A Comparison of Volunteering Rates from the 2006 Census of Population and Housing and the 2006 General Social Survey

The Voluntary Work module in the 2006 General Social Survey (GSS) estimated that about one in three Australian adults were volunteers, while the 2006 Census of Population and Housing reported one in five adults as being volunteers. This paper investigates the nature of the differences in measuring voluntary work between these two collections to explain differences in the voluntary work rates. It also uses Census and GSS data to identify characteristics associated with being more likely to volunteer.

This paper has found that the personal interview approach of the GSS provides a better quality estimate of the rate of volunteering compared with the single question in the Census. This is due to a number of reasons: the person answers on behalf of themselves and thus should know whether they volunteer; the detailed questions in the GSS are designed to elicit an accurate response; and the survey also allows for prompting and clarification by the interviewer which cannot be done in a self-completed Census form.

Using descriptive and multivariate statistical techniques to address the second objective, the paper compares a wide range of social and demographic characteristics of adults living in Australia. It finds that adults with the following characteristics are more likely to volunteer:

• Adults living in a family with a co-resident dependent child;
• Adults with a higher level of educational attainment; and
• Adults proficient in English.

Both collections were consistent in the direction of effects on volunteering for a broad set of socio-demographic variables. However, differences in the strength of these effects were observed of around 1.5 to 2 times, with the rates of volunteering higher in the GSS. The largest differences were found for proficiency in spoken English, family composition and highest level of educational attainment. The GSS provides a better quality estimate of the rate of volunteering, however analysis in this paper shows the Census can be used to understand differences between geographical areas, including for small areas. Although the absolute level of volunteering will be understated in the Census, differences between areas can still be observed.

Volunteers make a valuable contribution to society in both economic and social terms. They provide services which may otherwise have to be paid for or left undone. This allows organisations to allocate their often limited finances elsewhere. The value of the work contributed by volunteers to non-profit institutions in 2006–2007 was estimated to be $14.6 billion (ABS, 2009b).

The Australian Bureau of Statistics (ABS) defines a volunteer as ‘someone who willingly gives unpaid help, in the form of time, service or skills, through an organisation or group’ (ABS, 2006a). The focus on 'help willingly given’ is a key feature of this definition. It excludes activities done: to qualify for government benefits; under a Community Service Order; as part of a student placement; or as emergency work during an industrial dispute. It also excludes any activity which is done as part of paid employment or for a family business. This definition of a volunteer includes a wide range of people who provide unpaid help through organisations in their community, but excludes ‘direct volunteering’ where people willingly provide help, but not through an organisation (e.g. helping a neighbour move house).

In 2006, the ABS collected information on voluntary work in the Census of Population and Housing and three sample surveys: the General Social Survey (GSS), the Adult Literacy and Lifeskills Survey (ALLS) and the Time Use Survey (TUS). These surveys collected information on voluntary work undertaken within Australia for an organisation such as a community service provider, charity, sporting club or school. Each of the collections had a different purpose and used somewhat different methods to collect volunteer information. Despite these differences, the three surveys provided fairly similar estimates of the participation of adults aged 18 years and over in voluntary work activities, at around 35% (excluding adults in very remote areas). In contrast, the Census estimate for the same population was much lower at just over 20%.

The GSS is the primary source of detailed information on voluntary work in Australia. Because of this, and because it has the largest sample size of all sample surveys which collected information on volunteering in 2006, the GSS was chosen to compare to the Census in order to investigate reasons why the Census estimate differs considerably from the survey estimates.

This paper investigates the disparity in results from the two collections, what the socio-demographic characteristics associated with volunteering are, and how Census data on volunteering can be used for small area estimates. To do this, the paper briefly reviews literature on volunteering, describes differences in the collection methodology between the GSS and Census and presents findings of statistical analysis. The analysis comprises descriptive univariate statistics on the characteristics of volunteers, small areas analysis, and the results of a multivariate logistic regression.

