Review of 2021 Census Topics

This is a consolidated submission from three Australian Government agencies (the Department of Jobs and Small Business (DJSB), Comcare, and the Fair Work Commission or FWC) in the Jobs and Innovation Portfolio.

There are three main areas where we have suggested improvements in the Census under the the ‘income and work’ topic, as follows.

Separating wage income from total income

In this Submission, we recommend that a question be added to the 2021 Census form to distinguish wage (including salary and supplement) income from non-wage income. This would make Census information much more useful to economists and labour market researchers, particularly those analysing wage developments at the unit-record, local or regional levels.

The current collection of individual income is relatively broad. These data are likely to provide more benefit if it included a breakdown of components, such as wages and salaries, and allowed respondents to provide a specific dollar amount rather than ranges, which would allow for additional analysis such as the calculation of averages.

Work participation
Summary
Through this submission, we seek to include the collection of data that relates to work participation. This information is intended to inform better targeted policy and program design to support work participation, particularly for those with a physical or mental health condition or illness.

This information will provide valuable evidence for private and not-for-profit organisations, and government portfolios, such as employment, social services and health, to address these issues.

Proposed inclusion in Census
At a high level, we seek to identify the cohorts and characteristics of those with a temporary or permanent physical or mental health condition or illness, and to measure the extent to which this affects their ability to work.

As such, we suggest the following be incorporated into the Census:
• whether an individual has experienced a temporary or permanent physical or mental health condition or illness – if so, what type of condition or illness, and
• whether this condition or illness has affected the individual’s ability to work, and if so:
o the type of impact, and/or
o the type of income support being received.

We acknowledge that there are a number of work and health-related questions within the 2016 Census that could be modified to capture this information. However, a working example of how this information could be collected based on the structure of the 2016 Census would be the inclusion of three new questions and adjustments to two existing questions. These new questions and adjustments are noted below in red.

X1 In the last year, has the person had any of the following:
(include definitions) o Short-term health condition (lasting less than six months)
o Long-term health condition (lasting six months or more)
o Disability (lasting six or more)
o No health conditions  Go to XX
X2 Does your health condition result in any of the following: o Permanent restriction from work
o Short term restriction from work
o Limited work (e.g. shorter hours)
o No impact with appropriate supports (e.g. flexible arrangements)
o No impact
o Other

Question 23 (2016 Census) to be adjusted to read:
23 What were the reasons for the need for assistance or supervision shown in Questions 20, 21 and 22? o No need for help or supervision
o Health condition or disability (shown in Question X1)
o Old or young age
o Difficulty with English language
o Other cause

Question 34 (2016 Census) to be adjusted to read:
34 Last week, did the person have a job of any kind?
(need to include definition of paid leave or income support) o Yes, worked for payment or profit
o Yes, but absent on holidays, on paid leave, on strike, or temporarily stood down
o Yes, but was on paid leave or income support  Go to X3
o Yes, unpaid work in a family business  Go to 38
o Yes, other unpaid work  Go to 46
o No, did not have a job, but was on paid leave or income support  Go to X3
o No, did not have a job  Go to 46
Note: the following question should sit above Question 46 (from 2016 Census)
X3 Which of the following best describes the paid leave or income support you were on last week
(definitions may need to be included) o Employer provided entitlement (e.g. sick leave)
o Social security payment
o Workers’ compensation payment
o Life insurance payment
o Superannuation payment
o Motor vehicle accident compensation payment
o Other

 

Evidence of work participation
There is significant and compelling evidence to support, and general agreement about, the importance of work participation, including the Australian and New Zealand Consensus Statement on the Health Benefits of Work released by the Australasian Faculty of Occupational and Environmental Medicine first released in 2011. This consensus statement presents compelling international and Australasian evidence that work is generally good for health and wellbeing, and that long-term work absence, work disability and unemployment generally have a negative impact on health and wellbeing.

