Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

The Index of Relative Socio‐economic Disadvantage: general population views on indicators used to determine area‐based disadvantage

The Index of Relative Socio‐economic Disadvantage: general population views on indicators used to... Abstract Objective: To ascertain general population perceptions of the importance of indicators comprising the Index of Relative Socioeconomic Disadvantage (IRSD). Ruth Walker and J.E. Hiller Department of Public Health, University of Adelaide, South Australia Methods: Data for this study came from a face-to-face Health Omnibus survey of 3,001 residents in metropolitan and country South Australia, conducted in 2003. Results: Overall, respondents viewed the IRSD indicators as important. Of the 14 indicators, seven were seen as important by more than two-thirds of respondents (ranging from 90% perceiving the number of families with children and a low income important to 68% perceiving the number of one-parent families with dependent children as important). Younger respondents and those of lower educational attainment were more likely to perceive the indicators as unimportant, compared with older people. For example, 14% of people aged 15-24 vs. 5% of people aged 55-64 (p≤0.001) viewed the indicator ‘number of one-parent families and dependent children’ as unimportant. Conclusions: While the general population generally recognises the IRSD indicators as important measures of area-based disadvantage, there were systematic age differences in the degree to which individual indicators were deemed important. There was a general lack of support for several indicators (such as proportion of people separated/divorced, houses with no cars). Implications: This research raises the question of which factors are important in representing area-based disadvantage for young people and equally the use of this index when examining variations in the health of young Australians. (Aust N Z J Public Health 2005; 29: 442-7) he importance of the place in which a person lives in affecting their health is well established, although exactly how this occurs remains a matter of debate and speculation. Specifically, shared characteristics of an area (such as high unemployment, low income, etc.) have been argued to affect the health outcomes of residents, independent of individual characteristics.1-3 In Australia, area-based health research relies on socio-economic conditions of areas derived from national Census data, collected every five years. Four summary measures, or Socio-Economic Indexes for Areas (SEIFA), stem from these Census data and are used to measure different aspects of socio-economic conditions by geographic areas. These are: Index of Relative Socio-economic Disadvantage; Index of Relative Socioeconomic Advantage/Disadvantage; Index of Economic Resources; and Index of Education and Occupation.4 In 2001, the Australian Bureau of Statistics reviewed the indexes and methods used to construct them, which had formerly been based on “people’s experiences with other indexes, rather than referring to a theoretical model”4 (p.1). A new variable selection strategy was instigated, based on a theoretical model of disadvantage derived from the relevant literature. The Index of Relative Socio-economic Disadvantage (IRSD) is the most general of the four indexes and is obtained by measuring factors in the community that place an individual at a disadvantage, compared with someone else.5 The index comprises 20 variables that either directly measure disadvantage or reflect disadvantage. The IRSD consists of three levels of disadvantage: Level 1, or ‘core’, variables used to measure socio-economic status or a key aspect of socio-economic status (such as income, education, occupation); Level 2 variables, which are direct measures of an aspect of socioeconomic disadvantage (such as employment status, low fluency in English); and Level 3 variables, which reflect measures of disadvantage, or “signal that an area has some disadvantage”, but are not in themselves direct measures (such as Indigenous status, and being divorced/ separated)5 (p.3). The IRSD covers the whole of Australia and is designed to have a mean value across Australia of 1,000. Relatively advantaged regions have high index values (i.e. the area has a low percentage of families of low income or people with little training and in unskilled occupations) and relatively disadvantaged areas have low index values (i.e. the area has a higher percentage of lowincome families, people with relatively lower educational attainment, or people in unskilled occupations).4 As in Australia, New Zealand, Canada and the United Kingdom have developed indexes of ‘area deprivation’ (which can be likened to socio-economic disadvantage) typically acquired from national Census data.6-8 In the United Kingdom, debate has centred around the fact that measures of area deprivation do not recognise the many different domains of Submitted: October 2004 Revision requested: January 2005 Accepted: May 2005 Correspondence to: Dr Ruth Walker, Department of Public Health, Mail Drop 207, University of Adelaide, South Australia 5005. Fax: (08) 8303 6885; e-mail: ruth.walker@adelaide.edu.au AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2005 VOL. 29 NO. 5 Methods General population views on the IRSD disadvantage; for example, social exclusion (relational issues and barriers to inclusion), and officials in both England and Scotland have been keen to support new attempts to measure and identify area deprivation.9 The use of SEIFA indexes to measure socio-economic disadvantage in Australia has not been without criticism. Some perceive an over-reliance on the indexes as a broad-brush approach to measure health inequalities, with resultant loss of finer details that may underlie social and economic processes at the basis of such inequalities.10,11 Nevertheless, in much of current Australian public health research exploring the relationship between socioeconomic status and mortality, use of SEIFA indexes has become routine. Furthermore, studies have found that SEIFA indexes have different relationships with health outcomes in Australia, especially when focusing on suicide among women.12 Page and colleagues found that while male suicide was associated with all three area-based measures of SES (IRSD, Index of Economic Resources and Index of Education and Occupation), suicide differentials were less clear for females. The findings of this study clearly illustrate how SEIFA indices may not be as predictive of particular health outcomes as individual components of these indices and that the general IRSD may have limited usefulness. Table 1: Indicators comprising the Index of Relative Socio-economic Disadvantage.a The percentage of: 1. 