Introduction

Addressing the relationship between life and death is a delicate policy problem. Indeed, simultaneously a very personal situation rooted in ethics, values, emotions, and spirituality and, a health concern requiring a variety of services, the end of life implies social, economic and legal decisions relative to healthcare costs, rights, and responsibilities. This issue raises the immense question of individual and collective rapports to life and death where extended reflections on individual autonomy, patient–doctor relationships, control over medical treatments, roles and relationships with relatives, and authority over the quality of one’s own life are necessary (Castra 2013). For these reasons, and not surprisingly, this “wicked” problem (Raisio and Vartiainen 2015) has appeared recurrently on political agendas of Western countries. A number of them have already adopted legal frameworks to address the question (Government of Canada 2016), most of which include medical-aid-in-dying (MAID) along palliative care services. For example, the states of Oregon (1997), Vermont (2013), Washington (2008), and California (2015) allow only physician-assisted suicide (i.e., helping the patient commit suicide by providing lethal substances that the patient self-administers). The Netherlands (2002) and Luxembourg (2009) permit physician-assisted suicide and voluntary euthanasia (i.e., administration of substances, usually by a physician, that cause death). Belgium (2002), Columbia (2015), and Canada (2016; Québec in 2015) allow voluntary euthanasia.

Debates around euthanasia and assisted suicide periodically generate news coverage, particularly in tragic cases such as Terri Schiavo (Annas 2005) in the USA; Ramon Sampedro in Spain (Guerra 1999); and Sue Rodriguez (Koch 1996), Robert Latimer (Government of Canada 2013), and the Carter Case (Government of Canada 2015) in Canada. In Québec, the cases of two women called “Marielle Houle,” one who helped her 36-year-old son to die in 2004 and the other, a woman with a degenerative disease that left her severely handicapped and whose husband helped her die in 2005, revived the debate. It contributed to the introduction in June 2005 of a private bill in the House of Commons by Francine Lalonde, a Member of Parliament, to amend the federal Criminal Code to permit “the right to die with dignity” (Government of Canada 2005).Footnote 1 Although the bill was defeated, as a first attempt to decriminalize assisted suicide, it paved the way for the Québec law under study here. Ten years later, Bill 52, the Act Respecting End-of-Life Care (2014), was implemented in December 2015 in Québec (Québec 2017), the second-most populous province in Canada. It was the first in the country to allow voluntary euthanasia.

As a concrete example of a radical policy change, the Québec legislation is rather engaging from a problem construction perspective. It provides an opportunity to study the framing of a recent policy decision as opposed and complementary representations of an issue nested in what can be recognized as a paradigm change (Hall 1993). This is the primary goal of this paper, as narratives drawn from opinion articles (Letters to the Editor) were deconstructed through their arguments and reconstructed into different framings of the issue. Narratives refer to storytelling where “(…) plot, structure, meaning, resolution and so forth—are created by people conversing and arguing with others” (Fischer 2003, p. 162) and are therefore “considered integral to the social expression of cultural values” (Hampton 2004, p. 262). Framing is a process (van Hulst and Yanow 2016) that structure this information into organized principles (Fischer 2003). Ranges of stakeholders and stakeholder networks have been involved directly or indirectly in the formulation of the legislation, constructing different interpretations of the problem based on otherwise common themes whose meanings can vary widely (such as dignity, autonomy, sanctity of life). Moreover, the case, although itself interesting, goes beyond its geographical and social specificities and speaks to both end-of-life issues in general and MAID specifically as discourses share similarities with what can be witnessed elsewhere in the world. As such, this analysis contributes cumulatively to knowledge on end-of-life policy making.

Finally, but most importantly, this case offers an additional opportunity to question empirically the soundness of morality policy as a distinct category of public policy. Indeed, at first view, MAID fits the description, as it is an issue “primarily and predominantly shaped by values of the first principle” (Knill et al. 2015, p. 100) on which compromises are difficult (Mooney 1999; Mooney and Lee 1995). Yet, interestingly enough, MAID cases are barely studied from a morality policy perspective (Glick and Hutchinson 2001). Rather, morality policy research has mostly focused on other highly emotional issues, such as gambling (Engler and Dumig 2016; Ferraiolo 2013; Pierce and Miller 2001), gun control (Hurka and Nebel 2013; Godwin and Schroedel 2000), death penalty (Mooney and Lee 1999a, b, 2001), prostitution (Schmitt et al. 2013), abortion (Kreitzer 2015; Norrander and Wilcox 2001; Camobreco and Barnello 2008; Smith and Tatalovich 2003; Studlar 2001; Mylchreest 2001; Mooney and Lee 1995), embryos (Clifford et al. 2015; Snow, 2009, 2012, 2014a, b), LGBTQ-related policies (Mucciaroni 2011; Miceli 2005; Albæk 2003; Haider-Markel 2001; Haider-Markel and Meier 1996), and tobacco and drug use (Ferrairolo 2014; Omori 2013; Euchner et al. 2013; Bowen 2012; Studlar 2008). Moreover, among these, Canadian research objects are nearly nonexistent and restricted to one particular issue, that of Artificial Reproductive Technologies and commercial surrogacy (Snow 2014a, b). By adopting a framing perspective on morality policy to understand policy change, this study offers an analysis that not only contributes to the variety of policy issues investigated under that literature but also consolidates our understanding that so-called morality issues can be framed in ways other than through a moral perspective. The next section presents the case and a brief overview of the literature on MAID and policymaking in Canada and Québec.

