The duty lawyer model works because it is unbundled
A grant of aid is close to all-or-nothing: carriage of the matter, or nothing. The duty lawyer service is the opposite — advice and appearance on the day, no means test, no grant — and it is the most efficient part of the system.
Steps this would change
Rationale
The legal assistance process contains two very different models sitting side by side.
The grant model. Apply, pass a means test, pass a merits test, and receive carriage of the matter — or receive nothing. Three tests, each capable of ending the application, and no intermediate outcome.
The duty lawyer model. Turn up at court, be seen on the day, get advice and an appearance. No application, no means test, no grant, no ongoing carriage. It is modelled here at one day, and it is generally regarded as the highest-value service in the system per dollar spent.
The difference is that duty work is unbundled: it provides one discrete piece of legal help rather than the whole matter. That is why it can be given to everyone who turns up, and why it does not need the rationing machinery the grant model requires.
The idea is that discrete-task assistance be funded as a service in its own right beyond the courtroom door: drafting one document, advising on one decision, appearing at one listing, negotiating one settlement — available on a lighter test than a full grant, to people who currently receive nothing because they cannot be given everything.
For a great many problems in the turn-away population — a tenancy notice, a debt letter, a Centrelink review, a first family violence listing — one piece of competent help at the right moment is the difference between an outcome and a default judgment.
Filed as an idea rather than a proposal, because the professional and practical objections below are substantial and this project cannot resolve them.
What is being traded away
Unbundled legal work carries real professional risk, and lawyers know it. Advising on part of a matter without knowing the whole of it is how negligence happens. A practitioner who drafts one document without carriage may miss a limitation period, a cross-claim, or a family violence risk that only emerges from the full file. Professional indemnity insurers and regulators have been cautious about limited-scope retainers for exactly this reason, and that caution is not restrictive practice.
It may entrench a two-tier system. "One piece of help for the poor, full representation for those who can pay" is a defensible response to scarcity and an uncomfortable thing to build deliberately. The risk is that discrete-task assistance becomes the ceiling for legal aid rather than an addition to it — cheaper per person, worse per outcome, and politically easier to fund than the full grants it would partly displace.
The people who benefit most from full representation are the least able to use unbundled help. Discrete-task assistance assumes a client who can act on advice, manage their own matter between interventions, and recognise when to come back. That describes some of the turn-away population and not the most vulnerable part of it — people with cognitive disability, acquired brain injury, or in acute family violence crisis need someone to carry the matter, not to coach them through it.
And the saving is small and possibly illusory. Thirty days on a 130 day path that is short because it ends in refusal. If unbundling increased the number of people helped without increasing funding, the help each receives gets thinner — which is rationing by another name, and this database should say so rather than present it as efficiency.
Precedent
Duty lawyer services already operate this way at courts across the country, without a grant and without a means test, and are widely regarded as the highest-value part of the legal assistance system per dollar.
Sources
- Duty lawyer services operate without a grant of aid or a means test
- Legal aid grants assessed against means, merits and jurisdictional tests
- Community Legal Centres Australia: more than 1,000 people turned away daily