There are numerous theories to explain who engages in voluntary work. The Dominant Status model originally developed by Lemon et al. (1972) is frequently cited. According to this model, participation in voluntary activity is associated with a ‘dominant status’ position in society. That is, people who are highly educated and who have a high status job are likely to have a higher propensity of volunteering than people who do not (Lemon et al., 1972). Bales (1996) also comments that volunteering is associated with higher levels of education, income, and belonging to the dominant ethnic group. In Australia, level of educational attainment has also been found to be positively associated with the propensity to volunteer (Evans and Kelly, 2000). This association has been found in a number of studies (for example, Goss, 1999; Zwart and Perez, 1999; Davis Smith, 1998). However, while education level is associated with the propensity to volunteer, it may not be a determinant among people who commit a large amount of time to volunteering. Lyons and Hocking (2000) reported that among highly committed volunteers (those volunteering more than 300 hours a year) there was no difference in education level.

Other research has shown that people volunteer for activities that benefit their children, in areas such as sports teams or scouting (Smith, 1994). Furthermore, some researchers argue that motivational aspects (Clary and Snyder, 1991) and church attendance (Evans and Kelley 2000) have strong associations with volunteering, however this theory could not be tested in this analysis because similar information about religion and church participation were not available from both the GSS and the Census.

Analysis already published by the ABS is consistent with the themes of this research. People who volunteer generally have higher levels of education and higher incomes than those who do not volunteer. There are high rates of volunteering among people who are employed and parents with school-aged children. People in rural areas volunteer at slightly higher rates than people in major urban areas. The organisations that attract the highest number of volunteers include sports clubs, educational institutions and community/welfare organisations. While the rate of volunteering had risen over the ten or so years to 2006, the average number of hours people volunteered had fallen in 2006 (ABS 2006a; 2008).

As with collecting many social statistics, collecting information on volunteering is not a straightforward task. Different people have different perceptions of volunteering. Rooney et al. (2004) reported that longer, more detailed prompts led respondents to recall volunteering at higher incident rates than questions with fewer prompts. Therefore, it is very important to adequately define and explain the term ‘volunteer’ in a way that is easily understood in a consistent way by respondents. People may not identify themselves as volunteers even though they perform voluntary activities. Does a parent coaching their child’s sporting team or helping out at their child’s school think of themselves as a volunteer? International research suggests not, unless questions are appropriately structured (Bailie, 2006; Rooney et al., 2004).

There is a need to consider whether the ABS definition of a volunteer can be conveyed to respondents in ways that elicit accurate and consistent responses. Research has shown that prompting respondents with examples of volunteer activities results in more accurate identification by respondents of volunteering activities and leads to higher estimates of volunteer participation (Bailie, 2006; Rooney et al., 2004).

The name of the organisation or type of work may be useful to prompt people to respond appropriately to questions about voluntary work. Some activities are commonly regarded as voluntary work, such as volunteering for a well-known community organisation such as the Red Cross or Lifeline. Other activities, such as assisting people in one’s local church community, may be regarded as part of their church commitment rather than formal voluntary work, even though it encompasses help willingly given through an organisation, and therefore falls within the scope of the definition. Alternatively, people may not see this as different to informal assistance within their community.

The proficiency in English of respondents also has an impact on the identification of volunteers, especially when collection instruments in Australia are presented in English. People who do not speak or read English well may receive help from a trained interviewer or a member of their family in a survey setting; however they need to actively seek help when completing their Census form. The ABS offers translation services to assist people to complete the Census, but respondents may choose not to take advantage of these services. There is a wide body of research into the effects of language proficiency on data quality (Bowling, 2005) and response patterns (Harzing, 2006; McCrae, 2002; Gibbons et al., 1999; Church et al., 1988). The results show that survey concepts are more refined in the minds of respondents when the person is allowed to respond in his or her own language. It is expected that there would be a similar impact on questions on volunteering.

The GSS is the primary source of information on voluntary work in Australia. This survey collects information about the household and about one person 18 years and over in the household in non-very remote areas. It collects information on many different topics, such as living arrangements, health, education, work, income, housing, and social participation. Information collected on participation in voluntary work is collected directly from the respondent and includes: time spent; frequency; number and types of activities; whether people are reimbursed; and the organisation for whom they volunteer.

