The “Good Character” Problem

The recent appeal decision AA v Law Society of Ontario upheld the Law Society Tribunal’s 2023 decision to licence to applicant “AA” after finding him to be of “good character”—even though AA had admitted to have sexually abused three young children in 2009 (and to hiding this information from the Law Society in an earlier licensing application, which he withdrew in 2017 following an anonymous tip disclosing the abuse).

The AA case and other good character hearings stemming from sexual misconduct involving minors have generated considerable discussion both inside and outside the legal profession about how law societies should assess licensing applicants’ good character—and whether we should assess it at all.[1]

I pause here to note: I had been working on a column about the good character requirement more generally when the AA appeal decision was released and I decided it would be more interesting and timely to focus on the particular issue it raises about public confidence in the legal profession. So, I will be writing about the good character requirement in a two-part series, with my next column speaking more generally to its purpose, process, and problems, as well as potential ways forward.

For the past several years, I have assigned the 2013 decision in LSUC v Melnick as required reading to the upper-year students in my Legal Ethics class at Osgoode. That matter concerned an applicant whose teaching licence had been revoked after the 28-year-old middle school teacher engaged in a sexual relationship with a 14-year-old former student in 2004 (he applied to and was accepted to law school before pleading guilty to two criminal charges in 2006, and began law school in 2007 while still serving his non-custodial sentence).[2] The Tribunal’s hearing panel initially found that the applicant had not established he was of good character, but this was reversed by the Tribunal’s appeal panel, which held that the hearing panel had “misapprehended or failed to appreciate the totality of the evidence respecting the applicant’s good character” such that its findings were “unreasonable”. The appeal panel concluded that the applicant “had overcome a gross dereliction of his responsibilities and established his present good character” [emphasis added].[3]

Perhaps unsurprisingly, various media outlets were critical of the Melnick appeal decision, reporting it with headlines such as “Sex offender allowed to practice law in Ontario after he lost teaching licence over child luring conviction” and “London-area teacher who seduced his 14-year-old student now of ‘good character’ to law society”.

Ten years later, we are witnessing the same phenomenon, not only with the AA case but also Tribunal’s recent decisions in Colangelo that granted licensure to a paralegal applicant who similarly had lost her teaching licence after pleading guilty to child luring (this time, in connection with sexually explicit messages and photos she had shared with 16- and 17-year-old students at the high school at which she worked. Sample headline: “Child sex offender can practise as Ontario paralegal”).[4]

The LSO has already announced that it will seek judicial review of the AA decision before the Divisional Court, and seek a stay of the order precluding AA’s licensure until the matter is finally resolved.

If we gauge public opinion based on reader comments on the Toronto Star website (e.g. “How can this man be of good character and he sexually abused children?” and “This decision is a travesty and I strongly hope that the Law Society wins the judicial review”),[5] it appears the public agrees AA should not be a licensed lawyer and supports the Law Society’s decision to keep pushing back through escalating layers of appellate review.

These cases invite the question: why do we bother having a “good character” requirement if we are going to grant licences to former sex offenders (etc.) and anger the public—and perhaps erode public confidence in the legal profession—anyway?

In my opinion, while there are very good reasons for the imposition of a good character requirement, it is all but impossible to apply such a requirement meaningfully and fairly.

More specifically, I suggest it is not possible to apply the good character requirement in a way that will effectively promote public confidence in the legal profession while also holding firm to the ideals of our profession and of our justice system in a free and democratic society.

Although this is not how the Law Society framed its arguments, it seems to me that public outcry in matters like AA, Melnick, and Colangelo is at its core premised in the belief that people who committed certain offences should simply never be granted the privilege of a law licence.

The problem with such a rule, of course, is that it flies in the face of a fundamental tenet of our justice system that rehabilitation is possible. If our law societies were to adopt or apply such a rule, they would be tacitly undermining or objecting to this important principle.

Indeed, good character jurisprudence is clear that the question to be determined is whether the applicant is currently of good character—and expressly (and repeatedly) emphasizes that what is to be assessed is the applicant’s character at the time of the hearing, and not their prior character.

