Speech by Mervyn King, Deputy
Governor
Cardiff Business School, Cardiff
University, 18 June 2001
"Hanes Dwy Ddinas" or "A Tale of Two Cities"
Vice-chancellor, ladies and gentlemen.It was once said that economic forecasters are the unfathomable in pursuit of the unpredictable. As a student I found this odd. Whenever England played Wales at rugby the result was wholly predictable. The reason was simple. My fellow-student, Gerald Davies, ran faster than Englishmen, even when he carried the ball and they didn't. Economics is, unfortunately, less straightforward. Yet any decision that involves an elapse of time between its implementation and its effects requires a forecast. Time lags are the essence of the transmission mechanism of monetary policy. And that is why the Monetary Policy Committee spends much time thinking about the likely future path of the economy.
The severe limits on our ability to anticipate future events means that forecasts can be no more than a description of the relative likelihood of a range of possible outcomes. Anyone who presents you with a point forecast for the future path of the economy is either concealing the most interesting part of their analysis, or suffering from self-deception. So the Monetary Policy Committee presents forecasts in terms of probabilities. And it is the balance of risks to the economy which determines monetary policy.
So what are the balance of risks to the British economy at present? Overall, the past five years have been a period of remarkable stability for the UK. Annual GDP growth has averaged 2.8%, and there have been 35 consecutive quarters of positive economic growth. Since the MPC was set up in May 1997, inflation has averaged 2.4%, very close to our target of 2½%, and been in the remarkably narrow range of 2%-3% in 43 out of the past 49 months. This degree of stability reflects well on the new framework for monetary policy in this country. But if I focus today on the risks ahead, it is not simply because as a central banker I know that to every silver lining is attached a cloud, but because the small deviations of inflation from the target over the past four years give a misleading impression of the greater volatility of inflation that we might reasonably expect to see in future.
There are two sources of uncertainty facing the UK at present, one domestic and the other international. Both are related to imbalances. The correction of these imbalances will pose challenges for policy over the next few years.
To see the domestic imbalance look beneath the relative tranquillity of the aggregate data to the significant turbulence among different sectors of the economy. It is obvious that there is an imbalance between those sectors of the economy most exposed to international competition and those sectors facing predominantly domestic competition. To describe this as a difference between manufacturing and services is too simple, but the point is clear. For that part of the economy producing internationally tradable, not just traded, goods and services, output is falling while consumer spending is buoyant. As a result, there is a "tale of two cities" in which some industries are experiencing the worst of times, with declining manufacturing output, and others the best of times, with the latest retail sales figures showing an increase in the volume of sales over the past twelve months of 6.4%.
Such imbalances are not uncommon. Indeed, I first used the phrase "tale of two cities" in early 1995 to describe the contrast between strong demand for manufacturing output and weak retail sales, the opposite of the position today. Since then the economic see-saw has lurched from external to domestic demand, partly as a result of the continuing weakness of the euro which has led to a large and persistent rise in the sterling effective exchange rate.
The scale of the imbalances at home can be seen in figures for domestic and external demand. Whereas aggregate demand and output have risen at a fairly steady rate of around 2¾% over the past five years, domestic demand, supported by rapid money and credit growth, has risen much more rapidly than net external demand. Indeed, private final domestic demand grew at an average rate of nearly 5% over this period, and its growth was in excess of 4% a year in four of the last five years. This imbalance between domestic demand and output has resulted in a rising trade deficit. Net external demand, exports less imports, actually made a negative contribution to total economic growth in each of the past five years. This is the first time that the external contribution to growth has been negative for a period of five consecutive years since the 1870s. And that negative contribution is likely to continue for the foreseeable future.
Imbalances on this scale are unsustainable, although it is far from obvious for how long they can continue. At some point, the weakening balance sheet positions of the private sector will lead to a slowing of consumption and investment as households and firms, respectively, reduce spending to lower the ratio of debt to prospective income. Household debt to income and corporate debt to profit ratios are at historically high levels. Moreover, within the corporate sector the debt levels of the most highly indebted companies - those in the top decile of the distribution of debt to profits ratios - are even higher than in the early 1980s and early 1990s. That holds true even when the telecoms sector is excluded. Spending might, therefore, be expected to decelerate in response to these balance sheet positions. But, for the time being, final domestic demand continues to grow at above trend rates, exacerbating the imbalances within the economy and adding to the risk of a large adjustmen t at some point in the future. The continuing shift of resources to meet the demand for better public services means that private demand must grow much more slowly over the next few years. In turn, that would enable the trade deficit to stabilise and eventually fall back.
