RDP 2014-01: Macroeconomic Consequences of Terms of Trade Episodes, Past and Present 4. Economic Policy Settings and Terms of Trade Episodes

This section contrasts past policies and their effects with those in the current episode. Historically, macroeconomic policy frameworks were more rigid and not well equipped to handle shocks. In some cases this led to economic outcomes that were highly disruptive. Today's policy frameworks and institutional structures allow greater flexibility within the economy. While not all parts of the economy have directly benefited, these policies have assisted in producing a relatively smooth adjustment process in a macroeconomic sense relative to past booms. These settings and frameworks may help facilitate the necessary macroeconomic adjustments as the terms of trade declines.

4.1 Monetary Policy

Monetary policy in Australia has operated under various frameworks during the past 150 years. These have included fixed exchange rate regimes, money supply targets and inflation targeting under a floating exchange rate. Each framework has sought, with varying degrees of success, to limit monetary expansion and contain inflation (Cornish 2010).

Policy in earlier episodes was constrained by the fixed or pegged exchange rate regime. This was supported by the use of instruments such as capital controls and qualitative and quantitative controls on credit. In addition, the prevailing wisdom among policymakers in the 1960s and 1970s was that full employment was the primary goal, with less emphasis on inflationary consequences, limiting the scope for the central bank to respond to inflationary pressures stemming from terms of trade booms. There was also little communication with the public regarding monetary policy initiatives (Cagliarini, Kent and Stevens 2010).

The real exchange rate and the terms of trade have moved together quite closely (Figure 11). In every past episode there has been a sizeable real appreciation. The appreciation is a key part of the adjustment mechanism to an upswing in the terms of trade. A reduction in the competitiveness of the other exposed sectors helps to free up resources for the booming commodity sector. However, the real appreciations in the 1950s and 1970s under a fixed or a managed exchange rate were enabled by a sharp increase in inflation of prices and wages (Figures 12 and 6).

Figure 11: Real Exchange Rate and Terms of Trade
Figure 12: Consumer Price Inflation

In the case of the early 1920s and 1970s, the nominal exchange rate was adjusted either prior to, or following, a large movement in the terms of trade. For example, in the 1970s there were successive changes to the exchange rate regime, including a shift to pegging to a trade-weighted basket of currencies; however, these realignments were generally insufficient to insulate the Australian economy. According to Pagan (1987, p 118), ‘… there seems little doubt that the recession [of the mid 1970s], although probably inescapable, was intensified by a policy that opted for an overvalued exchange rate’.[14]

In addition to exchange rate policy, Australian monetary policy used other instruments during most previous terms of trade cycles. For example, following the large spike in wool prices during the Korean War boom, monetary policy was tightened by changing the availability of credit through increasing trading banks' required capital holdings (Schedvin 1992). Credit restrictions also took the form of qualitative controls, but their effectiveness was limited (Schedvin 1992). In 1951/52, as the terms of trade decreased sharply, the stance of monetary policy was relaxed, primarily through central bank loans and the release of ‘special account’ holdings to trading banks (Schedvin 1992).

The rise and subsequent fall in the terms of trade in the 1970s coincided with large fluctuations in net capital inflows, stagflation, political instability and the collapse of the Bretton Woods system of fixed exchange rates, all of which posed difficulties for monetary policy.[15] In the early 1970s, the RBA had difficulty managing the increased capital inflows and capital controls were used.[16] Meanwhile, the banking sector was still highly regulated and the RBA continued to use both qualitative and quantitative controls.[17]

Prior to the current terms of trade cycle, the RBA adopted an inflation-targeting framework, with the goal of keeping consumer inflation between 2 and 3 per cent on average over the medium term. This framework has provided a clear nominal anchor which, together with changes to the structure of the labour market described in Section 4.5, limited the feedback of the boom to economy-wide prices and wages. Notwithstanding the very significant rise in the terms of trade in the current episode, inflation has been well contained, especially when compared to the experience of the 1950s or 1970s. The floating exchange rate and inflation-targeting framework has also provided the RBA with the flexibility to maintain an accommodative stance of monetary policy more recently to support domestic demand since the terms of trade has started to fall.

