Archive for the 'Larry White' Category

Hayek’s Rapid Rise to Stardom

For a month or so, I have been working on a paper about Hayek’s early pro-deflationary policy recommendations which seem to be at odds with his own idea of neutral money which he articulated in a way that implied or at least suggested that the ideal monetary policy would aim to keep nominal spending or nominal income constant. In the Great Depression, prices and real output were both falling, so that nominal spending and income were also falling at a rate equal to the rate of decline in real output plus the rate of decline in the price level. So in a depression, the monetary policy implied by Hayek’s neutral money criterion would have been to print money like crazy to generate enough inflation to keep nominal spending and nominal income constant. But Hayek denounced any monetary policy that aimed to raise prices during the depression, arguing that such a policy would treat the disease of depression with the drug that had caused the disease in the first place. Decades later, Hayek acknowledged his mistake and made clear that he favored a policy that would prevent the flow of nominal spending from ever shrinking. In this post, I am excerpting the introductory section of the current draft of my paper.

Few economists, if any, ever experienced as rapid a rise to stardom as F. A. Hayek did upon arriving in London in January 1931, at the invitation of Lionel Robbins, to deliver a series of four lectures on the theory of industrial fluctuations. The Great Depression having started about 15 months earlier, British economists were desperately seeking new insights into the unfolding and deteriorating economic catastrophe. The subject on which Hayek was to expound was of more than academic interest; it was of the most urgent economic, political and social, import.

Only 31 years old, Hayek, director of the Austrian Institute of Business Cycle Research headed by his mentor Ludwig von Mises, had never held an academic position. Upon completing his doctorate at the University of Vienna, writing his doctoral thesis under Friedrich von Wieser, one of the eminent figures of the Austrian School of Economics, Hayek, through financial assistance secured by Mises, spent over a year in the United States doing research on business cycles, and meeting such leading American experts on business cycles as W. C. Mitchell. While in the US, Hayek also exhaustively studied the English-language  literature on the monetary history of the eighteenth and nineteenth centuries and the, mostly British, monetary doctrines of that era.

Even without an academic position, Hayek’s productivity upon returning to Vienna was impressive. Aside from writing a monthly digest of statistical reports, financial news, and analysis of business conditions for the Institute, Hayek published several important theoretical papers, gaining a reputation as a young economist of considerable promise. Moreover, Hayek’s immersion in the English monetary literature and his sojourn in the United States gave him an excellent command of English, so that when Robbins, newly installed as head of the economics department at LSE, and having fallen under the influence of the Austrian school of economics, was seeking to replace Edwin Cannan, who before his retirement had been the leading monetary economist at LSE, Robbins thought of Hayek as a candidate for Cannan’s position.

Hoping that Hayek’s performance would be sufficiently impressive to justify the offer of a position at LSE, Robbins undoubtedly made clear to Hayek that if his lectures were well received, his chances of receiving an offer to replace Cannan were quite good. A secure academic position for a young economist, even one as talented as Hayek, was then hard to come by in Austria or Germany. Realizing how much depended on the impression he would make, Hayek, despite having undertaken to write a textbook on monetary theory for which he had already written several chapters, dropped everything else to compose the four lectures that he would present at LSE.

When he arrived in England in January 1931, Hayek actually went first to Cambridge to give a lecture, a condensed version of the four LSE lectures. Hayek was not feeling well when he came to Cambridge to face an unsympathetic, if not hostile, audience, and the lecture was not a success. However, either despite, or because of, his inauspicious debut at Cambridge, Hayek’s performance at LSE turned out to be an immediate sensation. In his History of Economic Analysis, Joseph Schumpeter, who, although an Austrian with a background in economics similar to Hayek’s, was neither a personal friend nor an ideological ally of Hayek’s, wrote that Hayek’s theory

on being presented to the Anglo-American community of economists, met with a sweeping success that has never been equaled by any strictly theoretical book that failed to make amends for its rigors by including plans and policy recommendations or to make contact in other ways with its readers loves or hates. A strong critical reaction followed that, at first, but served to underline the success, and then the profession turned away to other leaders and interests.

The four lectures provided a masterful survey of business-cycle theory and the role of monetary analysis in business-cycle theory, including a lucid summary of the Austrian capital-theoretic approach to business-cycle theory and of the equilibrium price relationships that are conducive to economic stability, an explanation of how those equilibrium price relationships are disturbed by monetary disturbances giving rise to cyclical effects, and some comments on the appropriate policies for avoiding or minimizing such disturbances. The goal of monetary policy should be to set the money interest rate equal to the hypothetical equilibrium interest rate determined by strictly real factors. The only policy implication that Hayek could extract from this rarified analysis was that monetary policy should aim not to stabilize the price level as recommended by such distinguished monetary theorists as Alfred Marshall and Knut Wicksell, but to stabilize total spending or total money income.

This objective would be achieved, Hayek argued, only if injections of new money preserved the equilibrium relationship between savings and investment, investments being financed entirely by voluntary savings, not by money newly created for that purpose. Insofar as new investment projects were financed by newly created money, the additional expenditure thereby financed would entail a deviation from the real equilibrium that would obtain in a hypothetical barter economy or in an economy in which money had no distortionary effect. That  interest rate was called by Hayek, following Wicksell, the natural (or equilibrium) rate of interest.

But according to Hayek, Wicksell failed to see that, in a progressive economy with real investment financed by voluntary saving, the increasing output of goods and services over time implies generally falling prices as the increasing productivity of factors of production progressively reduces costs of production. A stable price level would require ongoing increases in the quantity of money to, the new money being used to finance additional investment over and above voluntary saving, thereby causing the economy to deviate from its equilibrium time path by inducing investment that would not otherwise have been undertaken.

