A press release yesterday from the QLD Govt, and associated social media posts like this from Jackie Trad, are suggesting that there is a jobs boom underway in regional QLD. While we steer away from making political comment on this forum we will make comment about data, and whether it is being used properly. So is this claim true? The short answer is no. The data being presented by the Govt to support their claim is confused and misleading (at best). I’ve added some comments to the official media release (in red) below which hopefully explains why, and then presented the picture as it really is.
First the media release (with comments)….
—
Palaszczuk Government backing 75,900 regional jobs (not regional jobs…this number refers to seasonally adjusted employment increase across whole of QLD)
Regional Queensland is enjoying a jobs boom under the Palaszczuk Government (not true…see below).
Hot on the heels of last week’s jobs data which revealed that labour force participation in Queensland has hit a seven-year high (true), figures released today confirm that the Palaszczuk Government is creating jobs in every corner of the state (hmmm).
Premier Annastacia Palaszczuk said that 75,900 jobs have been created outside Greater Brisbane since her Government came to office (not true…this number refers to whole of the state).
“Over the past year alone, 60 per cent of the jobs created in Queensland have been outside Greater Brisbane,” Ms Palaszczuk said. (this is only true if we use 12-month averages across the State and regions. The 75,900 number previously quoted is NOT a 12-month average…we are now comparing apples and oranges. When we use 12-month averages consistently the employment increase in QLD falls to 39,300 of which 60% are from “outside Greater Brisbane” as quoted)
Deputy Premier and Treasurer Jackie Trad said data from the Australian Bureau of Statistics reveal that the state’s lowest regional unemployment rate was in Cairns, where the creation of 20,900 jobs since January 2015 has driven unemployment down to just 4.8 per cent. (again, this is 12-month average data)
“The regional jobs boom in Queensland means that the Cairns labour market is now as tight as the labour market in the state of Victoria,” Ms Trad said. (if we use a 12-month average for VIC then unemp rate there is 4.7%…so nearly true)
“Over the past year, Cairns, Townsville, Darling Downs, Central Queensland, the Outback and Wide Bay regions have all enjoyed an increase in jobs or a decrease in unemployment – or both. (again, this is 12-month average data)
“The Palaszczuk Government is backing regional Queensland. The same cannot be said of the LNP, which oversaw the loss of 11,400 regional jobs while in office.” (I cannot be bothered to waste time trying to verify this number as there is no indication of which measure they are using nor over which precise period…it’s largely irrelevant anyway)
Minister for Employment, Small Business, Training and Skills Development Shannon Fentiman welcomed the continued evidence of strong jobs growth under the Palaszczuk Government.
“We are committed to creating jobs right across our state, especially in our regions,” Ms Fentiman said.
“Just today we announced Skills Boost which is slashing the cost of TAFE in the industries where we know there will be jobs.”
“This will ensure we have the skilled workforce we need and continue our 36th month of consecutive job growth.”
More than 226,800 jobs have been created in Queensland since January 2015. (true, based on Trend ABS data…the preferred measure)
As regular readers will be aware, we dislike using the extremely lagged 12-month averages (mainly) quoted in this release and that is why we created the Conus/CBC Staff Selection Trend for analysis at the regional level. By using the ABS and Conus/CBC Staff Selection Trends consistently across the State and regions we can get a more accurate, timely and relevant picture. Doing so tells a rather different story to the official line.
Trend employment increases over past 12 months.
QLD +58,500
Greater Brisbane +36,800 (63%)
Rest of QLD +21,700 (37%)
SEQ +48,100 (82%)
(Greater Brisbane + Gold and Sunshine Coasts + Darling Downs)
Outside SEQ +10,400 (18%)
It’s clear that the ‘regions’, especially outside the SE corner, are NOT enjoying a “jobs boom”.
The full Conus/CBC Staff Selection Regional Employment Trend data is available for download below. Please feel free to use this data but kindly acknowledge Conus/CBC Staff Selection when you do so.
Conus/CBC Staff Selection Trend Regional Employment QLD – Sept 2019
October 29th, 2019 at 2:10 am
Hi Pete, looks like the quality control on government media releases has gone out the window! Good work. I must say though that I think there is just too much sampling error in the regional data to say much that’s meaningful except where the numbers are extreme.
October 29th, 2019 at 3:06 am
Thanks Gene. Obviously I disagree with you to an extent re the regional data. I believe that our Trend analysis, particularly when looked at over a reasonable time-frame (say 4+ months) can provide useful insight into movements within the regional labour markets even allowing for the high levels of volatility inherent. Certainly it provides us with a much more timely and usable data-set than a simple annual average.
November 4th, 2019 at 3:44 am
Always have to go back to comments from Ricardian Ambivalence on the employment data and that the employment number is lower quality than the unemployment rate. Particularly in regional areas where the monthly population assumption input even beyond sampling error is really just a guess. This is why Bill Cummings is right and the census numbers demonstrate significant problems with the official LFS numbers for employment on a regional basis. I know Pete disagrees on this.
November 4th, 2019 at 4:18 am
I don’t disagree that there are issues around the regional employment data particularly given the input of estimated population data. But I do think that suggesting we have to rely on a 5-year snap-shot from the Census and can therefore ignore other indicators (even if they have issues) makes little sense.
I know Bill’s argument about the disparity between the ABS LFS and Census data; but there are huge disparities between these 2 at the national level (far more than when comparing to our regional Trend data)…do we therefore ignore the LFS data at the national level too?
For example the 2016 Census shows a national labour force just under 11.5 million while the LFS for that month had it at 12.7 million…more than 10% higher. Yet in Cairns the Census shows a labour force of 115,200 while our Trend data had it at 117,400…a difference of less than 2%.
The Census had the unemployment rate in AUS at 6.9%, the LFS 5.6%; in QLD it had 7.6%, the LFS 6.1%; in Cairns it had 7.8%, our Conus Trend was at 7.9%.
I continue to believe that regional data (when seen through the lens of our Trend analysis) provides useful insights.
November 5th, 2019 at 1:49 am
First, like Gene I agree with the general conclusion of your post but maintain concerns around the employment numbers at a regional level. At a national level the difference between the Census and LFS number is consistent within reasonable parameters between census periods. At a regional level it is not and becomes unreliable even on employment trends over even a five year period. The census data is also supported by the irregular ABS series based on ATO data which I think is actually the best quality data albeit delayed and annual.
The LFS data for Cairns shows a decline in employment between the 2011 and 2016 census. The census and ATO data contradict this. These measures are consistent at a national level. The increase in Cairns employment since 2016 should be observed in this context. Significant employment revisions in Mackay and Townsville are also noted.
I agree that your regional trend data provides useful insights. The limitations of the LFS data though remain material.
November 5th, 2019 at 2:00 am
Mark, actually our Trend series shows an increase of 5%+ in employment in Cairns between the 2006 and 2016 Census.