Introduction to the SHS longitudinal data

The Specialist Homelessness Services (SHS) Collection (SHSC) commenced on 1 July 2011. Data are provided by over 1,700 government-funded agencies that deliver homelessness services to people in need of support. These data are based on interactions between clients and service providers and are collected at fixed points in time:

  • at the start of a support period
  • at the end of every month during a support period
  • at the end of a support period.

SHS Agencies provide these data to the AIHW on a monthly basis. For detailed information about the SHSC see Specialist Homelessness Services Collection.

More information about the SHS data

The following resources provide important details of the AIHW SHSC that assist in interpreting the longitudinal analyses of the data:

SHS National Minimum Data Set – describes the collection and usage attributes of the data, including definitions of the variables collected.

SHS data quality statements – published annually and provide key insights into the coverage, relevance and accuracy of the data.

SHS Collection Manual (PDF 2.6 MB) – provides a ‘user friendly’ guide to collecting data for agency workers including how the questions are structured.

SHS annual report – summary report on clients and services provided within a financial year. These reports provide useful information on the characteristics of clients and of their service use, including for key client groups and are accompanied by interactive data cubes.

SHS monthly data – high level summary data released quarterly on the number of clients receiving support each month.

In addition to this, the SHSC data have potential to support longitudinal analyses. Specifically, the types of support provided to clients over time can be examined and can provide valuable insights into the experiences of SHS clients over time and as well as important evidence for future policy development.

The AIHW has explored this previously in reports that examined patterns of service usage across multiple years, for example:

SHS clients can be either experiencing homelessness or at risk of homelessness

When interpreting SHSC results, it is important to note that not all clients in the data are homeless. Many are ‘at risk’ of homelessness and some of these never experience formal homelessness. For example, more than half (56%) of clients with closed support started their first support period of 2020–21 at risk of homelessness and 67% finished their last support period of 2020–21 at risk of homelessness (AIHW 2021).

Longitudinal analyses

The analyses of SHSC data presented here are based on longitudinal data constructed from support-period level data from 2011–12 onwards. Analyses of these data examine aggregate client characteristics within 3 study periods (see Box Intro.1). These study periods facilitate the analysis of client characteristics within a set cohort period (defining study period), over periods of past SHS support (the retrospective study period) and their future SHS experiences (the prospective study period).

The SHS longitudinal data contain information for over 1 million individual clients nationally since 1 July 2011. The data offer a unique opportunity to quantitatively explore, and in some cases, validate or replicate, the myriad qualitative studies on homelessness that, while rich in depth (being based on interviews, questionnaires, or case notes), are often based on relatively small sample sizes.

Longitudinal analyses of SHSC data cannot answer all questions about people experiencing insecure housing. The analyses presented here make the most of the available data by providing unique insights into the support profile of SHS clients over long periods of time. They also highlight areas that may warrant further and more detailed investigation through further research or other analyses of data.

Box Intro.1: SHS longitudinal analyses - Study period approach

The SHSC longitudinal data are analysed by partitioning a client’s time receiving SHS support into 3 periods (also Figure Intro.1):

Defining study period

This is the key period for which client characteristics and experiences are aggregated. It can vary in length for each analysis but is typically 12 months, to include an adequate sample of clients and to cover a long enough period to capture client outcomes. All clients will have at least one support period in this period, because the cohort is defined by clients who met a certain set of conditions within that time.

For example, a defining study period may be limited to women with a family or domestic violence experience/service needs within the period of 2015–16. Characteristics of these clients and their service use patterns would be aggregated within the 12-month defining study period. The 12-month period is different for each client, beginning on their first day of support that meets that inclusion criteria – in this case, being an FDV client and support starting in 2015–16.

Most analyses will also look at the patterns of service use before (retrospective period) and after (prospective period) the defining study period.

Retrospective study period

This period examines data for clients that needed SHS support before the defining study period. Analyses in the retrospective period examine whether a client experienced particular circumstances (for example, homelessness, unemployment, used a particular service type) in the past. A client’s presence in the retrospective period is itself a characteristic that can be studied. For example, clients that are not in the retrospective period may be considered ‘new’ to SHS services in the defining study period.

Prospective study period

This period examines data and outcomes for the clients that continue receiving support into the future. The most basic outcome is whether a client needed SHS support in this period – that is, whether they can be considered as an ongoing or long-term user of SHS (continuous or otherwise), compared with clients whose SHS engagement finished during the defining study period. More detailed analyses examine whether specific events occurred during the prospective study period, such as whether the client was homeless or what services the client used.

Figure Intro.1: Study period approach to longitudinal analyses

The diagram shows how the SHSC data are partitioned into the three study periods for longitudinal analysis. The middle box demonstrates the key defining study period, which is the length of time within which a specific client cohort is selected and client support and characteristics are analysed. An arrow to the left demonstrates the retrospective study period which includes analysis of historical support and client characteristics for this cohort. An arrow to the right demonstrates the prospective study period which includes analysis of future support and client characteristics of the client cohort.