RDP 2013-03: Implications for the Australian Economy of Strong Growth in Asia 2. Adjustment to a Commodity Price Boom: A Theoretical Perspective

2.1 The ‘Model’

Before turning to the question of how the Australian economy has adjusted to the increase in commodity prices, we first describe how a small open economy such as Australia's might adjust in theory to such a boom, drawing on earlier work by Gregory (1976), Corden and Neary (1982) and Corden (1982, 1984). A key part of the adjustment process is the appreciation of the real exchange rate. Much of that can occur via the nominal exchange rate in the case of an inflation-targeting regime with a flexible exchange rate such as that which exists in Australia. The real exchange rate must still appreciate in the case of a fixed nominal exchange rate regime, but more of this occurs through inflation of wages and prices, as was the case in Australia during the Korean War wool boom of the 1950s.^[4] By helping to insulate the domestic economy from these inflationary pressures, the flexible exchange rate is an integral part of a smooth transition, and one that provides important benefits to the economy given that, in reality, the future path of the terms of trade is not certain.

To understand the elements of the adjustment process, we briefly describe a small open economy model with three sectors: the resource sector, the ‘other tradable’ sector and the non-tradable sector; two factors of production – labour and capital, and (mineral and energy) resource endowments. Most of the resource commodities produced are assumed to be exported. Resource endowments can be thought of as being in fixed supply, but of varying quality (that is, some bodies of ore or deposits of oil and gas are more costly to extract than others). Extraction requires building capital ‘infrastructure’, removing any overburden, digging wells, and building extraction, processing and transport facilities. All of this takes some years to put in place but then lasts for many years.

In the case of Australia's recent experience, the shock to commodity prices that underpinned the increase in the terms of trade is best thought of as being quite persistent. Even though the exact extent of the boom has been, and remains, uncertain, the key point is that commodity prices rise to a high level for quite some time. They are then expected to decline as more supply is brought forth, at home and elsewhere in the world.

A persistent increase in commodity prices leads to a rise in investment in the resource sector. This takes place over the course of a number of years because it is quite costly, or even impossible, to adjust the capital stock rapidly.

The process of adjustment to the boom in the terms of trade can usefully be characterised as occurring in three overlapping phases.^[5] First, commodity prices, and hence the terms of trade, increase significantly. Second, the investment phase entails a build-up in the capital stock in the resource sector over a number of years. This phase is relatively intensive in the use of labour when compared to the third, longer-lived operational phase when production and exports of resources increase significantly. The terms of trade is expected to eventually decline in response to the extra capacity coming on line around the world.

Exactly how the economy responds to the increase in the terms of trade will depend on the response of the exchange rate, interest rates, wages and prices. This will depend in part on the degree of substitutability of labour across the different sectors, as well as the substitutability between tradables and non-tradables in consumption. In what follows, we assume that nominal wages are quite sticky downwards (that is, it takes relatively high levels of unemployment to bring down nominal wages), and that there is some substitutability across sectors, but it may not be not perfect, in which case relative wages across the different sectors might change over time. Similarly, we assume that tradable and non-tradable goods and services are substitutable, but not perfectly so. Finally, we assume that prices adjust gradually, at least in comparison to the nominal exchange rate in the case where the latter is flexible.

There is a long history of research on how a small open economy adjusts to a boom in its resource sector. The canonical model, which divides the economy into a tradable and non-tradable goods sector, was first described by the Australian economists Salter (1959) and Swan (1960). This model was later extended to include three sectors – a booming tradable sector, a lagging tradable sector and a non-tradable sector – in the so-called ‘Dutch Disease models’ (Gregory 1976; Corden and Neary 1982; Corden 1982, 1984). The development of these models was motivated by the need to understand better the linkages between structural changes that were taking place in Australia and a number of other countries in the 1960s and 1970s and the development of the resource export sector.^[6] A number of commentators have drawn on the predictions of these models to explain the adjustments the economy is currently going through (Gruen 2006, 2011; Henry 2006, 2008; Banks 2011; Connolly and Orsmond 2011; Corden 2012). The contribution of the present study is not to ‘reinvent the wheel’, but rather to outline the various adjustments that are taking place in Australia, both from a theoretical and empirical standpoint, and to consider the likely transition path of the economy once the ‘investment phase’ of the boom peaks and the economy moves into the ‘production phase’ of the boom.

2.2 The Adjustment to the Shock

To think about how the economy responds to the positive terms of trade shock in the context of the theoretical model, it is helpful to start by considering the forces acting on the various ‘prices’ in the economy (that is, consumer prices, wages, interest rates and the exchange rate) at the onset of the shock. This provides a guide as to how these variables will need to adjust (relative to their baseline paths) to bring about equilibrium.