As Paul Zimmerman and I have pointed out in our paper on Hayek’s response to Piero Sraffa’s devastating, but flawed, review of Prices and Production (the published version of Hayek’s LSE lectures) Hayek’s argument that only an economy in which no money is created to finance investment is consistent with the real equilibrium of a pure barter economy depends on the assumption that money is non-interest-bearing and that the rate of inflation is not correctly foreseen. If money bears competitive interest and inflation is correctly foreseen, the economy can attain its real equilibrium regardless of the rate of inflation – provided, at least, that the rate of deflation is not greater than the real rate of interest. Inasmuch as the real equilibrium is defined by a system of n-1 relative prices per time period which can be multiplied by any scalar representing the expected price level or expected rate of inflation between time periods.

So Hayek’s assumption that the real equilibrium requires a rate of deflation equal to the rate of increase in factor productivity is an arbitrary and unfounded assumption reflecting his failure to see that the real equilibrium of the economy is independent of the price levels in different time periods and rates of inflation between time periods, when prices levels and rates of inflation are correctly anticipated. If inflation is correctly foreseen, nominal wages will rise commensurately with inflation and real wages with productivity increases, so that the increase in nominal money supplied by banks will not induce or finance investment beyond voluntary savings. Hayek’s argument was based on a failure to work through the full implications of his equilibrium method. As Hayek would later come to recognize, disequilibrium is the result not of money creation by banks but of mistaken expectations about the future.

Thus, Hayek’s argument mistakenly identified monetary expansion of any sort that moderated or reversed what Hayek considered the natural tendency of prices to fall in a progressively expanding economy, as the disturbing and distorting impulse responsible for business-cycle fluctuations. Although he did not offer a detailed account of the origins of the Great Depression, Hayek’s diagnosis of the causes of the Great Depression, made explicit in various other writings, was clear: monetary expansion by the Federal Reserve during the 1920s — especially in 1927 — to keep the US price level from falling and to moderate deflationary pressure on Britain (sterling having been overvalued at the prewar dollar-sterling parity when Britain restored gold convertibility in March 1925) distorted relative prices and the capital structure. When distortions eventually become unsustainable, unprofitable investment projects would be liquidated, supposedly freeing those resources to be re-employed in more productive activities. Why the Depression continued to deepen rather than recover more than a year after the downturn had started, was another question.

Despite warning of the dangers of a policy of price-level stabilization, Hayek was reluctant to advance an alternative policy goal or criterion beyond the general maxim that policy should avoid any disturbing or distorting effect — in particular monetary expansion — on the economic system. But Hayek was incapable of, or unwilling to, translate this abstract precept into a definite policy norm.

The simplest implementation of Hayek’s objective would be to hold the quantity of money constant. But that policy, as Hayek acknowledged, was beset with both practical and conceptual difficulties. Under a gold standard, which Hayek, at least in the early 1930s, still favored, the relevant area within which to keep the quantity of money constant would be the entire world (or, more precisely, the set of countries linked to the gold standard). But national differences between the currencies on the gold standard would make it virtually impossible to coordinate those national currencies to keep some aggregate measure of the quantity of money convertible into gold constant. And Hayek also recognized that fluctuations in the demand to hold money (the reciprocal of the velocity of circulation) produce monetary disturbances analogous to variations in the quantity of money, so that the relevant policy objective was to hold the quantity of money constant, but to change the quantity of money proportionately (inversely) with the demand to hold money (the velocity of circulation).

Hayek therefore suggested that the appropriate criterion for the neutrality of money might be to hold total spending (or alternatively total factor income) constant. With constant total spending, neither an increase nor a decrease in the amount of money the public desired to hold would lead to disequilibrium. This was a compelling argument for constant total spending as the goal of policy, but Hayek was unwilling to adopt it as a practical guide for monetary policy.

In the final paragraph of his final LSE lecture, Hayek made his most explicit, though still equivocal, policy recommendation:

[T]he only practical maxim for monetary policy to be derived from our considerations is probably . . . that the simple fact of an increase of production and trade forms no justification for an expansion of credit, and that—save in an acute crisis—bankers need not be afraid to harm production by overcaution. . . . It is probably an illusion to suppose that we shall ever be able entirely to eliminate industrial fluctuations by means of monetary policy. The most we may hope for is that the growing information of the public may make it easier for central banks both to follow a cautious policy during the upward swing of the cycle, and so to mitigate the following depression, and to resist the well-meaning but dangerous proposals to fight depression by “a little inflation “.

Thus, Hayek concluded his series of lectures by implicitly rejecting his own idea of neutral money as a policy criterion, warning instead against the “well-meaning but dangerous proposals to fight depression by ‘a little inflation.’” The only sensible interpretation of Hayek’s counsel of “resistance” is an icy expression of indifference to falling nominal spending in a deep depression.

Larry White has defended Hayek against the charge that his policy advice in the depression was liquidationist, encouraging policy makers to take a “hands-off” approach to the unfolding economic catastrophe. In making this argument, White relies on Hayek’s neutral-money concept as well as Hayek’s disavowals decades later of his early pro-deflation policy advice. However, White omitted any mention of Hayek’s explicit rejection of neutral money as a policy norm at the conclusion of his LSE lectures. White also disputes that Hayek was a liquidationist, arguing that Hayek supported liquidation not for its own sake but only as a means to reallocate resources from lower- to higher-valued uses. Although that is certainly true, White does not establish that any of the other liquidationists he mentions favored liquidation as an end and not, like Hayek, as a means.

Hayek’s policy stance in the early 1930s was characterized by David Laidler as a skepticism bordering on nihilism in opposing any monetary- or fiscal-policy responses to mitigate the suffering of the general public caused by the Depression. White’s efforts at rehabilitation notwithstanding, Laidler’s characterization seems to be on the mark. The perplexing and disturbing question raised by Hayek’s policy stance in the early 1930s is why, given the availability of his neutral-money criterion as a justification for favoring at least a mildly inflationary (or reflationary) policy to promote economic recovery from the Depression, did Hayek remain, during the 1930s at any rate, implacably opposed to expansionary monetary policies? Hayek’s later disavowals of his early position actually provide some insight into his reasoning in the early 1930s, but to understand the reasons for his advocacy of a policy inconsistent with his own theoretical understanding of the situation for which he was offering policy advice, it is necessary to understand the intellectual and doctrinal background that set the boundaries on what kinds of policies Hayek was prepared to entertain. The source of that intellectual and doctrinal background was David Hume and the intermediary through which it was transmitted was none other than Hayek’s mentor Ludwig von Mises.

