Heikensten: Stabilisation policy: experiences and challenges
First of all, congratulations on your anniversary. For an 80-year-old you look relatively fresh. The organisation I represent first saw the light of day 333 years ago, though despite appearances I haven't been there from the start.
Today I shall be dwelling a good deal on the past to see what lessons we can draw from 80 years of stabilisation policy. Economic developments impose new policy requirements. Over the years, the instruments and objectives of economic policy have been put to the test and renewed in the light of experience. I shall be taking you on a conducted tour of the business cycles to see what they can teach us about the current situation.
Strong belief in spontaneous recovery
Let me start in 1921, the year in which Marcus Wallenberg Senior initiated what subsequently became the Swedish Taxpayers Association. That was in a period when both economic growth and inflation were fluctuating widely. Not counting wars and bad harvests, the crisis Sweden went through was the deepest in modern times.
When trade picked up after World War One and demand grew, prices had risen sharply. The strong economic upswing led in time to overheating, with shortages of fuel, raw materials and labour. The high demand for labour brought unemployment down to 4-5 per cent of trade union membership. In the period from 1917 to 1920 the annual rate of wage increases averaged about 25 per cent. In 1919 and 1920 the balance of trade showed massive deficits of about 5-6 per cent of GDP.
Sweden's inflationary economy elicited mounting criticism. Knut Wicksell, one of our best-known political economists, took an active part in pointing to the grave injustices that rising inflation caused by skewing contracts that had been concluded earlier. Something ought to be done, he said, but no serious efforts to stabilise the economy were actually made.
The subsequent adjustment occurred suddenly and unexpectedly. When the international boom came to an end, demand slumped. From 1920 to 1922, manufacturing output in Sweden dropped about 40 per cent and GDP about 35 per cent (1). Unemployment shot up to 25 per cent. The Swedish economy went into a deflationary spiral. The price level fell about 35 per cent from 1920 to 1922. Wages also plummeted; for an industrial worker, for instance, the wage level fell more than 35 per cent in the course of about 18 months. Firms were caught in a vice, squeezed on all sides by steeply falling prices. As always, the firms that suffered most were those with inflated profits, over-valued assets and large misplaced investments.
The abrupt plunge into a depression was followed by an equally sudden recovery. Manufacturing output rose around 1922 and annual growth in the following years up to 1929 averaged as much as 4 per cent.
So what was the attitude to economic policy in those days? On the whole, the depression, with falling prices and rising unemployment, was seen as a necessary evil, a purgatory that would result in a spontaneous recovery to economic balance.
As regards stabilisation policy in one form or another, the primary responsibility rested with the Riksbank. In practice, however, policy was passive and tardy. A look at the wholesale price index from 1913 to 1920 shows that the increase in Sweden was almost twice that in the United States, for example. In time, the interest rate was raised and the supply of credit was restricted. However, the tighter monetary stance was then retained even after activity had turned downwards. The reason was that the Riksbank tried to bring the Swedish krona back to its par value in 1913 in order to rejoin the gold standard. As a result, at the beginning of 1921 the interest rate was around 6 per cent at a time when prices were falling about 20 per cent.
Fiscal policy was based on other criteria than those of stabilisation policy. Built-in stabilisers did exist, for instance in the form of unemployment support, but the benefits were low in order to strengthen the incentive to search for new employment. This did work - when the upturn came in February 1922 the number of persons on relief work or the dole was over 100,000; by the end of 1923 the figure had dropped to 6,000 - but the flexibility did, of course, exact a price in the form of social hardship.
Stabilisation policy's emergence
In contrast to the lack of an active economic policy in the period around 1920, decision-makers in the 1930s had high ambitions, underpinned by a nascent stabilisation policy theory.
The course of economic events in connection with the crisis of 1930-33 partly resembled what had happened in Sweden a decade earlier. An international boom with unduly high hopes of future profits had led to markedly inflated asset prices. The party ended in a financial crash and Wall Street's collapse in October 1929. A number of banks had to close their doors. On top of all this, the US Federal Reserve imposed credit restrictions.