2. 3. 4. 5. 6. 7. 8. 9. People aged 15 and over with no qualifications Families with children and having parental income less than $15,600 per year Unemployed females Unemployed males Employed males classified as ‘labourer & related workers’b Employed females classified as ‘labourer & related workers’b One-parent families with dependent children People aged 15 years and over who left school before they were 16 years old Employed males classified as ‘intermediate production & transport workers’ A further issue with the SEIFA indexes, and the IRSD in particular, is that by using ethnic composition as a measure (e.g. proportion of Indigenous residents in an area, or those of nonEnglish speaking background), the index is then unable to evaluate the relative influence of SES versus ethnic factors in contributing to area-based disadvantage. To the best of our knowledge there has been no systematic attempt in Australia to explore general population perceptions of the IRSD indicators, in terms of importance in determining whether an area is disadvantaged. As previously mentioned, these indicators have been developed by the Australian Bureau of Statistics and continue to be widely used with little scrutiny.10 This study therefore aimed to assess the use of this index, in particular whether certain groups in the general population were more or less likely to perceive the IRSD indicators as important. While SEIFA indexes measure area characteristics and can be used as proxy for individual or household SES, it should be pointed out that this study aimed to evaluate population views of the validity of the items for an area measure, not as a proxy for SES. The aims of the study were to explore: • Whether the general population perceived the IRSD indicators as important in measuring disadvantage in an area. • Which indicators were seen as more/less important. • Whether socio-demographic characteristics of the general population influenced which indicators were seen as more/less important. Methods Data for this study came from the 2003 South Australian Health Omnibus Survey.13 Structured face-to-face interviews were conducted in households throughout metropolitan and country areas of South Australia on a range of health-related topics. Sample selection The metropolitan sample was derived from collection districts (CDs), a geographic area classification used by the Australian Bureau of Statistics for Census collections. Each CD consists of approximately 220 dwellings. 5 Three hundred and forty metropolitan CDs were selected, with probability of selection proportional to size. A predetermined selection process based on a ‘skip pattern’ of every fourth household meant that 10 dwellings were chosen per CD (see footnote). The interview was conducted with one person aged 15 or over per household. When more than one eligible person resided in the household, the respondent was the last person to have his/her birthday. The country sample was derived from all cities/towns with a population size of 10,000 or more in the 2001 Census. The balance of the country sample was chosen from centres with a population of 1,000 or more in the 2001 Census, with probability proportional to size.13 One hundred CDs were included and 10 dwellings per Footnote: It should be noted that as population sizes and dwelling numbers of CDs differ, e.g. within Adelaide city from 1,070 persons to 158 persons per CD, the probability of households being selected would have varied from CD to CD. 10. Families with low incomes (less than $15,600 per year) 11. Households renting (government authority) 12. People aged 15 years and over separated or divorced 13. Dwellings with no cars 14. Employed females classified as ‘intermediate production & transport workers’b 15. People aged 15 years and over who did not go to school 16. Aboriginal or Torres Strait Islanders 17. People who do not speak fluent English 18. Employed females classified as ‘elementary clerical, sales & service workers’b 19. Occupied private dwellings with two or more families 20. Employed males classified as ‘tradespersons’b Notes: (a) Grouped according to weighting to indicate the contribution of each variable to the index. (b) Indicators excluded in the current study. 2005 VOL. 29 NO. 5 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Walker and Hiller Article Table 2: Socio-demographic characteristics of respondents (n=3,001). Characteristic Age range Age (years) 15-24 25-44 45-64 65+ Gender Male Female Country of birthd Australia UK and Ireland Other European Asian country New Zealand Other Marital status Married/de facto Never married Separated/Divorced Widowed Educational attainment Still at school Left before 15 years of age Left after 15 years of age Trade qualifications Tertiary qualifications Employment status Work part/full time Retired Home duties Student Unemployed Other Household incomea <$12,000 $12,001-$40,000 $40,001-$80,000 $80,001+ Not stated Household sizeb 1 2 3 4+ CD chosen for interview. Again, one person aged over 15 years was selected per dwelling for interview. 