The act respecting end-of-life care

Discussions leading to the adoption of the Act Respecting End-of-Life Care accelerated after a publication from the Québec Physicians Board in 2009. Their report reframed the debate around decriminalizing euthanasia into a question of appropriate end-of-life care, arguing that approaching this issue from a legal perspective only was insufficient to address all possible situations that physicians faced. This report motivated the Government of Québec to appoint a special commission that held public and targeted consultations between 2009 and 2011. The commission heard from 32 experts, 239 people at public hearings across the province, and 114 people during open microphone periods, and received 272 written submissions, 6558 online questionnaires, and more than 16,000 comments by various other means of communication (Assemblée Nationale 2012). In 2012, the Québec government issued the report Dying With Dignity (Mourir dans la dignité). This report was reviewed by a committee of experts who concluded that medical aid in dying should be included, in certain circumstances, as part of the continuum of care. The law, implemented in December 2015, establishes rights with respect to end-of-life care and rules for the providers as well as regulations related to continuous palliative sedation, to the powers of the Minister of Health and Social Services, and to advance medical directives. It defines MAID as “care consisting in the administration by a physician of medications or substances to an end-of-life patient, at the patient’s request, in order to relieve their suffering by hastening death” (Québec 2017, p. 4).

The implementation of the provincial act was followed by an amendment to the Canadian Criminal Code (and other acts) in June 2016. In Canada, provinces have authority over several policy areas, comprising health, and MAID was included in the Québec bill as a complementary healthcare provision. However, criminal law falls under the federal jurisdiction and, at the time of the Québec bill’s passing, prohibited assisted dying. At about the same time as the provisions of the act were to come into effect, the Supreme Court of Canada heard an important case (the Carter caseFootnote 2) and was asked to determine whether the current prohibition in the Criminal Code on assisted dying was unconstitutional under the Canadian Charter of Rights and Freedoms. The Supreme Court heard the appeal in October 2014, and in February 2015 ruled unanimously that it was indeed unconstitutional. The court gave the federal government 12 months to prepare a new legislation on MAID. That decision was favorable to the province of Québec because it removed any legal barrier preventing its bill from going forward. However, on December 1, 2015, the Québec Superior Court ruled that certain provisions of the act were in conflict with the federal Criminal Code and that until the Supreme Court decision on the Carter case came into force, the federal law would prevail. This decision had the effect of rendering the provincial provisions inoperative. Yet, 8 days later, a leave to appeal was granted by the Québec Court of Appeal and the proceedings in the Superior Court of Québec were suspended.

Because of its recent implementation, only a few studies have focused on the act itself or on the debate preceding its adoption and implementation. Landry et al. (2015) discussed the ethical considerations that should be taken into account in policy development around this issue. Comparing the cases of Belgium, the Netherlands, Oregon, and Québec to contextualize the recent development of MAID in Ontario, the authors raised questions about patient autonomy, age of consent, autonomy of the service providers and other professionals (e.g., pharmacists), and the preferred governmental level for overseeing the implementation of these new laws. In another comparative perspective, Boivin et al. (2015) evaluated euthanasia practices in 19 countries and advised caution to the potential effects of MAID policies on medical practices. The authors noted that methodological variations and the biases induced by the different policy contexts are rarely taken into account, and so several factors (e.g., policies, health systems, cultural and religious characteristics) must be accounted for when considering the transferability of different international experiences.

On the general topic of MAID in Canada and aside from research analyzing legal aspects (Schafer 2013; Judo 2013; Meisel 2003; Steunenberg 1997), part of the literature focuses on the meaning of certain expressions, such as “dying with dignity”. Saint-Arnaud et al. (2007), using a philosophical anthropology perspective, shows that it is not so much death itself that is feared but the accompanying pain and suffering. Dignity is a concept that is rather elusive because it relates to individual perceptions about what dying with dignity might mean. The semantic ambiguity of certain concepts related to euthanasia has led some authors to propose that debates around that issue should proceed carefully (Volant 2012). In the same vein, Marcoux et al. (2007) concluded that the public’s lack of understanding of what constitutes euthanasia was positively related to opinions about its acceptability. For the authors, the importance of the wording of surveys and education of the population on these questions should not be minimized if a sound public debate on the legalization of the practice is to take place.

Analyzing the controversy, several authors who have taken up the philosophical and legal arguments have been interested in their historical evolution as demonstrations of either the danger of adopting MAID legislations (Giroux 2012) or the difficulty to confirm empirically the “slippery slope” argument (Schafer 2013). Others (Lamarche 2007) have reviewed highlights of the debate surrounding euthanasia in Canada to understand the evolution of the Québec discussions initiated in 2005. More recently, Marchand (2011) examined the Québec Physicians Board’s reflection on the End-of-Life Act and suggested that it has made a welcome change in an otherwise legal debate by directing it in another direction, toward care and decision-making. Others have considered the social construction of the “physician-assisted suicide” concept by demonstrating how stakeholders use existing evidence to influence the debate (Karsoho et al. 2017).

Very few have spoken to the utility of framing and discourses surrounding MAID to inform public debate (Kaufert et al. 2013). Verbakel and Jaspers (2010), for example, tested four main arguments (religious, slippery slope, patient autonomy, dignity) and showed that the religious argument was the most prevalent (as religious citizens were shown to be the most opposed to euthanasia). Conversely, those who attach great importance to individual autonomy, or those living in countries where it is greatly valued have more favorable attitudes towards euthanasia. Finally, although some research can be found on media representations of the issue (Birenbaum-Carmeli et al. 2006; Haller and Ralph 2001; Marcoux et al. 2007, Worten and Yeatts 2000), Canadian analyses on the subject have been rare and focus mostly on physician perceptions (Wright et al. 2015).

