Chapter 4 sets out the findings of the Joint Committee of Public Accounts and Audit (JCPAA) inquiry into the Commonwealth performance framework, based on Audit Report No. 31 (2015–16), Administration of Higher Education Loan Program Debt and Repayments. The Australian Taxation Office (ATO) and the Department of Education and Training (DET) were the audited agencies. The chapter comprises:
Committee conclusions and recommendations
Committee conclusions and recommendations
Comprehensive performance measures are a key requirement of the Commonwealth performance framework and support transparency in how Commonwealth entities are meeting their purposes. DET has developed some initial performance measures for the Higher Education Loan Program (HELP), focusing on student access and debt repayment. The Committee notes, however, that the department could improve its performance reporting by including measures of program sustainability. This area is increasingly important given HELP’s transition to a demand-driven system and the department’s projection that HELP debt will total approximately $192.5 billion by 2024–25, with $55 billion not being repaid.
The collection and interrogation of processing data is a widely-used and long-established way of informing decision making. In this context, the Committee was concerned by the Australian National Audit Office (ANAO) finding that DET had not established a robust program of evaluation and review based on rigorous analysis of sound data and did not have a complete dataset of HELP payees on which to base policy analysis. However, the Committee notes advice that, following legislative changes allowing sharing of tax file numbers with the ATO, the department is now building a longitudinal database with de-identified information. This development should better support analysis and evidence-based policy formulation.
The Committee strongly endorses the ANAO recommendation that DET and the ATO broaden the range of HELP information reported publicly. If a broad range of data is available, research institutions and other interested parties can use it to identify drivers of trends and propose improvements. The Committee therefore recommends that the department consult with stakeholders on what program data would be most useful for analytical purposes and release as much of that data as possible within privacy constraints.
The Committee recommends that the Department of Education and Training, in responding to Recommendation 4 in Audit Report No. 31 (2015–16), Administration of Higher Education Loan Program Debt and Repayments, consult with the research sector and other interested parties about the Higher Education Loan Program data they would like to see released and publish as much data as is possible within privacy constraints.
Organisational capability is another focus area of the new Commonwealth performance framework arrangements. The ANAO pointed to the significant level of manual intervention used in existing HELP administrative processes, and the extent to which this had slowed processing and added to the administrative cost burden. However, the Committee notes advice provided by DET and the ATO that new tax file number sharing arrangements and data exchange software will resolve some key administrative capability issues. This in turn will create a more efficient process for transferring HELP debt data between the entities, validating it and uploading it to the ATO’s taxpayer records. DET and the ATO had also not established sound administrative arrangements for processing the HECS-HELP Benefit. However, it is noted that this Benefit will cease from 1 July 2017.
Risk management has been a focus in the public sector over many years and remains so in the new Commonwealth performance framework. Of concern to the Committee was that DET and the ATO had not been fulfilling their program risk assessment and management responsibilities, despite establishing a memorandum of understanding requiring them to do so. While the ANAO found that the ATO and the department had independently implemented activities to detect and deter non-compliant behaviour in HELP, it was evident that a shared approach to identifying and managing program risks had been lacking. The Committee notes advice from DET and the ATO that they are now developing a joint risk-based compliance and enforcement strategy for the program, in response to the ANAO’s audit recommendation on this matter. Once the strategy has been developed, it will be important for the entities to periodically refresh their risk assessments and review the effectiveness of mitigation strategies.
The Australian Government provides income-contingent loans to eligible students through the Higher Education Loan Program (HELP), which removes up-front cost barriers to tertiary education and training. HELP provides students with loans to cover relevant costs associated with higher education. Students are required to repay the loan once their income reaches a minimum threshold. HELP is administered by DET, and HELP debt and repayments are collected and managed by the ATO.
The inclusion of Audit Report No. 31 2015–16, Administration of Higher Education Loan Program Debt and Repayments, within the scope of the Committee’s inquiry into the Commonwealth performance framework provided an early opportunity to examine issues relevant to performance reporting. The ANAO’s audit conclusions pointed to the need for improvements in performance monitoring and evaluation, risk management and cooperation between Commonwealth entities.