Advertisements

Hayek, Deflation and Nihilism

In the discussion about my paper on Hayek and intertemporal equilibrium at the HES meeting last month, Harald Hagemann suggested looking at Hansjorg Klausinger’s introductions to the two recently published volumes of Hayek’s Collected Works containing his writings (mostly from the 1920s and 1930s) about business-cycle theory in which he explores how Hayek’s attitude toward equilibrium analysis changed over time. But what I found most interesting in Klausinger’s introduction was his account of Hayek’s tolerant, if not supportive, attitude toward deflation — even toward what Hayek and other Austrians at the time referred to as “secondary deflation.” Some Austrians, notably Gottfried Haberler and Wilhelm Roepke, favored activist “reflationary” policies to counteract, and even reverse, secondary deflation. What did Hayek mean by secondary deflation? Here is how Klausinger (“Introduction” in Collected Works of F. A. Hayek: Business Cycles, Part II, pp. 5-6) explains the difference between primary and secondary deflation:

[A]ccording to Hayek’s theory the crisis is caused by a maladjustment in the structure of production typically initiated by a credit boom, such that the period of production (representing the capitalistic structure of production) is lengthened beyond what can be sustained by the rate of voluntary savings. The necessary reallocation of resources and its consequences give rise to crisis and depression. Thus, the “primary” cause of the crisis is a kind of “capital scarcity” while the depression represents an adjustment process by which the capital structure is adapted.

The Hayekian crisis or upper-turning point of the cycle occurs when banks are no longer willing or able to supply the funds investors need to finance their projects, causing business failures and layoffs of workers. The turning point is associated with distress sales of assets and goods, initiating a deflationary spiral. The collapse of asset prices and sell-off of inventories is the primary deflation, but at some point, the contraction may begin to feed on itself, and the contraction takes on a different character. That is the secondary deflation phase. But it is difficult to identify a specific temporal or analytic criterion by which to distinguish the primary from the secondary deflation.

Roepke and Haberler used the distinction – often referring to “depression”” and “deflation” interchangeably – to denote two phases of the cycle. The primary depression is characterized by the reactions to the disproportionalities of the boom, and accordingly an important cleansing function is ascribed to it; thus it is necessary to allow the primary depression to run its course. In contrast, the secondary depression refers to a self-feeding, cumulative process, not causally connected with the disproportionality that the primary depression is designed to correct. Thus the existence of the secondary depression opens up the possibility of a phase of depression dysfunctional to the economic system, where an expansionist policy might be called for. (Id. p. 6)

Despite conceding that there is a meaningful distinction between a primary and secondary deflation that might justify monetary expansion to counteract the latter, Hayek consistently opposed monetary expansion during the 1930s. The puzzle of Hayek’s opposition to monetary expansion, even at the bottom of the Great Depression, is compounded if we consider his idea of neutral money as a criterion for a monetary policy with no distorting effect on the price system. That idea can be understood in terms of the simple MV=PQ equation. Hayek argued that the proper criterion for neutral money was neither, as some had suggested, a constant quantity of money (M), nor, as others had suggested, a constant price level (P), but constant total spending (MV). But for MV to be constant, M must increase or decrease just enough to offset any change in V, where V represents the percentage of income held by the public in the form of money. Thus, if MV is constant, the quantity of money is increasing or decreasing by just as much as the amount of money the public wants to hold is increasing or decreasing.

The neutral-money criterion led Hayek to denounce the US Federal Reserve for a policy that kept the average level of prices essentially stable from 1922 to 1929, arguing that rapid economic growth should have been accompanied by falling not stable prices, in line with his neutral money criterion. The monetary expansion necessary to keep prices stable, had in Hayek’s view, led to a distortion of relative prices, causing an overextension of the capital structure of production, which was the ultimate cause of the 1929 downturn that triggered the Great Depression. But once the downturn started to accelerate, causing aggregate spending to decline by 50% between 1929 and 1933, Hayek, totally disregarding his own neutral-money criterion, uttered not a single word in protest of a monetary policy that was in flagrant violation of his own neutral money criterion. On the contrary, Hayek wrote an impassioned defense of the insane gold accumulation policy of the Bank of France, which along with the US Federal Reserve was chiefly responsible for the decline in aggregate spending.

In an excellent paper, Larry White has recently discussed Hayek’s pro-deflationary stance in the 1930s, absolving Hayek from responsibility for the policy errors of the 1930s on the grounds that the Federal Reserve Board and the Hoover Administration had been influenced not by Hayek, but by a different strand of pro-deflationary thinking, while pointing out that Hayek’s own theory of monetary policy, had he followed it consistently, would have led him to support monetary expansion during the 1930s to prevent any decline in aggregate spending. White may be correct in saying that policy makers paid little if any attention to Hayek’s pro-deflation policy advice. But Hayek’s policy advice was what it was: relentlessly pro-deflation.

Why did Hayek offer policy advice so blatantly contradicted by his own neutral-money criterion? White suggests that the reason was that Hayek viewed deflation as potentially beneficial if it would break the rigidities obstructing adjustments in relative prices. It was the lack of relative-price adjustments that, in Hayek’s view, caused the depression. Here is how Hayek (“The Present State and Immediate Prospects of the Study of Industrial Fluctuations” in Collected Works of F. A. Hayek: Business Cycles, Part II, pp. 171-79) put it:

The analysis of the crisis shows that, once an excessive increase of the capital structure has proved insupportable and has led to a crisis, profitability of production can be restored only by considerable changes in relative prices, reductions of certain stocks, and transfers of the means of production to other uses. In connection with these changes, liquidations of firms in a purely financial sense of the word may be inevitable, and their postponement may possibly delay the process of liquidation in the first, more general sense; but this is a separate and special phenomenon which in recent discussions has been stressed rather excessively at the expense of the more fundamental changes in prices, stocks, etc. (Id. pp. 175-76)

