Yesterday morning, a Modern Monetary Theorist tweeted part of a 2017 interview he did with Warren Mosler, one of that school's leading proponents and benefactors. As the clip was titled "Against Free Banking: The Liability Side Isn't The Place For Market Discipline," yours truly couldn't resist a look-see.
Having heard just about every conceivable argument against free banking, I frankly did not expect to gain much; but I got even less than I expected. The promised argument "against free banking" consisted of little more than the tired old assertion that, without comprehensive deposit insurance, banks, no matter how sound, are bound to fall victim to runs:
What happens is, you know, somebody starts a rumor in social media that they saw the bank president's wife taking money out of the bank, so all of a sudden there's a panic, everybody tries to take their money out, and the bank fails… And the market in all its wisdom has decided that bank needs to fail. It's always for some reason that's highly suspect… Even the banks that failed in '08 and were liquidated, on a look back, they were not insolvent. It was just liquidity issues…
Apart from the quaint bit about the banker's wife, and misinformed particulars concerning bank failures in 2008, there's nothing special about Mosler's statement, except for its source. Many believe as he does. But most don't pretend to be experts on money and banking, or to have given any serious thought to the topic. What little they know they might have learned by watching It's a Wonderful Life or Mary Poppins (or, in the case of the more studious, both).
In contrast Mr. Mosler, a hedge-fund founder and engineer by trade, who began his career working at a small S&L, does claim to be an expert on money and banking. What's more he commands a large following. So one might expect him to display a little more knowledge of his subject than one might gather from a couple old movies. Alas, he doesn't. And so, instead of enhancing others' understanding of runs, he throws the weight of his authority, such as it is, behind what amounts to little more than a hackneyed folk tale.
That folk tale — call it "The Great American Banking Myth" — has played an outsize role in shaping bank regulations, first in the U.S. itself, and eventually worldwide. And not for the better. On the contrary: we have the tale to thank for much of the rot afflicting so many of the world's banking systems today. Consequently it cries out for a thoroughgoing debunking. Though a single essay can hardly suffice, I hope this one can stand as an abstract.
More to Fear than Fear Itself
There are basically two theories concerning why someone lacking full insurance runs on a bank. One, which informs the Great American Banking Myth, says that runs happen because depositors have no idea whether their banks are sound or not. The other says that they happen because depositors do know. According to the first theory, runs are ignorance based. According to the second, they're knowledge based.
Not surprisingly, evidence from past uninsured banking systems suggests that neither theory is strictly correct. But that evidence also suggests that the vast majority of historical bank runs have been knowledge based. Even runs on sound banks, when they've occurred, have seldom been due to sheer panic. Instead, when sound banks suffered runs, it was often because they had dealings with unsound borrowers or banks that gave their depositors some reason to worry about them.
Because, panic aside, it's rare for all or even most banks in a banking system to be in hot water at the same time, knowledge-based runs seldom translate into system-wide banking panics. Instead, people rush to take money from unsound banks, so as to put it back in sound ones. Because of this, knowledge based runs can actually be beneficial: they can serve as a very effective form of "prompt corrective action" that closes insolvent banks quickly, before they exhaust their capital cushions, and, therefore, well before their mangers have any incentive to play "shoot the moon." The same information based runs that close insolvent banks quickly also see to it that scarce savings are expeditiously re-deployed by solvent ones.
According to George Kaufman, perhaps the foremost authority on this topic, between the end of the Civil War and 1919, there were only nine years in which more than 100 banks failed. The year 1893 witnessed the largest number of failures–326 in all. Yet that was still just 4 percent of the country's grand total of 7614 banks at that time. Despite the lack of deposit insurance, annual depositors' losses for the full period 1875-1933 averaged just .20 percent of the failed banks' total deposits–mute testimony to the fact that, whether owing to runs for some other reason, troubled banks tended to be shut down before they had a chance to become deeply insolvent.
Although the years between the Civil War and 1914 witnessed several serious financial crises–in 1873, 1884, 1893, and 1907–during which many banks suspended payment, none can be said to have involved a system-wide banking panic in the true sense of the term. Instead, regulations that limited banks' ability to issue paper currency while denying them the power to branch led to occasional, severe systemic liquidity shortages. Knowledge of those shortages, in turn, triggered runs. But the fundamental problem wasn't that people ran on U.S. banks. It was that bad regulations gave people a good reason to run on them. Had there been no regulations, there wouldn't have been any reason, and the runs simply wouldn't have happened.
So, at least, says the theory that runs are knowledge-based. But here again, the evidence–this time from Canada–supports it. Canada's banks of the same era were free of the foolish regulations that hampered U.S. banks; and, sure enough, it saw very few bank runs, and suffered from no severe banking crisis. (For a more detailed account of the causes of pre-Fed U.S. banking crisis, and how Canada avoided them, see my Cato Policy Analysis, New York's Bank.)
