The judgment of Moldaver, Gascon, Brown and Rowe JJ. was delivered by
Moldaver, Gascon & Brown JJ.
 This case concerns the proper approach for considering whether a guilty plea can be withdrawn on the basis that the accused was unaware of a collateral consequence stemming from that plea, such that holding him or her to the plea amounts to a miscarriage of justice under s. 686(1)(a)(iii) of the Criminal Code, R.S.C. 1985, c. C-46.
 The decision of an accused to plead guilty is plainly significant. By pleading guilty, an accused waives his or her constitutional right to a trial, relieving the Crown of its burden to prove guilt beyond a reasonable doubt. Taking this step is of such significance that it represents one of the very few decisions in the criminal process which an accused must personally take. Indeed, defence counsel are ethically bound to ensure that the ultimate choice is that of the accused.
 The plea resolution process is also central to the criminal justice system as a whole. The vast majority of criminal prosecutions are resolved through guilty pleas and society has a strong interest in their finality. Maintaining their finality is therefore important to ensuring the stability, integrity, and efficiency of the administration of justice. Conversely, the finality of a guilty plea also requires that such a plea be voluntary, unequivocal and informed. And to be informed, the accused 'must be aware of the nature of the allegations made against him, the effect of his plea, and the consequence of his plea' (R. v. T. (R.) (1992), 10 O.R. (3d) 514 (C.A.), at p. 519).
 We agree with our colleague Wagner J. that for a plea to be informed, an accused must be aware of the criminal consequences of the plea as well as the legally relevant collateral consequences. A legally relevant collateral consequence is one which bears on sufficiently serious legal interests of the accused. Here, Mr. Wong was not aware of the immigration consequences of his conviction and sentence. Immigration consequences bear on sufficiently serious legal interests to constitute legally relevant consequences. His guilty plea was therefore uninformed.
 We respectfully disagree with our colleague, however, as to the prejudice that must be shown to establish a miscarriage of justice and vacate a guilty plea. Our colleague proposes that whether an accused has shown prejudice should be determined by way of a 'modified objective' analysis. On that approach, prejudice giving rise to a miscarriage of justice is established where the court is satisfied of a 'reasonable possibility that a similarly situated reasonable person would have proceeded differently if properly informed' (Wagner J.’s reasons, at para. 80). As we discuss below, this approach does not account for the fundamentally subjective and deeply personal nature of the decision to plead guilty. Further, it will likely be difficult for courts to apply.
 In our view, the accused should be required to establish subjective prejudice. Meaning, accused persons who seek to withdraw their guilty plea on the basis that they were unaware of legally relevant consequences at the time of the plea must file an affidavit establishing a reasonable possibility that they would have either (1) opted for a trial and pleaded not guilty; or (2) pleaded guilty, but with different conditions. To assess the veracity of that claim, courts can look to objective, contemporaneous evidence. The inquiry is therefore subjective to the accused, but allows for an objective assessment of the credibility of the accused’s subjective claim.
A. Modified Objective Framework
 Under the modified objective approach as stated by our colleague, a guilty plea may be withdrawn if the accused shows
(1) that he or she was not aware of a legally relevant collateral consequence and (2) that there is a reasonable possibility he or she would have proceeded differently if properly informed of that consequence.
(Wagner J.’s reasons, at para. 44).
 While this statement may appear to call for a subjective assessment (since it refers to what the accused would have done), our colleague clearly affirms that his approach entails applying a modified objective standard - one which requires the reviewing court to consider what 'a reasonable person in the circumstances of the accused' would have done (para. 104). The requisite standard of proof he states is 'reasonable possibility': specifically, is there a reasonable possibility that awareness of the legally relevant consequence would have 'sufficiently influenced' the decision of someone in the accused’s circumstances, or that a reasonable person in the same situation as the accused would have 'proceeded differently' (para. 81 (emphasis deleted)).
 We agree that the accused must first show that he or she was unaware of a legally relevant collateral consequence at the time of pleading guilty, and endorse a broad approach to evaluating the relevance of a collateral consequence in the assessment of whether a guilty plea was sufficiently informed. We also agree that a legally relevant collateral consequence will typically be state-imposed, flow from conviction or sentence, and impact serious interests of the accused. And, like our colleague, we do not see it as necessary to define the full scope of legally relevant collateral consequences nor the characteristics of such consequences for the purposes of this appeal. We see two problems, however, with the second step as our colleague states it.
 First, a modified objective framework fails to account for the fundamentally subjective nature of the guilty plea. As the Attorney General of Alberta observed before us:
. . . the decision whether or not to plead guilty is inherently personal and an accused at first instance can decide to simply roll the dice whether or not they are advised by their lawyer they have a realistic prospect of conviction and whether or not it’s going to have a deleterious effect upon sentence.
. . .
. . . sometimes people can decide to run trials in a very ill-advised manner.
(Transcript, at pp. 122-23)
 We agree. The decision to plead guilty reflects deeply personal considerations, including subjective levels of risk tolerance, priorities, family and employment circumstances, and individual idiosyncrasies. For this reason, it is one of the few steps in the criminal process where defence counsel are ethically required to seek their client’s direct instruction (R. v. G.D.B., 2000 SCC 22,  1 S.C.R. 520, at para. 34).
 Simply put, pleading guilty is the decision of the accused, not a reasonable accused, or someone like the accused. To permit reviewing courts to substitute their own view of what someone in the accused’s circumstances would have done is to run a serious risk of doing injustice to that accused. An example from United States caselaw suffices to make the point. In Lee v. United States, 825 F.3d 311 (6th Cir. 2016), the accused sought, as Mr. Wong seeks, to withdraw his plea on the basis that he was unaware of its consequences for his immigration status. The Sixth Circuit Court of Appeals denied the accused’s motion. Even taking into account the accused’s particular circumstances, the Sixth Circuit wrote:
. . . no rational defendant charged with a deportable offense and facing 'overwhelming evidence' of guilt would proceed to trial rather than take a plea deal with a shorter prison sentence. [para. 2]
 The accused in Lee had deposed that he would have proceeded to trial, with the effect of near certain deportation, rather than taking a plea deal with certain deportation, even if conviction at trial meant a longer prison sentence. Despite what the Sixth Circuit saw as the only rational course of action, the accused’s right to remain in the United States was more important to him than any jail sentence, no matter its length. The Sixth Circuit’s decision was ultimately overturned by the Supreme Court of the United States in Lee v. United States, 137 S. Ct. 1958 (2017), in which the objective approach for assessing prejudice was rejected.
 Our colleague quite rightly notes that this Court has applied a modified objective test in other contexts, such as when assessing the availability of the defences of necessity (R. v. Latimer, 2001 SCC 1,  1 S.C.R. 3, at para. 32) and duress (R. v. Ruzic, 2001 SCC 24,  1 S.C.R. 687, at para. 61) (Wagner J.’s reasons, at para. 88). But it does not follow that a modified objective test is equally suited to assessing the considerations surrounding the decision to plead guilty. It is true that much of the criminal law is itself premised upon objective considerations, in that it 'reflect[s] society’s values as to what is appropriate and what represents a transgression' (Latimer, at para. 34; see also Perka v. The Queen,  2 S.C.R. 232, at p. 248). The availability of a valid defence such as necessity or duress cannot therefore be approached purely subjectively, lest it 'become simply a mask for anarchy' (Latimer, at para. 27, citing Southwark London Borough Council v. Williams,  Ch. 734 (C.A.), at p. 746). The same cannot, however, be said of the accused’s decision to plead guilty. That decision does not purport to reflect society’s values as to what is right or wrong. Rather, it reflects the accused’s subjective choice. Unlike the considerations applied to the availability of a valid defence, no mischief follows from letting considerations personal to the accused determine whether he or she would have, in the circumstances, made an informed plea of guilty. Ultimately, what matters is the accused’s decision to plead guilty or to proceed to trial, and not whether that decision is, to someone else, reckless or irrational.
 We acknowledge that our colleague does not advocate for a purely objective framework, but for a modified objective framework. It allows for some consideration of the general 'situation and characteristics of the accused' insomuch as it undertakes the analysis from the perspective of someone 'in the particular circumstances of the accused' (see Wagner J.’s reasons, at paras. 80 and 87). But it nonetheless suffers from the same drawback as a purely objective inquiry: it focusses upon what a judicially constructed hypothetical person would do, instead of how the particular accused would have proceeded.
