Thursday, September 12, 2013


Management of the macro economy has moved to a more sophisticated and thus more medium-term approach, which has increased the focus on the supply side. I want to discuss some less familiar technical terms that have become part of the debate: ‘trend’ growth, the ‘potential’ growth rate, the ‘output gap’ and how macro policy and micro policy fit together.

The meaning of ‘trend’ growth

It has become common to hear the RBA or the treasurer saying the economy (or private consumption or some other key macro variable) is growing at, below or above ‘trend’ - and expecting people to know what this means.

It can be a bit confusing because the Bureau of Statistics uses the word ‘trend’ to mean something quite different. So let’s get that out of the way first. The bureau presents its estimates of key indicators on three bases: original, seasonally adjusted (which allows valid comparisons to be made between adjacent months or quarters) and ‘trend’, which would be better described as ‘smoothed seasonally adjusted’. To remove some of the ‘noise’ in the seasonally adjusted figures - and thus make their underlying direction more readily apparent - you subject the series of observations to an averaging process called a 5-term (for quarterly data) or 13-term (for monthly data) centred ‘Henderson’ moving average. The advantage of this statistical technique is that it makes the direction in which the variable is moving very clear without moving any turning-points. The disadvantage is that you don’t know the final value of an observation until two further quarters (or six further months) have passed. Until then, the estimate is subject to regular revision.

But as the term is used by macro economists (as opposed to the statisticians), ‘trend’ means an indicator’s medium-term to long-term average, with ‘medium-term’ meaning a period of more than two or three years, and long-term meaning maybe 10 or 20 years. For instance, the Rudd government’s economic statement before the 2013 election campaign showed the 30-year average growth rate for real GDP is 3.25 per cent a year.

But here’s where it gets a bit tricky. As well as using ‘trend’ to refer to this backward-looking historical average growth rate, the macro managers also use it to refer to the forward-looking average rate of growth over the future medium term. This forward-looking ‘trend’ is the same as the economy’s ‘potential’ rate of growth.

The meaning of the ‘potential’ growth rate

The potential rate of growth in production (output or real GDP) is the maximum average rate at which it can grow over the medium to longer term without worsening inflation. It thus has similarities to the non-accelerating-inflation rate of unemployment (NAIRU), which is the lowest rate to which unemployment can fall without worsening inflation. Both are measures of full employment or full capacity or the ‘sustainable’ rate of growth.

The three Ps of economic growth

A common way to examine economic growth is to decompose it into ‘the three Ps’: population, participation and the productivity of labour. It is, in fact, a labour-market oriented way of thinking about economic growth. Over the medium to longer term, the economy’s rate of growth can be seen as determined by the rate of growth in the population (in particular, the population of working age), any change in people’s participation in the workforce, and the rate at which the productivity of labour is growing.

Now, when economists treat the backward-looking medium-term ‘trend’ rate of growth as essentially the same as the forward-looking ‘potential’ growth rate they’re implicitly assuming there’s no reason to expect any of the three Ps to change in the coming decade or so from what they’ve been over the past few decades.

But that’s not a reasonable assumption, particularly at present. And this is why Treasury’s more forward-looking, potential-oriented approach to trend has led it to keep revising down its estimate for trend. We can make whatever guesses we like about whether the medium-term rate of improvement in labour productivity will be faster, slower or about the same as it’s been in the past. But the two other Ps - population of working age and the rate of participation - being demographic variables, can be projected with more confidence. And one thing that leaps out from the demography and the ageing of the population is that the great population bulge which is the baby-boomers (those born between 1946 and 1961) is rapidly entering an age cohort with a significantly lower rate of participation than the cohort they are leaving. (This remains true even if it’s also true that many baby-boomers are retiring later than had been expected). If you’ve noticed that Treasury has been revising down its estimate of trend growth, this is the main reason for it. Actually, this process has been occurring for some time. According to Treasury’s PEFO issued early in the 2013 election campaign, forward-looking trend growth was taken to be 3.5 per cent from 1998, then lowered to 3.25 per cent in 2005 and to 3 per cent in 2006. These downward revisions also reflected ‘lower projected productivity growth’. The trend rate is expected to be lowered further to 2.75 per cent some time ‘early next decade’.

The way to determine what Treasury’s estimates are for trend rates of growth in other key macro variables is to look at its annual mechanical ‘projections’ used for the last two years of the budget’s ‘forward estimates’. Trend growth in real GDP of 3 per cent is consistent with annual growth in total employment of 1.5 per cent and an unemployment rate steady at 5 per cent. This implies Treasury’s estimate of the NAIRU - full employment or unemployment’s ‘long-term sustainable level’ - is also 5 per cent. It further implies that, on average, employment needs to grow by 1.5 per cent just to hold the rate of unemployment steady. These estimates, in turn, fit with a CPI inflation rate of 2.5 per cent (ie the mid-point of the RBA’s medium-term target range) and annual growth in nominal GDP of 5.25 per cent (implying inflation as measured by the GDP deflator averages a fraction less than as measured by the CPI because, in the present environment, the terms of trade are assumed to trend down, which reduces growth in the GDP deflator but not the CPI).

