Reflecting on the Supreme Court’s Reassertion of Judicial Control Over Lawyer Withdrawal and Its (Non) Impact on the “Perjury Trilemma”

Cases Considered: R. v. Cunningham, 2010 SCC 10; R. v. White, 2010 ABCA 66

PDF version: Reflecting on the Supreme Court’s Reassertion of Judicial Control Over Lawyer Withdrawal and Its (Non) Impact on the “Perjury Trilemma”

In its March 26, 2010 decision in R. v. Cunningham, 2010 SCC 10, the Supreme Court of Canada rejected the position of the Yukon Territory Court of Appeal that courts have no jurisdiction to prevent counsel from withdrawing from a scheduled criminal proceeding, even if the withdrawal is only for non-payment of fees (Cunningham v. Lilles, 2008 YKCA 7). The Supreme Court affirmed the position taken by most other provincial courts of appeal, that both superior and statutory courts may require that counsel apply for leave when seeking to withdraw from scheduled criminal proceedings, and that in exceptional circumstances the application to withdraw may be denied (See R. v. C (D.D.) (1996), 110 C.C.C. (3d) 323 (ABCA); R. v. Deschamps, 2003 MBCA 116); Bernier v. 9007-1474 Québec Inc., [2001] J.Q. No. 2631 (Que. CA); Mireau v. Canada (1995), 128 Sask. R. 142 (C.A.); R. v. Brundia, 2007 ONCA 725; Contra see Re Leask and Cronin (1985), 18 C.C.C. (3d) 315 (BCSC)).

The issue presented to the Court in R. v. Cunningham was, and is, incapable of entirely satisfactory resolution. On the one hand, if courts assert and exercise control over counsel decisions to withdraw there is a material risk that in the course of exercising that control they will breach solicitor-client privilege either directly or indirectly. There is also a risk of prejudicing the proceedings against the accused. On the other hand, if a court fails to assert and exercise control over counsel decisions to withdraw, it leaves criminal accused vulnerable to abuse by lawyers who may withdraw unethically – that is, in circumstances not permitted by codes of professional conduct – and without the prospect of meaningful control of that abuse by provincial law societies.

Through discussion of Cunningham and the recent judgment of the Alberta Court of Appeal in R. v. White, 2010 ABCA 66, this comment will argue that the Supreme Court provided a reasonable answer to the problem of counsel withdrawal from scheduled criminal proceedings. It will suggest, however, that the judgment did not (and could not) resolve the problems that arise for criminal accused when counsel withdraw mid-proceeding for “ethical reasons,” and in particular because of anticipated or completed client perjury. As long as codes of professional conduct require lawyers to withdraw in such circumstances, solicitor-client privilege and confidentiality will be violated, and the accused may well be prejudiced in the substantive matter before the court.

R. v. Cunningham
Jennie Cunningham was a private counsel retained by Clinton Morgan to defend him on charges of sexual offences against a young child. Mr. Morgan’s fees were to be paid by legal aid; however, legal aid withdrew financial assistance after Mr. Morgan failed to respond to a request from legal aid for updated financial information. After the denial of legal aid, Ms. Cunningham sought to withdraw from representing Mr. Morgan; this was shortly before the commencement of a preliminary inquiry. The preliminary inquiry raised issues to be addressed in relation to the presentation of evidence by the child complainant.

Ms. Cunningham’s application to withdraw was refused by the Territorial Court of the Yukon Territory because the only issue between her and her client was financial, and considerable prejudice would arise given the seriousness of the charges, the importance of timely proceedings given the age of the complainant, the complexity of the issues raised at the preliminary inquiry, and the prejudice to the accused were he to proceed without the benefit of counsel. Ms. Cunningham appealed first to the Supreme Court of the Yukon Territory, where she was unsuccessful, and then to the Court of Appeal of the Yukon Territory, where she was successful. In a judgment which I discussed in an earlier blog post (here), the Court of Appeal held that courts have no jurisdiction to require a lawyer to continue with representation, relying on the disciplinary authority of law societies to control improper withdrawal, the potential risk to solicitor-client privilege from court inquiries into the reasons for withdrawal and the potential for a lawyer-client conflict of interest were representation to continue.