As such, it is important to understand the reasons for, and characteristics of, individuals who are unable to participate in the workforce.

Value of work participation
In 2018, the Collaborative Partnership (see the “Any further comments” Section at the end of this submission for an explanation of this Partnership) commissioned research by Monash University that identified that a total of $37.2 billion was spent on income support in one financial year (2015-16), through a complex mix of government systems, private sector insurers and employers:
• Employer provided entitlements, e.g. sick leave (6.5 million recipients, $18.7 billion expenditure)
• Social security (469,000 recipients, $8.6 billion expenditure)
• Workers’ compensation (156,000, $2.5 billion expenditure)
• Life insurance (95,000 recipients, $4.4 billion expenditure)
• Motor vehicle accident (6,000 recipients).

This information would provide important evidence to authorities and agencies that are responsible for supporting work participation, health, employment, social services and welfare, workers’ compensation, life insurance and superannuation. In addition to influencing government and policy, this information will also enable employers and industry to address barriers and issues within the workplace to support work participation that can be targeted accordingly.

Understanding the barriers to work participation, particularly regarding health conditions that impact on an individual’s ability to work is an important priority for the Australian government and community at large. Addressing ways to support Australians to participate in work is important particularly in the context of other trends such as the economic impacts of an ageing population, the future of work (expected changes to job design and availability), and increasing prevalence of mental illness (for example, 1 in 5 Australians will experience in their lifetime a diagnosable mental health condition).

Method of setting pay
The information that we seek is required for research and to assist with monitoring outcomes related to wage setting and other workplace relations arrangements.
The primary benefit of the Census is the ability to obtain information at a fine level of detail. In particular, agencies in the Jobs and Innovation portfolio require information at the most detailed level of industry and occupation. This information can be used as a proxy for modern award coverage and is cross-classified by gender, age, hours worked, labour force status and educational attainment.
A question on method of setting pay (modern award, collective agreement, individual arrangement) would assist in better understanding a wide range of characteristics on employees that receive an award rate of pay, compared with other employees. This would specifically assist with the Fair Work Commission’s Annual Wage Review and four- yearly review of modern awards.
Additional data items that would be of interest include:
• Number of jobs held; and
• Whether the respondent worked in the last year, to obtain some measure of long-term unemployment.
We also recognise that there would be benefits in linking the Census to other household surveys, such as the Labour Force Survey, and to continue to promote the longitudinal dataset format by increasing the sample size.

Separating wage income from total income
Currently the ABS Census specifically ask each respondent ‘What is the total of all income the person usually receives?’ The ABS defines Total Income as ‘Total (or gross) income, is the sum of income from all sources, whether monetary or in kind, before income tax, the Medicare Levy and the Medicare Levy surcharge are deducted.’

As such, the information collected includes all sources of income such as wages, government allowances and pensions, investment income and non-cash receipts such as goods and services, use of a vehicle or subsidised housing. The separation of these sources of income into wage (wages and government supplements) and non-wage income (investment, non-cash) would provide a clearer picture of the distribution of wage earnings across the nation.

The data collected on wage and non-wage income would be incredibly useful to economists and labour market analysts to understand better the changes in and distribution of wages over time both at an aggregate and regional level. It would also provide an opportunity for further research on those who receive income at or around the national minimum wage (NMW), which would inform policy decisions about setting this minimum wage.

Separate information on wage and total income would help to inform labour market research on this topic and help shape future policy development, better target Government funding, program planning and education and training, through identifying areas of wage disparity and identifying those at risk of poorer socio-economic outcomes and reliance on Government subsidies.

Work participation
As noted in the submission summary above, work participation and barriers to participation are priorities for the Australian government and broader community due to the financial and social impacts of poor participation and poor health/social outcomes. Understanding and targeting of issues in these areas has impacts on the requirements noted above (i.e. for electoral, legislative, policy, program and research purposes).