15-95 491 1,060 903 547 1,471 1,530 2,286 311 190 54 24 134 1,866 719 227 189 160 457 893 418 1,073 1,692 543 358 244 87 77 227 961 874 545 394 415 1,685 511 390 Questions included The Department of Public Health, University of Adelaide, included 14 questions in the omnibus survey relating to general population perceptions of the importance of the indicators that make up the Index of Relative Socio-economic Disadvantage. Participants were asked: “The Australian Government uses a range of indicators to determine the levels of disadvantage within areas where people live – could you please tell me how important you think these factors are in determining whether an area is disadvantaged”. These ‘factors’ or indicators are shown in Table 1. Participants were asked to rate the perceived importance of 14 of the original 20 indicators that make up the index. Indicators relating to specific occupation were not included. The indicators that were included related to income (n=2), education (n=3), unemployment (n=2), dwellings and living conditions (n=2), motor cars (n=1) and ‘other’ (n=4) (e.g. relating to number of Indigenous persons and those lacking fluency in English). Participants were asked to rate the perceived importance of each indicator on a five-point scale ranging from very important to very unimportant. Data analysis Data analyses were performed using SPSS (Statistical Package for the Social Sciences). Descriptive statistics were used to determine the proportion of respondents who perceived each indicator as important, and to describe the study sample. Responses were re-coded from five categories to three, i.e. ‘very important/important’ recoded to ‘important’, ‘unimportant/very unimportant’ recoded to ‘unimportant’. Chi-square tests were conducted to test for an association between perceived importance of the indicators and sociodemographic characteristics of respondents. Due to the sample size and hence likelihood of small differences being statistically significant, the level of significance was set at p≤0.001. Results From a possible total sample of 4,400 households, interviews were conducted with 3,001 individuals (70.3% response rate). At the outset, 132 households were either vacant, businesses, or did not have permanent tenants in residence. Of the remaining sample, non-response tended to be due to ‘refusal’ (716/4,268, 17%) (i.e. individual not interested, too busy, etc.) or contact not able to be established after six visits at different times of the day/evening and different days of the week (315/4,268, 7%). Table 2 presents the socio-demographic characteristics of the sample. Of the respondents, 49% were men; ages ranged from 15 to 95 years; approximately 16% were aged younger than 25 years and more than one-third were aged between 25 and 44 years of age; 35% of individuals had higher education qualifications. The sample distributions for age and education were similar to those of the general population found in the 2001 Census.14 It is evident from Table 3 that, generally, the respondents 2005 VOL. 29 NO. 5 Aboriginal or Torres Strait Islander originc No 2,252 Aboriginal 29 Torres Strait Islander 1 Don’t know 4 Notes: (a) Total annual household income of all household members before tax. (b) Including the participant, how many people aged 15 or over who lived in household. (c) n=2,286 – i.e of those who were born in Australia. (d) Country of birth figures include missing data for two respondents. AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Methods General population views on the IRSD perceived the indicators as important in determining area-based disadvantage. Three indicators, the number of rented households, people who are separated or divorced, and houses with no cars were least likely to be viewed as important. When comparing general population perceptions of the importance of each of the IRSD indicators with Australian Bureau of Statistics weightings, as could be expected, respondents perceived those indicators with higher weightings (i.e. usually those relating to income, education and employment) as more important. Table 4 presents socio-demographic characteristics of respondents that were significantly associated with perceived unimportance of the IRSD indicators, using chi square tests for association. Although perceived importance of 12 out of 14 of the IRSD indicators were significantly associated at the p≤0.001 level with respondents’ socio-demographic background, only those eight indicators with an index weighting of >20 are presented in Table 4 (i.e. those indicators which are said to contribute at a higher level to the IRSD). Overall, the socio-demographic characteristics most often significantly associated with perceived unimportance of the indicators were age and educational attainment. The younger the respondent and the lower their educational attainment, the more likely they were to perceive the indicators as unimportant. Discussion Overall, there was general agreement that the indicators used by the ABS to measure area-based socio-economic disadvantage were important. Of 14 indicators used by the ABS, seven were viewed by more than two-thirds of the general population as being important in determining area-based disadvantage. The indicators that were perceived by less than half of those surveyed as being important were: the proportion of rented households; people who are separated or divorced; and houses with no cars. Degree of importance ranged from nearly 90% of those surveyed perceiving the proportion of families with children and a low income important in determining area-based disadvantage, to 35% perceiving the proportion of houses with no cars as important. As expected, the ‘core’ variables used in the IRSD to measure areabased socio-economic status, i.e. income and education, were seen as most important. Ethnic indicators were not seen as important to respondents, with half either undecided or thinking the proportion of Aboriginal or Torres Strait Islanders was unimportant in determining an area’s level of disadvantage. As mentioned earlier, findings such as these reinforce the question of whether ethnic indicators should be included together with area or socioeconomic status measures in indexes measuring disadvantage. It could be argued then that the IRSD is not effective in evaluating the influence of SES versus ethnic influences. The most important socio-demographic predictors of perceiving the IRSD indicators as unimportant were younger age and lower educational attainment. Younger people (who in this study were defined as aged between 15-24) were significantly less likely to rate the indicators as important. These findings suggest that the indicators used by the ABS to measure area-based disadvantage are not as relevant to young people. It should be noted that although young people tended not to rate the indicators as important, we can not discount the fact that the indicators may