|
|
Enron Mail |
Thida also contributed this article from the Economist 25-Sep-99
A new economy for the New World: Inflation may not be as dead as it seems ONCE upon a time sailors went to sea in the belief that the earth was flat, putting them at risk of dropping off the edge. Fortunately the discovery that it was round solved that problem. Economic thinking about monetary policy and inflation has undergon= e a similar about-turn. In the 1960s, the conventional economic wisdom was that monetary policy could reduce unemployment. The theoretical underpinning for this was the Phillips curve, named after Bill Phillips, an economist from New Zealand based at the London School of Economics. Mr Phillips, also a trained engineer, constructed a machine to demonstrate the workings of the economy, using water to represent liquidity. In 1958 he published a study showing that between 1861 and 1957 some kind of trade-off between wage inflation an= d unemployment seemed to have been operating in Britain: when unemployment wa= s high, inflation was low, and vice versa. This seemed to suggest that centra= l banks could permanently reduce unemployment by tolerating a bit more inflation. A decade later, however, two American economists, Milton Friedman and Edmun= d Phelps, challenged this theory. The trade-off between inflation and unemployment, they argued, was only short-term, because once people came to expect higher inflation, they would demand higher wages, and unemployment would rise back to its =0F"natural rate=0F", which depended on the efficien= cy of the labour market. There was no long-term trade-off between inflation and unemployment: in the long run, monetary policy could influence only inflation. If policymakers tried to hold unemployment below its natural rat= e (also known by an acronym, NAIRU , the non-accelerating inflation rate of unemployment), inflation would be pushed ever higher. Just as Messrs Friedman and Phelps had predicted, the level of inflation associated with a given level of unemployment rose through the 1970s, and policymakers had to abandon the Phillips curve. Today there is a broad consensus that monetary policy should focus on holding down inflation. But this does not mean, as is often claimed, that central banks are =0F"inflati= on nutters=0F", cruelly indifferent towards unemployment. If there is no long-term trade-off, low inflation does not permanently chok= e growth. Moreover, by keeping inflation low and stable, a central bank, in effect, stabilises output and jobs. Don Brash, governor of the Reserve Bank of New Zealand, explains how this happens, using 4. The straight line represents the growth in output that the economy can sustain over the long run; the wavy line represents actual output. When the economy is producing below potential (ie, unemployment is above the NAIRU ), at point A , inflation will fall until the =0F"output gap=0F" is eliminated. When output= is above potential, at point B , inflation will rise for as long as demand is above capacity. If inflation is falling (point A ), then a central bank wil= l cut interest rates, helping to boost growth in output and jobs; when inflation is rising (point B ), it will raise interest rates, dampening dow= n growth. Thus if monetary policy focuses on keeping inflation low and stable= , it will automatically help to stabilise employment and growth. The Fed clearly understands this relationship between inflation and the output gap, but initial signs suggest that the ECB may underestimate the extent to which it can safely hold interest rates low in the short run to boost output and jobs. It has repeatedly said that lower interest rates cannot reduce unemployment. That is undoubtedly true in the long run, but i= f an economy is operating below potential and the jobless rate is above the NAIRU , then interest rates can safely be cut, and hence output boosted, without pushing up inflation. Statements by ECB officials often appear to suggest that all of the 10% unemployment rate in the euro area is structura= l and none of it cyclical. However, estimates from the IMF suggest that outpu= t in the area is currently almost 2% below potential, ie, the economy is operating at point A . With growth in the euro area starting to pick up, th= e ECB is already muttering about a rise in interest rates. But it is far too soon: the area still has a significant output gap that will continue to hol= d down inflation. Is a bit of inflation good for you When inflation was in double digits, central banks had a simple rule: bring it down. But now that the rate in almost all rich economies is 2% or less, they are being forced to ask: how low? This is hotly debated, because some economists believe that inflation has economic benefits as well as costs. Consider, first, the costs. When inflation is high, people find it difficul= t to distinguish between changes in average prices and changes in relative prices. For example, a firm cannot tell if a rise in the price of copper reflects general inflation or a scarcity of the metal. This distorts important price signals, leading to a misallocation of resources. Inflation also creates uncertainty about the future, which reduces investment. And lastly, because of the way inflation interacts with tax systems, which are never fully indexed for