In the March 2021 issue of Econ Journal Watch, Farley Grubb, a professor of economics at the University of Delaware and a research associate of the NBER, published a paper with the most extraordinary abstract. I quote it here in full:
“In the last issue of Econ Journal Watch, Ronald Michener (2020) published his seventh critical comment on my research. In my replies to his previous six comments, I demonstrate that Michener is misguided (see Grubb 2005; 2006a; b; 2018b; 2019b; 2020a). I will continue that demonstration here in my reply to his seventh comment. I will demonstrate that Michener does not understand basic microeconomic theory; that Michener does not understand rational expectations or how to make it operational; that Michener does not understand my model of monetary performance; that Michener does not understand how colonial New Jersey redeemed its paper money; and that Michener does not know how to evaluate quotation evidence. In summary, Michener thinks downward sloping marginal revenue curves are demand curves, thinks that rational expectations means you must be able to expect the unexpected, does not consider risk when evaluating an asset’s present value, erroneously thinks that the redemption of colonial paper money was on demand, and considers antecedent and subsequent context irrelevant for evaluating quotation evidence.
Uncommonly harsh words! Yet not unexpected. Ronald Michener, a professor of economics at the University of Virginia, and Farley Grubb spent almost 25 years arguing over the monetary system of the American colonies. It has been a vicious battle, but a farcical one. Do not think that there are no stakes, though! At the core of their arguments is whether the US Constitution itself was good for the citizens of America, and whether the bar on states emitting their own currency raised or lowered welfare.
Michener spoke first. In 1987, he published “Fixed Exchange Rates and the Quantity Theory in Colonial America”, which seeks to answer a mystery: why did only some of the American colonies get inflation? Many of the colonies printed great sums of paper money; only in New England was there serious inflation. Running simple regressions over the quantity of money printed and price changes, as West (1978) did, reveals no correlation for most quantities. The quantity of money was also subject to unexpected and highly unusual shocks — for example, in 1768, the property tax revenue for two years in east Jersey was stolen overnight. These bills were to be taken out of circulation; such an immense increase in money supply should surely lead to inflation. It did not! In another instance, after the death of the treasurer of Virginia, John Robinson, it was discovered that he had embezzled no less than half of all paper currency then in circulation. (He had given out notes that should have been destroyed as loans to friends). Where was the effect on prices?
Michener sets out to reconcile the quantity theory with the evidence. He argues that there were two kinds of money, paper and specie, available to colonists, and that the two were perfect substitutes. An increase in the number of paper bills leads to an equal and opposite outflow of specie through trade. Merchants in London would regard paper money as worth less, and demand specie in payment. As Benjamin Franklin wrote in 1764, “On the emission of the first paper money, a difference soon arose between that and silver; the latter … being fit for a remittance. This property having soon found its value by the merchants on one another for it, and a dollar thereby coming to be rated eight shillings in paper money of New York, and [seven shillings six pence] in paper of Pennsylvania, it has continued uniformly at those rates in both provinces now near forty years, without any variation upon new emissions.” Only when all of the specie is driven out will inflation result. The apparent failure of the quantity theory of money is simply a failure of data — we can measure the money in bills, but not in coins.
Farley Grubb’s 2001 paper, which kicks off the whole fracas, sets out to measure it. No one has a comprehensive record of specie money; but perhaps you could make one by counting the mentions of money in newspaper advertisements. One of the most common ads at the time were runaway slaves, convicts, and servant ads. These gave a reward, partly dictated by statute, and partly in an award above that from the owner. The ratio of paper money to specie offered should tell us the ratio of paper money to specie; and if you don’t believe that, then at the very least changes in the ratio indicates something.
What Grubb found is that a very high percentage of the money was offered was in Pennsylvania pounds — 80%, in fact. 13% was in silver, specifically the Spanish pieces of 8, and gold pistoles amounted to 3 to 5% of the circulating currency. Paper money from other colonies basically never circulated in other colonies, and the ratios were the same both for urban and rural residents. Residents would routinely offer paper money to anyone in their province, and specie to anyone without.
Grubb then extends this paper in a 2003 article published in the American Economic Review, where he argues that the institution of centrally issued money did not improve the status quo, and was in fact driven by personal avarice by well-connected bankers. The stability of colonial money was due to the deliberate actions of the colonial governments in promising to take money out of circulation through taxes. State money didn’t fail, it was banned.
Michener, with Robert E. Wright, thinks this is astoundingly wrong, and show why in a 2005 comment. Grubb commits an elementary error – he confuses units of exchange with units of account. When someone advertised that they would pay such and such in Pennsylvania pounds, they clearly did not mean that they would pay in that specific unit of exchange. Rather, they would pay an amount equal to that, in whatever currency they had on hand. Michener and Wright argue, quoting Charles M. Andrews, that the money “was not money at all, but only a method of reckoning values, a statement of the amount in shillings at which a Spanish dollar would be accepted in a given colony.” Exchange was done in a multitude of currencies, but they would all be converted into Pennsylvania pounds when being written down in the account books. Michener and Wright adduce considerable textual evidence that it was specie which circulated, particularly after 1781, when paper money almost entirely disappeared.
Farley Grubb wrote back quickly, in an article published in the next edition of the AER. The key argument is simply reiterating that while colonists did use units of account for noting transactions, for arms-length transactions colonists would specify the unit of exchange too. His key evidence is that people would often list out the coins that they would pay in. In addition, Grubb accuses Michener and Wright of selectivity in quoting evidence too, and adduces quotes supporting his viewpoint.
