How has Australia responded to the terms of trade decline?

How has Australia responded to the terms of trade decline?

Domestic Policy Strengthening the Budget and the Economy
Thursday, 02 February 2017

Dr David Gruen, Department of the Prime Minister and Cabinet

David Gruen

A couple of years ago, I gave a speech in which I responded to Professor Ross Garnaut’s 2013 claim that, with the end of the mining investment boom, the Australian economy faced a hard landing (or ‘dog days’ in Ross’s parlance) in the absence of bold productivity-enhancing reform, because the unprecedented size of the downswing in mining investment would prove too disruptive for the domestic real economy and the labour market to absorb.[1]

At the time of my speech (mid 2014), my view of the transition was cautiously optimistic, or as I put it: ‘so far, so good’. But that was early in the mining investment downswing, and it is of interest to re-visit that conclusion with the benefit of a further two-and-a-half years of lived experience, which has put us in sight of the end of the investment downswing.

So as not to keep you in suspense, my punchline is that this further two-and-a-half years has been kind to the cautiously optimistic view. While it is always possible to find ways in which economic outcomes could have been more favourable, in aggregate the Australian economy has adjusted remarkably well to what has been an extremely large shock – certainly the Australian economy has suffered substantially less disruption than the economies of the other major commodity exporting countries.  

My presentation today involves running through a series of charts which puts the boom in context, and provides a sense of how the economy has coped with the biggest shock of its kind in Australia’s economic history.

Let me begin by reminding you how big this boom has been. The first two charts continue to amaze me, even though I have watched as they have evolved over the past several years.

Chart 1, Australia’s terms of trade. The chart shows the index level of Australia’s terms of trade, calculated as export prices divided by import prices, for the period 1869 to 2016, where the index peaked in September quarter 2011 at 136.6 and the latest level for September quarter 2016 is 93.9. This compares to the average index level of 70.2 for the period 1960 to 2000, a period which exhibited much smaller variations in export and import prices. Sourced from ABS statistics and the long term historical index comes from the Reserve Bank of Australia paper by Gillitzer and Kearns, 2006.

The terms of trade boom we have just lived through is easily the biggest sustained boom in our history, as the 5-year centred moving average in Chart 1 makes abundantly clear. It has now been five-and-a-half years since its peak in the September quarter 2011, with the terms of trade having fallen by over 30 per cent since then.

As is widely understood, the boom was largely a consequence of big rises in the prices Australia received for many of its mineral commodity exports. What is less frequently commented upon, but was also a feature of the boom, was the significant rise in the price of a range of rural commodities. Both these developments owe a lot to rising demand from China and, to an increasing extent, India and Indonesia, as they continue their development, urbanisation, and strong growth of their middle classes, with rising disposable incomes to match.

Chart 2, Australia, resources investment as a share of GDP, over the period 1861 to 2018. The time series includes ABS data up to September 2016 and Treasury forecasts to 2017-2018. There is a sharp peak in the series in 2012-2013, at just over 9 per cent of GDP, having moved from a level under 2 per cent of GDP in the year 2000. The latest observation for September 2016, has the resource investment to GDP share around 3.5 per cent of GDP, with forecasts out to 2018 that see the share around 3 per cent of GDP. Sourced from ABS statistics and the long term historical index comes from the Reserve Bank of Australia speech Battelino 2010, which draws on the work of Butlin, 1964 and 1985, and Withers et al 1985.

Resources investment increased from less than 2 per cent of GDP before the boom to around 9 per cent in 2012-13. This resulted in something like a quadrupling in the capital stock in the resource sector, and a tripling in the sector’s productive capacity in the space of a decade. The largest investment was in LNG production capacity, with Australia on track to overtake Qatar as the world’s largest sea-based exporter of LNG. The economic activity and employment that accompanied the investment boom drove a significant reallocation of labour across industries that is now being unwound as investment projects reach completion.

Developments in per capita incomes have been almost as striking as those in the terms of trade and resources investment. The terms of trade boom provided a sizable boost to income growth in the 2000s, sustaining strong growth in per capita incomes. This was in the context of weaker labour productivity growth than we achieved in the 1990s and a weaker contribution from rising labour utilisation than we achieved in the 1980s.

