 |
Estimated Unemployment Rates by Electorate, December Quarter 1997
Andrew Kopras & Tony Kryger
Statistics Group
23 June 1998
Introduction
The Statistics Group of the Department of the Parliamentary
Library has been compiling unemployment rate statistics by electorate
since March quarter 1996. These statistics are derived from small area
labour market estimates produced by the Department of Employment, Education,
Training and Youth Affairs (DEETYA). DEETYA estimates are in turn based
on Australian Bureau of Statistics (ABS) labour force survey and population
data and Department of Social Security figures.
The information in this Research Note refers to the Electoral
Divisions as they will be at the next election, that is following the
1997 redistributions in Western Australia, Queensland and the ACT. This
means that for these States and the ACT the party shown is notional party
only.
Methodology
Each quarter DEETYA produces unemployment and labour
force estimates for around 700 Statistical Local Areas (SLAs). Using these
as the basic building blocks, it is possible to construct an unemployment
rate estimate for each electorate in Australia. DEETYA unemployment and
labour force estimates at the state level are exactly equal to the corresponding
ABS estimates.
To construct an unemployment rate estimate for an Australian
electorate it is first necessary to allocate SLAs to electorates. If an
SLA falls wholly within an electorate there is no problem. If an SLA overlaps
two or more electorates however, the unemployment and labour force estimates
for that SLA have to be apportioned between electorates. From the Census
it is known how the population of overlapping SLAs is distributed between
electorates and this proportion is used to distribute unemployment and
labour force totals. It is then a matter of aggregating the data for SLAs
and part SLAs within electorates and calculating an overall unemployment
rate for each electorate.
The problem of SLAs that overlap two or more electorates
and the crude method of dealing with this problem (using population data
as the basis for apportioning unemployment and labour force totals between
electorates) means that unemployment rates by electorate are approximations
only and therefore should be treated with caution.
Results
- Unemployment rates for the December quarter 1997 ranged from a low
of 1.6 per cent in Bradfield (NSW) to a high of 15.8 per cent in Bonython
(SA). The average unemployment rate for all Australia was 8.0 per cent.
- There is a very definite skewing of ALP and National Party seats toward
the upper end of the unemployment rate spectrum. For example, a quarter
of all seats have an unemployment rate of 10.5 per cent or more. Of
these, 16 seats are held by the ALP, 15 by the Liberal Party and 6 by
the National Party (or 33, 19 and 33 per cent of all seats held by the
respective parties). At the opposite end of the spectrum, a quarter
of all seats have an unemployment rate of 6.1 per cent or less. Of these,
only 5 seats are held by the ALP (10 per cent of all ALP seats), 27
by the Liberal Party (35 per cent of all Liberal Party seats), only
one seat is held by the National Party and 4 seats are held by Independents.
- Three quarters of all Western Australian seats have unemployment rates
that are below the national average. This contrasts with Tasmania in
which there are no seats with an unemployment rate below the national
average. In between are New South Wales and Victoria (majority of seats
below the average) and Queensland and South Australia (majority of seats
above the average).
|
 |
