Personal Income of Migrants, Australia methodology

1 The statistics in this publication relate to persons aged 15 years and over who have migrated to Australia under a permanent or provisional visa with an arrival date between 1 January 2000 and 6 September 2016. This release contains new estimates on the sources of personal income that migrants received in the 2013-14 financial year. The tables provide a breakdown of total personal income by the following sources - Employee income, Own unincorporated business income, Investment income, Other income (excluding Government pensions and allowances) and Total income (excluding Government pensions and allowances).

2 The statistics in this publication were compiled from the 2013-14 Personal Income Tax and Migrants Integrated Dataset (PITMID) produced by the Personal Income Tax and Migrants Integrated Dataset Project. The project used a purely deterministic linking strategy to combine the Australian Tax Office (ATO) PIT 100% data file for the 2013-14 financial year with an extract of permanent migrant settlement records from the Department of Immigration and Border Protection (DIBP). Estimates for the 2009-10, 2010-11 and 2011-12 financial years have been released previously.

3 For further information on the Project see Research Paper: Feasibility Study of Linking Migrant Settlement Records to Personal Income Tax Data, Aug 2014 (cat. no. 1351.0.55.051), Research Paper: Personal Income Tax and Migrants Integrated Dataset (PITMID) 2011-12 Quality Assessment, October 2016 (cat. no. 1351.0.55.060) and the project entry on the Public Register of Data Integration Projects on the National Statistical Service (NSS) website.

4 The Permanent Migrant Data (PMD), formally known as the Settlement Database (SDB), is an extract of administrative data compiled by the Australian Government from various departmental systems and a number of external sources, including the Department of Immigration and Border Protection (DIBP) and Department of Human Services (Medicare Australia). The Department of Social Services (DSS) has custodianship of the data and provide permission for the ABS to receive the data directly from DIBP. The PMD is a consolidated database of information on persons who have been granted a permanent or provisional/temporary visa. The PMD generally excludes temporary visa holders.

5 Previous iterations of this project used an extract of the PMD that included migrants who had arrived in Australia between 1 January 2000 and 6 March 2013. Starting with the 2012-13 PITMID, a new extract of the PMD was used that included migrants who had arrived in Australia between 1 January 2000 and 6 September 2016. The number of records on the new extract rose from approximately 1.8 million to 2.8 million permanent migrants.

6 The Personal Income Tax (PIT) data is sourced from the Australian Taxation Office (ATO).

7 These results are based, in part on tax data supplied by the ATO to the ABS under the Taxation Administration Act 1953, which requires that such data is only used for the purpose of administering the Census and Statistics Act 1905. Any discussion of data limitations or weaknesses is in the context of using the data for statistical purposes, and is not related to the ability of the data to support the ATO's core operational requirements.

8 Legislative requirements to ensure privacy and secrecy of this data have been adhered to. In accordance with the Census and Statistics Act 1905, the results have been confidentialised to ensure that they are not likely to enable identification of a particular person or organisation.

9 The data has been collected in compliance with Australian taxation laws. The unit record data was provided to the ABS for a variety of statistical purposes and so was not tailored specifically to this project. The unit record PIT dataset contains a range of key data items such as income and tax deductions. It also contains auxiliary socio-demographic data items such as age, sex and birth year. Information on the statistics contained in the dataset is generally available through the ATO website or via a combination of data dictionary and tax return form information.

10 Data provided to the ABS by the ATO are from taxation returns processed up to 16 months after the end of the financial year (i.e. returns processed up to 31 October 2015 for the financial year ending 30 June 2014). Due to the identifying nature of the data it contains, access to all ATO datasets is strictly regulated by the ATO. Both the ATO and the ABS handle personal information contained in the data in accordance with the Australian Privacy Principles contained in the Privacy Act 1988.

11 According to taxation laws, individuals whose income is below a certain threshold are not required to submit tax returns. However, amendments to the taxation laws can significantly alter the information that is required to be reported in the personal income tax returns and statistics derived from the PIT dataset will be influenced by tax regulation changes. Since the 2012-13 financial year, the tax-free threshold has been $18,200, after rising from $6,000. This may lead to a slight over-inflation of median incomes as those with incomes below this value are not included.

