The ATO Longitudinal Information Files (ALife): Individuals—A New Dataset for Public Policy Research

• Published date: 01 December 2022
• Author: Thomas Abhayaratna,Andrew Carter,Shane Johnson
• Date: 01 December 2022
• DOI: http://doi.org/10.1111/1467-8462.12486

The Australian Economic Review, vol. 55, no. 4, pp. 541–557 DOI: 10.1111/1467-8462.12486

Data Article

The ATO Longitudinal Information Files (ALife):

Individuals—A New Dataset for Public Policy Research

Thomas Abhayaratna, Andrew Carter and Shane Johnson*

Abstract

The Australian Taxation Off‌ice Longitudinal

Information Files: Individuals (ALife:

Individuals), is one of the most comprehensive

tax administrative datasets in the world. The

ALife: Individuals dataset, which currently

covers the period 1990‒1991 to 2017‒2018, is

based on a 10 per cent longitudinal sample of

administrative unit‐record personal income

tax data. This new, high quality, longitudinal,

de‐identif‌ied, research‐ready dataset is avail-

able to approved researchers through secure

environments that safeguard taxpayers'

information. The availability of ALife:

Individuals opens exciting new possibilities

for public policy research and evaluation that

will improve understanding of taxpayer beha-

viour and support policy development and its

administration.

1. Introduction

Recognising the limitations of the existing

available administrative taxation data and in

the face of growing demand to ensure the

eff‌iciency and effectiveness of government

programs and service delivery, the Australian

Taxation Off‌ice (ATO) embarked on the

development of the ATO Longitudinal

Information Files (ALife) to improve access

to individual‐level, high‐quality administra-

tive data.

In 2018, the ATO released the f‌irst instal-

ment of ALife, Alife: Individuals (from here

referred to as ALife), a 10 per cent sample of

annual longitudinally linked individual level

personal income tax records.

1

New waves of

ALife are released annually, early each

calendar year, with a 2‐year lag to allow for

the lodgement cycle in tax returns. The

current release, ALife 2018, released in early

2021, provides unit‐record personal income

tax data over the period 1990‒1991 through

to 2017–2018 and individually linked

superannuation records from 1996–1997 to

2017–2018.

2

Since the inception of the tax system,

Australian taxation data has been made pub-

licly available in various forms. Initially,

* Abhayaratna: Tax and Transfer Policy Institute,

Australian National University, Canberra, Australia;

Crawford School of Public Policy, Australian National

University, Canberra, Australia; Carter and Johnson:

Crawford School of Public Policy, Australian National

University, Canberra, Australia. Corresponding author:

Abhayaratna, email: <Thomas.Abhayaratna@anu.edu.au>

The authors wish to acknowledge the contribution of other

members of the team who assisted in developing and

curating the Australian Taxation Off‌ice Longitudinal

Information Files: Individuals (ALife: Individuals), in-

cluding current and past colleagues of the ATO's Tax

Policy Research and Analysis and Taxation Statistics teams.

We would like to thank Bruce Bastian, Hang To, Justin

Holland, Son Nguyen and Robert Breunig for helpful

comments. All f‌indings, opinions and conclusions are those

of the authors and do not represent the views of the

Australian Government or any of its agencies. Shane

Johnson would also like to acknowledge the support of

the Sir Roland Wilson Foundation. The proposal for this

research was approved by the Australian National

University's Human Research Ethics Committee, protocol

number 2015/477.

Correspondence on the ATO Longitudinal Information Files

should be directed to: ALife@ato.gov.au.

© 2022 The Authors. The Australian Economic Review publishedby John Wiley & Sons Australia, Ltd on behalf of The University of

Melbourne, MelbourneInstitute: Applied Economic & Social Research, Faculty of Businessand Economics.

This is an open accessarticle under the terms of the Creative CommonsAttribution‐NonCommercial‐NoDerivsLicense, which permits

use and distribution in any medium, provided the original workis properly cited, the use is non‐commercial and no modif‌ications or

adaptations are made.

aggregate taxation statistics were incorporated

into the appendices of the Taxation

Commissioner's Annual Report to Parliament,

and from 1959–1960 annual reports have been

supplemented by the publication of Taxation

Statistics.

3

The content in Taxation Statistics

has also grown signif‌icantly over time. One

signif‌icant inclusion has been a small sampleof

conf‌identialised, individual unit‐record data

that was f‌irst released with Taxation Statistics

2003–2004.

4

The availability of a small conf‌identia-

lised individual unit‐record dataset was a

signif‌icant step forward in providing data

access,andspurredgrowthinpolicy‐

relevant research. However, the dataset has

several limitations.

5

Only a small number of

variables are provided, limiting possible

areas of research. In addition, the datasets

are heavily ‘perturbed’, whereby measure-

ment error is added to income variables,

which limits the use of standard techniques

to answer straightforward questions, such as

those that rely on regression discontinuity

design and bunching analysis (Gong and

Gao 2018). In addition, the data are cross

sectional and therefore techniques such as

f‌ixed effects, which could be useful given

the limited demographic information avail-

able in administrative tax records, cannot be

utilised. Furthermore, fundamentally dy-

namic research questions such as income

mobility cannot be addressed with cross‐

sectional data.

In recent years, we have seen greater

demand for data. In part, this increased

demand ref‌lects increased emphasis on a

desire for evidence‐informed policy. This

increased demand for data is supported by

greater computing power, falling storage costs

and improved analytical techniques

(Productivity Commission 2017). At the

same time, we have also seen increased

expectations around the availability and use

of the vast amounts of data collected by

government (Henry et al. 2010; Productivity

Commission 2017).

ALife provides a new dataset of individual

unit‐record personal income tax return data.

This new, longitudinal, de‐identif‌ied dataset is

available to approved users through moni-

tored, secure environments that safeguard the

conf‌identiality and privacy of taxpayers'

information.

The remainder of this paper is structured as

follows. Section 2 provides an overview of the

Australian personal income tax system.

Section 3 brief‌ly outlines why ALife was

created. Section 4 provides information on the

construction of the dataset and related re-

sources and Section 5 outlines the information

available within the dataset. Section 6 de-

scribes access and privacy arrangements and

the article concludes with a discussion on

extensions and possible enhancements that

could improve the potential of the dataset for

research and policy analysis.

2. Australia's Personal Income Tax System

Australia f‌irst introduced a federal income tax

in 1915 to f‌inance involvement in the First

World War. The federal income tax was

modelled on the income tax systems that

applied in the Australian states at the time and

the comprehensive income tax system that

applied in the United States (see Reinhardt

and Steel 2006 for a brief history of the

Australian taxation system and Tilley 2020 for

a detailed history of the early tax policy

developments following Australia's federa-

tion). While Australia's personal income tax

system has gone through many changes since

its inception in 1915, the basic principle of

comprehensive taxation has remained.

Australia's current personal income tax

system shares many similarities to the personal

income tax systems adopted across Western

economies. The Australian personal tax system

is based on the individual, although family

circumstances are recognised for a number of

elements, including the Medicare Levy (and

surcharge), the dependent tax offset and the

Family Tax Benef‌it (a means tested payment for

families with children that could be claimed as a

tax offset until 2008).

6

Australian residents are required to report

all of their worldwide assessable income,

which includes wages and salaries, pensions

and allowances, interes, dividends, capital

542 The Australian Economic Review December 2022

© 2022 The Authors. The Australian Economic Review published by JohnWiley & Sons Australia, Ltd on behalf of The

University of Melbourne, Melbourne Institute:Applied Economic & Social Research, Facultyof Business and Economics.