Speech Monetary Policy, Growth and Unemployment

It is a pleasure to be here in Melbourne, and to be speaking to the Australia-Israel Chamber of Commerce. I was very pleased to be invited to address this distinguished group so soon after I was appointed. If at the end of this address, you feel that I have still left some important questions unanswered, my suggestion is that your next guest be my opposite number from Israel. Jacob Frenkel, the Governor of the Bank of Israel, is a distinguished economist, but a very busy man. My invitations to him to visit Australia have so far been unsuccessful – perhaps this Chamber can swing the balance.

I want to take this opportunity today to talk about the economy, monetary policy, and some aspects of unemployment. I propose to do this by looking over a longer period than normal. Most economic commentary concentrates on very current events and focuses on the latest monthly or quarterly economic statistics. This is an inevitable result of the news-gathering process and the need for financial markets to reassess after each new piece of information. In this environment, the outlook for the year ahead is considered a long time. In my view, even that is too short – for monetary policy purposes, we should be looking at a period as long as an economic expansion. That may not be a lot of immediate help to most people because they would be unsure of what timespan was encompassed by an economic expansion. As a point of reference, the last two expansions have lasted for about six-and-a-half years, but I think we can do a lot better on this occasion.

Lengthening the Expansion

The current expansion in Australia has been going on for about 5½ years, which is similar to the United States, the United Kingdom and New Zealand, and a fair bit longer than for countries in continental Europe or Japan. Table 1 shows that over that period we have grown at an annual rate of 3½ per cent and had inflation of 2½ per cent. Our growth rate has been higher than nearly all OECD countries, although not as strong as for the countries of east Asia. Our inflation rate is also very good and close to the average for OECD countries.

Table 1: GDP Growth and Inflation
GDP(a) Consumer Prices(a)
Ireland 6.1 2.4
Norway 3.8 2.3
Australia 3.6 2.3
New Zealand 3.6 1.7
United States 2.7 3.1
Canada 2.2 1.7
Denmark 2.2 2.0
Netherlands 2.2 2.9
United Kingdom 2.2 3.0
Finland 1.6 1.9
Germany 1.4 3.1
Japan 1.4 0.8
Spain 1.4 4.8
Belgium 1.3 1.9
France 1.3 1.9
Italy 1.0 4.3
Sweden 1.0 2.1
Switzerland −0.1 2.3
(a) Average annual rate since June 1991

The results are pretty good by international standards, particularly since there should be a lot more expansion still to come. The main reason that I think we still have a lot more to come is that this has been a low-inflation expansion and remains so.

An important reason why inflation has remained low is that monetary policy has been determined to keep it low this time. The centrepiece of our monetary policy approach is a medium-term inflation target endorsed both by the Government and the Reserve Bank. It was also endorsed by the Treasurer in the previous Government and was incorporated into the Accord Mark VIII agreed in June 1995. It is not just something we dreamt up at the Reserve Bank; it has enjoyed wide bipartisan support. And recently, the Industrial Relations Commission made it clear that they felt their own decision had to take note of this framework.

The statement agreed between the Treasurer and myself at the time of my appointment saw low inflation as a requirement for sustained growth, which is another part of the Bank's Charter. Containing inflation is not an end in itself; it is one of the pre-conditions for sustained growth. It is not the only one, but it is obviously the one for which central banks have to accept most responsibility.

Inflation and Growth

We do attract criticism, however, from people who think that because we have an inflation target it means we are not interested in economic growth. This is not true – as I have pointed out on a number of occasions. In fact, the reverse is true: we are interested in sustaining a good inflation performance because we are interested in growth. Let me make two points I have made before:

  • The other way of looking at an inflation target is to say that the aim is to grow as fast as possible consistent with maintaining low inflation, but no faster. We say no faster because post-war history has taught us that allowing inflation to rise actually harms growth prospects in the longer term. Strong growth and long expansions have always gone hand-in-hand with low inflation. Weaker growth and short expansions have occurred in the high-inflation era, particularly the 1970s but also in some parts of the 1980s. This ‘no faster’ idea does not mean that we will panic at unexpectedly strong growth over a short period. It does mean we intend to avoid a ‘full steam ahead and damn the torpedoes’ approach to policy.
  • As I said at the start of the talk, we want a long expansion this time. We do not want the expansion to last only six-and-a-half years and be followed by a recession. The damage, particularly to unemployment, occurs during recessions. The net rise in unemployment in Australia over the past 25 years has not been due to prolonged periods of weak economic growth; it has been concentrated into three relatively short periods of recession (see Graph 1). The best thing monetary policy can do to improve our employment prospects is to prolong the expansion and delay and reduce the size of any subsequent recession. On occasion, that means easing the growth rates back a bit to counter inflationary pressures early, as an alternative to more vigorous application of the brakes when inflation has built more momentum.
Graph 1
Graph 1: Unemployment Rate

Ideally, we should try to eliminate the business cycle (and recessions) altogether, but that would be too utopian an aspiration. The business cycle has been declared dead before, only to re-assert itself. I do believe, however, that if inflation is kept under control, this expansion will be a lot longer than its predecessors and any subsequent downturn kept milder than in the past. Less pressure would be placed on monetary policy in a late-cycle situation if the aim were only to slow the economy, than if it were necessary both to slow the economy and to reduce an entrenched inflation, as has been the case in the past.

