A Quantitative Analysis of Leave Applications to the Supreme Court of Canada in Constitutional Cases

Authors

Paul-Erik Veel, Lenczner Slaght LLP

Abstract

The requirement that a party must, in most cases, obtain leave to appeal in order to bring a case to the Supreme Court of Canada poses a significant barrier to cases being heard by that Court; indeed, the vast majority of cases are not granted leave to appeal. Because the leave process operates as a barrier, understanding the leave to appeal process and outcomes are in turn important to understanding the development of Canadian law. However, because the Supreme Court does not provide reasons for decision as to whether or not to grant leave in a particular case, conventional tools of legal analysis provide us with little insight into its decision-making. Quantitative analysis of leave applications can provide insights about leave applications, including the determinants of successful leave applications.

Citation Information

Veel, Paul-Erik. "A Quantitative Analysis of Leave Applications to the Supreme Court of Canada in Constitutional Cases." The Supreme Court Law Review: Osgoode’s Annual Constitutional Cases Conference 115. (2024).
DOI: https://doi.org/10.60082/2563-8505.1458
https://digitalcommons.osgoode.yorku.ca/sclr/vol115/iss1/16

Creative Commons License

This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

References

1 Supreme Court Act, R.S.C. 1985, c. S-26, s. 53.

2 These include provincial references determined by provincial Courts of Appeal, as well as various criminal appeals. See Supreme Court Act, R.S.C. 1985, c. S-26, s. 36 Criminal Code, R.S.C. 1985, c. C-46, ss. 691-693.

3 Supreme Court Act, R.S.C. 1985, c. S-26, s. 40. There are also provisions by which the provincial courts and the Federal Court of Appeal can grant leave for a case to be heard by the Supreme Court of Canada: Supreme Court Act, R.S.C. 1985, c. S-26, ss. 37 and 37.1. However, this power is rarely invoked.

4 Supreme Court of Canada 2022 Year in Review, online: https://www.scc-csc.ca/reviewrevue/2022/yr-ra2022-eng.pdf.

5 Supreme Court Act, R.S.C. 1985, c. S-26, s. 40(1). https://doi.org/10.1086/scr.1985.3109496

6 Alex Bogach, Jeremy Opolsky & Paul-Erik Veel, "The Supreme Court of Canada's From-The-Bench Decisions" (2022) 106 S.C.L.R. (2d) 251.

7 Supreme Court of Canada 2022 Year in Review, online: https://www.scc-csc.ca/reviewrevue/2022/yr-ra2022-eng.pdf.

8 Supreme Court of Canada 2022 Year in Review, online: https://www.scc-csc.ca/reviewrevue/2022/yr-ra2022-eng.pdf.

9 See, e.g., Bertha Wilson, "Leave to Appeal to the Supreme Court of Canada" (1983) 4:1 Advoc. Q. 1; Geoff R. Hall, "Applications for Leave to Appeal: The Paramount Importance of Public Importance" (1999) 22:1 Advoc. Q. 87; Bruce Ryder & Taufiq Hashmani, "Managing Charter Equality Rights: The Supreme Court of Canada's Disposition of Leave to Appeal Applications in Section 15 Cases, 1989‐2010" (2010) 51 S.C.L.R. (2d) 505 https://doi.org/10.60082/2563-8505.1203; Denise Cooney, "An Absence of Reason: Why the Supreme Court of Canada Should Justify Dismissing Applications for Leave to Appeal" (2012) 70:1 U.T. Fac. L. Rev. 41; Donald J. Netolitzky, "The Walking Wounded: Failure of Self-Represented Litigants in 2017 Supreme Court of Canada Leave to Appeal Applications" (2021) 58:4 Alta. L. Rev. 837 https://doi.org/10.29173/alr2654; Vanessa A. MacDonnell, "Justice Suzanne Côté's Reputation as a Dissenter on the Supreme Court of Canada" (2018) 88 S.C.L.R. (2d) 45 (suggesting that the leave process may have changed as a result of Côté J.). https://doi.org/10.60082/2563-8505.1360

10 Donald J. Netolitzky, "The Walking Wounded: Failure of Self-Represented Litigants in 2017 Supreme Court of Canada Leave to Appeal Applications" (2021) 58:4 Alta. L. Rev. 837. https://doi.org/10.29173/alr2654

11 Part I of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (U.K.), 1982, c. 11 [hereinafter "Charter"].

12 Bruce Ryder & Taufiq Hashmani, "Managing Charter Equality Rights: The Supreme Court of Canada's Disposition of Leave to Appeal Applications in Section 15 Cases, 1989‐2010" (2010) 51 S.C.L.R. (2d) 505. https://doi.org/10.60082/2563-8505.1203

13 All of the data in our database is obtained by a human review of publicly available sources. The primary sources consulted are: (1) the Supreme Court of Canada's news releases, online: https://scc-csc.lexum.com/; (2) case information from the Supreme Court of Canada's website, online: https://www.scc-csc.ca/; and (3) decisions of lower courts. We also obtain information about lawyers acting for parties from various publicly available sources, such as lawyers' websites and LinkedIn profiles.

