The rise and fall of automotive manufacturing in Australia

The auto industry has been an important feature of the Australian economy throughout the postwar era, with active industrial policy shaping that history.

The first global manufacturers commenced Australian production in the 1920s, reliant on the import of ‘kits’ of components which received only final assembly in Australia. ^{Footnote 7} After the Second World War, a Labor government implemented an ambitious industrial strategy to develop automotive manufacturing as part of a broader plan for economic diversification (Reference EmmeryEmmery, 1999). The plan centred on an effort to develop an ‘Australian car’: designed and manufactured in Australia, embodying Australian-made components. General Motors, through its iconic Australian subsidiary Holden, was the first to produce a uniquely Australian vehicle in 1948 (Reference Kennedy, Kennedy and DavisKennedy et al., 2007). Other global original equipment manufacturers (OEMs) followed suit, establishing comprehensive and vertically integrated Australian operations. The overall strategy was supported with active policy levers, including import-licensing, local content rules and tax and duty concessions (Reference PursellPursell, 2001).

Import licensing was eliminated in the 1960s, replaced with quotas and tariffs on imported vehicles and parts. Tariff rebates were paid to OEMs which met strong domestic value-added targets (eventually set at 85% of the value of vehicles assembled in Australia); this provided further stimulus to domestic components production (Reference PursellPursell, 2001). High tariffs (above 50% on finished vehicles) and quotas limited imports to around 20% of domestic sales. In the 1970s, the strategy was amended to encourage exports: companies received incremental relief from the domestic content threshold, reflecting the value-added in exported components and vehicles (Reference PursellPursell, 2001). By the 1970s, the industry produced close to one-half million vehicles annually (Figure 1). The country assembled almost as many vehicles (for both domestic and export use) as Australians purchased, making Australia largely self-sufficient in vehicle production on a net basis. ^{Footnote 8} Automotive manufacturing was the country’s largest manufacturing sector, employing over 100,000 workers (Australian Bureau of Statistics (ABS), 1975, Chapter 21); it was concentrated in two major clusters (in Victoria and South Australia), but with a supply chain that reached into all states.

Figure 1. Australian vehicle assembly.

Source: Ward’s Auto (2015).

Downsides of this strategy included a highly fragmented industry structure: 5 different OEMs produced up to 15 different models, but with small production runs (Reference PursellPursell, 2001). Combined with high tariff protection, this meant that vehicles in Australia were considerably more expensive than in Europe or North America (Automotive Review Secretariat, 2008; Reference PursellPursell, 2001). On the other hand, the focus on fostering unique Australian models spurred significant automotive design, engineering and research capacities within Australia. As a result, the Australian industry demonstrated a higher proportionate innovation effort than other countries which also lacked home-grown OEMs (Reference StanfordStanford, 2017).

As part of a broader vision of market-oriented reform, a subsequent Labor government in the mid-1980s dramatically changed its approach to macroeconomic, labour market, industry and trade policies (Reference PuseyPusey, 2003). This shift in policy direction constituted a controversial effort by Labor to implement core neoliberal policies such as anti-inflation measures, wage restraint, financial liberalisation, tariff reduction and deregulation, but conditioned by a unique political bargain that included a range of other measures intended to cushion the impact on working people. From the beginning, this policy direction embodied an uneasy coexistence of market and planning, as the government grappled with the competing objectives of enhancing business confidence and maintaining support from its core constituency. ^{Footnote 9}

A complex and contested approach to industry policy was a key component of this grand compromise, in which the auto industry obviously played a central role. In 1984, the government implemented a new Motor Industry Development Plan, called the ‘Button Plan’, after the Industry Minister of the day (Productivity Commission, 2014: 30). Its main features were the elimination of quantitative restrictions on imports, replaced by tariff quotas (which in turn were also phased out by 1992), and steep reductions in vehicle tariffs, which fell immediately from 57.5% to 45%, and then by 2.5% per year for the next 12years, reaching 15% by 2000 – by then in line with typical most favoured nation (MFN) tariffs levied by most other industrial countries. To support the domestic industry through this change, the government implemented direct subsidies and incentives to domestic production, supposedly on a temporary basis (Reference PursellPursell, 2001).

