Go to top of page

2.5 Delineating policy and service delivery

As noted in the Phase One Report, policy development is a core function of government and should be undertaken by portfolio departments. This helps to clarify accountability, encourage specialisation and reduce costs.

Government ‘service delivery’ refers to the range of methods used to deliver public policy. As noted in the Productivity Commission’s Report on Government Services 2014, it is largely concerned with:

  • the provision of public goods such as:
    • a legal framework that determines the rules for ownership of property and the operation of markets (including the arrangements which underpin the work of the courts, police and corrective services); and
    • helping citizens manage adverse events (including the work of emergency services and aspects of the health system); and
  • enabling higher levels, higher quality and/or more equitable consumption of services that governments consider to have particular merit or that generate beneficial spillover effects for the community (including education, health and welfare services).

The taxation, Commonwealth payment and regulatory systems are key government delivery interfaces for Australian citizens and businesses.

At the Commonwealth level, there are a number of high profile examples of policy development and service delivery separation. For example, the Department of Human Services (DHS) is the Commonwealth’s primary service delivery agency for social services on behalf of departments such as Health, Social Services, Employment and Education. The Australian Taxation Office, Australian Securities and Investments Commission, Australian Competition and Consumer Commission and Australian Prudential Regulation Authority provide taxation, financial and other regulatory services, while the Department of the Treasury undertakes the majority of policy work in these areas.

This separation has led to efficiencies and provided opportunities to consider new service delivery models, including those which increase competition on the supply side of public services.

The creation of specialised service delivery agencies has also provided opportunities to integrate and make greater use of information technology (IT) to streamline back office functions, improve online offerings for clients and business, improve access to services and reduce service delivery costs.

Policy development and service delivery, however, are not mutually exclusive and the relationship between the two can be complex.


As suggested in the Phase One Report, the Commission considers that portfolio departments should undertake policy work and that service delivery should, for the most part, be undertaken by delivery agencies. This is because separation helps to:

  • ensure policy development and service delivery are undertaken by specialised entities with appropriate expertise and accountability arrangements, including clear ministerial responsibility;
  • improve the focus on citizens needs, while reducing costs through consolidation of delivery functions (allowing economies of scale in shared services and other overheads); and
  • make it easier to open up service delivery to competition.

Between the categories of policy and service delivery, however, choices need to be made about how things get done. A key challenge, therefore, lies not so much in the separation of policy and delivery (which already exists at the Commonwealth level for the most part) but in working out how best to connect them more effectively.

Structure of Commonwealth service delivery

The Productivity Commission’s most recent Report on Government Services 2014 provides a breakdown of various services delivered by the Commonwealth and the States.

DHS has the most extensive Commonwealth service infrastructure with:

  • 598 service centres providing Medicare, Centrelink, Child Support and CRS Australia services;
  • 227 access points in regional, rural and remote Australia providing free self-help facilities (such as online access, photocopying, telephones);
  • 352 agents in regional, rural and remote Australia providing similar services to access points, as well as agents providing face-to-face information service to the community;
  • 140 additional visiting services provided by CRS Australia; and
  • Mobile Service Centres delivering payments and services to regional/rural areas (DHS, 2013).

As noted above, there is already considerable separation between Commonwealth policy and service delivery functions. The Australian Public Service Commission (APSC) 2012-13 State of the Service report captures this by classifying Australian Public Service agencies into the following categories: policy; operational; regulatory; or specialist.

The APSC classifies 25 per cent of agencies as operational, 15 per cent as policy, 16 per cent as regulatory and 44 per cent as specialist – see Chart 2.11 below.

Chart 2.11: Classification of Australian Public Service entities by function

The chart shows that 25 per cent of Australian Public Service agencies are classified by the Australian Public Service Commission as operational, 15 per cent as policy, 16 per cent as regulatory and 44 per cent as specialist.

Source: APSC, 2013a.

This provides an indication of the degree of separation of policy and service delivery, but it obscures the fact that a number of activities undertaken by the Commonwealth are neither strictly policy development nor service delivery, but fall somewhere between the two.

