There Was No Place Like Canada

Speaking of myths about U.S. banking, another that tops my list is the myth that the Federal Reserve, or some sort of central-bank-type arrangement, was the best conceivable solution to the ills of the pre-1914 U.S. monetary system.

I encountered that myth most recently in reading America’s Bank, Roger Lowenstein’s forthcoming book on the Fed’s origins, which I’m reviewing for Barron’s. Lowenstein’s book is well-researched and entertainingly written. But it also suffers from an all-too-common drawback: Lowenstein takes for granted that those who favored having a U.S. central bank of some kind (whatever they called it and however they chose to disguise it) were well-informed and right-thinking, whereas those who didn’t were either ignorant hicks or pawns of special interests. He has, in other words, little patience with history’s losers, whether they be people or ideas. Like other “Whig” histories, his history of the Fed treats the past as an “inexorable march of progress towards enlightenment.”

Don’t get me wrong: I’m no Tory, and I certainly don’t think that the pre-Fed U.S. monetary system was fine and dandy. I know about the panics of 1884, 1893, and 1907. I know how specie tended to pile-up in New York after every harvest season, and that by the time it got there not one but three banks were likely to reckon it, or make claims to it, as part of their reserves. I also know how, when the harvest season returned, all those banks were likely to try and get their hands on the same gold, and how this made for tight money, if it didn’t spark a full-scale panic. Finally, I know that one way to avoid such panics, on paper at least, was to establish a central bank, or “federal” equivalent, capable of supplying banks with emergency cash when they needed it.

Yet I still think that the Fed was a lousy idea. How come? My reason isn’t simply that the Fed turned out to be quite incapable of preventing financial crises, though that’s certainly true. It’s that there was a much better way of fixing the pre-Fed system. That alternative was perfectly obvious to many who struggled to reform the U.S. system in the years prior to the Fed’s establishment. It could hardly have been otherwise, since it was then almost literally staring them in the face. But it should be equally obvious even today to anyone who delves into the underlying causes of the infirmities of the pre-Fed National Currency system.

What were these causes? Essentially there were two. First, ever since the Civil War state banks were prohibited from issuing circulating notes, while National banks could issue notes only to the extent that they backed them with specified U.S. government bonds. Those bonds were getting harder to come by (by the 1890s National banks had already acquired almost all of them). What’s more, it didn’t pay for National banks to acquire the costly securities just for the sake of meeting harvest-time currency needs, for that would mean incurring very high opportunity costs for the sake of having stacks of notes sitting idle in their vaults for most of the year.

The other, notorious cause of trouble was the fact that most U.S. banks, whether state or National,didn’t have branch networks of any kind. Instead, ours was for the most part a system of “unit” banks. This was so mainly owing to laws that prohibited them from branching, even within their own states. But even had branching been legal, the restrictions on banks’ ability to issue notes would have made it less economical by substantially raising the cost of equipping bank branches with inventories of till money.[1]

That unit banking limited U.S. banks’ ability to diversify their assets and liabilities, and thereby made the U.S. banking system much more fragile than it might have been, is (or ought to be) well-appreciated. Unit banking also encouraged banks to deposit their idle reserves with “reserve city” correspondents, who in turn sent their own surplus cash to New York. The National Banking Acts actually encouraged this practice by letting correspondent balances satisfy a portion of banks’ legal reserve requirements. The set-up kept money gainfully employed when it wasn’t needed in the countryside; but it also made for a mad scramble when cash was needed back home.

Far less well appreciated is how unit banking also contributed to the notorious “inelasticity” of the pre-Fed U.S. currency stock. Before I explain why, I’d better first lay another myth to rest, which is the myth that complaints concerning the “inelasticity” of the pre-Fed currency stock were a hobbyhorse of persons who subscribed to the “real-bills” doctrine — that is, the view that the currency supply could and should wax-and-wane in concert with the total quantity of “real bills” or short-term commercial paper presented to banks for discounting.

It’s true that many persons who complained about the “inelastic” nature of the U.S. currency system, including many who were instrumental in designing (and later in managing) the Federal Reserve System, also subscribed to the real bills doctrine, and that that doctrine is mostly baloney. But that doesn’t mean that the alleged inelasticity of the U.S. currency stock was a mere bugbear. The real demand for currency really did vary considerably, especially by rising a lot — sometimes by as much as 50 percent — during the harvest season, when migrant workers had to be paid to “move” the crops. And U.S. banks really were unprepared to meet such increases in demand by issuing more notes, even if doing so was only a matter of swapping note liabilities for deposit liabilities, owing to the legal restrictions to which I’ve drawn attention. In short, you don’t have to have drunk the real-bills Kool-Aid to agree that the pre-Fed U.S. currency system wasn’t capable of meeting the “needs of trade.”

How, then, did unit banking contribute to the problem of an inelastic currency stock? It did so by considerably raising the cost banks had to incur to redeem rival banks’ notes, and thereby limiting the extent to which unwanted banknotes made it back to their issuers. In a branch-banking system, note exchange and redemption are mostly a local, and therefore cheap, affair; add a few regional clearinghouses to handle items not settled locally, and you’ve got all that’s needed to see to it that unwanted currency is rapidly removed from circulation.

In the U.S., on the other hand, banks had to bear substantial costs of sorting and shipping notes to their sources, or to distant clearinghouses, which costs were made all the greater by the sheer number of National banks — tens of thousands, eventually — and resulting lack of economies of scale. These factors would normally have caused National banks to accept the notes of distant rivals at discounts sufficient to cover anticipated redemption costs, as antebellum state banks had been in the habit of doing. The authors of the 1863 and 1864 National Banking Acts were, however, determined to give the nation a “uniform” currency. Consequently they stipulated that every National bank had to accept the notes of all other national banks at par. That got rid of note discounts, sure enough. But it also meant that National banknotes would no longer be actively and systematically redeemed.[2] As I like to say, any fool can fix most any problem — so long as he ignores the others.

