chemistry class--maybe 10th or 11th grade, maybe earlier than that--we had a lab room. Two faucets; attached to each was a hose. The hoses were long. You would take the hoses and fill up beakers and other scientific apparatus. There was one kid in the class, mischievous boy, who liked to get the hoses all tangled up in the basin of the sink. You'd be holding your hose; but it turned out that the handle you thought it was attached to was the other one. So the water would be coming out too fast in your hose, so you'd tell the guy or you'd reach yourself and turn it down. Turned out you were turning down the other guy's hose; and when he saw the volume of his was going down, he'd take his handle and turn it up. 'Course he was turning up his hose. We had what we'd call an unstable feedback loop. That's what the Fed has done many times. It thinks it has tight money, so it loosens, when actually it's got loose money. So it's making things worse. Other times, it thinks it's got loose money, so it tightens. Scott Sumner podcast. Informational problem to keep with the bathtub metaphor: you don't realize the tap is open. Murky in the bathroom, lights are off sometimes; you are not even sure whether you are turning on the sink or the bathtub tap. Ambiguity of the data the Fed is using and the way it has to talk about it, as if it totally in charge at all times would be a relevant thing for textbooks to talk about, and nightly news as well. Think if people really understood how the Fed makes its decisions, the finger-in-the-wind manner in which it is forced to determine the course of monetary developments that can have overwhelmingly important effects on the whole economy, they would be shocked. It is not like a smoothly operating machine. This is not a matter of not having the right people in charge. It's just the way the system is. It does perhaps do the best it can; perhaps it doesn't merely cater to special interests in the banking industry or to the whim of the Administration, but the tools it has available to it are deeply flawed and limited in their ability to deliver. We need to think about a better mechanism. Should mention that when I talk about this cynical view of people helping their politically powerful friends, not talking about particular change in the Federal funds rate; I'm thinking about its discretionary activity when it goes out and orchestrates the LTCM bailout or encourages the Treasury to do things that I think are extremely destructive.
1:01:31Alternatives. We could imagine two types. One type is to change the institution utterly: replace the Fed with something else. The other would be to restrain the Fed in particular ways. Let's talk about large changes. What might we do if we eliminated the Fed? What would be the alternative arrangements, if any? What would we need to worry about if we got rid of a central bank? Fine line; impossible to define a line between a reform that gets rid of the Fed and one that doesn't. Becomes a semantic issue. Let me say that if we are to keep a fiat money--a money, the value of which doesn't depend on convertibility in some commodity like gold--then we have to have a means for guaranteeing the scarcity of this stuff which is not naturally scarce. Paper is cheap to print. In a minimal sense, having a fiat money means having a public monetary authority if only in order to safeguard the paper stuff from being augmented to an extent that causes it to steadily lose value. Or unsteadily, erratically. When you have a fiat standard, you need an institution that makes sure you are not printing oodles of the stuff. Minimal task. Need to make sure private agents aren't making replicas of the stuff, either. If that counts as the central bank, then you can't get rid of a central bank. You can only do so according to that very basic definition by having a commodity-based standard, like a gold standard. In that case there is absolutely no need for a central authority to guarantee the value of the monetary unit. As people claim in justifying the Fed, a sudden discovery of gold could arbitrarily change the amount of money in the system, and without a Fed you wouldn't be able to offset that. Very good example of how people, economists included, fail to distinguish between some ideal performance of the central bank, ideally managed paper money, and reality. The relevant comparison should be not between how much the price level might change under a gold standard because of supply shocks and how much it would change if you had an ideal central bank making sure to manage the fiat money correctly. It should be between what the gold standard is likely to generate by way of instability in the price level and what central banks seem likely to generate. The most extreme examples of gold supply shocks that people usually come up with--don't know if they are actually more extreme ones--are the Gold Rush in the late 1840s in the United States and the great price revolution of the 16th century when all the gold was found in the New World and started finding its way to Europe. Leading to inflation. If you take the inflation rates that occurred in those episodes and annualize them, they are trivial. They are less than the Fed's current supposed inflation target of 2%. And much less than the 3% they are talking about for a target of 3% for expected inflation of prices. People don't know their history. There's never been a supply shock that caused an inflation rate under a gold or silver standard that we would consider today a high rate of inflation. What about deflation? Argument would be through private or government decision-making, possibly as Doug Irwin talked about on a podcast recently, research on the way France hoarded gold in other