14 In the majority of appeals as of right, the Supreme Court's jurisdiction arises from the fact that the case is a criminal law matter where there was a dissenting judge at the Court of Appeal. However, in those cases, the Supreme Court will only hear an appeal as of right on the particular issue(s) on which a judge dissented at the Court of Appeal. Consequently, if an appellant who has an appeal as of right wishes to raise issues other than the issues on which a judge dissented, they must seek leave to appeal in respect of those issues.

15 For example, in the case of Trial Lawyers Assn. of British Columbia v. British Columbia (Attorney General), [2022] S.C.C.A. No. 262 (S.C.C.), the Supreme Court's summary was as follows: Constitutional law - Courts - Constitutional law - Courts - Tribunals - Provincial jurisdiction over administration of justice - Role of superior courts - Provincial legislation granting the Civil Resolution Tribunal jurisdiction over certain disputes relating to accidents caused by vehicles - Tribunal given exclusive jurisdiction to determine whether an injury is a minor injury and specialized expertise to determine liability and damages for claims of up to $50,000 - Is the grant of jurisdiction to the Tribunal to decide certain tort cases, pursuant to s. 133 of the CRTA and regulations, contrary to s. 96 of the Constitution Act, 1867, by conferring a judicial power exclusively within the jurisdiction of the superior courts at the time of Confederation, contrary to the principles set out in Reference re Residential Tenancies Act, 1979, [1981] 1 S.C.R. 714 - Is that same grant of jurisdiction contrary to s. 96 by infringing upon the core jurisdiction of the superior court, contrary to the principles set out in Reference re Code of Civil Procedure (Que.), art. 35, 2021 SCC 27 - Civil Resolution Tribunal Act, S.B.C. 2021, c. 25 ("CTRA"), s. 133 - Insurance (Vehicle) Act, R.S.B.C. 1996, c. 231 (IVA) - Constitution Act, 1867, ss. 92(14), 96.

16 The p-value of this difference using a one-tailed difference in proportions test in the percentage of constitutional cases vs. non-constitutional cases granted leave is p= 0.0003, indicating a high probability that this difference is a real one rather than a product of random fluctuations in the data.

17 The p-value of this difference using a one-tailed difference in proportions test is p=0.00029.

18 These values total up to 437, while there are only 434 cases that raise constitutional issues. This is because if a case raises both Charter and federalism issues, it will be counted in both fields.

19 On dialogue theory in the Charter context, see, e.g., Peter W. Hogg & Allison A. Bushell, "The Charter Dialogue Between Courts and Legislatures" (1997) 35:1 Osgoode Hall L.J. 75 https://doi.org/10.60082/2817-5069.1612; Peter W. Hogg, Allison A. Bushell Thornton & Wade K. Wright, "Charter Dialogue Revisited - Or 'Much Ado About Metaphors'" (2007) 45:1 Osgoode Hall L.J. 1. https://doi.org/10.60082/2817-5069.1254

20 A logistic regression need not be limited to a binary outcome in the dependent variable. There are logistic regression models that can accommodate more than two outcomes in the dependent variable. However, because the number of remands is very low, and because the factors driving a case getting remanded are likely very different from the factors that drive whether a case is likely to get leave, we exclude remands here for simplicity.

21 This potential for overfitting is further exacerbated where (as here) the outcomes in the dependent variable are imbalanced. In this case, the vast majority of cases involve a leave application being dismissed rather than granted, rather than something closer to a 50/50 balance.

22 In contrast to a simple measure like accuracy, ROC curves and AUC curves are insensitive to disparities in the proportions of the different classes: Max Kuhn & Kjell Johnson, Applied Predictive Modeling (New York: Springer Science+Business Media, 2013) at 264.

23 D.W. Hosmer & S. Lemeshow, Applied Logistic Regression, 2d ed. (New York: John Wiley and Sons, 2000), c. 5 at 160-164. https://doi.org/10.1002/0471722146

24 See, e.g., Asaf Bitton & Thomas Gaziano, "The Framingham Heart Study's Impact on Global Risk Assessment" (2010) 53:1 Prog Cardiovasc Dis. 68. https://doi.org/10.1016/j.pcad.2010.04.001

25 Because each of the three models uses different metrics to assess the importance or significance of a particular variable, a direct comparison between the models is not possible. However, the chart below provides a qualitative indication of how each model treated various variables, based on whether the model identified a feature as statistically significant (in the case of the logistic regression) and whether and to what extent the model used the feature (in the case of the decision tree and xgboost models).