The industry consolidated down to four OEMs, producing fewer models in just four assembly plants. Imports surged in the wake of unilateral trade liberalisation, providing Australian consumers with more choice and lower prices. The domestic industry’s share of domestic sales fell steadily. Exports increased, but not nearly enough to offset the decline in domestic sales, so overall production fell about 20% and a large and chronic automotive trade deficit emerged. ^{Footnote 10} The decline in domestic production was all the more substantial when measured in relative terms – that is, as a proportion of the domestic vehicle marketplace (which continued to expand on the basis of a growing population and rising living standards). By the early 1990s, Australia was assembling barely half as many vehicles as its consumers purchased (Figure 2) – a sharp decline from the net self-sufficiency that was attained in the 1970s.

Figure 2. Vehicle production-to-sales ratio, Australia.

Source: Author’s calculations from Ward’s Auto (2015).

Later, in the 1990s and early 2000s, vehicle assembly stabilised (at around 350,000 units per year), and employment in automotive manufacturing expanded slightly. This was in part due to the lower value of the Australian dollar at the time, which enhanced the relative cost competitiveness of Australian production, encouraging OEMs to produce more in Australia (including for export sales). The Australian dollar depreciated by one-third between 1989 and 2001 (Figure 3), reaching an all-time low of just over 50 cents US in 2001. ^{Footnote 11} Modest growth in exports also supported production levels: by the mid-2000s, exports accounted for 40% of all vehicle output. Some of this export growth reflected OEMs’ allocation to Australian assembly plants of models with broader international sales potential, rather than traditional iconic Australia-specific models (Automotive Review Secretariat, 2008). However, the industry’s share of domestic sales continued to shrink, and total production lagged far behind ongoing growth in domestic sales, so the production-to-sales ratio continued to fall steadily.

Figure 3. Australian dollar exchange rate.

Source: Reserve Bank of Australia (2017).

After 2000, the Australian government introduced another automotive support programme, based on duty remission credits, again supposedly on a temporary basis (Automotive Review Secretariat, 2008). But Australian production began to decline sharply after 2004, reinforced by a dramatic appreciation in the Australian currency (Figure 3). The Australian dollar doubled in value between 2001 and 2012, eventually exceeding parity with the US dollar. This badly damaged the cost competitiveness of Australian auto production. However, the government’s response was to embark on further unilateral trade liberalisation: in 2005, the tariff on vehicle imports was cut by another 5 percentage points to 10% (Automotive Review Secretariat, 2008), and Australia began signing free trade agreements (FTAs) with auto-exporting jurisdictions, starting with the US and Thailand in 2005. ^{Footnote 12} Imports reached close to 90% penetration of the domestic market, but automotive exports actually declined (reflecting Australia’s high costs, and inconsistent export marketing of Australian-made vehicles by the OEMs).

Total domestic assembly fell by half between 2004 and 2010. ^{Footnote 13} Mitsubishi shuttered its Australian engine and assembly plants in 2008 (Reference Beer and EvansBeer and Evans, 2010). The 2008–2009 global financial crisis (GFC) posed more challenges. A ‘cash for clunkers’ vehicle purchase incentive helped maintain domestic demand, and special loans were offered to GM-Holden in response to its liquidity crisis and bankruptcy restructuring. But even after the global economy exited from recession and financial conditions stabilised, Australian auto production continued to decline. Ford was the next assembler to announce its departure in May 2013, followed by GM-Holden later that year and Toyota in February 2014. At the time of the latter closures, the federal government was led by conservative Prime Minister Tony Abbott, who stated bluntly that ‘the government’s role is not to prop up private business’ (Reference TaylorTaylor, 2013). Traditional bipartisan support for active automotive industry policy had evaporated, hastening the decisions by GM and Toyota to leave (Senate of Australia, Economics References Committee, 2015).

Australia retains a modest truck and specialty vehicle manufacturing industry; some automotive components manufacture will also continue, based on sales to the domestic aftermarket and OEM assembly operations in other countries (Federal Chamber of Automotive Industries (FCAI), 2015; Senate of Australia, Economics References Committee, 2015). Australia’s strong automotive research and engineering capabilities (the legacy of the ‘Australian car’ strategy) may support some continued work in that sphere for the global supply chains of OEMs. For example, Ford has announced plans to maintain significant Australian engineering facilities. Policy-makers continue to explore options for preserving and nurturing remnants of the industry (Reference Clibborn, Lansbury and WrightClibborn et al., 2016; FCAI, 2015; Senate of Australia, Economics References Committee, 2015). Nevertheless, with the cessation of large-scale passenger vehicle assembly, Australia’s status as a major automotive producer is ending.