The Law Society’s arguments on good character matters make clear that they understand and accept the governing principles. Rather than argue that the test should be different, or that an applicant can never earn the public’s confidence because of their past conduct, law societies tend to argue that, as a factual matter, the applicant has not yet demonstrated that they have been rehabilitated or are remorseful for their past conduct.

In the AA and Colangelo matters, the hearing panel disagreed with the LSO’s factual arguments and accepted the applicants’ evidence that they deeply regretted their past misconduct and they were no longer the person who committed those misdeeds. In both cases the LSO appealed, and the appeal panel deferred to the factual findings made below and upheld the conclusion that the applicants were of good character. In both cases, the LSO sought or is seeking judicial review.

Based on these examples, I am becoming concerned that the LSO may be pursuing good character matters not because there is necessarily an arguable legal issue to be decided, but because of the optics with the public. That is, I question if the LSO is appealing these matters not because they have a reasonable chance of success on an appeal on the merits, but because it wishes to be seen by the public as taking a hard line and trying to keep apparently unsavoury characters out of the profession (regardless of the applicable legal test or whether public opinion is reasonably based or consistent with our legal system’s commitment to the possibility of rehabilitation).

To elaborate: The LSO’s principal argument before the Tribunal’s Appeal Division in the AA matter was that the hearing panel had misapprehended the evidence in arriving at its conclusion that AA was of good character. It did not suggest that the hearing panel had applied the wrong test for good character or had otherwise made an error in principle.

Accordingly, the Appeal Division’s reasons are predominantly about why it was deferring to the findings of fact made by the hearing panel below, which are subject to the high “palpable and overriding error” standard of appellate review. It held:

  • “On all of the points raised by the Law Society, what we have is contested evidence with respect to findings of fact… evidentiary review was the hearing panel’s task, not the role of the appeal panel.”[6]
  • “…there was no palpable error. In assessing misapprehension of evidence, it was not our role to ‘dissect, parse or microscopically examine the reasons’ of the fact-finder. The misapprehension of evidence must be plainly seen.”[7]
  • “In addition, it is not at all clear that there was any misapprehension of the evidence that affected the result.”[8]

Given the nature of the issues on the appeal and the Appeal Panel’s cautious and deferential approach to them, it seems to me the LSO will face an uphill struggle in seeking to have AA’s good character finding overturned by the Divisional Court.

To be clear, I think the LSO was right not to argue the appeal on the basis that Hearing Panel had made a legal error (which would have been subject to the less stringent “correctness” standard of review). There simply wasn’t an argument there; the Hearing Panel had applied the correct test for assessing a licensing applicant’s good character, which is well established in the Tribunal’s jurisprudence.

Interestingly, however, the LSO’s Notice of Appeal in the AA matter includes an additional ground of appeal that (based on the Appeal Division’s reasons, at least) does not appear to have been advanced at the hearing:

The Hearing Division erred in fact and law in not placing sufficient weight on the importance of maintaining public confidence in the legal professions.

The well-established test considers current good character by evaluating (i) the nature and duration of the past misconduct; (ii) whether the applicant is remorseful; (iii) the extent and success of the applicant’s rehabilitative efforts; (iv) the applicant’s conduct since the misconduct; and (v) the passage of time since the misconduct.[9]

While this test may be intended to serve the purpose of promoting public confidence in the legal professions (which exists alongside other regulatory objectives), it does not consider “public confidence in the legal professions as an independent factor entitled to weight. This would represent a fundamental change to the good character analysis.

Moreover, it is entirely unclear how and on what basis “public confidence in the legal professions” could or should be considered. Would it be appropriate if the Law Society made decisions about which cases to pursue and how to pursue them based on concerns about potential headlines and reader comments in the Toronto Star? Is the question simply: Will the public at large be OK with this person becoming a lawyer? If so, how do we address concerns about biases in public opinion; how public opinion may be based on misinformation, misunderstandings, or a rush to judgment; or how public opinion may be inconsistent with fundamental legal principles such as procedural fairness, non-discrimination, and the possibility of rehabilitation?