Hayek thus draws a distinction between two possible interpretations of liquidation, noting that widespread financial bankruptcy is not necessary for liquidation in the economic sense, an important distinction. Continuing with the following argument about rigidities, Hayek writes:

A theoretical problem of great importance which needs to be elucidated in this connection is the significance, for this process of liquidation, of the rigidity of prices and wages, which since the great war has undoubtedly become very considerable. There can be little question that these rigidities tend to delay the process of adaptation and that this will cause a “secondary” deflation which at first will intensify the depression but ultimately will help to overcome those rigidities. (Id. p. 176)

It is worth noting that Hayek’s assertion that the intensification of the depression would help to overcome the rigidities is an unfounded and unsupported supposition. Moreover, the notion that increased price flexibility in a depression would actually promote recovery has a flimsy theoretical basis, because, even if an equilibrium does exist in an economy dislocated by severe maladjustments — the premise of Austrian cycle theory — the notion that price adjustments are all that’s required for recovery can’t be proven even under the assumption of Walrasian tatonnement, much less under the assumption of incomplete markets with trading at non-equilibrium prices. The intuitively appealing notion that markets self-adjust is an extrapolation from Marshallian partial-equilibrium analysis in which the disequilibrium of a single market is analyzed under the assumption that all other markets remain in equilibrium. The assumption of approximate macroeconomic equilibrium is a necessary precondition for the partial-equilibrium analysis to show that a single (relatively small) market reverts to equilibrium after a disturbance. In the general case in which multiple markets are simultaneously disturbed from an initial equilibrium, it can’t be shown that price adjustments based on excess demands in individual markets lead to the restoration of equilibrium.

The main problem in this connection, on which opinions are still diametrically opposed, are, firstly, whether this process of deflation is merely an evil which has to be combated, or whether it does not serve a necessary function in breaking these rigidities, and, secondly, whether the persistence of these deflationary tendencies proves that the fundamental maladjustment of prices still exists, or whether, once that process of deflation has gathered momentum, it may not continue long after it has served its initial function. (Id.)

Unable to demonstrate that deflation was not exacerbating economic conditions, Hayek justified tolerating further deflation, as White acknowledged, with the hope that it would break the “rigidities” preventing the relative-price adjustments that he felt were necessary for recovery. Lacking a solid basis in economic theory, Hayek’s support for deflation to break rigidities in relative-price adjustment invites evaluation in ideological terms. Conceding that monetary expansion might increase employment, Hayek may have been disturbed by the prospect that an expansionary monetary policy would be credited for having led to a positive outcome, thereby increasing the chances that inflationary policies would be adopted under less extreme conditions. Hayek therefore appears to have supported deflation as a means to accomplish a political objective – breaking politically imposed and supported rigidities in prices – he did not believe could otherwise be accomplished.

Such a rationale, I am sorry to say, reminds me of Lenin’s famous saying that you can’t make an omelet without breaking eggs. Which is to say, that in order to achieve a desired political outcome, Hayek was prepared to support policies that he had good reason to believe would increase the misery and suffering of a great many people. I don’t accuse Hayek of malevolence, but I do question the judgment that led him to such a conclusion. In Fabricating the Keynesian Revolution, David Laidler described Hayek’s policy stance in the 1930s as extreme pessimism verging on nihilism. But in supporting deflation as a means to accomplish a political end, Hayek clearly seems to have crossed over the line separating pessimism from nihilism.

In fairness to Hayek, it should be noted that he eventually acknowledged and explicitly disavowed his early pro-deflation stance.

I am the last to deny – or rather, I am today the last to deny – that, in these circumstances, monetary counteractions, deliberate attempts to maintain the money stream, are appropriate.

I probably ought to add a word of explanation: I have to admit that I took a different attitude forty years ago, at the beginning of the Great Depression. At that time I believed that a process of deflation of some short duration might break the rigidity of wages which I thought was incompatible with a functioning economy. Perhaps I should have even then understood that this possibility no longer existed. . . . I would no longer maintain, as I did in the early ‘30s, that for this reason, and for this reason only, a short period of deflation might be desirable. Today I believe that deflation has no recognizable function whatever, and that there is no justification for supporting or permitting a process of deflation. (A Discussion with Friedrich A. Von Hayek: Held at the American Enterprise Institute on April 9, 1975, p. 5)

Responding to a question about “secondary deflation” from his old colleague and friend, Gottfried Haberler, Hayek went on to elaborate:

The moment there is any sign that the total income stream may actually shrink, I should certainly not only try everything in my power to prevent it from dwindling, but I should announce beforehand that I would do so in the event the problem arose. . .

You ask whether I have changed my opinion about combating secondary deflation. I do not have to change my theoretical views. As I explained before, I have always thought that deflation had no economic function; but I did once believe, and no longer do, that it was desirable because it could break the growing rigidity of wage rates. Even at that time I regarded this view as a political consideration; I did not think that deflation improved the adjustment mechanism of the market. (Id. pp. 12-13)

I am not sure that Hayek’s characterization of his early views is totally accurate. Although he may indeed have believed that a short period of deflation would be enough to break the rigidities that he found so troublesome, he never spoke out against deflation, even as late as 1932 more than two years the start of deflation at the end of 1929. But on the key point Hayek was perfectly candid: “I regarded this view as a political consideration.”

This harrowing episode seems worth recalling now, as the U.S. Senate is about to make decisions about the future of the highly imperfect American health care system, and many are explicitly advocating taking steps calculated to make the system (or substantial parts of it) implode or enter a “death spiral” for the express purpose of achieving a political/ideological objective. Policy-making and nihilism are a toxic mix, as we learned in the 1930s with such catastrophic results. Do we really need to be taught that lesson again?