Michener and Wright continue the argument in Econjournalwatch, the AER being unwilling to stomach substantially similar arguments. While some advertisements would list the coins, 83.3% of ads in the sample leave the pounds and shillings unclarified. While it seems bizarre to us, the “pound” in different colonies were really exchange rates for Spanish dollars. In addition, any promise to pay in specie would not be legally enforceable. Applying the same techniques to other arms length transactions – servant contracts – produces estimates wildly at odds with Grubb’s runaway slave ads.
The strongest piece of evidence that Grubb is confused comes from the ads from 1720. There were 34 unique ads, 29 of which offered pounds and shillings. Yet, Pennsylvania did not issue bills of credit until 1723. What could they possibly have meant, other than a unit of account?
Grubb responded, focusing on the quality of the quotations, but also making strong arguments with regard to the assumption of fixed exchange rates, which are necessary for the quantity theory to work. For example, where Michener (1988) quoted Mazzei (an Italian contemporary) in saying “In 1773, all transactions were made almost entirely in specie”, Grubb tracks down the original source as saying “In 1773, the year disorders began, that is, ten years after the end of the previous war, all transactions were made almost entirely in specie, which, however, did not abound.” Grubb accuses Michener and Wright of serious unprofessionalism, and intimidating other writers. I quote, “The attack here by Michener and Wright (2006) is not an isolated event. Over the last 20 years the debate over colonial money has become strident, in no small part due to Michener and Wright. … In addition, other writers on colonial money have privately related stories to me of being accosted by Michener after they published a book or article touching on colonial money, e.g. Newell (1998).”
Michener and Wright write back in the next issue, in “Farley Grubb’s Noisy Evasions on Colonial Money: A Rejoinder”, and really hit the nail on the head. I quote the central paragraph.
Our central disagreement is very simple. We believe that when colonists promised “six shillings Pennsylvania money” they usually meant “I will give you bills of credit (not necessarily Pennsylvania’s), gold, silver, tobacco, hogs, credit in my account book, etc. to the value of six shillings.” Grubb believes that in certain transactions “six shillings Pennsylvania money” invariably meant “I will give you six shillings in Pennsylvania bills of credit.” Grubb’s supposition that pounds, shillings and pence in runaway ads must refer to Pennsylvania’s bills of credit, has, in our view, been thoroughly undermined. Grubb never explains why there were profuse references to Pennsylvania pounds and shillings in runaway ads before there were Pennsylvania bills of credit; his argument that ads promising specie were legally binding commitments to pay in specie is convincingly refuted by Pennsylvania’s legal tender laws; he never explains why his technique, when applied to the early Federal period, produces results that contradict his own work in Grubb (2003, 2006b). A host of other evidence could be presented—advertisements and contemporary documents that simply make no sense if pounds and shillings meant only paper money.
That was the end of the first battle. The second battle, clearly given life by personal animus, is so tedious that I will not recount it, being altogether bored of this topic.
So who was right, in the end? Before I give my answer, I would like to give some comments of an insulting and scurrilous nature. First, I would respect Michener considerably more if he did anything, anything at all, besides publishing criticisms of one author. Perusing his CV gives the impression of someone who received tenure, and stopped doing anything at all useful. Notably, despite being at the same university for almost 45 years, he was never promoted to Professor, a fact reflected in his pay, which is considerably lower than his colleagues. This was basically the only thing he published for the past 25 years. I cannot imagine being so unproductive, and frankly it makes me feel somewhat disgusted. Many of their comments had an ugly sheen of envy about them; take, for instance, Robert Wright’s blog post on the whole fracas from 2020, with its frequent complaints about how the system is skewed toward big shots with top name advisors from top name schools, and allegations that Grubb unfairly reviewed his book in the Journal of Economic History. I can certainly agree, however, with his claim that “I have neither the time nor the inclination to reprise the entire Michener-Grubb debate”. This has been deeply tiresome to recount.
Nevertheless, I think Michener’s basically right. Grubb’s evidence base is far too weak to make the claims he makes. That he has been defending it doggedly does not make it right. Neither does the thin evidence base for anything at all related to the colonial period necessitate thin methods. There are some things we are not made to know. Yes, colonists doubtless listed some transactions in the unit of exchange, not the unit of account. We have no idea what percentage of people did so, or whether that percentage
This isn’t to say that the paper on colonial money stock shouldn’t have been published to begin with, or that Grubb was wrong for trying. It is good to try! And to spend one’s entire career in a purely negative sense, doing nothing but attacking the attempts of one economist to tackle hard problems, is a waste of effort and talent. I came away from reading this fracas profoundly annoyed at everyone involved, and happy that they have put this behind them in retirement.
Interesting post. But I think your exasperation is a bit too hasty; this is what debates in monetary history are like. There are a lot of strong incentives for writers with certain political agendas to rush to draw conclusions from sloppy treatments of evidence, and they are indeed very cliquish and resistant to criticism.
As to the topic itself, the confusion seems to me to be generated by a lack of appreciation for the dynamics of a monetary system based on bills of exchange without par clearing, in which you can observe structural spreads between the value of the unit of account in different exchange places. the local moneys being quoted are the pound as it valued on the local exchange, which is necessarily different in different places because of the exchange hierarchy.