Understanding Australia’s automotive policy surrender

The disempowerment and ultimate failure of Australian automotive industrial policy over the past 30years must be understood in the context of parallel developments in the national political economy during that period — together, of course, with marked changes in the contours of global automotive competition. Beginning in the early 1980s, a bipartisan consensus emerged that the previous state-supported development strategy, based on import-substituting industrialisation and a corporatist approach to income distribution and industrial relations, would be disassembled in favour of a more deregulated, market-driven approach (Reference KellyKelly, 2008; Reference PuseyPusey, 2003). The first major steps in that epochal shift were taken under the leadership of a Labor government which initially attempted to combine neoliberal trade, deregulation and financial policies with continued structures of tripartism and an expansion of some social protections. Supporters of the approach hoped that by trading off traditional labour movement demands as part of the compromise, support could be won for continued interventions in industrial policy and other dimensions of economic and social planning such as medical and retirement insurance. The Scandinavian model of active tripartite industrial planning was invoked (as, for example, in the ‘Australia Reconstructed’ exercise: see Australian Council of Trade Unions/Trade Development Council Mission to Western Europe (ACTU/TDC), 1987).

This compromise was contested, and ultimately unsustainable. Neoliberal approaches came to dominate the policy direction: visible in the pursuit of trade liberalisation, ratification of financialisation in the domestic economy, and the replacement of targeted industrial policy with a much more diffuse vision of innovation and upskilling, such as finds current representation in the National Innovation and Science Agenda (Australia Department of the Prime Minister and Cabinet, 2015). Over the years, Coalition governments extended the market-oriented aspects of the compromise such as trade liberalisation, while diluting or rolling back its more interventionist dimensions. Initial bipartisan support for the overall neoliberal framework of the plan helped to insulate its main features from subsequent revision. The gradual decline of automotive manufacturing, clearly attributable to features of that overall policy framework, was not fundamentally challenged by the Labor party and only by a dissident minority within its union base (Reference Ewer, Hampson and LloydEwer et al., 1991).

An incremental neoliberal ‘ratchet effect’ is thus clearly visible in Australian automotive industry policy after the Button Plan. Liberalised trade policy continued to advance under both Labor and Coalition governments. A misplaced expectation that unilateral, bilateral and multilateral liberalisation arrangements would result in a stronger, more rational automotive industry strongly influenced automotive policy reform beginning in the 1980s. The postwar development strategy of meeting domestic demand was initially effective, and reinforced the development of domestic research and engineering capabilities. Later, however, as consumer preferences diversified, it became impossible to sustain viable domestic production on the basis of domestic demand alone. Production runs were fragmented, productivity and costs suffered and more consumers turned to imports.

Thus, automotive industry policy later tried to transition towards a greater emphasis on exports. That transition was made more difficult by the Australian industry’s concentration on unique, relatively large vehicles with limited appeal in foreign markets. In retrospect, policy needed to more powerfully stimulate reorientation of domestic production toward export sales. Given the strong economies of scale inherent in this industry, that would have required winning global or at least regional production mandates from OEMs for export-oriented models for their Australian plants. ^{Footnote 14}

Policy-makers correctly identified the importance of nurturing a viable export market for Australian-made vehicles. But their main strategy for accomplishing that goal – namely, unilateral and multilateral trade liberalisation – failed utterly in stimulating automotive exports. Figure 4 indicates that the decline in automotive tariffs was closely associated with the continuing decline of domestic production. ^{Footnote 15} Trade negotiators argued that dismantling tariff barriers would, of itself, promote a mutual reorientation of production and the growth of exports. But, while automotive imports grew, reductions in foreign tariffs did not cause the expected automatic growth in exports. The approach underestimated the fluidity of international investment location decisions. With disincentives for imports mostly eliminated, global OEMs could simply service Australia’s lucrative vehicle market entirely from offshore production.

Figure 4. Australian vehicle tariff and production per capita.

Source: Automotive Review Secretariat (2008); Ward’s Auto (2015); ABS (2017).