Method of setting pay
The collection of data on method of setting pay (award rate of pay, collective agreement, individual arrangement) would assist with the Fair Work Commission’s Annual Wage Review and 4 yearly review of modern awards. The Expert Panel for annual wage reviews is required to conduct each annual wage review within the legislative framework of the Fair Work Act 2009 (Cth). Acquiring these statistics based on a population would help provide a better understanding on the proportion of employees that are affected by the Commission’s decisions and to expand the amount and type of research that could be undertaken to assist the Panel’s deliberations. In particular, the data would provide a better understanding of the groups of employees that are more likely to receive award rates of pay. The collection of this information would also assist analysts to conduct more detailed analysis of the effects of methods of setting pay on wages, incomes, labour productivity and other outcomes.

Separating wage income from total income
Currently ABS statistics on wages are only disaggregated down to the state and territory level, for example, median weekly and median hourly earnings in main job from the ABS release, Characteristics of Employment (Cat. No.6333.0); or at best, the SA4 level.

Data at the national level which can be disaggregated down to regional levels, will help to identify areas of disadvantage including wage income of those with low levels of total income, education, employment, remote areas, Indigenous Australians and those aged over sixty five years.

In addition, information on employees earning at or around the National Minimum Wage rate is limited. If the collection of individual income included a breakdown of components and allowed respondents to provide specific dollar amounts, this would allow a much more comprehensive analysis of these types of workers to be undertaken. Particularly as the number of workers earning the National Minimum Wage rate is likely to be small, this collection would benefit from a larger survey sample to enable some analysis at a disaggregated level.

Work participation
Additional information on work participation from the 2021 Census would enable analysts and policy developers to leverage other data items collected in the Census (e.g. geography, age, area of employment/skill) to enable targeted policy and program consideration.

Method of setting pay
A data item on ‘method of setting pay’ would be useful in estimating the proportion of groups of employees reliant on award rates of pay based on the whole population. This data could be combined with information on household composition and income.

Separating wage income from total income
Census Respondents are already asked to provide information on their total income. Our proposal is simply to add another question on wage and salary income. This would then enable non-wage income to be calculated by subtracting wage and salary income from total income.

Total income information already collected in the census is easily obtained from many written or online sources such as employer group certificates and other investment activity that is collected for taxation purposes.

A possible version of the question which could be added is “What is the total of wage and salary income the person usually receives?” There would also be a need to include a small amount of explanatory description of the definition of wage and salary income, such as: ‘Wage and salary income - Income paid directly to you by your employer or wages you receive from your own or family business or partnership’. However, expert survey question designers in the ABS may be able to suggest a more appropriate wording of this question and explanatory description.

Work participation
The suggested changes in this submission (see introductory Section) have been considered using simple language that would be easy and quick to consider and answer. However, the ABS might be able to suggest other ways that could represent an improvement.

Method of setting pay
The ABS (or alternative surveys) already applies methods in which these data are collected and can be understood by respondents through the Survey of Employee Earnings and Hours

Separating wage income from total income
The question is not likely to be considered intrusive, offensive or controversial and should be relatively straightforward for respondents to answer, as it simply disaggregates the income information question already included in the Census.

Work participation
The questions and responses noted in this submission seeking details regarding limitations and income sources are of similar vein to those that were in the 2016 Census, and are unlikely to be unacceptable to respondents.

Method of setting pay
These topics are no more intrusive than that is currently collected in the Census and respondents are likely to be willing and able to answer accurately.

Separating wage income from total income
With the additional question, the ABS will ask the respondents to separate their total income and provide a total for wage and salary income, which will be accompanied with a brief definition explaining the criteria for defining this amount (suggested amendments were provided against assessment criteria 3). This will require an additional code to distinguish between the two different income types, as such; this addition will not require extensive processing. As the total income question is already asked during the current Census the data collection required in a subset question to separate total income from wage and salary income would be minimal.