still be relevant to the health of young Table 3: General population perceptions of the importance of IRSD indicators in determining whether an area is disadvantaged (n=3,000a). Indicators: (How important is the number of . . . in determining whether an area is disadvantaged?) n Families with children and having a low income (<$15,600 per year) Families with low incomes (<$15,600 per year) People aged 15+ who did not go to school Unemployed males People aged 15+ with no qualifications People who left school before they were 16 years old One-parent families with dependent children Unemployed females Houses with two or more families People who do not speak fluent English Aboriginal or Torres Strait Islanders Households renting (government authority) People aged 15+ who are separated or divorced Houses with no cars 2,671 2,552 2,483 2,443 2,326 2,244 2,037 1,901 1,764 1,759 1,517 1,377 1,226 1,032 Important General population perceptions Neither important Unimportant nor unimportant % n % n % 242 331 321 406 458 483 645 754 852 780 1,016 1,064 1,122 1,116 8.1 11.0 10.7 13.5 15.3 16.1 21.5 25.1 28.4 26.0 33.9 35.5 37.4 37.2 87 117 196 151 216 273 281 345 384 461 467 559 652 852 2.9 3.9 6.5 5.0 7.2 9.1 9.3 11.5 12.8 15.3 15.5 18.6 21.7 28.4 ABS weighting 0.29 0.23 0.18 0.27 0.31b 0.25 0.25 0.27 0.13b 0.15 0.18 0.22 0.19 0.19 Notes: (a) One respondent did not answer these questions, therefore n=3,000 for this table. (b) Highest and lowest weightings respectively. 2005 VOL. 29 NO. 5 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Walker and Hiller Article Australians. There is a lack of research focusing on local area perceptions of young adults (aged 15-24), with the majority of research focused on ‘young people’ being conducted with adolescents and children. Of the few studies conducted, it is suggested that social disorganisation and levels of crime in the neighbourhood are particularly pertinent to the health of young adults.15,16 A limitation inherent with research asking participants to reflect on the ‘area’ within which they or others live is that we do not know what ‘area’ means to respondents. SEIFA scores are constructed at many different spatial scales (i.e. at the Collectors District, Statistical Local Area, Local Government Area levels, etc.) and we do not therefore know exactly what spatial scale respondents may have been drawing on when deciding on the importance of the various indicators for area disadvantage. To conclude, this study adds to international research that highlights limitations associated with the use of area-based socioeconomic measures drawn from national Census data.7,9,17 In particular, this study questions the relevance of the indicators used to determine area-based disadvantage for younger people and the implications this might have for our understanding of variations in the health of young people. This survey data has highlighted the possibility of further qualitative exploration of the SEIFA indexes to provide a more in-depth analysis of respondents’ perceptions, particularly the perceptions of young adults. A further possibility is the notion of constructing a range of different area indexes compared in a multilevel context using multiple outcomes for young people, making it possible to evaluate the predictive power of the various indices. The issue raised earlier, regarding the ‘loss of finer detail’ in objective measures of area-based disadvantage,10 is central in our rationale for conducting this research. This study has clarified areas for further study if we are to improve our understanding of health inequalities. By finding out whether perceptions of IRSD indicators varied according to socio-demographic variables of respondents, we were interested in the sorts of factors that might differentiate between those who think SEIFA-type indicators are Table 4: Respondent socio-demographic characteristics associateda with perceived unimportance of IRSD indicators. IRSD indicator/ Respondent socio-demographic characteristic No. people 15+ no qualifications Educational attainment Still at school Left before 15 years of age Left after 15 years of age Trade qualifications Tertiary qualifications No. perceiving indicator as unimportant n % IRSD indicator/ Respondent socio-demographic characteristic No. of unemployed males Age 15-24 years 25-34 years 35-44 years 45-54 years 55-64 years 65+ years Marital status Never married Separated/divorced Married/de facto Widowed Educational attainment Still at school Left before 15 years of age Left after 15 years of age Trade qualifications Tertiary qualifications Employment status Work part/full time Home duties Unemployed Retired Student Other No. perceiving indicator as unimportant n % No. families with children and low income (<$15,600b) Country of birth Australia 71 UK and Ireland 9 Other 7 Educational attainment Still at school 0 Left before 15 years of age 20 Left after 15 years of age 35 Trade qualifications 15 Tertiary qualifications 17 Household incomeb <$12,000 $12,001-$40,000 $40,001-$80,000 $80,001+ Not stated No. unemployed females Age 15-24 years 25-34 years 35-44 years 45-54 years 55-64 years 65+ years 6 41 21 8 10 No. of one parent families with dependent children Age 15-24 years 68 25-34 years 80 35-44 years 62 45-54 years 42 55-64 years 19 65+ years 47 Table 4 continued next page AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2005 VOL. 29 NO. 5 Methods General population views on the IRSD relevant in measuring disadvantage and those who do not. We see this study contributing to the emerging body of knowledge about subjective indicators of the influence of place on health, in particular, in exploring what disadvantage means to young people. There are interesting complexities associated with the use of SEIFA indexes in measuring health outcomes12 and our research highlights that factors contributing to area-based disadvantage, and hence their impact on health, may hold different meanings according to age. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australian and New Zealand Journal of Public Health Wiley

The Index of Relative Socio‐economic Disadvantage: general population views on indicators used to determine area‐based disadvantage

Download PDF
6 pages

Loading next page...