Speed Limits

Some people would characterise our approach to monetary policy as incorporating ‘a speed limit’. In a very broad sense this is true. We do not want the economy to grow so fast that it pushes up wage and price inflation to unacceptable rates. I have already explained the reason for that.

However, the concept of a ‘speed limit’ is unhelpful if we try to use it mechanistically. It has been claimed that the Reserve Bank has a rule that the economy cannot be allowed to grow faster than 3½ per cent. I have seen this claim on a number of occasions, but can assure you that it is not true. I do not know where it came from; perhaps from observing historical growth rates, or perhaps from some misinterpretation of things that the Bank has said on occasion. The figure of 3½ per cent is also sometimes regarded as the Government's target. Again, I do not know where this claim came from, but it is probably derived from the fact that government projections, such as those underlying the Budget forward estimates, often assume average growth rates like this, particularly for the out-years of multi-year projections. These assumptions should be seen as what they are – an attempt to base government planning on prudent, conservative assumptions rather than wishful thinking – not as iron-clad limits to growth or short-term ‘targets’.

There are a number of reasons why it does not make sense to specify a particular number as the ‘speed limit’ for the coming year. Among those reasons are the following:

  • Even if we have a clear idea of a long-term average rate of growth, an economy can clearly grow faster than that when it is taking up excess capacity. This would be the case if it was coming off a year when growth was a bit on the soft side (as was 1996, for example) and where inflation was very well contained.
  • It is not possible to ‘fine-tune’ an economy to grow at the rate you want over the space of one year. Neither forecasting accuracy, nor the monetary policy process, permits that degree of accuracy. Even with very good economic policy, the economy will grow faster than average half the time and slower the other half.
  • Just because the economy averaged a particular figure in the past, it does not mean that this average is the one we should project into the future when thinking about potential growth. There is some evidence, as is shown in our Semi-Annual Statement on Monetary Policy, that productivity growth has increased, and hence it is possible that potential growth may have risen. On the other hand, population and labour force growth may well be slower than in the past and this would work in the opposite direction.

I hope I have made it clear that we do not regard economic growth as excessive because it has breached a particular figure – say, 3½ per cent. We would regard it as excessive if it is pushing up wages and prices in a way which would cause our medium-term target to be exceeded. Obviously, the faster is growth in any particular period, all other things equal, the higher the risk of that problem becomes. But we do not have, and it is not sensible to have, a doctrinaire view of particular growth thresholds which, if breached, are assumed to generate problems. The policy process is rather more pragmatic than that, and proceeds by the rather simple technique of examining the evidence for what it says about inflation pressures in the period ahead.

That, to repeat, is the test: not whether a growth rate matches or exceeds a particular pre-determined figure, but whether low inflation performance is being maintained over time. At any point in time we are constantly monitoring wages and prices and making an assessment of the outlook over the next year or so. If we get good news and our assessment is for lower growth in wages and prices, we can, other things equal, let the economy run faster. This may be done by reducing interest rates or, depending on the circumstances, refraining from raising them. If on the other hand we are getting bad news, we may have to tighten or forgo an easing.

This idea is not all that novel. It is very similar to the way the monetary policy process is described in the United States. Our newspapers are full of the latest speculation about what Alan Greenspan – the Chairman of the Federal Reserve Board – is going to do this month, next month and thereafter. The big story of 1996 in the United States was that nearly everyone expected monetary policy to be tightened. The US economy was already at full capacity and still growing strongly, people expected that inflationary pressures would soon emerge, and the Fed would have to tighten. On three occasions – July, September and December – the markets were convinced it was going to happen, but on each occasion the Fed was able to sit on its hands. Why was it possible to delay the tightening for nearly a year? It was not because the economy weakened; in fact, it grew faster than initially anyone had forecast, expanding by over 3 per cent in 1996, well above previous forecasts and economists' estimates of its long-run potential growth rate. It was because the Fed continued to receive constant feedback from the economy indicating that rises in prices and wages remained moderate. None of this means that the US economy is somehow entirely freed from all the constraints it has experienced in the past; only that, to date, inflation performance has been a bit better than they expected. To that extent, the Fed has been able to allow the growth to run just a little longer.