As seen from the limited research addressing MAID in Canada from a policy perspective, there is a clear need for more consideration on this issue. In order to fill the gap, our research borrowed from both morality policy and policy framing literature to inform the analysis.

Morality policy by framing

Morality policy is linked to conceptions of morals or mores, “which are commonly held beliefs among communities derived from cultural norms” (Studlar 2001, p. 37). A conflict between mores necessitates a particular codification of values in society via legislation (Ibid.). Bowen (2012) explains that the role of morality policy is to provide “a lens that can be used to gain insight into the creation, implementation, and effects of policies that attempt to regulate personal and moral behavior” (p. 122). An important part of the literature defines and describes morality policy as a particular category of policy because of its specific features: issues of first principle, technical simplicity, high salience, public interest, and public participation (Mooney 2001; Mooney and Schuldt 2008; Studlar 2001; Bowen 2012; Arsneault 2001). Issues of first principles mean that a policy addresses questions of right and wrong, producing value-based conflicts wherein compromises are difficult (Mooney 1999; Mooney and Lee 1995). As matters of first principle, such issues naturally draw on criminal law, which is distinctive of this policy type (Knill 2013, p. 315). Morality policy is often understood to be technically simple because, as Mooney and Lee (2001) explains, “the debate is about first principles, not instrumental policy impacts, almost anyone can legitimately claim to be well informed” (p. 7). On such issues, expertise is not required in order to form an opinion. Instead, morality policy comes down to questions of right and wrong and fundamental beliefs that individuals possess. As well, partly due to their simplicity and personal relevance, such issues may be more salient to the public and thus can be engaged with more easily. This enhances their political significance and visibility, leading to greater-than-normal public participation. These features also award the media a heightened role in the policymaking process, as they can inform and sway public opinion (Mooney and Lee 2001; Haider-Markel and Meier 1996; Studlar 2001, p. 39).

Authors categorize a policy as moral either because it has a moral a priori status (substance) or through the communications involved (framing), whereby actors engage on the issue on the basis of, or with appeal to morality (Heichel et al. 2013, p. 319).Footnote 3 Substance definitions, also referred to as “content-based understanding” definitions (Engler and Dümig 2016, p. 3), are those that define morality policy by the value-based nature of its issues (Heichel et al. 2013, p. 319). As described by Camobreco and Barnello (2008), “these policies have in common a perception that much more than just policy questions are at stake in the political struggle over them. Rather, conflicts over morality policies are conflicts over deeply held fundamental beliefs about right and wrong” (p. 49). It is the role of deep, moral beliefs that, in the substantive view, gives these policies their particular character (Clifford et al. 2015, p. 229). Regulation of values, rather than the material questions of economic policies, is at the core of the substantive perspective (Studlar 2001, pp. 38–39; Studlar et al. 2013; Heichel et al. 2013, p. 320; Braun and Jörgens 2013, p. 1).

For those who subscribe to a framing definition of morality policy, it is the agency of the actors involved that determines what constitutes morality policy based on how they present and discuss the issue. Framing processes, as such, create morality policy instead of possessing an a priori existence (Euchner et al. 2013, p. 373). Actors invoke frames to defend their values against a perceived threat (Rapp et al. 2014, p. 422; Omori 2013, p. 520) or against sin (Meier 2001). The types of actors who apply frames is up for debate, given that authors such as Mucciaroni (2011) focus on interest groups, whereas Mooney and Schuldt (2008) classify the morality policy framework as “an assessment of people’s beliefs and attitudes about it”, referring to the general population (p. 201).

Mucciaroni (2011) maintains that morality policy does not exist per se and that the moral dimension is only “one of two broad strategies for framing issues” (p. 188). Unlike other authors (Haider-Markel and Meier 1996; Mooney and Lee 2001; Kreitzer 2015), Mucciaroni considers a morality frame to entail how actors prioritize certain values. Specifically, “morality policies concern “threats to core values” not because the values are “core” but, at a more fundamental level, because those who frame the issues place adherence to moral principles above alternative considerations” (Mucciaroni 2011, p. 191). The moral frame is contrasted against another strategy, the rational-instrumental policy frame, “in which we judge policies as valuable only if they help to produce desired results” (Ibid.). Instead of specific principles, it is the products of policy that are of interest. For instance, a moral frame may address drug use as either condoning or preventing immoral behavior, whereas a rational-instrumental frame may treat the matter as a public health issue, looking at how to improve overall societal outcomes in terms of lowering overdosing rates. With rational-instrumental framed issues, compromise and collaboration are possible, unlike moral issues that are indivisible and may result in an intractable controversy. Studying opposition to gay and lesbian rights, Mucciaroni (2011) found that the morality frame was deemphasized to the benefit of instrumental and procedural arguments.

According to Mucciaroni (2011), issues can be further differentiated as “fully framed” if both sides of an issue use the same type of frame, or “hybrid” if each side uses a different frame, as well as in relation to the behavior at the center of the moral argument (pp. 193–94). Arguments focusing on private behaviors concern individuals engaged in a particular behavior and the other people affected. Social arguments focus on the transgression of important moral principles or social values. Government arguments are directed at judging (as good or bad) the actions of governmental bodies that promote or undermine certain moral principles such as justice, equity, and order. Ferraiolo (2013), adopting Mucciaroni’s (2011) argument on framing to evaluate “moral talk” on gambling in the USA, found that both rational-instrumental and morality arguments frame the debate and that personal behavior frames are less common than expected whereas governmental morality frames are more prevalent. Complementing this research, Ferraiolo (2014) studied the nature of the debate over marijuana decriminalization in the United States and concluded that the issue is framed upon three sets of arguments: negative effects on youths, negative effects on families and society, and negative effect on the individual (i.e., user).