In Sweden, the depression was not as deep as the one in the early 1920s, though unemployment did rise to much the same level, about 25 per cent. The major difference was that the economy was successfully piloted past a steep price fall. Sweden left the gold standard in September 1931. The Finance Minister, Felix Hamrin, declared that monetary policy's objective would be to "use all the available instruments to maintain the Swedish krona's domestic purchasing power". Deflation was to be combated just as vigorously as inflation. So the Riksbank seems to have been the world's first central bank to base its policy on price stability (2). Our work today with an inflation target is accordingly set in a proud tradition.
When Sweden came off the gold standard the krona rapidly depreciated by about 30 per cent. Export producers found this particularly welcome as it immediately improved their competitive position.
The increased unemployment in the early 1930s elicited a debate about economic policy. Ernst Wigforss, the Social Democrat Finance Minister, saw unemployment as a function of insufficient purchasing power in the economy. To rectify this, he proposed public work projects with market wages, financed with short-term government borrowing. This was a departure from the earlier concentration on monetary policy as perhaps the only effective instrument for stabilisation policy.
Wigforss' proposal also differed from the previous policy of keeping unemployment benefits at a low level so as not to compete with regular wages. Another innovation was the focus on domestic demand instead of on a need to stimulate employment by adjusting prices and wages.
The new direction of policy was nurtured with ideas from the Stockholm school of economics, which included Erik Lindahl, Bertil Olin, Gunnar Myrdal and Erik Lundberg. Their ideas had been developed concurrently with those of John Maynard Keynes. The notion that government ought to intervene in a market economy in order to rectify imbalances permeated economic policy for many years.
Diminishing confidence in the new approach
Keynesianism and its analytical methods spread rapidly. Economists in the 1950s and 1960s devoted a lot of time to computing the multiplier effects of various government measures. For a time it looked as though the economy could be steered by pulling different levers and thereby overcoming the problem of a market economy's instability. The Phillips curve, presented in the late 1950s, demonstrated a negative relationship between nominal wages and unemployment. This further boosted the belief that the economy could be adjusted at will, leaving politicians to opt for their preferred combination of unemployment and inflation. The significance for growth of a fixed value of money fell into oblivion.
In the 1950s and 1960s, Keynesian stabilisation policy seemed to be successful in many countries, not least Sweden. Economic growth was high, supported for example by the post-war reconstruction of Europe, and it was also stable, with only small cyclical fluctuations. Growing budget deficits were tolerated during downward phases but it then proved increasingly difficult for governments to balance the budget over the business cycle.
In the early 1970s, the collapse of the Bretton Woods system for international exchange rate cooperation meant that the earlier constraints on economic policy no longer applied. Countries with growing current-account and budget deficits were now able to obtain finance in the emerging international capital markets.
Economic policy in many Western countries, Sweden included, drifted further and further away from Keynes's basic tenets. Inflation, which in the Bretton Woods era had averaged around 4 per cent, soon rose to twice that. When unemployment began to increase in connection with the oil crises in the 1970s it was evident that Keynesianism faced a crisis. It offered no solution to the problem of stagflation.
Earlier ideas were now revived and new approaches were developed. The difficulties decision-makers face over adjusting imbalances in good time had already been highlighted by economists, above all in the United States, in the late 1960s. There is a considerable risk, for example, that a demand stimulus when activity is low will not have time to act before an upturn has occurred. Policy is therefore liable to accentuate cyclical fluctuations instead of smoothing them.
Perhaps the most pungent criticism came from those who argued that in practice, firms and households act in the light of their economic policy expectations. If, for example, the government tries to stimulate demand by increasing transfer payments and thereby the budget deficit, households will simply save the additional benefits because they expect that, sooner or later, the public sector deficit will have to be financed and this entails either increased taxes or decreased transfers. There would be no increment to demand because households strive to keep consumption stable over time. The only effect is liable to be increased uncertainty about the future.
The criticism, which presupposed that economic agents are rational, also concerned the labour market. In that the Phillips curve disregards a central explanatory variable - inflation expectations - it is not stable over time. A demand stimulus that pushes inflation up erodes real wage levels. Employees will therefore demand compensation for the price increases and the attendant loss of purchasing power. That means that the additional employment which the stimulus was designed to generate will never materialise.
Moreover, Keynesian economic policy had already been criticised in the 1930s on the grounds that stimulating demand in a downturn is considerably easier than tightening policy in an upswing.
Economic developments in Sweden from the 1970s onwards provided plenty of support for Keynesianism's critics. Inflation rose at the same time as keeping total unemployment down proved increasingly difficult. Public spending expanded continuously and particularly in downward phases. Government debt grew even though tax pressure rose. A fundamental loss of competitiveness was countered with repeated devaluations.
