(Remarks at the Conference "Alternative Exchange Rate Regimes in the Globalised World", Tallinn)
The conquest of inflation
Those of us here today who once studied economics almost certainly did so in a period of at least mild, if not quite rapid inflation. And those of us who are central bankers have for the majority of our working lives been struggling to cope with inflation. So it's hardly surprising that back in 1983, when the Hong Kong authorities pegged the Hong Kong dollar to the United States dollar, we didn't seriously contemplate any deflationary scenario. The US inflation rate in the early 1980s was averaging some 4% pa and, not possessing any special forecasting capability in this field, we must have reckoned that it would remain near that figure. In addition, the textbooks would have told us that structural and developmental factors would tend to make Hong Kong's inflation rate somewhat faster than that of the US - as indeed it was for the ensuing fourteen years. Neither a significantly lower rate of US inflation nor a situation where Hong Kong's rate was significantly below the US rate would have been considered at all likely.
Yet Hong Kong now finds itself with a consumer price level that is 12% below the peak of four years ago, and with a GDP deflator some 15% below its peak and 22% lower relative to that of the US than it was four years ago. What is the explanation, and can one live with this?
There is general consensus, supported by both recent IMF analysis and research conducted within the HKMA, that Hong Kong's deflation is predominantly cyclical. I shall return to Hong Kong in a minute, but let me first dwell for a moment on the global slowdown in inflation.
Over the past four years the rate of consumer price inflation has averaged 2.5% in the USA, only 1.7% in Euroland, minus 0.3% in Japan, and minus 0.3% in Mainland China. In six other economies of east and south-east Asia taken together, the inflation rate fell to 2.7% on average in these four years as against 4.4% in the preceding four2. This widespread slowdown in inflation, and in some cases the emergence of deflation, is not just a cyclical phenomenon. There are, I believe, some more fundamental institutional and structural factors which have been contributing.
Underlying forces
a. Successful central banks
First of these is the commitment of monetary authorities to combatting inflation and their evident success in doing so (or even over-achievement in some instances). With so many of them represented here today, I do not need to elaborate further.
b. The supply side - history repeats itself
Second is the argument that the world is merely reverting to its more natural historical state of declining price levels, following the interruption of the second world war and ensuing decades3. The current deflationary tendency is, on this view, being driven not only by the IT revolution but also by the evolution of China and some other emerging economies as a huge competitive and price-reducing force in world markets, in a manner reminiscent of the productivity surge and market opening which produced falling price levels in other parts of the globe in the closing decades of the nineteenth century.
Allow me to digress for a moment into this history. As illustrated in the accompanying charts, in the thirty years 1865-95 consumer prices are estimated to have fallen by 48% in the United States and 27% in the United Kingdom - average compound rates of decline of about 2% and 1% respectively4. This is generally regarded as having been a supply-side-driven phenomenon arising from rapid productivity growth alongside abundant labour supply, aided by and the opening-up of new markets globally. Real GDP growth averaged 5-6% (boosted by immigration) in the USA and 1 % in the UK in that period. Subsequently, modest inflation took over, but inflation did not reach a serious scale until the first world war5.
The inter-war years were rather different from the late nineteenth century episode. There was a mixture of deflation and inflation - but with the former being more the result of monetary policy mistakes, such as those associated with the Great Depression, than of supply-side forces such as were so plainly evident half a century earlier. Inflation did not really become re-established on a significant scale until the second world war, after which it persisted for most of the remainder of the twentieth century.
Most recently, the potential problems posed by deflation, particularly in a financially developed world where debt and balance sheets interact, have become apparent, most notably perhaps in Japan.
All in all, there is perhaps a mixed message from these historical examples. It may be possible to live comfortably with a modest pace of deflation, especially if it reflects benign supply-side influences. Nevertheless, monetary policy might usefully be engaged to avert it.