The pertinence of Mucciaroni’s (2011) proposition on morality policy is supported by the literature on policy framing as its study facilitates the recognition of competitive conceptions and interpretations of a social problem and possible solutions by interest groups, policymakers, experts, and citizens. As such, it sheds light on dimensions that will be included and excluded from public discourse (Fischer 2003). This impacts both the nature of the issue and the policy. Indeed, Schoën and Rein (1994) showed how public policies are built on perceptions, beliefs, and subjective evaluations, and demonstrated that not only do preferences for a given framework provide a justification for the proposed solutions, they also change the problem itself. Frames, as Entman (1993) described, define problems—determine what a causal agent is doing with what costs and benefits, usually measured in terms of common cultural values; diagnose causes—identify the forces creating the problem; make moral judgments—evaluate causal agents and their effects; and suggest remedies—offer and justify treatments for the problems and predict their likely effects (p. 52; emphasis in original).

Frames structure information into organized principles (Fischer 2003) and can be compared with the concept of social representation (Negura 2006; Moscovici 1988), which fosters the idea that public policy emerges from a socially constructed world that provides explanations on the importance and democratic possibilities of civil society, on targets of public action, and on preferred values in the policy process (Ingram and Schneider 1993, 1997). Social representations are sociocognitive constructions (Abric 1994) because they originate from the communication (social dimension) of individual perceptions (cognitive dimension). Drawing on these cognitive aspects, framing provides an alternative explanation for decision-making, as opposed to relying on a rational model. Tversky and Kahneman (1986) argue that different solutions will be chosen depending on the affectivity and preferences of the actors. For example, presenting the same information either positively or negatively will influence the direction of opinions on an issue (Druckman 2004). To that effect, Gamliel (2012) found that support for euthanasia was higher if “presented as not prolonging life relative to ending life” (p. 699).

Although little can be found in the recent literature on the influence of public opinion on MAID policy, a number of authors in the last 10 years have written on media construction of assisted suicide or euthanasia. Through a textual analysis of media coverage, Birenbaum-Carmeli et al. (2006) studied three cases of assisted suicide performed by family members in Great Britain between 1996 and 1998. The authors highlight the political importance of media representations in that “the narrative of the courageous individual who refused to go on living a hampered life dovetailed with neoliberal discourse of the entrepreneurial individual and the related motion toward State withdrawal from caring labour” (p. 2161). Similarly, Gailey (2003), in a book on the Kevorkian case, analyzed how the media placed the discussion into two main frames—medical and legal—and two categories of actors—pro-life or pro-right-to-die.Footnote 4 Similarly, Kalwinsky (1998) demonstrated that media portrayals of assisted suicide tend to emphasize the voices of doctors and courts rather than patients, as well as ignore class and gender issues. Likewise, Holody (2011), in a study on the links between media coverage of assisted suicide and personal opinions on the subject, showed that legal framework was the most common aspect of news coverage, but that respondents’ opinions were predicted by their preexisting prejudices and personal frames about assisted suicide, and not necessarily by the media’s frames. Pollock and Yulis (2004) offered a different explanation by taking a contrasting approach to studying physician-assisted suicide (PAS). By asking how society influences media, they concluded that “newspapers ‘align’ themselves rather precisely with social change, in direct proportion to the relative presence or absence of a multiplicity of key groups or sectors in their own communities that may have some stake, perceived or real, in that change” (p. 298). Similarly, in an analysis of medical journals, social movement organizations’ publications, and print media, Siu (2010) showed that these three stakeholders’ communities constructed different realities of euthanasia and assisted suicide through their respective interpretation and rendering of the Kevorkian case.

Our research focuses on public opinion contributions to the debate that led to the adoption of the Québec law. As a number of authors demonstrated (Norrander and Wilcox 2001; Mooney and Schuldt 2008; Camobreco and Barnello 2008), public opinion informs beliefs, preferences, and attitudes, which in turn condition framing. By documenting four dimensions identified by Mucciaroni’s (2011) and Ferraiolo’s (2013) (types of frames, types of arguments put forward; behavioral changes targeted by the moral frame and proximity of the arguments), this research reconstructed the four following framings as advanced and transmitted through the Letters to the editor’s narratives in selected newspaper articles: legality is not morality, no dignity without autonomy, the lucrative Pandora’s box, and liberating death.

Methods

A search for articles published between 2005 and 2015 (before the implementation of the act) in Québec newspapers (French and English) was conducted using two databases: Eureka for French newspapers and Factiva for English newspapers. The search query was based on a combination of keywords.Footnote 5 Articles retained had to deal with the legislation. The formulation and adoption of the bill is seen as a breaking point between policy continuity and policy change and, therefore, a catalyst for the expression of arguments/ideas/positions. Through their Letters to the editors, citizens participate in an embodied debate rather than in a pure ideological one. This logic permitted to circumscribe the narratives to “participants” that would (1) manifest their position toward the bill (opponents/proponents) (2) sustained their position with arguments and (3) potentially propose a solution. Positions regarding the proposed law were important to grasp in order to situate them in relation to the type of arguments (moral, rational-instrumental or mixed) citizens put forward. Redundant or off-topic articles were excluded, resulting in a total of 175 opinion articles (Letters to the editor).Footnote 6 Out of these, 39 were discarded because they did not present any explicit argument or position on the proposed bill. This resulted in a final selection of 136 opinion articles. The decision to analyze the Letters to the editors rests on the premise that they are fundamentally different from other newspaper articles and are an important forum for public discussion (Nielsen 2010). They give a voice to citizens (public opinion) as active participants in the debate and offer an appreciation of how end-of-life care (and MAID specifically) is understood and of how decisions about it are perceived in the population. We do not ignore that media have diverse effects on a policy issue (Iyengar and Reeves 1997; Entman 1993). However, when it comes to Letters to the editors, although we can assume that a certain strategic selection has been made by newspapers’ editors, the criteria upon which individual submissions were chosen for publications is unknown. Letters to the editors are therefore understood as a conduit for citizens to participate in the debate (Shanahan and al. 2008), albeit a mediated one.