Back to price stability
In a number of industrialised countries, economic policy was realigned in the years around 1980. The aim was to provide a good foundation for stable growth by keeping inflation low. In Sweden the corresponding realignment did not occur until the turn of the 1980s. After almost two decades of high inflation, the costs were considerable. The financial excesses that had characterised the late 1980s in particular were ruthlessly exposed when the inflation bubble burst. GDP fell about 5 per cent and unemployment rocketed from levels around 2-3 per cent to two-digit figures.
After a series of heavy attacks, the Swedish krona was left to float in 1992. The fixed exchange rate regime was replaced by a return to price stability as the policy objective, now in the form of a 2 per cent inflation target. Since 1999, price stability is enshrined in law and the Riksbank is more independent in the pursuit of this goal.
While low inflation is admittedly not an ultimate goal for economic policy in general, it is an important prerequisite for generating growth that is stable and high. History clearly demonstrates that a high and fluctuating rate of inflation is incompatible with a permanent improvement in standards of living. A commitment to price stability is therefore one of the foundation stones of any economic policy that aims for rising prosperity and high employment.
The years since the policy realignment have been successful. But this is not only because stabilisation policy is now being conducted differently, with a clear focus on price stability and a balanced budget. It is also a result of all the measures that were introduced around the turn of the 1980s to improve the workings of the economy. These included the deregulation of capital markets, "the tax reform of the century", a realignment of housing policy and so on. Since the upturn in 1993, annual GDP growth has averaged 3-4 per cent, while inflation has stopped at 1 per cent. This is impressive, not least compared with the 1970s and 1980s, when growth averaged little more than 2 per cent and inflation over 8 per cent. Moreover, open unemployment has fallen in just four years from 9 to less than 4 per cent.
Today, when Sweden again faces an economic downturn, the initial position is considerably better than earlier. Inflation is low. There is no sign of large financial imbalances. Foreign trade is generating surpluses. The government budget is also in surplus and government debt has been reduced from about 80 per cent of GDP to 50-55 per cent.
An economic policy synthesis
This brief review clearly illustrates the interaction of theory and practice over the years. New features in economic development raise new questions. New theories provide answers that are sometimes incorporated in practical politics. This influences developments and raises new questions. In view of this, there is every reason to be humble. We are no doubt as far from the end of history in this field as in others.
Perhaps one can discern the outlines of a synthesis between the classical approaches and the ideas of Keynes. As I see it, such a synthesis can rest on three pillars:
- The basic condition for a good economic development is markets that function properly, with rules that are clear and firm.
- A stable value of money combined with sustainable public finances makes an essential contribution to an economic development that is favourable and stable.
- A policy with this direction that is implemented consistently can, with the aid of certain direct cyclical measures, help to stabilise economic activity.
As I have chosen to focus today on issues to do with cyclical activity and stabilisation policy, I have not said much about the first of these pillars but that does not imply a value judgement. On the contrary, the reforms of taxation, regulations and so on have presumably played a crucial part in the good development in the 1990s, both in Sweden and elsewhere. At the same time, recent developments bring out some of the problems we now face. Markets are evidently liable to malfunction in the sense that prices deviate markedly from what is reasonable in the long term. Stock markets have provided evidence of this in recent years and it is also very much the case with the krona's exchange rate. The result can be considerable fluctuations not just in financial markets but in the real economy, too. It is here that I believe the major challenges lie today, both for economic theory and for us more practical decision-makers.
I have already discussed the other two pillars. Attempts to raise demand and employment in the short term entail considerable risks, as we saw in the three decades before 1990. The neo-classical theories, not least those based on rational expectations, have provided a number of explanations for this. But that hardly means that Keynes is dead. Within a policy framework for macroeconomic stability and given sound, long-term rules for the market economy, there ought to be room for cyclical stabilisation, though not for fine tuning.
With a fixed exchange rate regime, monetary policy has to concentrate on maintaining the given rate and leave the stabilisation of demand mainly to fiscal instruments. In the first place that implies creating enough room - sufficiently large reserves - for the automatic stabilisers to be able to act in downward phases. The European Community's Stability & Growth Pact is designed to this end, for example. It is also a question of economic agents being able to rely on inflation remaining low.
