Reporters Phil Izzo and Sudeep Reddy of the Wall Street Journal flagged BIS economist William White’s identification of a developing “Hayekian” artificial boom and inevitable crackup in June of 2007, more than a year after White first published on the topic in 2006. The following is from William White’s April of 2006 article, “Is Price Stability Enough” (pdf), BIS Working Papers #205:
The historical record provides stark evidence that a preceding period of price stability is not sufficient to avoid serious macroeconomic downturns. Perhaps the most telling example is that of the Great Depression in the United States in the 1930s. The period was characterised by massive and continuing losses in terms of both employment and output, accompanied by a cumulative deflation process and associated financial distress in response to accumulated debt. Indeed, almost one third of US banks failed over the course of the 1930s.14 The crucial point is that this out turn was not preceded by any noticeable inflation. Indeed, prices were essentially stable for most of the 1920s and were actually showing signs of measured deflation before the decade drew to a close. Rather, the period was characterised by rapid technological innovation, rising productivity, rapid increases in the prices of equity and real estate and strong fixed investment. Behind these developments were ongoing technical innovations in the financial sector, not least the much greater availability of consumer credit.
Turning to more recent history, Japan has been in a protracted period of sub-par growth for well over a decade, with the GDP deflator falling almost 10 per cent on a cumulative basis. With growth averaging only around 1 per cent annually between 1992 and 2004, the unemployment rate rose from a low of 2 per cent in 1989 to a high of 5.5 per cent in 2001. At the same time, the banking system showed increasing signs of stress, and a number of bankruptcies were recorded in spite of strong and continuous state intervention. As with the earlier US experience, this very poor performance was preceded by a sharp increase in credit, asset prices and fixed investment. Notably, however, there was again no prior acceleration of overt inflationary pressures. As for the Japanese financial sector in the 1980s, it was both subject to the ongoing influence of technological innovation and, more importantly, was in the process of financial deregulation. Still more recently, attention could be drawn first to the financial crisis in South East Asia in the late 1990s. For some countries the costs could be measured as double digits of GDP, with associated increases in unemployment, and the banking systems were also significantly affected. In a number of cases, deflation threatened to, or actually did, emerge. Similar to the US and Japanese cases, these difficulties were not preceded by any inflationary excesses but rather by sharp increases in credit, asset prices and fixed investment. On the financial side, an important influence was exerted by large-scale capital inflows, which subsequently and suddenly reversed as the crisis worsened.
Finally, the same general point could be made about the rather different stresses imposed on the real and financial system by the failure of LTCM and the Russian debt crisis in 1998, and the collapse of global stock markets in 2001. These disruptive incidents also took place in an environment of effective price stability. As with the episodes above, each was preceded by significant evidence of financial overreach (accelerating credit growth, rising leverage, rising asset prices). And, in both the United States and Europe, there was a sharp increase in business investment directed largely to the technology, telecommunications and media sectors believed to epitomise the “New Era” then thought to be emerging.
These facts are as easy to describe as their implications are hard to deny: price stability was not enough to ensure high, sustained growth. What is harder to do is to present an analytical explanation for these costly events, given the absence of the expected catalyst of rising inflation. In the following parts of this section, two relevant points are made. First, recourse is made to some of the central tenets of pre-war Austrian theory and how that model contrasts with the Keynesian analytical approach still used by most central bankers. Second, an attempt is made to show how structural changes in the economy, both real and financial, might have rendered these theoretical insights of more practical relevance today than they were during the 1950s, 1960s, 1970s, and perhaps even the 1980s. In short, history might still matter.
Lessons from the history of economic thought
A useful starting point might be the Keynes-Hayek debate of the early 1930s. This was conducted in the early days of the Great Depression against the backdrop of a previous half century or more of substantial business cycle variations.16 While Hicks (1967) contends that the debate captured the imagination of the economists of the time, it has since been generally forgotten. Both Keynes and Hayek began by accepting some common insights. The first is that a monetary economy is fundamentally different from a barter economy. The second is that both built upon the Wicksellian framework which emphasised the problems associated, in a monetary economy, with the financial rate of interest deviating from the so-called natural rate of interest.17 These similarities noted, their thinking subsequently led them in quite different directions.
