Study Name:Household, Income and Labour Dynamics in Australia (HILDA) Survey

Study Director:Mark Wooden

Principal Investigators:Peter Dawkins, Gary Marks, Ruth Weston, Mark Wooden

Host Organization:Melbourne Institute of Applied Economic and Social Research, University of Melbourne

Year Initiated:2000

 

 

Governance

 

Funding sources

 

The Commonwealth Department of Family and Community Services (FaCS).

 

Host organization

 

The study is based at the Melbourne Institute of Applied Economic and Social Research, University of Melbourne, but is managed by the Melbourne Institute in association with the Australian Institute for Family Studies and the Australian Council for Education Research.

 

Governing body and role of external research advisors

 

Overall responsibility for the project rests with the Project Director, Professor Mark Wooden. He reports directly to FaCS through a nominated Project Officer.

 

Advisory input comes from at least three different sources:

 

(i)      a Steering Group (mostly comprising representatives from Government) advising FaCS on the management of HILDA;

(ii)    a Project Management Group chaired by the Project Director, and also including a representative from each of the three partners in the consortium and the FaCS project officer; and

(iii)   an external reference group (ERG) chaired by Professor Bruce Chapman and consisting of 7 other academic researchers and a representative from the Australian Bureau of Statistics.

 

Sample Design

 

Introduction

 

HILDA is only at the planning stage. It is not expected to begin the first Wave of data collection until late 2001.

 

Sample selection

 

It is expected that the initial sample will involve at least 8000 households selected using a clustered sample design. Clusters will be selected by sampling “collection districts” (or CDs) with a probability proportional to their size, as measured by the number of households recorded in the 1996 Census (but hopefully with corrections for growth since 1996).

 

 

Following rules

 

The following rules that are to be used are expected to be identical to those used in the British Household Panel Survey. Thus persons who subsequently become members of a households containing an original sample (e.g., as a result of birth or marriage, or because of other changes in household composition) become eligible for sample inclusion. These persons will only remain in the sample for as long as they remain in the same household as an original sample, with the exception of children born to an original sample member and parents of those children.

 

Oversampling

 

There is not expected to be any over-sampling of any key group, through stratification by region will mean the less populous States are over-represented.

 

Weights and attrition bias

 

The sample will be re-weighted each wave. It is expected that longitudinal weights will be derived from the results of econometric modelling of the attrition process.

 

Sample “refreshing”

 

Immigration will cause a need for some sample refreshing. This is not expected to take place within the first three waves of the survey.

 

Content

 

The primary objective of the HILDA Survey is to support research questions surrounding the following three broad and inter-related areas:

 

(i)      income dynamics – focusing on low-to-middle income households and their responses to policy changes aimed at improving financial incentives, and interactions between changes in family status and poverty;

(ii)    labour market dynamics – focusing on low-to-middle income households (including the role of part-time and casual jobs in escaping poverty), female participation and work-to-retirement transitions; and

(iii)   family dynamics – focusing on separation/divorce and social/economic status, and on any links between income support and family formation/breakdown.

 

The data should enable analysts to: explore the interdependencies and interrelationships between the various choices made by individuals and households and the constraints/events faced by individuals and households; investigate the impact of various life events; and examine the contextual determinants of change.

 

To meet the above objective, the data gathered by the HILDA Survey will support:

 

(i)      the analysis of relatively sophisticated descriptive questions — such as:

·         what are the patterns of duration in various states over time?

·         what do the transitions between various states or across boundaries look like over time?

·         what are the patterns, correlates and associations between the major items of interest? and

·         how do various life events impact on individual, family and household members over time?

 

(ii)    the investigation of behavioural determinants and causal factors — the gathering of various events and transitions in a temporal order is expected to allow analysts to disentangle the direction of and magnitude of causal relationships between the major items of interest (particularly the roles of heterogeneity, state dependence and behavioural responses to different incentives); and

 

(iii)   evaluation of policy — the collection of baseline information on a range of variables prior to the implementation of a policy allows rigorous policy evaluation to occur.

 

Collection

 

Mode

 

Wave 1 will be conducted using face-to-face interviews employing paper and pencil questionnaires. Thereafter, it is expected that the majority of interviews will be conducted by computer assisted telephone interviewing.

 

Reference population

 

All adult residents of private households aged 15 years and over.

 

Instrument design

 

In Wave 1 the survey instruments will be of the pen-and-paper variety. Separate instruments will be delivered on a household basis (to be answered by one or more members of the household) and on an individual basis (to be answered by all adult members of the households). Consideration is also being given to a leave-behind supplementary self-completion questionnaire.

 

Dependent interviewing

 

Only persons aged 15 years and over will be interviewed.

 

Calendar year, survey year and point of survey measures

 

Most information collected is likely to relate to the point of survey. Annual calendar information, however, will be collected on income sources and labour market activity. These data will relate to a period covering a financial year (July to June) and extending up to the month prior to interview, thus providing data on anywhere between 14 and 18 months of activity.

 

Processing

 

A detailed plan for the processing and management of data is still to be designed.

 

Dissemination

 

Again a detailed plan for the dissemination of the data is still to be determined. It is expected, however, that at a minimum, data will be made available for a nominal charge on a CD-Rom to bona fide researchers. We are investigating the option of making the data downloadable from our web site.