Imagining the Gold Standard

The Marginal Revolution University has posted a nice little 10-minute video conversation between Scott Sumner and Larry about the gold standard and fiat money, Scott speaking up for fiat money and Larry weighing in on the side of the gold standard. I thought that both Scott and Larry acquitted themselves admirably, but several of the arguments made by Larry seemed to me to require either correction or elaboration. The necessary corrections or elaborations do not strengthen the defense of the gold standard that Larry presents so capably.

Larry begins with a defense of the gold standard against the charge that it caused the Great Depression. As I recently argued in my discussion of a post on the gold standard by Cecchetti and Schoenholtz, it is a bit of an overreach to argue that the Great Depression was the necessary consequence of trying to restore the international gold standard in the 1920s after its collapse at the start of World War I. Had the leading central banks at the time, the Federal Reserve, the Bank of England, and especially the Bank of France, behaved more intelligently, the catastrophe could have been averted, allowing the economic expansion of the 1920s to continue for many more years, thereby averting subsequent catastrophes that resulted from the Great Depression. But the perverse actions taken by those banks in 1928 and 1929 had catastrophic consequences, because of the essential properties of the gold-standard system. The gold standard was the mechanism that transformed stupidity into catastrophe. Not every monetary system would have been capable of accomplishing that hideous transformation.

So while it is altogether fitting and proper to remind everyone that the mistakes that led to catastrophe were the result of choices made by policy makers — choices not required by any binding rules of central-bank conduct imposed by the gold standard — the deflation caused by the gold accumulation of the Bank of France and the Federal Reserve occurred only because the gold standard makes deflation inevitable if there is a sufficiently large increase in the demand for gold. While Larry is correct that the gold standard per se did not require the Bank of France to embark on its insane policy of gold accumulation, it should at least give one pause that the most fervent defenders of that insane policy were people like Ludwig von Mises, F.A. Hayek, Lionel Robbins, and Charles Rist, who were also the most diehard proponents of maintaining the gold standard after the Great Depression started, even holding up the Bank of France as a role model for other central banks to emulate. (To be fair, I should acknowledge that Hayek and Robbins, to Mises’s consternation, later admitted their youthful errors.)

Of course, Larry would say that under the free-banking system that he favors, there would be no possibility that a central bank like the Bank of France could engage in the sort of ruinous policy that triggered the Great Depression. Larry may well be right, but there is also a non-trivial chance that he’s not. I prefer not to take a non-trivial chance of catastrophe.

Larry, I think, makes at least two other serious misjudgments. First, he argues that the instability of the interwar gold standard can be explained away as the result of central-bank errors – errors, don’t forget, that were endorsed by the most stalwart advocates of the gold standard at the time – and that the relative stability of the pre-World War I gold standard was the result of the absence of the central banks in the US and Canada and some other countries while the central banks in Britain, France and Germany were dutifully following the rules of the game.

As a factual matter, the so-called rules of the game, as I have observed elsewhere (also here), were largely imaginary, and certainly never explicitly agreed upon or considered binding by any monetary authority that ever existed. Moreover, the rules of the game were based on an incorrect theory of the gold standard reflecting the now discredited price-specie-flow mechanism, whereby differences in national price levels under the gold standard triggered gold movements that would be deflationary in countries losing gold and inflationary in countries gaining gold. That is a flatly incorrect understanding of how the international adjustment mechanism worked under the gold standard, because price-level differences large enough to trigger compensatory gold flows are inconsistent with arbitrage opportunities tending to equalize the prices of all tradable goods. And finally, as McCloskey and Zecher demonstrated 40 years ago, the empirical evidence clearly refutes the proposition that gold flows under the gold standard were in any way correlated with national price level differences. (See also this post.) So it is something of a stretch for Larry to attribute the stability of the world economy between 1880 and 1914 either to the absence of central banks in some countries or to the central banks that were then in existence having followed the rules of the game in contrast to the central banks of the interwar period that supposedly flouted those rules.

Focusing on the difference between the supposedly rule-based behavior of central banks under the classical gold standard and the discretionary behavior of central banks in the interwar period, Larry misses the really critical difference between the two periods. The second half of the nineteenth century was a period of peace and stability after the end of the Civil War in America and the short, and one-sided, Franco-Prussian War of 1870. The rapid expansion of the domain of the gold standard between 1870 and 1880 was accomplished relatively easily, but not without significant deflationary pressures that lasted for almost two decades. A gold standard had been operating in Britain and those parts of the world under British control for half a century, and gold had long been, along with silver, one of the two main international monies and had maintained a roughly stable value for at least half a century. Once started, the shift from silver to gold caused a rapid depreciation of silver relative to gold, which itself led the powerful creditor classes in countries still on the silver standard to pressure their governments to shift to gold.

After three and a half decades of stability, the gold standard collapsed almost as soon as World War I started. A non-belligerent for three years, the US alone remained on the gold standard until it prohibited the export of gold upon entry into the war in 1917. But, having amassed an enormous gold hoard during World War I, the US was able to restore convertibility easily after the end of the War. However, gold could not be freely traded even after the war. Restrictions on the ownership and exchange of gold were not eliminated until the early 1920s, so the gold standard did not really function in the US until a free market for gold was restored. But prices had doubled between the start of the war and 1920, while 40% of the world’s gold reserves were held by the US. So it was not the value of gold that determined the value of the U.S. dollar; it was the value of the U.S. dollar — determined by the policy of the Federal Reserve — that determined the value of gold. The kind of system that was operating under the classical gold standard, when gold had a clear known value that had been roughly maintained for half a century or more, did not exist in the 1920s when the world was recreating, essentially from scratch, a new gold standard.

Recreating a gold standard after the enormous shock of World War I was not like flicking a switch. No one knew what the value of gold was or would be, because the value of gold itself depended on a whole range of policy choices that inevitably had to be made by governments and central banks. That was just the nature of the world that existed in the 1920s. You can’t just assume that historical reality away.

Larry would like to think and would like the rest of us to think that it would be easy to recreate a gold standard today. But it would be just as hard to recreate a gold standard today as it was in the 1920s — and just as perilous. As Thomas Aubrey pointed out in a comment on my recent post on the gold standard, Russia and China between them hold about 25% of the world’s gold reserves. Some people complain loudly about Chinese currency manipulation now. How would you like to empower the Chinese and the Russians to manipulate the value of gold under a gold standard?