inflation, it reduces the real return on saving and hence reduces future growth. Inflation in double digits clearly does considerable harm, but what about lower rates? And does 5% inflation do more damage than 2%, or zero? Yes, answers Martin Feldstein, president of the National Bureau of Economic Research*. He believes that even modest inflation can do significant damage through its effect on saving. Tax is levied on nominal interest income, so as inflation rises, the real after-tax return on savings comes down and people save less, which depresses future growth rates. Mr Feldstein estimates that in America a cut in inflation from 2% to zero would permanently increase the level of GDP by 1%. To reduce inflation by two percentage points would involve a one-off loss in output of 5% of GDP , but the discounted present value of the future annual gains, of around 35% of GDP , would far outweigh the loss. Zero inflation, he concludes, is a worth= y goal. Yet there is little empirical evidence that lower inflation rates do noticeably improve growth performance once inflation is below 5%, say. A scatter plot of inflation against growth for a range of countries shows no clear trend (see 5). A study by Robert Barro, an economist at Harvard University, found that a reduction in inflation of one percentage point increases the annual growth rate by a paltry 0.02 of a percentage point. Some economists go much further, arguing that modest inflation, of perhaps 3-4%, is good for growth and employment. Nominal wages, they say, tend to b= e rigid downwards. Workers may be prepared to put up with flat wages when the inflation rate is 3%, which amounts to a decline in real income, but they are reluctant to accept a pay cut in money terms. So if the inflation rate is zero, real wages cannot be adjusted downwards in declining industries or regions, which means that unemployment will rise. Inflation, the argument runs, greases the wheels of the labour market, allowing real wages to adjus= t more smoothly. A widely cited study by the Brookings Institution=0F+, which examined pay i= n America since 1959, confirmed that very few people take nominal wage cuts i= n any year. However, the problem of wage =0F"stickiness=0F" may be overstated= . There have been few periods in the past when inflation has been less than 3% for an extended period, so it is not surprising that falling wages are rare. If inflation were to remain low, resistance to wage cuts might fade. Indeed, i= n Japan wages have been falling over the past two years. And as long as productivity is rising, allowing unit labour costs to fall, there may be no need for nominal pay cuts anyway. America=0F's recent experience certainly suggests that nominal wage rigidities are no cause for concern: unemploymen= t has continued to fall even though average inflation over the past three years has been only 2%. A second common worry about zero inflation is that interest rates cannot fall below zero, so there is no way of allowing real interest rates to become negative to help the economy out of a recession. But the need for negative real interest rates may be exaggerated. Mervyn King, a deputy governor of the Bank of England, argues that negative real interest rates have been rare in America during the past half-century=0F?. During most recessions, low real interest rates have been enough to boost demand. The only time when there might be a problem is if the economy suffers a shock when the output gap is already large and inflation below target because policymakers have failed to react to a previous slump in demand. A pre-emptive policy which aims to prevent inflation going below target as well as above, argues Mr King, minimises the need for negative real interes= t rates. The final and most serious concern is that if central banks aim for zero inflation, prices are more likely to fall for brief periods, and the experiences of the 1930s and Japan today show that deflation can be more dangerous than modest inflation. But falling prices are not necessarily a problem. Before the first world war, a decline in the average price level was quite common during periods of rapid technological change, such as the late 19th century; yet these were also periods of strong growth. This is quite different from the harmful sort of deflation seen during the Great Depression. Moreover, the odd year of falling prices does not matter so lon= g as it does not feed expectations that they will continue to fall, causing households to delay spending. On balance, the benefits of retaining some inflation are probably overstated. Price stability reduces uncertainty and so offers the best economic environment for firms and households. So why do central banks not invariably aim for zero inflation? The answer is that official consumer-price indices tend to overstate the true rate of inflation in all countries, because they do not adjust fully for improvements over time in the quality of goods and services. Estimates suggest that in most countries the official consumer-price index overstates inflation by around 0.5-2.0% a year. This is the best reason why central banks should aim for a slightly positive inflation rate. Flat-earth economics With rates of consumer-price inflation of 2% or under, most rich countries have therefore more or