The Supreme Court reversed. In a unanimous judgment written by Justice Marshall Rothstein, the Court noted that while a client has an “unfettered right to discharge his or her legal counsel” (at para. 9), the lawyer’s common law fiduciary and ethical obligations constrain the lawyer’s ability to withdraw from representation once it has commenced. The Court held that the inherent jurisdiction of superior courts to control their own processes and ensure the administration of justice gives those courts the right to both remove counsel and to require them to continue (at para. 18). Statutory courts – such as the first instance Territorial Court in this case – have the same jurisdiction by way of implication from the statutory authority they are given over certain legal proceedings (at para. 19-20).

The Court held that exercising that jurisdiction to require counsel to continue in some cases is necessary to prevent delay that might prejudice the accused and the administration of justice (at para. 22). Doing so does not improperly interfere with solicitor-client privilege. That the client could not pay the lawyer is not privileged in these circumstances. Fee information may be privileged, but only where it is “relevant to the merits of the case, or disclosure of such information may cause prejudice to the client” (at para. 30). Where the reason for withdrawal is either irreconcilable differences between the lawyer and client, or an ethical issue having arisen, the Court recognized the potential for a judge to ask questions of the lawyer that will violate privilege. It suggested that that was what had happened in the earlier British Columbia decision in Re Leask and Cronin (1985), 18 C.C.C. (3d) 315 (BCSC). The Court held, however, that it must be presumed that lawyers “know and respect their professional obligations” and that judges “know the law” (at para. 34). Judges should not “inappropriately attempt to elicit privileged information” (at para. 34); the chance that they may do so is not sufficient to “justify leaving the decision to withdraw exclusively to counsel” (at para. 34).

That provincial law societies may discipline lawyers for professional misconduct is not relevant to whether or not the courts should exercise jurisdiction over withdrawal by counsel. The Court noted both that the jurisdiction of law societies and courts differ, and that the authority over lawyers exercised by each is “not mutually exclusive” (at para. 38). In brief, courts respond prospectively to conduct by lawyers that might cause a problem to the administration of justice; law societies respond reactively to conduct by lawyers that has caused a problem to the administration of justice. The power exercised by the courts in this respect, and across numerous areas, does not pose any threat to the independence of the bar (at para. 39).

The Court rejected the suggestion that lawyers would be in an improper conflict if required to continue with representation of a client. Counsel must act diligently, even when not being paid, and it is reasonable to presume that they will do so – “the court should presume that lawyers act ethically” (at para. 40).

The Court held that this power in the courts was distinct from the power to order a stay of proceedings until the government provides legal aid (a Rowbotham order; see R. v. Rowbotham (1988), 41 C.C.C. (3d) 1 (Ont. C.A.)). It also held that an order requiring counsel to proceed was exceptional, and “should truly be a remedy of last resort” (at para. 45).

In sum, the Court held that where counsel seeks to withdraw sufficiently in advance of proceedings such that an adjournment would not be necessary, the court should permit withdrawal without requiring counsel to give reasons. Where timing is at issue, and counsel states that the withdrawal is for ethical reasons, then “the court must accept counsel’s answer at face value and not enquire further so as to avoid trenching on potential issues of solicitor-client privilege” (at para. 48). “Ethical reasons” include a request by the client that the lawyer violate ethical obligations or refusal of the accused to accept advice of counsel “on an important trial issue” (at para. 48). The Court must grant a request to withdraw for ethical reasons (at para. 49). Where withdrawal is for non-payment of fees, permission to withdraw lies in the court’s discretion; the discretion should take into account the feasibility of self-representation, the availability of alternative representation, the impact of delay, the conduct of counsel and the history of proceedings. Since none of these factors involve matters specific to the lawyer-client relationship, considering them will not breach privilege. The Court held that review of a decision refusing withdrawal could proceed either through an application pursuant to s. 40 of the Supreme Court Act, R.S.C. 1985, c. S 26, or by way of certiorari on the grounds that the refusal constituted an error of law on the face of the record.