4.2 Fiscal Policy

For much of the 20th century, fiscal policy has been relatively unresponsive to large movements in the terms of trade. Figure 13 shows the estimated fiscal impact of government spending and taxation policies, as measured by changes in the consolidated public sector fiscal balance as a share of GDP. It highlights the lack of a clear systematic pattern during previous terms of trade episodes.[18] In the most recent episode, fiscal policy has arguably been more countercyclical, especially at the time of the global financial crisis.

Figure 13: Fiscal Impact

Australia's fiscal position in the first half of the 20th century partly reflected a strong desire among policymakers to operate a balanced budget.[19] Outside periods of war, Australia's consolidated revenue fund (CRF) – the fiscal position net of all expenditures and receipts except those to and from loan accounts – remained broadly in balance through the Federation, Roaring Twenties and Korean War wool booms. Indeed, during the Great Depression, governments at the Premiers' Conference declared their determination to balance their respective budgets for 1930/31, and to maintain this stance in future years (Australia, House of Representatives 1930).

Public revenue and expenditure have both become larger as a share of the economy over time. Australia's tax take increased from around 5 per cent of GDP at the turn of the 20th century to more than 20 per cent currently. There was a shift towards income taxation away from indirect taxes, particularly following the uniform income taxation legislation of 1942 (Pincus 1987). Indirect taxes have declined from constituting virtually all federal revenue during the terms of trade boom around Federation to now be less than 20 per cent of the total.[20] Alongside the expansion in the tax base, personal and company income tax rates have declined and are lower than in previous episodes such as in the 1950s and the 1970s (Reinhardt and Steel 2006). Overall, these changes have had a significant effect on the fiscal position in the current episode, including increasing the sensitivity and cyclicality of revenue collections to fluctuations in economic activity and the terms of trade and lifting incentives for workforce participation and investment.

Over the 40 years after Federation, the use of indirect taxes was favoured partly for reasons of feasibility (Reinhardt and Steel 2006). As a consequence, the automatic responsiveness of tax collections to fluctuations in macroeconomic conditions was relatively minor in a number of earlier terms of trade episodes (Pincus 1988). During the Korean War wool boom the efficacy of revaluing the currency or using a tax on income from wool exports was actively discussed.[21] The Government responded in 1950/51 by imposing a levy and mandatory savings of receipts from wool sales (Waterman 1972).

Government expenditure was also relatively small as a share of GDP in the very early terms of trade episodes, limiting the ability to stimulate the economy. During the Federation and the Roaring Twenties episodes, there was a consensus among economists and policymakers that public expenditure only had a limited role in stabilising economic cycles (Garnaut 2005).[22] In the current episode, however, the larger size of the government ensures some buffer to activity via automatic stabilisers to the downturn in the terms of trade, and the attitude among some economists to a discretionary fiscal policy response has changed, particularly for spending on infrastructure (see, for example, Garnaut (2013) and Sheehan and Gregory (2013)).

The level of public debt has been much higher in historical cycles and this also constrained the government's ability to respond to weaker economic conditions with sizeable discretionary fiscal stimulus. Figure 14 shows that total gross public debt was well above 100 per cent of nominal GDP during Federation and up to the Korean War, while in the 1970s it remained above 50 per cent of nominal GDP. The primary reason for the large size of public debt in the past was the legacy of major conflict and the ‘settler nature’ of the Australian economy, the latter requiring high rates of social and economic infrastructure. In contrast, public debt in the current episode is at low levels.[23]

Figure 14: Public Debt in Australia

The availability of credit to the public sector has historically moved broadly with terms of trade cycles and limited the government's ability to borrow from overseas during downswings. During the depressions of the 1890s and the late 1920s, it was difficult to issue new debt, because external capital markets were less receptive and forced domestic capital to fulfil a greater requirement (Butlin, MW 1987; McLean 2013).[24] In contrast, Australia currently has a strong credit rating and external funding is readily available.

4.3 Population Policy

Large-scale immigration has been an enduring feature of Australian history. Booms in the terms of trade have generally coincided with strong growth in net overseas migration (Figure 15). The cyclical pattern partly reflects economic conditions and relative job opportunities. That some terms of trade upswings have occurred following major conflict means that geopolitical issues such as displaced workers after World War II have also contributed. In contrast, downswings have coincided with a slowing in growth in migration and increasing unemployment.

