Australia is redesigning its employment services system, and it’s a real opportunity. But success won’t be decided by structure alone. It will depend on whether assessment, funding, streams, employer engagement and frontline delivery work together for real people. NESA opens up the design questions underneath its reform submission.
That picture is worth keeping in mind, because Australia is now redesigning the system that shapes it. Australia’s current employment services reform presents a significant opportunity to reshape how support is designed, funded and delivered, and to build a system that is more personal, more responsive and better connected to employers and local labour markets. NESA has just lodged its submission to the Australian Government’s reform process. Rather than repeat it here, I want to open up some of the questions underneath it, the ones that will decide whether the new model will work in practice for the people it is meant to serve.
A quick note for readers outside the sector: Workforce Australia is the Australian Government’s current mainstream employment services program, delivered through employment services providers and online services. It sits alongside other programs such as Inclusive Employment Australia, Transition to Work and the Remote Australia Employment Service.
It’s worth saying plainly that this reform is a good thing to be doing. The chance to build a system that is more tailored to individual need, less weighed down by administration and better linked to real jobs does not come along often. At NESA, we come to this as a partner in change. We have worked alongside governments, providers, employers and communities through nearly three decades of reform, and this has given us a strong foundation on which to build.
Reform tends to get discussed as structure: how many streams, who delivers what, which box a person belongs in. Those questions matter. But the deeper question is whether the parts connect: whether assessment, funding, service streams, digital tools, performance measures and frontline delivery pull in the same direction. The challenge is not simply whether we change the system, but whether those parts work together once they involve real people, in real circumstances, not matter where they live within Australia’s urban, regional or remote landscape.
The model proposed by the Government suggests there will be different levels of support for different levels of need, which makes sense in principle. The design question is what happens when people move between those levels. NESA supports differentiated service intensity, but our submission recommends that integrated delivery of Streams 2 and 3 should be the default, with flexibility in how providers demonstrate that integration. The reason is practical: support intensity should be able to change without disrupting the participant’s relationship with the organisation and practitioner who already knows them.
The evidence in the submission shows why this matters. Roughly one in six people who leave Workforce Australia return within a year, and the share of people in services for five years or more has grown from around 8% in 2010 to over 20% today. Barriers to work rise and fall. If every change in circumstances triggers a change of provider and requires someone to retell their story, engage with a new provider and rebuild trust, each of those moments becomes a point where they may quietly disengage.
Continuity is not a nice-to-have. Research across health, mental health and social services consistently finds that the consistency of the worker–participant relationship is among the strongest predictors of whether people stay engaged and get results. For people who have experienced trauma or institutional distrust, a trusted, steady relationship is often what makes disclosure and persistence possible at all. The design task is therefore to preserve continuity while still allowing service intensity to change as a participant’s circumstances change.
A system can be well designed and still fall over if it isn’t funded to do what it asks. Here, frontline experience and the evidence line up closely.
Departmental viability modelling for the current program, cited in NESA’s submission, found that up to twelve employment regions were unlikely to support even a single provider, and the model became loss making across a 10 year contract once inflation reached 4%. An Australian National Audit Office review also found that in every scenario examined providers ran monthly losses through the first year of operation. With indexation of only 6.8% every three years failing to keep pace with real costs, the evidence presented in the submission points to significant financial pressure across parts of the provider market.
The point here is not about money for its own sake. It’s a design principle: work out what delivering a good service actually costs, benchmark it against comparable government services, and fund that, rather than setting a budget envelope and hoping for the best.
There is also a genuine choice for government about what it is buying. Is it purchasing a prescribed activity, or an outcome? The current model does a bit of both: it sets the rules tightly but pays mainly on results, so providers are held responsible for outcomes shaped by settings they were required to follow. NESA’s preference is clear: a genuinely outsourced service that purchases outcomes and trusts providers to determine how best to achieve them, though either model, coherently funded, would improve on that in-between state. Either way, much of the unfunded administrative burden providers carry could be designed out at the same time.