 The second problem we see in the modified objective framework is that it will likely be difficult for lower courts to apply. Our colleague refers to what 'a similarly situated reasonable person' would have done (para. 80). But this is qualified by his statement that such a reasonable person need not be presumed to 'have taken the ‘best’ or single most rational course of action' (para. 82). Given the highly contextual and even idiosyncratic nature of factors that influence important decisions (such as choosing whether or not to plead guilty), adopting a standard based on what a hypothetical reasonable person (who might not always act in the most rational way) would have done effectively confers upon reviewing courts unbounded discretion to reach whatever conclusion they see fit. It also runs squarely into the injustice that led to the United States Supreme Court’s intervention in Lee.
 The modified objective framework will also be difficult to apply because it adopts a variable standard of scrutiny. Our colleague maintains that where the collateral consequence of a plea is 'as serious as deportation', a more lenient standard of reasonableness would be applied, whereas 'less obviously serious consequence[s]' are scrutinized by 'a more exacting inquiry' (para. 100). To be sure, different accused - even different similarly situated accused - may ascribe varying levels of significance to different collateral consequences, based on their idiosyncratic values and preferences. Thus, and with respect, we would not treat the significance of a particular consequence as a 'matter of common sense' (ibid.). And because Justice Wagner’s approach is not tied to a particular accused, but rather to a reasonable person, reviewing courts may be left guessing as to what standard of scrutiny to apply to the consequence at issue.
 In sum, our colleague’s modified objective approach risks, in our view, resulting in vacated guilty pleas even where there is no evidence that the accused personally would have done something differently. Even further, an accused who admits under cross-examination that he would have proceeded identically would still be entitled to withdraw his plea if a reasonable accused in his circumstances would withdraw his plea. This would impose unnecessary and substantial demands on a criminal justice system that is already overburdened, to the detriment of other participants in the system, including accused persons, victims, and the public at large who seek efficient and just resolution of criminal complaints.
B. Subjective Prejudice Framework
(1) Forms of Prejudice
 In our view, an accused seeking to withdraw a guilty plea must demonstrate prejudice by filing an affidavit establishing a reasonable possibility that he or she would have either (1) pleaded differently, or (2) pleaded guilty, but with different conditions. This approach strikes what we see as the proper balance between the finality of guilty pleas and fairness to the accused.
 With respect to the first form of prejudice - where the accused would have opted for a trial and pleaded not guilty - there will of course be instances in which the accused may have little to no chance of success at trial, and the choice to proceed to trial may simply be throwing a 'Hail Mary'. But a remote chance of success at trial does not necessarily mean that the accused is not sincere in his or her claim that the plea would have been different. For certain accused, such as the accused in Lee, the certain but previously unknown consequences of a conviction made even a remote chance of success at trial a chance worth taking. In such circumstances, and where the court accepts the veracity of his or her statement, the accused has demonstrated prejudice and should be entitled to withdraw his or her plea.
 There remains the second form of prejudice - where an accused would have pleaded guilty, but only on different conditions. A guilty plea on different conditions will suffice to establish prejudice where a court finds that the accused would have insisted on those conditions to enter a guilty plea and where those conditions would have alleviated, in whole or in part, the adverse effects of the legally relevant consequence. We do not presume here to list every condition which, if raised by the accused, could give rise to prejudice. At minimum, however, these additional conditions may include accepting a reduced charge to a lesser included offence, a withdrawal of other charges, a promise from the Crown not to proceed on other charges, or a joint submission on sentencing.
 The mere possibility of different conditions on its own is not, we stress, automatically sufficient. A plea may be withdrawn only where an accused credibly asserts that he or she would have, during the plea negotiation phase, insisted on additional conditions, but for which he or she would not have pleaded guilty. In short, the accused must articulate a meaningfully different course of action to justify vacating a plea, and satisfy a court that there is a reasonable possibility he or she would have taken that course.
 Parenthetically, we observe that the accused need not show a viable defence to the charge in order to withdraw a plea on procedural grounds. '[T]he prejudice lies in the fact that in pleading guilty, the appellant gave up his right to a trial' (R. v. Rulli, 2011 ONCA 18, at para. 2 (CanLII)). Requiring the accused to articulate a route to acquittal is antithetical to the presumption of innocence and to the subjective nature of choosing to plead guilty. An accused is perfectly entitled to remain silent, advance no defence, and put the Crown to its burden to prove guilt beyond a reasonable doubt. It does not make sense to let an accused proceed to trial at first instance without any defence whatsoever, but to insist on such a defence to proceed to trial when withdrawing an uninformed plea. Though the decision to go to trial may be unwise or even reckless, we are not seeking to protect an accused from himself or herself. Rather, we seek to protect an accused’s right to make an informed plea.
 For the same reason, we agree with our colleague that the ineffective assistance of counsel framework has no relevance to this case (Wagner J.’s reasons, at para. 60). That framework focuses on the source of the misinformation (or incomplete information) rather than the misinformation itself. Assessing whether prejudice arises from misinformation does not depend upon its source. As Saunders J.A. explained at the Court of Appeal, the particular miscarriage of justice engaged in this case arises from the invalidity of Mr. Wong’s plea (2016 BCCA 416, 342 C.C.C. (3d) 435, at para. 24).
(2) Subjective Analysis
 Our framework is premised upon the view that judicial scrutiny must be directed to how the accused, and no one else, would have proceeded. The question to be answered is whether the accused would have acted differently, had he or she been armed with the knowledge of the legally relevant consequence.
 That the analysis focusses on the accused’s subjective choice does not mean that a court must automatically accept an accused’s claim. Like all credibility determinations, the accused’s claim about what his or her subjective and fully informed choice would have been is measured against objective circumstances. Courts should therefore carefully scrutinize the accused’s assertion, looking to objective, circumstantial evidence to test its veracity against a standard of reasonable possibility. Such factors may include the strength of the Crown’s case, any concessions or statements from the Crown regarding its case (including a willingness to pursue a joint submission or reduce the charge to a lesser included offence) and any relevant defence the accused may have. The court may also assess the strength of connection between the guilty plea and the collateral consequence, that is, whether the trigger for the collateral consequence is the finding of guilt as distinct from a particular length of sentence. More particularly, where the collateral consequence depends on the length of the sentence - keeping in mind that a guilty plea typically mitigates a sentence - the court may have reason to doubt the veracity of the accused’s claim.
 While our colleague refers to similar factors (at para. 105), he would consider them in assessing whether a reasonable person in the accused’s circumstances would have been influenced in their decision to plead guilty by the information. Again, we see the analysis differently. To reiterate, it properly operates from the standpoint of the accused, and what the accused would or would not have done, knowing of the legally relevant consequence.
 Of course, the basis for judicial scrutiny of the accused’s claim is not limited to objective circumstances contemporaneous with the original plea, since the accused’s idiosyncratic preferences may not always be reflected in those circumstances. A reviewing court must therefore also test the veracity of the accused’s assertions in their own right. A court may properly find an accused’s expressed preferences to be credible, and to establish a reasonable possibility of prejudice, based solely on the contents of the accused’s affidavit and on his or her withstanding of cross-examination.
 Throughout the process of testing the accused’s claim, however, the focus must remain upon what this accused - and only this accused - would have done. The basis for that subjective inquiry is found in the subjective nature of the initial decision to plea. Because the original guilty plea is an exercise of the accused’s own subjective judgment, it logically follows that the test for withdrawing that plea should also be directed to the accused’s subjective judgment. This approach properly balances society’s interest in the finality of guilty pleas and fairness to the accused by striking the accused’s plea only where he or she would have proceeded differently.
 We note parenthetically that adopting a subjective framework, which requires the accused to swear an affidavit in support, will not create a 'procedural bar' to striking a plea (Wagner J.’s reasons, at para. 93). First, our colleague’s modified objective approach itself will require an accused to depose to his or her 'particular circumstances' (para. 87) and to not having been informed of a legally relevant consequence. Second, any concern about an accused person who seeks to have their plea struck but who is unrepresented and unaware of the necessity of deposing that they would have proceeded differently if properly informed can be accounted for by the trial judge who should take steps to ensure that the accused obtains representation or, at the very least, is assisted by duty counsel (where available). And third, the accused need not speculate on how other participants in the justice system would have proceeded (ibid.). Our approach simply requires an accused to state how he or she would have acted differently. Though a condition sought may turn on another party’s response - e.g. the Crown’s willingness to agree to a joint submission on sentencing - the accused need only state that he or she would have insisted upon such a condition to plead guilty, or else would have proceeded to trial.