Review of evidence
This section reviews the evidence received by the Committee regarding:
Implementation of ANAO recommendations
Performance monitoring and reporting
Implementation of ANAO recommendations
DET and the ATO agreed to the ANAO’s five recommendations (see Appendix B for details). At the public hearing, DET noted that ‘most of the actions’ resulting from the audit were ‘well in hand or completed’. The department further advised the Committee that the ‘department and ATO senior executives have agreed to hold external governance meetings, comprising relevant line managers, to oversee the implementation of audit recommendations’. The ATO similarly noted that it was ‘on track to implement all of the ANAO recommendations’.
Specific advice from DET and the ATO about their progress in implementing the ANAO’s recommendations is set out in Table 4.1.
Table 4.1: Summary of reported progress against ANAO recommendations
(ANAO Rec 1)
A new IT platform supporting the transmission of student loan data between agencies is ‘expected to be operational from 1 July 2017’.
Risk assessments and meetings
(ANAO Rec 2)
ATO conducted an internal risk assessment and developed a risk plan for HELP. DET has ‘reviewed and updated its overarching risk plan for the HELP program’, exchanged information with ATO on risk plans, and added risk management and compliance as a standing agenda item at fortnightly meetings between agencies.
Compliance and enforcement strategy
(ANAO Rec 3)
DET is working with the ATO on ‘risk and compliance processes to ensure that risks are effectively identified, documented and addressed in one place. It is expected that by the end of 2016 there will be a joint, risk-based, compliance management and enforcement strategy in place’.
Reporting and analysing information
(ANAO Recs 4 and 5)
DET is working with ATO and the Australian Government Actuary (AGA) on a ‘data improvement project involving the creation of a database of de-identified HELP information’, and to establish the final database by early 2018. The database is to provide DET with ‘a richer analysis of factors affecting HELP debt repayment’ to inform program design and expand information that can be published on HELP. In consultation with ATO and AGA, DET is ‘exploring ways in which selected information generated by the data improvement project could be published on an annual basis’.
Performance monitoring and reporting
Current program-level performance reporting
In reviewing DET’s portfolio budget statements and annual reports, the ANAO identified targets and indicators addressing program access and student debt repayment, including:
targets for the number of higher education places to be delivered by each of the five types of HELP loans
key performance indicators, under the objective of ‘Students who are able to repay their HELP debts do so’, which set targets for the average dollar amount of HELP debts, the average number of years to repay debts and the proportion of new debt not expected to be repaid
Outside of these formal Commonwealth performance framework documents, the ANAO found a similar focus in DET’s Higher Education Report, noting that it:
included details of funding allocations and attainment targets for higher education
had a focus on the number of new loans being obtained and the rate of enrolment in higher education
While policy responsibility for HELP rests with DET, the ATO publishes some program-level information, consistent with its responsibility for administering and collecting HELP debts. For example, the ATO’s Taxation Statistics report (published every 16 months) includes demographic information on HELP debt and repayments. In its annual report, the ATO also reports the value of HELP debt repayments, but only as a part of an aggregated figure together with a separate program (the Student Financial Supplement Scheme). The ANAO suggested that the figures be reported separately, and also noted that the ‘introduction of indicators reflecting compulsory and voluntary repayments would assist to measure the impact of activities undertaken by the ATO aimed at encouraging voluntary repayments and targeting non-compliant payees’.
Reporting on sustainability of HELP
The enhanced Commonwealth performance framework encourages Commonwealth entities to examine their operating environment over a long-term horizon, with guidance noting that entities could set out the ‘main external and internal factors that affect or influence … performance. For example, the entity may describe the conditions, circumstances and trends that may affect its capacity to achieve its purposes or the demand for, or supply of, its services.’
A significant trend in the HELP environment is the rate at which HELP debt is accumulating and its potential implications for the financial sustainability of the program. DET estimates that the compound annual growth rate of HELP debt will be 14.5 per cent between 2015–16 and 2024–25, and that $55.0 billion of an estimated $192.5 billion in outstanding HELP debt will not be repaid by 2024–25.