The second uncertainty facing the UK is the slowdown in the world economy, and, in particular, the downturn in the United States. In 2000, the world economy grew by 4.7%, the highest rate for twelve years. And, stimulated by an increase in productivity growth, the US growth rate had reached almost 6% a year in the first half of 2000. Such a rate was unlikely to have proved sustainable.
Over the past five years productivity growth in the United States has risen markedly. Labour productivity growth rose by more than 1 percentage point a year to an average of 2½%-3% between 1995 and 2000. Some of this rise was the impact of new technology on efficiency, and some was the result of greater investment in IT - capital deepening - which added significantly to the amount of capital with which each employee was working. But demand rose even more rapidly than the supply of output, leading to a large current account deficit. The external imbalance was the mirror image of the internal imbalance.
So a slowdown in the US was not only unsurprising, it was desirable. But the speed of the turn-round in the US has taken most commentators by surprise. Investment in IT, which had been growing at around 20% a year, actually fell in the first quarter. As a result, and as in the UK, the US has experienced a fall in manufacturing output while consumption and the housing market appear more resilient. Prompt action by the Federal Reserve has so far limited the slowdown to a downturn in inventories and fixed investment. Whether the slowdown in the US will be more prolonged than is embodied in the current consensus view will depend critically upon personal consumption. And, in turn, that will depend upon expectations of future productivity and hence income growth in the US.
If there were to be a reappraisal of the extent of the increase in productivity growth in the US, then there might be a further downward revision of asset prices and a lower path for consumption spending. So far, the evidence, both from consumption directly, and financial markets indirectly, suggests that no major reappraisal has occurred. Share prices have fallen from their peaks of last year. But price to earnings ratios remain well above historical levels. The average P/E ratio in the US since 1900 was around 14. The ratio peaked at over 30, falling back to current levels of well over 20. Interestingly, the divergence of valuation between hi-tech and other companies has virtually disappeared in recent months. There has also been a reduction in the perception of downside risks to equity returns. Returns on shares are usually negatively skewed; that is, there is a greater chance of a return below rather than above its expected value. During the course of this year, the degre e of negative skew has fallen in the US, suggesting that market expectations of further sharp falls is share prices have diminished.
Of course a reappraisal of the future growth of productivity could yet occur. The recent fall in IT spending will, at least in the short run, reduce the contribution to higher productivity growth from capital deepening. It is never easy to detect changes in underlying productivity growth because a large quantity of data is needed to distinguish clearly between changes in trend and cyclical movements. Nevertheless, there is no evidence to suggest that the rate at which computing processing power is increasing - doubling every eighteen months according to Moore's Law - has diminished. Indeed, there have been suggestions that the pace of doubling of processing power is now closer to twelve months.
But there is one reason for caution about estimates of higher future productivity growth, and that is the conventions used to measure output and hence productivity. It might seem obvious that a higher level of gross domestic product, produced by the same number of person-hours, is an improvement in productivity. But gross domestic product, GDP, is not value added. It is the latter which is the correct measure of output and it differs from the former by an appropriate allowance for depreciation of capital goods. The proper measure of output is net domestic product, NDP. In normal circumstances, the growth rates of GDP and NDP are identical. But the two can differ when the average depreciation rate of the capital stock is changing, and that is exactly what has been happening recently. There has been a shift toward greater investment in short-lived assets, such as IT and computer software, and as a result the average depreciation rate has risen. Part of the additional gross output is simply replacing the higher proportion of capital that wears out each year. It makes no sense to include the higher depreciation as part of increased output and hence higher productivity. A change in technology that raises output today at the expense of output tomorrow is not an improvement in productivity. Using estimates of growth rates of GDP will lead to an overestimate of the rise in productivity growth during the transition to higher depreciation rates. The rise in average depreciation rates in recent years suggests that the magnitude of this overestimation could be non-negligible, although more research is needed. Moreover, as John Kay has pointed out, the measurement problem, and the overestimation of output growth, is even more acute when the price of capital goods is falling relative to that of final consumption goods.