 Our subjective framework is consistent with the approach taken by the Ontario Court of Appeal in R. v. Henry, 2011 ONCA 289, 277 C.C.C. (3d) 293, and R. v. Quick, 2016 ONCA 95, 129 O.R. (3d) 334. In Henry, Watt J.A. found prejudice where 'there was a realistic likelihood that he [the accused] would have run the risk of a trial' (para. 37 (emphasis added)). In Quick, Laskin J.A. also directed his focus to how the accused would have conducted him/herself with the knowledge of the legally relevant consequence (para. 35). And, as we have also recounted, the subjective approach to assessing prejudice was just last year adopted by the Supreme Court of the United States in Lee.
 In response, our colleague cites this Court’s guidance in R. v. Taillefer, 2003 SCC 70,  3 S.C.R. 307, as a core rationale for a modified objective approach. In particular, he says that the Court’s approach in Taillefer 'resemble[s]' his framework (para. 89). In our respectful view, Taillefer does not hold and should not be understood as holding that prejudice arising from an uninformed plea - as distinct from being uninformed - is assessed on an objective standard.
 Recall that the framework for striking an uninformed guilty plea involves two discrete steps: (1) the accused being misinformed about sufficiently serious information; and (2) that lack of information resulting in prejudice (Wagner J.’s reasons, at para. 44). While these steps are, at times, collapsed in LeBel J.’s reasons in Taillefer, in our view, its best reading maintains their separation.
 Whether an accused is uninformed - that is, whether the information unknown to the accused falls within the scope of what an accused must know to give an informed plea - is assessed objectively. Here, this step objectively assesses the seriousness of the unknown legal consequence. In Taillefer, this entailed assessing the 'undisclosed evidence . . . together with all of the evidence already known' (para. 90). Whether undisclosed evidence is sufficiently serious to render an accused misinformed is undeniably an objective question. And it is this objective inquiry to which LeBel J. referred when outlining the objective component of the framework in Taillefer for striking a plea. In his words, that framework considered 'the volume, weight and relevance of the undisclosed evidence and the new possibilities that the opportunity to use that evidence would have offered' (para. 111). In Taillefer, applying that objective assessment led LeBel J. to conclude that the non-disclosure 'led to a serious infringement of the appellant’s right to make full answer and defence' (para. 112). But, to be clear - that infringement flowed from the objective content of the undisclosed evidence, and not from the subjective view of the appellant in that case about the significance of that evidence to his or her plea.
 In contrast, prejudice - that is, whether the accused’s being uninformed impacted the plea - is assessed subjectively by considering whether the accused would have taken a meaningfully different course of action in pleading. This is entirely consistent with Taillefer, where prejudice was similarly assessed by considering whether the accused would have made the same plea. In particular, a subjective analysis conforms to the direction in Taillefer that 'the breach must bear on the accused’s decision to enter the guilty plea', that courts must assess 'the impact of the unknown evidence on the accused’s decision to admit guilt', and that the test is whether 'there was a realistic possibility that the accused would have run the risk of a trial, if he or she had been' informed (para. 90 (emphasis added)). We also note that Laskin J.A., when following the 'general approach in Taillefer', applied a subjective rather than objective test (Quick, at para. 35). Similarly, the authorities that LeBel J. endorses in Taillefer when describing the proper approach to assessing prejudice also adopt a subjective approach (paras. 88-90).
C. Application of the Framework
 We agree with our colleague that Mr. Wong’s plea was uninformed (see Wagner J.’s reasons, at para. 102). To establish prejudice, however, the accused seeking to withdraw a guilty plea must show a reasonable possibility that, having been informed of the legally relevant consequence, he or she would have either pleaded differently, or pleaded guilty with different conditions. Mr. Wong has not met this burden.
 Though he filed an affidavit before the Court of Appeal, he did not depose to what he would have done differently in the plea process had he been informed of the immigration consequences of his guilty plea (Affidavit of Mr. Wong, A.R., at pp. 67-69; C.A. reasons, at para. 14; Wagner J.’s reasons, at paras. 54 and 57- 58). We therefore see no basis to permit him to withdraw his plea.
 We recognize that, at the time Mr. Wong sought to withdraw his plea, the state of the law as to what he was required to include in his affidavit was not entirely clear. And, like our colleague (at para. 105), we also recognize that someone in Mr. Wong’s circumstances may have elected to proceed to trial, even with a plea deal for a sentence of less than six months, in order to avoid inadmissibility to Canada. We observe, however, that the principal thrust of his submissions before us suggested that his overriding (although not exclusive) concern was to avoid deportation. With that in mind, we note that Mr. Wong’s sentencing appeal is outstanding, and the Crown has conceded that a sentence of six months less a day would be appropriate in light of Mr. Wong’s deportation risk (see R.F., at para. 69). From this, it follows that his right to appeal the removal order will likely be preserved after the conclusion of his sentencing appeal.
 All of that said, because Mr. Wong did not state in his affidavit that he would have proceeded differently, we are of the view that he has not established prejudice giving rise to a miscarriage of justice.
 We would dismiss the appeal.
The reasons of McLachlin C.J. and Abella and Wagner JJ. were delivered by
 An essential criterion of a valid guilty plea is that the accused be informed of the consequences of entering the plea. The criminal consequences of such a plea include conviction and the imposition of a sentence. However, a guilty plea can also trigger serious consequences collateral to the criminal process that may significantly affect the fundamental interests of the accused. This appeal requires us to consider whether an accused person must be aware of such collateral consequences for a guilty plea to be sufficiently informed.
 The appellant, Wing Wha Wong, is a permanent resident of Canada. He immigrated to this country from China over 25 years ago. He lives in Canada with his wife and their Canadian-born child and is his family’s sole financial provider. In the spring of 2012, Mr. Wong was charged with a single count of trafficking in cocaine to which he ultimately pleaded guilty. When he entered his plea, Mr. Wong was not aware that his being convicted and sentenced could result in the loss of his permanent resident status and a removal order from Canada without any right of appeal. Mr. Wong now seeks to withdraw his plea on the basis that it was uninformed and therefore gave rise to a miscarriage of justice.
 A guilty plea may be accepted by a court only if it is voluntary, unequivocal and informed. For it to be informed, the accused must understand the nature of the allegations, the effect of the plea and the consequences of the plea. In addition to effects on the criminal process itself, significant non-criminal consequences can flow from a guilty plea. These are known as collateral consequences. The central question in this appeal is when a guilty plea can be set aside because the accused was not aware that it might have serious collateral consequences. The answer must strike a balance between core values of the criminal justice system by ensuring a procedurally fair trial and safeguarding the rights of the accused, while also preserving the finality and order that are essential to the integrity of the criminal process.
 In my view, a guilty plea may be withdrawn if the accused shows (1) that he or she was not aware of a legally relevant collateral consequence and (2) that there is a reasonable possibility he or she would have proceeded differently if properly informed of that consequence. A legally relevant consequence is one which bears on sufficiently serious interests of the accused. Where an accused enters a plea unaware of its legally relevant consequences, this raises a concern related to procedural fairness. However, not every guilty plea entered in such circumstances will result in prejudice that is serious enough to constitute a miscarriage of justice. Therefore, the court must also be satisfied of a reasonable possibility that the accused would have proceeded differently had he or she been aware of the legally relevant consequence in issue, either by declining to admit guilt and entering a plea of not guilty, or by pleading guilty but with different conditions. This question is to be assessed on a modified objective standard. If the court is so satisfied, the prejudice to the accused constitutes a miscarriage of justice and the guilty plea may be withdrawn.
 For the reasons that follow, I would allow the appeal, set aside Mr. Wong’s conviction and remit the matter to the court of original jurisdiction for a new trial. I am satisfied that the loss of permanent resident status and the risk of removal from Canada without any right of appeal constitute legally relevant consequences. Mr. Wong was unaware that his guilty plea might carry these immigration consequences which flowed directly from his conviction and sentence. I am persuaded that there is a reasonable possibility that a reasonable person in Mr. Wong’s circumstances would have proceeded differently had he or she been aware of such consequences. His guilty plea therefore gave rise to a miscarriage of justice and must be set aside.
 Mr. Wong is a Chinese citizen and a permanent resident of Canada. He immigrated to Canada in 1990, apparently having left China because of the crackdown on the pro-democracy movement in that country in the late 1980s. Mr. Wong and his wife have a young daughter, and they currently live in Kamloops, British Columbia. As I mentioned above, he is his family’s sole financial provider.
