The Fiscal Theory of the Price Level: True and False

There are two versions of the fiscal theory of the price level (FTPL); one true, the other false.  The true version holds that if the fiscal authority dominates the policy space, then fiscal deficits could be monetized by the central bank. This version is consistent with the quantity theory of money, because inflation is ultimately determined by excess growth in the money supply.  If money growth were constant, inflation could not occur—that is, there could not be a sustained rise in the average level of money prices. In this sense, Milton Friedman’s dictum that “inflation is always and everywhere a monetary phenomenon” cannot be refuted (Friedman 1970: 11).

The second version of the FTPL, the so-called strong version, holds that even if the money supply is held constant, inflation can occur if the fiscal authority is passive.[1] All that is needed is for the public to expect prices to rise. People will then spend their given money balances at a faster rate — increasing the velocity of money — and prices will rise until expectations change. If the fiscal authority is passive, velocity can explode, producing hyperinflation (see McCallum and Nelson 2005). This feature of the strong version is referred to as “speculative inflation” and is independent of monetary policy (Tutino and Zarazaga 2014: 3). The strong version also implies that fiscal action—not monetary reform—is the primary tool for ending a hyperinflation. This version of the FTPL is false: it ignores historical evidence that shows the determining factor in generating hyperinflations is explosive growth in the money supply (or the expectation that such growth will occur); and it fails to recognize that stabilization  results from credible monetary reform.[2]

Expectations about future inflation don’t appear like manna from heaven — businesses and households know that excess money growth causes inflation. They also know that large unfunded government liabilities and budget deficits risk having the central bank monetize debt. Although the strong version of the FTPL assumes away that possibility, history does not.

One notable example is the German hyperinflation of 1921–23, and the rapid stabilization that ended runaway inflation. We will see that it was monetary policy not fiscal policy that enabled the rapid rise in the price level and abruptly ended it.

In their essay, “Inflation Is Not Always and Everywhere a Monetary Phenomenon,” Antonella Tutino and Carlos E. J. M. Zarazaga (2014) use the strong version of the FTPL, as proposed by Christopher Sims (1994), to explain the German experience. They argue that fiscal policy can explain both the hyperinflation and the stabilization of the German currency: it was the passivity of the fiscal authority that ushered in the hyperinflation, while activist fiscal policy ended the inflation. Let us see why they are wrong and why Friedman’s dictum still holds.

The key point that Tutino and Zarazaga (hereafter, TZ) make is that “hyperinflation is fiscal in nature because it can only happen if the fiscal authority—the central government — remains on the sidelines” (p.3). When the government did intervene by introducing the rentenmark, a new currency backed by real estate, prices stabilized, according to TZ, because “the government’s ability to raise revenues from the real estate market … successfully broke the link between mutually reinforcing lower fiscal revenues — implying higher fiscal deficits — and rising price levels” (pp. 3–4).[3]

The problem with TZ’s argument is that the introduction of the rentenmark was not fiscal policy; it was monetary reform. Furthermore, the mortgage-backing of the rentenmark was not sufficient to change expectations of further inflation, but it did help the public accept the new currency. Expectations changed because the public knew there was a legal limit on the total value of rentenmarks that could be issued by the Rentenbank, which was under the jurisdiction of the Reichsbank (the central bank). The backing of the currency by real estate was not relevant for stabilizing prices. There was no official convertibility between the inflated paper marks and the rentenmark, and the latter was not legal tender. The rentenmark was a parallel currency, added to the circulation of existing paper marks.[4]

It is true that 500 rentenmarks could be converted into a bond with a nominal value of 500 gold marks, “which was guaranteed by a legal mortgage on German property and which yielded a rate of interest at 5 percent in gold (actually payable in paper at the exchange rate of the gold mark),” but as Bresciani-Turroni (p. 340) points out, “the stability of the value of the rentenmark could not be due to the possibility of converting the latter into mortgage securities.” The reason is simple:

The market value of the mortgage bonds was lower than the nominal value. The market rate of interest was then much higher than 5 percent… . Besides, the increase of the issues of rentenmarks would continually add to the Government’s burden on interest on mortgage bonds, for which the public would exchange increasing quantities of rentenmarks; and therefore, in a precarious financial position, the uncertainty of the Government being able to continue the payment of interest would increase [ibid.].

