Productivity Commission’s myopic failure on industry assistance

The government’s announcement that the age of entitlement is over may be a poisoned chalice for the Productivity Commission. Daniel Munoz/AAP

Every year the Productivity Commission (PC) produces an annual report card on Australian government assistance to industry. And every year it singles out the “usual suspects” for attention, particularly the manufacturing industry. This year’s Trade and Assistance Review 2012-13 is no different except that the government itself has announced an “end to the age of entitlement”.

Does this signal the end of debate about the government’s role in the economy, or the beginning of a new one? Is the PC, like the apocryphal military commander, simply fighting the last war? While governments around the world may be jettisoning traditional policy approaches, they increasingly recognise the importance of promoting economic complexity, innovation and competitiveness in global markets and value chains.

By contrast, the PC estimates the value of industry assistance, such as tariffs and tax concessions, with reference only to the cost to consumers and other industries that use the output of assisted industries as inputs in their own production. The analysis reflects the neoclassical economic orthodoxy in which economic welfare is best secured through the unimpeded market mechanism, and the legitimate role of government is confined to a very limited set of “market failures”.

This approach has dominated policy thinking in Australia not just because it channels the logic of a mathematical theorem, but also because it has self-evident intuitive appeal. However, both the logic and appeal of such an approach are increasingly under challenge by developments in the real world of production and trade. As a result, there are at least three areas of contention in the PC Review.

Comparative advantage

First, the review starts from the principle of “comparative advantage between nations”, which asserts that two nations will be better off if they specialise in producing and trading those commodities in which they have a relative cost advantage, even if one of them has an absolute cost advantage in all commodities. This advantage is governed by the relative proportions of labour, capital and land in each nation.

However, the assumptions required for this principle to operate are wholly unrealistic. They include no international movement of capital and labour; no scale economies in production; identical commodities produced and traded in each country; and identical technologies. Even Paul Samuelson, who established the mathematically rigorous form of comparative advantage 65 years ago, recognised its inapplicability to the real world of trade.

This world is characterised by large-scale capital flows, uncertainty, increasing returns, product differentiation, oligopoly and constant technical change. Comparative advantage predicts inter-industry trade (nations specialise in making and trading either clothes or cars), but in reality intra-industry trade dominates (nations produce and trade both clothes and cars). For example, Germany is not only the leading exporter of advanced motor vehicles and machine tools but also “the 2nd textile and 4th greatest clothing exporting country in the world”.

The federal government has supported the car manufacturing industry but the PC Review didn’t answer what it should do in the future. Julian Smith/AAP

It is precisely because of these characteristics that advanced manufacturing and services require government support through R&D tax breaks, public procurement, technology transfer programs, network brokering and big investments in school, university and technical education and specialist infrastructure. Such support creates high skill, high productivity jobs, builds the science and technology base and enables technology spill-overs across the economy.

The PC Review finds that since 1970 the effective rate of assistance to manufacturing has fallen from 35% to 4%. Clearly not all past assistance was well conceived, but does this justify indifference to manufacturing decline? First world living standards will be difficult to maintain in a third world economic structure with a small number of high paid jobs in mining and finance and the rest in low productivity, low wage services such as aged care and retail.

Arbitrary definitions

The second problem with the PC approach is that the definition of “government assistance” is entirely arbitrary. The review’s estimates are initially said to cover “those measures that selectively benefit particular firms, industries or activities, and that can be quantified given practical constraints in measurement and data availability”. This would seem to capture any kind of preferential support for industry, either directly or indirectly.

However, the review restricts assistance to tariffs, a limited number of tax concessions such as those for R&D, and direct budgetary assistance such as funding to the CSIRO and car industry. It explicitly excludes:

“regulatory restrictions on competition… tax concessions for superannuation… government programs affecting a range of services industries, mainly relating to the provision of health, education, and community services… [and] resource access arrangements including to mining, forestry and fisheries”.

The effect is to portray manufacturing as a mendicant industry imposing high costs on other industries, especially services.

If just some of these excluded measures were to be included, a completely different picture of the scale and distribution of government industry assistance would emerge. For example, imagine evaluating “regulatory restrictions on competition” in the context of the International Monetary Fund finding that Australia has the most concentrated and profitable banking system in the world.

Or to take another example, compare the estimate of A$7.12 billion net manufacturing assistance with the A$32 billion allocated over 2013-14 to the superannuation industry in foregone tax revenue, comprising capital gains concessions and tax concessions on contributions. It has been calculated that these tax concessions are equivalent to annual expenditure on the age pension. Similarly, the 30% private health insurance rebate is a A$1.5 billion tax concession intended to “distort” consumer spending towards this industry.

There are other examples but the most significant relate to the resources sector, again excluded from the definition of government assistance. Taken together with the huge public investment in mining infrastructure, for which the states alone have been estimated to provide A$3 billion annually for the last six years, the A$4.5 billion annual expenditure on the diesel fuel tax rebate and other tax subsidies ensures that mining qualifies for the “mendicant” status of manufacturing, but apparently without the stigma.

The mining industry has benefited from government support without being typecast as an industry that needs to be propped up. Newcrest Mining/AAP

The campaign by mining companies against a resources rent tax, which might have laid the foundations for a Norwegian-style sovereign wealth fund, has now been compounded by claims of legal but costly tax avoidance. It was reported last week that Australia’s largest coalminer “paid almost zero tax over the past three years, despite income of A$15 billion, as it radically reduced its tax exposure by taking large, unnecessarily expensive loans from its associates overseas”.

Costs and benefits

Finally, the modelling in the PC Review has the effect of maximising the costs of government assistance and minimising its benefits. For example, the Review does not attempt to estimate the increase in R&D and productivity arising from business innovation support measures. Similarly, government spending on the CSIRO is, in the accounting framework of the PC, a net cost to taxpayers and industry. These measures are simply assumed to be ineffective or just a “deadweight” cost.

Whatever the merits or otherwise of assistance to the local car industry, the cost of this assistance is not balanced in the review against the income, company and GST tax revenues arising from the output of the industry. As the local car makers had an income of A$19.6 billion in 2011-12, receipts from these sources would have more than offset the value of assistance estimated at A$1 billion for 2012-13.

To discount income received by the government from the car industry but only include the cost of assistance, the PC assumes that labour and capital in the car industry would be fully employed in the absence of assistance. What we know, however, is that when the local makers withdraw around 2016-17, capital will be either repatriated to the parent company or invested in car production in, say, Thailand.

Similarly, research on labour redundancy from manufacturing plant closures suggests that about one-third leave the labour force, one-third get a lower quality job and only one-third get an equivalent or better job. This finding has been confirmed by more recent studies of car plant closures.

Should the Australian government have increased its co-investment in car assembly manufacture to retain a local presence, should it have redeployed support to car components suppliers or to other industries with greater innovation, growth and export potential, or should it have wound back support to all industries, irrespective of status and significance? Unfortunately, a reader of the PC Review will be none the wiser.