After several bilateral FTAs were signed, beginning in the mid-2000s, Australia’s automotive exports went into outright decline: the generalised contraction of the industry was not prevented or even slowed by ‘better access’ to foreign markets. As summarised in Table 1, none of Australia’s FTAs produced any measureable increase in offshore demand for Australian-made vehicles – and that demand was small (for the US) to non-existent (for the Asian FTA partners) to begin with. Trade liberalisation thus contributed to an erosion of Australian exports, not an expansion. While the auto workers’ union continued to campaign against these deals, Labor governments accepted them in practice, cementing the liberalised trade policy context that contributed to the decline of automotive investment and production in Australia.

Table 1. Australia’s automotive trade under Free Trade Agreements (2015, $million).

Source: Author’s calculations from Australia Department of Foreign Affairs and Trade (2015). Includes sectors 781–784.

FTA: free trade agreement.

A core assumption of neoliberal comparative advantage trade theory is the notion that every participant in international exchange will automatically attain a position of beneficial specialisation by following its ‘comparative advantage’. This approach assumes away the general challenge of competitiveness: industries might decline in the wake of liberalisation, but idled resources should automatically be absorbed in other, relatively more productive industries. The emphasis on transitions and adjustment, and the corresponding faith that there is a sustainable and beneficial industrial structure to adjust toward, is a hallmark of neoliberal trade policy (see, for example, Staff of the World Bank, International Monetary Fund, and World Trade Organization, 2017), and this theme was important in the evolution of industry policy in the Australian automotive sector.

Indeed, every episode of trade liberalisation, from the Button Plan through to the bilateral FTAs signed after 2005, was accompanied by the offering of automotive investment and production subsidies to support the industry’s ‘adjustment’. The programmes were all initially described as temporary measures to assist the industry in adapting to international pressures (Australia Department of Industry, Innovation, and Science, 2016; Automotive Review Secretariat, 2008; Reference PursellPursell, 2001). Yet subsidies, whether for investment, research, or outright production, could not alter the fundamental asymmetry of Australia’s automotive trade and the reduced appeal of Australia as a location for incoming automotive foreign direct investment (FDI). There was no explicit idea of what the industry would be ‘transitioning’ to; hence, subsidies were repeatedly rolled over. Eventually, the payment of subsidies was not even tied to concrete production or export commitments, as had been the case under the original Button Plan (Reference StewartStewart, 1994). In practice, the subsidies simply deferred and prolonged the decline of the industry, rather than helping it find a truly viable equilibrium.

The liberalisation of foreign currency trading (first implemented by the Labor government in 1983) and the general financialisation of Australia’s economy also contributed to the collapse of industrial policy in the automotive sector (as in other segments of manufacturing). Currency misalignments, permitted and encouraged by financial liberalisation, greatly exacerbated the industry’s downturn beginning in the early 2000s. At the peak of the Australian resource boom in 2012, the Australian currency was trading 60% above its purchasing power parity benchmark. ^{Footnote 16} This made Australian relative production costs ‘seem’ 60% higher than nominal values (measured relative to domestic price indices) would suggest. The overvaluation did not reflect strong trade performance. In fact, Australia incurred large trade deficits and growing international debt throughout the resource expansion – driven in part by large inflows of real foreign direct investment, but even more by more volatile financial inflows seeking to profit from Australia’s uniquely high interest rates and the expected further appreciation of the currency. The Reserve Bank’s single-minded focus on inflation targeting (another core neoliberal macroeconomic precept), and corresponding refusal to intervene to moderate the dramatic swings in currency markets, ratified the currency’s extreme overvaluation. Far from acting to limit the dollar’s rise (and its negative impact on non-resource exports), the Reserve Bank actually celebrated the appreciation as a useful adjustment mechanism in supposedly preventing inflation (Reference StevensStevens, 2013).

As illustrated above, an undervalued exchange rate had temporarily supported Australian automotive output during the 1990s and early 2000s. The reverse occurred, however, once the resource boom gathered momentum and the exchange rate shot up. The problem of currency overvaluation, and failure of Australian policy-makers to address it, contributed to OEM judgments that Australia was an unsustainably expensive location for future production. The subsequent retreat of the currency after the commodity price downturn in 2014 significantly restored Australian cost competitiveness. But by then it was too late: the decisions to shut down Australia’s industry had already been made.