No-one would be more pleased than the Reserve Bank to see a similar outcome for the Australian economy over the next 12 or 18 months – that is, continuing good news on wages and prices, allowing us to grow faster for longer. It was an assessment of improving prospects for inflation, and the associated scope for additional growth, that lay behind the three easings in the second half of 1996. We would like to have seen that pattern of news continued, but the latest data on wages seems to have been a setback to that expectation.

Other Views

Of course, not everyone agrees with this approach. There are those who think a much higher figure for growth is achievable and should be pursued. Their approach would be to nominate a figure – I have seen 5 per cent mentioned – and expect the Reserve Bank to adjust monetary policy to achieve it in the long run (not just for one year).

The trouble with this approach is that it amounts to a gamble with rather poor odds. While we do not wish to take too strong a view on what is ‘too much’ in any particular short period, we need to keep a reasonably sober perspective on what is likely to be feasible over the long run. A growth rate of 5 per cent, for example, is so far above our historical performance that unless you can specify how improvements on the supply side of the economy have lifted the potential growth rate dramatically, it cannot be taken seriously. Now there may well be some significant benefits coming through on the supply side as a result of structural changes and micro-economic reform. For example, reductions in tariffs and increased competition within the domestic market are two that spring to mind. We may also be getting something out of increased flexibility in some parts of the workforce, though the jury is still out at present on exactly how far that process has brought us. But even though there is some tentative evidence of higher productivity, I think we have still got a lot further to go before we could be confident that supply-side reforms have significantly lifted our potential growth rate.

More importantly, our current monetary policy procedures, consistently applied, should guide us to the right answer without having to take such a gamble. If the economy really is capable of much faster non-inflationary growth than in the past, constant evaluation of the feedback we receive from the data should enable us to detect this change. If we receive a flow of information pointing to lower pressures on wages and prices, for example, we can let the economy run faster than earlier relationships might have suggested, as in the Greenspan example. This may not satisfy those with a liking for a punt, because the feedback process takes longer. But it carries less risk of miscalculation.

Longer-term Aspects of Unemployment

So far, we have been talking about monetary policy in relation to the business cycle. I have argued that we should try to maintain the economy in a position of growth for as long a time as possible. This is best achieved by seeking to remain free of major imbalances, and addressing those which do emerge as quickly as possible.

But there is a bigger dimension to Australia's unemployment problem than managing the business cycle. It becomes clear from a consideration of the longer-term history both of our own economy and those of other countries. What I want to suggest is that a common perception that if we could only grow faster, unemployment would go away, is missing an important part of the story.

Consider the data in Table 2, which is for the same group of countries as shown earlier in Table 1. The table sets out the average growth rate per capita in these countries and their unemployment performance. This latter concept is measured by seeing how much higher their current unemployment rate is than it was in the first half of the 1960s. There is one country where the unemployment rate is now lower, namely the United States. At the other end of the spectrum, we have the large European countries where the unemployment rate is enormously higher than it was in the 1960s. We are somewhere in the middle – better than the large Europeans, but not as good as the United States, the United Kingdom or New Zealand.

Table 2: Growth Rate and Change in Unemployment
GDP per capita
Average growth
Unemployment rate
Early 1960s Latest Difference
United States 2.1 5.7 4.9 −0.8
Norway 3.2 1.7 3.5 1.8
Japan 4.5 1.3 3.2 1.9
Canada 2.3 6.0 9.6 3.6
United Kingdom 2.0 1.7 6.1 4.4
Switzerland 1.5 0.0 5.6 5.6
Netherlands 2.3 0.5 6.2 5.7
New Zealand 1.2 0.1 6.4 6.3
Denmark 2.3 1.5 8.0 6.5
Australia 2.2 2.0 8.7 6.7
Ireland 3.7 5.0 11.7 6.7
Sweden 2.0 1.6 8.4 6.8
Italy 3.0 3.5 12.3 8.8
France 2.6 1.4 12.8 11.5
Germany 2.4 0.7 11.2 10.5
Belgium 2.7 2.2 12.9 10.7
Finland 2.7 1.4 15.1 13.7
Spain 3.4 2.3 21.7 19.4

We can then pose the question: have the Europeans suffered because their growth rates were inferior to the United States? Is that the reason for their higher unemployment rates? The answer seems to be no, because column 1 shows that most of the major European countries had quite respectable growth rates over this period; for example, Germany, France, Italy and Spain all grew faster than the United States (or the United Kingdom, New Zealand or us). Their performance on unemployment has been dismal, not because of their growth performance, but despite it. Graph 2 shows that, over this relatively long period, there is no systematic relationship between the growth of the economy and the change in the unemployment rate (if there was, the dots would be clustered around a downward sloping line).