While it is of course understandable that the Law Society is concerned with public confidence in the legal professions, there are real questions about how—and how much—to consider this objective in exercising its regulatory authority, and in particular how this objective interacts with the LSO’s statutory duties “to protect the public interest”, “to act in a timely, open, and efficient manner”, and “to maintain and advance the cause of justice and the rule of law”.

My comments above are not intended as an endorsement of the conclusions in the aforementioned cases that the applicants in question were “of good character”. My point is that the Tribunal’s findings were reasonable in light of the established legal test for (current) good character. But this merely begs the question.

Does the test for good character serve the objectives of legal regulation? Is there a fundamental disconnect between what the Law Society wishes the good character analysis would assess and what the legal test actually assesses? Is serious reform required? Should a legal regulator be in the business of assessing applicants’ “good character” at all?

In my next column, I plan to take a step back from these recent and controversial cases to consider the good character requirement more broadly—including its purpose, history, process, and problems—as well as potential reforms to consider going forward.

I am always interested in readers’ comments on my columns, but I am particularly interested in your views on this one seeing as I plan to follow up on this topic in a month or two. Thoughts on the good character requirement and cases mentioned above—both from members of the legal professions and of the public—are welcomed in the comments below.


[1] The AA decision raises additional interesting issues about (i) when a licensee or licensing candidate’s identity should be anonymized in good character and discipline proceedings and (ii) whether conditions on a licence restricting services that can be provided to a particular demographic group (e.g. children) can ever be appropriate, both of which are beyond the scope of this column (and likely each warrant columns of their own).

Because I can, however, I will address the latter point briefly. AA’s licensing is subject to a condition that he cannot meet with children unsupervised. My personal view is that if you can’t be trusted to provide services to everybody, you shouldn’t be allowed to provide services to anybody. In my view, such a restriction necessitates differential treatment or limits access to services based on a protected ground. For example, until about a decade ago it was not uncommon for male health professionals found to have committed serial sexual misconduct to be permitted to continue practising subject to restrictions or prohibitions on treating women—meaning that women either did not have access to all licensed providers that men had access to, or that women would receive services in a different manner than men (e.g. with one more person in the treatment room observing and listening to their private concerns, if the professional required supervision). To me, this is both fundamentally unfair and arguably discriminatory. To illustrate another way, I suggest it would be absurd if a professional found to have acted in a racist manner to be permitted to practice so long as they were supervised when meeting with BIPOC clients, or for a professional with a history of making anti-Semitic comments to be permitted to practice so long as they did not serve Jewish clients.

Having said this—and while I maintain my view that the guiding principle “if you can’t serve everybody you can’t serve anybody” is sound and should govern—I acknowledge that in the AA case this particular concern may be more theoretical than practical. I cannot recall a single occasion in my career (in civil and regulatory litigation) where I have interacted with minors in a professional context. Obviously, this does not apply to all areas of practice—family law and immigration being notable exceptions—but I suggest that most lawyers’ practices would not actually be affected by a restriction on meeting with children unsupervised.

[2] Melnick v. Law Society of Upper Canada, 2012 ONLSHP 178 at para 8, rev’d 2013 ONLSAP 27.

[3] In-class discussion of this case has always been lively, and the students’ opinions on the outcome have never been unanimous. What is most interesting to me about the Melnick case is how—much like my students—two panels reviewing the same evidence and applying the same principles came to disparate conclusions about whether Mr. Melnick had demonstrated his “good character” such that he should be licensed to practice law. This has always struck me as indicative that there might be something wrong with having panels of (mostly) lawyers exercising their subjective judgment to assess the “good character” of individual applicants.

[4] Incidentally, here’s a fun fact: both AA and Ms Colangelo were represented before the Law Society Tribunal by none other than Mr. Melnick.