countries. Regarding Doug Irwin's work, let's bear in mind that he's talking about an episode that took place not under the classical gold standard, but under the inter-war gold standard, which was a completely different beast. It was a jury-rigged setup that heavily depended on the central banks of especially former belligerent nations of WWI. It depended on them cooperating with each other by having them treat claims of the Bank of England is if it were cash itself, and allowing those claims to accumulate instead of cashing those claims in for gold. The idea was to allow England, especially, to get away with returning to its pre-war gold parity in the 1920s without having to go through the massive deflation that normally would have been called for to restore that parity because the English money stock and price level were much higher when it restored the parity than they had been for a given level of reserves prior to WWI. Whole set of trying to run a gold standard on the cheap depending on other banks' cooperating; and when the Bank of France stopped cooperating, it all came tumbling down. House of cards. Why did they do that? Why didn't they stick with the pre-war gold standard that didn't require this cooperation? England would bear too high a cost? Why was that? The only way England could have gone to the pre-war gold standard--a classical gold standard mechanism could have been restored in the 1920s, but there were only two ways to do that. First, if you chose to go back to the pre-war definitions, the pre-war convertibility or par values, then big deflation was necessary because pre-war parity required pre-war price levels in the absence of special arrangements like the Gold Exchange Standard aimed at economizing on gold but also subject to collapse. Alternatively, and this would have been the more sensible approach, England and other countries could have let bygones be bygones; could have reestablished a gold standard but with new parities reflecting the reality of new higher price levels since WWI. That would have been the easy solution. What Barry Eichengreen and others say when they claim the gold standard caused the Great Depression, what they mean is that the gold exchange standard collapsed and contributed to the general collapse. When they say that abandoning the gold standard was the only way out, they neglect the alternative of a classical gold standard with modified parities to reflect the reality of inflation post-WWI. What role does the war itself play in this? Why did WWI disrupt? Is that because they couldn't make the normal transfers? What was going on there? The belligerent countries wanted to resort to inflationary finance to cover the expenses of the war. They wanted to spend more than they could spend without printing money. To do that, they had to abandon the gold standard. The United States didn't go off entirely, but it did impose gold embargoes. On the other hand, it didn't inflate as much as the other countries. It did inflate a lot but nothing compared to England or Germany--at least Germany by 1923. Wartime finance that led to the temporary abandonment of the gold standard. A gold standard works best if you are going to actually stick to it. But if you are going to abandon it and have massive inflation--and there are cogent reasons for going back because the gold standard has undeniable advantages--then it may be prudent to go back but with a different exchange value, different par value for gold. This, England was reluctant to do. It wanted to have its cake and eat it, too; other countries to a lesser extent, too. Some good reasons. The ultimate consequence of their attempts to restore the pre-war gold standard parities without having to endure the price level adjustments that a true restoration would have required led to the creation of this house of cards known as the Gold Exchange Standard, that was very prone to collapse. Depended in a way the gold standard did not on central banks' practicing forbearance to the extent they were creditors.
1:13:25In the paper you talk about some of these alternatives; also about keeping the Fed in place, requiring it to follow some arbitrary rule such as Milton Friedman advocated at times, steady growth rate of 2-3% in high-powered money, or a Taylor rule, which would link Federal funds rates to inflation and growth. Problem with all these is they don't really solve the informational challenge. Textbook, don't have reliable measures at a point in time of what the price level is doing or what money is doing. What alternative--putting aside political realities--what would be consistent with the information challenge? The key word here is "centrally." The only way to avoid major errors of monetary change connected with informational deficiencies, if not with political forces, is to decentralize decision-making in the money supply, which means decentralizing the money supply itself. Long been an advocate to providing to the private market place responsibility in providing private money and restoring it not to any one private agency but to a system of private agents. We now have all the eggs in one basket. If the Federal Open Market Committee (FOMC) get it wrong, the whole country gets it wrong. Don't let one firm monopolize those markets, don't let the mistakes of one firm into the mistakes of the whole nation. Also depends on there being incentives for private behavior conducive to all stability; theory suggests it can be. Not going to work with a crummy banking system dependent on government bailouts such as exists today in the United States, so the first step has to be to wean our system off government guarantees, reducing economy's dependence on a central authority as a source of money.