Graph 2
Graph 2:Growth and Unemployment in OECD Countries

There are, no doubt, many aspects to this story, which I cannot go into here. But it seems reasonably apparent that while good economic growth is a very important part of reducing unemployment, growth alone will not be the whole answer. If it was, Europe's unemployment rates would be much more like that of the United States. There is something in the European system which means that growth is not converted into jobs in the same way as it is in the United States.

People who have spent a great deal of time studying these issues have come to the conclusion that the highly regulated labour markets that form part of the continental European world can often work against the interests of job creation. Such labour markets usually have an institutional framework which promotes job security, imposes relatively strict minimum wages and conditions, provides easily accessible sickness and unemployment benefits, and increases trade union involvement in decisions.

As a result, these systems tend to be good at protecting the interests of those in work – the insiders, but are often not helpful to the longer-term interests of the outsiders – those looking for work. Employers are less keen to offer new jobs because of the higher costs and the difficulty of reversing the decision if it becomes necessary. The unemployed, while having a more generous safety-net than under a US-type system, are less likely to be offered a job and less likely to look for one. A high proportion of long-term unemployed is a feature of continental European countries. Ironically, it seems that measures designed to promote security of employment may well have the effect of limiting the extent of employment.

I think we know that there are elements of this story in our own case too. It is not as though the Australian economy has suffered particularly bad growth in the last twenty years. It grew more slowly than in the 1960s, but everyone suffered that slowdown. We have had several periods of reasonably solid growth by international standards. On each occasion, we have, unfortunately, found the limits to growth. But we still have high unemployment by the standards of the United States, New Zealand and the United Kingdom and, more importantly, by the standards of what we ourselves regard as acceptable.

Why does our unemployment performance have rather more in common with the continental European countries than we would like? Is it perhaps because our labour market institutions also have much in common? Our labour market was never as deregulated as the United States and we have not gone through recent deregulations to the same extent as have the United Kingdom and New Zealand. Last year the Government enacted some changes to increase flexibility through the Workplace Relations Act. These changes, simplifying and streamlining arrangements and making it easier to strike bargains, seem to be a move in the right direction. It is too early to judge how substantial the beneficial effects of these changes will be (and it will of course depend on how much use the parties choose to make of the flexibility on offer). On the surface, these reforms were less ambitious than those in the United Kingdom or New Zealand.

Changes to labour market arrangements are, of course, politically difficult. One reason is that the insiders have to give something up. Job security is clearly reduced, and some hard-won entitlements may have to be relinquished. More flexibility in wage setting almost certainly entails a wider dispersion of wage rates – that is more wage inequality – and reform of unemployment benefits would place tougher conditions on the unemployed. It is understandable that many people are unwilling to make these changes and why they foresee such worrying phenomena as the ‘working poor’.

Most people in our society, in other words, see general fairness as a relevant criterion in choosing policy outcomes. As a result, we would not want to go completely down the US route, with such extreme differences in incomes between senior executives of companies and the shop floor. But we do not have to go as far as the United States – there is a happy medium in this, as in most things. All I am suggesting is that we should re-examine our labour market institutions to see whether they are protecting those in jobs at the expense of those looking for jobs.

At the same time, while income inequality may not seem fair, unemployment is not very fair either. Some of our labour market regulations, and many of our attitudes towards labour market outcomes, indeed some of our very ideas about fairness, stem from a time when the unemployment rate was 2 or 3 per cent. It was inconceivable then that unemployment would have an upward trend over twenty years, or that labour market regulation would have any significant role in that process. But the world has changed, and our ideas about what sorts of labour market arrangements are appropriate need to change too. This is already happening, as evidenced by recent reforms. My guess, however, is that this will not complete the process.


I began by talking about the need to lengthen the economic expansion. This, achieved by keeping inflation at bay, is the best contribution monetary policy can make to the long-term health of the economy. That long-term focus does not mean that we are too cautious, that we never take a risk; it simply means that we weigh the risks and rewards of policy actions over longer time horizons than many of the commentators on the economic scene which fill our newspapers and airwaves each day.

We want an expanding economy, and low inflation. That offers the prospect that progress can be made in addressing the rate of unemployment, which remains too high by any standard in Australia. We fear, however, that the debate about whether unemployment will or will not reach supposed ‘targets’ over the next year is missing the main point. Faster growth over the next year or two will help reduce unemployment. But it would be a great shame if that decline obscured the longer-term aspect to unemployment, which endures beyond the business cycle.