Prior to the inclusion of the voluntary work module in the GSS in 2002 and 2006, similar modules were included in the 2000 Population Survey Monitor and a supplementary survey to the June 1995 Labour Force Survey.

The voluntary work questions in the GSS are carefully designed to help respondents correctly identify whether or not they volunteered in the last 12 months and to collect information about any voluntary work undertaken. An adult aged 18 years or over is selected from those people living in the household to participate in a personal interview. In the voluntary work module, a trained interviewer presents the randomly selected adult with a list of the types of organisations for which they may have volunteered (not a list of the types of voluntary activity). After establishing that a person participated in activities that might be voluntary work, they are asked a series of questions to ensure that these activities meet the definition of voluntary work. Selected questions from the survey are included in Appendix A and more details can be found in the ABS publication General Social Survey: User Guide, 2006 (ABS, 2006b).

The Census of Population and Housing is the largest statistical collection undertaken by the ABS. Conducted every five years, the Census measures the number of people in Australia on Census night, and collects key information about their personal and household characteristics and the dwellings in which they live. The Census allows data collected to be used for analysis of small areas to a lower level than cannot be achieved from a sample survey.

A single question on voluntary work was included for the first time in 2006 (ABS, 2006c). It was collected for all persons 15 years and over. In the Census, households were provided (along with the Census form) with a separate Census Household Guide to help them complete the Census. The guide provided information on the importance of each question and included a ‘How to answer’ section to help people respond accurately. For the 2006 Census, the guide provided examples of activities, such as assisting at organised events and with sports organisations; or helping with organised school events and activities, which could be included as voluntary work. However, the use of the guide is up to each individual. For those households that completed the Census on-line, this information was provided electronically with the question. The Census question on voluntary work and information from the guide are included in Appendix A.

The difference in volunteer rates between the GSS and the Census may be partly explained by the differences in collection methods and questions. These are covered in more detail below.

While non-response is a factor when comparing the GSS and Census data, its impact on participation in voluntary work cannot easily be quantified. There are two dimensions of non-response in each collection the first due to the person or household not participating in the collection at all, and the second due to the person participating in the collection but failing to answer the voluntary work questions.

Of the dwellings selected in the 2006 GSS, 13.5% did not respond fully or adequately to the survey. As the non-response to the GSS was low for a sample survey, the impact of non-response bias is considered to be negligible (ABS, 2006a).

For the Census, the 2006 Post Enumeration Survey (PES), which is used to estimate the underenumeration in the Census, estimated that 2.7% of the population did not respond to the Census.

In addition to the underenumeration in the Census identified in the PES, the non-response to the voluntary work question in the 2006 Census was 2.3% for people 18 years and over (restricted to the common population between GSS and Census). As found for other Census questions, levels of non-response were higher for older people (4.2% for people aged 60 years and over compared with 1.8% for people aged 18 to 60 years). The level of non-response to the voluntary work question overall in the Census was similar to that of many other Census questions. For further details see the Census Data Quality Statement on Voluntary Work for an Organisation or Group at www.abs.gov.au.

Research by Statistics Canada suggests that people who do not respond to telephone based survey questions on voluntary work actually have a lower rate of voluntary activity than people who respond (Hall et al., 2006). Similar factors may affect responses in the Census. This would be expected to increase the voluntary work rate in the Census compared to the GSS but is not reflected in the volunteering rates. A similar issue may also affect the accuracy of the volunteering rate in GSS, but it doesn’t explain the differences between the GSS and the Census.

The following descriptive analysis compares the profile of volunteers between the Census and GSS, including rates of volunteering by selected demographics. In Section 4.1 potential variables for describing the profile of volunteers are outlined. In Section 4.2 the profile of volunteers from the two collections are examined and compared to the general population. Finally in Section 4.3 the volunteering rate is presented along with the rate ratio between the two data sources for various socio-demographic characteristics.

For the following analyses, records with any missing values on variables of interest were removed and Census was restricted to the population common to the GSS. This allowed for consistency in analyses throughout the paper. Because records with missing values on any variable of interest were removed, the estimates in this paper may not match those published in other ABS publications.

