Better late than never

Exactly one year ago, Henry called for a 50-basis-point rate hike as a pre-emptive strike against inflation.

Sadly, this advice was ignored and now, according to two commentators (and more will follow), Australia may be facing three 25-basis-point rate hikes. Scott Murdoch and Richard Gluyas said in this newspaper late last week that now there is a risk of "three rapid rate rises". "Too little, too late" in 2006, 2007 and probably also in 2008.

Excessive caution is the main reason. A nation of course needs cautious central bankers, but we are badly served when this tips into timidity and (effectively) lack of faith in the resilience of the nation's economy.

Closely related to overcooked caution is excessive conservatism in forecasting. As the boom gathers pace, the evidence is emerging almost every time a new economic statistic is released. Consider the evidence.

Most importantly, employment growth in Australia has exceeded everyone's expectations, as has the fall in the rate of unemployment. The tendency for the economy to outperform "reasonable" expectations is, in fact, endemic.

Other recent economic statistics underline the point -- the economy is running away from the economic analysts. The quotes below are from "market economists", names suppressed to avoid any unintended slurs.

"Retail trade turnover grew by 0.8 per cent in September, which was stronger than the 0.5 per cent gain markets had expected. For the year to September, retail turnover was up a very healthy 8.2 per cent." As a side comment we must say: "Healthy!? The sentence is OK if you put "un" in front of "healthy", or use the word "unsustainable".

There is more: "Australia's trade deficit widened to $1.862 billion in September, from a revised increase of $1.665 billion in August. This was much higher than the market consensus for a $1.0 billion deficit."

Are readers as tired as Henry is of reading that economic data "exceeded expectations" or is "much higher than the market consensus"?

Do those who form these "expectations" or shape the "consensus" ever stop to reflect on their forecasting record? Do they ever take a stroll in a shopping mall, or think about record growth of household debt and the basis for record bank profits? If Australia were a nuclear reactor the attendants would be heading for the hills after operating the closedown procedures.

Another recent economic story was equally aggravating. "Building approvals jumped 6.8 per cent per cent in September, sharply retracing the 1.8 per cent decline seen in August. This is well above market expectations, which had centred around a 0.7 per cent increase. Total credit provided by the private sector rose by 1.2 per cent in September, following a strong increase of 1.4 per cent in August. The market had been expecting an increase of 1.0 per cent in the month."

The latest report of goods and services inflation shows the same tendency. Goods and service (CPI) inflation was a relatively modest 0.7 per cent for the September quarter and 1.9 per cent over the year. This has led the Treasurer to say the Reserve Bank would be acting against his "contract" with its governor if it raised rates.

There are two points that modify this argument. First, actual annual CPI inflation is artificially lower because the dropping out of the "temporary" banana price hikes a year ago. And the main items well above the artificially low overall rate were alcohol, housing, health and finance and insurance services, which is why Struggletown is doing it so hard. However, the more important point is that "underlying" goods and services inflation was 0.9 per cent for the quarter and 2.9 per cent for the year. This keeps the Reserve Bank on the hook, and adds to the rate hike case for the meeting of its board on Melbourne Cup day.

To return to our main theme of forecasting errors, another quote: "Producer prices grew by 1.1 per cent in the September quarter, well above expectations".

Further support for the forecasting bias hypothesis comes from an unexpected source. Michael Bachelard reported at the weekend for The Age that "real" (i.e inflation adjusted) pay has gone backwards under the Howard Government's new wage setting system, with some workers on award pay rates now more than $15 a week worse off in the past 12 months. (This is part of the reason for such strong employment growth, incidentally.)

Bachelard said: "Recent higher inflation has overtaken the modest pay rises handed out by the Australian Fair Pay Commission last month, pushing most of the 1 million workers covered by award rates into the red, according to new figures.

"The commission confirmed the figures yesterday, saying it had relied on Reserve Bank inflation forecasts, which were too low."

"The inflation forecast was 1.6 per cent for the period December 2006 to October 2007. We now know the actual outcome for the period was a rate of CPI inflation of 2 per cent," Commissioner Ian Harper told The Age.

It is impossible to believe the clever economists at the Reserve Bank have not noticed the clear tendency for economic outcomes to exceed expectations, and the market consensus that they do so much to shape.

The graph provides a look at 50 years of inflation, including measures of inflationary expectations of consumers. There is an opposite bias in consumers -- often their forecasts for inflation are too high. (This tendency was particularly evident after the high inflation created by the Whitlam government.) However, the hike in inflationary expectations, to an annual rate of 4.6 per cent, in October is alarming. It follows a general upward drift of all main measures of inflation since the late 1990s. The latest rise of inflationary expectations will fuel further sizeable wage claims, as we have already seem in the mining and service industries, most recently with nurses in Victoria.

A brave central bank would implement a 50-basis-point pre-emptive strike now. If there is no hike at all, the incoming government -- of whatever colour -- should think hard about the terms of its agreement with the Reserve Bank.