Laidler (1999) notes that the IS/LM model, still the workhorse in the stable of most central bankers,18 is essentially a one period model in which the short run and the long run are effectively indistinguishable. Its central message is that deviations between the financial and natural rate will create either deficient or excessive aggregate demand leading to unemployment and (in a fuller model) inflation, respectively. Both are undesirable in themselves. In contrast, the passage of time is a central feature of Austrian theory and, short of the long run, credit creation need not lead to overt inflation. Rather, relative prices are the key to future outcomes. Deviations between the financial and natural rate lead the financial system to create credit which encourages investments that, in the end, fail to prove profitable. The underlying reason for this is that the investments tend to be directed to the production of goods and services for which the level of demand anticipated never in fact materialises. While many have rightly criticised the specifics of Austrian capital theory, the concept of erroneous investment processes driven by credit creation is still noteworthy. Moreover, while most Keynesian models assume a relatively smooth adjustment from one equilibrium to another, the Austrians stressed growing imbalances (cumulative deviations away from equilibrium) and an eventual crisis whose magnitude would reflect the size of the real imbalances that preceded it. The underlying reason for this last observation is that the capital goods produced in the upswing are not fungible, but they are durable. Mistakes then take a long time to work off.
As is now well known, the Austrian approach dropped from sight in most parts of the world, in part because it offered no hope in the face of the crisis of the 1930s. Moreover, the Keynesian approach subsequently offered highly satisfactory performance in the post-war period, barring the 1970s as discussed above. Indeed, since the early 1980s the conventional approach to macro policy has produced truly stellar macroeconomic outcomes. Growth in most industrial countries (excluding Japan) has been both higher and less volatile, while inflation has been sharply lower but also less volatile.
Against this historical background of success, it might seem strange to suggest that the pursuit of low, positive inflation by central banks should be complemented by concerns about financial “excesses” and “imbalances” that are more in the Austrian spirit. It is argued immediately below that there are plausible reasons to warrant such a re-evaluation. As a complement, it will be further argued in Section 4 that there is a reasonable chance that the good performance seen to date might not be sustainable.
Why history might still matter
One reason to warrant a reappraisal of the current conventional approach to monetary policy is that the structure of the global economy has changed remarkably in recent decades. In particular, financial liberalisation has increased the likelihood of boom-bust cycles of the Austrian sort. Moreover, integration of big countries into the world economy and the liberalisation and globalisation of the real economy, as discussed above, appears to have had material affects on the inflation process and the transmission mechanism of monetary policy. Consider each development in turn.
The structural changes in the financial sector in recent decades have been profound. Some combination of technological change and deregulation has led to a quickening process of disintermediation from banks, growing reliance on market processes, globalisation and institutional consolidation.19 In short, we now have a liberalised financial system which seems much more likely to show boom-bust characteristics than the previously repressed one.
Read the whole thing. See also the 2007 Annual Report of the Bank for International Settlements, which William White helped write, released June 24, 2007.
UPDATE: White remained true to his earlier understanding in a paper written exactly one year ago, William White, “Globalization and the Determinants of Domestic Inflation” (pdf), BIS Working Papers #250, March 2008. This section particularly recommends itself:
The US economy seems particularly exposed to the dangers of an economic slowdown due to unwinding “imbalances”. Nevertheless, many other economies could also be affected either directly, by similar imbalances, or indirectly through trade and other linkages. Most of the English-speaking countries would find themselves in the first camp, while to varying degrees virtually every other country could find itself in the second.43 Before turning directly to these exposures, it is worth noting what is meant by imbalances, what their source might be and what evidence there is for such imbalances existing today. A useful definition of “imbalances” might be a significant and sustained deviation from historical norms in important economic and financial variables; for example, asset prices and household saving rates. The supposed danger would be that these series might mean-revert, with effects on the economy and financial markets that could prove disruptive, perhaps seriously so. In the traditional IS/LM framework discussed above, such concepts play no role because it is a single-period model based on an equilibrium defined solely in flow terms. By definition, stocks and cumulative processes are not considered. In contrast, cumulative processes were at the heart of pre-War business cycle theory, particularly as embodied in the work of those of the Austrian school of thought.44 The crux of their thinking was that credit growth had the potential to lead to spending misallocations, particularly for capital investment, which would in the end have to be redressed in a disruptive way.
Charts 6 and 7 above indicated clearly how credit has been both cheap and easily available over the last decade at least. Given that the global economy, as modelled above, was operating for an extended period at below full capacity, generalised inflationary pressures were absent. This meant in turn that policy had to be tightened less during upturns and could be eased more vigorously during downturns.45 Moreover, over the last decade or so, the financial system has itself become much more “complete”, implying an increased diversity of credit channels. For all the advantage this brings, it also brings an increased danger that credit will be extended for purposes that in the end prove unproductive, or imply debt levels too great to be serviceable.46 In sum, over recent years, the fundamental credit factor required to generate imbalances does seem to have been very much in evidence.