12 The main functions and responsibilities of the ATO are to administer taxation legislation and to collect a wide variety of taxes. The ATO therefore collects data from its reporting population as part of its processes to calculate income tax liability for those persons who are required to lodge an income tax return.

13 The ATO database covers all individuals who submit an individual income tax return and includes persons with income from one or more of a range of sources, such as employee income, own unincorporated businesses, superannuation and annuities, investments and some Government pensions, benefits or allowances.

14 The data items requested from the ATO and presented in this release are those needed to create the ABS Income Standards.

15 For the purposes of providing statistical measures for the entire population, the ATO database has some limits to its coverage. Persons who receive an income below certain levels are not necessarily required to lodge a tax return. This can include persons who derive their income from government pensions and allowances. Consequently, the coverage of low income earners, including people receiving government pensions and allowances, is incomplete in ATO records. In addition, some Commonwealth of Australia Government pension, benefit and allowance payments are exempt from income tax and are therefore not required to be included in tax returns. As such, the ATO data should be regarded as an indicative though not complete picture of all income earners in Australia.

16 Due to the tax exempt status of many Government pensions and allowances, information on these income sources are under-reported on the PIT dataset. For this reason, Government pensions and allowances are not included in the calculation of Total income in this publication. Methods for including more complete Government pensions and allowances information sourced from different DSS datasets are being considered for future iterations of this project.

17 All data presented are gross income before deductions - and before tax and the Medicare levy are applied. Data are for migrant taxpayers aged 15 years and over.

18 The 2013-14 PITMID is restricted to people on the PMD who linked to a PIT record. The PMD records on the extract were limited to those with a date of arrival between 1 January 2000 and 6 September 2016 (inclusive) and the following exclusions also apply;

• Non-visa settlers (e.g. some New Zealand citizens who have migrated to Australia)
• Deceased persons
• Migrants on Temporary visas, such as Students and Temporary work (Skilled) visas (these visa holders are not on the PMD).
   

19 The PMD date of arrival on which the scope is based reflects an individual's latest arrival pertaining to their latest permanent visa. For visas that were granted onshore (i.e. in Australia), the Arrival Date refers to the latest date of arrival prior to the grant of that visa. For Settlement visas that were granted offshore (i.e. outside of Australia), the Arrival Date refers to the first date of arrival after the grant of that visa.

20 Of the 2,789,810 PMD records, 1,674,211 records (60%) were linked to a PIT record. After obtaining the PIT records for the 2013-14 dataset, there were 1,486,808 records on the 2013-14 PITMID.

21 The main reasons for not linking an PMD record to a PIT record are:

• There is no corresponding PIT record; or
• The data on either the PMD record or the corresponding PIT record is of insufficient quality to link.
   

22 Analysis of the PMD and the PIT datasets prior to linking indicated that they were both of a sufficiently high quality to link if a corresponding PIT record existed. However, there are a number of valid reasons why an PMD record would not have a corresponding PIT record. There are many legitimate reasons for there not being a corresponding PIT record as not every person would be required to submit a tax return. These include but are not restricted to a person:

• not having submitted a tax return by the time of extraction of information;
• not being in paid employment or earning an income in the reference period;
• living in Australia on a visa stream that places restrictions on employment;
• not being in Australia during the reference period;
• having an income that is within the tax free threshold; and
• not having any tax withheld.
   

23 As such, it was not expected that the linkage rate for this project would reach 100% for the entire population on the PMD. There has been no weighting or adjustment applied to the dataset but due to the high match rate, bias is anticipated to be small. For more information, see the "Reliability of estimates" section below.

24 While the 2013-14 release of the PITMID data has been designed to be as similar as possible to previous releases, there are a few changes which have been outlined below.

25 Records with income values in 2013-14 that were considered vastly different to income amounts reported in previous years were assessed and, in some cases, deleted.

26 In the 2009-10 data, an outlier was identified for the Employee income of a Humanitarian migrant. This outlier was not present in the 2010-11 data and as such, the slight decrease in average Employee income for Humanitarian migrants from 2009-10 to 2010-11 should not be considered a true representation of the change in income.

27 Publication tables include some data for migrants holding Provisional/temporary visas. In 2009-10, Provisional/temporary visa populations were included in totals only.

