Speech ‘Speed Limits’, Growth and Monetary Policy

In a month's time, the 1999/2000 financial year will end. It will be a further two or three months until we get statistics for economic performance in the June quarter, but we could expect that those data will show that the Australian economy has completed nine years of continuous growth, at an average rate of just over 4 per cent per annum. It has been well documented that this compares favourably with earlier history, and with performance in many, in fact most, other countries.

As good as this performance has been, attention is tending to focus on prospects for sustaining that expansion into its tenth and eleventh years. This is as it should be: that we have a debate of this nature says a lot about the nature of economic performance in this expansion, and the improvements in both the institutional structure of policy making and in policy execution.

The advertised topic for today's discussion is that of ‘speed limits’. I prefer to think in terms of managing the economy's expansion, and so will come at things from that perspective. I do not wish to make any comments about the short-term monetary policy considerations which may be upper-most in the minds of policy-makers. The Bank has recently released the Semi-Annual Statement on Monetary Policy and the Governor has made his regular appearance before the House of Representatives Committee on Economics, Finance and Public Administration, providing ample opportunity for people to understand the Bank's current thinking. I come with no new messages. So you can all relax from the start. Hopefully, I can do the same.

I would like to begin with some observations about the growth performance we have observed, as a way to giving us some basic building blocks for the purposes of discussion about how to sustain growth in the period ahead.

Growth in the 1990s

I begin with a few pertinent facts.

Average annual GDP growth has been remarkably similar over the past three decades. This average conceals substantial fluctuations, of course. The 1970s saw faster growth in the first few years than it did thereafter. The second half of the 1970s was a period of particularly poor economic performance. Each decade has included one serious recession, which explains why average growth rates here are lower than the averages which result from computing growth only in the periods when the economy was expanding.

Table 1: Macroeconomic Aggregates
Average annual rates
1970s 1980s 1990s
Real GDP growth 3.5 3.3 3.5
Standard deviation of quarterly GDP growth 1.4 1.1 0.7
CPI inflation 10.1 8.3 2.3
Population growth 1.7 1.5 1.2
Labour force growth 2.1 2.4 1.4
Employment growth 1.6 2.4 1.3
Productivity growth (per person) 1.8 0.8 2.2
GDP per capita growth 1.8 1.7 2.2

The latter calculation shows an average growth rate of GDP since the low point of the early 1990s recession of 4.1 per cent. This is somewhat lower than the corresponding figure for the 1980s, but higher than the 1970s figure. More importantly, the length of this expansion exceeds those of the previous two, with no sign that it is likely to end in the near term. The obvious point bears repeating: prolonging expansions, and being in a position to ameliorate downturns if and when they occur, makes a tangible difference to economic performance over the long run. The performance of the US economy, currently enjoying the longest ‘peace-time’ expansion in its history, has rightly received accolades in recent years. We should not forget, however, that this follows the second longest expansion (in the 1980s), and that these two were separated by what was a relatively mild downturn by historical standards. Two good expansions had done a lot to make the US economy the pacesetter, even before talk of the ‘new economy’ became common.

Another feature of Australia's performance in the 1990s was that the variability of growth, as measured by the standard deviation of quarterly growth rates in GDP, has been lower than in the 1980s, and about half that seen in the 1970s. In other words, the economy has been more stable. It is surprising that this is not remarked on more often.

Graph 1
Graph 1: Expansions in GDP

There are several possible explanations for this outcome, which are not mutually exclusive. One is that the world economy was more stable. It was, at least as measured by the standard deviation of annual growth rates for the IMF's series for world GDP. This is despite the Asian crisis, which made for more instability in the growth performance of Australia's trading partners. One element of that is that the US economy, in particular, was more stable in the 1990s than in the 1980s. Presumably at least partly as a result of these factors, the variability of Australia's export growth, and of our terms of trade, declined in the 1990s. So the world economy in this respect, though not perhaps in others, was more benign.

Another potential explanation is that policy impulses to the Australian economy were also more stabilising in the 1990s. This has played an important role. The story is partly one of improvements to the institutional structure of policy regimes – the gradual development of the inflation-targeting regime is one example. The story also includes, I am prepared to claim, better execution of policy. Forward-looking policy should usually result in less need for aggressive policy adjustments than otherwise, and therefore less economic instability. All this was helped by clearer articulation of goals and of the policy adjustments in pursuit of them, and by various reforms in product and labour markets which made the economy more flexible.

Of course, the domestic and international elements of the story may be related. For significant periods of time in the 1990s, developments abroad may have made the environment for policy-making easier, at least through the growth and trade channels mentioned above. If the shocks emanating from abroad were smaller, the policy actions responding to them could be as well. On the other hand, while overall exports and terms of trade movements were less variable, the 1990s was a decade of recurrent international financial crises. One of these, the Asian crisis, was a severe test of the robustness and flexibility of the Australian economy and of our policy processes. It was a test, I think everyone agrees, that we collectively passed in good fashion.