 On April 3, 2012, Mr. Wong was charged with one count of trafficking in cocaine under s. 5(1) of the Controlled Drugs and Substances Act, S.C. 1996, c. 19 ('CDSA'). The charge stemmed from what was apparently a one-off transaction in a 'dial-a-dope' operation in which Mr. Wong allegedly sold a small amount of cocaine to an undercover officer. In early 2014, he entered a plea of guilty to the charge. The sentencing judge accepted that Mr. Wong was a low-level player in the trafficking of cocaine. He also noted that Mr. Wong had a criminal record, having been convicted of possession of stolen property in 1994. He convicted Mr. Wong and sentenced him to nine months’ imprisonment.
 It is common ground that, before entering his plea, Mr. Wong was not made aware that a guilty plea might carry immigration consequences. He deposed that he had not known that his immigration status could be affected by the criminal process, and that his lawyer had neither asked about his immigration status in Canada nor explained to him that it could be affected by a criminal conviction or sentence. Mr. Wong had believed the worst penalty that could flow from his guilty plea was a jail sentence.
 Mr. Wong’s trial counsel deposed that his instructions had been to do everything possible to avoid a jail sentence for his client. He confirmed that he had not been told and had not asked about Mr. Wong’s immigration status, and that he had not fully advised Mr. Wong of the possible immigration consequences of a guilty plea.
 Because of Mr. Wong’s status as a permanent resident in Canada, his conviction and sentence did have two serious consequences under the Immigration and Refugee Protection Act, S.C. 2001, c. 27 ('IRPA'). As a first consequence, Mr. Wong was rendered inadmissible to Canada for serious criminality. Section 36(1) of the IRPA provides that a permanent resident will be 'inadmissible on grounds of serious criminality for (a) having been convicted in Canada of an offence . . . punishable by a maximum term of imprisonment of at least 10 years, or of an offence . . . for which a term of imprisonment of more than six months has been imposed'. A conviction for trafficking in cocaine carries a maximum penalty of life imprisonment under s. 5(3) of the CDSA. Mr. Wong was sentenced to nine months’ imprisonment. Thus, he was rendered inadmissible for serious criminality on both grounds.
 A permanent resident who is inadmissible to Canada may be referred to an admissibility hearing, at which the Immigration Division of the Immigration and Refugee Board will decide whether to allow the person to remain in Canada or to make a removal order requiring the person to leave Canada: IRPA, ss. 44(2) and 45. Where a removal order is made and is not subsequently stayed or set aside on appeal, the person loses permanent resident status and must leave Canada immediately: IRPA, ss. 46(1)(c), 48 and 49(1)(a).
 A removal order can be appealed to the Immigration Appeal Division, which may take into account humanitarian and compassionate considerations in favour of setting aside the order: IRPA, ss. 63(3) and 67(1). However, there is no right to appeal for a permanent resident who is inadmissible because of a crime that was punished in Canada by a term of imprisonment of at least six months: IRPA, s. 64(1) and (2). Therefore, the second immigration consequence of Mr. Wong’s nine-month sentence was the loss of his right to appeal any removal order made against him, on any grounds whatsoever, including humanitarian or compassionate considerations.
 Mr. Wong first learned of the immigration consequences of his plea while serving his prison sentence when a representative of the Canada Border Services Agency contacted him by telephone. About a month after being released from jail, he received a letter from the Agency informing him that an immigration hearing would be held to decide whether he would be required to leave Canada. At that point, after having already served his sentence, Mr. Wong appealed his conviction, asking that his guilty plea be set aside on the ground that he had not been informed of its full consequences.
III. Judicial History
British Columbia Court of Appeal (Saunders, Harris and Fitch JJ.A.), 2016 BCCA 416, 342 C.C.C. (3d) 435
 Mr. Wong formulated his appeal to the British Columbia Court of Appeal as a competence of counsel claim, arguing that his trial counsel’s ineffective assistance had resulted in a miscarriage of justice within the meaning of s. 686(1)(a)(iii) of the Criminal Code, R.S.C. 1985, c. C-46. He sought to withdraw his plea on that basis. Mr. Wong deposed in an affidavit filed in his appeal that he had been unaware of the possible immigration consequences of his conviction and sentence. However, he did not specifically assert that he would have declined to plead guilty had he been aware of those consequences. Mr. Wong also sought and was granted leave to appeal his sentence, but that appeal has been held in abeyance pending the outcome of the conviction appeal.
 The Court of Appeal dismissed Mr. Wong’s conviction appeal. Each member of the panel wrote separate concurring reasons. They all accepted that Mr. Wong had been unaware of the collateral immigration consequences of his plea, but concluded that his guilty plea could not be withdrawn as it had not resulted in a miscarriage of justice. The panel was divided on the analytical framework to be applied in disposing of the appeal.
 Saunders J.A. applied the framework for assessing whether ineffective assistance of counsel has resulted in a miscarriage of justice. This framework requires the accused to establish that trial counsel’s representation fell below a standard of reasonableness expected from professionals and that the deficient counsel work resulted in prejudice to the accused which constituted a miscarriage of justice. Saunders J.A. considered the validity of a guilty plea as a subset of this ineffective assistance of counsel framework.
 Saunders J.A. concluded that there were two reasons why Mr. Wong could not succeed in having his guilty plea set aside on the basis of a miscarriage of justice. First, she held that to show that an invalid guilty plea has resulted in a miscarriage of justice, the accused must establish an 'articulable route to an acquittal', or in other words, some prospect of success with respect to the verdict (para. 26). Mr. Wong had failed to articulate any basis to avoid conviction and had consequently failed to meet this requirement. Second, Saunders J.A. found that Mr. Wong had not established that an awareness of the possible immigration consequences would have made a difference to his decision to plead guilty.
 Fitch J.A. agreed in the result, but he would have analyzed the case solely on the basis of whether the guilty plea was valid, rather than applying the ineffective assistance of counsel framework. He accepted that Mr. Wong’s plea was uninformed, but found that the prejudice needed in order to establish a miscarriage of justice had not been demonstrated. Mr. Wong had failed to specifically depose that he would not have entered the guilty plea had he been aware of its collateral consequences, and his failure to do so was fatal to his appeal. Fitch J.A. expressed reservations about endorsing an added rule that an appellant seeking to have a guilty plea set aside on the ground of a miscarriage of justice must also establish an articulable route to an acquittal.
 In brief concurring reasons, Harris J.A. agreed with the analytical framework set out by Saunders J.A., but echoed the concerns of principle raised by Fitch J.A. regarding the requirement that, for a guilty plea to be withdrawn on the basis of a miscarriage of justice, an articulable route to an acquittal must be established.
A. Guilty Pleas in the Canadian Criminal Justice System
 The overarching question in this appeal is when a guilty plea will result in a miscarriage of justice on the basis of its being uninformed. I would not base the analysis of this question on the ineffective assistance of counsel framework. The central issue in this case is whether Mr. Wong’s guilty plea was informed and constituted a valid waiver of his rights. Focusing on whether the ineffective assistance of counsel was the source of the purported invalidity of the plea only confuses the analysis.
 Guilty pleas are of central importance to the Canadian criminal justice system. For many years, a substantial majority of criminal convictions in Canada have resulted from guilty pleas: O. E. Fitzgerald, The Guilty Plea and Summary Justice (1990), at p. 1; J. Di Luca, 'Expedient McJustice or Principled Alternative Dispute Resolution? A Review of Plea Bargaining in Canada' (2005), 50 Crim. L.Q. 14, at p. 15; S. N. Verdun-Jones and A. A. Tijerino, Victim Participation in the Plea Negotiation Process in Canada (2002), at p. 1. The guilty plea is one aspect of the plea bargaining process, in which Crown and defence counsel negotiate a joint submission on sentence and the accused agrees in exchange to plead guilty. As this Court recently stated, such agreements are 'commonplace and vitally important to the well-being of our criminal justice system, as well as our justice system at large': see R. v. Anthony-Cook, 2016 SCC 43,  2 S.C.R. 204, at para. 25. The plea bargaining process is fundamental to the administration of justice: the disposition of cases by means of plea bargains benefits all participants in the justice system, preserves limited resources and introduces certainty into the criminal process (Anthony-Cook, at paras. 35-40).
 While it is true that the plea bargaining process yields important benefits, it must also be fair. This Court has long recognized the importance of the rights waived by an accused in pleading guilty: Adgey v. The Queen,  2 S.C.R. 426, at p. 440. A guilty plea constitutes a formal admission of guilt to the crime with which the accused is charged. It relieves the Crown of its burden to prove guilt beyond a reasonable doubt and constitutes a waiver of essential procedural safeguards. An accused who pleads guilty forfeits such constitutionally enshrined protections as the right to make full answer and defence, the right to silence, the right against selfincrimination and the presumption of innocence.