While we’re on the three Ps, the charter of budget honesty requires Treasury to produce an ‘intergenerational report’ every five years. We’ve already seen three such reports. Treasury prepares projections of the major classes of government spending for the following 40 years (which always suggest massive growth in federal spending on health care), and estimates the size of the ‘fiscal gap’ if the growth in tax collections is to be capped as a percentage of GDP. But just as interesting are Treasury’s underlying estimates of the three Ps, past and future. In the 2010 IGR, the average annual rate of growth in real GDP over the past 40 years was 3.3 per cent, with growth in the population of working age contributing 1.7 percentage points, change in participation (broadly defined - see below) contributing minus 0.2 points and productivity improvement 1.8 points. Treasury projected that, over the coming 40 years, real GDP will grow at a slower average rate of 2.7 per cent a year, with growth in the working-age population contributing 1.3 percentage points, change in participation contributing minus 0.2 points and productivity improvement contributing 1.6 points. So most of the slowing in growth is explained by slower population growth and the rest by slower productivity growth. Of course, the reason we want to see the economy growing strongly is to increase our material standard of living, which is simply measured as growth in real GDP per person. Treasury estimates that the growth in GDP per person will slow by a smaller gap: 1.9 per cent over the past 40 years versus 1.5 per cent over the coming 40. In this case, the gap is explained equally by slower growth in the population of working age and slower growth in productivity.

Note that the expected slowdown in productivity improvement is no more than Treasury’s best guess. In both 40-year periods the working-age population grew a fraction faster than the population over all, but the difference between working-age and overall growth is expected to change sign in the coming 40 years because of the ageing of the population. It shouldn’t surprise you that participation is expected to fall and thus make a negative contribution to growth in the coming 40 years. But it may surprise you that participation also made a negative contribution in the past 40 years, at a time when we know the participation of women grew strongly. The explanation is that this improvement was more than offset by the higher rate of unemployment during the period and by a decline in average hours worked per worker as the proportion of part-time workers increased.

Speed limits and the ‘output gap’

One term you may be more familiar with is the ‘output gap’. This is the difference between the actual level (not the growth rate) of GDP and the level of potential GDP. In any particular year, actual growth is determined by the strength of aggregate demand, whereas the potential growth rate is the maximum sustainable rate of growth over the medium-term, set by the growth in aggregate supply.

So if in any particular year actual growth exceeds the trend (potential) growth rate this could be taken to mean that aggregate demand is growing faster than aggregate supply, and so is known as an ‘inflationary gap’. But it’s not that simple. Whether or not growth in excess of trend is inflationary depends on whether the economy is already operating at full employment (of all resources, not just labour) or full capacity. If it is then, yes, such growth will be inflationary (and will involve attempting to drive the unemployment rate below the NAIRU). If, however, the economy is operating well below full capacity (with factories and workers not fully employed) then short-term growth in excess of trend will not be inflationary. Indeed, the usual pattern in the first few years after a recession - in which, of course, actually capacity falls to levels way below full capacity - is for growth to far exceed trend. This is not a problem, it’s desirable.

So trend should be thought of as setting a ‘speed limit’ for the rate at which the economy should be growing only after the economy has returned to full capacity. Remember, too, that full capacity is not a fixed point. It grows every year. By how much? On average, by the trend rate of growth - this is what trend measures: the average rate at which the economy’s capacity to produce goods and services expands every year.

Factors leading to growth in the economy’s productive capacity

What factors cause the economy’s supply side - its capacity to produce goods and services - to grow each year? That’s easy: growth in the three Ps. Aggregate supply grows in line with the growth in the labour force, new business investment and public infrastructure, gains in the human capital of the workforce (education and training) and productivity gains arising from economies of scale and advances in technology.

Governments can exert some influence over the growth in the labour force by their policy on immigration and by reforming policies that have the effect of discouraging participation in the labour force. Government policies can also have an influence on the NAIRU. As well, governments can exert some influence over productivity improvement by the incentives and disincentives they create affecting primary research, research and development, and the tax and transfer system generally. They also exert influence by the size and effectiveness of their spending on education and training, as well as on economic infrastructure.

The Gillard government’s budget of May 2011 provided an interesting example of a government using its budget not just to influence aggregate demand but also to influence aggregate supply. It sought to add to supply, especially the supply of labour, particularly skilled labour. It involved a modest expansion of the immigration of skilled workers, the reform and expansion of vocational education and training (TAFE), and the use of sticks and carrots to encourage greater participation in the labour force by disadvantage workers (the long-term unemployed, sole parents and the disabled) and by wives in single-income couples who were under 40 and had no children to look after.

How macro and micro policy fit together

On the surface, the traditional instruments of macroeconomic management - monetary policy and fiscal policy - are very different to the relatively new idea of ‘microeconomic policy’. The traditional instruments are aimed at demand, whereas micro reform is aimed at supply; they work in the short term, whereas micro policies generally effect change only over the medium term. But despite appearances, microeconomic policy is actual an instrument of macro management.

Historically, the macro managers focused solely on managing demand, trying to get it up to the potential growth rate in the recovery phase after recessions, and trying to hold it down to the potential growth rate during booms. The era of micro-economic reform, however, represents the realisation by the economic managers that they can increase the economy’s rate of growth (rather than just keeping it as stable as possible) by improving the flexibility and efficiency of the supply side and also by actually increasing supply.

So the traditional macro instruments are aimed at achieving a stable rate of growth whereas micro policy aims at achieving a higher rate of growth over the medium term, at raising trend. Another way to think of it is this: if avoiding inflation means keeping demand and supply growing at the same rate, one solution (the traditional solution) is to ensure demand doesn’t grow too fast. But another, more creative solution is to do what you can to hasten the growth of supply. To the extent you can do this, you can get faster growth without inflation problems.