The strength of the resource boom also likely contributed to policy-makers’ complacency regarding the potential consequences of automotive downsizing. Unemployment fell to very low levels in the mid-2000s, thanks largely to strong employment growth in resource extraction and its supply chain, and broader macroeconomic conditions were strong. In that context, concerns over the impact of automotive closures were subdued. Policy-makers from the major political parties, and influential bodies like the Productivity Commission of the Australian Government (2014), tended to express strong faith in the automatic, beneficial nature of inter-sectoral economic adjustments. And with a resource boom in full swing, it was easier to believe that job losses in one sector negatively impacted by soaring net imports would be naturally offset by gains in more successful export-oriented sectors. Analysts forgot that extractive resource industries demonstrate strong cyclical patterns; the quick ending of the so-called commodities ‘super cycle’ after 2014 should not have been a surprise.

Australia’s impressive postwar achievement in building a viable and successful automotive sector was ultimately undone, as policy-makers (beginning in the 1980s) lost sight of both the rationale for, and the necessity of, active industrial policy in attracting and nurturing investment in strategic, globally mobile sectors. Evidence that the industry was locked in a downward trajectory was overlooked on the strength of undue faith that comparative advantage reallocation of resources would ultimately offset any sectoral losses and lead the national economy toward a more efficient equilibrium position. Policy-makers abandoned their previous conviction that industries like automotive manufacturing – once seen as an obvious and logical target for industrial policy – even ‘matter’ in a world of liberalised trade and global value chains. Hence, there was no particular reason to pay the sector any special attention.

Industrial policy activism in other automotive-producing countries

Meanwhile, governments elsewhere maintained a more active stance. The continuing effectiveness of industrial policy helps to explain the continuing viability of automotive manufacturing in other jurisdictions, even those facing challenges equally or more daunting than Australia’s. Of course, there has been considerable and painful automotive adjustment and restructuring in many countries, arising from the impacts of trade liberalisation (Reference Klier and RubinsteinKlier and Rubinstein, 2010; Reference StanfordStanford, 2010), the evolution of global value chains and their impact on supply relationships in the industry (Reference GereffiGereffi, 2014; Reference Newsome, Taylor and BairNewsome et al., 2015), and the generalised course of deindustrialisation in the advanced economies (Reference Doussard, Peck and TheodoreDoussard et al., 2009; Reference Rowthorn and RamaswamyRowthorn and Ramaswamy, 1999). But in other countries, despite tumultuous global industrial and financial conditions, governments (even conservative ones) retained a willingness to mobilise substantial resources and regulatory influence to support automotive production. In no other country did policy-makers effectively leave the fate of the industry solely to international market forces and the private decisions of global OEMs.

The major policy levers of automotive industrial policy

Despite the dominance of neoliberal ideas in most of the world, activist industrial policy has not generally receded in importance and effect (Reference RodrikRodrik, 2008; Reference Stiglitz and YifuStiglitz and Yifu, 2013). The continued relevance and effect of industrial policy remains especially visible in the automotive industry. Most governments continue to play an active role in attracting automotive investment to their respective jurisdictions. Interventions in the sector have been especially strong since the GFC (Organisation for Economic Cooperation and Development (OECD), 2009). As the financial crisis took hold, governments provided subsidies and incentives to boost vehicle demand (shocked by the impact of the crisis on consumer confidence), emergency financial aid was provided to distressed OEMs and suppliers, tariffs and other trade restrictions were imposed in several countries, and support was expanded for industry investments and research (Reference StanfordStanford, 2010). Even after the crisis abated, governments in most regions remained active in their efforts to attract and support domestic automotive manufacturing.

Surveying continuing automotive policy efforts, several broad themes are visible. Direct government incentives to defray capital costs are a common tool for influencing investment at critical decision points (Center for Automotive Research, 2015; Reference Molot, Eden and DobsonMolot, 2005). ^{Footnote 20} Reference Yates and LewchukYates and Lewchuk (2016) argue that investment subsidies are now seen by OEM executives as a normal, necessary feature of the policy environment in any automotive jurisdiction.

