A statistic that deserves more air-time

I've been tracking the New Zealand business cycle for quite a while, and thought I knew most of the statistical and data sources. Somehow or other, one of them slipped under my radar, and it may have slipped under yours, too, as it doesn't look like it gets a lot of media coverage. So with a grateful nod to Stephen Toplis, Head of Research at the BNZ, who put me onto it, a few words about the job vacancies data published by the Dept of Labour (now part of MBIE).

First of all, here's what the latest (June) data are showing, for all vacancies. The commentary from the DoL mostly focuses on the skilled vacancies, and there's logic to that, but in any event here's the overall picture. If you're interested in regional and occupational data, it's all there, too.

The graph shows what you'd expect: solid and ongoing expansion in vacancies, consistent with the strengthening of the overall economy, which on the vacancy series has been underway now since mid 2009. But we haven't reached boom times for job seekers: the series isn't quite back to the strong levels of 2007 (when this series starts). It's also worth remembering that while a rising vacancies series is generally a good thing, as it tends to be linked with positive things happening in the labour market, it can also be picking up some potential problems (eg more employers unable to find the skills they want).

What's the basis of the data? They're the total online job ads listed on the three big employment sites, SEEK, TradeMe and heraldjobs.co.nz. These days, it's a big number - some 20,000 distinct job ads each month, which is another little glimpse of the powerful disintermediation happening on the web.

For economists, and for businesspeople, it's worth knowing that the data also have some predictive power. In the 2009 background paper explaining the data, the DoL drew this graph. It shows the Skilled Vacancies Index leading employment growth. If you shunt the vacancies index one quarter to the right, you get an extremely close link, with a correlation coefficient of 0.95.

There's also a lot of value for both employers and employees in the detailed occupational and regional data. If, as a business, you were wondering whether to pay a bit over the odds to get a vacancy filled, I'd have a quick squizz at what this series is telling you about the state of the market for the skill  you are after, and conversely for employees.

I should have known about this series before, and now that I've finally made its acquaintance, I'd urge other people to make more use of it, too. With all due respect to Stats NZ, we're not blessed with an immense range of monthly economic data in New Zealand, and I understand why - priorities, costs, and sometimes the limited value-add of monthly data, which can contribute more statistical 'noise' than genuine new information.

All that said, anything additional on a monthly basis is a bonus. And as the DoL says, "In some instances it is a robust leading indicator, and has the considerable advantage of being very up-to-date" (the June data were published this morning, July 17). It's also fine work by the DoL to use data that was being prepared by businesses for their own commercial purposes in any event, a trend that is becoming more prevalent at statistical agencies everywhere as they make more use of administrative and commercial data, and for obvious reasons. As the DoL says, "Jobs Online brings together advertised job vacancy information from the major internet job boards, maximising the information value of the data they hold to create new labour market statistics without an increase in respondent burden or compliance costs".