Analysis
The Court’s judgment in Cunningham has two great strengths: first, it addresses an important instance of the vulnerability of criminal accused, a vulnerability which exists even in relation to their own lawyers. While the vast majority of lawyers will not withdraw from representation without proper reason, the requirement that the court approve withdrawal will provide an additional incentive to ensure that that is the case. Particularly notable in this respect was the emphasis of the Court on prejudice to the accused and to the administration of justice, and not on questions such as inconvenience to court scheduling, which have motivated courts in past cases (see for example R. v. C (D.D.) (1996), 110 C.C.C. (3d) 323 (ABCA)). The first three factors to be considered by a court in considering an application to withdraw relate to the circumstances of the accused; the Court expressly rejected the consideration of “whether allotted court time can be otherwise usefully filled” as irrelevant (at para. 51). Obviously, as the Court noted, it is only in the most extreme circumstances that counsel should be required to represent an accused without compensation. And, although the Court does not make this point expressly, where appropriate a Rowbotham application should be considered by a court on an expedited basis rather than leaving counsel – alone amongst everyone else in society – to bear the burden of ensuring another person’s access to justice. At the same time, endorsing judicial review of counsel requests for withdrawal on the eve of trial protects the person most vulnerable in a criminal proceeding – the accused.

Second, the Court did its best to address the risk that can arise from asserting judicial discretion over counsel withdrawal: the likelihood that courts will violate solicitor-client privilege by inquiring into the true reasons for withdrawal and – worse – in doing so impair the fairness of the proceedings against the accused whose prejudicial disclosures to counsel will be revealed. The Court was absolutely clear that when counsel states that the reasons are ethical in nature, the court must not ask further questions and must permit withdrawal. This does not prevent a violation of the solicitor-client privilege: the statement that “ethical reasons” have arisen communicate something private to the solicitor-client relationship to the court, even if nothing about those reasons is stated. The Court did, however, do its best to make that disclosure as minimal as possible while still exercising the supervisory jurisdiction necessary to protect the accused and the administration of justice.

What the Court could not do, though, is ameliorate the fundamental problem that arises when lawyers choose (or are required by their professional obligations) to withdraw because of “ethical reasons”. Take the most notable of the ethical problems likely to cause withdrawal mid-proceeding – where a criminal accused tells a lawyer that she is going to lie on the stand, the lawyer attempts to dissuade her, but she is adamant that she is going to do so (or, alternatively, where after testifying the client advises the lawyer that she has lied on the stand). The client has a constitutional right to testify but the lawyer has an ethical obligation not to mislead the court. Under most Canadian codes of conduct (and in the view of most Canadian commentators), the lawyer must withdraw from the representation if he cannot persuade his client to change her mind about lying, or disclose to the court that she has lied.

Under the judgment in Cunningham, the lawyer will now advise the court that “ethical reasons” require him to withdraw. With what result? The lawyer will have fulfilled his legal and moral obligation of candour to the court. He will have respected his client’s constitutional right to testify. But the lawyer will have violated two other legal and moral obligations – privilege and the client’s right to counsel. While the Court in Cunningham reduces the harm to privilege by preventing the court from inquiring into the nature of the ethical reasons that have arisen, as soon as a lawyer advises the court that ethical reasons require withdrawal, the lawyer has also placed a bright red “caution” flag over his client’s head. The court knows that the client has in some way behaved improperly – or, at best, is perceived by her lawyer to have behaved improperly. That is, it knows something that should legally be private between the lawyer and client, and which will potentially prejudice the proceedings against the accused. In addition, the accused will have no lawyer, and will be perceived as being at fault for that fact, which may lead the court simply to require the accused to proceed unrepresented.

R. v. White
The facts of R. v. White, 2010 ABCA 66 demonstrate what this scenario might look like. White was convicted of second degree murder in August 2008. His trial began in April 2007, and the Crown closed its case in May of that year. The defence opened its case by calling the accused, who testified that his girlfriend had committed the murder, that he had helped her clean up and then had lied about what happened in order to protect her. After the accused testified, his lawyer sought to withdraw for “irreconcilable differences” – that is, because of an ethical problem. White retained new counsel and the trial was adjourned. The new counsel was bringing a mistrial application, which was set down to be heard in December. Prior to the hearing of that and a related motion on its timing, the new counsel sought leave to withdraw (for reasons which are not stated in the judgment); a third counsel had been obtained for the accused, but that third counsel could not appear until June 2008. The trial judge rejected the application to adjourn, and ordered White to proceed in February 2008, with new counsel if he could find one, but unrepresented if he could not. White was unable to find counsel in the time provided, and proceeded unrepresented to present the main witness in support of his case – his girlfriend, who testified that she had committed the murder – as well as another witness to impugn the character of one of the Crown witnesses.

After the evidence had been called the trial was adjourned until April 2008, at which point White had new counsel. Argument took place in June 2008, and in August White was convicted, largely because the trial judge did not view the girlfriend or White as credible.