The Great Depression
While the Great American Banking Myth is partly informed by what people think happened in the decades preceding the Great Depression, it's mainly inspired by the bank failures of the 1930s. To be precise, it's inspired by things people know about those bank failures that ain't so.
Everyone knows, for example, that the Great Depression witnessed a sudden jump in the number of U.S. bank failures. Except it didn't. Instead, as Charles Calomiris explains, and as the chart below (taken from The Motley Fool) shows, the bank failures of the Great Depression were more like a continuation of and acceleration of "the severe banking sector distress that had gripped agricultural regions throughout the 1920s."
During World War I, farmers dramatically increased their farm acreage to take advantage of high wartime crop prices. Both established unit banks and large numbers of new ones formed for the purpose financed the farmland boom. When the bust came in 1920, those same banks started going under. In short, as a reviewer of a detailed study published by the American Bankers Association at the end of the 1920s put it (see Kaufman), the bank runs of that decade were "an effect of banking difficulties rather than a cause." The reviewer added, significantly, that this finding was "contrary to the fixed ideas of the public and even many bankers." It appears that The Great American Banking Myth has been with us for a long time.
Why does this matter? It matters because it suggests–and research (by Peter Temin, Eugene White, and David Hamilton, among others) has confirmed–that many of the banks that failed during the early 1930s were also victims of the land boom and bust, and not of any sheer panic on the part of their depositors. According to Elmus Wicker, 62 percent of the 5000 banks that failed between 1930 and 1932 did so outside of episodes that have been identified as banking crises. Even the Brooklyn-based Bank of United States, the spectacular December 1930 failure of which has been blamed (by Milton Friedman and Anna Schwartz, among others) on a combination of ignorant bank depositors and anti-Semitic Fed officials–appears, in light of subsequent research, to have been in trouble before the run that did it in.
What seems to have been true of the Bank of United States was also true of most of the banks caught up in the wave of panic that claimed many other U.S. banks during November and December 1930. That wave got going as a result of the failure of Caldwell & Company, a Nashville bond-trading house, and its subsidiary, the Bank of Tennessee. "When the market crashed in late 1929," a May 2000 article in the Nashville Post explains,
it set into motion a series of events that would eventually bankrupt Mr. Caldwell's teetering empire. Some of the industrial companies owned by Caldwell & Co. lost huge amounts of money or went bankrupt. The bond issues that Caldwell had bought, but not yet sold to the public, became impossible to sell. A plan to publicly sell a mutual fund comprised of Caldwell's insurance companies became no more than a pipe dream.
Panic, evidently, wasn't to blame for these failures. Nor, according to Elmus Wicker, did the failure of the Caldwell group, including what was then Tennessee's largest bank, cause panic to spread indiscriminately among bank depositors during the weeks that followed. Instead "poor loans and investments in the twenties was [sic] the principal factor contributing to the weakness of the more than 120 banks that closed."
Much the same can be said about the Chicago banking panic of June 1932. According to research by Charles Calomiris and Joseph Mason, although many Chicago banks suffered a decline in deposits, the 40 banks that actually failed all turned out to be insolvent. None, in other words, failed simply because panic-stricken depositors were confused about which banks were, and which ones weren't, in trouble. In short, prior to February 1933, bad fundamentals, rather than fear, appear to have been responsible for most Great Depression bank failures, giving the lie to Mosler's claim, early in the linked clip, that "nobody can tell which bank is safe and which isn't."
And what about the banking crisis of February-March 1933? Here, for once, was a truly nationwide panic, culminating in a nationwide bank holiday that suspended operations at all of the nation's banks. Surely, they weren't all insolvent! Indeed, we know that many reopened afterward. Panic must therefore have broken out.
Indeed, it did. But it was a far cry from the sort of panic that informs The Great American Banking Myth. What people feared most on that occasion wasn't that their banks might go bust. It was that FDR, who would take office in March, might devalue the dollar. The possibility of such a devaluation, which FDR was increasingly loathe to rule out as inauguration day approached, gave people a reason to rush to cash-in their bank deposits for gold, or for Federal Reserve notes they could then redeem in gold, even if they were certain their banks were sound. In short, as Barry Wigmore has put it, the banking panic of 1933 was only superficially a run on the banks: fundamentally it was "a run on the dollar."
It was, indeed, officials at the Federal Reserve Bank of New York, which bore the brunt of public demands for gold and was on the verge of falling short of the Federal Reserve Act's gold-backing requirements, who first pleaded for a national bank holiday, and who were relieved when New York State Governor Herbert Lehman declared a state banking holiday that anticipated the national one. The significance of this becomes all the more apparent when one considers that officials at several of New York’s biggest banks, fearing such a holiday would damage their banks' reputations, argued against it, telling Lehman that they'd "rather stay open and take their beating.”