Variables were selected that were identified in the literature as correlating with volunteerism and were available in both collections. As a result, the analysis did not include religion or church participation. Country of birth was not included in the analysis because language proficiency was highly correlated with country of birth so only language proficiency was used.

Groupings within variables were selected to enable comparison of differences between the two collections. For example, four age group categories were used to provide a generational view and look at people in stages of the life course. The four categories selected were people aged: 18-29 years; 30-39 years; 40-59 years; and 60 years and over. Research shows that members of different generations differ from each other in attitudinal, educational and behavioural characteristics (ABS, 2005; ABS, 2009a; Gainsford, 2005). Members of younger generations are often highly educated, and people aged 18-29 in 2006 were more likely than other generations to have started adult life with credit cards and mobile phones. While each generation shares certain characteristics, it should be acknowledged that within each generation a great deal of variety occurs.

Respondents’ family composition was categorised into three categories-families with children aged under 15 years; families with dependent children aged 15-24 years; and families with no dependent children. The literature on volunteering suggests that a person from a family which has at least one dependent child is more likely to volunteer than a person who lives with a family where there are no dependent children. For this reason, single people, and people living in group houses were included in the ‘family with no dependent children’ category.

The proficiency in spoken English variable refers to the main language spoken at home. For those whose main language spoken at home wass English, they are in the category ‘English only’. For those who spoke a different main language at home, they are in the groups ‘Speaks English well or very well’ or ‘Speaks English not well or not at all’.

The highest educational attainment variable has been coded into three categories for the purposes of this analysis. These three education categories are: those with Advanced Diploma/Diploma or above (e.g. Advanced Diploma, Bachelor Degree etc.), those with Year 12 or Certificate III/IV, and those with Year 11 or below (Certificates I/II and Certificates not further defined are included in Year 11 or below).

Four employment categories were used. The categories used in this analysis separate those employed full-time, those employed part-time, those who were unemployed and those not in the labour force. In this analysis of volunteering, we analysed the incidence of volunteering, not the frequency or the duration of volunteering. Some of the effects of employment status might have had a different result if frequency or duration of volunteering could be used as opposed to the simple incidence rate. For example, those with more time available, such as those who are not in the labour force may volunteer for more hours than those who are employed.

Overall, the profile of volunteers by selected socio-demographic characteristics from the two collections shows some differences in their characteristics, albeit relatively small (table 4.1). Comparing the characteristics of volunteers from the GSS and Census, we see that a greater proportion of volunteers were female in the Census than in the GSS (57.2% compared with 53.9%). These proportions were higher than the proportion of the Australian population that was female (51.2%) indicating that females were more likely to volunteer than males.

Volunteers were only slightly different in their age structure between the two collections. There were higher proportions of volunteers who were in the middle age groups in the GSS and higher proportions of volunteers in the older age groups in the Census.

The Census showed proportionally more volunteers were outside the capital cities compared with the GSS collection (41.4% compared with 39.2%). Both were greater than the proportion of the total population outside capital cities (34.4%). The Census also showed large differences across highest level of educational attainment, with a high proportion of volunteers having Advanced Diploma/Diploma qualifications or above. This was expected as educational attainment has been shown to be a positive predictor of volunteering (Evans and Kelly, 2000).

There were differences between volunteer profiles and the population profile, with the profile of volunteers including a higher proportion of people with high educational attainment, who are married (registered or social) and who are in a family with children under 15 years.

Rates of volunteering by selected social and demographic characteristics are presented in table 4.2. The measure of volunteering obtained from 2006 GSS (34.2%) was 1.7 times that measured in the 2006 Census (20.1%). This is displayed in the rate ratio column of table 4.2. As discussed previously, differences in collection methodology, particularly the impact of proxy reporting, the use of prompting techniques and multiple questions, may impact greatly on the discrepancy between these two measures of volunteering.

Table 4.2 shows that volunteer rates in the Census were consistently 1.5 to 2 times lower than the GSS across the range of socio-demographics. Exceptions to this are where people spoke English well or very well (2.1), or not well or not at all (2.9); those aged 60 years and over (1.3); and those who were unemployed (the rate was not significantly different from the Census rate).