Despite the growing trend in HELP debt, sustainability indicators have been absent from DET’s reported performance measures. The ANAO has also noted that the department’s Higher Education Report does not include information about the fastest growing component of HELP debt, VET FEE-HELP, which provides loans for students studying higher level vocational education and training courses. The ANAO suggested that ‘measures of sustainability could include deliverables relating to the level of debt and amount of new debt and repayments.’
When asked about the sustainability of HELP debt repayments, DET advised that:
There are a few moving parts to it. In terms of the growth, there is the simple demand, so there has been a lot of increase in VET and higher education through the emergence of the demand-driven system in the last few years. So simple volumes have been a big driver in that. There is the aspect of policy design that says: ‘What ought the rules to be? On what basis should people repay?’ That goes to things like repayment thresholds and amounts. … then there are questions about collection and enforcement. We have seen changes to all these.
Specific sustainability reforms cited by DET included:
removing some upfront payment discounts and repayment bonuses that go to diminish the value of collectable debt; the closure of the VET FEE-HELP program to be replaced by VET student loans … and, very recently, new lower minimum repayment thresholds so that people start repaying more quickly.
Compiling, analysing and disclosing HELP data
In terms of collecting data and analysing HELP debt, the ANAO identified that there was ‘limited information available regarding the factors affecting the debt repayments’ and that DET was ‘unable to demonstrate to the ANAO that it routinely monitored and analysed factors affecting the repayment of HELP debt’. The ANAO recommended that DET more fully analyse characteristics of debt and repayments, and use this information to inform program design.
Collecting a broad range of relevant data is a prerequisite for undertaking detailed policy analysis, but the ANAO found that DET did not have a complete dataset of HELP payees. At the public hearing, the department advised that:
the correctly tight restrictions under legislation have always been applied to the use of tax file numbers for specific purposes. So, whilst we have had a database—with such data as we can get into it; and the ATO obviously has the complete picture—we have had to rely on a variety of de-identified data, the use of the Australian Government Actuary and so on to create datasets from which we can derive policy. There has been some indirectness in all of this. With the passage of the legislation, it now allows the use of the tax file number for these purposes. Then the technology can support all sorts of things we have not been able to do in the past.
The legislative changes referred to by the department relate to the establishment of an enhanced HELP database. DET advised the ANAO that it was:
working with the AGA and the ATO to build an expanded longitudinal database of HELP debtors. This database will include de-identified information relating to debtors, their likelihood of repaying that debt and factors that influence their capacity to repay. This will enable greater analysis of factors affecting HELP debt repayment and inform future policy and program design.
At the public hearing, DET explained how the new dataset might change the nature of its analysis:
Inherent in an income-contingent loan scheme is the fact that you are creating the conditions in which people will not repay if they do not achieve the outcomes in terms of employment and future lifetime income … the granular detail we will now be able to examine about the experience of particular cohorts of students in particular subsets by education field, by state or by any other demographic delimiter we might choose to explore will be richer than it has been to date, and we will then be able to start to think about whether, in these schemes, we want to have a smaller or bigger risk of not returning the loan funds based on what the benefits overlaying that student access to higher education will be.
A related matter discussed at the public hearing was whether HELP data should be made more widely available for analysis. The ANAO recommended that DET and the ATO better inform stakeholders by expanding the information provided publicly, to include data on growth in HELP debt and collection of repayments. In explaining the basis for its recommendation, the ANAO noted that:
We are aware of a lot of academics and others who work in that space who contribute to the policy debate, so we were looking for some of that to come out. We did some comparisons internationally of where there was more disclosure of the modelling of the debt. Again, some of the discussion was about what the key factors are that contribute to the debt. If some of that information can come out publicly, it can be analysed by academics and others and it can contribute to policy improvements.
As to what approach it would take in responding to the ANAO’s recommendation, Mr Dom English, Group Manager, Department of Education and Training, advised that:
The paramount consideration will always be privacy, so any data we release has to be constructed in a way that would not allow individuals to be identified. But beyond that our intent is to examine and provide as much publicly available data as is relevant and useful to support an understanding of what is happening with this debt and the management of it by the sector and the government. That is a matter we are still exploring … But I would expect to see further [student] cohort breakdowns of the data. Amount, accumulation and repayments would be pretty readily publishable … There will be interesting differences across cohorts in terms of both labour market participation rates and the success of students from various academic backgrounds and various levels of academic preparation … We have a bias to try and get as much out as we can …
DET also advised the Committee that it wanted to consult with the research sector and other interested parties to gather views about what data would be useful.
