Date created:27/8/04 Last modified:27/8/04 Maintained by: John Quiggin John Quiggin
5 June 2004.
The debate over the proposed Free Trade Agreement with the United States took an interesting turn last week, with a report from the Centre for International Economics, predicting that adoption of the FTA would raise Australian Gross National Product by $5.6 billion per year, or around 0.7 per cent of GNP.
An earlier CIE study, predicting benefits of $4 billion per year, largely on the basis of the assumption that the FTA would include improved access to US markets for sugar, beef and dairy products. Since our negotiators got nowhere on sugar, and accepted bad deals on beef and dairy products, we might expect that benefits should have declined, not risen.
A closer look at the report clarifies the issue. The estimated benefits from the 'free trade' part of the agreement have indeed declined. The estimated gain from changes in merchandise trade is around 0.1 per cent of GNP, about what the economy would gain from the ordinary processes of economic growth every ten days or so. And this depends on optimistic technical assumptions about elasticities of substitution. With different assumptions, the result could be smaller, or perhaps even negative.
The paradox is resolved by the observation that the new report includes additional, highly speculative sources of benefit that were not included in the original analysis. In the new analysis, most of the benefit is derived from these speculative assumptions.
The first assumption relates to so-called 'dynamic productivity gains', supposed to arise from trade liberalization. Belief in these dynamic gains is something of an article of faith for Australian supporters of microeconomic reform. They are undeterred by the fact that their position is inconsistent with mainstream economic theory, and unsupported by empirical evidence.
More interesting are the gains attributed to the supposed reduction in the risk premium for equity arising from capital market liberalisation. The risk premium is the difference between the rate of interest on government bonds and the rate of return expected by investors in equity and is between 6 and 8 percentage points on most estimates.
As the CIE modelling results indicate, the equity premium is far more important than is often appreciated. The CIE modelling implies that a reduction of 5 basis points (0.05 percentage points) in the equity premium would induce a permanent increase in GNP of around 0.5 percentage points. A linear extrapolation (not entirely satisfactory, but a reasonable first approximation), shows that eliminating the equity premium altogether would raise GNP by between 60 and 80 per cent.
Although the equity premium reflects the riskiness of investing in equity it is far larger than would be expected on the basis of the efficient markets hypothesis (which is implicit in the CIE analysis). Estimates based on the efficient markets hypothesis suggest that the equity premium ought to be no more than 1 percentage point. Most of the premium is due to some combination of market failures, investor irrationality and distortions arising from taxes and regulations.
Among other things, the size of the equity premium explains the fact that, contrary to what might be expected, mixed economies (with a combination of public and private enterprise) have generally outperformed those with a minimal public sector. In areas such as infrastructure investment, the lower cost of capital to governments, which can spread risk through the tax system, frequently offsets any improvements in operating efficiency associated with private ownership. As a result, privatisation often worsens the financial position of governments, rather than improving it as predicted by analyses based on the efficient markets hypothesis.
Thus, the CIE is right to focus on the equity premium. The difficulty is in the assumption that capital market liberalisation will reduce the equity premium and will have no offsetting adverse effects. The proposed changes are tiny by comparison with the floating of the dollar and the associated removal of exchange controls over the 1970s and 1980s, not to mention the associated domestic liberalisation. Yet there is no convincing evidence that these changes had any net effect on the risk premium for equity.
Then there's the question of offsetting effects. The most important changes arising from the FTA relax restrictions on foreign takeovers. The CIE analysis is based on the assumption that such restrictions are necessarily harmful. Yet there's ample evidence at every level to contradict this. Takeovers reduce economic welfare as often as they increase it.
The aggregate economic effects of the trade liberalisation proposed in the FTA will be negligible. The effects of capital market liberalisation will be much greater, but are as likely to be harmful as beneficial.
John Quiggin is an Australian Research Council Federation Fellow in Economics and Political Science at the University of Queensland.
Read more articles from John Quiggin's home page
Go to John Quiggin's Weblog