Introduction and scope of the debate
My recent exchange with my old friend Guido Hülsmann has now reached its conclusion. I am grateful for his invitation to offer final remarks and respond to his objections. As always, his comments are friendly and rhetorically elegant. Yet I had hoped for an engagement with the core arguments of my previous rejoinder — a hope that was only partially met throughout the debate.
Prof. Hülsmann presents the discussion as a counterfactual exercise concerning price inflation in Argentina in December 2023:
“What would have happened if the Milei government had closed down Argentina’s central bank right after taking power in December 2023?”
The Role of Credit-Money expectations in Argentina
However, the debate did not begin with a counterfactual. It began with my application of Mises’s concept of credit money to the Argentine case. I argued that the peso retained a credit-money character rooted in its convertibility experience (1991–2002),1 and in Milei’s original dollarization proposal, which envisaged using central bank reserves to redeem pesos into dollars. If the central bank had been closed immediately upon Milei’s inauguration, this convertibility expectation would have vanished. The peso would have ceased to function as a credit-money and become a pure fiat money. Ceteris paribus, this would imply a decline in the demand for pesos.
Under Argentine conditions in late 2023, such a decline would have been deeply destabilizing. The peso was already subject to accelerating monetary distrust, dollar preference by Argentinians, collapsing demand, and a monthly price inflation rate approaching 25 percent. The surging dollar exchange rate further accelerated the flight from the peso and the loss of confidence. A further drop in money demand—triggered by the closure of the central bank—would not have halted the loss of confidence. It would have intensified it, risking a complete collapse of the peso’s purchasing power: hyperinflation, or demonetization if one prefers that term.
Professor Hülsmann acknowledges part of this analysis:
“Ceteris paribus, as Bagus explains, and we agree with him, this would have reduced the demand for pesos.”
He then adds:
“But how important is this factor? As Philipp Bagus presents it, there is only a short step between this decline in demand for money and hyperinflation. But hyperinflation does not arise from a loss of confidence alone, but from a loss of confidence combined with an enormous increase in the money stock.”
The core theoretical issue: Money demand and purchasing power
For me, this is where the true disagreement begins. I believe the claim is incorrect. The issue at stake is a fundamental theoretical matter:
Is a complete destruction of a money’s purchasing power possible through a collapse in money demand alone—or must it be accompanied by a large increase in the money stock?
The praxeological answer is clear: a fall in demand has the same effect as an increase in supply on the prices of non-money goods.2 If the demand for fiat money evaporates, its purchasing power approaches zero regardless of what happens to the quantity of money. This is why Rothbard, Salerno, and Hazlitt — as noted earlier in this exchange — emphasize that a sudden loss of confidence, independent of the money supply, can lead to the abandonment of a good as a medium of exchange.
The process of demonetization and the primacy of ideas
When a commodity becomes monetized, it acquires an additional demand beyond its direct uses as a consumer good or a producer good. When a good is demonetized because people cease using it as a medium of exchange and turn to another good they consider superior, its demand falls back to the demand for it as a consumer good or a producer good. In the case of fiat money the demand falls back to virtually zero, which means that prices in terms of this fiat money approach infinity. The physical supply of the fiat money or the expectation of the evolution of the physical supply becomes irrelevant at this point. Fiat money depends critically on trust. Once this trust erodes, its monetary role can diminish very rapidly.
Economics is about ideas that actors have about the world. It is not about physical things. Ludwig von Mises repeatedly has pointed out the importance of ideas.3 The power of government, Mises points out in Omnipotent Government (2010, p. 7), does not rest on material conditions such as arms, but ultimately rests on ideas:
“A lasting order cannot be established by bayonets. A minority cannot rule if it is not supported by the consent of those ruled; the rebellion of the oppressed will overthrow it sooner or later, even if it were to succeed for some time.”
Also, the purchasing power of money, and in effect its status as money, rests ultimately on ideas. Once actors lose confidence in it and no longer use it as a medium of exchange, the material question of the physical supply ceases to matter.
Argentina’s extreme monetary fragility
In his concluding remarks, Prof. Hülsmann concedes that such a scenario is ‘not logically impossible,’ which suggests that we may share more common ground on the theoretical mechanism than initially apparent. Yet instead of addressing the theoretical structure, he draws attention to the empirical case of Argentina in December 2023.
Here, we simply attribute different weight to institutional expectations under extreme uncertainty. My argument is context specific: in Argentina’s circumstances—deep monetary distrust, dollarization preferences, the credit-money character of the peso, excessive public debt, a precarious banking system, a long history of defaults, and a vocal opposition prepared to overturn the government while claiming to be the only force capable of stabilizing the country—the abolition of the central bank would have plausibly triggered a collapse in money demand.