Reporting on capability is a focus area within the new Commonwealth performance framework, with Commonwealth entities being required to identify (in their corporate plan) key strategies and plans to be used to achieve their purposes. In articulating their strategies and plans, entities are encouraged to describe their current capability, assess how their capability needs may change, and outline their strategies to build capability in relevant areas such as staffing, capital investment and communication technology (ICT). The ICT infrastructure and processing capability supporting HELP is extensively canvassed in the ANAO report, including in relation to:
recording and processing HELP debts
recording and processing the HECS-HELP Benefit
managing debts and repayments
Recording and processing HELP debts
The ICT capability associated with recording HELP debts is based on the following flow of program information:
education providers submit student debt information to DET via an online tool—this data contains the personal details of students, type of study undertaken and value of loan debt incurred
DET collates the data into files and transfers them to the ATO four times a year
the ATO uploads the data to record debt information against ATO client records and is responsible for administering and collecting the debts
The ANAO report found that DET and the ATO have a number of controls in place to support the collection, transfer and upload of reliable and accurate HELP debt data, but that issues with the integrity of data contained in files transferred from DET ‘complicate the processes to upload the student debt data into the ATO’s Integrated Core Processing information technology system, and sometimes result in extensive manual data matching processes’. Specific data integrity issues cited by the ANAO include files containing:
incorrect or missing tax file numbers (TFNs)
duplicate and/or shared unique identifying numbers called Commonwealth Higher Education Student Support Numbers (CHESSNs)
data formats that the ATO system is not able to accept (eg, apostrophes in addresses)
The practical impact of the data integrity issues is that the ATO undertakes a significant amount of data validation and does not use the TFNs and CHESSNs as the unique identifiers they were designed to be. Instead it attempts to match the department’s debt records to ATO client records using programmed combinations of information fields (eg, given and family names, date of birth and postcode or TFN, given name and date of birth).
At a specific level, the ATO’s upload and matching process involves five stages, with three data validation checks and a debt matching process occurring prior to uploading the student files into the system. At each stage, manual intervention may be required, such as where data cannot be decrypted correctly, invalid or missing data is identified, or debt records cannot be matched to ATO payee records. Unmatched records may require further follow-up through letters to students and providers seeking information. The ANAO found that the ‘full extent and cost of these issues is unknown as their occurrence is not routinely recorded’ by the ATO or DET. The ATO was unable to quantify for the ANAO the number of records that failed checks at each of the five matching stages. However, the ATO advised the ANAO of 11 851 records that failed the final upload stage in 2014–15 (28 940 in 2013–14), and the ANAO estimated the approximate cost to the ATO to manage these records at $193 000 ($450 000 in 2013–14).
To address these issues, the ANAO recommended that DET and the ATO ‘simplify and improve the debt recording process’ by redesigning ‘the controls for the initial recording, transfer and upload of HELP debts’. The ANAO also noted that there was further scope to address the underlying data integrity and system misalignment issues through improving the alignment of data field rules between ATO’s and DET’s systems and improving the completeness of student data.
The ATO’s submission to the inquiry advised that, following legislative changes, the ATO and DET are developing a ‘real time data exchange channel’ to:
streamline and automate the process for reporting of HELP obligations to the ATO;
allow for the redesign of the controls for more simpler and efficient recording, transfer and upload of HELP obligations;
allow real time transfer of HELP loans from [the department] to the ATO;
include an automatic TFN verification and validation process which will further minimise fraud.
The ATO similarly advised the ANAO that only ‘limited manual intervention will be required’ following the introduction of the data exchange channel as it will ‘require TFNs to be verified before providers can submit student files and intervention will only be necessary for records with an incorrect or compromised TFN. TFN verification is expected to improve the efficiency of the debt matching process’.
