Your editorial yesterday highlighted important aspects about labour force figures in the Northern Territory.
It is true that NT unemployment figures are subject to greater volatility than those compiled in other states and territories. This is a consequence of the NT's smaller population, and hence smaller sample size. Unfortunately, solutions are not as obvious as they may appear to be.
While increasing the sample size is one option, such change needs to take into account increased costs of running the survey and the increased burden placed on the community to provide survey responses.
Using Centrelink 'dole' payments data to measure the level of unemployment would provide a very narrow measure. It would exclude, for example, those people who are not entitled to such government benefits even though they are actively looking for work.
Including people employed through Community Development Employment Programs (CDEP) in the employment numbers is in line with international standards, and recognises that the CDEP arrangement provides a wage paid by an employer for work undertaken under the program.
ABS has always been upfront in alerting users of its data if there are particular issues relating to quality which we think are important enough to raise. If a warning about data is provided it does not mean the data are flawed or useless. It means they need to be interpreted with care.
At the end of the day NT labour force figures are compiled using well tested methodologies that stack up with the best in the world. NT labour force figures are a good indicator of change over time, but there are times when caution in interpretation is warranted, and additional information from other sources, such as benefit recipient series can help fill out the picture.
ABS is continually looking for ways to improve the quality of its survey estimates, and will keep users informed of any proposed enhancements to the methods that we use.
Australian Bureau of Statistics, Northern Territory
(08) 8943 2130
This page first published 18 October 2004, last updated 24 July 2008