Automotive investment is responsive to international and especially intra-regional (Reference Klier and RubinsteinKlier and Rubinstein, 2010) differences in production costs, as evidenced by the migration of investment toward lower-cost regions of Europe and North America. Governments join this race by invoking measures to reduce relative production costs in their respective jurisdictions. Such measures include policies to suppress wages and other labour costs, measures to enhance productivity (like publicly funded training, or subsidies for automation), public provision of tailored infrastructure and transportation facilities (to reduce logistics expenses), lower corporate and payroll taxes and in some locations, direct subsidies to production.

Despite the constraints of FTAs, which discipline some (but not all) of the traditional tools of industrial policy, ^{Footnote 21} governments also continue to use active trade policy interventions to support their domestic industries. The use of non-tariff barriers to limit imports, and various institutional and fiscal measures to manage trade, are common; both Japan and Korea, for example, effectively utilise such measures to minimise import penetration. Indirect supports for exports (including preferential financing, publicly financed export infrastructure and more) are also common. Several countries, including Russia, China and Brazil, have imposed targeted tariff protections against automotive products (both vehicles and components), notwithstanding trade disciplines – often invoking safeguard or anti-dumping loopholes. Indeed, the safeguard clauses which exist in most trade agreements (permitting governments to limit imports when demonstrated harm is occurring to domestic production) is another avenue through which the Australian government could have moderated the impact of unbalanced import flows on domestic production.

Active management of exchange rates has been invoked in many locations to protect the competitiveness of domestic production and discourage imports. Once again, this widespread practice stands in sharp contrast to Australia’s ‘hands off’ approach. Governments and central banks in several key auto jurisdictions, including Japan, Korea, China and Brazil, regularly use market interventions or direct regulations to suppress exchange rates. In other countries, the same goal has been achieved through the use of unconventional monetary policy such as quantitative easing in the US, the UK, Japan and the Eurozone.

The increasing technology-intensity of automotive manufacturing, and the accelerating pace of both product and process innovation, have made government innovation policy a potent and commonly-used lever in attracting and nurturing investments. Government investment subsidies in most jurisdictions are now tailored to emphasise innovation inputs to domestic production, subsidising R&D activity at a higher rate than other forms of automotive investment or providing direct government funding for key innovation missions (such as the generous support for fuel efficiency and battery technology offered by the US Department of Energy). Broader innovation policies – such as the tax treatment of R&D investments, partnerships between industry and public research institutions, and the supply of well-trained innovation professionals – have special relevance in the auto industry, given its reliance on research and engineering inputs. The idea is that these innovation supports will entice OEMs and top-tier suppliers to locate innovation activity (and the high-quality jobs that come with it) in the domestic region. Moreover, it is hoped that anchoring innovation activity may spillover into localised assembly and production opportunities, as well. ^{Footnote 22}

Direct public ownership of productive firms fell out of favour in most countries during the neoliberal era, but there remains a surprising degree of public intervention in ownership structures in the auto industry. In emerging economies, total or partial public ownership is still common in automotive manufacturing. China’s rapid automotive development has been spurred in part by government rules requiring foreign OEMs, anxious to establish production in China, to make joint-venture investments with Chinese partners. This strategy facilitates the transfer of technology and management expertise, hence fostering all-round domestic capabilities. Direct public ownership is also present in Europe (such as France’s partial ownership of Renault, and the German state of Saxony’s continued equity share in Volkswagen); governments wield equity shares to push OEMs to retain core operations in their home jurisdictions. Public ownership was even used temporarily in the US during the GFC to support the restructuring of General Motors and Chrysler. Other forms of public ownership – including sovereign wealth fund and state development bank holdings in OEMs – also influence location decisions.

In sum, the global automotive policy toolbox remains filled with a variety of important and powerful instruments. While the evolution of technology and globalisation has certainly altered, and in some ways constrained, the application of these tools, it is obvious that investment and industrial development in this strategic sector are hardly being left to the unconstrained discretion of private firms and market signals. To the contrary, the engagement of governments in actively shaping the growth and location of the industry is, if anything, stronger than ever. Australia’s decision to abandon active industrial policy in this sector, and tolerate the disappearance of mass vehicle assembly altogether, cannot therefore be ascribed to some universal trend in policy approach, nor to some set of irresistible economic constraints. To the contrary, Australia’s trajectory has been unique and extreme.