GtDep/ItHadToHappen2 jw May 2009
THE GREAT DEFLATION OF 1929-33 (ALMOST) HAD TO HAPPEN1
John H. Wood
Wake Forest University
[T]he value of money … does not depend permanently on the quantity of it possessed by a given community, or on the rapidity of its circulation, or on the prevalence of exchanges, or on the use of barter or credit, or, in short, on any cause whatever, excepting the cost of its production.
Other causes may operate for a time, but their influence wears away as the existing stock of the previous metals within the country accommodates itself to the wants of the inhabitants. As long as precisely 17 grains of gold can be obtained by a day's labour, every thing else produced by equal labour will, in the absence of any natural or artificial monopoly, sell for 17 grains of gold; whether all the money of the country change hands every day, or once in four days, or once in four years; whether each individual consume principally what he has himself produced, or supply all his wants by exchange; whether such exchanges are effected by barter or credit, or by the actual intervention of money; whether there be 1,700,000 or 170,000 grains in the country.
Nassau Senior, "On the quantity and value of money."
Figure 1 shows price movements during British and American gold suspensions and resumptions arising from the Napoleonic Wars (War of 1812 in the U.S.), the American Civil War, and World War I. (The lower dotted gold production lines are discussed below.)2 The American experience of 1913-33 differed from the others in that, because of large inflows of gold from abroad, suspension was not required. Nevertheless the gold standard required the same ultimate price-level adjustment as in the other cases. The similarity of the price movements in all five cases was not accidental. Given the unchanged official prices of gold and the relative costs of gold production, they had to agree. Monetary and fiscal policies and other events following the wars only affected the timing of the price-level restorations. It is worth emphasizing at the outset that the British and other resumptions of pre-war parities with the dollar in the mid-1920s were insufficient because the United States had also had substantial inflation.
There are many other explanations of the world-wide Great Deflation/Depression of 1929-33, perhaps beginning with the Austrian view that the excesses of the 1920s credit boom and consequent overinvestment had to be liquidated (Hayek 1935, pp.161-61; Robbins 1934, pp.35-54). For several years the dominant (Keynesian) explanation was a collapse in demand (Gordon 1952, p.388; Temin 1976, p.178). Schumpeter (1939, pp.161-74) pointed to the coincidence of long and short cyclical downturns. Monetary explanations have dominated since Friedman and Schwartz (1963, pp.299-419) emphasized the collapse of money after a routine downturn had been triggered by tight Federal Reserve policy to stem the stock-market boom and prevent gold losses following tight French monetary policy (Hamilton 1987). The renewed interest in monetary factors was expanded to focus on the gold standard as the chief culprit, domestically and internationally. Commitment to fixed rates of gold convertibility of their currencies inhibited expansionary responses by the monetary authorities and transmitted deflation across countries, exacerbated by their competition for limited gold reserves (Temin 1989, pp.1-40; Eichengreen 1992, pp.12-21, 222-86). This is characterized as “the breakdown of the gold standard,” and was probably the main reason for its abandonment in the 1930s.
These explanations begin with the late 1920s, although a large literature finds the origins of the depression in the disruptions of the Great War, whose legacy was “new political borders drawn apparently without economic rationale; substantial overcapacity in some sectors (such as agriculture and heavy industry) and undercapacity in others …; and reparations claims and international war debts that generated fiscal burdens and fiscal uncertainty” (Bernanke and James 1991). The consequent international claims and conflicts would have been difficult even for the more-smoothly functioning pre-1914 gold standard, but that had also changed. It has been argued that the prewar gold standard was hegemonic, with Great Britain the “international conductor” (Keynes 1930b ii, p.306; Kindleberger 1973, ch.14). In the 1920s, however, “the relative decline of Britain, the inexperience and insularity of the new political hegemon (the United States), and ineffective cooperation among central banks left no one able to take responsibility for the system as a whole” (Bernanke and James 1991). So on top of the fundamental economic disruptions and political instability (Svennilson 1954, p.321), to which Hansen (1938) added secular stagnation, was the mismanagement of the gold standard, particularly the competition for gold reserves discussed below.
However, even those who find the sources of the Great Depression in the political and structural fragility of the system left by the Great War believe that its effective cause lay in events beginning in the late 1920s. Tight monetary and fiscal policies “were due to the adherence of policymakers to the ideology of the gold standard” (Temin 1989, p.7). Gold flows to France and the United States “provided the contractionary impulse that set the stage for the 1929 downturn … because of the foreign reaction it provoked through its interaction with existing imbalances in the pattern of international settlements and with the gold standard constraints” (Eichengreen 1992, pp.12-13).
The present paper seeks to reinforce the view that the gold standard, or rather its mismanagement, was the sufficient cause of the Great Deflation of 1929-33. I say Great Deflation because the analysis is limited to price levels. Real effects may be surmised but are not treated here.3 Nor do I try to explain the deflation’s timing. I accept that it was made inevitable by the return to prewar gold parities but argue that the mismanagement of the gold standard lay not in the failure of central banks to prop up domestic money stocks and price levels after the resumptions of the mid-1920s, but in the initial overvaluations of the currencies involved.