In both collections, participation in voluntary work was associated with education level and employment status. People with Advanced Diploma/Diploma or above reported a higher rate of volunteering than people with Year 12 or Certificate III/IV qualifications. Likewise, people with Year 12 or Certificate III/IV qualifications reported a higher rate of volunteering than those with Year 11 or below. Those who were employed part-time reported the highest rates of volunteering across employment status.

Proportionally fewer people from capital cities reported volunteering compared with people who lived outside the capital cities.

The rate ratio between the GSS and Census was relatively low for the oldest age group (1.3), which could be due to some of the issues described above under Collecting information on volunteers. For example, a high proportion of people in the oldest age category live in single, or two person households and are therefore likely to have lower rates of proxy reporting. It could also be that people in the oldest age group have a lower rate ratio due to them having a greater proportion of non-respondents who are less likely to volunteer (Hall et al., 2006).

Looking at the rate ratios, it can be seen that for many characteristics, the GSS volunteer rate is 1.5 to 2 times that of the Census rate. This reflects the ratio of 1.7 found for the overall volunteer rate. This result indicates that, in general, the estimates from the Census were consistently lower for all socio-demographic characteristics compared with those from the GSS.

It appears that English language proficiency has a large impact on reporting voluntary work activities in the Census relative to the GSS. The rate ratio was higher for all people who speak a main language other than English at home, but was particularly high for people who speak English not well or not at all (with the highest rate ratio as 2.9). Some explanations for this may include the use of a translator (trained interviewer or family member) in GSS, the usefulness of more prompting in conveying voluntary work, and a higher non-response rate to the question on voluntary work on the Census (3.6% for people who speak English not well or not at all compared with 3.0% for people who speak English very well or well and 2.1% for people who speak English only). The rate ratio was lowest for those who were unemployed (GSS rate not significantly different from the Census rate), people not in the labour force (1.5) and people aged 60 years and over (1.3).

Table 4.3 presents rates of volunteering by state and territory as well as state and territory by capital city and rest of state.

The state and territory rates for GSS show that the volunteering rate was consistently between 1.4 to 1.9 times greater in GSS than Census. Exceptions to this were South Australia rest of state (1.1 times) and for Perth (2.1 times), although the rate for South Australia was not significantly different between the GSS and Census.

The Census also showed lower rates of volunteering in each capital city than the GSS, between 1.5 (South Australia) and 2.1 (Western Australia) times. The rates for the rest of State were also lower than the GSS, ranging from 1.4 (Northern Territory) to 1.8 (Queensland) times. The exception to this was the rest of State for South Australia for which the rates were not significantly different between collections.

Factors associated with volunteering may be inter-related. For example, age structure and family composition may affect the volunteer rates of people living in rural or urban areas. A multivariate analysis helps to understand these inter-relationships by looking at effects while holding all other factors constant.

For this multivariate analysis GSS data was chosen. This is because the characteristics of volunteers were similar between the two datasets, and because the GSS provides more accurate estimates for voluntary work. The descriptive analysis in tables 4.1 and 4.2 show that characteristics such as sex, education, employment and proficiency in spoken English are all associated with volunteering. However, these characteristics may also be associated with each other. In order to investigate the individual effects of each factor on volunteering, independent of the other factors, logistic regression has been used. This determines the effect of individual characteristics (such as sex, age, education, employment etc.) on volunteering while holding the effects of all the other variables constant.

Logistic regression is a type of multivariate analysis. By assigning a value of 1 if the person volunteered and 0 otherwise, a binary dependent variable is created that is suitable for logistic regression modelling. The results of this type of modelling are easier to interpret than other non-linear techniques (such as probit modelling). The listing of logistic coefficients and model diagnostic statistics used in this analysis are in Appendix C and detailed technical information on the logistic model is in Appendix D.