The problem of recreating a gold standard was beautifully described in 1922 by Dennis Robertson in his short classic Money. I have previously posted this passage, but as Herbert Spencer is supposed to have said, “it is only by repeated and varied iteration that alien conceptions can be forced upon reluctant minds.” So, I will once again let Dennis Robertson have the final word on the gold standard.

We can now resume the main thread of our argument. In a gold standard country, whatever the exact device in force for facilitating the maintenance of the standard, the quantity of money is such that its value and that of a defined weight of gold are kept at an equality with one another. It looks therefore as if we could confidently take a step forward, and say that in such a country the quantity of money depends on the world value of gold. Before the war this would have been a true enough statement, and it may come to be true again in the lifetime of those now living: it is worthwhile therefore to consider what, if it be true, are its implications.

The value of gold in its turn depends on the world’s demand for it for all purposes, and on the quantity of it in existence in the world. Gold is demanded not only for use as money and in reserves, but for industrial and decorative purposes, and to be hoarded by the nations of the East : and the fact that it can be absorbed into or ejected from these alternative uses sets a limit to the possible changes in its value which may arise from a change in the demand for it for monetary uses, or from a change in its supply. But from the point of view of any single country, the most important alternative use for gold is its use as money or reserves in other countries; and this becomes on occasion a very important matter, for it means that a gold standard country is liable to be at the mercy of any change in fashion not merely in the methods of decoration or dentistry of its neighbours, but in their methods of paying their bills. For instance, the determination of Germany to acquire a standard money of gold in the [eighteen]’seventies materially restricted the increase of the quantity of money in England.

But alas for the best made pigeon-holes! If we assert that at the present day the quantity of money in every gold standard country, and therefore its value, depends on the world value of gold, we shall be in grave danger of falling once more into Alice’s trouble about the thunder and the lightning. For the world’s demand for gold includes the demand of the particular country which we are considering; and if that country be very large and rich and powerful, the value of gold is not something which she must take as given and settled by forces outside her control, but something which up to a point at least she can affect at will. It is open to such a country to maintain what is in effect an arbitrary standard, and to make the value of gold conform to the value of her money instead of making the value of her money conform to the value of gold. And this she can do while still preserving intact the full trappings of a gold circulation or gold bullion system. For as we have hinted, even where such a system exists it does not by itself constitute an infallible and automatic machine for the preservation of a gold standard. In lesser countries it is still necessary for the monetary authority, by refraining from abuse of the elements of ‘play’ still left in the monetary system, to make the supply of money conform to the gold position: in such a country as we are now considering it is open to the monetary authority, by making full use of these same elements of ‘play,’ to make the supply of money dance to its own sweet pipings.

Now for a number of years, for reasons connected partly with the war and partly with its own inherent strength, the United States has been in such a position as has just been described. More than one-third of the world’s monetary gold is still concentrated in her shores; and she possesses two big elements of ‘play’ in her system — the power of varying considerably in practice the proportion of gold reserves which the Federal Reserve Banks hold against their notes and deposits (p. 47), and the power of substituting for one another two kinds of common money, against one of which the law requires a gold reserve of 100 per cent and against the other only one of 40 per cent (p. 51). Exactly what her monetary aim has been and how far she has attained it, is a difficult question of which more later. At present it is enough for us that she has been deliberately trying to treat gold as a servant and not as a master.

It was for this reason, and for fear that the Red Queen might catch us out, that the definition of a gold standard in the first section of this chapter had to be so carefully framed. For it would be misleading to say that in America the value of money is being kept equal to the value of a defined weight of gold: but it is true even there that the value of money and the value of a defined weight of gold are being kept equal to one another. We are not therefore forced into the inconveniently paradoxical statement that America is not on a gold standard. Nevertheless it is arguable that a truer impression of the state of the world’s monetary affairs would be given by saying that America is on an arbitrary standard, while the rest of the world has climbed back painfully on to a dollar standard.

HT: J. P. Koning

Larry White on the Gold Standard and Me

A little over three months ago on a brutally hot day in Washington DC, I gave a talk about a not yet completed paper at the Mercatus Center Conference on Monetary Rules for a Post-Crisis World. The title of my paper was (and still is) “Rules versus Discretion Historically Contemplated.” I hope to post a draft of the paper soon on SSRN.

One of the attendees at the conference was Larry White who started his graduate training at UCLA just after I had left. When I wrote a post about my talk, Larry responded with a post of his own in which he took issue with some of what I had to say about the gold standard, which I described as the first formal attempt at a legislated monetary rule. Actually, in my talk and my paper, my intention was not as much to criticize the gold standard as it was to criticize the idea, which originated after the gold standard had already been adopted in England, of imposing a fixed numerical rule in addition to the gold standard to control the quantity of banknotes or the total stock of money. The fixed mechanical rule was imposed by an act of Parliament, the Bank Charter Act of 1844. The rule, intended to avoid financial crises such as those experienced in 1825 and 1836, actually led to further crises in 1847, 1857 and 1866 and the latter crises were quelled only after the British government suspended those provisions of the Act preventing the Bank of England from increasing the quantity of banknotes in circulation. So my first point was that the fixed quantitative rule made the gold standard less stable than it would otherwise have been.

My second point was that, in the depths of the Great Depression, a fixed rule freezing the nominal quantity of money was proposed as an alternative rule to the gold standard. It was this rule that one of its originators, Henry Simons, had in mind when he introduced his famous distinction between rules and discretion. Simons had many other reasons for opposing the gold standard, but he introduced the famous rules-discretion dichotomy as a way of convincing those supporters of the gold standard who considered it a necessary bulwark against comprehensive government control over the economy to recognize that his fixed quantity rule would be a far more effective barrier than the gold standard against arbitrary government meddling and intervention in the private sector, because the gold standard, far from constraining the conduct of central banks, granted them broad discretionary authority. The gold standard was an ineffective rule, because it specified only the target pursued by the monetary authority, but not the means of achieving the target. In Simons view, giving the monetary authority to exercise discretion over the instruments used to achieve its target granted the monetary authority far too much discretion for independent unconstrained decision making.