28 Employee income includes the main forms of payments made to employees for their work or services. Employee income, as reported on the income tax return, includes:

• Q1-CDEFG Total income from wage and salary (before tax and application of Medicare levy) as shown on the 'PAYG payment summary - individual non-business';
• Q2-K Allowances, earnings, tips, director's fees, etc;
• Q3-H Employer lump sum payments - 5% of amount B;
• Q3-R Employer lump sum payments - amount A;
• Q4-I Employment termination payments;
• Q9-O Attributed personal services income;
• Q12-B Employee share schemes;
• Q20-N Exempt foreign employment income;
• Q20-T Other net foreign employment income;
• Q20-U Net foreign employment income - Payment summary
• IT1-W Total Reportable fringe benefits; and
• IT2-T Reportable employer superannuation contributions.
   

29 Employee income includes employer social contributions. However, income for many components of social contributions are not reported on the PIT dataset and hence are not included in the calculation of Employee income for this publication.

30 Reportable employer superannuation contributions include contributions for both employees and own unincorporated business owners. However, the proportion of people reporting as own unincorporated business owners is considered to be sufficiently small for the gross amount to be included as employee income.

31 Other net foreign employment income was separately identified by the ATO on the 2009-10 tax form for individuals (Item Q20-T) for the first time. Items Q20-N, Q20-T and Q20-U on the tax return are included as foreign employment income. Amounts in T are for foreign employment income where an Individual PAYG form has not been issued by the foreign employer.

32 Amounts in Q20-N are for foreign employment income where tax exemptions apply (e.g. employment income for some defence and police work). Where the foreign employment income is 100% tax exempt, it will not appear on the tax return or the supplementary form. Where the foreign employment income is not 100% tax exempt, the amounts reported in Q20-N are taxable amounts, that is gross amounts less deductions.

33 Own unincorporated business income includes the following data items on the individual income tax return:

• Q13-NL Distributions from partnerships and trusts for primary production activities;
• Q13-O Distributions from partnerships for non-primary production activities;
• Q14-A Net personal services income; and
• Q15-BC Net income (or loss) from business.
   

34 "Net personal services income" does not include income a person received as an employee, making it different from "Attributed personal services income".

35 Investment income includes the following data items on the individual income tax returns:

• Q10-L Gross interest;
• Q11-S Dividends unfranked amount;
• Q11-T Dividends franked amount;
• Q11-U Dividends franking credit;
• Q13-C Franked distribution from trusts - Non-primary production;
• Q13-U Distribution from trusts less net capital gains and foreign income - non-primary production;
• Q20-F Australian franking credits from a New Zealand company;
• Q20-R Net foreign rent; and
• Q21-P (less FQU) Net rent.
   

36 Superannuation and annuity income includes the following data items on the individual income tax returns:

• Q7-JN Australian annuities and superannuation income streams;
• Q7-YZ Australian annuities and superannuation income streams - lump sum in arrears;
• Q8-QP Australian superannuation lump sum payments; and
• Q22-W Bonuses from life insurance companies and friendly societies.
   

37 A change to legislation relating to superannuation, taking effect from 1 July 2007, means that people aged 60 years and over who receive superannuation income in the form of a lump sum or income stream (such as a pension) from a taxed source, will receive that income tax free. Therefore, if a person has no other income, or their total income is below the tax-free threshold, or any tax payable is mitigated by a tax offset (such as Senior Australian Tax Offset), then this person will not be required to lodge a tax return. Therefore, there is a known undercount for this data item.

38 Methods for including more complete Superannuation and annuities income from additional datasets are being considered for future iterations of this project.

39 Due to the low response rates to this income source, Superannuation and annuities income was included in Other income.

40 Other income (excluding Government pensions and allowances) includes the following data items on the individual income tax returns:

• Q19-B Foreign entities - Transferor or trust income;
• Q19-K Foreign entities - Controlled foreign company income;
• Q20-LD Foreign source income - Net foreign pension or annuity;
• Q20-M Foreign source income - Other net foreign source income; and
• Q24-VY Other income.
   

41 Due to the tax exempt status of many Government pensions and allowances, information on these income sources are under-reported on the PIT dataset. For this reason, Government pensions and allowances are not included in the calculation of Total income in this publication. Methods for including more complete Government pensions and allowances information sourced from different DSS datasets are being considered for future iterations of this project.