All articles were read twice and manually coded in two rounds with NVivo11.Footnote 7 A total of 71 codes were created to deconstruct the newspapers’ articles. In the first round, all articles were loosely coded, looking for end-of-life care/MAID definitions/descriptions and any framing indications by both proponents and opponents of the debate. This first round facilitated a familiarity with the material, upon which a plan was developed for the second round. In the first round, articles were also coded for the year of publication, the typeFootnote 8 and gender of the contributor, as well as for the standpoint regarding the adoption of the bill (for, against and, nor for or against). In the second round, coding was narrowed to document Mucciaroni’s (2011) dimensions. Codes were developed to capture the spirit of the arguments put forward. Any deontological arguments, for example, allusions to sanctity of life, trivialization of death, immorality or loss or morals, dignity, humanity, respect, were coded as moral. Conversely, arguments relative to the potential effects (positive or negative) of adopting the bill were coded as rational-instrumental. Among those, reduction or loss of services, slippery slope, setting a legal precedent and consequences for the person or the families were the most common. Twenty articles included both moral and rational-instrumental frames. Once completed, the coding made it possible to distinguish moral framings from rational-instrumental ones, in and of themselves and according to the opponents and proponents in the debate, as well as the degree of proximity between the arguments (fully framed or hybrid). Each article was also coded for the level (individual, social, or governmental) and nature of the behavior that was the focus of the moral argument. For example, personal behaviors arguments related to obligations toward the person or the family. It is worth noting that personal behaviors were only mentioned by 14 contributors. However, arguments targeting social and governmental behaviors were common. Social behaviors’ arguments reflected the potential transformation of society (detrimental or positive). Increase in services (new or existent), citizen security and general respect for citizens are examples of targeted governmental behaviors. Repetitive codes from the first and second rounds were deleted and similar codes were merged together under a common identifier. Following Ferraiolo (2013), the relative importance of each frame was assessed for both opponents and proponents by the number of mentions of a particular frame (or both) as a percentage of the total number of articles as well as a percentage of the total number of articles for each position.

Four framings were drawn from the analysis, representing both sides of the issue: “Legality is not morality” (opponents of the bill) and “No dignity without autonomy” (proponents of the bill) are established on moral arguments whereas “The lucrative Pandora’s box” (opponents to the bill) and “Liberating death” (proponents of the bill) rest upon rational-instrumental ones.

Findings: four framings to end-of-life care

Overview

As shown in Table 1, one thing that stands out in the 136 opinions articles published between 2005 and 2015 is the appreciable gap between the number of positions “for” and “against” regulating end-of-life care by including MAID, with opponents almost twice as numerous (60 vs. 36%). This is surprising given that an Angus Reid poll in 2010 found that 63% of Canadians (78% in Québec, the highest percentage in the country) were favorable to legalizing euthanasia (Angus Reid 2010). On the other hand, aside from having presented a brief at the targeted and public consultations of the special commission, letters to the editors are an effective way to (at least) balance the policy narrative as they provide important public exposure. In theory, opponents to the debate would then privilege this additional participation vehicle to voice or re-voice their views on such a sensitive issue. Another explanation, which unfortunately cannot be sustained empirically by only analyzing the articles, might be that newspapers themselves, in filtering Letters to the editors, privileged a “counter-narrative”. Not uncommon for this type of article, a lower number of women (31%) voiced an opinion on the matter between 2005 and 2015 (among which 65% were opponents) compared to men.

Table 1 General observations
Full size table

Although it is not necessarily surprising, 49% of articles framed the issue on moral terms, 36% referred to rational-instrumental arguments, and 15% framed the issue as both a moral and a rational-instrumental question. What is more interesting, however, is that, contrary to what was expected for MAID, both proponents and opponents framed the issue on moral and rational-instrumental grounds in relatively similar proportions. This suggested that, contrary to some morality policy literature, the MAID debate seemed to encompass more than just conflicting basic moral values. These last observations led us to dig further into the details of the arguments put forward and compare them to Mucciaroni and Ferraiolo’s research findings.

For those who framed the debate in moral terms, four main themes were invoked: sanctity of life, individual autonomy, dignity, and humanity/compassion. On the other hand, abuse/slippery slope, service reduction, protection of the most deprived, and individual and social consequences composed the rational-instrumental position. The four framings presented below were reconstructed along those themes.Footnote 9

Moral framing 1: legality is not morality

As illustrated in Fig. 1, sanctity of life is one of the central arguments employed in opposition to the bill. This argument, based mainly on religious beliefs, puts forward the idea that regulating end-of-life is inadmissible for the simple reason that life is lent to human beings before being transformed into something greater—immortality. Sanctity of life is defined as one of the most important values in society that cannot be sacrificed for the sake of a few who evoke the right to freedom. As one article puts it, “assisted suicide and euthanasia undermine the duty to live that mobilizes all our daily actions”. Other interests than those preserving life are seen as suspect, hiding self-serving motivations detrimental to society. The expression “assisted suicide” was also deconstructed to demonstrate that it is, in fact, a veiled murder. In that perspective, the physician’s role is to protect life and integrity of patients.