If my dog is limping, and I discover that she’s got a pebble wedged between her paw pads, I don’t think of calling for a team of stretcher bearers: I just pull the pebble out. In the same way any reasonable person, knowing the underlying causes of the infirmities of the pre-Fed U.S. currency system, would first consider removing those causes. And that was precisely what many advocates of currency reform tried to do before any dared to suggest anything like a U.S. central bank. That is, they tried to get bills passed — there must have been at least a dozen of them — calling for some combination of (1) repeal of the bond-backing requirement for National banknotes; (2) allowing National banks to branch, and (3) restoring state banks’ right to issue currency. The restrictions on note issue had, after all, been put into effect for the sake of helping the Union government fund the Civil War — a purpose now long obsolete. The restrictions on branching, on the other hand, were widely understood to be another deleterious consequence of the unfortunate decision to model the National Banking Acts after earlier, state “free banking” laws.

Might deregulation alone, as was contemplated in such “asset currency” reform proposals (so-called because they would have allowed banks to issue notes backed by general assets, rather than by specific securities), really have given the U.S. a perfectly sound and stable currency and banking system? Yes. How can I be so confident? Because it would have given the U.S. a currency system like Canada’s. And Canada’s system was, in fact, famously sound and famously stable.[3]

“Don’t mention the war!” is what Basil Fawlty tells his staff, out of concern for the sensibilities of his German guests. (Basil himself nevertheless can’t help referring to it again and again.) “Don’t mention Canada!” is what a Whig historian of the Fed must tell himself, assuming he knows what went on there, lest he should broach a topic that would muddle-up his otherwise tidy epic. For to consider Canada is to realize that there was, in fact, no need at all for all the elaborate proposals, hearings, secret meetings, and political wheeling-and-dealing, that ultimately gave shape to the Federal Reserve Act, if all that was desired was to equip the United States with a currency system worthy of a nation already on its way to becoming an economic powerhouse. Like Dorothy’s ruby slippers, the solution to the United States’ currency ills had been at hand, or at foot, all along. Legislators had only to repeat to themselves, “There’s no place like Canada,” while taking steps that would tap obstructive legal restrictions out of the banking system.

Of course that didn’t happen, thanks mainly to a combination of banking-industry opposition to branch banking and populist opposition — spearheaded by William Jennings Bryan — to any sort of non-government currency. “Asset currency” was, if you like, “politically impossible.”

So reformers at length turned to the alternative of a central bank. And how was that supposed to work? Though buckets of ink have been spilled for the sake of offering all sorts of elaborate explanations of the “science” behind the Federal Reserve, the essence of that solution, once considered against the backdrop of the “asset currency” alternative, couldn’t have been simpler. It boils down to this: instead of allowing already existing U.S. banks to branch and to issue notes backed by assets other than government bonds, the government would leave the old restrictions in place, while setting up a dozen new banks that would be uniquely exempt from those restrictions. If National banks (or state banks, if they chose to join the new system) wanted currency, but lacked the necessary bonds, they still couldn’t issue more of their own notes no matter what other assets they possessed. But they might now take some of those other assets to the Fed, to exchange for Federal Reserve Notes. The Fed was, in short, a sort of stretcher corps for banks lamed by earlier laws.

To an extent, the more centralized reform resembled an asset currency reform one step removed. But there were two crucial differences. First, by setting the “discount rate” at which they would exchange notes for commercial paper and other assets, the Federal Reserve Banks could either encourage or discourage other banks from acquiring their notes. Second, because member banks could count not just gold and greenbacks but Fed liabilities as reserves, the Fed’s discount rates influenced the overall availability of bank reserves and, hence, of money and credit. These differences, far from having been innocuous, were, as we now realize, portentous.

Still the Fed did have one incontestable advantage over previous reform proposals. For it alone was politically possible. It alone was a winning solution.

But the fact that the Fed won in 1913 doesn’t mean that other, rejected options aren’t worth recalling. Still less does it warrant treating the Fed as sacrosanct. History isn’t finished. Just a few years before the Federal Reserve Act was passed, most people still believed that Andrew Jackson had put paid once and for all to the idea of a U.S. central bank. Today most people still consider the Federal Reserve Act the last word in scientific monetary control. As for what most people will think tomorrow, well, that’s partly up to us, isn’t it?

[1] Although they typically appreciate the debilitating consequences of unit banking, many U.S. economists and economic historians appear unaware of the crucial role that freedom of note issue played historically in facilitating branch banking. That banking systems involving relatively few restrictions on banks’ ability to issue banknotes, like those of Scotland before 1845 and Canada until 1935, also had extremely well-developed branch networks, was no coincidence.

[2] On the limited redemption of National banknotes and attempts to address it see Selgin and White, “Monetary Reform and the Redemption of National Bank Notes, 1863-1913.” Business History Review 68 (2) (Summer 1994).

[3] For a very good review of the features and performance of the Canadian system in its heyday, see R.M. Breckenridge, “The Canadian Banking System, 1817-1890,” Publications of the American Economic Association, v. X (1895), pp. 1-476. Not long ago, when I spoke favorably of Canada’s system at a gathering of economic historians, one asked afterwards, rather superciliously, whether I realized how large Canada’s economy had been back around 1913. Apparently my interrogator thought that Canada’s small size made its success irrelevant. I can’t see why. Nor, evidently, could the many persons who proposed and lobbied for various asset currency proposals over the course of over a decade or so.

[Cross-posted from]