Other explanations of the occurrence of the Great Deflation, although not necessarily to its timing or real effects, are irrelevant. Given the decision to return to (or in the case of the United States, to maintain) the gold standard at the pre-war parity, the Great Deflation was inevitable, whatever the propensities to spend, social and political disruptions, stock prices, reparations, distributions of gold and foreign exchange, locations and relative importance of financial centers, or central bank policies. The return of price levels to their 1914 values (adjusted for changes in parity) was dictated by the gold standard, according to which, as stated by Senior (1829) in the opening quotation, the relative price of gold and other goods (the price level) equals the ratio of their marginal costs of production. A necessary assumption is that the relative cost of gold production did not change significantly between 1913 and 1933, which will be shown is supported by the data.
The remainder of the paper is organized as follows. Section 1 lays out the theory of the price level under the gold standard, Section 2 describes the postwar monetary policies that sought to escape the implications of the gold standard by stabilizing prices at their end-of-war levels, and the inevitable deflation is described in Section 3. Section 4 concludes.
1. The Workings of the Gold Standard
I use a variant of Barro’s (1979) model in which, in the steady state, following Senior (1829) and Mill (1848, pp.501-503), the equilibrium general price index of commodities (P) relative to the fixed price of gold (Pg) equals their relative marginal costs of production -- relation (4) in Table 1.
The supply and demand for money determine the price ratio at any point in time, as indicated by (1)-(3). Shifts in the supply and demand for money (H and K) alter P (and therefore P/Pg) temporarily, but set in motion forces that restore the equality of relative prices and costs. For example, if banks reduce reserve ratios and/or the public holds less currency relative to bank deposits, so that H and P rise, gold production falls and its nonmonetary use rises, both depressing the monetary gold stock, Gm. The final position includes a return to the original P and HGm. This case is developed for the 1920s in Section 3.
Table 1. The Gold Standard
The supply and demand for money are
(1) and (2) Md = K(R)Py,
where M is the medium of exchange (in dollars, note and deposit claims on gold plus gold coin), Gm is the monetary gold stock (in ounces), Pg is the arbitrarily fixed dollar price at which the monetary authority stands ready to buy and sell an ounce of gold, H is the money multiplier, i.e., the ratio of money to the monetary base, P is the general price level of commodities, y is output, and K is a (negative) function of the rate of interest. H is inversely related to the demand for gold (see Appendix).
Equating (1) and (2) gives
(3) , which holds at all times.
The real (commodity) cost of producing gold is c(g) (c′ , c″ > 0). Profit-maximizing behavior by gold producers gives
(4) c′(g) = Pg/P, which implies the supply function for new gold:
(5) , gs′ < 0.
The flow demand for non-monetary (jewelry and industrial) gold is
(6) , f1 > 0, f2 < 0, so that the change in the monetary gold stock is
(7) , where < 0.
Substituting a specific form of (7) into (3) for Pg = 1 and fixed H and K gives
(8) where a, b, and P* govern the rate of change of monetary gold.
A steady state exists if Pt = Pt-1 = P* and a = , i.e., if the rates of change of Gm and output are constant and the same.
2. Monetary policies in the 1920s
Cooperation and reform. In August 1918, the Cunliffe Committee on Currency and Foreign Exchanges enunciated the initial goal of postwar British monetary policy:4
In our opinion it is imperative that after the war the conditions necessary to the maintenance of an effective gold standard should be restored without delay. Unless the machinery which long experience has shown to be the only effective remedy for an adverse balance of trade and an undue growth of credit is once more brought into play, there will be grave danger of a progressive credit expansion which will result in a foreign drain of gold menacing the convertibility of our note issue and so jeopardizing the international trade position of the country.
This meant the “cessation of Government borrowing as soon as possible after the war,” return to an effective Bank of England discount rate (we would now say an independent central bank), and gold convertibility of the currency, at least in foreign transactions because gold coinage was discontinued. It was assumed that “the aim of policy must be to restore the pre-war Gold Standard in essentials,” including the reestablishment of the prewar $/£ exchange rate of $4.86 (Moggridge 1969, pp.12, 14). These assumptions were shared by several other industrial nations, particularly Sweden, Denmark, Netherlands, Norway, Switzerland, Canada, Australia, New Zealand, and South Africa, who resumed prewar pars with the dollar between 1922 and 1926, and maintained them at least to September 1931 (Federal Reserve Board, 1943, pp.662-81; League of Nations 1920; 1946, pp.92-93).
They understood that the currency inflations of 1914-20 made resumption difficult, especially since gold production had fallen continuously since 1915. Conferences at Brussels in 1920 and Genoa in 1922 looked for solutions, but got no further than imprecise promises of cooperation to economize on and share the gold stock (Eichengreen 1992, p.153). Proposals for international supervisory and lending bodies similar to those established in 1944 at Bretton Woods were rejected. The obvious choice – between devaluation, enormous deflation, or abandonment of the gold standard – was suppressed. The 28 percent British deflation relative to the United States implied by the 3.81 exchange rate at the end of 1920, although recognized as a difficult task in itself, greatly underestimated the two-thirds deflation (Table 2) necessary to a return to the prewar gold parity.
