Odd comments yesterday in the ANZ’s analysis of its latest job ads figures, which raises questions about its relevance, or rather its interpretation.

And they also raise the question of whether the survey could reveal more about internet job advertising, to establish where the growth in online jobs is occurring, either by state or by industry sector.

In its May job ads series yesterday, the ANZ said there had been a fall in newspaper job ads in the month, which was given more weight than the rise in internet advertising.

The ANZ said in commentary “the strength in job advertisement numbers in May was not broadly based, instead driven entirely by a 5% rise in internet advertising. Newspaper job advertisements in contrast fell 6.8%. This was driven by falls in all states and territories except for New South Wales.

“Hence, while the overall outlook appears bright, the fall in newspaper job advertising in May does suggest that some businesses are beginning to adopt a more cautious stance. This is not unexpected given three consecutive interest rate rises in the first half of 2010 as well as the now uncertain global backdrop.

“With employment growth racing ahead of output growth in the first part of 2010, it may also be that some firms begin to slow the addition of new workers until growth in activity picks up. This would boost Australia’s hitherto subdued productivity performance.

“Despite the strong rise in total job advertising, the slowdown in newspaper job advertising growth does imply some impending moderation in Australia’s recent strong employment growth.”

Just how can a fall in advertising in a sector covering some 5.5% of all job ads last month justify those comments when there was overwhelming growth in the other 94.5%, i.e. internet job advertising?

Newspaper job ads in the month totalled 9245, which is more than half the level it was at its peak in April, 2008 of 20,207 ads.

Internet ads were 158,389, compared to the peak in April 2008 of 258,042.

Looked at another way, in April 2008 newspaper ads were 7.8% of all job ads in that month. In May of this year they were down to 5.5%. That’s bordering on the irrelevant as an indicator.

That’s further emphasised by the size of the fall from April to May in newspapers, just 639 fewer ads (and in only the major capital city papers. Papers in places such as Newcastle, Wollongong, Geelong, the Gold Coast, Cairns, etc, are not measured, nor are suburbans). That’s just 160 fewer ads each week in May in the major metro papers.

And the fall in the number of newspaper ads was piddling compared to the 7509 new jobs advertised on the net in the month, or that’s nearly 1900 new jobs every week.

And with newspaper ads broken down state by state, the ANZ could do better analysis on net advertising by state or by industry to give us a better reading of where the growth is occurring.

There is the assumption in the survey that internet ads are national, which is true. If they are being filled by people outside the geographic location of the advertising employer, that would indicate a greater degree of mobility in the labour force than it’s thought. If the ads are principally being filled by people from within the geographic location of the advertiser, then that could indicate less mobility that thought.

Either way the absolute dominance of job ads by the internet demands far greater analysis each month than we now get, more than the newspaper ads, which are fading.

Trends in the bigger and more comprehensive internet job ads market would seem to be better indicators of the direction of activity in the economy rather than newspaper display and classifieds, which tend to concentrate on higher-paying management or executive roles. The net is more blue collar, clerical, etc, employees rather than managers.

Finally, that 6.8% drop in internet advertising is minor news in the national jobs market, but it’s a sign of continuing revenue pressures for the publishers, Fairfax Media, News Ltd and West Australian Newspapers (Kerry Stokes).