Bresciani-Turroni tells us that confidence in the rentenmark was buoyed by the fact that it was a new currency and the public “believed in the efficacy of the mortgage guarantee.” But that confidence “would have been quickly dissipated if the public had been led to expect that, despite the obligation imposed on the rentenbank by decree, the Government would exceed the pre-arranged limit to the issues” (p. 348). It was the limitation on the quantity of money — not the expected revenue from the mortgage securities—that was an important factor in stabilizing the value of money.[5]

It is also noteworthy that the public’s confidence in the rentenmark currency was reinforced by the “constant-value clause,” which obligated those who took out loans from the Rentenbank to repay their debts in the same quantity of gold marks as represented in the original loan. That clause was intended to prevent the speculation that occurred during the hyperinflation when businesses and others took out bank loans in nominal paper marks but repaid them using greatly depreciated marks, thus giving speculators a strong incentive to support runaway inflation (Bresciani-Turroni 1953: 353).

The rentenmark did not begin to circulate until November 16, 1923, and was added to the existing stock of paper marks, which were still the only legal tender. At the same time, the Reichsbank stopped monetizing government debt by ending the discounting of Treasury bills. Bresciani-Turroni (p. 337) calls that monetary reform “a fact of fundamental importance” — yet it too is ignored by TZ.

Even though newly created paper marks could not be used to finance government profligacy, the central bank continued to supply marks for commercial uses. Between November 16, 1923, and November 30, the amount of paper marks in circulation increased from 93 trillion to more than 400 trillion, and reached 1,211 trillion by July 31, 1924. Meanwhile, the quantity of rentenmarks went from 501 million on November 30, 1923, to 1,803 million on July 31, 1924. Consequently, “the stabilization of the German exchange was not obtained by means of contraction, or even by a stoppage of the expansion of the circulation of legal currency” (ibid.).

Most notably, and in contrast to the FTPL as stated by TZ, “The exchange was stabilized before there existed the conditions (above all the equilibrium of the Reich Budget) which alone could assure a lasting recovery of the monetary situation” (Bresciani-Turroni, p. 355).

Germany faced hyperinflation because the Weimar leaders chose to finance World War I reparations and other spending by money creation — rather than cut spending and increase taxes — and to reduce the real value of public debt. Once the monetary printing presses started rolling, it was hard to stop them. Moreover, the Reichsbank was under the influence of the real bills doctrine and met all demands for credit with newly minted paper marks, believing that inflation was unlikely if the bank only discounted short-term bills that reflected real output (Nurkse 1946: 16–17).[6] The problem is that bank credit is expressed in nominal terms. Thus, as prices rose because of rapid money growth, the demand for credit increased and businesses repaid debts in depreciated currency.

What Bresciani-Turroni teaches us is that models like the strong version of FTPL are not sufficient to inform us of the forces that underlie hyperinflation and stabilization. A close study of the policy actions taken in Weimar Germany shows that inflationary expectations are grounded in the credibility of central banks, as well as fiscal authorities. The competing theory that “explosive expectations” can generate runaway inflation without any change in the money supply cannot be supported by the experience of the German hyperinflation. Likewise, there is no evidence to support TZ’s claim that the hyperinflation was ended by fiscal action — that is, backing the rentenmark by real estate revenues. Rather, it was ended by fundamental monetary reform and a credible commitment to return to price stability as well as fiscal fortitude.