 In recognition of the importance of these rights, the law has imposed certain requirements that must be met for a guilty plea to be accepted as valid, namely that the plea be voluntary, unequivocal and informed. A plea will be informed if the accused is aware of the nature of the allegations made against him or her, as well as of the effect and consequences of the plea: R. v. Taillefer, 2003 SCC 70,  3 S.C.R. 307, at para. 85, quoting R. v. T. (R.) (1992), 10 O.R. (3d) 514 (C.A.), at p. 519, per Doherty J.A.; Criminal Code, s. 606(1.1).
 A person who is convicted on the basis of a guilty plea may appeal that conviction and seek to withdraw his or her plea. In Adgey, at p. 431, this Court held that an accused may change a plea in this manner if the court is persuaded there are 'valid grounds' to do so. The Court expressly declined to define the scope of this expression, though it suggested that evidence that the accused had not intended to admit to an essential fact constituting the offence, had misapprehended the effect of the guilty plea or had simply not intended to plead guilty at all might constitute valid grounds (p. 430). These examples are not exhaustive. They simply illustrate the possibility of withdrawing a guilty plea if it does not meet the criteria for validity set out above.
 The onus is on a person who appeals a conviction on the ground of an invalid plea to show that the plea was in fact invalid: T. (R.), at p. 519. The integrity of the plea bargaining process and the certainty and order which are essential to the criminal process depend on the finality of guilty pleas. The benefits associated with guilty pleas will be lost and the very functioning of the criminal justice system will be threatened if such pleas are set aside lightly. Accordingly, there is a considerable public interest in preserving the finality of guilty pleas, and the burden of showing that a guilty plea was invalid falls to the accused.
 By pleading guilty, Mr. Wong relinquished the important rights mentioned above. He argues that he did so without being properly informed of the consequences of his plea, and that his plea therefore did not meet one of the requirements for validity set out above, which have been developed to protect the rights of the accused in a fair process. He thus appeals his conviction on the ground that there was a miscarriage of justice under s. 686(1)(a)(iii) of the Criminal Code. As a result, Mr. Wong’s appeal requires this Court to consider a question of procedural fairness: Was he sufficiently informed of the full consequences of his plea such that the process in which he surrendered his rights was fair? To answer this question, the Court must determine (1) when a guilty plea will be considered uninformed because the accused was not aware of the possible collateral consequences of entering it, and (2) the circumstances in which there has been prejudice such that an uninformed guilty plea has given rise to a miscarriage of justice and should be set aside on appeal.
B. When Does an Uninformed Guilty Plea Result in a Miscarriage of Justice?
(1) The Accused Was Not Aware of a Legally Relevant Consequence
 It is well established that for a plea to be informed, the accused must be aware of its consequences: Taillefer, at para. 85. At a minimum, this entails awareness of the criminal consequences of a plea, and thus awareness that conviction and a penalty may follow: T. (R.), at p. 523. At issue is whether collateral consequences must also be known to the accused in order for his or her plea to be informed.
 Collateral consequences are consequences that are secondary or collateral to the criminal process and that have an impact on the offender: see R. v. Pham, 2013 SCC 15,  1 S.C.R. 739, at para. 11. This Court has already held that collateral immigration consequences may be relevant in the sentencing context: Pham, at para. 13. As I wrote in Pham, although the sentence must always be proportionate to the gravity of the offence and the degree of responsibility of the offender, collateral consequences such as deportation may be relevant factors in determining the fitness of the sentence (para. 24). However, the simple fact that a collateral consequence is relevant at the sentencing stage does not mean that it necessarily bears on the validity of a guilty plea. In determining whether a sentence is fit, a court must consider all relevant factors, which may include collateral consequences of the sentence. The validity of a sufficiently informed guilty plea engages different considerations. In the latter context, the ultimate issue is whether the accused forfeited his or her rights, by pleading guilty, in a process that was fundamentally fair.
 Provincial appellate courts have been divided on whether, in order for a guilty plea to be informed, the accused must be aware of its collateral consequences. Courts in Alberta and Quebec have taken a narrow approach, holding that an awareness of collateral consequences is not relevant and does not affect the validity of an otherwise informed plea. According to this approach, the assessment of whether a guilty plea was sufficiently informed, and therefore valid, is concerned only with whether the accused understood the consequences of the guilty plea for the criminal proceedings themselves: R. v. Slobodan, 135 A.R. 181 (C.A.); R. v. Hunt, 2004 ABCA 88, 346 A.R. 45; R. v. Nersysyan, 2005 QCCA 606; R. v. Raymond, 2009 QCCA 808, 262 C.C.C. (3d) 344.
 In other provinces, a broader approach has been taken to the relevance of collateral consequences in the assessment of whether a guilty plea was sufficiently informed. Courts in British Columbia and Ontario have accepted that a guilty plea may be set aside on the basis that the accused was not aware of its collateral consequences: R. v. Quick, 2016 ONCA 95, 129 O.R. (3d) 334; R. v. Aujla, 2015 ONCA 325; R. v. Shiwprashad, 2015 ONCA 577, 337 O.A.C. 57; R. v. Sangs, 2017 ONCA 683; R. v. Tyler, 2007 BCCA 142, 237 B.C.A.C. 312; R. v. Kitawine, 2016 BCCA 161, 386 B.C.A.C. 24. Although the broader approach means that awareness of collateral consequences may be relevant to the validity of a guilty plea, there is no consensus on the scope or the precise nature of the collateral consequences which must be known to the accused in order for his or her guilty plea to be informed.
 I would not endorse the narrow approach according to which collateral consequences are irrelevant to the assessment of whether a guilty plea is sufficiently informed. The requirement that a guilty plea be informed is intended to ensure that an accused who gives up his or her procedural rights does so in a manner that preserves the integrity and fairness of the criminal process. The narrow approach focuses solely on whether the accused was aware of the consequences of a guilty plea for the criminal proceedings and excludes the consideration of collateral consequences which might affect his or her fundamental interests. To endorse the narrow approach would be to compromise the ability of the accused to make an informed decision. Such an approach would be incongruous with the principled rationale underlying the requirement of an informed plea to ensure procedural fairness.
 Collateral consequences that affect the accused person’s fundamental interests could have a more significant impact on the accused than the criminal sanction itself. As a result, it may be essential for an accused to be aware of such consequences in order to enter an informed guilty plea. This is particularly true in the immigration context, in which an accused may be exposed to a collateral consequence as serious as deportation. People who are to be deported may experience any number of serious life-changing consequences. They may be forced to leave a country they have called home for decades. They may return to a country where they no longer have any personal connections, or even speak the language, if they emigrated as children. If they have family in Canada, they and their family members face dislocation or permanent separation.
 The seriousness of these consequences has led Canadian courts to adopt the broader approach and accept that an accused person’s awareness of immigration consequences is relevant to the determination of whether his or her plea is sufficiently informed. As a matter of practice, it is also well established in Canada that defence counsel should inquire into a client’s immigration status and advise the client of the immigration consequences of a guilty plea, and that counsel should raise the immigration consequences that might result from the client’s being convicted or from a particular sentence that might be imposed at a sentencing hearing. This practice is reflected in the following criminal practice form, checklists and guidelines prepared by Legal Aid Ontario and various law societies to ensure that accused persons are entering informed guilty pleas: Legal Aid Ontario, Plea Comprehension Inquiry, October 2017 (online), at pp. 2-3; Law Society of British Columbia, Sentencing Procedure, updated September 1, 2017 (online), at p. C-3-4; Barreau du Qubec, Dtermination de la peine, updated December 2013 (online), at p. 1; Law Society of Ontario, How to Prepare and Conduct a Sentencing Hearing, updated December 2016 (online), see Step 9: Prepare the client for the sentencing hearing. The provision of aids such as these by institutions of the legal profession illustrates an increasing acceptance that awareness of collateral immigration consequences is highly relevant in the criminal context and forms part of an informed guilty plea. As a point of comparison, the United States Supreme Court has also recognized the profound impact of deportation, describing it as a 'particularly severe ‘penalty’', and has imposed a requirement on defence counsel to advise non-citizen clients of the risk of deportation a guilty plea might entail: Padilla v. Kentucky, 559 U.S. 356 (2010), at p. 365, quoting Fong Yue Ting v. United States, 149 U.S. 698 (1893), at p. 740.
