Thread by @GeorgeSelgin

My Gorton and Zhang critique series continues with a look at their discussion of antebellum U.S. currency (pp. 23ff.) Although their discussion of that history is better than many other recent ones (see alt-m.org/2021/07/06/the…), it is still misleading.

Although GZ recognize that most bank failures back then had nothing to do with "wildcat" banking, they still exaggerate the extent to which antebellum currency was (literally) "subpar."

Just as CZ "cherry pick" by focusing on US currency experience, ignoring what happened elsewhere, they also cherry pick from the diverse US record. I've already noted how they ignore the Suffolk System's successful achievement of a uniform New England currency--a "good" cherry.

In this thread I'll mention some examples of their "bad" cherry picking and some other disturbing inaccuracies.

On p. 23, citing a 1995 article by Ken Ng, they say that as a result of "free banking laws, "starting in 1837...anyone could open a bank." In fact Ng concludes that "the theory that free banking led to lower barriers to entry must be rejected." jstor.org/stable/2121621…

Not good!

On. p. 25, GZ claim that private banknotes only circulated because "the alternative was a bewildering array of different coins from around the world." That's one way to account for the popularity of nonpar private money, which is otherwise hard to reconcile with the GZ analysis.

The trouble is that, while foreign coin made up a large share of total US coins in the early 19th c., by the 1840s, the situation was changing rapidly, with most foreign coins being reminted into US coins. By 1853, David Martin reports (jstor.org/stable/2119352…),...

"The U.S. had at last provided a sufficient national gold and silver specie currency." It follows that, after 1853 at least, nonpar (and non-NQA) private banknotes somehow managed to circulate despite the availability of perfectly good NQA U.S. government money.

That coexistence seems especially paradoxical in view of GZ's "example" (p.. 25) of a private note issued by a Tennessee bank that "might circulate at a 20 percent discount in Philadelphia." But look at the table on p.. 26, and you'll see what this means.

Tt means that _during a major panic_, when many banks failed or were in danger of failing, like that of 1857, their notes might be so heavily discounted. Usually, for most banks, that wasn't the case. It wasn't even the case for many during panics.

Normally, banknote discounts weren't so high. In fact they were only high enough to reflect the costs of getting notes back to their sources for redemption. Hence the positive relation between quoted discounts and an issuers' distance from market from which the quote is taken.

As more railroads and telegraph lines crossed the country, these discounts declined. That's another fact GZ ignore--though G certainly knows it! As I've noted in many places, by late 1863 discounts mostly ran between 0 and 2%!

That's not so bad, really. And those discounts persisted not because different banks held different assets some of which were opaque, but owing to unit banking and a lack clearinghouses, New York's (est. 1853) having been the first to be established after the Suffolk System.

The only reasonable conclusion any monetary historian who considers _all_ the evidence of past private currency (banknote) systems can draw is that, far from tending to be, let alone always being, nonpar money, competitively-supplied banknotes _tend_ to circulate at par.

Nonpar banknotes aren't normal. They're pathological results of legal restrictions, underdeveloped note-exchange and clearing infrastructure, and actual or probable bank insolvencies. Otherwise simple redemption-based arbitrage MUST suffice to keep banknotes at par.

So what does this tell us about stablecoins? It tells us that they are _not_ like banknotes in crucial respects. One obvious difference is the lack of unlimited and low-cost redeemability on demand into USD. This is abundantly clear in the case of Tether. news.ycombinator.com/item?id=256843…

Instead of misleading comparisons that stress stablecoins' resemblance to _bad_ banknotes, we could use some comparisons of stablecoins with _good_ banknotes, stressing the differences between them. That would be the right way to use history to inform stablecoin regulation.

@threadreaderapp, kindly unroll.