For example, the Department of Health is classified as a policy entity, but it also undertakes service delivery activities. The Office of Hearing Services (OHS) is a branch in the Department, which is responsible for managing and administering the Australian Government Hearing Services Programme. The OHS provides a number of activities including: call centre services; a complaints management process; client voucher applications processing; management of contracted service providers as well as a memorandum of understanding with Australian Hearing Services; audit and compliance activities; and policy advice to government to inform the strategic direction of hearing loss prevention (Department of Health, 2012).

In addition, in some organisations, the line between policy and service delivery is not clear. This is a claim often made with respect of national regulators where there is a perception that some regulators seek to test the extent of the law in addition to enforcement.

For example, a report by Deloitte Access Economics commissioned by the Australian Food and Grocery Council notes that several national regulators operating in the food and grocery sector can initiate new rules, make rules (decide on what should be regulated) and enforce rules (process and grant applications for registration) resulting in the potential for regulatory activism: the ability to expand the range of activities to be regulated and to create their own work programme (Deloitte Access Economics, 2013).

The Australian Competition and Consumer Commission will often test the law (including Australian Consumer Law) by taking on legal cases where the outcome is less certain. The fact that this organisation has incurred operating losses partly reflects this strategy as there have been significant financial impacts when unsuccessful court cases result in costs being awarded against the Commission (Australian Competition and Consumer Commission, 2013).

Trends in policy and service delivery arrangements

The clearest example of the delineation of policy and delivery functions at the Commonwealth level is the separation of social policy from the delivery of social services.

The current delivery arrangements have their origins in the creation of Centrelink in 1997. Centrelink was established as a statutory agency in the social security portfolio with responsibility for delivering services and benefits to unemployed Australians and social welfare recipients.

Centrelink was not set up as a Budget-funded agency, but as one funded directly by the departments and agencies for which it delivered services, where each negotiated a purchase price for the services Centrelink agreed to provide (Halligan, 2008).

These purchaser-provider arrangements were governed by negotiated contractual agreements (Business Partnership Agreements (BPAs)) with each department. Centrelink was required to report to its client departments on its performance against a series of detailed indicators. As noted by Halligan (2008):

The use of BPAs was a fairly new experience for Commonwealth Government agencies. They required all parties to understand fully what services were to be provided, when, how and at what cost, without making the relationships between the organisations too rigid in a politically and publicly sensitive environment.

DHS was established in 2004, bringing together a range of service delivery agencies, including Centrelink, Medicare, Health Services Australia and the Child Support Agency to improve service delivery arrangements. The Department was tasked with ensuring the Commonwealth secured the best value for money in service delivery while emphasising continuous improvement and a whole-of-government approach (APSC, 2007).

The Human Services Portfolio was created partly in response to the Uhrig review of corporate governance of statutory authorities and office holders (APSC, 2007). The direct purchaser provider model and independent board were replaced by more traditional governance arrangements, including the heads of the service delivery agencies (Medicare Australia, Centrelink, Health Services Australia and Australian Hearing) now reporting directly to the Minister through the Secretary of DHS. The other two service delivery agencies, the Child Support Agency and CRS Australia, were legally part of DHS and reported to the Secretary (APSC, 2007).

In December 2009, the then Commonwealth Government announced significant reforms to government service delivery to generate further efficiencies and improvements. As part of these reforms, in 2011 the former Human Services agencies – Centrelink and Medicare – were integrated into DHS, bringing together the Department’s frontline service delivery networks into a single customer facing network to provide coordinated support to people (DHS, 2011).

The evolution of DHS has also led to greater use of digital delivery channels. As noted in the Strategic Review of Future Directions for Australian Government Service Delivery:

Emerging internet technology, both hardware, in terms of capacity, speed and geographical reach, and software, in terms of interpretive data manipulation and communication, is rapidly changing modes of communication and expectations about information, immediacy and functionality

Australian Government, 2009

As shown in Chart 2.12 below, just over half of people contacting government (54 per cent) use e‑Government (internet and telephone) as their preferred method. A smaller proportion (37 per cent) mainly makes contact with government in person, while a small group mainly use mail (7 per cent).