The Alberta Court of Appeal overturned White’s conviction and ordered a new trial on the basis that White did not receive a fair trial. In particular, the trial was unfair because the failure to grant an adjournment required White to present the central part of his case without counsel to address the numerous legal issues raised by the girlfriend’s testimony, and in particular her ability to claim immunity under the Canada Evidence Act, R.S.C. 1985, c. C-5. The Court noted that White was not to blame for the absence of counsel and that the girlfriend’s evidence was “truncated and fractured” (at para. 21); “competent counsel [may] have chosen not to call Ms. Eliuk” (at para. 24, emphasis in original).

The Court of Appeal did not address, however, the relationship, if any, between the trial judge’s decision to refuse White’s adjournment request, and the decision of his original counsel to withdraw as soon as his client had testified on the basis of “irreconcilable differences.” It seems reasonable to speculate that a judge in that situation might think – whether or not this was true – that White had lied, that White’s lawyer knew he had lied and could not say so given his duties under the Alberta Code of Professional Conduct, and that White’s lawyer knew that the pending testimony of the accused’s girlfriend was also false. It is true that White’s lawyer never said, “I am withdrawing because my client is a liar”. But when a lawyer withdraws for ethical reasons immediately following his client’s testimony what inference can a rational judge help but draw other than that one? This may not have affected the trial judge’s decision not to grant an adjournment. On the other hand, given the seriousness of the charges, White’s youth and inexperience, the critical nature of the girlfriend’s testimony, the complexity of the issues that testimony presented and the lack of culpability of White in the withdrawal of other counsel, the refusal of the adjournment seems otherwise hard to understand.

This is the sort of problem that Cunningham did not – and could not – resolve. Had it followed the judgment of the Yukon Court of Appeal and allowed counsel to leave without giving any reasons, the Court would have avoided inviting judges to draw the inference that was available to the judge in White: since the judge would not know why White’s first lawyer withdrew, the judge could more easily avoid the inference that it was because White had lied. The Court would, though, still have left White and others like him without a lawyer when ethical difficulties arise. Moreover, it would have done no more than fail to ameliorate a problem that arises not from the rules on withdrawal, but from the rules that require counsel to withdraw when faced with client perjury in criminal cases.

As set out by Monroe Freedman (see: “Professional Responsibility of the Criminal Defense Lawyer: The Three Hardest Questions” (1964) 64 Mich. L. Rev. 1469; Monroe H. Freedman and Abbe Smith, Understanding Lawyers’ Ethics 3d ed. (LexisNexis, 2004) pp. 159-195), a lawyer faced with client perjury faces a moral trilemma: sacrifice solicitor-client trust and confidentiality, sacrifice the client’s right to counsel, or sacrifice the lawyer’s duty of candour to the court. Faced with knowledge of the client’s dishonesty, the lawyer cannot fulfill all the moral obligations simultaneously, and must choose between them. The decision in Cunningham does not alter that fact, and it does not alter the fact that should the lawyer choose to withdraw, the client’s right to counsel and the right to confidentiality will be compromised. The better resolution to that problem – and what is necessary to avoid the problem faced by White – is for counsel to sacrifice instead candour to the Court. Assuming that the ethical problem in White was that the accused was lying, the better outcome would be for his lawyer to carry on representing him as normal. Not necessarily presenting the testimony of the girlfriend (which the lawyer would, in any event, be under no obligation to present), but representing the client as if he had told the truth. Doing so may not have resulted in White’s acquittal, and would have resulted in the trial court being presented with testimony that may have been dishonest, but it would also have left White represented, would have prevented a 14 month delay in the conclusion of the trial and would have left the Court in the position it ended up in in any event – assessing the credibility of White as to his role in the murder.

This may not be the appropriate resolution of every ethical dilemma faced by the lawyer; it is not every case of “irreconcilable differences” that can or should be resolved by the lawyer staying the course. But in my view where the problem is client perjury, and where any other solution will result in a violation of confidentiality, active prejudice to the client and the denial of counsel, the solution of continued representation, however messy and unsatisfactory, is the best solution on offer.

Like Freedman, I do not embrace this solution without serious reservations. Such reservations are unavoidable given the result that the lawyer violates his ethical obligations. But once the lawyer knows of the client’s anticipated or completed perjury, violation of some ethical obligation is necessary – the only question is which one.