 
/lp/wiley/the-index-of-relative-socio-economic-disadvantage-general-population-uwoX2vMHdx

References (19)

Publisher
Wiley
Copyright
Copyright © 2005 Wiley Subscription Services, Inc., A Wiley Company
ISSN
1326-0200
eISSN
1753-6405
DOI
10.1111/j.1467-842X.2005.tb00224.x
Publisher site
See Article on Publisher Site

Abstract

Abstract Objective: To ascertain general population perceptions of the importance of indicators comprising the Index of Relative Socioeconomic Disadvantage (IRSD). Ruth Walker and J.E. Hiller Department of Public Health, University of Adelaide, South Australia Methods: Data for this study came from a face-to-face Health Omnibus survey of 3,001 residents in metropolitan and country South Australia, conducted in 2003. Results: Overall, respondents viewed the IRSD indicators as important. Of the 14 indicators, seven were seen as important by more than two-thirds of respondents (ranging from 90% perceiving the number of families with children and a low income important to 68% perceiving the number of one-parent families with dependent children as important). Younger respondents and those of lower educational attainment were more likely to perceive the indicators as unimportant, compared with older people. For example, 14% of people aged 15-24 vs. 5% of people aged 55-64 (p≤0.001) viewed the indicator ‘number of one-parent families and dependent children’ as unimportant. Conclusions: While the general population generally recognises the IRSD indicators as important measures of area-based disadvantage, there were systematic age differences in the degree to which individual indicators were deemed important. There was a general lack of support for several indicators (such as proportion of people separated/divorced, houses with no cars). Implications: This research raises the question of which factors are important in representing area-based disadvantage for young people and equally the use of this index when examining variations in the health of young Australians. (Aust N Z J Public Health 2005; 29: 442-7) he importance of the place in which a person lives in affecting their health is well established, although exactly how this occurs remains a matter of debate and speculation. Specifically, shared characteristics of an area (such as high unemployment, low income, etc.) have been argued to affect the health outcomes of residents, independent of individual characteristics.1-3 In Australia, area-based health research relies on socio-economic conditions of areas derived from national Census data, collected every five years. Four summary measures, or Socio-Economic Indexes for Areas (SEIFA), stem from these Census data and are used to measure different aspects of socio-economic conditions by geographic areas. These are: Index of Relative Socio-economic Disadvantage; Index of Relative Socioeconomic Advantage/Disadvantage; Index of Economic Resources; and Index of Education and Occupation.4 In 2001, the Australian Bureau of Statistics reviewed the indexes and methods used to construct them, which had formerly been based on “people’s experiences with other indexes, rather than referring to a theoretical model”4 (p.1). A new variable selection strategy was instigated, based on a theoretical model of disadvantage derived from the relevant literature. The Index of Relative Socio-economic Disadvantage (IRSD) is the most general of the four indexes and is obtained by measuring factors in the community that place an individual at a disadvantage, compared with someone else.5 The index comprises 20 variables that either directly measure disadvantage or reflect disadvantage. The IRSD consists of three levels of disadvantage: Level 1, or ‘core’, variables used to measure socio-economic status or a key aspect of socio-economic status (such as income, education, occupation); Level 2 variables, which are direct measures of an aspect of socioeconomic disadvantage (such as employment status, low fluency in English); and Level 3 variables, which reflect measures of disadvantage, or “signal that an area has some disadvantage”, but are not in themselves direct measures (such as Indigenous status, and being divorced/ separated)5 (p.3). The IRSD covers the whole of Australia and is designed to have a mean value across Australia of 1,000. Relatively advantaged regions have high index values (i.e. the area has a low percentage of families of low income or people with little training and in unskilled occupations) and relatively disadvantaged areas have low index values (i.e. the area has a higher percentage of lowincome families, people with relatively lower educational attainment, or people in unskilled occupations).4 As in Australia, New Zealand, Canada and the United Kingdom have developed indexes of ‘area deprivation’ (which can be likened to socio-economic disadvantage) typically acquired from national Census data.6-8 In the United Kingdom, debate has centred around the fact that measures of area deprivation do not recognise the many different domains of Submitted: October 2004 Revision requested: January 2005 Accepted: May 2005 Correspondence to: Dr Ruth Walker, Department of Public Health, Mail Drop 207, University of Adelaide, South Australia 5005. Fax: (08) 8303 6885; e-mail: ruth.walker@adelaide.edu.au AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2005 VOL. 29 NO. 5 Methods General population views on the IRSD disadvantage; for example, social exclusion (relational issues and barriers to inclusion), and officials in both England and Scotland have been keen to support new attempts to measure and identify area deprivation.9 The use of SEIFA indexes to measure socio-economic disadvantage in Australia has not been without criticism. Some perceive an over-reliance on the indexes as a broad-brush approach to measure health inequalities, with resultant loss of finer details that may underlie social and economic processes at the basis of such inequalities.10,11 Nevertheless, in much of current Australian public health research exploring the relationship between socioeconomic status and mortality, use of SEIFA indexes has become routine. Furthermore, studies have found that SEIFA indexes have different relationships with health outcomes in Australia, especially when focusing on suicide among women.12 Page and colleagues found that while male suicide was associated with all three area-based measures of SES (IRSD, Index of Economic Resources and Index of Education and Occupation), suicide differentials were less clear for females. The findings of this study clearly illustrate how SEIFA indices may not be as predictive of particular health outcomes as individual components of these indices and that the general IRSD may have limited usefulness. Table 1: Indicators comprising the Index of Relative Socio-economic Disadvantage.a The percentage of: 1. 