Chart 2.12 Main service delivery channel used to contact government

As noted in the text above, this chart shows that just over half of people (54 per cent) use e government (internet and telephone) as their preferred way of communicating with government. A smaller proportion (37 per cent) mainly makes contact with government in person, while a small group mainly use mail (7 per cent).

Source: AGIMO, 2011

As noted in the Phase One Report, DHS has recently introduced the myGov service, which allows people to access government services for Medicare, Centrelink, Child Support, the Department of Health, the Department of Veterans’ Affairs and the National Disability Insurance Agency using one user name and password online and, in some cases, via mobile phone applications.

This trend towards greater use of online service delivery is expected to continue. The Australian Taxation Office is scheduled to join myGov in 2014 through its e-tax product and new web and mobile services. MyGov recently added a digital mailbox service to provide a secure electronic mail delivery channel for official correspondence, with the capacity to link with the Australia Post Digital Mailbox facility and, potentially, commercial providers.

Alternative service delivery models

The public service does not have a monopoly on the provision of services to the public. Governments in Australia and overseas use a mix of methods to deliver services to the community. These are summarised in the Report on Government Services 2014 as follows:

  • delivering or providing the services directly (a ‘delivery/provider’ role);
  • funding external providers through grants or the purchase of services (a ‘purchaser’ role);
  • subsidising users (through vouchers or cash payments) to purchase services from external providers;
  • imposing community service obligations on public and private providers; and
  • providing incentives to users and/or providers, such as reducing tax obligations in particular circumstances (known as ‘tax expenditures’) (Productivity Commission, 2014).

In addition, there is a range of hybrid arrangements, including different partnership arrangements such as collaborative partnerships and public private partnerships in various guises, as well as newer approaches such as ‘social impact/benefit bonds’. Social impact/benefit bonds are described briefly in Section 10.17 of Appendix Volume 2.

In a 2012 report, Diversity and Contestability in the Public Service Economy, Professor Gary Sturgess argued for the introduction of greater competition in the delivery of public service through use of ‘choice-based’, ‘commissioning’ and ‘contestability’ models to improve productivity and financial performance.

  • Choice-based models involve service users themselves selecting from a range of alternative providers financed, for example, through government vouchers.
  • Commissioning models involve public officials purchasing services on behalf of the community through competitive tendering and contracting. While this option includes simple outsourcing models, it also captures public-private partnerships, public-private joint ventures and integration contracts.
  • Contestability models involve benchmarking service providers, with failing institutions facing a credible threat of competition (although perhaps not actual competition).

Sturgess’s view is that commissioning models have the widest application.

The Commission has already made a number of recommendations in the Phase One Report to increase competition in the delivery of services using these approaches. These include:

  • choice-based models – the introduction of vouchers for Indigenous education;
  • commissioning models – outsourcing visa processing and potentially aspects of the DHS payment system; and
  • contestability – benchmarking the performance of the Australian Broadcasting Corporation and Special Broadcasting Service.

Moves away from direct Commonwealth service provision

The 1996 Commission of Audit recommended the separation of policy formation from programme delivery by ‘clarifying the separation between purchaser – the Commonwealth Government – and service provider, including promoting competitive tendering for services by “contracting out”‘.

Two years after the Commission’s report, the Commonwealth Employment Service was replaced as a provider of employment services by a new competitive market, known as the Job Network.

The outsourcing of service delivery in employment services followed a purchaser‑provider model, with the then Department of Employment and Workplace Relations purchasing employment services from a competitive market place while welfare services were purchased from Centrelink by the then Department of Family and Community Services. MacDermott (2008) suggests:

The efficiency argument underpinning the purchaser/provider model was that the contractual arrangement itself would force the purchaser to take responsibility for policy implementation by defining the outputs being sought, enabling the provider to focus on responsiveness to the public and value for money more generally.

As noted in the Phase One Report, international evidence suggests that the Australian approach is seen as being at the frontier of good policy in this area. Job Services Australia and its predecessors demonstrate the benefits of using contracting arrangements compared to traditional government provision of these services.