All of this was accompanied by a very good inflation performance. After twenty years of high single or low double-digit inflation, inflation as measured by the CPI averaged 2.3 per cent in the 1990s, or 2.8 per cent excluding interest changes. This was a good performance. For the period of the inflation target, since mid 1993, inflation has averaged about 2 per cent.

To be sure, inflation has fallen in virtually every country in the 1990s. But it was already on the way down globally in the 1980s, yet Australia shared in that only at the margin. It was not until the 1990s, and probably the second half of the 1990s, that inflation as a part of life was removed from the Australian economy.

These results – both on growth and inflation – should give economists more confidence in insisting that, even if getting inflation down is costly, maintaining low inflation carries no growth penalty, and in fact is likely to assist long-term growth by helping the price mechanism allocate resources more effectively.

A few other basic macro statistics of the past decade are of interest. Working-age population growth slowed in the 1990s, a result of the baby boom effect waning, and a lower immigration intake. Together with an apparent lessening in the rate at which female participation rates increased compared with the 1980s, this meant that labour-force growth was considerably lower in the 1990s than in the 1980s. Employment growth was similarly lower.

Productivity growth, on the other hand, increased. This meant that even though the increase in the number of workers each year was smaller, output growth was actually slightly higher on average than in the 1980s. With slower overall population growth, per capita GDP growth picked up quite noticeably, and in fact was at its highest since the 1960s. This is another statistic which should get more attention. A pick-up in per capita GDP growth of half a percentage point per year makes a tangible difference to living standards if it is extended over a long period.

Speed Limits

The title of today's seminar talks about speed limits. Here there is potential for ambiguity, if not confusion, but I shall make an attempt to give a perspective.

The discussion of this kind of issue has to confront, at least at the conceptual level, the notion of the economy's potential output. It is clearest to think of the level of potential output as being that which could be sustained, over an extended period, without any upward or downward pressure on inflation.[1]

Note that this is not the engineering concept of potential output – which would perhaps be the maximum physical quantity of goods and services that the economy could produce if every worker and every piece of capital equipment were pushed to their limits. That notion lacks the sustainability element, calibrated by price (or rather inflation) stability, that I think is critical in an economic concept of potential output. Note also that this is explicitly a level concept, but because the stock of productive resources grows each year, that level is increasing all the time: the path of potential output is an upward sloping one. We can then say that the slope of that potential path is the potential growth rate in the longer term – assuming that all resources are ‘fully’ utilised.

Graph 2
Graph 2

The distinction between levels and rates of change here is very important. At any given time, the economy's level of production may be below its potential. In such instances, there is the possibility of the economy growing faster than the long-run growth rate of potential for a while, in order to narrow the gap between the levels of actual and potential output. That is what happens in the recovery from recession. It can happen, in fact, for quite extended periods if the surplus capacity at the starting point is sufficiently large. It is what has been happening, on average, for most of the past nine years in Australia.

In mid 1991, output was a long way below potential. Initially, the upswing in GDP was weak, so that the gap between actual and potential output probably continued to widen until growth picked up. But the overall growth rate since mid 1991 of just over 4 per cent appears to have been a little above the economy's long-run potential growth rate, as best that can be judged from the data of the 1990s, so that the extent of spare capacity has gradually declined – as was intended.

What then are the ideas people have in mind when they talk of speed limits?

The Governor has laid to rest several times perceptions which apparently existed in some quarters a few years ago that the Bank thought there was a GDP speed limit for Australia of 3.5 per cent growth. As the above analysis suggests, even if long-run potential growth was thought to be 3.5 per cent – and I do not endorse that (or any other) particular number – the economy can be allowed to grow faster than long run potential if the level of output is below potential to begin with. In fact it should be, and would be, encouraged to do so in an inflation-targeting framework, unless inflation was forecast to exceed the target for some other reason.

Having considered that issue, are there genuine speed limits?

Clearly, at some point there must be. Economies can overheat if the level of aggregate demand outstrips potential supply persistently. Concerns over this have increased in the US in the past six months, even with greater optimism about supply potential than used to be the case. Additional resources cannot be conjured up indefinitely even in very flexible economies. So if an economy were thought to be operating above potential, growth would need to be held below the long-run potential growth rate to re-align the level of demand with that of potential output.

Speed limit effects might also occur in economies where there is not a general sense of persistent excess demand, but simply because the expansion in demand over a particular short period is so great that supply cannot be increased quickly enough to match it in the short term, even though supply could be increased in the long term. This, I think, is the way to understand the increase in inflation which occurred in Australia in the mid 1990s. A period of exceptional strength in demand and output growth put upward pressure on the growth of labour costs, even though the rate of unemployment was still quite high (at 9 to 10 per cent). The economy was almost certainly below its long-run sustainable level of output, but the short-term dynamics resulted in a noticeable increase in inflation. In other words, even with a large degree of spare capacity, there may sometimes be a limit to how fast that spare capacity can be reduced, without creating inflation problems.