[5] See Jacques Gallant, “Law Society loses appeal and must let child sex abuser practise law—and we can’t reveal his name”, Toronto Star, March 21, 2024,

[6] AA v Law Society of Ontario, 2024 ONLSTA 6 at para 67.

[7] Ibid at para 70.

[8] Ibid at para 69.

[9] Known as the “Armstrong factors” following the decision in LSUC v Armstrong, 2011 ONLSAP 1.


  1. The biggest problem with the “Good Character” requirement is not that we have it, but that it is only applied once (on admission).

    Most lawyer/client interaction is between lawyers who have already been called to the bar. And sometimes racism, misogyny, and other prejudices will present themselves when lawyers are no longer on their best behaviour.

    And then it is from our (admittedly imperfect) profession that appointments to the bench are made.

    There is a theory that “judge school” sanitizes and reforms appointees to enable them to suddenly behave in a manner worthy of the label “Good Character” even if they did not before their appointment; but there is also a competing theory that we all just grow into adult versions of our 3yo selves.

    And given the amount of American media we consume in Canada, no one watching the SCOTUS drama of the last decade can pretend that that court is politically unbiased. With little or no knowledge of Canadian vetting processes, how can the public believe that our courts are different?

    To trust the outcome, we have to trust the process. And for us to trust the process, it has to be visibly unbiased.

  2. This is very interesting and I will look forward to Part 2. You might want to look at the cases and commentary on the “tertiary” grounds for denying bail. My understanding is that even if an accused person poses no risk of fleeing or reoffending before trial, they can still be detained if necessary “to maintain public confidence in the administration of justice.”

    I think this is somewhat analogous to “public confidence in the legal professions” as a basis for finding lack of good character. In both cases it’s really about preventing people from saying “I can’t believe they…”( let him out on bail/let her become a lawyer.).

    The question you raise of whether this sort of thing is legitimate or not touches on some deep questions. I think views on this may map onto the divide between classically liberal versus classically conservative conceptions of the relationship between state and citizen.

  3. Also…

    There is a push from some contingents to infiltrate the self-governing professions (including the practice of law and therefore the courts) with religious elements.

    Perhaps we will eventually have lawyers and judges failing to provide unbiased services to the public based on religious convictions (in the way that some physicians will not prescribe birth control or refer patients to abortion services providers).

    And when we are steeped in someone else’s theocracy, what confidence can the public have that our profession remains of “Good Character”?

  4. Anne McGillivray

    I continue to find it odd that a tribunal, having found that a candidate for the bar has established good character in the present, should then set restrictions on the candidate’s future practice based precisely on the candidate’s bad character in the past. This suggests that the tribunal lacks all confidence in in its own decision.

    As I agree that the principle “if you can’t serve everybody you can’t serve anybody’ is sound and should govern”, I take exception to your exception regarding clients who are children. Rarity is not the issue. The issue is the principle itself. Children, like any other rights-bearers, are fully entitled to approach any lawyer. They are no more in a class of their own in this respect than any other group to whom a lawyer may impermissibly take exception, or of whom a lawyer may impermissibly take advantage. I hope you will rethink this point.

    The good character requirement is a last gasp of the lawyer-as-gentleman standard which premised and concluded early Canadian codes of ethics. Gentlemen were gentlemen because they were presumed to be of good character which is of course both tautological and outdated. Even so, ‘good character’ continues to capture something fundamental about undertaking the practice of law. We need to figure out much more precisely what that something is. I look forward to part II.

  5. This discussion and these issues tend to quickly diffract, and I don’t want to bore anyone, so I’ll offer just a couple of incomplete thoughts for consideration. I don’t think it’s accurate to say that rehabilitation is *the* “fundamental tenet” of the criminal justice system – it’s certainly an important principle (or goal), but I don’t think it is accurate to say that it is the organizing principle of the entire system. (By way of comparison, Section 718 of the Criminal Code, which sets out the objectives of sentencing, lists six different objectives, and rehabilitation comes in fourth.)

    Further to that, it’s not clear why a/the fundamental tenet of the criminal justice system has any significant bearing on the lawyer licensing system – maybe it does/should, but I think that needs to be spelled out in further detail.