Fig. 1
figure 1

Moral framing 1: legality is not morality

Full size image

Connected with sanctity of life, dignity is also an important argument to those opposing implementation of the bill. As will be discussed in the next narrative, proponents who framed the issue as a moral one put forward this argument as well. For opponents, however, dignity is the basis upon which respect for life is possible. Dignity is intrinsic to human nature, and so no loss of function through illness or loss of autonomy can fundamentally affect it. Some extend the notion of dignity as compassion and respect for humanity that physicians need to exhibit instead of participating in the fatal act.

Critics of the legislation supported their moral position mainly with arguments targeting the spirit of governmental actions and attitudes. Regarding attitudes, arguments concern the necessary respect government must show toward patients that are not able to voice their opinion, as well as for the policy process itself as decision-makers are blamed for being insensitive to opposing views and determined to short-circuit the discussions. With respect to actions, alluding to the negative effects of service cuts (decreased accessibility and lack of funding for palliative care establishments) or pressing for palliative care services as the only humane alternative to the proposed legislation, were the most common arguments.

Moral framing 2: no dignity without autonomy

In this narrative, the notion of dignity touches on something different than in the previous one. It is not conceived as a universal concept but rather as a very personal principle. As excerpts in Fig. 2 displays, dignity is dependent on the individual right to enact one’s will. In likening dignity to a right, contributors relied on personal stories of family members, usually parents, who did not die as they would have wanted to (as if that final life stage had been taken out of their control). As such, dignity goes hand in hand with individual autonomy, as the latter becomes the expression of dignity. In fact, not respecting individual autonomy leads to indignity. Personal autonomy means taking control over one’s own destiny, as well as over the medical system. As one contributor puts it “I consider that it is the strictest right of every human being to refuse to be subjected to the dictates of doctors, to refuse to be subjected to the therapeutic persistence of which modern medicine abuses with impunity, to refuse to have to suffer an atrocious death because ‘medicine can no longer do anything’.” Individual autonomy, similar to sanctity of life in the first narrative, therefore rests on important moral values, albeit individualistic ones. If proponents do not dispute the fact that physicians cannot be compelled to perform assisted suicide, they, however, clearly state that a patient cannot be refused the right to decide how and when to end his/her life. As such, respect for individual autonomy is considered a core component of humanity. Opposing the sanctity of life argument of the first narrative (that life has been lent to human beings), freedom is put forward as something that is given. One’s life, as such, is entirely one’s own and can be disposed of as one wishes. From that perspective, the person is sacred, not life, and “it is up to the person to adopt the strategy that suits him/her best, to relieve suffering as much as possible or quickly put an end to his/her life”.

Fig. 2
figure 2

Moral framing 2: no dignity without autonomy

Full size image

In this second narrative, humanity and compassion are seen as the necessary quality for understanding what dignity truly means. Only through compassion, the argument runs, is it possible to see how suffering and lack of autonomy undercuts dignity. Contrary to the first narrative, compassion and humanity are used in opposition to the sanctity of life, which is positioned as an absolute that could lead to “inhumane postures or cruelty” through the prolonging of suffering: “although compassion must be an absolute priority, it must not unconditionally have the last word, especially if it translates into a ‘condemnation to live’ against any desire of the dying. … When suffering resists all our efforts, that death precedes the intimate values of the dying and that he asks for it lucidly, euthanasia is not ‘a resignation of our human communities from suffering’”. Humanity and compassion are also seen as a necessary condition for considering the human being behind the sickness.

This narrative targets social behaviors, presenting the adoption of the bill as a step toward a better society or, contrarily, denouncing opponents as agents responsible for society’s future decline into abuse and discrimination of patients. In other words, as conceived in this narrative, protecting dignity lead to a more humane society.

As will be illustrated through the two other narratives below, the MAID debate is equally framed on rational-instrumental arguments, emphasizing the negative consequences of adopting or rejecting the proposed law.

Rational-instrumental framing 1: the lucrative Pandora’s box

This narrative against the implementation of the legislation is mainly constructed around the slippery slope/abuse argument. MAID is presented as a path that decision-makers should not take because of the probable drift toward an ever-growing expansion of assisted suicide cases, possible abuse, as well as toward the unduly broadening of health professionals’ powers with consequent damages to the province’s population. To back up this idea, “heard examples” of problems experienced by certain countries that have legalized MAID are recalled. Pushed to the extreme, this position is supported by an argument about a “large number” of ill-intentioned people and “accomplices” who would take malicious advantage of such a permissive law. This argument relates to protecting the most vulnerable, where the sick or elderly are seen as inevitably bound to accept decisions made by others about their life.

In this narrative, governmental behaviors are targeted. As illustrated in Fig. 3, the palliative care services shortages are used to justify the slippery slope/abuse argument. Indeed, some claims that the underlying motive of the bill is to reduce services to alleviate pressures on the health system. This assertion is substantiated by reference to the fact that the bill, by including terminal palliative sedation, redefined the meaning and purpose of palliative care “unilaterally and without consultation with palliative-care specialists”. Increasing the number of efficient palliative services is thus preferred to the adoption of the bill. This call for more services is also reinforced by the certitude that adopting any law that would render assisted suicide accessible would change the essence of the medical practice for the worse.

Fig. 3
figure 3

Rational-instrumental framing 1: the lucrative Pandora’s box

Full size image

Rational-instrumental framing 2: liberating death

The fourth narrative is also constructed through the slippery slope/abuse argument, but, contrary to narrative 3, as a cautionary statement against the artificial prolongation of life. In essence, the narrative decries the social construction of dying as something that needs to be remedied, of perceiving death “as a weakness”. Medical progress, rather than providing relief, is unduly transforming a natural life stage by prolonging the agony.