2. 3. 4. 5. 6. 7. 8. 9. People aged 15 and over with no qualifications Families with children and having parental income less than $15,600 per year Unemployed females Unemployed males Employed males classified as ‘labourer & related workers’b Employed females classified as ‘labourer & related workers’b One-parent families with dependent children People aged 15 years and over who left school before they were 16 years old Employed males classified as ‘intermediate production & transport workers’ A further issue with the SEIFA indexes, and the IRSD in particular, is that by using ethnic composition as a measure (e.g. proportion of Indigenous residents in an area, or those of nonEnglish speaking background), the index is then unable to evaluate the relative influence of SES versus ethnic factors in contributing to area-based disadvantage. To the best of our knowledge there has been no systematic attempt in Australia to explore general population perceptions of the IRSD indicators, in terms of importance in determining whether an area is disadvantaged. As previously mentioned, these indicators have been developed by the Australian Bureau of Statistics and continue to be widely used with little scrutiny.10 This study therefore aimed to assess the use of this index, in particular whether certain groups in the general population were more or less likely to perceive the IRSD indicators as important. While SEIFA indexes measure area characteristics and can be used as proxy for individual or household SES, it should be pointed out that this study aimed to evaluate population views of the validity of the items for an area measure, not as a proxy for SES. The aims of the study were to explore: • Whether the general population perceived the IRSD indicators as important in measuring disadvantage in an area. • Which indicators were seen as more/less important. • Whether socio-demographic characteristics of the general population influenced which indicators were seen as more/less important. Methods Data for this study came from the 2003 South Australian Health Omnibus Survey.13 Structured face-to-face interviews were conducted in households throughout metropolitan and country areas of South Australia on a range of health-related topics. Sample selection The metropolitan sample was derived from collection districts (CDs), a geographic area classification used by the Australian Bureau of Statistics for Census collections. Each CD consists of approximately 220 dwellings. 5 Three hundred and forty metropolitan CDs were selected, with probability of selection proportional to size. A predetermined selection process based on a ‘skip pattern’ of every fourth household meant that 10 dwellings were chosen per CD (see footnote). The interview was conducted with one person aged 15 or over per household. When more than one eligible person resided in the household, the respondent was the last person to have his/her birthday. The country sample was derived from all cities/towns with a population size of 10,000 or more in the 2001 Census. The balance of the country sample was chosen from centres with a population of 1,000 or more in the 2001 Census, with probability proportional to size.13 One hundred CDs were included and 10 dwellings per Footnote: It should be noted that as population sizes and dwelling numbers of CDs differ, e.g. within Adelaide city from 1,070 persons to 158 persons per CD, the probability of households being selected would have varied from CD to CD. 10. Families with low incomes (less than $15,600 per year) 11. Households renting (government authority) 12. People aged 15 years and over separated or divorced 13. Dwellings with no cars 14. Employed females classified as ‘intermediate production & transport workers’b 15. People aged 15 years and over who did not go to school 16. Aboriginal or Torres Strait Islanders 17. People who do not speak fluent English 18. Employed females classified as ‘elementary clerical, sales & service workers’b 19. Occupied private dwellings with two or more families 20. Employed males classified as ‘tradespersons’b Notes: (a) Grouped according to weighting to indicate the contribution of each variable to the index. (b) Indicators excluded in the current study. 2005 VOL. 29 NO. 5 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Walker and Hiller Article Table 2: Socio-demographic characteristics of respondents (n=3,001). Characteristic Age range Age (years) 15-24 25-44 45-64 65+ Gender Male Female Country of birthd Australia UK and Ireland Other European Asian country New Zealand Other Marital status Married/de facto Never married Separated/Divorced Widowed Educational attainment Still at school Left before 15 years of age Left after 15 years of age Trade qualifications Tertiary qualifications Employment status Work part/full time Retired Home duties Student Unemployed Other Household incomea <$12,000 $12,001-$40,000 $40,001-$80,000 $80,001+ Not stated Household sizeb 1 2 3 4+ CD chosen for interview. Again, one person aged over 15 years was selected per dwelling for interview. 