Bank Runs Since the '30s
Although the introduction of nationwide deposit insurance dramatically reduced the number of bank runs after 1933, because coverage was limited, it didn't eliminate runs altogether. But in practically every case when runs did occur, the affected banks were pre-run insolvent. That was the case with Franklin National Bank, which was run upon during May 1974, with Continental Illinois a decade later, with several Texas banks between1987 and 1989, and with the Bank of New England at the start of 1990. In none of these instances was there anything approaching a systemic banking panic: in all of them most of the depositors who withdrew money from the unsound banks did so in order to place it in others they considered sound.
Nor, despite Mosler's assertions, was the story much different during the 2007-8 financial crisis. According to the FCIC, the crisis saw runs on just three significant deposit-taking institutions: IndyMac, WaMu, and Wachovia. Other sources add Citibank to the list. In every case save Wachovia's, the runs occurred only after evidence emerged of the institutions' deteriorating financial condition; and in every case the institutions were found to have pursued excessively risky strategies or to have been otherwise poorly managed. Wachovia alone appears to have been sound, which is presumably why Wells Fargo was happy to swoop it up without any government encouragement. So even in Wachovia's case "the market in all its wisdom" pretty much got it right.
Furthermore, and once again contrary to what Mosler says, none of the major depository institutions that experienced runs during the Great Financial Crisis was "liquidated." Instead, like most other banks and S&L's that were resolved during the GFC, they were acquired by other firms. Several non-depository mortgage lenders were liquidated; but Countrywide, the only one named by the FCIC as having suffered a major run, was itself acquired by Bank of America. Several insurance companies were also liquidated, as was Lehman Brothers–an investment bank. But insurance companies aren't banks by any reckoning; and Lehman's was "deeply insolvent" when it went under.
After some digging, I've only managed to unearth three cases of depository institutions that were both run upon and liquidated during the GNC. These were: the Cape Fear Bank of North Carolina, Colorado's New Frontier Bank, and Colonial Bancgroup, a bank holding company based in Montgomery, Alabama. Colonial Bancgroup failed after buying $400 million of fake mortgages. Cape Fear Bank failed after lending $18 million to one of its directors, who turned out to be a dead beat. New Frontier, finally, failed because…well, let's just say it was dirty in too many ways to list here.
To claim that these banks merely suffered from liquidity issues, as Mosler does, is not just wrong: it's ludicrous. So, for more than one reason, is Mosler's claim that "the Fed's" "outrageous" and "inexcusable" decision to liquidate them "shows a complete lack of understanding of banking." (For those who don't know, U.S. bank failures are administered not by the Fed but by the FDIC. Consequently the Fed couldn't have liquidated a solvent bank even if it wanted to.)
The Freest Banking Systems were the Safest
It's evident, in short, that Warren Mosler doesn't know much about bank runs. As for free banking, his remarks don't even touch the topic, for to do that they'd have to refer, not to any U.S. experience, but to that of any of several banking systems that really was relatively free of government regulations.
The U.S. itself has never come close to exemplifying genuinely free banking. On the contrary: banking here has always been heavily regulated in one way or another. Nor was the so-called "free banking" era an exception. And it has been largely thanks to stupid U.S. bank regulations that the U.S. banking system has witnessed far more bank runs and failures than the systems of most other industrialized nations. Elsewhere bank failures have been relatively infrequent, bank runs have been rare still, and banking crises have been rarer still.
I've already mentioned the case of Canada, which didn't experience any full-fledged banking crises between 1870 and 1914, while the U.S. faced roughly one crisis every decade. The Scottish banking system was equally crisis free for most of the 19th century. And what was special about the Canadian and Scottish banking systems? Simply, that they were among the least regulated banking systems of their epochs. Although each stood on the firm foundation of English common law, the two systems benefited by being free of the many foolish legal restrictions that rendered so many other nations' banking systems, and the U.S. system in particular, less stable.
Nor were the Scottish and Canadian cases unique. Other relatively free banking systems also had relatively good track-records. Those who really wish to understand the historical causes of banking instability should start by studying the ill-effects of misguided banking regulations, rather than by assuming that banking systems are inherently fragile.
Although I've taken issue with various MMT claims in the past (see, e.g. here and here), I've grown to respect several Modern Monetary Theorists. Far from being ill-informed, people like Eric Tymoigne and Nathan Tankus (the list is by no means exhaustive–these happen to be two whose work I know best) know a lot more than many orthodox economists do about the workings of the U.S. monetary system. Knowing this, I'm not inclined to accuse Modern Monetary Theorists of being ignorant just because I disagree with many of the school's positions and arguments.
But on the subject of bank runs, at least, Warren Mosler shows no signs of being well informed. Although his talk is laced with knowing chuckles, along with disparaging references to "so-called neo-liberals" who are so foolish as to think markets work in banking, his attitudinizing is so much bluff and bluster. The truth is rather that, so far as knowledge of runs is concerned, Mosler is no less destitute than some of the banks whose demise he so heedlessly laments.