Why the jobs figures don't add up
Strong February data and upwardly revised estimates for 2013 point to a strengthening labour market. But the ABS suggests that you shouldn’t be popping champagne corks just yet.
Many analysts had expected the labour market to pick up but nobody expected employment to expand as strongly as it did in February. Employment rose by 47,300 in February, around three times higher than expectations, and is now up around 70,000 over the year.
Employment growth was driven by a stunning turnaround for full-time employment, which rose by 80,000 in February. By comparison, part-time employment fell by 33,000. In fact there were large moves in all major categories.
However, there is a huge caveat to these exceptional data. The ABS estimates that around 37 per cent of employment growth in February was driven by sample changes and 29 per cent of the decline in people not in the workforce. The ABS instead suggests that we focus on trend estimates.
On trend basis, employment rose by 8,500 in February, with three quarters of that monthly gain reflecting full-time employment growth. Employment growth for men exceeded the modest rise in female employment. Suddenly the data seems far more sensible and consistent with other measures of activity.
The seasonally-adjusted employment data was heavily revised for this publication, with employment growth over the year to January shifting upwards by 0.4 percentage points. As a result, the labour market was not as weak during 2013 as initially expected. Nevertheless by historical standards, annual employment growth remains subdued.
The unemployment rate rose by 0.1 percentage points to 6.0 per cent. The good news however is that the participation rate has begun to trend upwards.
This is largely due to a return of disgruntled former workers to the labour force -- in addition to some sampling issues. By comparison, the retirement of ‘baby boomers’ continues to weight on participation but at least for now it is being offset by other factors.
RBA deputy governor Philip Lowe had plenty to say last night on the long-term outlook for the labour market and participation (Coming to terms with Australia’s stark economic reality; March 13). The reality is that any immediate recovery in labour market participation will likely prove short-lived and the participation rate will recommence its downward trajectory soon. Naturally that will weigh on employment growth and also push the unemployment rate downwards in coming years.
Obviously the headline data is fantastic and suggests that some of the recent improvement in retail spending and lending is finally feeding its way through to employment. The economy appears to be gathering momentum as it faces the upcoming challenge posed by the imminent collapse in business investment and potential austerity measures enacted in this year’s federal budget.
However, I’d caution readers from getting too far ahead of themselves. By the ABS’s own admission the strong February data is unreliable and the trend data point to a far more modest improvement in labour market conditions. Quite frankly, this is a confusing monthly report and one that will be misinterpreted by many analysts.
Last year there was also a remarkable increase in employment for February; however, much of it was revised away and a number of analysts were left feeling fairly foolish. Initial estimates for February 2013 pointed to monthly employment growth of over 70,000 people – the estimate for today's report was for a far more modest rise of 27,000.
So for now we should be cautiously optimistic about the labour market but recognise that there remain significant challenges for the Australian economy. The long-term outlook for employment remains weak but conditions appear to have improved a bit during February and may improve further over the first half of the year.
But the February data was certainly not as strong as the headline data indicates and, consistent with the ABS recommendation, I recommend that readers focus more on the trend than the less reliable seasonally adjusted data. I’m fairly sure that is how the RBA will approach the data and they won’t be jumping to any conclusions just yet.