Figure 15: Net Overseas Migration

Migration policy has been used as a form of stabilisation policy during previous cycles (Butlin, Barnard and Pincus 1982). Immigration was used to meet skill shortages, affect demand for non-tradeables and to foster growth in trade-exposed and import-competing industries (such as manufacturing and agriculture). For example, skilled immigration was encouraged in the 1920s, which was important in the textiles and iron & steel industries (Pope 1987). In the 1950s, foreign migrant workers participated extensively in public works, while in the 1960s and early 1970s they were a key input into trade-exposed manufacturing and non-tradeable industries such as construction (Meredith and Dyster 1999).

4.4 Trade Protection

Trade protection in Australia has had various motives over the past century and therefore estimates of tariff levels do not display a close relationship to terms of trade episodes (Figure 16).[25] However, one of the main motivating factors for increasing protectionism, particularly in the first three episodes, was to diversify the economy, in order to shield it from swings in commodity prices.[26] Other reasons, such as to promote exports more generally, to support greater employment at given real wages and other distributional goals, ‘defence considerations’ and ‘national pride’ were also sometimes espoused (Anderson 1987; Henry 2007).

Figure 16: Average Tariff Rates

At times in Australia's history, protectionist policies have been a partial and implicit response to swings in the terms of trade. For example, tariff levels were raised sharply in the early 1920s episode in response to lower transport and production costs, the cessation of war-related demand and the resumption of global trade (Anderson 1987). Further sizeable increases in tariffs subsequently occurred in the early 1930s after the terms of trade had fallen. This coincided with a global shift towards protectionism and a collapse in global trade (Eichengreen and Irwin 1995). Another influence is likely to have been the Brigden Enquiry of 1929, which sought redistribution of income to labour away from capital, in part to support increased immigration (Anderson 1987).[27] In the current episode, Australia's tariff rates have been the lowest since Federation.

4.5 Labour Market

Another way a commodity price upswing can affect the broader economy is via the labour market and wages. While terms of trade booms have generally resulted in a reduction in the unemployment rate, many episodes have also resulted in a sharp increase in wages due to relatively inflexible labour market arrangements (Figure 17).[28]

Figure 17: Average Weekly Earnings

A federal basic (social) wage played a central role until the late 1960s, with automatic wage indexation introduced in the early 1920s. Margins above the social wage differed across jobs, based partially on skills. The centralised institutional structure probably limited the scope for sectoral wage differentials in past major terms of trade upswings (Freebairn and Withers 1977).[29] In addition, automatic indexation meant that high inflation could feed into wages, further increasing inflation, although in some downswings wages policy was adjusted.[30]

The institutional structure of the labour market during the current episode has been the most flexible over any expansion since Federation; a considerable increase in relative wages in the resources sector and a more decentralised industrial system facilitated a relatively low unemployment rate during the upswing in the terms of trade without creating substantial inflationary pressures.

The dynamics of labour demand in the current episode are also considerably different to those of the past because the phases in a mining boom are more distinct than in an agricultural-based boom (see, for example, Corden and Neary (1982) and Plumb et al (2013)). In particular, in the construction phase, when capacity is expanded, there is both strong investment and demand for labour. However, once the expansion in capacity is complete and mines are operational, it is likely that labour needs will be relatively less.[31]

The responsiveness of labour supply to a booming sector depends on factors such as the extent of specialised skills required and its geographic location. D'Arcy et al (2012) reported that as at February 2012 nearly one-fifth of employed people had begun their job in the previous year, suggesting that there was considerable labour mobility. They also note that around half of all job movements at this time involved job changes between industries. Alternatively, some studies have found the responsiveness of labour to intra-industry wage differentials to be limited (e.g. Kilpatrick and Felmingham (1996a, 1996b), using data from 1989 and 1992). D'Arcy et al (2012) argue that in the current episode the non-financial costs of relocating and skills mismatch have to some extent restricted geographic mobility; Bhattacharyya and Williamson (2011) examine differences in unemployment rates across the states, and find that since the 1920s boom labour has been geographically immobile.[32]