In this narrative, the slippery slope/abuse argument leads to positive and negative individual and social consequences. On the positive side, the adoption of the bill is seen as a liberating care for all, including the patient (end of agony) and his/her family (end of feelings of helplessness). Society is also seen as socially and financially benefiting from this ending of agony. On the negative side, the slippery slope/abuse argument warns opponents about the danger of inducing even more dramatic social situations, such as suicide in patients and consequent distress among caregivers and health professionals.

As schematized by the outer circle in Fig. 4, the slippery slope and the individual and social consequences arguments are strengthened by a third one, the decrease in palliative care services, which is also used in the third narrative. Targeting governmental behaviors, this narrative implies that the decrease or insufficient provision of palliative care services is a situation that will not change in the near future, thus requiring additional end-of-life options for patients. Based on this observation, the adoption of the bill is seen as liberating.

Fig. 4
figure 4

Rational-instrumental framing 2: liberating death

Full size image

Discussion and conclusion

The four framings drawn from the opinion articles offer an initial, albeit important entry into the construction of MAID in Québec as a policy problem. Because of the nature of the arguments put forward by opponents and proponents regarding the formulation of the Act Respecting End-of-Life Care, we believe our analysis goes beyond this specific case. Indeed, considering the comparable support for MAID in the rest of Canada, in addition to converging views about the recognition that the debate is neither uniquely moral nor solely a health issue (Angus Reid 2016), we expect the results to be relatively comparable in terms of the distribution of the arguments between moral and/or rational-instrumental framings.

Because of the nature of the issue at stake, it was anticipated that opponents to the adoption and implementation of the bill would have framed their arguments mostly on religious terms as Verbakel and Jaspers (2010) noted. Although this is more apparent in Narrative 1, the religious position did not overwhelm the debate and is only observable through the sanctity of life argument. One possible explanation rests on the fact that Québec, although Catholic by tradition, is considered to be more secular than religious. We suspect that religious arguments might be more visible in the rest of Canada and in the United States due to a lower secularization than in Québec. The 2016 Angus-Reid poll showed, for example, that opinions regarding the possible obligation of religious partner organizations in the provision of care services to comply with providing MAID, despite their moral objections, are markedly different for Québec than for the rest of Canada (3 in 5 Canadians against vs. 3 in 5 Quebecers for) (Angus-Reid 2016). We suspect this last speculation would also hold for the United States as some authors have proposed that although a secularization movement may be present, the “United States remains one of the most religious in the club of rich countries, alongside Ireland and Italy, and indeed as observed earlier, this makes America one of the most religious countries in the world” (Norris and Inglehart 2012, p. 94). Another explanation for the marginal presence of religious arguments might be that, due to the nature of the data, linking individual positions to religious beliefs was only possible to the extent that it was exposed in the article. The relationship between the argument and the secular or religious background of the contributor was therefore difficult to grasp. However, the fact that opponents of the bill framed the issue in rational-instrumental terms in a similar proportion speaks to the relatively low influence of religious beliefs on the debate.

Some arguments upon which the narratives were reconstructed, namely sanctity of life and dignity and autonomy, are similar to what others have found (Green-Pedersen 2007; Haider-Markel and Joslyn 2004; Haller and Ralph 2001). Yet, our study departs from authors (Gailey 2003) who found the legal frame to be an important one. Assertions relative to the “right to die” and the “criminalization of families” for proponents, to “euthanasia as a crime” or to the illegality of legislation with respect to the Canadian Criminal Code for the opponents, were not dominant in the opinion articles. This underrepresentation of the legal frames might be in part due to the nature of the Letters to the editors, their purpose being to voice an opinion, to influence the debate or to share a personal experience. Articles, therefore, reflect individual positions on the degree of permissiveness MAID implies and on its consequences (be they positive or negative). Another explanation may be found in the context surrounding the formulation stage of the legislation as MAID was included in a continuum of care services in the Québec legislation, therefore forcing a debate on health away from the legal one.

Using the Letters to the editors as “concrete incarnation of the abstract ‘public debate’” (Nielsen 2010, p. 32), our study also shows how policy actors—in this case, public opinion—moralize or de-moralize the issue to influence both the direction of the debate and the outcomes of the decision-making process. For instance, this research shows that specific arguments deployed as different framings are also carriers of powerful policy images. This is particularly striking with the first two framings where dignity is both the vector of life and death. Authors have argued that the criteria of dignity may be of little use for policy decision and action (Mattson and Clark 2011). Indeed, if dignity is intrinsic because God-given, it cannot be lost due to public decisions to allow MAID. On the other hand, if policies are the means by which dignity can be embodied, then dignity is not intrinsic as an individual principle. Nevertheless, dignity, as a policy image, becomes accessible knowledge, providing potentially convincing and certainly over persistent understandings of the issue (Roe 1994).

Our findings also illustrate how arguments build on each other (for example, dignity as the basis upon which respect for life is possible or humanity and compassion as prerequisites for dignity) and how some are shared by both sides of the issue (albeit with different intentions). In the third and fourth framing, for instance, the notion of slippery slope partly takes shape in the criticisms toward the state of palliative care services in the province. In essence, both sides are advocating, ideally, for increased services, one side seeing the shortage in palliative care services as the context that rendered MAID inevitable and the other considering the sustained lack of services as a convenient justification for the expansion of MAID services. Both sides are also preoccupied with individual and social consequences that may result from either maintaining the status quo or implementing the legislation. From a policy instrument perspective, identifying the nature of the behaviors targeted by both the moral and rational-instrumental frames is worth noting. Opponents to the bill directed their arguments toward governmental behaviors, whether they framed the issue on moral or rational-instrumental grounds. Lack of existing palliative services, fear of service cuts and requests for a more diversified palette of end-of-life care options are put forward as the only reasonable alternatives to MAID. Proponents, for their part, argue that adoption of the bill is in the province’s best interest as it contributes to its evolution as a progressive society. These findings support Ferraiolo’s (2013) conclusion that private behavior morality frames are less important than social and governmental ones.