A trend to increased delivery of government services by non-government service providers is also evident overseas, in particular in the United Kingdom. For example, one of the objectives of the United Kingdom’s 2010 Spending Review was to:

...increas[e the] diversity of provision in public services through further use of payment by results, removing barriers to greater independent provision, and supporting communities, citizens and volunteers to play a bigger role in shaping and providing services.

Coordination across services and policies

As noted above, policy development and service delivery are not mutually exclusive and the relationship between the two can be complex. A good understanding of service delivery and customer needs, often best learnt ‘by doing’, can produce the highest quality and most practical policy.

The 2009 Strategic Review of Future Directions for Australian Government Service Delivery argued that to improve service delivery there is a need to strengthen horizontal linkages across portfolios and policies and service delivery networks (Australian Government, 2009). It adds that linking policies and programmes is a critical driver of citizen-centric service delivery (and the achievement of better outcomes), noting that ‘programs rarely operate in isolation’.

This is, however, a challenging and potentially expensive task. As the Strategic Reviewnotes, joined-up government goes against the grain of Westminster government and the traditions of the Commonwealth bureaucracy. Information on services and programmes is generally agency-specific, incomplete (in terms of all relevant services or programmes), not functionally linked and dependent on citizens’ navigating multiple information sources.

In addition, the world as it is imagined by policymakers rarely resembles the world as those who deliver services experience it day-to-day.

Front line managers report that they are often consulted late in the process, once policy development has been largely completed and there is insufficient time for effective feedback. Also policy agencies are not always receptive to advice from the front line.

For example, the recent Capability Review of the Treasury noted that stakeholders described Treasury officers as dismissive of the expertise of line agencies or business and, instead, tended to revert to first principles to build an understanding from scratch (APSC, 2013b).

This is not a trait that is limited to Treasury. It is evident that significant cultural change is required and that policymakers should be more prepared to learn from and access the ‘lived experience’ of front line workers. Practical issues around delivery should be built into the foundations of policy design.

A range of strategies to improve feedback loops between policy and service delivery, including encouraging mobility between policy and service delivery networks is outlined in Section 3.4 of the Phase Two Report.

Addressing these issues is particularly important in cases where more innovative and ‘citizen centric’ service delivery models are being pursued because these require greater flexibility for service providers and, ideally, much better integration across policy and service delivery silos and better use of administrative data to reduce costs and improve outcomes.

Potential areas for reform

‘Commissioning’ of services

As indicated above, governments have a variety of instruments at their command for the implementation of policy – direct delivery, contracts and grants, loans and loan guarantees, taxes and tax expenditures and market design and regulation.

Which of these is best, and thus how the relationship between policy and implementation should be constructed and navigated, will differ depending on the circumstances.

A ‘commissioning’ arrangement has the potential to improve aspects of the traditional relationship between policy and delivery.

Commissioning is the process through which providers are granted authority to deliver services designed to meet specified social and economic outcomes. It is a relatively new concept in public administration, having been developed over the past decade or more in the United Kingdom at a time when public officials were starting to take a more strategic view of public sector markets (Sturgess, 2012).

The process of commissioning involves the identification and prioritisation of outcomes and the design of measurable performance objectives. This will inform government as to whether those outcomes are being met and whether they are being delivered in an effective and efficient manner.

A key challenge in commissioning public services is the specification of outcomes and the design of performance measures that serve as appropriate surrogates for those outcomes.

Sturgess (2012) notes the United Kingdom’s Institute for Government notion of ‘system stewardship’ to capture the task and challenges involved in the commissioning process, including the extent to which a service will be devolved, the need for uniformity in delivery and the complexity of the service in question.

By specifying outcomes or high level outputs, rather than processes or inputs, performance measurement can be employed in a way that gives the providers of services the room and incentive to innovate.

While commissioning has drawn on lessons from performance contracting, it has not particularly focused on privatisation or outsourcing. Rather, it has focussed on the design, development and management of contestable systems. Contestability involves performance benchmarking across the entire system and formal processes for intervening in failing institutions (public or private).