A positive shock to the terms of trade, resulting from a rise in commodity prices, increases income accruing to the resource sector and increases that sector's demand for productive inputs. This exerts a degree of inflationary pressure, implying that:

• At initial wages, there will be excess demand for labour emanating from the resource sector. Assuming that the economy begins this adjustment at close to full employment, this will put upward pressure on wages as the resource sector seeks to draw in labour from the ‘other tradable’ and non-tradable sectors.
• Similarly, at the initial exchange rate, there will be an excess demand for Australian dollars, reflecting an increase in the overall demand for Australian resources and assets that will begin to generate a higher return. The pressure for the appreciation can also be viewed as a forward-looking response to the demand for offshore funds that will be needed to help finance investment in the resource sector. To the extent that the exchange rate appreciates, demand for the exports of the ‘other tradable’ sector will decline.
• At initial prices of goods and services, extra income associated with the terms of trade shock implies an excess demand for the output of the non-tradable sector, which will tend to put upward pressure on prices and wages in that sector. To the extent that the exchange rate appreciates, downward pressure on the prices of imported goods could induce substitution away from non-tradable goods and services, or free up income to spend on non-tradable goods and services. Given these opposing forces, the net outcome for demand in the non-tradable sector is not clear a priori. Upward pressure on prices will also be reduced to the extent that profits associated with the terms of trade shock accrue to foreigners and investment is sourced from imports.
• Upward pressure on wages and prices associated with demand from the resource sector, and any excess demand in the non-tradable sector, might require an increase in interest rates in order to contain inflation; this will depend in part on the extent of the exchange rate appreciation. Also, the need for interest rates to rise will be lessened to the extent that inflation expectations remain well anchored and wage pressures in stronger parts of the economy do not spill over to other parts.

Pulling all of this together implies the following. The positive terms of trade shock will cause the nominal exchange rate to appreciate. By itself this pushes up Australian wages in foreign currency terms and thereby frees up labour from the ‘other tradable’ sector – that is, the rise in wages in foreign currency terms represents a loss of competitiveness for this sector. What happens to wages in Australian dollar terms will depend on the extent of the nominal exchange rate appreciation, whether demand for non-tradable goods and services rises or falls, and the substitutability of labour across sectors. It is likely that wages in Australian dollar terms will rise in the resource and non-tradable sectors relative to the ‘other tradable’ sector. To the extent that wages in Australian dollar terms rise overall (again, relative to the baseline), the real exchange rate appreciates further than is implied by the nominal exchange rate appreciation alone. In any case, there is a clear signal for labour to leave the ‘other tradable’ sector and move towards the resource sector (and possibly the non-tradable sector if its demand also rises).

In theory at least, at the time of the terms of trade shock, the nominal exchange rate will jump higher and interest rates will rise. The interest rate differential (vis-à-vis the rest of the world) means that the exchange rate can then be expected to depreciate in a smooth fashion along the path of adjustment. This is because arbitrage in the foreign exchange market requires the expected return on holding Australian dollar assets to be equal to the expected return on foreign currency assets.^[7] The higher interest rate during the course of adjustment will also discourage some investment outside of the resource sector.

What can we say about prices? Domestic inflationary pressures, associated with higher wages and incomes, will lead to higher inflation for non-tradable goods and services but, at the same time, the gradual pass-through of the initial exchange rate appreciation will lead to lower inflation for tradable goods and services (whose prices in foreign currency terms depend to an extent on global considerations). In this way, the appreciation of the exchange rate helps to offset the inflationary impulse from the terms of trade shock, and assists in maintaining inflation in line with the inflation target.

These changes will unwind to some extent as the adjustment process runs its course. As the capital stock in the resource sector reaches its new equilibrium level, demand for labour in the resource sector overall will decline as the relatively labour-intensive investment there declines. Even so, labour demand in the resource sector will still remain somewhat higher than it was prior to the terms of trade shock given the operational needs associated with producing more output. The net reduction in labour demand in the resource sector when moving from the investment to the operational phase of the boom implies that labour needs to be reabsorbed into the non-resource sector. This will tend to put some downward pressure on wage growth, although this effect will be mitigated somewhat to the extent that the demand for non-tradable goods has increased since the onset of the terms of trade shock (that is, if the income effect has outweighed the substitution effect). The reduction in wage growth that does occur, when coupled with some improvement in the competitiveness of the ‘other tradable’ sector due to the exchange rate depreciation (that begins, in theory, immediately after the unexpected boom to the terms of trade and initial appreciation), will help the non-resource sector to absorb the labour being shed by companies that undertook the resource investment.

The price changes seen during the investment phase of the boom will be at least partly reversed during the operational phase of the boom. In particular, the gradual depreciation of the exchange rate (following the initial appreciation) will put upward pressure on tradable inflation over time, while a reduction in wage growth will put downward pressure on non-tradable inflation. These opposing effects may result in little net change to overall inflation. In this case, inflation can again remain in line with the target and the interest rate can return to its neutral level. In line with this discussion, Figure 3 provides a stylised illustration of how key variables might respond to a persistent positive shock to global commodity prices.

In reality, there are a number of reasons why an economy might not adjust in accordance with the theory outlined above; for example, households, firms and authorities do not have perfect knowledge of the timing and magnitude of the terms of trade boom at the outset, and it takes time for firms to adjust to changing circumstances. There are also many forces affecting the economy at a given point in time, making it difficult to isolate the impact of a single shock. Nonetheless, the theory provides a useful heuristic framework for analysing the response of the Australian economy to the resource boom.

Footnotes

For a discussion of the Korean War wool boom episode, see Appendix A of Plumb, Kent and Bishop (2012). [4]

To our knowledge, Gregory (2011b) was the first to cast the current resources boom as one that takes place in three distinct phases. See also Sheehan and Gregory (2012), which is forthcoming in the Australian Economic Review. [5]

As Gregory (2011a) has noted, another purpose of these models was to increase understanding of the potential effects of two policy instruments that had not generally been used in Australia – a large across-the-board tariff cut and changes in the nominal exchange rate. [6]

The theory of uncovered interest parity (UIP), which connects expected changes in the exchange rate to interest differentials, is central to many international macroeconomic models. Yet empirically, UIP is consistently found not to hold (for a survey, see Engel (1996)). [7]