    Additionally, I share your skepticism about the usage of formulations such as “public confidence”, particularly, as is invariably the case, in the absence of any attempt to actually ascertain what the public thinks about any particular matter and any willingness to implement what the public’s preference might be (see also, in the criminal justice context, the rhetorical device of bringing the administration of justice into disrepute). Such formulations are ultimately a distraction and dodge for the exercise of power – power should be exercised honestly, and it is fundamentally dishonest to say that the views of the public are in any sense meaningfully taken into account in these kinds of decisions; no one asks the public, and in many cases those exercising the power would not be inclined to accept the public’s “verdict” in any event. None of which is to say that the public’s views *shouldn’t* be taken into account – they should; if lawyers and judges turn away from meaningful engagement with the public’s views and concerns, they risk creating an even more solipsistic system, which the public becomes (justifiably) ever more disenfranchised with and from.

  6. Brooke MacKenzie

    Thank you both for your comments.

    @Bob Tarantino, I agree that the better characterization is “a fundamental tenet”, which is what I’d intended to say–thanks for catching that.

    As for the larger question of how principles from the criminal justice system bear on the lawyer licensing system, this is something I plan to address in my follow-up column on these issues.

  7. Thanks for your take on these issues. I do think part of the conversation needs to address the lack of knowledge around, and acceptance of myths about, sexual abuse, and in this case the sexual abuse of children and adolsecents. While judges receive some training, most other administrative decision makers do not. It is deeply problematic, for example, to describe Mr. Melnick as having had a “sexual relationship” with a 14 year old student, as in your summary above. It would be more accurate to say that he used his position of trust and authority as an adult in a regulated profession to sexually exploit a teenage girl, and very shortly thereafter, claimed that he should be allowed into another regulated profession which would also put him in a position of trust or authority toward clients and staff. I have taught students with serious criminal records, but they spent many years rebuilding their lives, earning trust, and giving back to the community before going to law school. And they were always honest about what they did.

    It is also a myth that those who sexually offend against kids only target kids. That is sometimes the case, but many offenders target both children and adults. Or they groom parents to get at their kids. So the practice restriction offers the illusion of doing something, rather than really getting serious about sexual assault and harassment in the legal profession.

  8. Brooke MacKenzie

    @Janine Benedet: I had taken the term “relationship” from the underlying decisions without much thought, and did not intend to minimize or characterize the underlying conduct. Thank you for pointing out the importance of language used in these matters–I am not sure that in this case it bears on the ultimate analysis, but I agree that words matter and their implications can nevertheless affect readers and their views on the administration of justice.

    @Anne McGillivray: Thank you for your comment. By way of clarification, I was not suggesting or condoning any exception regarding children–only observing that the restriction placed on AA, in all likelihood, has little practical effect. I think we agree that the principle is the issue and should govern regardless. One of the things I am grappling with for part two is what the good character requirement accomplishes if public confidence may nevertheless call for or justify licence restrictions based on the communities that can be served.

  9. Hi Brooke:

    Excellent article. I find the conduct of AA abhorrent and I do not believe anyone who could do something like should ever be allowed to be a lawyer even if they prove they are rehabilitated. Some conduct and some crimes should permanently prevent people from becoming members of trusted professions.

    You mention the principle of rehabilitation a few times. Perhaps you unpack this a bit more in the second part of this, but I wonder if the principle of rehabilitation is something we need to be concerned with outside of criminal law. I agree it should be an important principle in criminal law. If nothing else, it is efficient to turn people who can be rehabilitated into contributing members of society.

    But is there truly an argument that the principle of rehabilitation requires us to let reformed criminals access their highest and best possible professional life? Or is it simply enough that they rejoin and productively contribute to society at some fair threshold level? If rehabilitation to the point of allowing someone to be free from the consequences of their crimes in their career aspirations is important enough to require a professional regulator to consider it, shouldn’t we get rid of criminal records for everyone? Isn’t that one of the biggest barrier to rehabilitation into good careers for basically everyone who has ever committed a crime?