 In my view, the procedural fairness concerns that the informed plea requirement was originally intended to address may mean that for a guilty plea to be informed, awareness of serious collateral consequences such as these is required. I would therefore endorse a broader approach to the effect that whether a guilty plea was sufficiently informed may depend on whether the accused was aware of such collateral consequences and whether the accused, in entering a guilty plea, thus forfeited his or her rights in a process that was fundamentally fair.
 Courts that have adopted a broader approach have used the expression 'legally relevant' to describe a collateral consequence which must be known to the accused in order for his or her plea to be informed: see T. (R.), at p. 524; Quick, at paras. 28-30. I find that this expression is appropriate to describe the types of consequences that are sufficiently serious to bear on the validity of a guilty plea. For a collateral consequence to be legally relevant and capable of supporting a determination that a guilty plea is sufficiently informed, it will typically be state- imposed and flow fairly directly from the conviction or sentence, and it must have an impact on serious interests of the accused.
 A guilty plea will therefore be uninformed if the accused establishes on a balance of probabilities that he or she was unaware of a collateral consequence that is legally relevant. Legally relevant collateral consequences are not limited to the immigration context. Possible collateral consequences are so varied that what is legally relevant defies simple classification. The characteristics enumerated above are not meant to be prerequisites for legal relevance, but are simply factors a court should consider when an accused seeks to set aside a guilty plea on the basis of a claim that he or she was unaware of a collateral consequence.
 I would also emphasize that for a plea to be informed, the accused need not be informed of every conceivable consequence of the plea. While a guilty plea can trigger myriad collateral consequences which arise in a variety of circumstances, only those that are legally relevant are germane to this inquiry. Some consequences may be too remote or trivial to constitute information which must be known to the accused in order for his or her guilty plea to be informed. In my view, it would be neither necessary nor wise in this appeal to exhaustively define the scope of legally relevant consequences. The content of this concept must evolve incrementally as new cases are considered.
 I note that an assessment of legal relevance does not require a fact-specific inquiry into the significance of a collateral consequence to the accused before a court. Rather, at this step of the inquiry, the only concern is whether the consequence is sufficiently serious that it would constitute a legally relevant consequence. I am satisfied that a state-imposed consequence such as the risk of deportation without any right of appeal, which flows directly from a criminal conviction and sentence, bears on serious interests and constitutes a legally relevant collateral consequence.
(2) There Is a Reasonable Possibility That the Accused Would Have Proceeded Differently Had He or She Been Aware of the Collateral Consequence
 Even if it is shown that a guilty plea was uninformed because the accused was unaware of a legally relevant collateral consequence, that alone does not establish a miscarriage of justice. An uninformed guilty plea may raise the possibility of a breach of procedural fairness, but the court must go on to consider the effect of the lack of awareness. An uninformed guilty plea may only be set aside on the basis of a miscarriage of justice if it has resulted in prejudice to the accused.
 Therefore, at the second stage of the inquiry, a court must be satisfied of a reasonable possibility that the accused would have proceeded differently had he or she been aware of the collateral consequence, either by declining to admit guilt and entering a plea of not guilty, or by pleading guilty but with different conditions. This must be determined by applying an objective standard, modified such that a court can take the situation and characteristics of the accused before it into account. Thus, the inquiry is not concerned with whether the accused before the court would actually have declined to plead guilty. Reviewing courts must objectively assess the impact of the missing information in the particular circumstances of the accused. The question, therefore, is whether there is a reasonable possibility that a similarly situated reasonable person would have proceeded differently if properly informed.
 The applicable standard of proof is a reasonable possibility, which falls between a mere possibility and a likelihood: Strickland v. Washington, 466 U.S. 668 (1984), at pp. 693-94, per O’Connor J., cited in R. v. Joanisse (1995), 102 C.C.C. (3d) 35 (Ont. C.A.), at p. 64. Thus, a court need be satisfied only of a reasonable possibility that a reasonable person in the same situation as the accused would have proceeded differently had he or she been aware of the collateral consequence. It need not be satisfied of a likelihood that a similarly situated accused would in fact have chosen to plead not guilty: see e.g. Taillefer, at para. 111. At its heart, the inquiry is concerned with the effect of the unknown collateral consequence on the ability of the accused to make an informed decision. In other words, it is concerned with preventing the prejudice that results where information, if known, would have sufficiently influenced a decision whether to plead guilty, to the extent that there is a reasonable possibility that a similarly situated accused would have proceeded differently; it is not concerned with determining whether such an accused would actually have declined to plead guilty.
 Further, it need not be presumed that a reasonable person in the same situation as the accused would have taken the 'best' or single most rational course of action based on the likelihood of success at trial. The inquiry is not concerned with whether it would have been reasonable to plead guilty. Instead, the inquiry considers whether there is a reasonable possibility that a similarly situated reasonable person would have proceeded differently, in light of the circumstances and the seriousness of the collateral consequence at issue. This assessment does not confer 'unbounded discretion' upon reviewing courts: majority reasons, at para. 16. The reasonableness inquiry operates within the particular framework set out above.
 This inquiry reflects the reality that awareness of the possibility of legally relevant collateral consequences affects not only decisions made by the accused, but also those of other participants in the criminal justice system. For example, if Crown counsel were aware that a guilty plea on a particular charge might expose the accused to a legally relevant collateral consequence, this might have an impact on Crown counsel’s decisions in the plea bargaining process. Awareness of that fact might influence Crown counsel in deciding whether to proceed on certain charges, whether to accept a plea to a lesser included offence or what sentence to agree to in joint submissions. This, in turn, is relevant to the accused person’s strategy and tactics. Where Crown counsel offers to accept a plea to a lesser included offence, or agrees to recommend a reduction in the sentence, the accused must then assess these new developments and decide whether to plead guilty, run the risk of trial or continue the negotiation process. All of these reciprocal considerations bear on the ultimate decision whether to plead guilty.
 In short, the decision whether to plead guilty is part of an extended process that involves other decisions made by multiple players. Some of these decisions are simply out of the accused person’s control. Prejudice arises where a person, if properly informed, might have proceeded differently - either by declining to admit guilt and entering a plea of not guilty, or by pleading guilty only after seeking different conditions or a different charge. An objective inquiry focused on the prejudice flowing from a reasonable possibility that a similarly situated accused would have proceeded differently, rather than an evaluation of a subjective assertion as to a conclusive outcome, more realistically captures the nature of the plea bargaining process and the cumulative effect of the multiple decisions made within that process. Ultimately, it is compatible with the integrity of the plea bargaining process as a whole.
 I readily accept that the initial decision whether to plead guilty is fundamentally subjective for an accused and reflects deeply personal considerations. However, it does not necessarily follow that, where the accused later seeks to vacate a guilty plea on the basis of invalidity, the prejudice arising from that plea can be assessed only subjectively. As my colleagues accept, any approach to the withdrawal of a guilty plea must strike a balance between the finality of guilty pleas and fairness to the accused: majority reasons, at para. 19. Thus, while the initial decision to enter a guilty plea reflects a subjective choice made by an accused, the decision whether to strike that plea on the basis of invalidity is no longer strictly personal to the accused. It must also consider society’s interest in the finality of guilty pleas. But this is not to suggest that the public interest in the finality of guilty pleas may override the prejudice suffered by an individual accused as a result of an uninformed plea. I agree with my colleagues that where the prejudice flowing from an uninformed plea has given rise to a miscarriage of justice, the plea must be struck. Where our approaches diverge is with respect to how that prejudice is to be assessed.
 The modified objective approach strikes a proper balance between the competing interests when an accused seeks to withdraw a guilty plea on the ground that he or she was not aware of a legally relevant consequence. This test allows a court to take the situation and characteristics of the accused into account in order to properly assess whether the uninformed plea had a prejudicial effect in his or her circumstances. At the same time, the objective nature of the test reflects society’s interest in the finality of guilty pleas and militates against an accused seeking to strike a plea for capricious or trivial reasons which may in fact be unrelated to his or her being unaware of a particular consequence. It also ensures that an accused cannot seek to strike a plea on the ground that he or she was deprived of information that would have been unlikely to have an impact on the decision in the circumstances.