42 Total income (excluding Government pensions and allowances) is calculated by summing the values of Employee income, Own unincorporated business income, Investment income, Superannuation and annuities, and Other income (excluding Government pensions and allowances).

43 Given the migrant focus of this publication, a non-standard income source for income from foreign sources was calculated.

44 Income from foreign sources include the following data items on the individual income taxation return:

• Q19-B Foreign entities - Transferor or trust income;
• Q19-K Foreign entities - Controlled foreign company income;
• Q20-U Foreign source income - Net foreign employment income payment summary;
• Q20-N Exempt foreign employment income;
• Q20-T Other net foreign employment source income;
• Q20-R Net foreign rent;
• Q20-LD Foreign source income - Net foreign pension/annuity income; and
• Q20-M Foreign source income - Other net foreign source income.
   

45 For an individual, the taxable income is the amount remaining after deducting from assessable income all deductions allowed under the Income Tax Assessment Act for that year. It is the amount to which tax rates are applied.

46 The Australian taxpayer population is ranked in ascending order according to the Taxable income or loss reported on their 2013-14 income tax return (ITR), and then divided into ten equal groups (i.e. deciles), each comprising 10% of the population. Extreme values have not been excluded from the individual income tax data and therefore contributes to medians and decile cut offs . Individual taxation return lodgers whose age or taxable income or loss was unknown were excluded from the calculation of decile boundaries.

Note that the deciles are based on all 2013-14 ITR lodgments and differ from the ATO Taxation statistics percentile table which is based on Taxable individuals only. Taxable individuals only include those with a net tax amount greater than zero.

47 Individuals may receive income from a number of sources. Net income from a specific source may be positive or negative. For example, an individual may have positive net income from Employee income but negative net income from Investment. The number of individuals for each income source includes all persons with either positive or negative net income from that source.

48 Readers should note that individuals can receive income from more than one source. The total number of persons receiving total income cannot be calculated as the sum of the individual income source components. For example, an individual could derive income from Employee income, Investment and their Own unincorporated business and therefore contribute to the person count in all three income categories. However, they would only be counted once in Total income.

49 There are a few data consideration that users should be aware of when interpreting or analysing the statistics.

50 Generally, the ATO considers someone to be an Australian resident for tax purposes if they:

• have either always lived in Australia or have come to Australia to live permanently;
• have been in Australia for more than half of the financial year (unless their usual home is overseas and they don't intend to live in Australia);
• have been in Australia continuously for six months or more and for most of that time have been in the one job and living in the same place; or
• are an overseas student enrolled in a course of study for more than six months duration.
   

51 Percentage changes in the annual income received by migrant taxpayers from 2009-10 to 2013-14 are in 2013-14 dollars i.e. have been adjusted by the Consumer Price Index (CPI) to enable comparison 'in real terms'.

52 In order to represent values from previous years in 2013-14 dollars, income amounts were multiplied by the following values.

53 A job is determined by an Individual Pay-As-You-Go (PAYG) record linked to an ABN record. A person with more than one PAYG link to an ABN is considered as having more than one job in the financial year. A person with more than one job in the financial year can have held these jobs consecutively, concurrently or both. The presence of "Period start date" and "Period end date" on the 2013-14 PAYG dataset enables the identification of those records who held more than one job at the same point, i.e. multiple job holders.

54 Where there is a job splitting arrangement in place for two employees performing the one job, this would be counted as two jobs based on the Individual PAYG rather than one. Given this definition and the available tax data, there will be an over-count for the number of jobs.

55 The ABS uses an economic statistics model on the Australian Business Register (ABR) to describe the characteristics of a business and the structural relationship between related businesses. Within large and diverse business groups, the units model is used to define reporting units that can provide data to the ABS at suitable levels of detail.

56 In mid 2002, the ABS commenced sourcing its register information from the ABR and at that time changed its business register to a two population model. The two populations comprise what is called the Profiled Population and the Non-Profiled Population. The main distinction between businesses in the two populations relates to the complexity of the business structure and the degree of intervention required to reflect the business structure for statistical purposes.

57 The majority of businesses included on the ABR are in the Non-Profiled Population. Most of these businesses are understood to have simple structures. For these businesses, the ABS is able to use the Australian Business Number (ABN) as the basis for a statistical unit. One ABN equates to one statistical unit.