In fact the classical gold standard was abandoned in everything but the use of gold in international transactions. The new gold standard was envisioned as a managed currency with the objective of price stability regardless of the cost of gold production, which was hardly mentioned. If the objective is to stabilize the currency “in terms of commodities [it] is natural to ask,” Keynes (1923, p.139) wrote, “why it is necessary to drag in gold at all.”
Significant cooperation took place in the 1920s, such as in the stabilization of the German mark in 1924 and the resumption and support of the British pound in 1925 and after (Clarke 1967, pp.45-107), but it was limited. The first duties of central banks are to the economic well-being of their own countries (Clarke 1967, pp.40-41). “The rules of the gold standard game” (Keynes 1925) that were supposed to have made the prewar system work was that central banks did not resist, and even reinforced, gold flows. If a country with stable prices lost gold through to a balance-of-payments deficit to country undergoing deflation, neither central bank offset the gold flow so that prices were brought into equality, lower in the former and higher in the latter, as must be the case since both were tied to gold. Behaving otherwise would lead cause maldistributions of reserves and price distortions. However, countries did not follow the rules in the 1920s (Nurkse 1944, pp.66-69), but offset gold flows in the interests of domestic stability; just as they had, in fact, before 1914 (Bloomfield 1959). The survival of the gold standard in the earlier case can be explained as follows. A country may preserve financial stability by moderating reserve gains and losses without endangering the short-term international value of its currency if traders trust its commitment to that value in the long term. The mispricing of gold interfered with that credibility in the 1920s.
Accounting for gold production. A critic of the postwar system, Gustav Cassel (1920), stressed the importance of the endogenous quantity of gold.
I wish … to call attention [to a] difficulty which seems to have escaped the attention … even of those who have become aware of the importance of the problem of stabilising the value of gold. This difficulty arises in connection with the production of the metal.
If we have a stabilised monetary demand for gold, we must, of course, have an annual production of gold corresponding to the general rate of progress of the world, and, in addition, sufficient to cover the yearly waste of gold. This normal annual demand for gold amounted, during the period 1850-1910, on an average to about 3 per cent of the total accumulated stock of gold …. Of this sum, 0.2 per cent covered the loss of gold and 2.8 per cent was added to the world’s stock.
Since 1914, however, “the rise of prices of commodities in terms of gold … has hampered … production and brought it down considerably.” Furthermore, “it is only natural that the demand for gold as a material for articles of luxury should have increased substantially. [T]he danger of a quite insufficient supply of gold is much more imminent than seems to be generally recognized. [A]ssuming the production of gold to remain about constant, we have to face a growing scarcity of gold and a continued depression of prices.”
The implications that Cassel drew from this situation was that countries should not go back to prewar prices, or if the objective is price stability, to the prewar system at all. A much talked of advantage of the prewar system was its “high degree of stability,” which “we should now endeavour to restore” (1922, p.254). Adopting mispriced currencies, fighting over a shortage of gold reserves, and entering into a long period of deflation are not ways to do it.
Cassel’s analysis and recommendations were ignored by policymakers and rejected by most economists.5 Palyi (1972, p.88) accused Cassel of twisting Ricardo’s Purchasing Power Parity idea into an “infallible guide to policy.” What was needed to make the system work was rather “a better technique” for the use of existing reserves, which are neither “too large or too small,” except when prevented from going where they were needed (Mlynarski 1929, p.31). “The gold scarcity has never been a decisive factor in the postwar era. Until 1929 it was more than offset by the rise in efficiency,” that is, by more economical reserves (Neisser 1941). These and other writers (Hardy 1936, p.18; Phinney 1933) thought Cassel’s data irrelevant, except possibly in relation to “an archaic monetary system” (Mlynarski 1929, p.28). R.G. Hawtrey preferred devaluation but thought “continuous co-operation among central banks” as envisioned in the Genoa Resolutions might work (1922).
This abstraction from the realities of the gold standard is bewildering. The dollar was already inflated, and unless the real resource cost of gold had fallen, would itself have to return to its prewar level. The economization of gold – that is, the continued increase in the already greatly expanded quantity of credit on the basis of a declining monetary gold stock relative to income – could not go on indefinitely. Eichengreen (1992, p.155) noted that “curiously little mention of gold was made” at the Brussels and Genoa conferences, but even he blamed the Great Deflation on the failure of cooperation regarding the distribution of reserves.
An explanation of their attitude might be found in the failure of officials and economists to comprehend the development of gold mining during the quarter century before 1914. From an accidental and irregular phenomenon, even through the California discovery of 1849, gold mining had become an industry dominated by large firms who explored and developed reserves in the manner typical of other systematic profit-seekers (Rockoff 1984). Senior’s theory had become fact.