15-95 491 1,060 903 547 1,471 1,530 2,286 311 190 54 24 134 1,866 719 227 189 160 457 893 418 1,073 1,692 543 358 244 87 77 227 961 874 545 394 415 1,685 511 390 Questions included The Department of Public Health, University of Adelaide, included 14 questions in the omnibus survey relating to general population perceptions of the importance of the indicators that make up the Index of Relative Socio-economic Disadvantage. Participants were asked: “The Australian Government uses a range of indicators to determine the levels of disadvantage within areas where people live – could you please tell me how important you think these factors are in determining whether an area is disadvantaged”. These ‘factors’ or indicators are shown in Table 1. Participants were asked to rate the perceived importance of 14 of the original 20 indicators that make up the index. Indicators relating to specific occupation were not included. The indicators that were included related to income (n=2), education (n=3), unemployment (n=2), dwellings and living conditions (n=2), motor cars (n=1) and ‘other’ (n=4) (e.g. relating to number of Indigenous persons and those lacking fluency in English). Participants were asked to rate the perceived importance of each indicator on a five-point scale ranging from very important to very unimportant. Data analysis Data analyses were performed using SPSS (Statistical Package for the Social Sciences). Descriptive statistics were used to determine the proportion of respondents who perceived each indicator as important, and to describe the study sample. Responses were re-coded from five categories to three, i.e. ‘very important/important’ recoded to ‘important’, ‘unimportant/very unimportant’ recoded to ‘unimportant’. Chi-square tests were conducted to test for an association between perceived importance of the indicators and sociodemographic characteristics of respondents. Due to the sample size and hence likelihood of small differences being statistically significant, the level of significance was set at p≤0.001. Results From a possible total sample of 4,400 households, interviews were conducted with 3,001 individuals (70.3% response rate). At the outset, 132 households were either vacant, businesses, or did not have permanent tenants in residence. Of the remaining sample, non-response tended to be due to ‘refusal’ (716/4,268, 17%) (i.e. individual not interested, too busy, etc.) or contact not able to be established after six visits at different times of the day/evening and different days of the week (315/4,268, 7%). Table 2 presents the socio-demographic characteristics of the sample. Of the respondents, 49% were men; ages ranged from 15 to 95 years; approximately 16% were aged younger than 25 years and more than one-third were aged between 25 and 44 years of age; 35% of individuals had higher education qualifications. The sample distributions for age and education were similar to those of the general population found in the 2001 Census.14 It is evident from Table 3 that, generally, the respondents 2005 VOL. 29 NO. 5 Aboriginal or Torres Strait Islander originc No 2,252 Aboriginal 29 Torres Strait Islander 1 Don’t know 4 Notes: (a) Total annual household income of all household members before tax. (b) Including the participant, how many people aged 15 or over who lived in household. (c) n=2,286 – i.e of those who were born in Australia. (d) Country of birth figures include missing data for two respondents. AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Methods General population views on the IRSD perceived the indicators as important in determining area-based disadvantage. Three indicators, the number of rented households, people who are separated or divorced, and houses with no cars were least likely to be viewed as important. When comparing general population perceptions of the importance of each of the IRSD indicators with Australian Bureau of Statistics weightings, as could be expected, respondents perceived those indicators with higher weightings (i.e. usually those relating to income, education and employment) as more important. Table 4 presents socio-demographic characteristics of respondents that were significantly associated with perceived unimportance of the IRSD indicators, using chi square tests for association. Although perceived importance of 12 out of 14 of the IRSD indicators were significantly associated at the p≤0.001 level with respondents’ socio-demographic background, only those eight indicators with an index weighting of >20 are presented in Table 4 (i.e. those indicators which are said to contribute at a higher level to the IRSD). Overall, the socio-demographic characteristics most often significantly associated with perceived unimportance of the indicators were age and educational attainment. The younger the respondent and the lower their educational attainment, the more likely they were to perceive the indicators as unimportant. Discussion Overall, there was general agreement that the indicators used by the ABS to measure area-based socio-economic disadvantage were important. Of 14 indicators used by the ABS, seven were viewed by more than two-thirds of the general population as being important in determining area-based disadvantage. The indicators that were perceived by less than half of those surveyed as being important were: the proportion of rented households; people who are separated or divorced; and houses with no cars. Degree of importance ranged from nearly 90% of those surveyed perceiving the proportion of families with children and a low income important in determining area-based disadvantage, to 35% perceiving the proportion of houses with no cars as important. As expected, the ‘core’ variables used in the IRSD to measure areabased socio-economic status, i.e. income and education, were seen as most important. Ethnic indicators were not seen as important to respondents, with half either undecided or thinking the proportion of Aboriginal or Torres Strait Islanders was unimportant in determining an area’s level of disadvantage. As mentioned earlier, findings such as these reinforce the question of whether ethnic indicators should be included together with area or socioeconomic status measures in indexes measuring disadvantage. It could be argued then that the IRSD is not effective in evaluating the influence of SES versus ethnic influences. The most important socio-demographic predictors of perceiving the IRSD indicators as unimportant were younger age and lower educational attainment. Younger people (who in this study were defined as aged between 15-24) were significantly less likely to rate the indicators as important. These findings suggest that the indicators used by the ABS to measure area-based disadvantage are not as relevant to young people. It should be noted that