Work participation
Though the submission includes potentially one or two new question/answer fields, it is unlikely that this information would be difficult to code, or require significant instruction or explanation.

Method of setting pay
There may be some difficulty in accurately collecting this data. Employees are less likely to know their method of setting pay than their employers who have access to payroll records. However, research on methods of setting pay using the Household, Income and Labour Dynamics in Australia (HILDA) survey has shown that the difference in coverage of employees reliant on awards is not too different to that collected by the ABS from employers. However, both of these two existing resources are limited by either their sample size or the amount of variables that can also be analysed.

The ABS already has instructions on how to answer questions on methods of setting pay (from its Employee Earnings and Hours survey) that can be applied to the Census.

Separating wage income from total income
Time series data that distinguishes wage (including salary) income from total income at the local and regional level would be necessary to see the changes over time to wages at the local level and for disadvantaged groups in the population such as Indigenous Australians. This would be used for analytical purposes such as assessing wage inequality for disadvantaged groups and wage and salary differences among regions of Australia. In turn, this could be used to improve policy development and evaluation, including of wage-setting arrangements, place-based policies to get unemployed people into work and educational and training policies and programs.

Work participation
It is likely that this information will be required in future Censuses as there will be an ongoing need to capture work participation trends and information relating to barriers. The topic of work participation has been, and will continue, being a significant policy issue for the Australian population, particularly in the context of an ageing population, rise in mental health issues, and future of work design.

Method of setting pay
The topics discussed will continue to remain relevant for future annual wage reviews. A longitudinal dataset would better enable these issues to be analysed over time to assess developments in wage-setting and incomes.

Separating wage income from total income
The Household Income and Labour Dynamics survey (HILDA) collects information on weekly earnings (typically gross) which can then be disaggregated into hourly earnings using information collected on the usual weekly hours of work of all those employed in the household, resulting in ‘current usual earnings per hour worked’. HILDA is also broken down into major urban (population 100,000+), non-major urban (population between 1,000 and 99,999) and non-urban regions (populations less than 1,000 and regional and remote areas). However, the sample size is relatively small with only around 17,000 households participating in the survey, which can produce high sampling errors. Partly due to its relatively small sample size, HILDA can tend to under-represent certain groups such as Indigenous Australians, with its limited reach into regional areas, and recent migrants, which become less recent as they continue through the survey and may not necessarily be replaced in the next wave. Unlike the Cenus, HILDA survey participants can have high rates of dropping out where they are then replaced by a household with similar (but not the same) attributes. This could provide inconsistencies in future responses compared with the responses from the previous household making a consistent time series difficult.

As mentioned previously, the ABS’s own Characteristics of Employment survey does collect information on wages before tax, however the data is not disaggregated beyond the State and Territory level, or SA4 using microdata, and cannot be cross-tabulated with at-risk groups unlike Census data.

Work participation
The Census presents the only opportunity for an Australia-wide snapshot of work participation trends and issues. Though smaller studies can be, and have been, undertaken, there are limited opportunities to capture this information and data from a national perspective.

Method of setting pay
While most of this data is collected in other ABS surveys, the amount of analysis that can be undertaken is limited by the sample sizes of these surveys and there continues to remain scope for further research that could be considered if these topics are added to the Census and expanding the types of disaggregation.

The Collaborative Partnership is a national effort by public, private and not-for-profit organisations to improve work participation for people with temporary or permanent physical or mental health conditions which may impact their ability to work. The Collaborative Partnership is focused on aligning systems and services in workers’ compensation, life insurance, superannuation, disability support and employment services.

Comcare is currently the Chair of the Collaborative Partnership which includes:
• Department of Jobs and Small Business
• Department of Social Services
• Insurance Council of Australia
• Employers Mutual Limited (EML)
• Royal Australasian College of Physicians
• Australasian Faculty of Occupational and Environmental Medicine
• Australian Council of Trade Unions.