Recording and processing HECS-HELP Benefit
Another ICT capability issue identified by the ANAO relates to the ATO’s processing of the HECS-HELP Benefit. The HECS-HELP Benefit ‘aims to encourage graduates of maths, science, education and nursing study courses to gain employment in specified occupations’. Payees in those fields may be eligible to apply for a discount reducing their compulsory HELP repayment amount or total HELP debt, subject to complex eligibility requirements, including the type of employment undertaken and certain geographical eligibility requirements. The total value of discounts provided in 2014–15 was $31.9 million.
The ANAO report notes that it was not possible to test the ATO’s processing of revised HECS-HELP Benefit thresholds (benefit amounts are indexed to CPI on 1 June each year) as the ATO could not provide reliable data. The ANAO also concluded that the ATO’s system ‘does not support accurate processing of the HECS-HELP Benefit in all circumstances, and the ATO estimates that some 12 000 payees are affected by errors including incorrect account balances’.
System issues were first identified by the ATO in 2011, and the ANAO report noted that, ‘while there have been internal requests for the system to be improved since this time, they have not been actioned’. The ATO advised the ANAO that the processing issues cost approximately $180 000 per annum in staff time to manage and that the potential impact for payees is the receipt of an incorrect benefit amount or incorrectly being denied a benefit. These issues also reduce the speed of processing, with the ATO not achieving its service standard of processing applications within 28 days.
The HECS-HELP Benefit has been phased out, with the ATO’s submission to the inquiry noting that ‘at the time of the ANAO audit there was legislation before Parliament to remove the HECS-HELP benefit. This law passed Parliament and was given Royal Assent in September 2016, with the HECS-HELP benefit ceasing from 1 July 2017, although claims for the benefit will be available until 30 June 2019’.
Managing debts and repayments
In terms of the ATO’s management of HELP debts and repayments, the ANAO obtained ‘assurance that the ATO updated the Integrated Core Processing system with the correct income repayment thresholds and accurately calculated HELP compulsory repayments’, and concluded that ‘HELP voluntary repayments have largely been effectively managed’.
However, the ANAO pointed to issues regarding discounts applied to voluntary repayments. Up until 1 January 2017, some voluntary repayments attracted a debt discount of five per cent of the value of the repayment, but the ANAO found that the ATO incorrectly provided bonuses for two types of HELP loans, despite a 2003–04 Budget announcement indicating that the bonus would not apply to those loan types.
The failure to implement this budget measure relates to another ICT capability gap—the ATO’s Integrated Core Processing system is not able to differentiate repayments by HELP loan type. The ANAO noted that DET could not provide documentation to demonstrate there was a formal decision to not implement the Budget measure. The ATO advised the ANAO that if legislation had been introduced to provide for the measure it would have created this capability within its system.
In relation to the cost of not implementing the measure, the ANAO estimated that the value of bonuses incorrectly processed in 2014–15 was around $2 million and around $10 million since the loans were introduced in 2004–05. DET advised the Committee that ‘one of the more recent changes in the program has been the removal of the voluntary repayment bonus ... That scheme is now finished’.
HELP compliance activities
Commonwealth entities are required under the Public Governance, Performance and Accountability Act 2013 (PGPA Act) to ‘establish and maintain … an appropriate system of risk oversight and management for the entity,’ and responsibilities are further defined by the Commonwealth Risk Management Policy. The guidance supporting the enhanced Commonwealth performance framework encourages entities to demonstrate that ‘effective systems of risk oversight and management have been implemented’. For HELP, non-compliance by participants with program requirements is the most significant risk addressed in the ANAO report.
At a broad level, the ANAO found that the ATO and DET had implemented a range of activities to detect and deter non-compliant behaviour among HELP participants, including business-as-usual and targeted compliance efforts:
Business as usual activities (those managed as part of the ATO’s wider income tax compliance program) raised $34.2 million in compulsory HELP repayments in 2013–14 and 2014–15. The vast majority of these additional repayments ($29.1 million) were raised as a consequence of the ATO targeting under-reported income.