How people are assessed and directed to support is one of the quiet decisions that shapes everything downstream. Get it right and people reach the help that fits. Get it wrong and some end up in light-touch or digital-only services when they actually need intensive, human support, and stay there.
The submission argues for an assessment process that avoids duplication and works across programs: a fast, evidence-informed government triage that determines entry into a stream, followed by a comprehensive, provider-led assessment, ideally conducted face to face, that identifies needs and circumstances that may not emerge during an initial triage. It also names a group worth watching, the “missing middle” of people who look straightforward on paper but are facing barriers that need more help to address than Stream 2 is generally intended for, yet they aren’t complex enough to meet the criteria for a Stream 3 level of support. Then there’s also the question of when whether digital services will have a clear pathway to a real person when a self-serve digital service just isn’t working for them.
It’s easy to talk about employment services as though the work is only about preparing jobseekers. It isn’t. A future model has to invest just as deliberately in the demand side, understanding employers, developing suitable roles, supporting workplace adjustments and maintaining relationships over time.
The submission is direct about this being the under-built half of the system. NESA recommends funding providers to build genuine employer-facing capability across all streams, acting as workforce-planning partners rather than routing employers to a new government or third-party broker. In practice that could mean a single trusted point of contact, help with workplace adjustments and onboarding, and continuing support after the initial hire, the practical support that gives employers confidence to recruit and retain the right person. Small businesses, in particular, often have limited recruitment and human resources capacity, and value a provider that can operate as an extension of their team.
The submission keeps returning to frontline experience for a reason. Implementation problems tend to be visible first to the people closest to delivery, participants, practitioners, employers and community organisations. They notice the referral that goes nowhere, a payment rule that creates unintended incentives, the assessment that misses the point. Consulting the frontline is not merely stakeholder management; it is one of the most reliable ways to identify design risks early, before they become embedded in the system.
Two themes in the submission are really about stewardship, the government’s ongoing role in keeping a national system healthy.
The first is technology. Artificial intelligence is already reshaping service delivery and, used well, can absorb administrative work and give frontline staff more time with people. NESA’s position is that AI should augment the practitioner relationship, not substitute for it. Decisions affecting a person’s service stream, payments or obligations must remain subject to meaningful human oversight, supported by clear governance and transparency about when AI is being used. Done thoughtfully, technology handles the paperwork and people deliver the service.
The second is the workforce, because you cannot run a relationship-based system without valuing the people who hold those relationships. The submission argues for demonstrated competence and applied capability, built through practical, portable micro-credentials, rather than relying on formal qualifications alone. Beyond training, a national Workforce Plan should also address workforce attraction, retention, safety, remuneration and the stability organisations need to invest in capable frontline teams.
If there is one implementation lesson worth holding onto, it’s that reforms of this scale should be tested before they are relied on everywhere. The submission makes a strong case for staging the change: publishing region-level market-viability and participant-flow modelling before going to market, and using pilots to test how assessment, funding, streaming and frontline practice interact before the model is implemented nationally. As the new model is developed, it will be important to test its structural assumptions while sequencing the broader reform carefully, rather than locking every element in at once and discovering too late that the joins do not hold.
If there’s a thread running through all of this, it’s that the system needs to work as a whole. Drawing on the submission, a handful of practical principles stand out:
These questions are intended to help strengthen the reform’s design, not to stand in its way. The opportunity in front of us is real, and getting the design right, so the parts hold together in practice, not just on paper, is how we make the most of it.
Reform of this scale will take continued collaboration. government, providers, employers, participants and communities each hold a different piece of the picture, and at NESA we will keep contributing evidence and frontline experience as the model develops. This submission is one contribution to a conversation that is far from over.
I invite you to read NESA’s full submission on employment services reform in Australia and share your perspective on the questions it raises.
Dr Marion Byrne BA LLM MDiv