There are many inexpensive solutions that won’t compromise their integrity
All of us who work in the administrative tribunal world know we have an access to justice problem. The Statistics Canada analysis of the 2021 Canadian Legal Problems Survey shows how big a problem it is.
We also know that governments won’t ride to the rescue with bags full of cash. If you were a politician, would you really put this issue ahead of climate change, health care, or clean water on First Nations’ lands?
There are solutions adjudicative tribunals can embrace right now without compromising their impartiality. Some cost nothing, and some are low cost. Most importantly, tribunals can implement many without asking anyone’s permission – not ministers, bureaucrats, or courts.
Here are my top four:
Active adjudication from start to finish, not just in hearings
Our adversarial process works with lawyers on both sides. It breaks down when one or both parties represent themselves. The courts have gradually recognized this. A common law duty has evolved: adjudicators must ensure unrepresented parties understand how a hearing works.
Beyond explaining pure process questions, many adjudicators now set the issue agenda in a hearing and take the lead in questioning witnesses. As the jurisprudence develops, the frontiers of active adjudication expand.
But they must expand well outside the hearing room. If a party does not know how to prepare their case beforehand, how much help is an empathetic, active adjudicator at the hearing? By then, it’s too late. Active adjudication must start when a case is opened and continue until a final order is issued.
Tribunals can provide navigators to guide unrepresented parties through the process. The federal Social Security Tribunal, and British Columbia’s Workers Compensation Appeals Tribunal and Human Rights Tribunal are already doing it. They have transformed tribunal registries from passive recipients of case information to active agents of assistance. Training registry officers to play a navigator role raises costs incrementally but is relatively affordable.
And the solution is not exclusively human. Some tribunals offer automated navigation, such as the BC Civil Resolution Tribunal’s Solution Explorer. The Justice Bot developed by the Cyberjustice Lab at Université de Montréal displays the rapidly increasing capacity of artificial intelligence to deliver customized legal information to litigants. Tribunals need to start developing both human and machine resources to fight the inequities of the static, adversarial model.
You can’t participate meaningfully in your own case if you can’t understand what you’re expected to do. Tribunals need to start looking now at the ability of large language models to rewrite text in plain language. There is enormous scope for a tribunal to harness AI to rewrite its public-facing materials: rules of procedure, forms and letters. This can be done at a low cost and great speed.
But we should proceed with caution. Otherwise, it’s “garbage in – garbage out.” The tribunal needs to critically assess what a chatbot spits out. It must retain control of the pen.
First, tribunals need to learn what plain language means. The common notion that it’s just about short words in short sentences is wrong. But the training investment in understanding the interrelated workings of structure, wording and design is cheap. Second, how a query is framed for an AI chatbot makes a big difference to the quality of its output, so some thought on how to pose the task matters.
Finally, tribunals should consider whether and how large language models can be harnessed to help edit (not draft) decisions for plain language. There are ethical and legal risk issues around deliberative secrecy and the confidentiality of information in draft decisions. But we need to figure out how to mitigate them, and this conversation has yet to begin in the tribunal world meaningfully.
Most tribunals have been designed through a top-down exercise. It starts with the tribunal’s statutory mandate, general administrative law principles, and the tribunal’s powers and rules. Decisions around process design and the resourcing of the tribunal then follow. While this may benefit justice system insiders, the legacy of this approach is our current access to justice challenge.
Tribunals have a better chance of pre-empting access to justice problems if they make user needs the starting point in designing processes. The costs of looking outward when making design choices are minimal. There are a lot of excellent free resources – the Stanford Legal Design Lab stands out. This is less about money than about a culture shift. Still, it demands time and effort to reach and engage users. It also depends on managing user expectations about what a tribunal is capable of, legally, operationally and financially.
Honestly, who reads a tribunal’s annual report? What do users want from an administrative justice service? They want to know how the tribunal works and how long they will have to wait for their case to be heard and resolved. Tribunals can’t offer guarantees, but they can give users a realistic idea of how long they will have to wait for a written decision. Every tribunal has a (usually) small minority of complex – and therefore unpredictable – cases, but this does not justify refusing to tell people what to expect in the great majority of cases. Tribunals have the data. It’s time to start being upfront about what users can expect from a justice service. Their taxes pay for it.
Trust in public institutions matters. It’s fundamental to democracy. So is access to justice. A tribunal that starts now will see results soon enough. It’s time for Tribunal 2025.