German monetary experts, writing in the Dawes Report, viewed the “liquid cover” for the rentenmark as “insufficient to guarantee a permanent [monetary] system.” They argued for the removal of the rentenmark and the introduction of a convertible currency. Their proposal was accepted with the passage of the monetary law of August 30, 1924, which made the reichsmark the new legal tender (Bresciani-Turroni, p. 338). When the monetary law became effective on October 11, 1924,  the Reichsbank abolished the constant-value clause, which was deemed unnecessary as the paper mark was now convertible into the reichsmark at an exchange rate of 1 reichsmark = 1 billion paper marks (1 billion = 1,000,0002), and the rentenmark was convertible into the new currency at a rate of 1 to 1. On June 5, 1925, the legal tender status of the old paper mark ended and it was taken out of circulation (Bresciani-Turroni 1953: 353–54).[7]

In conclusion, one must agree with Willem Buiter (2002: 459) when he says that the FTPL model, in its strong version, “is fatally flawed”—it “has feet of clay.” More telling, by arguing that “inflation is not always and everywhere a monetary phenomenon,” Tutino and Zarazaga undermine the responsibility of central banks to maintain the long-run value of fiat money, which increases the danger of inflation.

__________________________

References

[1] For a more formal discussion of the weak and strong versions of the FTPL, see Carlstrom and Fuerst (2000).

[2] By assuming that the money supply is constant, and there is no opportunity for excess money growth by the central bank, the strong version of the FTPL sets up a strawman. It by-passes the dynamic theory of money (also known as “the theory of monetary disequilibrium”), which holds that large increases in the quantity of money relative to the trend rate of real output depreciate the value of money and lead to a subsequent rise in the velocity of money, accentuating the rise in prices and further reducing the real money stock (see Warburton 1966: 4–5). This inflationary spiral will continue until the monetary authority changes expectations by adopting fundamental reform that ends excessive money creation.

[3] The rentenmark was introduced by the decree of October 15, 1923, and began circulating on November 15.

[4] See Bresciani-Turroni ([1931] 1953: 334–37).

[5] The government tried to circumvent the legal limit on the issuance of rentenmarks in December 1923, but “was confronted by a determined refusal by the management of the Rentenbank.” That episode “helped to strengthen confidence in the new money. The limitation of the quantity was then of primary and fundamental importance” ( Bresciani-Turroni, p. 348).

[6] Humphrey (1980) carefully lays out the major fallacies that misguided monetary policymakers, blindsiding them to the dangers of excess money growth, and points to the significance of monetary reform in quickly stabilizing the value of the German currency.

[7] The reichsmark had a fixed gold content but was not convertible into gold until April 1930, at the discretion of the Reichsbank (Bresciani-Turroni 1953: 354).

Bresciani-Turroni, C. ([1931] 1953) The Economics of Inflation: A Study of Currency Depreciation in Post-War Germany, 1914–1923.  Foreword by L. Robbins; translated by M. E. Savers.  London: George Allen and Unwin.

Buiter, W. H. (2002) “The Fiscal Theory of the Price Level: A Critique.”  The Economic Journal 112 (July): 459–80.

Carlstrom, C. T., and Fuerst, T. S. (2000) “The Fiscal Theory of the Price Level.” Federal Reserve Bank of Cleveland Economic Review 36 (1): 22–32.

Friedman, M. (1970) “The Counter-Revolution in Monetary Theory.”  IEA Occasional Paper No. 33. London: Institute of Economic Affairs.

Humphrey, T. M. (1980) “Eliminating Runaway Inflation: Lessons from the German Hyperinflation.” Federal Reserve Bank of Richmond Economic Review (July/August):  3–7.

McCallum, B. T., and Nelson, E. (2005) “Monetary and Fiscal Theories of the Price Level: The Irreconcilable Differences.” Oxford Review of Economic Policy 21 (4): 565–83.

Nurkse, R. (1946) The Course and Control of Inflation: A Review of Monetary Experience in Europe After World War I. Geneva: League of Nations. (Nurkse wrote Part 1 of this report.)

Sims, C. A. (1994) “A Simple Model for Study of the Determination of the Price Level and the Interaction of Monetary and Fiscal Policy.” Economic Theory 4 (3): 381–99.

Tutino, A., and Zarazaga, C. (2014) “Inflation Is Not Always and Everywhere a Monetary Phenomenon.” Federal Reserve Bank of Dallas Economic Letter 9 (6): 1–4.

Warburton, C. (1966) Depression, Inflation, and Monetary Policy: Selected Papers, 1945–1953.  Baltimore: The Johns Hopkins Press.

[Cross-posted from Alt-M.org]