What Do Human Rights Tribunals Actually Do?

And why do so few of their cases ever reach a conclusion?

We may not all agree on a single, complete definition of “human rights”, but there’s probably plenty about which there’s no debate. No one should ever lose a job solely because of race or sex, for instance.

Since the potential for such discrimination exists, there should be mechanisms through which injured parties can seek justice that don’t require the time and costs associated with civil courts. That, I assume, was the original inspiration behind the creation of human rights tribunals across Canada.

On the one hand, keeping up with the demand that’ll be attracted to any “free” service can be a challenge. As Tribunal Watch Ontario recently noted, the Ontario Human Rights Tribunal (HRTO) has failed pretty spectacularly on that count:

“In 2023/24, the HRTO dismissed over 1,380 applications on ‘jurisdictional and procedural’ grounds, without affording the applicant an opportunity to make oral submissions at a summary hearing, up from 1,124 such dismissals in 2022/23.”

But such systems - just because they’re easier to access - are particularly vulnerable to abuse. Any time you lower the barriers to entry for litigation, you’re bound to attract scammers and grifters. And any system requiring a relatively light burden of proof must, by force, rely on human discretion for rulings - which will always produce higher rates of bias and false convictions.

The Ontario tribunal has attracted it’s share of controversy. In a much-publicized 2008 case against Macleans Magazine, they eventually ruled that claims of “Islamophobia” in certain published articles fell outside the tribunal’s mandate. But the case was built on (to put it charitably) fragile legal foundations and consumed significant legal and process costs for the magazine and its authors.

None of which is what this particular post is really about. Instead, we’re here today to enjoy some data-driven insights into the way HRTO really works. All this, by the way, is thanks to the Canadian Legal Information Institute - a public service provided by the Federation of Law Societies of Canada whose goal is making Canadian court operations data freely accessible to developers. Free data: what’s not to love?

After briefly reading through a few case summary pages on the HRTO site, I began to see hints of interesting patterns, but I realized that useful observations would require visualizing the big-picture. For that, I’d need to take a few steps back.

So I pulled the metadata associated with a (fairly) randomly-selected set of 239 cases. No less than 86 of those cases (36 percent) had one of these three words included as keywords:

• Undeliverable

• Abandoned

• Unresponsive

Many - perhaps most - of the indicated full case descriptions showed that the applicants had simply dropped out of contact with the tribunal and were no longer responding to emails and phone calls. The “decisions” in such cases will often read:

“The applicant is deemed to have abandoned the Application.”

Some of the “orphaned” cases, however, seemed to result from a failure of the respondents to reply to any attempts by the tribunal to establish communication. There are examples where the tribunal claims to have reached out through street addresses and email addresses provided by the application but received no response.

In such cases there’s the risk that a malicious applicant might be setting up a victim with a spurious claim associated with false contact information. It doesn’t take much of an imagination to guess where that could end.

But it’s also possible that many targets of HRTO just can’t be bothered to respond. While in theory, the tribunal has the power to compel participation - even through the courts - it seems that that rarely happens.

My guess is that some respondents are counting on the tribunal’s crazy case backlog, coupled with similar wait times for the attention of real courts, and are just ignoring the system altogether.

I’m not sure I’d want to bet my future on those assumptions, but the results point to a system that’s at least partially dysfunctional.