The odds ratios obtained from the logistic regression analysis are presented in table 5.1. As the name suggests, odds ratios are the ratio of two odds. The odds ratio for a female is obtained by dividing the odds of being female in the volunteering group by the odds of being male in the volunteering group. Any variable that has an odds ratio of greater than 1.0 indicates an increased propensity to volunteer, while a variable with an odds ratio of less than 1.0 has a reduced propensity to volunteer. If the confidence interval for an odds ratio includes the value 1.0, then we consider the estimated odds ratio not to be significantly different from 1.0 (i.e. the resulting propensity to volunteer based on that characteristic is not significant).

In the GSS model, the odds ratio for a female is 1.28. Since the confidence interval surrounding this estimate (1.18–1.38) does not include the value 1.0, it can be asserted that, at a 95% significance level, being female increases the propensity to volunteer, all other variables being held constant. Another way to put this is that the odds ratio result shows the effect of being female increases the odds of volunteering by 28% over males, while controlling for the effects of all other variables (age, family composition, marital status, proficiency in spoken English, education level, employment status and area of residence).

The size of the odds ratio indicates which variables have the strongest association with volunteering. However, since the odds ratios are not linear, it is not possible to make relative comparisons between characteristics. For example, an odds ratio of 2.42 for having an Advanced Diploma/Diploma qualification or above does not mean these people are 1.6 times more likely to volunteer than people whose highest qualification was Year 12 or a Certificate III or IV (odds ratio 1.52).

The variables that contribute most strongly to the propensity to volunteer are educational attainment and family composition, based on the model. Having progressively higher levels of educational attainment contributed markedly to the propensity to volunteer compared with those whose highest level of educational attainment was Year 11 or below. The results also show that respondents from a family where there was a dependent child under the age of 15 years had twice the odds of volunteering than a person from a family with no dependent children (the base category).

People who do not speak English well, or do not speak English at all, had a lower propensity to volunteer than people who speak English, even after controlling for variables such as education and employment status.
Being employed part-time increased the propensity to volunteer over being employed full-time. In addition, as described above, women had a higher propensity to volunteer than did men.

Finally, the results indicate that people who lived outside the capital cities had an increased propensity to volunteer compared with their capital city counterparts. Similar results have been observed in Canada, where both formal and direct personal volunteering were found to be highest in intermediate urban, small town and rural communities (Reed and Selbee, 2001).

Although the GSS is the best source for data on volunteering, the sample size prevents analysis of voluntary work for small geographic areas below capital city / balance of state (see Table 4.3 for rates). Differences in socio-demographics between communities may also result in differences in voluntary work rates between communities and using the Census can be useful to understand volunteering at small areas. Even though the Census cannot be used to estimate absolute levels of volunteering in different areas, it could be used to identify where there are differences between areas.

Statistical Local Areas (SLAs) within Sydney were chosen for comparison of characteristics. Table 6.1 compares volunteer rates for people in Sydney across selected characteristics.

Due to the sampling error for GSS estimates for Sydney, rate ratios were not found to be significantly different from each other, therefore differences in volunteering between different demographic groups cannot be identified at this level of geographic area. The exception is for the unemployed, which was significantly lower than those employed or not in the labour force (the rate for GSS and Census for unemployed was not significantly different).

Although differences within groups cannot be identified, the estimates for the rate ratios tend to lie within the bounds described above under the Descriptive analysis (i.e. the GSS volunteer rate is 1.5 to 2.0 times larger than the Census volunteer rate). This suggests the Census can be used to compare differences in the propensity to volunteer at lower geographic regions.

Differences in rates between SLAs were consistent with differences in the predictors of volunteering that were previously described above under the Results from the logistic regression. Therefore, differences between SLAs in volunteering rates can be compared with reasonable confidence, even though the absolute level of volunteering will be understated within SLAs.

Table 6.2 presents the volunteering rates for the eight SLAs within Sydney – those with the four lowest and four highest volunteering rates. All SLAs are presented in Appendix E.

Fairfield East has the lowest volunteer rate at 7.9% followed by Bankstown North-East (9.0%). Ku-ring-gai has the highest rate (29.2%), followed by the Blue Mountains (27.1%). This compared to 16.8% for all of Sydney.

Table 6.3 presents the population profiles for selected SLAs to allow for comparison to groups of interest.

