Senior economic policy makers, economic analysts, academics and commentators have been concerned about the daunting challenge of structural and budget adjustment facing Australia due to the decline in mineral prices from the levels reached in the boom period of 2003-04 to 2011-12.
The crux of the matter lies in the distinction between production and income. Edwards' focus is on the production side, on the 'contribution to GDP from the mining boom to 2011-12'.
Almost all other commentators and modellers look beyond GDP, to the real income of Australians. Essentially, over and above GDP, during the boom a large boost was given to the income (or purchasing power) of Australians. Now that mineral prices have fallen, this will not recur, and it presents an adjustment challenge.
Economists have two standard methods to estimate the boost to economic income delivered by a boom in export prices:
- The Australian Bureau of Statistics (ABS) way, using the System of National Accounts. This method yields a benefit equal to 14% of the GDP of 2011-12.
- The modelling way, using quantitative representations of the economy: this yields a benefit equal to 13% of the GDP of the year 2012-13 (argued in recent RBA Research Paper authored by Peter Downes et al. Although I discussed the paper in my Minerals Council critique of Beyond the Boom, it is not mentioned in Edwards' response as it was published after Edwards' book).
The ABS's national accounts tell us that over the decade to 2011-12, production as measured by real GDP grew more slowly than did the real purchasing power of GDP, as measured by national income. There was a 14% gain in income over and above the boost to GDP itself. This difference is usually called the 'trading gain', because it measures the benefit of a rise in the international terms of trade.
Before proceeding, here is a clarification to reassure Edwards: everybody has reported the size of the benefit of the boom as a ratio, equal to the benefit (however conceived) divided by some understandable denominator. Everyone, including Edwards, uses the GDP (or GNP) of the end year (2011-12 or 2012-13) as the denominator. Nobody who has contributed to the debate thinks that the ABS has estimated that the boom caused a 14% rise in GDP. Everybody knows that, because it takes GDP as given, the ABS method attributes no boost to GDP on account of the boom.
The Downes et al paper does, however, estimate that the boom did indeed boost the GDP of 2012-13 by 6%. National income was boosted by 13%, with the extra 7% mainly due to the trading gain.
Edwards steadfastly refuses to accept the standard interpretation of the trading gain as showing the rise in income, over and above the rise in GDP. His ground is that, for the 'trading gain' to become actual rather than merely hypothetical, the income boost must be spent on additional imports.
This assertion suggests a failure to understand national accounting. Edwards does not seem to appreciate that the ABS can estimate the size of the gain in income without having to show what uses were made of the income gain. Income can be used for saving, as well as for spending. Savings rose dramatically. Edwards amply documented the rise in saving, but he did not seem to understand that it refutes his argument that the ABS's 14% trading gain in income was merely hypothetical until it manifested as imports.
Moreover, Edwards has misinterpreted the facts about imports: 'Households, anyway, do not seem to have responded gladly to flat import prices. They have not very vigorously increased their purchases of imports compared with earlier trends.' But import prices were not flat. Instead, due to the rise in the exchange rate, they fell greatly, relative to Australian prices generally. That is what a sharp rise in the exchange rates does.
As to quantities, imports volumes rose about 70% faster than GDP in the boom (and the composition of imports did not change much). The additional imports were financed out of the higher national income and were stimulated by the large fall in the real price of imports.
The Downes et al paper confirms that the mineral boom caused a large rise in import volumes. The methodology used, a fully specified model of the economy, is far superior to Edwards' method, which estimated 'counterfactual' import volumes by simply comparing the decade before 2002-03 with the decade following.
Before going further, it is necessary to state that Edwards and all other commentators must realise that some benefits of the boom will last beyond its end. These lasting advantages are not captured in the production or income numbers discussed so far. They consist in huge rise in the capital stock, and a significant increase in private wealth, both of which will ease the adjustment to lower terms of trade.
There is a revealing lapse in Edwards' use of national accounts nomenclature. Although Edwards repeatedly states that his 3% is the mining boom's contribution to GDP, strictly his is an estimate of the mining boom's contribution to GNP, not GDP.
He first calculates the contribution as 6% of the GDP of 2011-12, but then discounts that by half because of foreign shareholdings in mining. But this is how the national accounts go from GDP to GNP, by deducting from GDP what is owned by foreigners on account of their productive and financial services (net of what is owing in the other direction). Thus, the ABS estimate of the trading gain takes account of foreign ownership of GDP and therefore, contrary to what Edwards asserts, should not be further discounted for foreign ownership.
Finally, there is something contingent in the way Edwards arrived at his 6% boost to GDP or 3% boost to GNP. If the rise of the exchange rate had been less, his estimate would have been more. Consequently, regardless of the cause of the rise in the exchange rate — and most observers say the mineral boom was a major, if not the prime factor — Edwards' method understates the national gain.
This is because the rise in the exchange rate reduced his estimate of the benefits of the boom, but did not reduce the benefits to Australia as a whole. Edwards measured the AUD-denominated benefits that flowed to the Australian owners of mining shares to state governments as royalties and to the ATO as company tax payments from miners. The rise in the exchange rate reduced these AUD flows (regardless of the causes of the rise in the exchange rate). But the rise in the exchange rate did not reduce the national gain. It merely redistributed some of it away from the Australian owners of mining shares and towards other Australians, including those purchasing imports which were now much cheaper in AUD terms.
Edwards needs to understand that (to the first order of approximation) a rise or fall in the exchange rate redistributes income and wealth, rather than increases or decreases national income and wealth.
Unfortunately, in a table I misreported the dates for which I estimated the effect of the exchange rate on Edwards' calculation.
In conclusion, Edwards seems to have had difficulty in applying the System of National Accounts to the issue, he confuses GDP, GNP and economic welfare and he does not understand the redistributive effect of a rise in the exchange rate. As a result, Edwards greatly understates both the benefit of the boom and the burden of adjustment caused by the fall in the international terms of trade.
The lone dissenter is sometimes right, but not in this case.
Photo courtesy of Flickr user Robyn Jay.