Finding better indicators to read the health of the economy

I could attempt to explain to you why the Bureau of Statistics is having such embarrassing trouble with its monthly estimate of employment, but I won't bother. It's horribly complicated and at a level of statistical intricacy no normal person needs to worry about.

What this week's labour force figures now tell us is that, though the rate of unemployment has been slowly drifting up since mid-2011 - when it was 5 per cent - it seems to have steadied this year and, using the smoothed figures, has stayed stuck at 6 per cent for the past three months.

What numbers should we be looking at, asks Ross Gittins.
What numbers should we be looking at, asks Ross Gittins. Photo: Tamara Voninski

This is reasonably consistent with what we know about other labour-market indicators, such as job advertisements and vacancies, claims for unemployment benefits and employers' answers to questions about hiring in the National Australia Bank's survey of business confidence.

It also fits roughly with what the national accounts have been telling us about the strength of growth in the economy. We know that when the economy is growing at its trend rate of about 3 per cent a year, this should be sufficient to hold the rate of unemployment steady.

The accounts told us real gross domestic product grew by 3.4 per cent over the year to March, and by 3.1 per cent over the year to June.

But now let me tell you something that, while a bit technical, is much more worth knowing than the gruesome details of the bureau's problem with the labour force survey.

One of our smartest business economists, Saul Eslake, of Bank of America Merrill Lynch, has reminded us that GDP is only one of various summary indicators of overall economic activity provided by the national accounts. And the economy's peculiar circumstances over the past decade and for some years to come mean GDP is not the least misleading of the various measures.

Eslake says real GDP measures the volume (quantity) of goods and services produced within a country's borders during a particular period. (Actually, it doesn't include the many goods and services produced within households, which never change hands in a market.)

To estimate real GDP the bureau takes the nominal, dollar value of the goods and services produced, then "deflates" this figure by the prices of those goods and services relative to what those prices were in the base period.

We commonly take the value of the goods and services we produce during a period to be equivalent to the nation's income during that period. This easy assumption works for most developed economies most of the time.

But Eslake reminds us that "for an economy like Australia's, the prices of whose exports are much more volatile than those of other 'advanced' economies, abstracting from swings in the prices of exports (and imports) obscures a significant source of fluctuations in real incomes".

We've experienced a series of sharp swings in our "terms of trade" - export prices relative to import prices - over the past decade of the resources boom, which was interrupted by the global financial crisis in 2008-09. For the past three years, of course, mining commodity prices have been falling.

Trouble is, real GDP doesn't capture the effects of these swings. So the values of our production and our income have parted company, as they do every time our terms of trade change significantly. An improvement in our terms of trade causes our income to grow faster than our production, whereas a deterioration has the opposite effect.

This matters because of the chicken-and-egg relationship between production and income: we use the income we earn from our part in the production process to buy things and thus induce more production. So if our real income slows or falls, soon enough this dampens our production.

However, the national accounts include a measure of overall economic activity that does capture the effects of movements in our terms of trade: real gross domestic income, GDI. It grew a lot faster than real GDP for most of the time between 2002 and 2011, but since then has grown much more slowly than real GDP (a big reason for our slowly rising unemployment).

Next Eslake says that as the resources boom moves into its third and final phase - with mining investment winding down and exports ramping up - real GDP growth will be an even less useful guide to what's happening to domestic income and employment.

This is because maybe 80 per cent of the income generated by resources exports will be paid to the foreigners who own most of our mining companies and financed most of the new investment.  It's also because the depreciation of Australia's greatly enlarged stock of capital equipment and structures as a result of all the mining investment spending will now absorb a greater share of our gross income.

(A separate issue Eslake doesn't mention is that the highly capital-intensive nature of mining means the increased production of mineral exports will create far fewer jobs than you would normally expect.)

If you've ever wondered about the difference between gross national product and gross domestic product it's that the former excludes all the income earned on Australian production that's owed to the foreign suppliers of our debt and equity financial capital, making it a more appropriate measure for us given our huge foreign debt and foreign investment in our companies.

If you've ever wondered what the "gross" in GDP, GNP, GNI etc means, it's short for "before allowing for the depreciation of our stock of physical capital".

So gross national income (GNI) is a better measure than gross domestic income (GDI), and net national disposable income (NNDI) is a better measure than GNI.

Which, by the way, explains why real NNDI is used as the base for all the further non-national-accounts-based modifications included in Fairfax Media's attempt to calculate a broader measure of economic welfare, the Fairfax-Lateral Economics wellbeing index, released each quarter soon after the publication of GDP.

Ross Gittins is the economics editor.