58 For a small number of businesses, the ABN unit is not suitable for ABS economic statistics purposes and the ABS maintains its own units structure through direct contact with businesses. These businesses constitute the Profiled Population. This population consists typically of large or complex groups of businesses. For more information on how the ABS codes these businesses, refer to the Explanatory Notes of publication Australian Industry, 2013-14 (cat. no. 8155.0).

59 The data presented in this publication were compiled before the processing of all income tax returns for any given year may have been completed. Data provided to the ABS by the ATO are from returns processed up to 31 October, 16 months after the end of the financial year. Any returns lodged after this date are not included. Therefore for 2013-14, returns processed after 31 October 2015 are not included. This enables a consistent basis for comparison across financial years.

60 Due to the late lodgement of tax returns in any tax reference year, the data provided in this report under-estimates the total taxable income for any given financial year.

61 Annual revised data is published by the ATO in the Taxation Statistics publication (Individuals' Tax, Table 1) for selected income items. One of these items is 'Salary and Wages'. Although this data item is different to the data contained in this release (as it does not include all the items for Employee income listed in paragraph 23), it can be used to give an indication of the likely direction of change in the number of 'Salary and Wages' earners and total 'Salary and Wages' income as more tax returns are lodged.

62 As an example, Table 8 below shows that for the 2012-13 income year, an additional 4.3% of taxpayers earning income from 'Salary and Wages' (included in Employee income) lodged their income tax returns in the two years after the initial processing cut off of 31 October 2014. This translated to a further 4.2% of 'Salary and Wages' income being reported, two years onwards.

63 The ATO provides information annually in Taxation Statistics on their website about changes that may affect taxation statistics. Changes relating to personal income tax are in each edition of Taxation Statistics.

64 For the 2012-13 financial year, the tax-free threshold increased from $6,000 to $18,200. This may have an impact on the change in median income between 2011-12 and 2012-13.

65 In accordance with the Census and Statistics Act 1905, all published estimates are subjected to a confidentiality process before release. This process is undertaken to minimise the risk of identifying particular individuals, families, households or dwellings in aggregate statistics, through analysis of published data.

66 To minimise the risk of identifying individuals in aggregate statistics, a technique is used to randomly adjust cell values. This technique is called perturbation. Perturbation involves small random adjustment of the statistics and is considered the most satisfactory technique for avoiding the release of identifiable statistics while maximising the range of information that can be released. These adjustments have a negligible impact on the underlying pattern of the statistics. After perturbation, a given published cell value will be consistent across all tables. However, adding up cell values to derive a total will not necessarily give the same result as published totals.

67 The introduction of perturbation in publications ensures that these statistics are consistent with statistics released via services such as TableBuilder. Caution should be exercised by users when deducing that there are nil people in an area with certain types of income. In general, no reliance should be placed on table cells with small values. Due to limitations of systems used to produce the publication, the median and average values in the data cubes have not been perturbed.

68 Error in estimates produced using the Personal Income Tax and Migrants Integrated Datasets may occur due to false links, missed links and the non-random distribution of unlinked records.

69 In order to assess the success of the linking, the following subpopulation of linked records was identified as those most likely to have submitted a tax return:

• Aged 15-64 years;
• Skilled, Family or Humanitarian visa stream;
• Permanent residence start date was prior to the start of the reference period; and
• TRIPS departure information showed they did not permanently leave Australia prior to the reference period.
   

70 This subpopulation of linked records were assessed against secondary sources. The Estimated Resident Population (ERP) as at 30 June 2012 and the number of tax returns submitted according to the 2011-12 Taxation Statistics publication were used to estimate the match rate. In 2011-12, the match rate was estimated to be 100.5%. This suggested that the number of expected links was underestimated. Nonetheless, the high match rate suggests a high quality dataset with few missed links. For more information, see Research Paper: Personal Income Tax and Migrants Integrated Dataset (PITMID) 2011-12 Quality Assessment, Oct 2016 (cat. no. 1351.0.55.060).

71 See Research Paper: Feasibility Study of Linking Migrant Settlement Records to Personal Income Tax Data, Aug 2014 (cat. no. 1351.0.55.051) for more information on the estimated match rate for the 2009-10 and 2010-11 PITMID.