Our research speaks to the complexity of the debate, an important consideration to the extent that MAID is often described as a polarizing issue, implying that proponents and opponents respective positions are irreconcilable. That may be so for some partakers, but at its face value, this affirmation does not give any measure of how and why these positions do not meet, nor does it leave any room for a more nuanced assessment of the so-called divided debate. Our research provides some leads into the “how” question by taking a closer look at the intricacies of the narratives. By deciphering what the main arguments of sanctity of life, dignity, slippery slope and individual autonomy are made of, it was possible to determine that the moral frame is not the prerogative of the opponents, as both sides use it to discuss the issue, nor is it the only frame they use to criticize the adoption and implementation of the law. In Mucciaroni’s term, the MAID debate is fully framed rather than hybrid and this, on both moral and rational-instrumental types of arguments. This is well captured by the scope the notion of dignity covers in the first and second framings or that of slippery slope/abuse in the third and fourth ones. Explaining why both sides framed the issue in moral and rational-instrumental terms, although more difficult to do, brings us to speculate in line with Mucciaroni’s findings on the best framing strategy to influence audiences. Moral frames express personal convictions and appeal to those who share similar views. They also serve as a reminder that the end of life cannot be discussed in merely programmatic terms. As such, although unavoidable considering the nature of the issue, moral frames are not sufficiently mobilizing to be internalized by citizens and public decision-makers and are, by themselves, therefore not effective enough to make a difference in the whole debate.

Conversely, rational-instrumental frames are contextualized strategies, at macro-, meso-, and micro-levels, complementing the moral framings. At the macro-level, whether used by proponents or opponents, they serve the function of questioning the social consequences of future governmental decisions regarding life and death. As such, they bring attention to the necessary safeguards decision-makers should include in the policymaking process. At the meso-level, rational-instrumental frames emphasize the necessary interdependence between MAID and the current situation of healthcare services, specifically the lack of palliative care services throughout the province. Both opponents and proponents are either concerned by a potential decrease in services or accepting of the fact that these services are, in all likelihood, doomed to remain underfunded or even disappear. At the micro level, rational-instrumental frames bring up the consequences for the individual (or the patient) considered the most important target of governmental actions. Rational-instrumental arguments are therefore multifaceted and reach out to possibly more citizens. Resorting to rational-instrumental framings may then be a more effective strategy to influence the policy process than only invoking moral ones. For example, because both moral and rational-instrumental frames coexist in the opponents’ narratives may partly explain why the bill could actually be adopted. While the introduction of MAID was an important policy change, the legislation also emphasized the need for palliative care services as an important option for end-of-life patients along restrictive regulations for accessing MAID. Since rational-instrumental arguments are more amenable to compromise and collaboration than moral ones, the costs of going forward with a potential unpopular option were downplayed.

In conclusion, our analysis reinforces Mucciaroni’s (2011) assertions that sensitive issues classified as morality policy cannot be apprehended solely through the frame of morality. Our findings align with Ferraiolo’s (2013) that opponents do rely on morality and rational-instrumental arguments to frame the issue. Our study would have missed an important dimension in how public opinion framed MAID in Québec if we had only tackled the issue as a morality policy by substance instead as a framing dynamic. Explaining the construction of MAID as a policy problem simply as an opposition of values is an oversimplification of an otherwise complex debate that took place in the province.

We believe that our research also engages with Lasswell by emphasizing the necessary inquiry into the nature of the problem (knowledge of the policy process), although, as Turnbull (2008) proposed, the underlying assumption in this paper is that the relationship between knowledge of and knowledge in the policy process is contingent, and not necessarily causal. Nevertheless, from a theoretical standpoint, morality policy may help understand conflict and difficulties in policy implementation. As outlined by Mooney (2001), noncompliance may be championed by groups who “believe that the very basis of their identity and their most fundamental values are threatened by a law”, such as activists who endure arrests to make a political statement (Mooney 2001, p. 13). Even though this topic is beyond the scope of this research, it is fair to say that the implementation process of the Québec legislation is certainly punctuated by resistance from the medical profession (although the law allows for a physician to refuse to provide MAID).

Our study also shares Laswell’s preoccupations with the need to address fundamental issues in society and to be concerned with context (Lasswell 1970)  and values (Lasswell 1951, pp. 8–10). By paying attention to the politics of problem formulation when it comes to issues that are prima facie polarized such as MAID, we build upon and extend on current knowledge about what makes an issue a moral one or not. Here, it is the policy statement that may be of interest, rather than the “outcome,” because of the manner in which the statement itself indicates the values of society, sending out “a signal to the world….that proponents of a certain policy proposal hold the right position on the issue” (Wagenaar and Altink 2012, p. 284). Our research helps understand how the narratives and counternarratives triggered by the issuing of Québec legislation’s statement cocreate the issue. This last point touches on the normative foundation of problem construction for sensitive policy issues by documenting some of the cognitive dimensions intrinsic to this important policy change, which lead to reconstruct the meanings of end-of-life care in relation to the contexts in which they unfold and from which they depend.