c. Globalisation and the pervasiveness of the law of one price
Returning to the factors behind current price trends, there is one aspect of globalisation which may be especially relevant. This is the rapid expansion of the range of globally traded output (in contrast to efficiency gains within some pre-defined range), with resultant implications for prices. In the past it was common to observe more rapid domestic inflation in economies enjoying faster productivity growth, as wages rose and the prices of non-tradables accelerated, dragging up the consumer price index. This mechanism, based on an assumed clear analytic distinction between price behaviour of tradables and non-tradables, may still be at work, but surely it is not as powerful as it used to be. The range of tradables is expanding relative to that of non-tradables. While it has long been acknowledged that progressively more goods are becoming tradable because of, for example, advances in freight transportation, the revolution in respect of services may now be more significant. Thus, many services which were in the past produced only in the country where they were consumed can nowadays be out-sourced offshore or accessed from somewhere else in the world at the click of a mouse - back-office processing, multi-functional call-centre services, legal advice, logistics management, design services for everything from clothes to buildings, and even certain forms of medical consultation or education, to name but a few. Moreover, buyers of services increasingly elect to travel elsewhere to consume them. Thus the international price elasticity of a wedding ceremony, a game of golf, hospital or health spa treatment, or conference facilities is ever rising. Soon there will remain only very few location-specific activities - heritage tourism perhaps, but even that is not immune from the pressure of the law of one price, if one considers the competing heritage sites around the world.
I see some evidence in Hong Kong of the impact of this process of more broad-based globalisation, and I am sure it is happening elsewhere too. Liberalisation of trade policies and the achievements of the WTO will also have contributed. Thus, deflation is being driven in part by the expansion and intensification of competition across an ever larger range of sectors and countries. This may explain why some price surprises on the macro-policy front in the recent past have been on the downside, but if central banks can digest this ingredient there is no reason why they should be more than temporarily knocked off course by it.
Hong Kong
In Hong Kong, other contributors to deflation over the past few years have been the general strength of the US dollar and the process of price equalisation across the border with the Mainland, although both the IMF and we in the HKMA have found that this latter effect is rather weaker overall than a few well publicised examples might lead one to believe6.
Hong Kong's deflation, of some 12% in CPI terms over four years, seems, however, to be considerably more than could be accounted for by the above list of essentially non-cyclical factors. We therefore infer that much of it has been cyclical. To what extent has the currency board been responsible for this cycle?
About half of the fall in our CPI can be attributed to a collapse of property prices, which feed through both directly and indirectly to the CPI. This in turn was largely the result of the bursting of a previous asset price bubble. This is the part of our deflation which may appear most obviously to be linked to domestic monetary conditions rather than to any global cyclical factors. However, although it is clear that monetary conditions in Hong Kong were, as a consequence of being tied to those in the US through the currency board mechanism, looser during the mid-1990s than one might ideally have wished, and although empirical research suggests that there is a relationship between the property cycle and interest rates, this relationship is not found to be a particularly robust or powerful one. Moreover, there were other factors contributing to the developing property price bubble in the mid-1990s, such as the strong overall performance of the economy, linked to rapid development in Mainland China, some euphoria ahead of the return of Hong Kong to China in 1997, demographics, and the impact of local land release and public housing policies7. The cessation or reversal of such influences helps explain the subsequent downturn.
All in all, therefore, there is quite a range of factors which have come together to produce deflation in Hong Kong. There would be little point in asking precisely how much of deflation can be blamed on the currency peg since, before answering, one would need to specify what sort of monetary policy the authorities would have pursued instead of a currency board. In the absence of a liquidity trap (possibly a significant proviso in this age), it should be possible to choose a policy which could succeed in the narrow objective of eliminating deflation, at whatever cost or benefit in terms of other policy objectives. But the currency board regime rules out such a choice, and in this sense it cannot escape from being ascribed some significance in the explanation of Hong Kong's price deflation.
Does price deflation matter?
In addressing this question, some attention tends to focus on the fact that negative inflation may in certain circumstances imply undesirably high real rates, and that, in particular, it rules out negative real interest rates (at least when measured simplistically by adjusting nominal rates for immediate past price movements), which may be deemed appropriate during a period of deflation. But it is not obvious that real rates of themselves have much of an influence on the intertemporal consumption decision - for example, there is little evidence of people having abstained from purchasing electronic goods over the past several decades despite fairly certain knowledge that prices would be lower tomorrow, and there is a limit to the extent that one can delay buying food or clothing. And, as regards the investment spending, it is not the real interest rate calculated by reference to the CPI which is relevant, but the user cost - the nominal interest rate relative to the expected future net yield from the asset in question, adjusted for depreciation. Prices of different asset classes adjust according to market forces and the user-cost arithmetic, which may not be particularly sensitive to changes in the broader price level per se. In this light, it is at least questionable whether mild overall deflation would be a material impediment to investment decisions.