less achieved =0F"price stability=0F". However, that= does not mean that central bankers can go out on the town and celebrate. Despite subdued inflation, the big three central banks all face their own particula= r dilemmas. Japan=0F's economy is still at risk from deflation. The ECB is ha= ving to deal with the added uncertainty of a completely new currency, which has made the bank=0F's monetary policy extra cautious at a time when output in = the euro area is stuck below its potential. As for the American economy, the Fe= d is baffled because inflation is currently lower than any economic model would have predicted. Under the old economic rules, if unemployment fell below the NAIRU (thought to be around 5.5%), inflation would start to rise. This, in turn, implied that the maximum rate of growth the American economy could safely sustain (adding up growth in the labour force and in productivity) was about 2.25-2.5%. Today America=0F's jobless rate stands at a 30-year low of 4.2% = and GDP has grown at an average rate of almost 4% a year over the past three years. According to the textbooks, inflation should be rising. Instead, America=0F's underlying inflation rate has remained subdued. Hence all the = talk about the Goldilocks economy=0F-neither too hot nor too cold. One popular explanation of why strong growth has not pushed up inflation is that the American economy is undergoing a paradigm shift. In this =0F"new economy=0F", it is argued, information technology and increased global competition have opened up new investment opportunities and boosted productivity growth. Mr Greenspan himself frequently sings the praises of the IT revolution, and talks of a =0F"once-a-century phenomenon=0F". Other economists go further and argue that in this new era of low inflation the Fed can throw away its rulebook: the notions of a NAIRU and a maximum sustainable rate of growth have become redundant. Inflation is dead. Then again, perhaps not. America=0F's productivity growth in the non-farm sector has indeed increased, to an annual average of 2% over the past three years, twice the average over the previous 25 years. The big question is whether this is just a cyclical blip or a permanent increase in trend growth. If (a big if) this increase is maintained, it may raise the economy= =0F' s non-inflationary speed limit to 3% a year, say. But that does not mean there are no limits to growth: if it continues at its recent pace and the labour market continues to tighten, inflation is bound to rise. Some people may want to believe in economic miracles, but there is an alternative explanation for America=0F's low rate of inflation. The country= has benefited from four favourable factors. First, weak oil and commodity price= s until earlier this year, and a strong dollar and weak overseas demand, especially in Asia, have held down import prices. Second, an investment boo= m has created excess capacity which has made it hard for firms to pass on wag= e increases in higher prices. Although unemployment is at a 30-year low, America=0F's capacity utilisation is well below its historical average. Thi= rd, non-wage labour costs have been curbed by a booming stockmarket, which has allowed firms to reduce their contributions to employee pension plans. And fourth, the inflation rate has been nudged down by changes in the way of measuring it. By holding down prices, these factors have in turn helped to restrain wages= . But the old relationship between a tight labour market and wage growth remains alive and well: real wage growth has accelerated over the past few years, as you would expect when the unemployment rate is below the NAIRU . This may have been temporarily offset by the favourable factors, but as these effects wear off, inflation is likely to creep up again. Oil prices are already rising; the dollar has slipped; and recovery in the rest of the world will push up import prices. There are also signs that nominal wage costs are now starting to pick up. Goldilocks is in danger of burning her tongue, and the Fed may come to regret that it did not raise interest rates sooner. Despite the increases this summer, real interest rates still remain lower than a year ago. Reports of the death of inflation are therefore much exaggerated. Far from being dead, inflation may have taken on a new, more dangerous guise. *=0F"The Costs and Benefits of Going from Low Inflation to Price Stability= =0F", by Martin Feldstein. NBER working paper No 5469, 1997. =0F+ =0F"The Macroeconomics of Low Inflation=0F", by George Akerlof, Willia= m Dickens and George Perry. Brookings Papers on Economic Activity 1, 1996. =0F? =0F"Challenges for Monetary Policy: New and Old=0F", by Mervyn King. F= ederal Reserve Bank of Kansas City symposium, August 1999. A note from David to fulfill my mission on name-dropping: Martin Feldstein is another old boss. George Akerlof is and Bill Dickens was a Berkeley professor. Feldstein is G.W. Bush's economic adviser, and would probably be a President Bush's nominee for Secty of Treasury or to replace Greenspan. Akerlof is married to Janet Yellen. David I. Levine Associate professor Haas School of Business ph: 510/642-1697 University of California fax: 510/643-1420 Berkeley CA 94720-1900 email: levine@haas.berkeley.edu http://web.haas.berkeley.edu/www/levine/
|