When money demand dominates the quantity of money
Prof. Hülsmann calls my scenario “out of proportion,” but he does not identify which causal step is disproportionate:
- That abolishing the central bank alters institutional expectations,
- That expectations influence money demand,
- That money demand affects purchasing power
- That this effect can dominate the influence of a shrinking money stock.
My argument proceeds through four explicit causal steps.
If Prof. Hülsmann disagrees, it would be highly valuable—both for readers and for the development of the argument—to know at which precise step our analyses diverge. Describing my scenario as ‘romantic’ is picturesque, but it would be helpful to know which specific step in the praxeological chain of reasoning he finds implausible, and for what reason.
Possibly, Prof. Hülsmann considers the last step implausible, as he asserts that “the demand for money is itself largely dependent on the perceived evolution of the money stock.” I respectfully disagree. Money demand depends on the perceived usefulness of holding money as a medium of exchange and as a hedge against uncertainty—that is, on the quality of money. Expected changes in the money stock can influence this perception, but they are not its foundation. People demand money because of what money is and what it does for them.
Expectations about the money stock may modify this demand, but they do not create it; they are secondary rather than primary, and merely one factor among many that shape the demand for money. Since expectations about the money stock do not generate the demand for a generally accepted medium of exchange, it follows that the evolution of the money stock does not, by itself, sustain its value. In this respect, Prof. Hülsmann’s emphasis on the stock of money appears to rely on a somewhat romanticized view of its importance. While the expected evolution of the money stock is only one variable in the determination of money demand, institutional changes—such as a revolution, a lost war, or the transition from a credit money to a pure fiat money in a distrustful monetary environment—can shift the valuation of money far more abruptly.4
Demonetization and Hoppe´s view
It is noteworthy that Hans-Hermann Hoppe independently arrived at a conclusion quite close to mine, describing the peso as likely to ‘disappear.’
This suggests that the mechanism I outline is not idiosyncratic, but grounded in widely shared Austrian reasoning. More specifically, Prof. Hoppe stated in the context of a discussion on the closure of the Argentinian central bank and the collapse of the banking system that: “the peso would disappear, that´s good…then people would use dollars, instead of pesos. Pesos die out and they use dollars.” (PFS, Min. 36)
Hoppe clearly agrees that the peso would lose purchasing power rapidly and disappear as money. He does not employ the term “hyperinflation” and probably would prefer the term “demonetization.” However, the logical mechanism is identical. Prof. Hülsmann, by contrast, described Milei as a “very bad economist” for predicting such an outcome. This also raises the question of how consistently such judgments should be applied within Austrian analysis.
The Role of Central Bank Assets
There is a certain tension in Prof. Hülsmann’s argument regarding the importance of central bank assets for the quality of money, and therefore for the demand for money.5 He cites with approval Kristoffer Hansen’s argument that the elimination of the Argentine central bank would not have caused a hyperinflation. Hansen explicitly claims that the assets of the central bank are irrelevant for the demand for a currency, stating:
“The value of the peso is, as we have established, independent of the assets of the Argentinian central bank…”
However, central bank assets do matter for the quality of a currency—especially in situations of monetary stress, such as when confidence in the currency is falling, when a flight from it is underway, when the exchange rate is collapsing, or in cases of monetary reform such as dollarization or the liquidation of a central bank.
Professor Hülsmann himself writes:
“Existing peso notes would either have remained in circulation or been liquidated in exchange for assets held by the central bank. Existing cash accounts held at the central bank would either have been exchanged for peso notes, or liquidated in exchange for assets held by the central bank.”
This passage implies that the assets of the central bank are important, because the higher the quality of these assets, the better the position of peso holders in such a scenario.6 This stands in clear contrast to Hansen’s claim that the assets are irrelevant. It is difficult to see how both assertions could be maintained simultaneously.
Moreover, Prof. Hülsmann indicates that the peso (that is, the base money stock) is “a liability of the central bank”. While peso notes and central-bank cash accounts are not legally redeemable debt instruments today, there was such an obligation during the convertibility period (1991–2002), and there may be one again in the future. Thus, the assets are particularly relevant—especially in periods of transition toward monetary reform.
Must a libertarian politician close the central bank immediately?
Professor Hülsmann takes issue with the fact that Milei did not close “Argentina’s central bank, as promised, right after taking power in December 2023.”
He acknowledges that there were reasons for this decision, such as the need to prevent “a terrible economic crisis.”
Yet he does not consider this a valid reason, stating:
“Considerations of this sort are the hallmark of the middle-of-the-road social democrat or Christian-democrat “conservative” leader, the proverbial party politician, honourable member of the political caste.”
In response, three clarifications are in order.
First, based on personal conversations, it is true that Milei believed that closing the central bank on his first day in office would have precipitated a hyperinflation and major economic crisis. It would have forced him out of office within days.
