This blog post first appeared in the Economic Forces newsletter. Its author is a Senior Fellow at BPI and Chair of the Department of Economics at Ole Miss.
This semester, I am teaching a course about Bitcoin and monetary economics. To some, that seems odd. Many economists aren’t interested in Bitcoin. Others dislike it. Those who dislike it tend to think it is either “bad” or useless. To me, it seems obvious why I would teach such a course. From the day I first heard about it on an April 2011 episode of EconTalk, I have been fascinated by it. I have been following its development since that day. I have also written a lot about it.
Part of the motivation for teaching the class is just a response to demand. There is probably no other topic that brings as many random students to my office door than Bitcoin.
However, another motivation for teaching the class was partly to help me organize my thinking. When I talk to other economists, a lot of them will ask me why I’m interested in Bitcoin. For example, a stereotype of people interested in Bitcoin is that they are libertarian or even an anarcho-capitalist. I am neither of the two. In fact, I was somewhat libertarian back in 2011. I have become substantially less libertarian since then and yet my interest in Bitcoin hasn’t declined.
Another stereotype is that people only care about Bitcoin because they believe that the price is going to go up (with good reason, there are many on social media who refer to Bitcoin as an “NGU (number go up) technology”). This isn’t me either. I never talk about the price other than to make jokes and share memes on social media during periods of extreme volatility.
The typical answer that I give people is that I am interested in it because I am interested in monetary economics. Most economists find that answer unsatisfying, especially considering that many monetary economists don’t find it interesting.
I have also been thinking about this as it relates to Economic Forces. A question that I sometimes get from readers is why I write so much about monetary topics when the Substack is explicitly about price theory.
This has really made me think about my approach to monetary economics and what, if anything, is unique about that approach. Thus, I thought I would write a post explaining how I think about monetary economics as well as why I think that explains my interest in Bitcoin.
My approach to monetary economics is distinct, but I don’t think that it is unique. It cannot be unique since my own views on monetary economics have been shaped by a number of other economists, like Jurg Niehans, Earl Thompson, David Glasner, Scott Sumner, Larry White, and George Selgin. It might be useful to expand on why they have had this influence, what I’ve learned, and what makes this approach distinct.
Early on in graduate school, my dissertation advisor, Bob Rossana, told me that if I wanted to write a dissertation about money, I first needed to read Jurg Niehans’s The Theory of Money. Initially, I thought this was somewhat odd. In class, we had spent all of our time on contemporary models and at that time, this book was already 30 years old — and was rarely, if ever, cited by any of that literature. Furthermore, he offered no explanation for why I was to read this book. Instead, he simply made it a requirement if I wanted to write about money.
Dissertation advisors work in mysterious ways, and for that I am grateful (I miss you Bob!). Niehans’s book is excellent. Niehans was a price theorist. He just happened to be interested in money. The book is really just a guide to using price theory to understand money and other imperfectly substitutable assets. The book also possibly includes one of the best chapters in any book using price theory to understand commodity money.
This book not only had a profound effect on the way that I thought, but it also had a profound effect on the things that I read. At that point, I already had an affinity for UCLA economists like Alchian and Demsetz and Klein and Chicago economists like Friedman and Becker and Murphy. Thus, I was interested to find more economists who took a price theoretic approach to monetary economics. And that is how I found Larry White and George Selgin’s work on competition in the supply of currency (for example, see here and here, but there’s much more where that came from!).
What was interesting about Larry’s and George’s work is that it applied the same tools for thinking about firm behavior to banks and, in particular, competitive note issuance. In contrast, a lot of contemporary monetary economics (and macroeconomics, more generally) took central banks as given and offered somewhat historically dubious justifications for their existence. Thus, by articulating a theory of how private banks would behave under competitive note issuance and using historical evidence to test their theory, they opened up new questions to contemplate. One such question was why the state has a monopoly on money — a question I’ve written about here on multiple occasions (see here, here, and here).
Of course, any discussion of my thinking must also mention the work of Earl Thompson. Earl was a price theorist’s price theorist. He was willing to make unique arguments, but always backed by price theoretic reasoning. He made several important and interesting contributions to monetary economics. His paper, “The Theory of Money and Income Consistent with Orthodox Value Theory” articulates a theory of a competitive money supply that he subsequently applies to the gold standard era. His model is able to explain several implications of a competitive model and the gold standard, more generally. In doing so, he resolves some puzzles in monetary theory and presents a monetary theory of the Great Depression distinct from that articulated by people like Friedman and Schwartz. David Glasner expanded on some of this work in his book, Free Banking and Monetary Reform.
