Spatial Variables

From Release 5.0 onwards, the household addresses from each wave have been geocoded and assigned a 2001 Census Collection District (CD). Where the address details were not sufficient to geocode exactly, the nearest cross section or street segment was used. Further, some fuzzy matching and manual look-up of maps were employed where the street name or suburb did not provide a reasonable match. We are able to build up from CD level to the following geographic regions:

• Statistical Local Area ( SLA );
• Local Government Area (LGA);
• Statistical Sub-Division (SSD);
• Statistical Division (SD);
• Section of State (SOS); and
• Major Statistical Region (MSR).

The General Release (confidentialised) HILDA files do not include geographical descriptors for CD, postcode, SLA, LGA, SSD and SD. These files only include State, Section of State and MSR. The In-confidence Release (unconfidentialised) files include all geographical descriptors mentioned above.

The table below lists the derived spatial variables. Aside from the area identifiers, several other spatial variables are included on the file such as:

• Remoteness area – this is derived based on the assigned SLA;
• Socio-Economic Indexes for Areas (SEIFA) – deciles are assigned for four types of SEIFA scores based on the assigned SLA ;
• The distance moved from the last wave –this is calculated from the geocoded addresses. Where the geocoding had to be approximated and the household moves close by, there may be some households who have moved but the distance moved is calculated as zero.

Other related spatial variables which are not derived that you should be aware of are State (_hhstate) and whether the household has moved from the last wave (_hhmove).

Table 15: Derived spatial variables