Targeted (HELP-specific) compliance measures used by the ATO included targeting individuals with HELP debts who had not lodged tax returns. The program ran over three financial years (2011–12 to 2013–14) and resulted in approximately one-third of the non-lodgers targeted subsequently lodging a tax return. This raised $207.7 million in compulsory repayments and $9.7 million in voluntary repayments. The activity has been discontinued as a HELP-specific compliance measure and is now part of the ‘standard non-lodger program’ conducted by the ATO. The ANAO suggested that the ATO monitor outcomes from the new arrangements to ensure that they are as successful as the HELP-specific non-lodger program.
While the ANAO report largely focused on compliance by students in repaying debts, the ANAO also cited advice from DET that it undertakes both education provider and student compliance activities, including: verifying student data; monitoring the lifetime access limits that apply to two HELP payments; and reconciling funds distributed to education providers.
Cross-entity risk management
At the public hearing, the Auditor-General advised the Committee that shared responsibilities had been a focus for recent ANAO performance audits, including HELP, noting: ‘we identified that when there was a joined-up approach, the risk of governance failure was higher and therefore it was something that agencies should focus on as an important area’. Despite finding the ATO and DET each had compliance activities in place, the ANAO report identified that ‘these activities are not supported by a risk-based compliance and enforcement strategy’.
In areas of shared administration, such as HELP, the Commonwealth performance framework is intended to ‘support a more joined-up view of government activity by providing the Parliament and the public with the means to relate contributions by different Commonwealth entities in shared policy areas’. The joined-up arrangements for HELP are set out in a Memorandum of Understanding (MoU) agreed between DET and the ATO. The MoU provides that responsibility for managing HELP risks is shared by the two entities, and requires that they each:
maintain risk management plans for the HELP services they are responsible for delivering
advise the other party of any risk management impacts likely to affect their administration of HELP
In relation to risk management plans, the ANAO found that the ATO had not developed a risk management plan for HELP or conducted an assessment of compliance risks associated with the services it is responsible for delivering under the MoU, including the collection and administration of HELP debt. Instead it relies on its broader ‘ATO corporate risk management strategy for the income tax system’. Similarly, the ANAO advised that DET ‘had specific risk plans for three of the five [HELP] programs … but not the other two. So it was incomplete from a HELP perspective’.
With respect to the importance of such plans, the ANAO noted that they had:
done some auditing of [the ATO’s] smaller programs and we can accept that there is not going to be explicit risk in compliance strategies every time. But on this occasion we thought, given the scale of it, there should have been a more coordinated risk approach between the two agencies and also compliance strategies that straddled both the agencies so they could get together to prioritise the risk and the compliance activity.
In this regard, the ANAO report found that DET and the ATO meet on a regular basis to discuss ‘a range of issues’ but ‘there has been no formal and explicit risk assessments and mitigation strategies discussed and shared in these meetings’.
The ANAO made two recommendations relating to risk management, including that the ATO and DET:
work together to develop a risk-based compliance and enforcement strategy for HELP, with the objective of providing assurance that compliance risks associated with HELP are effectively identified and mitigated
improve the management of HELP risks and support MOU risk management requirements, by developing risk management plans for their respective MOU services and introducing a standing agenda item for risk management at their regular meetings to facilitate discussion of relevant risks
DET’s submission to the inquiry identified that the department and the ATO held a risk workshop following the audit at which the entities discussed both shared risks and those specific to each entity, including in the areas of ‘provider verification, debt calculation, implementation of policy/legislative change, debt write-offs and waivers, administration of tax file numbers, compliance (both by resident and overseas debtors), client communications and stakeholder management’. DET also advised that:
we have been looking at the sorts of risks that we are responsible for managing, including improving the validation of data by the students as their tax file numbers and student enrolment details are collected. That will feed into a project that is currently being rolled out to implement a new online registration process for students in that commencement of the HELP debt process.
In implementing the ANAO’s recommendations, DET further noted that it was working with the ATO on ‘risk and compliance processes to ensure that risks are effectively identified, documented and addressed in one place. It is expected that by the end of 2016 there will be a joint, risk-based, compliance management and enforcement strategy in place’. The department had also ‘reviewed and updated its overarching risk plan for the HELP program’ and exchanged information with the ATO on respective risk plans.
6 December 2017