For summaries of the exchange rate regimes in the 1970s, see Blundell-Wignall, Fahrer and Heath (1993) and Debelle and Plumb (2006). [14]

Schedvin (1992) characterised the degree of difficulty in conducting monetary policy in the 1970s to that in the 1960s as a ‘quantum leap’. [15]

The capital controls included an embargo on short-term capital inflows and a Variable Deposit Requirement, which required a fraction of overseas borrowings to be deposited at the RBA in an interest-free account, effectively serving as a tax (Treasury 1999). [16]

For example, a policy statement from the RBA's 1969–70 annual report (RBA 1970, p 48): ‘It was announced that, in the course of its contacts with savings banks, trading banks, and life offices, the Bank had asked these institutions to maintain, and in the case of savings banks, to the extent practicable to increase, the volume of their housing loans in coming months. This action followed the recent falling-off in dwelling approvals by local government authorities and indications of a decline in housing commencements’. See also Battellino and McMillan (1989). [17]

The advantage of this approach to estimating the fiscal impact is that it is transparent, requires few assumptions and is easily compared from year-to-year. The disadvantage is that it can be influenced by small changes in the timing of expenditure, although this is unlikely to have had a material effect on the timing of economic activity. [18]

Achieving a balanced budget was considered to be ‘the hallmark of good government’ (Australia, House of Representatives 1959). [19]

In 1901, customs and receipts from public enterprises accounted for nearly all Commonwealth revenue (Vamplew 1987). [20]

A revaluation was opposed by exporters of agricultural products other than wool, while graziers denounced the tax on wool incomes as a ‘class tax’ (Waterman 1972). [21]

The so-called ‘horror’ budget of 1951 explicitly recognised for the first time in Australian history that the budget should be used for countercyclical policy purposes (Waterman 1972). Most of the initiatives were focused on the revenue side and included increases in income, sales and company tax (Schedvin 1970). [22]

Government gross debt is the total liability the government owes to creditors. This is comprised largely of outstanding Commonwealth Government securities. Net debt is gross debt less the debt owed to the government. For a summary of the historical trends of public debt in Australia, see Di Marco, Pirie and Au-Yeung (2009). [23]

In the 1890s and in the aftermath of the Roaring Twenties, rescheduling debt repayments was limited due to institutional and cultural reasons (McLean 2013). A statutory reduction in the interest on public debt for domestic investors was implemented as part of the Premiers' Plan in the early 1930s (Pincus 1985). [24]

The 25 per cent cut to tariffs in 1973 is not particularly evident in Figure 16 as some items were exempt (where the duties were primarily to raise revenue, rather than protection), and tariffs were quickly unwound for some industries (e.g. automobiles) (Lloyd 2008). [25]

The ‘Dutch disease’ literature (stemming from Gregory (1976)) emphasises that other tradeable firms, particularly manufacturing, may not be able to increase production quickly when a terms of trade boom is unwound. In addition, an increase in commodity prices may have a permanent component, in which case some structural change is inevitable. [26]

The sharp increase in tariffs following the Korean War episode reflects an increase in dutiable imports, rather than tariff rates (Lloyd 2008). [27]

For overviews see Withers (1987) and Bhattacharyya and Hatton (2011). [28]

Pope (1982) finds similar results over 1900–1930. Some studies considering later periods, however, argue that the institutional structure had little impact on the responsiveness of real wages (Hughes 1973; Withers, Pitman and Wittingham 1986). [29]

In January 1931, the Arbitration Court implemented a sizeable nominal wage cut, which assisted the recovery in economic activity (Valentine 1987); in 1953, automatic wage adjustments were abolished, which combined with other policies helped slow inflation (Boehm 1971). [30]

In both types of boom, however, there may be considerable increases in employment in industries servicing the booming sector (for example, the business services or construction sectors). [31]

Watson (2011) (using Household Income and Labour Dynamics in Australia (HILDA) data) found that in both 2002 and 2008 only 6 per cent of people changing jobs moved 500 kilometres or more. Debelle and Vickery (1999) also found little role for wage differentials in influencing interstate labour mobility. [32]