although young people tended not to rate the indicators as important, we can not discount the fact that the indicators may still be relevant to the health of young Table 3: General population perceptions of the importance of IRSD indicators in determining whether an area is disadvantaged (n=3,000a). Indicators: (How important is the number of . . . in determining whether an area is disadvantaged?) n Families with children and having a low income (<$15,600 per year) Families with low incomes (<$15,600 per year) People aged 15+ who did not go to school Unemployed males People aged 15+ with no qualifications People who left school before they were 16 years old One-parent families with dependent children Unemployed females Houses with two or more families People who do not speak fluent English Aboriginal or Torres Strait Islanders Households renting (government authority) People aged 15+ who are separated or divorced Houses with no cars 2,671 2,552 2,483 2,443 2,326 2,244 2,037 1,901 1,764 1,759 1,517 1,377 1,226 1,032 Important General population perceptions Neither important Unimportant nor unimportant % n % n % 242 331 321 406 458 483 645 754 852 780 1,016 1,064 1,122 1,116 8.1 11.0 10.7 13.5 15.3 16.1 21.5 25.1 28.4 26.0 33.9 35.5 37.4 37.2 87 117 196 151 216 273 281 345 384 461 467 559 652 852 2.9 3.9 6.5 5.0 7.2 9.1 9.3 11.5 12.8 15.3 15.5 18.6 21.7 28.4 ABS weighting 0.29 0.23 0.18 0.27 0.31b 0.25 0.25 0.27 0.13b 0.15 0.18 0.22 0.19 0.19 Notes: (a) One respondent did not answer these questions, therefore n=3,000 for this table. (b) Highest and lowest weightings respectively. 2005 VOL. 29 NO. 5 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Walker and Hiller Article Australians. There is a lack of research focusing on local area perceptions of young adults (aged 15-24), with the majority of research focused on ‘young people’ being conducted with adolescents and children. Of the few studies conducted, it is suggested that social disorganisation and levels of crime in the neighbourhood are particularly pertinent to the health of young adults.15,16 A limitation inherent with research asking participants to reflect on the ‘area’ within which they or others live is that we do not know what ‘area’ means to respondents. SEIFA scores are constructed at many different spatial scales (i.e. at the Collectors District, Statistical Local Area, Local Government Area levels, etc.) and we do not therefore know exactly what spatial scale respondents may have been drawing on when deciding on the importance of the various indicators for area disadvantage. To conclude, this study adds to international research that highlights limitations associated with the use of area-based socioeconomic measures drawn from national Census data.7,9,17 In particular, this study questions the relevance of the indicators used to determine area-based disadvantage for younger people and the implications this might have for our understanding of variations in the health of young people. This survey data has highlighted the possibility of further qualitative exploration of the SEIFA indexes to provide a more in-depth analysis of respondents’ perceptions, particularly the perceptions of young adults. A further possibility is the notion of constructing a range of different area indexes compared in a multilevel context using multiple outcomes for young people, making it possible to evaluate the predictive power of the various indices. The issue raised earlier, regarding the ‘loss of finer detail’ in objective measures of area-based disadvantage,10 is central in our rationale for conducting this research. This study has clarified areas for further study if we are to improve our understanding of health inequalities. By finding out whether perceptions of IRSD indicators varied according to socio-demographic variables of respondents, we were interested in the sorts of factors that might differentiate between those who think SEIFA-type indicators are Table 4: Respondent socio-demographic characteristics associateda with perceived unimportance of IRSD indicators. IRSD indicator/ Respondent socio-demographic characteristic No. people 15+ no qualifications Educational attainment Still at school Left before 15 years of age Left after 15 years of age Trade qualifications Tertiary qualifications No. perceiving indicator as unimportant n % IRSD indicator/ Respondent socio-demographic characteristic No. of unemployed males Age 15-24 years 25-34 years 35-44 years 45-54 years 55-64 years 65+ years Marital status Never married Separated/divorced Married/de facto Widowed Educational attainment Still at school Left before 15 years of age Left after 15 years of age Trade qualifications Tertiary qualifications Employment status Work part/full time Home duties Unemployed Retired Student Other No. perceiving indicator as unimportant n % No. families with children and low income (<$15,600b) Country of birth Australia 71 UK and Ireland 9 Other 7 Educational attainment Still at school 0 Left before 15 years of age 20 Left after 15 years of age 35 Trade qualifications 15 Tertiary qualifications 17 Household incomeb <$12,000 $12,001-$40,000 $40,001-$80,000 $80,001+ Not stated No. unemployed females Age 15-24 years 25-34 years 35-44 years 45-54 years 55-64 years 65+ years 6 41 21 8 10 No. of one parent families with dependent children Age 15-24 years 68 25-34 years 80 35-44 years 62 45-54 years 42 55-64 years 19 65+ years 47 Table 4 continued next page AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2005 VOL. 29 NO. 5 Methods General population views on the IRSD relevant in measuring disadvantage and those who do not. We see this study contributing to the emerging body of knowledge about subjective indicators of the influence of place on health, in particular, in exploring what disadvantage means to young people. There are interesting complexities associated with the use of SEIFA indexes in measuring health outcomes12 and our research highlights that factors contributing to area-based disadvantage, and hence their impact on health, may hold different meanings according to age.

Journal

Australian and New Zealand Journal of Public Health – Wiley

Published: Oct 1, 2005

There are no references for this article.