What might be more significant than impact of deflation via the interest rate is the effect of the deflationary process on real incomes and profits. If prices are driven down by competitive forces, profits are squeezed, which may impact adversely on future investment. It may serve as a harsh signal of the need to restructure or diversify. At the same time, if nominal wages remain unchanged, workers will have a gain in real income. We witnessed this in Hong Kong, but it did little to sustain consumption because of caution over job security. After a while, both nominal pay began to adjust downwards and unemployment began to rise. Although it may still be true that many of those in work are better off (having suffered reductions in pay of less than the amount of price deflation over the four year period), many others have lost their jobs or have had to settle for lower-paid positions. Meanwhile, the reduction in real profits corresponding to the buoyancy of real wages may have dampened investment spending, although supporting empirical evidence is scarce.
Overall it has, in practice, proved difficult to identify what impact deflation of goods and services prices, of itself, may have had on activity. In general, however, because the degree of downward wage flexibility in Hong Kong is probably greater than elsewhere, the economy is believed to be better able than others to cope with this type of deflationary pressure and to avoid severe distributional disequilibria.
Deflation of asset prices is a rather different issue. We have certainly experienced some negative wealth effects, particularly on household expenditure, as a result of the sharp drop in property prices.
One factor enabling Hong Kong to live with deflation perhaps better than some other economies is the resilience of the banking system and its ability to cope with debt deflation. A long tradition of fairly conservative prudential oversight of the banks has left them strong enough to weather the recession.
Although we have demonstrated that we can live with deflation, a modest positive rate of inflation might nevertheless be preferred to our current modest negative rate. And it would certainly be desirable to avoid gyrations of property prices of the sort which have brought such a sharp deflation there in recent years. However, we do not regard the exchange rate as a viable instrument for trying to dispel deflation. We would not consider it worthwhile sacrificing our long track record of disciplined monetary stability for that purpose.
Conclusion - lessons from the 1990s
Allow me to set my concluding remarks in the contest of the specific question posed for this panel, namely "Fix versus float - what have we learned from the 1990s?" My answer falls into two parts.
First, the 1990s served to confirm what had already become apparent from our experience with a currency board in the 1980s, namely that a currency board can be a viable and, arguably, optimal regime for certain types of economy, but not all. The necessary conditions for success include sufficiently flexible costs and prices, a sound banking system, adequate fiscal discipline and the requisite foreign exchange reserves to back the monetary base, preferably free of other encumbrance. Not all these desiderata were satisfied in one instance of a 1990s' currency board, which has since passed into history.
The second lesson concerns the novel feature to emerge late in the 1990s and persist through to the current decade - namely deflation. Society globally has been re-learning how to live with this, after a gap of a lifetime. It seems perfectly possible to live with modest deflation, as our ancestors did, once you have adapted to it. We are reasonably confident that the US economy, to whose monetary stance we in Hong Kong are tied, will not develop in such a way as to make our problem - if that is indeed what it is - any worse. While a floating regime plainly provides greater flexibility to attempt to influence the price level (although such attempts may not always succeed), Hong Kong sees no reason to abandon its monetary regime, which has served the economy well for almost twenty years.
Source: International historical statistics: Europe 1750-1993 / B. R. Mitchell, London: Macmillan Reference; New York, N.Y. : Stockton Press, 1998
1 Remarks delivered at the conference on "Alternative exchange rate regimes in the globalised world", organised by Eesti Pank to mark ten years of the currency board in Estonia, in Tallinn on 11 June 2002.
2 Average for Malaysia, Philippines, Singapore, South Korea, Taiwan and Thailand, weighted by nominal GDP. Indonesia is excluded because of its considerable institutional and political instability in the period - it suffered particularly intense inflation in 1998-99.
3 I am grateful to Eisuke Sakakibara for a lucid exposition of this point in an address last month to the Hong Kong Foreign Exchange and Market Practices Committee. The 69th Annual Report of the Bank for International Settlements (pages 77-81) provides an interesting snapshot of the topic, as seen three years ago.
4 For the UK, the peak-to-trough comparison would be 1873-95, during which the average compound deflation rate was 1.7% pa.
5 My thanks to HKMA colleagues Daryl Ho and Lu Dai for researching the historical data.
6 See Philip Schellekens, "Deflation in Hong Kong SAR", IMF Working Paper, forthcoming; and Jiming Ha and Kelvin Fan, "Price convergence between Hong Kong and the Mainland", HKMA Research Memorandum, forthcoming.
7 For a full analysis see Wensheng Peng, "What drives property prices in Hong Kong?", HKMA Research Memorandum, forthcoming.