The principle of commissioning has been likened to the notion of a supply chain, where client and supplier organisations are deeply interconnected, as opposed to that of a market where the regulator stands at arm’s length from free-standing firms engaged in trade with one another.

The process of commissioning differs from procurement or purchasing, which are technical functions not usually concerned with the identification and specification of outcomes, system design or the exploration of alternative service models.

For example, on infrastructure projects, the vast majority of savings are delivered in the early stages of commissioning where there is still an opportunity to fundamentally alter the scope of what is being commissioned. Once the process has progressed to procurement, the scope for fundamental savings is quite limited.

In some cases services should be commissioned through an iterative process, weighing up the intended outcomes, the ease of implementation and affordability through repeated rounds of development and re-assessment. This can lead to innovation, as tends to happen in the private sector.

To illustrate, Sturgess (2014) notes that:

Some aspects of commissioning are strongly commercial in nature, and in the UK, serious consideration has been given to the possibility of contracting some of them to private firms. This included the management of the Defence Equipment and Support Agency, which was to be contracted through a Government Owned, Contractor Operated model. This procurement was abandoned in December 2013 when one of the two bidding consortia withdrew.

However, in the UK Work Programme (the UK equivalent of Job Services), the prime contractors have established networks of small and medium sized providers from the public, private and not-for-for-profit sectors, developing and managing this supply chain with a high degree of transparency. Several of the prime contractors, or ‘integrators’ as they have come to be known, deliver none of services themselves, so that, in effect, they are managing a diverse supply chain on behalf of government.

Strengthened implementation planning

The Cabinet Implementation Unit (CIU) was established within the Department of the Prime Minister and Cabinet in 2003 to report to Cabinet on progress with the implementation of Cabinet priorities. Where implementation problems were identified, they were brought to the Cabinet’s attention and the CIU worked with agencies on remediation planning.

The concept of an Australian implementation unit was inspired by the Prime Minister’s Delivery Unit in the United Kingdom. It recognises that in Australia, as it is elsewhere, government service delivery is increasingly complex, often involving multiple portfolios and jurisdictions (Department of the Prime Minister and Cabinet, 2013).

In recent years the CIU has maintained regular Cabinet reporting, while taking an active role in advising departments on implementation planning early in the policy process. This has involved:

  • assessing all submissions and new policy proposals going to the Cabinet for implementation and delivery issues planning;
  • building capability across the Australian Public Service on implementation planning through training and the development of guides; and
  • maintaining an implementation network for senior executives to share implementation and delivery experiences.

The CIU also operates a secondment programme which supports implementation capability building by increasing networks and sharing implementation expertise. It consults with departments and agencies on tailored implementation planning for significant implementation exercises (Department of the Prime Minister and Cabinet, 2013).

Randomised controlled trials

Randomised controlled trials (RCTs) are a potentially powerful evaluation tool, which can be used to test the efficacy of a proposed intervention or alternative approaches to implementation.

Box 2.2: Randomised controlled trial for job seekers in the United Kingdom

In 2012, the United Kingdom Cabinet Office conducted a six-month randomised controlled trial with a Job Centre in Essex. The trial tested the impact of changing processes used with job seekers to help them find work. The trial tested the difference between the existing process and an alternative approach, which involved:

  • streamlining administration to ensure job seekers talked about getting back to work on their first day (not after two weeks); and
  • changing meetings to focus on future job search activity rather than checking previous efforts.

Within the job centre, clients were randomly assigned to a treatment group receiving the new processes and a control group which was subject to the existing processes.

The results demonstrated that job seekers in the treatment group were 15 to 20 per cent more likely than those in the control group to be off benefits 13 weeks after signing on (see Chart 2.13 below).

Chart 2.13: Proportion of jobseekers finding work after 13 weeks – results of randomised controlled trial

The chart shows that those in the treatment group were more likely than those in the control group to come off benefits 13 weeks after signing on in each of the trial/comparison months of June, July and August.

