Did Dunsmuir Simplify the Law on Judicial Review?

My aversion to the Supreme Court of Canada’s recalibration of administrative law doctrine in Dunsmuir v. New Brunswick is well known: in my view, the Dunsmuir approach simply sweeps a whole lot of substantive analysis under a formal rug.

But I thought I would nonetheless ask the question: did Dunsmuir simplify the law on judicial review?

Answering this question is not easy: what does it mean to say that a decision “simplified” the law? One means, admittedly a very crude means, is to measure the length of decisions. Of course, the length of a decision does not always reflect the difficulty of reaching it. As the adage goes: I wanted to write a short note, but I didn’t have the time. Nonetheless, common sense suggests that longer decisions are a sign that the concepts at issue are difficult. Certainly Dunsmuir responded to a concern that the law was too complex, with its multi-factor test and multiple standards of review, something that would naturally manifest itself in longer judgments.

So I looked (more accurately, I had a student look*) at the Court’s decisions in two periods: 2003-2008 and 2008-2013. They are apt for comparison because the old ‘pragmatic and functional’ framework was stable in the first period and the second period came right after the Dunsmuir recalibration. The sample size is small, but there are roughly the same number of cases in each period, with roughly the same number of dissents and concurrences. There was some doctrinal work to do in both periods: for Khosa (2009) and Newfoundland Nurses (2011) in the second period, read LeBel and Deschamps JJ concurring reasons in CUPE (2003) and Via Rail (2007).

The total word count in 2003-2008 was 369,341, with an average of 18,482 per decision

The total for 2008-2013 was 299,846, with an average of 14,278 per decision.

That’s a difference of 69,495 words, which is 18% of the 2003-2008 total. Based on the averages, the decrease is more impressive (22%).

I thought it possible that the law might have become more complex between 2003 and 2013. Apparently not. My numbers are broken down under several headings: facts, law, parties’ submissions, administrative law principles, analysis and conclusion. Any increased complexity would show up in the facts and the law, but even taking these into account, the decease is dramatic.

Some caveats: I still have to check whether the Court’s decisions in general are shorter. A preliminary look suggests that they might be, but I need more data.

I can’t seem to upload my Excel worksheets onto this platform yet. I will try to do so.

At some point, I will have to extend the study to other Canadian courts. The Supreme Court is not necessarily representative of what is going on across the country.

Lastly, I am the living embodiment of the adage that lawyers are relatively intelligent people who are terrified of blood and numbers, so there may be some mathematical howlers even in this very basic analysis. I’m happy to be corrected!

* Thanks to Pablo Laplanche for research assistance.



This content has been updated on October 8, 2014 at 21:45.