Scott Sumner’s views are probably best known from his blog. While Scott’s recent book, The Money Illusion, is often presented as a theory of the Great Recession, it is much more than that. The book clearly articulates a price theoretic view of the determination of the price level and highlights the type of empirical evidence that supports that view. Scott’s views are probably most similar to David’s and Earl’s work, but sufficiently unique that I think one could argue that it is Scott who has carried on the Chicago tradition in monetary economics more than any other monetary economist.
But another lesson from studying Larry and George and Earl and David relates more to the UCLA brand of price theory. They emphasize the role of institutions. These rules shape incentives and therefore economic outcomes. Failure to take account the role of institutions can result in very different predictions. However, their focus on institutions shouldn’t be mistaken for simply taking institutions as given and reasoning from there (although there is some of that and that is valuable). Consistent with the UCLA approach, there’s also analysis of how those institutions come to be.
Along these lines, Earl explained why states would suspend the gold standard during wars and return to the gold standard at the previous parity after the war when other economists concluded this was silly and costly. David and Larry and George developed theories of the state’s involvement in money.
All of this is not to say that there aren’t others working on these topics and contributing to our understanding of these topics. In fact, I can think of many of my contemporaries who are working on these things and from whom I have also learned. However, in many ways these were my primary influences in thinking through a wide variety of topics and issues and I think that they represent a sufficiently unique brand of monetary economics that emphasizes price theoretic foundations and the study of institutions (both as they are and why they exist). And that influence is why I have adopted the approach that I have.
So what does any of that have to do with Bitcoin?
The focus on institutions and on explaining the mechanics of different monetary regimes likely made me more willing to take Bitcoin seriously. I enjoyed learning how the gold standard worked even though I had no expectation that regime was coming back. So why wouldn’t I be interested in the development of a new type money or competition in an unexpected form?
This is also true of the importance of historical evidence emphasized by practitioners of this price theoretic approach. Focusing on lessons from monetary and financial history leave one less susceptible to the lazy historical narratives that are sometimes (often?) found in introductory textbooks.
Nonetheless, it is not just that. The price theoretic approach that these economists employ offer direct lessons for the study of Bitcoin.
For example, Earl Thompson’s theory emphasizes the role of the last period problem when it comes to the production of private money. A bank that produces its own bank notes faces a last period problem in the sense that if there is some final period in which the notes will be willingly held, then no one will want to hold them in the subsequent period. If they won’t be accepted in the subsequent period, then no one should be willing to accept them in the final period. Through backward induction, this means that no one should be willing to hold them now. Earl emphasized that one would need some sort of commitment mechanism. He proposed that the commitment mechanism would be a promise to buy these notes back in exchange for some fixed quantity of a commodity. This resolves the last period problem because people know they won’t be stuck holding a worthless piece of paper because they can redeem the paper for the commodity. This also resembles a gold standard under competitive note issuance.
Ben Klein developed a competing theory. In Klein’s theory, he argued that people would be willing to hold money without any future promise of redemption as long as that money was expected to retain its purchasing power. Since banks would issue their own brands of notes, they would have an incentive to maintain stable purchasing power to maintain their brand name. Banks would maximize the present discounted value of profits associated with the brand. Of course, there is one caveat. A bank that built up a large balance sheet and a respected brand name might be able to engineer a large one-time transfer to itself. The bank could do this by secretly printing a large quantity of bank notes. The owner(s) of the bank could then spend these notes on the basis of their strong brand name in the market and receive a variety of goods and services in return. Over time, the general public would realize what happened and the massive increase in the supply of notes would render them worthless. However, this would only happen after the owners enriched themselves at the expense of note holders.
So how do we know that we can trust the bank? Klein’s argument was that as long as the present discounted value of future profits from the brand exceeded the value of the large, one-time transfer, the bank would be willing to operate honestly. Although that economic argument is correct, there is no way for such a bank to commit to its future actions. As a result, it is unlikely that such a system could actually work in the real world.
When Bitcoin (the protocol) emerged on the scene, I think that a lot of people evaluated bitcoin (the asset) in terms of a Thompson-type argument. If we start in some last period in which bitcoin is no longer valued, then backward induction implies that it is worthless today. If people view it that way, then it will never get off the ground.
