Review of 2021 Census Topics

Inclusion of enhanced questions about volunteer work in the Census, beyond the current "Yes/No" style response.

A question on Volunteering has been asked since the 2006 Census, but the question is very broad, and most of our clients gain little use out of it. Apart from broad spatial patterns (generally, volunteers are older, more likely to be born in Australia, and live in regional areas) there is little that can be gleaned from this question due to the broad nature of the question. It only asks whether the person has done volunteer work in the last 12 months for any organisation or group, nothing about the type of organisation, frequency etc.

A 2012 Flinders University study estimated the value of volunteering in the community across Australia at $290 billion, which would be around 20% of GDP if it was counted in economic terms. There is certainly a national importance in understanding this group, and the value of the sector is recognised in celebrations such as the National Day of Volunteers. Quite a number of Census questions relate to paid employment (labour force status, hours worked, industry, occupation, sector etc.) but only one relates to the volunteer sector and it is very broad. Volunteering is an important contributor to the Health Care and Social Assistance industry, which is the fastest growing in terms of employment, so understanding the contribution of volunteers would enable decision makers to adequately resource this area.

Suggested improvements to the volunteering question, as well as “Has the person volunteered in the last 12 months?” include “What is the frequency of volunteering (weekly, monthly, etc.)?”, “How many hours has the person volunteered in the last month?”, and “For what type of organisation does the person volunteer?”. The last would be a write-in question with mark boxes similar to industry and occupation currently.

Volunteering information is most important at a local level. It is one of the most commonly requested datasets from our Local Government clients at .id – mainly because Local Government has broad responsibility for a lot of community organisations which use volunteers, and often have strategies to engage with the community and encourage volunteering at a range of levels. The current “Yes/No” question doesn’t really provide a lot of insight beyond broad statements about whether the level of volunteering is high or low in the local area. The volunteering page on .id's community profile is not one of the most used pages - with about 12,000 page views in the past year, it ranks 29th of 38 topic pages - but this may be due to the lack of usefulness of the topic in its current form.

Finding out the types and frequency of volunteering in the local community would help these organisations target the programs and increase the level of volunteering with flow on effects for community wellbeing. This can only be done with local small area information available from the national Census.

Volunteering is currently collected on the Census and has been for the past 3 Censuses. The question is very broad, and this may be a response to the difficulty of collecting the information on a self-responded questionnaire. It may be that there is a subset of population who can remember that they volunteered, but not remember for whom or how long. The more complexity is added to these questions, the more difficult this would be to self-report.

However similarly detailed questions are asked successfully in the employment section of the Census – eg. Hours worked and occupation. The difficulty may be in the recency of this – eg. The question on employment asks how many hours are worked in the past week. For volunteering it’s likely the question would need to relate to a longer period to capture all activity. This may cause issues with memory, but could probably be field tested. It seems unlikely that the question would be considered too intrusive or sensitive.

The main issue may be a perception of what constitutes volunteering – and the question may need some guidelines for this. This is already an issue with the question in its current form. Some parts of the community may volunteer by an objective definition, but may just consider this helping our friends or relatives and not formal volunteering, so there’s a chance that this isn’t captured. The lower rates of volunteering seen in culturally diverse communities may provide evidence of that. Some cultures don’t have a tradition of volunteering in a formal sense, but a high level of participation in the community which they may call something else.

Our respondents in the Local Government and community sectors are most concerned with this type of volunteering. Volunteering Australia states that any time given to the community is a form of volunteering – so in order to formally represent this for advocacy in a local area – a change in wording of the question may be required. In its current form the question makes this difficult. Even a change in wording of the broader current question would be welcomed by these groups. Of course this would constitute a break in series for time series data, but in its current form, if no changes are proposed, the question is of limited use and the space on the Census may be better off given to something else.

Volunteering is not controversial and is unlikely to be considered intrusive by most respondents. The main issue may be some frustration with the question, perhaps a lack of understanding by respondents about what constitutes volunteering and perhaps a higher non-response rate due to that. This could be alleviated by question wording and guidelines, but inevitably moving towards a more complex question set will lead to more difficulty here and further field testing would be required.

Questions on the frequency and hours of volunteering could be collected without significant new classifications or coding. A question on the type of organisation volunteered for would require substantial coding and possibly a new classification. It may be able to be coded to ANZSIC using an existing framework similar to the Industry topic. However it still represents another question with substantial human coding required to interpret responses. It is possible a classification different to ANZSIC with more resolution around the community sector rather than businesses might be needed. I’m not sure whether this classification exists. If not it would be substantial work to create this.

It’s worth noting that the current volunteering question has shown minimal change over time. The rate of volunteering nationally has changed within 1 percentage point (17 to 18%) over the past 10 years. So it is fairly stable and the only real use for this has been broad comparisons between areas. It’s likely however that this stability is representative of the broad nature of the question. While the overall rate of those who volunteer in the past 12 months hasn’t changed much, there may be structural change in the hours, types of organisations etc. related to the changes in the labour market and availability of leisure/free time in the population. Time series would therefore most likely continue to be useful to bring out these patterns.

However if the question remains the same as currently, it’s less clear that time series will be useful in the future, and the question could potentially be dropped or asked less frequently.

As noted by our client feedback at .id – some information is available as an alternative. Agencies such as Volunteering Australia do keep track of “formal” volunteering and a survey/admin collection from such agencies could provide substantial information on the sector. However this will still miss the “informal” sector, which is likely to make up a majority of volunteering. Such data could be collected via a survey, but there is probably no other reasonable source due to the ad-hoc nature and lack of a need to register as a volunteer in these informal settings.

Thanks for the opportunity to submit suggestions for the volunteering topic. Below is the feedback verbatim from one of our client correspondents which led to this submission.

Hi Glenn and Lailani. You may already collect data on the number of people who volunteer, if you don't please collect volunteering data. In 2012, Flinders University estimated the number of volunteers and assigned a $ benefit to the community of $290 billion. The challenge will be the type of volunteering that is done and where. If a person volunteers for a community agency such as a local op shop, that is recorded and agencies like mine that find and place volunteers keep that data. But I suspect that thousands of hours are not being recorded for the informal volunteering that occurs in a community. For example, many people "help out" at local sports clubs but don't see that as volunteering and that never gets recorded. As a result, volunteering is under reported and seen by governments as largely irrelevant and this gives them permission to not properly fund a service that is vital to communities. The Department of Social Services has indicated that it will not fund volunteering agencies beyond 2021. At South East Volunteers, we have been promoting the Volunteering Australia definition that states that any time given to a community or activity is volunteering. So it would be useful to ask if people volunteer formally or "help out" and record that as volunteering too.