My third point was that Henry Simons himself recognized that the strict quantity rule that he would have liked to introduce could only be made operational effectively if the entire financial system were radically restructured, an outcome that he reluctantly concluded was unattainable. However, his student Milton Friedman convinced himself that a variant of the Simons rule could actually be implemented quite easily, and he therefore argued over the course of almost his entire career that opponents of discretion ought to favor the quantity rule that he favored instead of continuing to support a restoration of the gold standard. However, Friedman was badly mistaken in assuming that his modified quantity rule eliminated discretion in the manner that Simons had wanted, because his quantity rule was defined in terms of a magnitude, the total money stock in the hands of the public, which was a target, not, as he insisted, an instrument, the quantity of money held by the public being dependent on choices made by the public, not just on choices made by the monetary authority.

So my criticism of quantity rules can be read as at least a partial defense of the gold standard against the attacks of those who criticized the gold standard for being insufficiently rigorous in controlling the conduct of central banks.

Let me now respond to some of Larry’s specific comments and criticisms of my post.

[Glasner] suggests that perhaps the earliest monetary rule, in the general sense of a binding pre-commitment for a money issuer, can be seen in the redemption obligations attached to banknotes. The obligation was contractual: A typical banknote pledged that the bank “will pay the bearer on demand” in specie. . . .  He rightly remarks that “convertibility was not originally undertaken as a policy rule; it was undertaken simply as a business expedient” without which the public would not have accepted demand deposits or banknotes.

I wouldn’t characterize the contract in quite the way Glasner does, however, as a “monetary rule to govern the operation of a monetary system.” In a system with many banks of issue, the redemption contract on any one bank’s notes was a commitment from that bank to the holders of those notes only, without anyone intending it as a device to govern the operation of the entire system. The commitment that governs a single bank ipso facto governs an entire monetary system only when that single bank is a central bank, the only bank allowed to issue currency and the repository of the gold reserves of ordinary commercial banks.

It’s hard to write a short description of a system that covers all possible permutations in the system. While I think Larry is correct in noting the difference between the commitment made by any single bank to convert – on demand — its obligations into gold and the legal commitment imposed on an entire system to maintain convertibility into gold, the historical process was rather complicated, because both silver and gold coins circulating in Britain. So the historical fact that British banks were making their obligations convertible into gold was the result of prior decisions that had been made about the legal exchange rate between gold and silver coins, decisions which overvalued gold and undervalued silver, causing full bodied silver coins to disappear from circulation. Given a monetary framework shaped by the legal gold/silver parity established by the British mint, it was inevitable that British banks operating within that framework would make their banknotes convertible into gold not silver.

Under a gold standard with competitive plural note-issuers (a free banking system) holding their own reserves, by contrast, the operation of the monetary system is governed by impersonal market forces rather than by any single agent. This is an important distinction between the properties of a gold standard with free banking and the properties of a gold standard managed by a central bank. The distinction is especially important when it comes to judging whether historical monetary crises and depressions can be accurately described as instances where “the gold standard failed” or instead where “central bank management of the monetary system failed.”

I agree that introducing a central bank into the picture creates the possibility that the actions of the central bank will have a destabilizing effect. But that does not necessarily mean that the actions of the central bank could not also have a stabilizing effect compared to how a pure free-banking system would operate under a gold standard.

As the author of Free Banking and Monetary Reform, Glasner of course knows the distinction well. So I am not here telling him anything he doesn’t know. I am only alerting readers to keep the distinction in mind when they hear or read “the gold standard” being blamed for financial instability. I wish that Glasner had made it more explicit that he is talking about a system run by the Bank of England, not the more automatic type of gold standard with free banking.

But in my book, I did acknowledge that there inherent instabilities associated with a gold standard. That’s why I proposed a system that would aim at stabilizing the average wage level. Almost thirty years on, I have to admit to having my doubts whether that would be the right target to aim for. And those doubts make me more skeptical than I once was about adopting any rigid monetary rule. When it comes to monetary rules, I fear that the best is the enemy of the good.

Glasner highlights the British Parliament’s legislative decision “to restore the convertibility of banknotes issued by the Bank of England into a fixed weight of gold” after a decades-long suspension that began during the Napoleonic wars. He comments:

However, the widely held expectations that the restoration of convertibility of banknotes issued by the Bank of England into gold would produce a stable monetary regime and a stable economy were quickly disappointed, financial crises and depressions occurring in 1825 and again in 1836.

Left unexplained is why the expectations were disappointed, why the monetary regime remained unstable. A reader who hasn’t read Glasner’s other blog entries on the gold standard might think that he is blaming the gold standard as such.

Actually I didn’t mean to blame anyone for the crises of 1825 and 1836. All I meant to do was a) blame the Currency School for agitating for a strict quantitative rule governing the total quantity of banknotes in circulation to be imposed on top of the gold standard, b) point out that the rule that was enacted when Parliament passed the Bank Charter Act of 1844 failed to prevent subsequent crises in 1847, 1857 and 1866, and c) that the crises ended only after the provisions of the Bank Charter Act limiting the issue of banknotes by the Bank of England had been suspended.

My own view is that, because the monopoly Bank of England’s monopoly was not broken up, even with convertibility acting as a long-run constraint, the Bank had the power to create cyclical monetary instability and occasionally did so by (unintentionally) over-issuing and then having to contract suddenly as gold flowed out of its vault — as happened in 1825 and again in 1836. Because the London note-issue was not decentralized, the Bank of England did not experience prompt loss of reserves to rival banks (adverse clearings) as soon as it over-issued. Regulation via the price-specie-flow mechanism (external drain) allowed over-issue to persist longer and grow larger. Correction came only with a delay, and came more harshly than continuous intra-London correction through adverse clearings would have. Bank of England mistakes boggled the entire financial system. It was central bank errors and not the gold standard that disrupted monetary stability after 1821.

