A further look at the Cape York employment data following previous post on Aurukun. There was reform to indigenous employment schemes for remote communities from July 2015 with the implementation of the Community Development Program (CDP). This would correlate with the trend change in the Small Area Labour Market (SALM) data from that date.
The SALM explanatory notes do seem to indicate that some account is taken for this in the Centrelink data applied in the methodology. The increased unemployment rate in the SALM data for Aurukun is derived from both an increase in the number unemployed and a fall in the labour force size.
However the SALM methodology is also intended to be consistent with and also derived from the ABS Labour Force Survey for the SA4 region. The SA4 data for Outback follows the same clear trend from July 2015. This is Outback SA4 with a 12 month moving average and also the Conus Trend (blue):
Again the trend break in July 2015 is significant across all factors reported by ABS (unemployed, labour force, participation, etc.) contributing to the increased unemployment rate. The 12 month average just like the 4 quarter SALM data tends to smooth what was a sharp break from July 2015. It isn’t clear to me why the labour force survey results for the entire SA4 region should be influenced so significantly by any CDP changes.
As suggested previously the Outback SA4 region is statistically difficult with a relatively low population (and sample size) covering a vast geography from Torres Strait to Birdsville with not much economic connection between those. It is also possible that much of what we are seeing in the trend reflects issues in the labour force sample for Outback.
The ABS sample rolls over eight months with only an eighth of the sample rolling in and out for successive months. The recent period has seen some extraordinary volatility even for Outback. In the recent March quarter of the SALM data the raw ABS unemployment rate for Outback SA4 has varied between 1.5% and 19.0% in successive months. Which is of course nonsense.
When the SALM methodology is applied to all the SA2 areas in Outback SA4 the unemployment rates look like this.
I don’t know. It may be worth a query to ABS and Dep’t of Employment.