With a flexible exchange rate, a large part of the responsibility for stabilisation policy rests in practice with the central bank. Strong confidence in monetary policy and in the ability of the central bank to fulfil its objectives is of primary importance here. It will then normally be feasible to lower interest rates when economic downturns are imminent and vice versa. In that way, a policy focused on price stability will normally also help to stabilise economic activity.
When opinions differ, it is now usually a question of practical judgements. Are the budget reserves sufficiently large for the automatic stabilisers to be given their head? Can one be sure that confidence in the inflation target will not weaken if the interest rate is lowered still more? Perhaps demand can be stimulated deliberately. Timing these matters is still not easy and there is a large risk of cyclical tendencies being reinforced instead of the opposite. But few of us would be prepared to relinquish the possibility entirely (3).
Current challenges
So what does this imply in relation to the present situation in Sweden? How should we handle the combination of clearly slowing activity, persistently large uncertainty as to the direction in which we are travelling, an exchange rate that at present is unprecedentedly weak and inflation above the Riksbank's target?
First let me repeat that the initial position is relatively favourable, even though the strong upswing in recent years would have been an appropriate time to press forward with the work of improving our economy's long-term growth potential. Just a few years ago, not many people would have believed, for instance, that our public debt would fall below the EU average. The remarkably high growth has contributed to this but so has a relatively tight stance based on clear principles for budget policy. Given a downturn of the magnitude most people now foresee, there is a risk of the budget surplus slipping to less than 2 per cent. That need not be dramatic, however. What matters is that the 2 per cent target is met over the business cycle. But the deficit that would be permissible - without endangering long-term confidence in fiscal policy - if activity were to go on weakening is an open question that we should be reluctant to put to the test. Budget policy's other foundation stone, in addition to the targeted surplus, is the spending ceiling. This has been observed to date but the safety margins have shrunk. Additional savings will therefore be needed if activity weakens as much as now seems likely. To exceed the spending ceiling the first time it is exposed to serious strains would harm confidence in the budget system that has been established.
For monetary policy the situation is also relatively good. Inflation has been low for a number of years, frequently lower in fact than the Riksbank's 2 per cent target. Now that it has crept up above 2 per cent in recent months and thus exceeds our target, I am not aware that anyone has taken this as a sign that we do not take the target seriously. Inflation expectations, as measured by a variety of indicators, have also been well in line with our target. This is the fruit of the policy we have followed. Had we continually stretched the limits in the past, there would have been less room in which to take prompt action, as we did a fortnight ago, for instance.
As you are all probably aware, the interest rate hike of 0.25 percentage points in July, to 4.25 per cent, was controversial. Let me therefore underscore that the move was taken just because we were concerned that the confidence we had established in the commitment to low inflation would be eroded when inflation rose rapidly and unexpectedly, clearly exceeding our target, at the same time as the exchange rate weakened dramatically. In many respects this development was confined to Sweden. When we lowered the interest rate in September, the situation was different. It is true that the krona was still weak and inflation above the target, but the terrorist attacks in the United States threatened to almost paralyse an entire world and subdue activity even more.
Looking ahead, there are no grounds for ruling anything out. The repo rate depends as usual on our assessment of inflation one to two years ahead. The current assessment is being made against the background, for example, of international economic activity that threatens to be weaker than we foresaw in the Inflation Report in May. By itself, this points to a more favourable picture of inflation. On the other hand there is, for instance, the weak exchange rate. For a time, moreover, inflation has been higher than we counted on. To some extent this has to do with more transitory price increases for certain goods. But even when these are excluded, inflation has continued to creep up. The ultimate conclusion from our assessment is something to which we will be returning in connection with the publication of the year's third Inflation Report on 16 October.
Notes:
1) Schön, L. (2000), En modern svensk ekonomisk historia (A modern Swedish economic history), SNS Förlag, Stockholm.
2) Berg, C. & Jonung, L. (1999), Pioneering price level targeting: the Swedish experience 1931-1937. J. of Monetary Economics, 43:3.
3) The current debate in the United States is of interest in this context. Even before the tragic events on 11 September, policy had considerable elements of neo-Keynesianism. In a column in Business Week (1 October 2001), Robert Barro, a leading neo-classicist, put forward arguments that I am sure would have earned a gold star from Professor Keynes.