The most notable difference for these SLAs can be seen in proficiency in spoken English. The Blue Mountains had the highest percentage of people who spoke English only (95.0%), with Ku-ring-gai (80.8%) being close to the total adult Australian population (82.3%). In contrast, Bankstown North-East (23.0%) and Fairfield East (25.4%) had much lower percentages of people who spoke English only.

Another large difference between SLAs can be seen across highest level of educational attainment. The largest percentage of people who had completed an Advanced Diploma/Diploma or above could be seen in Ku-ring-gai (60.6%), followed by the Blue Mountains (38.3%). These two SLAs were above the average for Australia (28.2%). Conversely, Bankstown North-East and Fairfield East had below average percentages of people who had completed an Advanced Diploma/Diploma qualifications or above (21.3% and 14.7%).

Both proficiency in spoken English and highest level of educational attainment were found to be strong predictors of volunteering above in the Results from the logistic regression. Family composition was another strong predictor; however the variations between the selected SLAs for family type were much smaller than those for proficiency in spoken English and highest level of educational attainment.

Bankstown North East and Fairfield East had greater percentages of people in family compositions (such as those with dependent children under 15 years) that are more likely to volunteer compared with the Blue Mountains and Ku-ring-gai.

Differences were also found in other demographic characteristics that predict volunteering, namely age, social marital status and part-time employment status. However, while these differences were smaller than for proficiency in spoken English and highest level of educational attainment, they were in the direction that would be expected. For example, being employed part-time was a positive predictor of volunteering and a greater percentage of people were found to be employed part-time in the selected SLAs with higher volunteer rates compared with the Australian average and the selected SLAs with lower volunteer rates.

Tables 6.4 and 6.5 show that for SLA comparison the Census can be used to compare differences in the propensity to volunteer between SLAs where there are large differences in volunteer rates. These results suggest that the propensity to volunteer involves interactions between a number of the demographic and socio-economic characteristics identified as strong predictors of volunteering.

Analysis of the 2006 GSS and the 2006 Census data showed that the overall difference in volunteer rates of 1.5 to 2 times was relatively consistent across a wide range of characteristics including age, sex, social marital status, family composition and educational attainment. The difference between GSS and Census volunteer rates was more pronounced for who those whose main language spoken at home was not English, particularly those who do not speak English well, or not at all.

Overall the profiles of volunteers obtained from the two collections were similar and the logistic regression on the GSS yielded similar results. Factors associated with volunteering included age, sex, educational attainment, proficiency in spoken English, family composition and employment status. These results are consistent with previous research on volunteering (Zappala and Burrell, 2001; Reed and Selbee; 2001; Smith 1994; Evans and Kelly; 2000; Lyons and Hocking; 2000, Cox 2000).

It has been shown through analysis of lone person households that the difference between Census and GSS estimates for voluntary workers is not solely due to proxy reporting. The difference may also be due to personal interview approach, using more detailed questions and prompting in the GSS. This finding is also consistent with previous research (Bailie, 2006; Rooney et al., 2004).

Some groups are more likely to be identified as volunteers in the Census, such as people who are highly educated. The results of the Census small area SLA comparison showed that variations in the Census voluntary work rates are consistent with demographics of the population of the small areas and in line with the predictors of voluntary work, although the volunteering rates will be understated in the Census.

This leads to the question of how the two data sources can be used to provide information on volunteers. In most cases the high quality and detailed information contained in the GSS will be most appropriate for the needs of users, especially for estimates at the national and state level. The GSS has an extensive range of social and economic data items available for use in the analysis of volunteers.

As a survey, however, the GSS is subject to sampling error which can be large if the number of observations for a particular characteristic is small. Therefore, the Census may be suitable for analysis on voluntary work for differences between small areas or for small population groups. However given the relative and consistently lower voluntary work rates this should be used with caution. To provide context and accuracy, users could analyse Census results for voluntary work in conjunction with other information that might be known about the area or group under study, for example residents’ educational profile or the number of families with young children. This combined information could be useful for: planning; directing support services to volunteers in local areas; targeting publicity to encourage people to become volunteers; and the development of a local area indicator for the strength of communities or social inclusion.

This release previously used catalogue number 4441.0.55.002.