 In my view, the modified objective inquiry also mitigates, to a greater extent than a subjective assessment, the inherently speculative nature of the assessment of prejudice flowing from an uninformed plea. As set out above, the plea bargaining process involves nuanced and interdependent considerations. The requirement that a guilty plea be sufficiently informed is meant to protect an accused person’s right to make informed decisions within that process. It is artificial to require accused persons to state exactly how they would have proceeded had they been informed of the consequences of their plea. In other words, it is one thing to ask a judge to assess whether there is a reasonable possibility that a similarly situated accused would have proceeded differently. But it is quite another to require an accused person to specifically 'articulate a meaningfully different course of action' and then defend that speculative assertion sufficiently to withstand a rigorous credibility determination: majority reasons, at para. 22. It will often be difficult for an accused person to say precisely how he or she would have behaved differently - let alone how other participants in the justice system would have proceeded - if the consequences of the plea had been known. Presumably, an accused who chooses to appeal a conviction in light of new information will generally have a subjective belief that he or she would have proceeded differently if sufficiently informed at the time. However, in my view, prejudice is best assessed by considering objectively how the information would have mattered in the particular circumstances of the accused, on a standard of reasonable possibility, rather than by evaluating how compellingly the accused is able to describe subjective prejudice by way of affidavit and how well the accused is able to withstand cross-examination.
 As set out above, the objective inquiry should be modified to take into account the particular situation of the accused. This Court has applied a modified objective standard in a number of contexts. In R. v. Latimer, 2001 SCC 1,  1 S.C.R. 3, at para. 32, the Court applied a modified objective test to elements of the defence of necessity. It described the standard as one that 'involves an objective evaluation, but . . . takes into account the situation and characteristics of the particular accused person' (para. 32). A similar standard has been applied in relation to the defence of duress: in R. v. Ruzic, 2001 SCC 24,  1 S.C.R. 687, at para. 61, the Court drew on Latimer and applied an 'objective-subjective standard' to assess the gravity of threats in respect of that defence. It explained that in applying this standard, 'courts will take into consideration the particular circumstances where the accused found himself and his ability to perceive a reasonable alternative to committing a crime, with an awareness of his background and essential characteristics' (para. 61).
 In Taillefer, although this Court did not explicitly use the expression 'modified objective standard', the approach it took resembled that standard. In that case, the Court considered the circumstances in which an accused may withdraw a guilty plea on the basis of the discovery of fresh evidence that had not been disclosed by the prosecution. LeBel J., writing for the Court, rejected a subjective approach that would have required a court to ask whether the accused before it would have declined to plead guilty had the Crown not breached its duty to disclose. Instead, LeBel J. preferred an objective test, that is, whether there was a reasonable possibility that the fresh evidence, had it been disclosed, would have influenced the decision of a reasonable and properly informed person whether to plead guilty. But LeBel J.’s approach was not a purely objective one. Rather, his test was whether there was a reasonable possibility that a person 'in the same situation, would have run the risk of standing trial': Taillefer, at para. 90
 My colleagues suggest that Taillefer should not be taken as having endorsed a modified objective approach. With respect, I cannot agree. It is telling that in Taillefer, the appellant had filed an affidavit stating that he would not have pleaded guilty if he had known of the existence of the undisclosed evidence and his counsel had also made declarations to this effect (para. 110). Nevertheless, LeBel J. held that the appellate court had erred in applying a subjective test in determining the impact of the non-disclosure on the appellant’s decision to plead guilty. As LeBel J. concluded on this point, 'that is not the applicable test. The test is what the reasonable person in the same situation would have done' (para. 111). In the circumstances, LeBel J. held that it was not unreasonable to think that a similarly situated accused would have hesitated to admit guilt. On this basis, he assessed the prejudice flowing from the failure to disclose the evidence by applying a modified objective test, as set out above, rather than by evaluating the subjective assertions made by the appellant by way of affidavit.
 In my view, the reasoning in Taillefer unambiguously indicates that an objective framework should be applied in assessing the impact of undisclosed evidence - or analogously in this case, unknown collateral consequences - on the accused person’s decision to admit guilt. My colleagues assert that 'prejudice - that is, whether the accused’s being uninformed impacted the plea - is assessed subjectively', which they conclude is 'entirely consistent' with Taillefer (para. 35 (emphasis in original)). With respect, this proposition ignores the specific dictate in Taillefer that 'the impact of the non-disclosure on the appellant’s decision to plead guilty' - in other words, the prejudice - is assessed not by applying a subjective test, but instead by considering 'what the reasonable person in the same situation would have done': Taillefer, at para. 111. The fact that this objective framework is used by a court to assess prejudice in relation to the accused before it does not transform the inquiry into a subjective one.
 As a closing concern, I note that requiring subjective prejudice to be demonstrated in the manner suggested by my colleagues might act as a procedural bar to an accused who did not understand - or was not instructed of - the need to specifically depose that he or she would have declined to plead guilty, or would have pleaded guilty only on specific conditions, had he or she been sufficiently informed. Such a procedural bar would operate despite the obvious presence of significant prejudice to the accused flowing from the uninformed guilty plea. My colleagues insist that adopting a subjective framework will not create such a procedural bar (para. 30). However, they also implicitly accept that such a procedural bar may exist by indicating that its effects will hopefully be attenuated by attentive trial judges (para. 30).
 That the requirement that an accused demonstrate subjective prejudice by way of affidavit acts as a procedural bar is evident in my colleagues’ disposition of this very appeal. In this case, Mr. Wong deposed that he was unaware that his conviction carried any immigration consequences. He appealed his conviction only after having already served his sentence. He now faces deportation without any right of appeal, consequences which flow directly from his uninformed plea. If deported, Mr. Wong will be forced to leave the country he has called home for over 25 years and will face either permanent separation from, or relocation of, his family - including his Canadian-born child. My colleagues accept that Mr. Wong was unaware of these serious consequences and that his plea was uninformed (para. 4). They acknowledge that someone in Mr. Wong’s circumstances may have elected to proceed to trial, even if offered a plea deal for a sentence of less than six months, in order to avoid inadmissibility to Canada (para. 38). Indeed, they accept that, based on Mr. Wong’s submissions, his overriding concern was to avoid deportation (para. 38). Despite these findings, my colleagues are of the view that Mr. Wong has not established prejudice giving rise to a miscarriage of justice under their subjective framework.
 My colleagues come to this conclusion because, although Mr. Wong filed an affidavit before the Court of Appeal, he did not depose that he would have entered a different plea, or insisted on different conditions, had he been informed of the consequences of his plea. In the absence of this specific incantation, my colleagues conclude that there is no basis to permit Mr. Wong to withdraw his plea (paras. 37 and 39). There is no further inquiry into whether Mr. Wong was, in fact, subjectively prejudiced by this lack of information. On this approach, in future cases, the ability of trial judges to assess the prejudice flowing from an uninformed plea will be wholly contingent on whether there is sufficiently specific language in an affidavit as to how the accused would have proceeded if properly informed. This will be so even if the inescapable conclusion, in light of the circumstances of the accused and the seriousness of the consequence, is that clear injustice flowed from the uninformed plea. In my view, to endorse such an approach risks favouring form at the expense of substance.
C. An Articulable Route to an Acquittal Is Not Required
 I would reject the approach taken by some provincial appellate courts to the effect that an accused must establish an articulable route to an acquittal before a guilty plea can be set aside: see e.g. Hunt; Nersysyan.
 In my view, the functional role of the guilty plea in the context of the criminal justice system explains why an articulable route to an acquittal should not form part of the inquiry. The guilty plea has been described as having a 'dual nature', operating as both a procedural and an evidentiary device: Fitzgerald, at p. 103. A guilty plea is an evidentiary device insofar as it substitutes for proof beyond a reasonable doubt. It is a procedural device in that it obviates the need for a trial on the merits and results in a renunciation of the accused person’s rights: Fitzgerald, at p. 103. The validity of a guilty plea must be understood in the context of these twin roles. Whether a guilty plea is valid as an evidentiary device depends on whether it represents an admission to the essential elements of the offence. Whether it is valid as a procedural device depends on whether it is voluntary, unequivocal and sufficiently informed such that the accused relinquishes his or her rights in a fair process that compensates for the absence of further procedural protections afforded by the criminal process.
 A concern that a guilty plea may be invalid because there was confusion over the factual basis of the plea and because the accused did not intend to admit to the essential elements of the offence, for example, goes to the evidentiary role of the guilty plea. In other words, the plea may not properly substitute for proof beyond a reasonable doubt with respect to elements of the offence. By contrast, a concern such as Mr. Wong’s - that a plea was not sufficiently informed - relates to the guilty plea as a procedural device. The assessment of the validity of his guilty plea must be considered in the context of the question whether he surrendered his fundamental rights in a fair process, and not that of whether his plea substantively operates as proof beyond a reasonable doubt of the elements of the offence.