72 Given the high quality of the variables used (name and address) for the deterministic and probabilistic linking and the strict acceptance conditions used, the expected number of false links in 2013-14 PITMID is not expected to be significant. However, the number of false links in the dataset has not been calculated or estimated.

73 If the linking process is conducted in the future, PMD records linked in previous years will not be reprocessed. They will remain linked to the original PIT record. Due to the proposed enduring nature of the links, it was required that strict acceptance conditions be imposed on the linkage process to minimise the potential number of false links.

74 The integration process used to link the 2012-13 and 2013-14 PITMID datasets differed from that used to integrate the 2009-10 to 2011-12 datasets. However, analysis conducted on the linked files compared the linked results and found that approximately 4,400 PMD records that had linked to a PIT record in the 2011-12 dataset linked to a different PIT record in the 2012-13 dataset. However, none of these records were recorded as having significantly changed income characteristics and are not considered to have an impact on the statistics presented in this publication.

75 Due to the strict acceptance conditions imposed during the linkage process, it is possible that some legitimate records could have been missed. However, given the high quality of the linking variables used, it is likely that a high proportion of the unlinked records are simply non-tax lodgers and not missed links. The estimated high match rate indicates that the number of missed links is likely to be minimal.

76 Error introduced by under or over representation of characteristic based groups in unlinked records has not been mitigated by a calibration process. Due to the high quality of the linking variables, it is likely that a high proportion of the unlinked records are simply persons who did not lodge a taxation return.

77 Estimates from the 2013-14 PITMID will differ from the estimates produced from other ABS collections and estimates produced from the PMD for several reasons. The estimates are a result of integrating data from two administrative data sources. The linked records are uncalibrated as there were no known population totals to benchmark to, and the resulting dataset is unique from both the PIT data and the PMD. Due to the quality issues mentioned in the Reliability of Estimates section, estimates should generally be treated with caution.

78 Further information about the data and the linking methodology used for the 2011-12 PITMID is available in the Research Paper: Personal Income Tax and Migrants Integrated Dataset (PITMID) 2011-12 Quality Assessment, Oct 2016 (cat. no. 1351.0.55.060). Information on the data and linking methodology used to create the 2009-10 and 2010-11 PITMID can be found in the Research Paper: Feasibility Study of Linking Migrant Settlement Records to Personal Income Tax Data, Aug 2014 (cat. no. 1351.0.55.051). This research paper provides a summary of the Migrant PIT Feasibility Study.

79 The Australian Census and Migrants Integrated Dataset (ACMID) was created by integrating records from the SDB who had been granted a permanent visa between 1 January 2000 and Census night (9 August 2011) with the 2011 Census of Population of Housing. Approximately 75% of the records on the 2013-14 PITMID could also be present on the 2011 ACMID.

80 The ACMID contains information such as Occupation of main job and Personal weekly income. However, this information is limited to persons aged 15 years and over on Census night. The information on Occupation and income on the PITMID includes all persons who linked to a PIT record, regardless of age. The ACMID also includes all persons aged 15 years and over on Census night, while the PITMID may only include those who earned an income above the tax free threshold. Therefore, while similar statistics can be derived from both the ACMID and the PITMID, they are not directly comparable due to differing population scopes.

81 Estimates from the PITMID are not comparable to the PIT data included in the regional statistics publications; Wage and Salary Earner Estimates for Small Areas (cat. no. 5673.0.55.003) and Estimates of Personal Income for Small Areas (cat. no. 6524.0.55.002). These publications include the PIT data for the whole population, not just migrants. Consequently, personal income data for permanent migrants cannot be identified in these publications' data.

82 Tables 10, 11 and 12 in this PITMID publication contain data for the Australian taxpayer population. This data is not directly comparable with data contained in the most recent Estimates of Personal Income for Small Areas publication. Data in the PITMID tables include only those people aged 15 years and over at the beginning of the reference period, whereas data in the Estimates of Personal Income for Small Areas publication include all taxpayers.

83 The ABS acknowledges the continuing support provided by the Department of Immigration and Border Protection (DIBP), the Department of Social Services (DSS) and the Australian Taxation Office (ATO) for the Migrant PIT Data Integration Project. The provision of data as well as ongoing assistance provided by our stakeholders is essential to enable this important work to be undertaken. The enhancing of migrant related statistics through data linkage by the ABS would not be possible without their cooperation and support.