Second, although Milei campaigned on closing the central bank, he never committed to a specific date, as Prof. Hülsmann suggests. His plan always envisaged a phased process, as we discussed in our previous contribution.
Third, Prof. Hülsmann’s formulation appears to imply that any politician who does not immediately abolish the central bank—or even the state itself—and instead opts for a phased reduction to avoid social turmoil and the return of socialists to power must be regarded as a Christian or social democrat. I do not believe this inference is warranted.
First, this interpretation seems at odds with Prof. Hülsmann´s own earlier statement that “the main task of any reformer” is to “take temporary measures to help the most vulnerable during the transition phase.” Would such considerations not also, by his logic, be “the hallmark of the middle-of-the-road social democrat”?
Second, and as stated elsewhere, I beg to differ. One can be a libertarian politician as long as one pushes in the right direction. According to Jesús Huerta de Soto, the libertarian politician must follow a dual strategy:
(1) he must rigorously study libertarian principles and disseminate them to the public without compromise; and
(2) he must develop transition plans toward the ideal that do not violate those principles.
Short-term compromises may be acceptable only if they move society in a more libertarian direction. In no case may a set of measures move away from it.
There is, in my view, an essential distinction between a libertarian politician implementing principled, prudentially sequenced reform and a social democrat with no interest in reducing the scope of the state at all.
Concluding remarks
In sum, I am grateful for this exchange. In my view, it has highlighted the importance of the demand for money. The purchasing power of fiat money can disappear even without an increase in its supply — the central praxeological point at issue, and one whose importance the present exchange has helped to highlight. The debate has underscored the decisive role of institutional expectations in shaping money demand under extreme uncertainty, and the possibility that a collapse in demand can dominate the influence of the money stock.
In addition, important questions have been raised concerning central bank assets, remunerated central bank liabilities, the loss of credit-money characteristics, fiscal demand for money, the inadequacy of the term ‘price level,’ the nature of libertarian political strategy, and the respective roles of a priori theory and contingent historical circumstances, among other important issues.
One remaining disagreement concerns the weight we assign to institutional expectations in shaping money demand under extreme uncertainty and in the dynamics of hyperinflation. In this respect, our divergence lies in the empirical importance each of us attributes to institutional expectations under conditions of acute monetary fragility. More fundamental, however, is our deeper disagreement over the logic of money demand and, consequently, over the process of demonetization.
I hope this exchange contributes to clarifying these important issues within Austrian monetary theory. I remain grateful for the opportunity to discuss them in depth.
- 1
The credit-money character is rooted not only in the convertibility experience, but also in the fact that throughout most of the period from 2002 to 2023 Argentina operated under a fixed exchange rate system which—given high monetary and price inflation—was accompanied and reinforced by the CEPO (capital controls). I owe this point to Bernardo Ferrero and would like to thank him for his helpful comments and suggestions throughout the exchange.
- 2
The same logic applies to any good: a rise in supply and a fall in demand exert equivalent effects on its price. In the case of money, the two processes are not fully symmetrical. For example, they differ in their redistributive effects and in their impact on social real cash balances. A fall in money demand (at a constant money stock) reduces aggregate real cash balances, whereas an increase in the money supply (at a constant money demand) leaves them unchanged, ceteris paribus.
- 3
See, for instance, Human Action, (1998), Chapter IX, “The Role of Ideas.” In Liberalism. In the Classical Tradition, (2002), p. 51, Mises states: “The ultimate outcome of the struggle, however, will not be decided by arms, but by ideas. It is ideas that group men into fighting factions, that press the weapons into their hands, and that determine against whom and for whom the weapons shall be used. It is they alone, and not arms, that, in the last analysis, turn the scales.”
- 4
On abrupt changes in the quality of money, see Bagus (2009), “The Quality of Money,” The Quarterly Journal of Austrian Economics 12, no. 4: 22–45. In Bagus (2015), “The Quality of Monetary Regimes,” in The Next Generation of Austrian Economics: Essays in Honor of Joseph Salerno, ed. Per Bylund and Per Howden, 19–34 (Auburn, Ala.: Ludwig von Mises Institute), I analyze the effects of changes in the monetary regime on the quality of money.
- 5
On the importance of central bank assets for the quality of money, see Bagus and Howden. (2016). “Central Bank Balance Sheet Analysis.” Betriebswirtschaftliche Forschung und Praxis 68(2), pp. 109-25.
- 6
Indeed, most of the central bank’s assets consisted of government bonds. In the event of a government default, the banking system would have been left with worthless claims, triggering financial collapse and a sharp depreciation of the peso, thereby further eroding confidence in the currency. It certainly would have made a difference for the peso if, instead of Argentine government bonds, the central bank’s assets had consisted of gold (or especially dollars).