Chart 3, Australia, gross national income capita, for each decade over the period 1960 to 2016. The chart decomposes the percentage points contribution for each component: labour productivity; labour utilisation; terms of trade; and net foreign income. In the decade for the 1960s, labour productivity contributed 2.3 percentage points to the average growth in GNI per capita of 3.0 per cent. In the 1990s, labour productivity contributed 2.19 percentage points to the average growth in GNI per capita of 2.20 per cent. From the year 2000 to 2012, labour productivity contributed 1.4 percentage points, and the terms of trade contributed 1.1 percentage points, to the average growth in GNI per capita of 2.6 per cent. For the most recent period since the peak in the terms of trade, 2012 to 2016, labour productivity contributed 1.6 percentage points, and the terms of trade detracted 1.8 percentage points, compared to average GNI per capita at minus 0.60 per cent.

In the years since its peak in 2011, the terms of trade have detracted from income growth by so much that, even with reasonable labour productivity growth, gross national income (GNI) per capita has been falling. This has also been reflected in gradually falling real average earnings per hour across the economy over the past four years, for the first time in living memory.

Chart 4, Australia, real average earnings per hour, over the period 1978 to 2016. The time series shows compensation of employee earnings per hour worked in real 2015-2016 Australian dollars. Real earnings moves from around 29 dollars per hour to 41 dollars per hour over the period June 1993 to June 2012, an average growth rate of 1.9 per cent per year. Since June 2012 to September 2016, real earnings have been reasonably flat. Sourced from ABS national accounts and consumer price index statistics.

It is as well to recognise that this anomalous behaviour of income and earnings growth is overwhelmingly driven by the rapid decline in the terms of trade Australia experienced over this time. With an end to the trend decline in the terms of trade now in prospect, positive real income growth should return – with its rate of growth again strongly influenced by the rate of labour productivity growth.

Chart 5, Australia, average wages by industry, over the period 1999 to 2016. All indexes are based at 1.0 in March 2004, the start of the terms of trade boom. The chart shows for each industry, the ratio of each industry specific wage price index, relative to the aggregate wage price index. Resource industries include mining and metals manufacturing, and they increase from 1.0 in March 2004 to just under 1.10 by September 2016. The construction industry, remains just under 1.05 by September 2016. Retail trade decreases to 0.95 by September 2013 and remains around this level, while accommodation and food services decreases to 0.93 by September 2011 and remains flat into 2016. Sourced from ABS wage price index statistics.

The adjustment of real wages across industries has supported the economic adjustment. Relative wages adjusted more freely over this boom than was possible in earlier decades, resulting in less disruption to the real economy and, importantly, the labour market.

As the resources boom gathered strength from around 2004, strong demand for labour in the resources, construction and professional services sectors saw wages strengthen relative to those in other sectors. And then, as the boom receded after 2011, wage growth in these sectors slowed to less than the average, enabling the wages in other sectors to catch up somewhat. This pattern seems likely to continue as the resources investment downswing continues to run its course.

As the boom has receded, overall wage moderation (with gradually falling real earnings as shown in Chart 4) has provided strong support for employment, and helped translate the sizable nominal depreciation of the exchange rate into a real depreciation, which has been crucial for the rebalancing of the economy.

The sort of relative wage adjustment shown in Chart 5 didn’t occur in the 1970s or early 1980s, and the result was significant increases in unemployment – an outcome we’ve succeeded in avoiding during the latest episode.

Patterns of employment growth across sectors have mirrored changes in the economy. It is revealing to split the past decade into the five years leading up to the terms of trade peak, 2006-11, and the five years since, 2011-16.

Chart 6, Australia, change in employment by industry, over the decade to November 2016. For each industry, the chart shows the change in the level of employment from 2006 to 2011 and 2011 to 2016, aswell as the percentage increase in the level for the entire period 2006 to 2016. Healthcare is a standout industry where employment has increased by 278,000 for the period 2006 to 2011 and by 204,000 for the period 2011 to 2016, a cumulative increase of around 482,000 persons. For the healthcare industry, this represents an increase of around 45.0 per cent over the entire decade. Retail trade has seen a cumulative increase of around 62,000 persons, with a growth of only 5.4 per cent over the entire period.