To what extent are either of these two sorts of limits affecting Australia at present?

We cannot be definitive, only approximate. There is less surplus capacity now than there was, say, two or three years ago. At the sorts of output growth rates which appear to be in prospect for the next year or so, we will probably continue to eat into the remaining spare capacity to some extent, though not as quickly as we have done in the past couple of years. But while we have to be alert for signs of bottlenecks and pressure on prices and wages, there is no compelling reason to think that we will significantly exceed the potential level of output in the near term. Likewise, the anticipated growth rates do not seem likely to break the other kind of short-term speed limit – on most people's current forecasts they are likely to be well below what was seen in 1994.

So there are no reasons to think, based on the outlook and policy settings prevailing at present, that the expansion will be cut off by having broken ‘speed limits’.

Sustaining the Growth

There are good prospects for the expansion continuing. What is the role of policy in helping to realise these prospects?

The answer the Bank has given is that monetary policy settings have to be adjusted, as the economy's circumstances change. As the press release of 5 April stated:

The fundamental reason why interest rates have risen in Australia and in most other developed countries over the past year is that the degree of monetary stimulus that was appropriate to earlier circumstances is no longer needed.

You know the general history of the past few years well, so I can be brief. Beginning in mid 1996, and continuing through mid 1997, monetary policy was eased gradually as inflation appeared to be falling back to, and then below, the target. By the time cash rates had fallen to 5 per cent, in July 1997, we had become quite confident that policy was clearly expansionary, and would facilitate stronger growth and, in time, a return of inflation to the target.

As it turned out, the Asian crisis hit at about that time. This reduced Australia's national income sharply, and pushed down the exchange rate. Since the economy had ample capacity for growth, and inflation was still below the target, policy accepted the lower exchange rate, which we expected to put some upward pressure on inflation, and accommodated very strong growth in domestic demand from 1997 to 1999. The expansionary setting was maintained for a considerable time. In fact, at the time in late 1998 when prospects for the world economy appeared to be most grim, rates declined a little further.

By late in 1999, the scenario looked quite different. It was pretty clear that the world economy was in a fully-fledged upswing, with above-average growth rates on the cards for 2000 and 2001. This growth is being led by, but is not confined to, the United States. The Australian economy had continued to record very robust growth. Inflation had increased moderately, and was looking as though it would be back in the target zone in the period ahead. The Bank took the view, as did central banks in many countries, that the low levels of interest rates appropriate for circumstances of downside risks to growth, inflation below target, fears in some quarters of global deflation (and financial distress in some countries, though not Australia) should not be maintained in a period of strengthening world growth, likely waning of disinflationary forces and renewed financial optimism. The interest rates which had facilitated inflation returning to our target were not likely to be the ones which would keep it there. Accordingly, rates have been increased over the ensuing six months.

We do not believe that this sequence of adjustments will stop growth in its tracks. It does mean that growth will not be as strong as it probably would have been, had we left rates at their lows. But it is appropriate that domestic expenditure – where interest rates have their main impact – slow from its 5–6 per cent growth pace of last year, given that the trade sector of the economy will be turning from a major drag on growth to having a neutral or positive impact. I do not for a moment want to imply that we think we can fine-tune this shift in the sources of growth. But think about a scenario involving already strong growth getting a boost from the world recovery, not at an early stage of the expansion, but at quite a mature stage, with inflation already tending to rise, however gradually. Add to this a tendency at times for the exchange rate to weaken. Should monetary policy hold overnight interest rates at the same levels as seen in the immediate aftermath of the early 1990s recession (and even lower in the case of most borrowing rates)?

The Board considered that a much more prudent approach was to make early, moderate adjustments to monetary policy settings, taking them to levels more in keeping with changed circumstances. These moves were made not because things had gone wrong, but to reduce the probability that inflationary imbalances would develop in future. That is, after all, the point of a forward-looking approach to policy. Of course policy makers can't, and don't, promise to be able to avoid all potential future problems. But they have acted in a way which will lessen at least some of the risks, and hence will help to sustain growth in the medium term.


Australia's macroeconomic performance in the 1990s ought to give us reason for optimism about the future. Good growth, rising living standards, declining unemployment and low inflation have coexisted. These are good macroeconomic outcomes. A sensible framework for monetary policy has played a part in this, and will continue to do so.

Are there, as today's topic implies, speed limits to growth? There are some, but they still allow us to make pretty good progress along the economic roadway. Adjustments in monetary policy have to play their part in keeping a reasonable speed for the economy. That means interest rates move as circumstances change: the sensible driver, having crested the hill, eases up on the accelerator as the car starts down the other side. A careful approach to policy maximises our chances of sustaining economic expansion, helping set the pre-conditions for a durable further fall in unemployment, and building on the progress made in the 1990s.


Note that this discussion takes expectations of inflation as given. [1]