Bail Pending Appeal

R. v. Farinacci — Ontario Court of Appeal — Justice Arbour J.A. (Dubin C.J.O., Catzman, Osborne and Doherty JJ.A. concurring) — November 8, 1993 | [1993] O.J. No. 2627, 86 C.C.C. (3d) 32

Background

Leonard Farinacci, Guy Paul Morin, and William Wade were each convicted of serious criminal offences and sought bail pending appeal under s. 679 of the Criminal Code. Each had been denied release at the single-judge level. Farinacci brought a review before the full court pursuant to s. 680(1), and Morin and Wade were granted intervener status, with their own bail applications held pending resolution of the constitutional question Farinacci raised.

The constitutional question was this: in the wake of the Supreme Court's 1992 decision in R. v. Morales, which struck down the words "in the public interest" from s. 515(10)(b) of the Criminal Code as unconstitutionally vague in the pre-trial bail context, could the identical language in s. 679(3)(c) — governing bail pending appeal — survive? The applicants said no. The Crown said the contexts were fundamentally different. The Attorney General of Canada intervened on the Crown's side.

Issues

1. Does s. 11(e) of the Charter — the right not to be denied reasonable bail without just cause — apply after conviction?
2. Does a constitutional right to bail pending appeal exist under ss. 7 or 9 of the Charter?
3. Is the term "public interest" in s. 679(3)(c) of the Criminal Code unconstitutionally vague?

The Legal Framework

Section 679(3) of the Criminal Code permits a judge of the court of appeal to release a convicted appellant pending the determination of their appeal if the appellant establishes three things: (a) the appeal is not frivolous; (b) they will surrender into custody as required; and (c) their detention is not necessary in the public interest. The applicants accepted the burden structure. Their target was criterion (c).

In Morales, the Supreme Court had found "public interest" in s. 515(10)(b) unconstitutionally vague — it lacked ascertainable boundaries, permitted a standardless sweep, and failed to give sufficient guidance for legal debate, all in a context where s. 11(e) of the Charter was directly engaged and where Parliament had already carved "public safety" out as a separate ground, leaving "public interest" without intelligible residual content.

The applicants argued the same reasoning applied to s. 679. Arbour J.A., writing for a unanimous five-judge panel, disagreed at every level of the analysis.

The Court's Analysis

The starting point was s. 11(e). The rights in s. 11 of the Charter belong to persons "charged with an offence" — they are pre-trial and trial rights, exhausted by verdict. A conviction is not a presumption of guilt; it is a conclusive legal finding of guilt, enforceable unless reversed. The presumption of innocence that anchors the right to reasonable pre-trial bail is substantially spent by the conviction. Section 11(e) does not follow the accused through the appellate process.

On ss. 7 and 9, the court declined to resolve whether a constitutionally entrenched right to bail pending appeal existed, choosing instead to assume — for the purpose of the application only — that such a right existed, and to assess s. 679(3)(c) against that assumed right. Even on that generous assumption, the analysis led to the same place.

The critical distinction from Morales operated on two levels. First, the constitutional context had shifted: without s. 11(e) in play, the vagueness of "public interest" had to be measured against the demands of reviewability and enforceability, not against the presumption of innocence. Second, unlike the legislative structure in s. 515 — where "public safety" and "public interest" sat side by side, leaving the latter without intelligible independent content — s. 679 contained no parallel drainage. "Public interest" in the appellate context had a coherent, judicially developed meaning that had been consistently applied.

That meaning, drawn from a line of cases including R. v. Demyen and R. v. Pabani, reflects the inherent tension between two legitimate demands: the enforceability of judgments and the reviewability of convictions. Public confidence in the justice system requires both. A serious offender advancing weak grounds of appeal may properly be denied bail because enforceability should not yield. A first offender facing a short sentence who would complete it before an appeal is even heard may equally and properly be granted bail, because denying it would render the statutory right of appeal a nullity. That is the balancing exercise s. 679(3)(c) demands — and it is a principled exercise, not a standardless one.

Arbour J.A. also noted the practical safety valve: where the balance is close, a court may often do justice to both interests by granting expedition of the appeal rather than granting bail. The statute supports that approach.

Decision

Appeal dismissed. Section 679(3)(c) is constitutionally valid. The expression "public interest" in the post-conviction, bail-pending-appeal context is not unconstitutionally vague. Farinacci's application for bail was denied on the merits.

Key Takeaways

Farinacci is the definitive authority on bail pending appeal in Ontario — and it draws a line that matters every time a convicted accused argues the system should treat them the same as someone awaiting trial.