Here, I think, we do arrive at a basic theoretical disagreement, because I don’t accept that the price-specie-flow mechanism played any significant role in the international adjustment process. National price levels under the gold standard were positively correlated to a high degree, not negatively correlated, as implied by the price-specie-flow mechanism. Moreover, the Bank Charter Act imposed a fixed quantitative limit on the note issue of all British banks and the Bank of England in particular, so the overissue of banknotes by the Bank of England could not have been the cause of the post-1844 financial crises. If there was excessive credit expansion, it was happening through deposit creation by a great number of competing deposit-creating banks, not the overissue of banknotes by the Bank of England.

This hypothesis about the source of England’s cyclical instability is far from original with me. It was offered during the 1821-1850 period by a number of writers. Some, like Robert Torrens, were members of the Currency School and offered the Currency Principle as a remedy. Others, like James William Gilbart, are better classified as members of the Free Banking School because they argued that competition and adverse clearings would effectively constrain the Bank of England once rival note issuers were allowed in London. Although they offered different remedies, these writers shared the judgment that the Bank of England had over-issued, stimulating an unsustainable boom, then was eventually forced by gold reserve losses to reverse course, instituting a credit crunch. Because Glasner elides the distinction between free banking and central banking in his talk and blog post, he naturally omits the third side in the Currency School-Banking School-Free Banking School debate.

And my view is that Free Bankers like Larry White overestimate the importance of note issue in a banking system in which deposits were rapidly overtaking banknotes as the primary means by which banks extended credit. As Henry Simons, himself, recognized this shift from banknotes to bank deposits was itself stimulated, at least in part, by the Bank Charter Act, which made the extension of credit via banknotes prohibitively costly relative to expansion by deposit creation.

Later in his blog post, Glasner fairly summarizes how a gold standard works when a central bank does not subvert or over-ride its automatic operation:

Given the convertibility commitment, the actual quantity of the monetary instrument that is issued is whatever quantity the public wishes to hold.

But he then immediately remarks:

That, at any rate, was the theory of the gold standard. There were — and are – at least two basic problems with that theory. First, making the value of money equal to the value of gold does not imply that the value of money will be stable unless the value of gold is stable, and there is no necessary reason why the value of gold should be stable. Second, the behavior of a banking system may be such that the banking system will itself destabilize the value of gold, e.g., in periods of distress when the public loses confidence in the solvency of banks and banks simultaneously increase their demands for gold. The resulting increase in the monetary demand for gold drives up the value of gold, triggering a vicious cycle in which the attempt by each to increase his own liquidity impairs the solvency of all.

These two purported “basic problems” prompt me to make two sets of comments:

1 While it is true that the purchasing power of gold was not perfectly stable under the classical gold standard, perfection is not the relevant benchmark. The purchasing power of money was more stable under the classical gold standard than it has been under fiat money standards since the Second World War. Average inflation rates were closer to zero, and the price level was more predictable at medium to long horizons. Whatever Glasner may have meant by “necessary reason,” there certainly is a theoretical reason for this performance: the economics of gold mining make the purchasing power of gold (ppg) mean-reverting in the face of monetary demand and supply shocks. An unusually high ppg encourages additional gold mining, until the ppg declines to the normal long-run value determined by the flow supply and demand for gold. An unusually low ppg discourages mining, until the normal long-run ppg is restored. It is true that permanent changes in the gold mining cost conditions can have a permanent impact on the long-run level of the ppg, but empirically such shocks were smaller than the money supply variations that central banks have produced.

2 The behavior of the banking system is indeed critically important for short-run stability. Instability wasn’t a problem in all countries, so we need to ask why some banking systems were unstable or panic-prone, while others were stable. The US banking system was panic prone in the late 19th century while the Canadian system was not. The English system was panic-prone while the Scottish system was not. The behavioral differences were not random or mere facts of nature, but grew directly from differences in the legal restrictions constraining the banks. The Canadian and Scottish systems, unlike the US and English systems, allowed their banks to adequately diversify, and to respond to peak currency demands, thus allowed banks to be more solvent and more liquid, and thus avoided loss of confidence in the banks. The problem in the US and England was not the gold standard, or a flaw in “the theory of the gold standard,” but ill-conceived legal restrictions that weakened the banking systems.

Larry makes two good points, but I doubt that they are very important in practice. The problem with the value of gold is that there is a very long time lag before the adjustment in the rate of output of new gold will cause the value of gold to revert back to its normal level. The annual output of gold is only about 3 percent of the total stock of gold. If the monetary demand for gold is large relative to the total stock and that demand is unstable, the swing in the overall demand for gold can easily dominate the small resulting change in the annual rate of output. So I do not have much confidence that the mean-reversion characteristic of the purchasing power of gold to be of much help in the short or even the medium term. I also agree with Larry that the Canadian and Scottish banking systems exhibited a lot more stability than the neighboring US and English banking systems. That is an important point, but I don’t think it is decisive. It’s true that there were no bank failures in Canada in the Great Depression. But the absence of bank failures, while certainly a great benefit, did not prevent Canada from suffering a downturn of about the same depth and duration as the US did between 1929 and 1933. The main cause of the Great Depression was the deflation caused by the appreciation of the value of gold. The deflation caused bank failures when banks were small and unstable and did not cause bank failures when banks were large and diversified. But the deflation  was still wreaking havoc on the rest of the economy even though banks weren’t failing.


About Me

David Glasner
Washington, DC

I am an economist in the Washington DC area. My research and writing has been mostly on monetary economics and policy and the history of economics. In my book Free Banking and Monetary Reform, I argued for a non-Monetarist non-Keynesian approach to monetary policy, based on a theory of a competitive supply of money. Over the years, I have become increasingly impressed by the similarities between my approach and that of R. G. Hawtrey and hope to bring Hawtrey's unduly neglected contributions to the attention of a wider audience.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 1,571 other followers

Follow Uneasy Money on WordPress.com