 I find that an approach that requires an articulable route to an acquittal is wrong in principle because it confuses the evidentiary and procedural functions of the guilty plea. If an accused seeks to withdraw a guilty plea on the ground that he or she was unaware of its collateral consequences, the complaint goes to procedural fairness. In other words, the core failing of an uninformed guilty plea is a flawed process that has resulted in a miscarriage of justice. The need for an articulable defence, on the other hand, relates to the evidentiary role of a guilty plea - that is to say, whether it properly operates as a substitute for proof beyond a reasonable doubt. The evidentiary role of a guilty plea is not at issue in this appeal. An articulable route to an acquittal is not required before the plea can be set aside on the basis that it was uninformed.
 I do note that the strength of the Crown’s case and the viability of a defence may be factors in determining whether the fact that the accused was unaware of possible collateral consequences has resulted in prejudice. Where an accused seeks to withdraw a guilty plea on the ground that it was uninformed and he or she has a strong defence, or the Crown has a weaker case, it is more likely that the information would have influenced the decision of a similarly situated accused whether to plead guilty. On the other hand, if the accused has no discernible defence, it may be more difficult to establish that he or she would nevertheless have proceeded differently had he or she been aware of a legally relevant consequence.
 There will be circumstances, however, where the Crown’s case will be irrelevant to the assessment of prejudice. In other words, the amount of work done at the second step of the test as set out above, which assesses the prejudice flowing from the uninformed plea, may differ depending on the legally relevant consequence at issue. For example, as a matter of logic, the strength of the Crown’s case and the viability of a defence diminish in relevance when balanced against a collateral consequence as serious as deportation. Where an accused is subject to a consequence with such severe ramifications, the prejudice flowing from not being informed of that consequence will likely be easy to establish. Ho
Please Login To View The Full Judgment!
wever, where a different and perhaps less obviously serious consequence is at issue, a more exacting inquiry to assess prejudice at the second step of the test may be required. This does not create a 'variable standard of scrutiny': majority reasons, at para. 17. With respect, it is simply a matter of common sense; the more serious the consequence, the more easily prejudice is likely to be established. Of course, this always depends on the relevance of the consequence in the particular circumstances of the accused. Similarly, the strength of the Crown’s case may also become largely irrelevant in circumstances where the only way for the accused to avoid the collateral consequence in issue is by pleading not guilty and going to trial, no matter how unlikely an acquittal may be: see e.g. Lee v. United States, 137 S. Ct. 1958 (2017).  In closing, I would observe that all participants in the justice system share the concern that when an accused enters a guilty plea, it is sufficiently informed. I reject any view that requiring accused persons to be informed of the legally relevant consequences flowing from a guilty plea might impose too high a burden on the justice system. It is an accepted practice in Canada that defence counsel should be alert to their clients’ immigration status and to the potential immigration consequences of a guilty plea, and trial judges would be wise to raise the question of such collateral consequences whenever an accused pleads guilty. V. Application  Mr. Wong sought to withdraw his guilty plea on the ground that he had not been aware that a conviction and sentence flowing from that plea could expose him to serious immigration consequences. The result of his guilty plea is that he has become inadmissible to Canada. His case has been referred for an immigration hearing, and he faces the risk of deportation. He has also lost the right to appeal any removal order against him and to raise humanitarian or compassionate considerations to prevent his deportation. These state-imposed immigration consequences flowed directly from Mr. Wong’s conviction and clearly bear on his serious interests. I am satisfied that these consequences for Mr. Wong’s immigration status constitute legally relevant consequences. It is common ground that Mr. Wong was not aware that the conviction and sentence flowing from his guilty plea could affect his immigration status. I am therefore satisfied that his plea was uninformed.  I am further satisfied that there is a reasonable possibility that Mr. Wong would have proceeded differently had he been properly informed of these consequences. I come to this conclusion based on a consideration of whether a reasonable person in Mr. Wong’s situation might have proceeded differently had he or she been aware of the collateral consequences. Mr. Wong’s particular circumstances are as follows: he is a permanent resident of Canada and has lived in this country for over 25 years, he has a wife and a young Canadian-born child, and he now faces the loss of his permanent resident status and deportation from Canada. I accept that these immigration consequences would have mattered significantly to someone in similar circumstances in deciding whether to plead guilty. Indeed, these consequences may well have mattered more than any criminal sanction in the form of a custodial sentence.  In my view, the Court of Appeal erred in dismissing Mr. Wong’s appeal on the basis that he had not specifically deposed that he would have entered a different plea had he been aware of the collateral consequences of a guilty plea. First, as I explained above, the test is not a subjective one. Second, the question to ask is not whether a court can be conclusively satisfied that the accused would have entered a different plea had he or she been informed of the relevant collateral consequences. Rather, the question is whether there is a reasonable possibility that a reasonable person in the circumstances of the accused would have proceeded differently had he or she been aware of those consequences. The answer to this question must be based on considerations related to the plea bargaining process as a whole.  In this case, there are numerous ways in which this information would have influenced the decision of a reasonable person in Mr. Wong’s circumstances, such that he or she might have proceeded differently, either by declining to admit guilt and entering a plea of not guilty, or by pleading guilty but with different conditions. I am satisfied that had a similarly situated accused been aware that such immigration consequences were possible, that knowledge would have affected the course of plea bargaining negotiations. It could have changed the approach taken by the Crown in this case. That, in turn, would have affected subsequent decisions of the similarly situated accused. For example, an accused in Mr. Wong’s circumstances might have tried to negotiate a joint submission on a sentence of less than six months in order to avoid losing the right to appeal a removal order. It is possible that such an accused would have pleaded guilty only with that condition, but would otherwise have declined to admit his or her guilt and decided to go to trial. It is also entirely possible that an accused in Mr. Wong’s circumstances would have proceeded to trial even if the Crown offered a sentence of less than six months. This is because a person convicted of the offence with which Mr. Wong was charged would become inadmissible to Canada, no matter the length of the sentence imposed on him or her. A sentence of less than six months would merely preserve the right to appeal a removal order. Such an accused might therefore have proceeded to trial even if offered a plea deal with a sentence of six months or less in the hope of avoiding a deportation order. It can therefore be concluded that there is a reasonable possibility that Mr. Wong would have proceeded differently had he been properly informed of the immigration consequences of his plea.  Mr. Wong was deprived of the ability to make informed decisions on such matters in the plea bargaining negotiations. He was ultimately deprived of a fair process. I pause to note that while the prejudice flowing from Mr. Wong’s guilty plea is assessed objectively, this inquiry cannot be reduced to a mechanical assessment of the likelihood of conviction at trial. I do not accept that a reasonable person would necessarily plead guilty when faced with a strong chance of conviction at trial, even in light of the fact that a guilty plea would operate as a mitigating factor at sentencing. Regard must be had to the particular circumstances of the case and the seriousness of the collateral consequence at issue. Thus, as I mentioned above, although the strength of the Crown’s case can operate as a factor in the analysis, it is not determinative. The relative importance of this factor is reduced when compared to immigration consequences as serious as those faced by Mr. Wong following his guilty plea. I am satisfied that a reasonable person may choose to run the risk of trial, even where there is a high likelihood of conviction, rather than plead guilty and face almost certain deportation.  The Court of Appeal further erred by requiring Mr. Wong to show an articulable route to an acquittal as a condition for having his plea set aside. As I explained above, such a requirement is wrong in principle, and an accused seeking to withdraw a plea on the ground that he or she was not informed of a collateral consequence is not required to show an articulable route to an acquittal.  In closing, I note that my colleagues remark that Mr. Wong’s sentence appeal is outstanding. They say that because the Crown has conceded before this Court that a sentence of six months less a day would be appropriate, it is likely that Mr. Wong will be successful on his sentence appeal and thereby preserve his right to appeal any removal order made against him (para. 38). I would simply observe that even if Mr. Wong is successful on his sentence appeal, he will still be subject to a removal order as a consequence of his conviction. In any event, the likelihood of success at a sentence appeal has no bearing on the merits of a conviction appeal. VI. Disposition  Mr. Wong’s plea was uninformed and it gave rise to a miscarriage of justice. I would allow the appeal, grant leave to withdraw the guilty plea, quash the conviction and remit the matter to the court of original jurisdiction for a new trial. Appeal dismissed, MCLACHLIN C.J. and ABELLA and WAGNER JJ. dissenting.