It doesn't. The presumption of innocence is not a renewable resource. It does real constitutional work before and during trial, and it is the foundation of the right to reasonable pre-trial bail under s. 11(e). Once a jury or judge returns a finding of guilt, that presumption is spent. The conviction is a legal fact. What follows is an appellate process with its own internal logic — and its own constitutional demands, centered not on innocence, but on the competing imperatives of enforcing judgments and correcting errors.

The "public interest" standard in s. 679(3)(c) survives because it has been given coherent content by decades of appellate jurisprudence. Judges are not operating in a vacuum. They are weighing the strength of the appeal, the nature of the offence, the sentence imposed, the projected appellate timeline, and the risk that denial of bail renders the appeal pointless in practical terms. That is a principled exercise. It can be reviewed. It can be criticized. It can be wrong in individual cases. But it is not a standardless sweep — and that is the constitutional threshold.

For anyone defending a bail-pending-appeal application, Farinacci is the map. The stronger the grounds of appeal, the longer the projected wait, and the more the sentence will be served before the court ever hears the case — the stronger the argument that the public interest favours release. Get those three variables in front of the judge clearly, and the framework works in your favour.

R. v. H.K. Court of Appeal for Ontario — Justice Gillese J.A. (Motion Judge) August 5, 2022 | 2022 ONCA 578 | Docket M53658 (C70924)

Background

H.K. is a 67-year-old Canadian citizen with no criminal record. He immigrated in 1981, became a citizen in 1986, and built his life here — home, family, companies, community. Everything he has is in Canada. He is not a flight risk by any measure.

He was convicted and sentenced to ten months' imprisonment. He appealed both conviction and sentence. Then he applied for release pending appeal. The Crown opposed it, arguing his appeal had no merit and that detention was necessary in the public interest.

Gillese J.A. heard the application by video conference on August 4, 2022, granted it from the bench, and promised brief written reasons. This is that decision.

The Issue

Whether H.K. should be released pending appeal under the three-part test: that the appeal is not frivolous, that he will surrender into custody, and that his detention is not necessary in the public interest.

The Analysis

The flight risk question was the easiest part of the analysis. H.K. has no criminal record, deep roots in Canada, and nothing waiting for him anywhere else. There was no serious argument that he would fail to surrender. That prong was answered before it was asked.

On the merits, Gillese J.A. rejected the Crown's position that the appeal was without merit. The appeal more than clears the "not frivolous" threshold — a deliberately low bar that requires only that the grounds raised are arguable, not that they will succeed. The sentence appeal was the stronger branch: there was a real and compelling argument that relevant evidence of extraordinary hardship may not have been properly considered when the sentencing judge determined whether a conditional sentence was appropriate. That is a live issue, not a manufactured one.

The public interest analysis is where the decision does its most important work. H.K. received a ten-month sentence. Under that sentence, he would become eligible for parole after approximately 3.5 months and eligible for release on earned remission after approximately 6.5 months. The time between the application and the likely hearing and disposition of his appeal would almost certainly exceed that window. In plain terms: if the application were denied, H.K. would likely finish serving the custodial portion of his sentence before his appeal was ever decided.

That makes the sentence appeal moot. And mootness is not a technicality — it is a substantive injustice. The entire purpose of a sentence appeal is to obtain meaningful relief from a sentence the accused argues is wrong in law or principle. If the sentence has already been served by the time the appeal is heard, there is nothing left to remedy. The court cannot give back time. Releasing H.K. pending appeal was the only way to preserve the appeal's functional purpose.

That risk of mootness alone, Gillese J.A. held, weighs heavily in favour of release. Reinforced by the genuine strength of the sentence appeal grounds, the balance was clear.

A publication ban was also ordered at the Crown's request under ss. 486.4(2.2) and 486.5(1) of the Criminal Code. H.K. did not oppose it.

Decision

Application granted. Release ordered in accordance with the draft release order provided to the court. Publication ban imposed.

Key Takeaways

H.K. is a clean, efficient application of the release pending appeal framework, and its most instructive lesson is about mootness as a factor in the public interest analysis.

The public interest component of the test is not one-dimensional. It doesn't just ask whether the public has an interest in seeing sentences enforced. It also asks whether there is a public interest in the integrity of the appellate process itself - and there plainly is. An appeal that is rendered meaningless by the passage of time is not justice delayed. It's justice eliminated.

Where a sentence is short and an appeal is genuinely arguable, the math often runs against detention. By the time the court can hear and decide the matter, the custodial portion may already be served. At that point, releasing the accused pending appeal isn't a favour to the accused — it's the only way to keep the appellate remedy alive.

Read H.K. alongside Antic and the through-line is consistent: liberty is the default, restrictions require justification, and courts must be alert to the ways procedural mechanics can quietly swallow substantive rights. Keeping H.K. in custody pending appeal wouldn't have served justice. It would have made the appeal pointless. That's not a public interest. That's a structural failure — and Gillese J.A. declined to let it happen.