However, I interpreted Bitcoin as a test of Klein’s theory. This is a privately-supplied asset, but the protocol explicitly limits the supply of bitcoin to 21 million. That limit on the supply is the commitment mechanism that is lacking from the type of money issuers described by Klein. Furthermore, it is a credible commitment since it is written into the code. Not only that, changing the commitment is not incentive compatible for users. Since Bitcoin is decentralized, any change to the code must be adopted by the consensus of all involved. However, holders of bitcoin don’t have an incentive to adopt a change that increases the supply because that increase in supply would dilute the value of their own holdings. In other words, Bitcoin represents a possible solution to the commitment problem evident in Klein’s model. That is a remarkable innovation! Critics might see it as a minor innovation, but my view has always been that the market can determine what that innovation is worth.
This also helps to explain a difference I have with people who dislike Bitcoin’s supply schedule (including some of the people I listed as an inspiration above). The general consensus of the people I discussed above is that an ideal monetary system is one in which the supply of that money is entirely demand-determined. Such a system would prevent monetary disequilibrium and the (potentially) severe macroeconomic consequences thereof. With a modern central bank, this would mean adopting a nominal GDP target (see my forthcoming paper for more on this topic, here or here).
According to that view, bitcoin is not ideal as money. Nonetheless, Bitcoin cannot work in that way and still solve the Klein problem. It is not possible to have a commitment not to inflate the supply if the supply is allowed to fluctuate based on market forces. This is due to the inability to solve the oracle problem, or the inability of the network to access external data. Thus, one would not be able to incorporate market-based targets into the code and would thereby reintroduce an element of trust. The Klein problem returns!
Even a constant money supply rule for bitcoin, as suggested for the Federal Reserve by Milton Friedman, would suffer from this problem. By having a fixed supply, the Bitcoin protocol creates an anchoring that makes any change incompatible with the incentives of its users. Thus, while it might not be ideal, it is the only commitment mechanism that solves the Klein commitment problem.
In addition, fluctuations in supply could take place using bitcoin-backed assets rather than bitcoin itself. Just as bank notes were redeemable for gold and had an elastic supply, the same could be true with bitcoin in the place of gold.
In fact, the earliest thought that I had when I initially heard about the fixed supply is that it reminded me of George Selgin’s proposal for free banking in the modern world. A simplified version of his proposal was to fix the supply of the monetary base and allow banks to issue notes redeemable for base money. And I’m not the only one. One of the earliest adopters of Bitcoin, Hal Finney, replied to those worried about how Bitcoin would scale if it gained widespread adoption. Finney replied by explaining George’s work and how it would apply to Bitcoin.
Finally, Bitcoin raises issues that often don’t get much attention, but that seem important to those who use money. Physical cash provides some value to those who use it as a result of its privacy characteristics. When one person gives cash to another person in exchange for a good or service, that limits knowledge about that exchange to the parties involved. There is no third party with a record of that transaction.
As commerce increasingly takes place over the internet and involves the use of credit and debit cards, the amount of transaction information that third parties have is growing. To the extent to which people value privacy, that entails a cost. In addition, greater availability and access to this data creates potential costs associated with data leaks or even censorship.
Bitcoin itself grew out of a concern over privacy in a digital world by a disparate group known as Cypherpunks. They emphasized the difference between secrecy and privacy and worried about the degree of record-keeping that would be possible in a digital world. To increase privacy, they posited the need for an electronic version of cash. There were a number of precursors to Bitcoin that sought to solve that problem, but failed for various reasons. In many ways, the development of Bitcoin reflected the lessons learned in that process. It is also no coincidence that the title of the Bitcoin whitepaper is “Bitcoin: A Peer-to-Peer Electronic Cash System.”
There is much remaining to explore about privacy.
In short, I think that Bitcoin represents an important test of a variety of arguments known to monetary theorists and, in particular, to those who take a price theoretic approach. My interest is therefore in seeing how all of this plays out. We had a new type of money emerge out-of-sample. We now get to test these results using something that didn’t exist when the theory was written down. I find that fascinating. In fact, I cannot understand why others do not.
I’ve decided to teach this class for a number of reasons. First, there seems to be genuine interest and excitement about this topic. Why not teach students about things they are interested in?! Second, organizing this course has really helped me to organize my own thoughts and focus on the distinct lessons I want students to understand. And finally, teaching is learning. By teaching this topic, it will force me to think about things a little more carefully and deeply and hopefully pose new questions along the way. More opportunities to learn.
Those who think this topic to be unworthy of study are entitled to that opinion. Undeterred, I’m excited for the semester.