Service sectors (health care, professional services and education) have seen strong employment growth over both five-year periods. Unsurprisingly, employment in the resources sector expanded strongly in the first five-year period, but not subsequently. Manufacturing saw a fall in employment over the first five-year period (small in percentage terms), a result of both strong productivity growth reducing the need for labour inputs to produce a given amount of manufacturing output, and the strong real exchange rate reducing competitiveness. By contrast, since 2011, employment in manufacturing has risen slightly.

Chart 7, Australia’s export shares, by destination. The chart shows two pie charts, one for 2005-2006 and one for 2015-2016. Exports to China represented around 11 per cent of our total exports of goods and services in 2005-2006 and this increased to 27 per cent in 2015-2016. Japan’s share moved from 18 per cent in 2005-2006, to only 12 per cent by 2015-2016. The European Union moved from 14 per cent back in 2005-2006, to only 9 per cent by 2015-2016.

The change in the destination of Australia’s exports over the past decade has also been striking. A decade ago, 11 per cent of Australian goods and services exports went to China; now it’s around 27 per cent, with the higher share predominantly coming at the expense of Japan and the European Union. The share of exports to many other countries in our region has remained broadly stable, as those countries provide some of the primary markets for our coal and LNG exports.

Chart 8, Australian exports to China, nominal values in billions of Australian dollars, for the period 2000 to 2016. The chart provides a time series for key components including: iron ore and coal; rural goods; services; and the remaining components are included in other goods. The value of iron ore and coal started at just over 1.0 billion dollars in 2000-2001, peaking at 66.3 billion dollars in 2013-2014, and has now settled at 44.4 billion dollars in 2015-2016. The value of services also started at just over 1.0 billion dollars in 2000-2001, and continues to grow strongly, now at 10.7 billion dollars in 2015-2016.

The huge rise in exports to China was initially dominated by iron ore and coal. Since 2013, however, the value of iron ore and coal exports to China has fallen significantly, with big falls in prices dominating significant rises in volumes. At the same time, the fall in the exchange rate has facilitated a significant diversification of Australia’s exports to China, with big rises in the value of rural exports, as well as tourism and education exports.

Let me now conclude with some reflections on the future.

In common with most advanced economies, we are faced with a period of relatively weak labour productivity growth outside the resources sector, and the ageing of the population is gradually reducing labour participation rates. At a fundamental level, the continued shift in economic weight towards Asia, the disruptive effect of new technology, ageing populations, mass people movements across borders, overlayed with the pressures of sustainability are reshaping the global environment and complicating the strategic challenges facing policy makers.

With Asia moving up the value-added chain, there is competition for more service jobs in the global market. Slow global growth and elevated levels of uncertainty have contributed to more moderate economic growth and job creation in Australia.

Forecasts for global growth, like those produced by the IMF, continue to imply a delayed timetable for global economic recovery. The latest forecasts for 2017 and 2018 show a gradual global recovery, though with growth remaining somewhat below trend. It is worth noting, however, that the growth rate of global potential output appears to be markedly lower that before the financial crisis, so that global trend growth is also lower.

In the United States, labour market indicators – in particular, wages growth and the behaviour of quits – suggest an economy now close to full employment, and certainly closer than at any time since the onset of the financial crisis nearly a decade ago. With the US macro-economy therefore in a stronger cyclical position than the Australian economy, it seems likely that US monetary policy will be tightened several times before the Australian monetary authorities see the need to respond similarly.

If that analysis is correct, it seems likely to be accompanied at some point by further significant weakening of the Australian dollar, which should help support the continued rebalancing of the Australian economy.


[1] I am grateful to Tobias Beckmann and Dan Smith for enthusiastic assistance preparing these remarks, and to Jason McDonald and Nigel Ray for helpful comments.