    The fact is that even in the criminal law context, we do not enact policies that let people rehabilitate to a level that guarantees they can get even the least desirable jobs, let alone climb into a highly regulated and competitive profession.

    No one—child sex offenders or not—has a right to be in a regulated profession. Can a child sex offender only become “rehabilitated” if he can join a regulated profession? Not all people with law degrees become lawyers. There are many other great job options open to those with law degrees.

    The majority of people who want to be lawyers and who have never sexually assaulted their own children do not get there—often through no fault of their own. Think about all the barriers to being a lawyer that are already not fair. First, you need to be able to access ~$100,000-300,000 for your education. Second, you need to do better than most other people in school for 6-7 years (or have the time and money to learn how to ace the LSAT). Third, you need to get an articling position. Isn’t it unfair that people who do not have enough money or academic ability can never be lawyers? Most of the barriers to becoming a lawyer are not within the average person’s control at all. Isn’t that more concerning than prohibiting child sex offenders from becoming lawyers? (I know this isn’t a great comparison because AA already passed most of these hurdles.)

    So why do we suddenly care so much about clearing barriers to becoming lawyers for someone who committed disgusting sex crimes that were intentional and within their control? Why would we owe them rehabilitation into what is supposed to be a highly competitive, regulated, and trusted profession?

    I think there is a big difference between the principle of rehabilitation and forgiveness; some of the arguments about this case seem more like arguments in favour of mandatory forgiveness.

    I just don’t see the relevance of the principle of rehabilitation from criminal justice in the context of professional regulation. It should be a privilege to be a lawyer—not a right and certainly not something regulators owe to criminals for the sake of their own rehabilitation.

    If rehabilitation is not a principle we are required to following in professional regulation, what other reasons exist for not simply bending to the public’s obvious desire to not have to worry about whether their lawyer is a child sex offender, murder, domestic violence offender, or any of the other types of criminals that can become lawyers under such a permissive test as the present good character test? Why not bend to the majority of lawyers who find the application of the good character test in these kinds of cases morally offensive and a blight on their own professional achievement (anecdotal, but I’m sure data would support this)?

    I feel ashamed to be a lawyer when I think that someone who sexually assaulted his own kids and was dishonest with the regulator in an official application is now presumptively as trusted as me (especially because he was allowed to benefit from the disgusting nature of his own crimes by hiding his name to protect his children from being retraumatized).

    I think the current interpretation of the legislation does probably compel the result in AA. So I think the real debate is how we amend the legislation and strike strike a fair balance here. But I don’t think rehabilitation should play any role in the test.

  10. Alastair Clarke

    Excellent commentary. I have also appreciated the thoughtful comments above. From my point of view, the case of AA and the ensuing analysis from the LSO reminds me that bad facts make bad law. The admitted facts that AA had engaged in sexual abuse of young children are objectively abhorrent. I can understand Joel’s comment that he feels “ashamed” that someone (suffering from a “paedophilic disorder in remission”, as described by Dr Federoff) is presumptively as trusted as he. I can see how this situation could ensue strong public opinion and juicy headlines.

    Notwithstanding the above, I share Brooke’s concerns regarding the expansion of the “good character” requirement and how it was applied in this particular case. The five (5) factors outlined seem sufficient and it is not clear expanding the requirement was necessary. This requirement reminds me of the factors outlined in Ribic v. Canada (Minister of Employment and Immigration), [1985] I.A.B.D. No. 4 (QL). I regularly refer to these factors when I have a client who is facing removal from Canada. The contexts are completely different; however, it seems the 2 tests are designed to answer a similar question. Neither the current test, as described by Brooke, nor the Ribic factors give any weight to the “public confidence” in their respective systems. The core of both tests is to try to come up with a semi-objective, reasonably consistent standard to decide if the person concerned be allowed to be a member of a group. These are difficult decisions to be sure.

    I look forward to reading part 2.

Leave a Reply

(Your email address will not be published or distributed)