What this argument suggests is that economists should be cautious about drawing strong conclusions about the future path of productivity growth. The changes in technology are real and evident in many aspects of our home and business lives. But their implications for aggregate productivity growth are much more uncertain.
The two uncertainties facing the UK economy at present, namely the imbalances both at home and overseas, suggest that the Monetary Policy Committee may be in for a slightly bumpier ride over the next four years than during its first four. So far, the Monetary Policy Committee has managed to achieve an overall balance in the economy despite the contrasting fortunes of the internationally exposed and more sheltered sectors of the economy. But if these large imbalances are to unwind, then the Committee is likely to face a difficult challenge in trying to maintain a balance between total demand and supply. Controlling inflation in future may prove more difficult than over the past four years, and the longer the imbalances persist the greater the risk that the subsequent adjustments to demand, output and inflation are larger than those experienced recently.
Any rebalancing of the economy is likely to be associated with a fall in the real exchange rate. But the magnitude and timing of that are not only uncertain, but also difficult for the MPC to influence. As Sir Samuel Brittan wrote recently about expected changes in exchange rates: "These developments are most likely to happen when they are least expected and least welcome". Although a sharp fall in sterling has been the fervent wish of many manufacturers over the past four years, I recall the words of the person who told me: "think very carefully before you make a wish - it might come true". The only problems worse than those of an excessively strong currency are those of an excessively weak currency.
The existing imbalances pose risks also to the inflation outlook. Prices of services have tended to rise faster than those of goods - about 2% a year faster since 1990. But over the last twelve months the gap between the two inflation rates has exceeded three percentage points. Services inflation has been in the 3-4% range for some time, but goods inflation declined steadily from over 3% in late 1995 to close to zero in the spring. The strength of sterling was the main factor that lay behind this fall in goods inflation. But there are now tentative signs of a pick up in goods prices and the retail sales deflator. RPIY inflation has also risen during the course of this year from 1.5% to 2.8%. The MPC will monitor the price data extremely carefully. But there is likely to be significant short-run volatility in the inflation figures over the coming months, and it would be unwise to read too much into the latest inflation figure because the rise from 2.0% to 2.4%, although the lar gest one-month increase since October 1996, was largely accounted for by jumps in the prices of seasonal food and petrol.
There is, I believe, broad-based support for the objective of setting interest rates to meet the inflation target. Quarterly opinion polls since November 1999, commissioned by the Bank of England and published in our Quarterly Bulletin, provide support for this proposition. But it is clear that the MPC still has much work to do to explain how interest rates affect the economy. Building a constituency for low inflation is a primary objective of the Bank, and one to which we attach great importance. Support for low inflation, and stability more generally, cannot be taken for granted, nor based solely on fading recollections of boom and bust. There is a new generation with little memory of the high and unstable inflation rates of the past. The Bank has, therefore, started an annual competition for schools in which students play the role of the Monetary Policy Committee. The first year's competition attracted over 200 entries, and culminated in the national final at the Bank of England in March. Next year's competition has just been announced, and I would encourage as many schools as possible to enter the competition and pit their wits against the MPC. I am delighted to report that Bassaleg Comprehensive School in Newport, who won last year's South Wales regional heat, were the first team in the UK to enter this year's competition.
Success in explaining the objectives of monetary policy, building the constituency for low inflation, and setting policy in order to meet the inflation target, are all key parts of the role of the Monetary Policy Committee. But the MPC cannot fine-tune the economy quarter by quarter. Its contribution to stability is to meet the symmetric inflation target of 2½%, looking beyond the immediate short-term fluctuations in monthly inflation rates. The Committee will have to watch carefully developments in the economy as they unfold, and be ready to act promptly in either direction. There will always be unpredictable events to which the MPC will need to respond. What is essential for monetary policy is the transparent framework for meeting the inflation target in the medium term, and, in so doing, helping to anchor inflation expectations as close as possible to the 2½% target. At the beginning of next month the MPC will meet for the 50th time. Reaching our half-century has not been w ithout incident. No doubt there will be more surprises on the way to our century. But the symmetric and transparent inflation target provides the best defence against unexpected events.