Source: United Kingdom Cabinet Office Behavioural Insights Team, 2012a

RCTs work by randomly dividing a population into two or more groups, giving one intervention to one group and another to a different group, and then testing for a pre‑specified outcome for each group. As the United Kingdom Cabinet Office’s Behavioural Insights Team (2012b) has noted:

What makes RCTs different from other types of evaluation is the introduction of a randomly assigned control group, which enables you to compare the effectiveness of a new intervention against what would have happened if you had changed nothing.

This is possible because random selection addresses two common issues for more traditional programme evaluation techniques: how to account for uncontrolled, external factors, such as strong economic growth; and selection bias – the fact that the very people who want to participate in a programme are systematically different to those who do not.

RCTs are used extensively internationally in medical research and international development. They are not used extensively in public policy more broadly, even though administrative information and systems required to randomly assign members of a particular population into groups, and assess outcomes, are often already in place.

RCTs could be applied to public policy problems in Australia, particularly with respect to social policy. For example, as noted in the Phase One Report, there is a critical lack of robust evidence and evaluation about the effectiveness of Indigenous programmes at all levels of government. RCTs could be usefully incorporated into the design of Indigenous programmes to ensure programmes will be effective in achieving desired outcomes.




Australian Competition and Consumer Commission 2013, Australian Competition and Consumer Commission and Australian Energy Regulator Annual Report 2012-13, Australian Competition and Consumer Commission, Canberra.
Australian Government 1996, Commission of Audit, Australian Government, Canberra.

Australian Government 2009, Strategic Review of Future Directions for Australian Government Service Delivery, Australian Government, Canberra.

Australian Government Information Management Office 2011, Interacting with Government Australians’ use and satisfaction with e-government services, Australian Government, Canberra.

Australian Public Service Commission 2007, Building Better Governance, Australian Government, Canberra.

Australian Public Service Commission 2009, Policy Implementation through Devolved Government, Australian Government, Canberra.

Australian Public Service Commission 2013a, State of the Service Report, State of the Service Series 2012-13, Australian Government, Canberra.

Australian Public Service Commission 2013b, Capability Review – the Treasury, Australian Government Canberra.

Deloitte Access Economics 2013, Reforming regulation of the Australian food and grocery sector, Deloitte Access Economics, Canberra.

Department of Health 2012, About the Office of Hearing Services, Australian Government, Canberra, viewed January 2014, http://www.health.gov.au.

Department of Human Services 2011, Service Delivery Reform: Transforming Government Service Delivery, Australian Government, Canberra.

Department of Human Services 2013, Annual Report 2012-13, Australian Government, Canberra.

Department of the Prime Minister and Cabinet 2013, Ten Years Strong – PM&Cs Cabinet Implementation Unit, Australian Government, Canberra.

Halligan, J 2008, The Centrelink Experiment - Innovation in Service Delivery, Australian National University Press, co-published with the Australia and New Zealand School of Government, Canberra.

MacDermott, K 2008, Whatever Happened to Frank and Fearless? The Impact of New Public Management on the Australian Public Service, ANU Press, co-published with the Australia and New Zealand School of Government, Canberra.

Organisation for Economic Co-operation and Development (OECD) 2012, Value for Money in Government: Australia 2012, OECD Publishing, Paris.

Productivity Commission 2014, Report on Government Services 2014, Steering Committee for the Review of Government Service Provision Australian Government, Canberra.

Sturgess, G 2012, Diversity and Contestability in the Public Service Economy, NSW Business Chamber, Sydney.

Sturgess, G 2014, Policy and Delivery – A Paper for the National Commission of Audit, correspondence and advice to the National Commission of Audit.

United Kingdom Government 2010, Spending Review 2010, Her Majesty’s Treasury,London.

United Kingdom Cabinet Office, Behavioural Insights Team 2012a, Blog: New BIT trial results: helping people back into work, Cabinet Office London.

United Kingdom Cabinet Office, Behavioural Insights Team 2012b, Test, Learn, Adapt: Developing Public Policy with Randomised Controlled Trials, Cabinet Office, London.