Autonomy Promotion in a Multiethnic Context: Reflections on some normative issues

The concept of health promotion enshrined in the Ottawa Charter grants an important place to autonomy. However, it is not clear that health broadly defined and autonomy are related in the sense intended by the authors of the Charter. Moreover, promotion of autonomy towards groups who do not consider it as a core value remains a problem. This paper presents a concept of autonomy and promotion that provide a partial answer to this problem. Examples of public policy in a multiethnic context are given to illustrate how the proposed conception of autonomy promotion differs from more coercive policies. VOLUME 5 NUMÉRO 1 PRINTEMPS/SPRING 2010 AUTONOMY PROMOTION IN A MULTIETHNIC CONTEXT: REFLECTIONS ON SOME NORMATIVE ISSUES


ABSTRACT
The concept of health promotion enshrined in the Ottawa Charter grants an important place to autonomy. However, it is not clear that health broadly defined and autonomy are related in the sense intended by the authors of the Charter. Moreover, promotion of autonomy towards groups who do not consider it as a core value remains a problem. This paper presents a concept of autonomy and promotion that provide a partial answer to this problem. Examples of public policy in a multiethnic context are given to illustrate how the proposed conception of autonomy promotion differs from more coercive policies.
The field of action specific to public health has broadened considerably over the last century. The reasons for this expansion are multiple. One of the main reasons is that the range of problems we seek to eradicate through such action is no longer -at least in Western countries -composed of infectious diseases and parasitic infections whose causes were identified and treated with the help of scientific breakthroughs in the field of medicine and epidemiology 1 . Today, a substantial portion of the problems being addressed by public health workers stem from people's living habits, specifically those habits considered to be harmful to people. We have only to think of obesity. We understand well enough, for example, what causes this problem (inactivity, among other things), but we have trouble grasping how to properly remedy such a problem, because it is largely caused by multiple distal factors (including social and environmental factors) that influence our living habits, and these factors seem too far removed causally from said problems to be clearly isolated as the causes of the problems in question. Moreover, and more importantly, these problems are, for the most part, the harmful consequences of activities freely engaged in by the persons concerned. Consequently, public health interventions now go beyond efforts to prevent problems stemming from poor living habits, to include actions aimed at promoting good health; the idea is that such promotion should target common factors that are positively associated with good health through a multidisciplinary approach, currently defined as health promotion. 2 However, if we seek to promote specific behaviours and values associated with good health -if we embrace a strong perfectionist position in public health -then, how do we deal with the fact that, firstly, people are, under certain circumstances, free to harm themselves, and secondly, that good health might be embodied in an indefinite number of lifestyles? How does perfectionism in health promotion deal with diversity, on the one hand, and with permissible harm to oneself, on the other? The goal of the present article is to offer a partial solution to this problem, by considering the role of autonomy in health promotion and by proposing a conception of autonomy that is able to properly deal with these problems.

HEALTH AND AUTONOMY
At this point, a clearer conception of what health promotion seeks to promote is required. The redefinition of the very notion of what it means to be healthy can be seen as evidence of a shift in the paradigm alluded to in the previous section. Indeed, this shift was recently given concrete expression in the 1986 Ottawa Charter, whose proposed definition of health extends far beyond the classic -negative -conception of health (i.e. health defined as the absence of pathologies). The Charter stipulates that: Health is, therefore, seen as a resource for everyday life, not the objective of living. Health is a positive concept emphasizing social and personal resources, as well as physical capacities. Therefore, health promotion is not just the responsibility of the health sector, but goes beyond healthy lifestyles to wellbeing (WHO, 1986).
According to the Charter, the goal of health promotion is to help individuals "reach a state of complete physical, mental and social wellbeing". The preferred means of achieving this state of wellbeing and, therefore, the goal of much health promotion action, is to ensure the capacity of individuals to exercise control over their lives, to direct their own lives -in short, to be autonomous. In fact, many studies have demonstrated that individuals who have little or no control over their lives are generally less healthy than others. 3 Autonomy is thus seen as crucial to an individual's health, understood here in the broad sense of the word. Of course, the importance of autonomy is not limited to the area of health: it is one of the central political values of liberal democracies, for quite identical reasons: mainly, its contribution to human flourishing. Nevertheless, it is mainly the place of autonomy within a certain conception of health and its promotion that I wish to consider.
Cleary, the move from a narrow to a broad conception of health -or health as wellbeing -can pose problems on many levels. 4 Among other things, a broad definition of what constitutes health is valueladen. 5 If we consider autonomy, it is clear that it is not only the capacity to make meaningful choices, but also a moral value, the importance about which reasonable people can disagree. Thus, if we focus specifically on the central role assigned to autonomy in the broad conception of health, are we led to conclude that the health of persons who do not appear to live autonomously is therefore inevitably diminished as compared to that of those around them? This problem derives from the fact that there seem to be a good number of people for whom a life in which autonomy does not occupy a central place is desirable and for whom this does not seem to produce harmful health effects. This is particularly true for persons who belong to conservative religious groups. In fact, there are people for whom obedience and self-denial are central values; for such people, the promotion of autonomy by public institutions may very well conflict unduly with their lifestyle. If this is true, as I maintain, then the perspective that positions autonomy as the cornerstone 6 of health promotion must at the very least be clarified, if not substantially modified to accommodate such cases.

TWO MEANINGS OF THE CONCEPT OF AUTONOMY
How can the concept of autonomy be clarified to shed some much needed light on this problem? Part of the difficulty surrounding its elucidation stems from the fact that autonomy has two distinct meanings. 7 Firstly, there is procedural autonomy, which refers to an individual's effective capacity to make choices independently of the influence of others. The formulation of this definition is somewhat inadequate, given the large degree to which our choices are causally (directly or indirectly) influenced by our social environment, but we nevertheless consider individuals to be independent who have ostensibly exercised this capacity, even if their choices resemble those of their peers. Thus, we generally consider that procedural autonomy may be fully exercised in the absence of undue external constraint. If we consider the Charter's definition of health, it seems, on the face of it, to consider autonomy as a capacity. The second meaning refers to substantive autonomy, which is not the capacity of individuals to make choices, but rather the quality of the choices made. The substantive conception of autonomy is linked to the idea that there are values which should be upheld by individuals if they are to be considered healthy in the broad sense of the term. Thus, substantively autonomous individuals make the right choices with respect to their lives, the choices that correctly embody what it means to be autonomous; indeed, from this perspective, there are particular lifestyles or values that concretely express what it means to live autonomously. It isn't obvious whether or not the definition of autonomy adopted by the Charter would consider it as encompassing its substantive meaning; references to physical, mental and social wellbeing are not explained. But, in parallel, it's worth noting that public health workers would hardly consider people who are clearly not making healthy choices about their lives as being healthy. That may be because the broad definition of health is generally understood as involving the narrow one; if we consider our two definitions of autonomy, then we could say that procedural autonomy requires substantive autonomy. In other words, we couldn't be considered as having the capacity for autonomy if we did not display apparent autonomous choices 8 .
However, that is precisely the problem. It seems quite plausible that there are deeply religious persons who choose autonomously, in the procedural sense of the word, to live lives that do not assign a central role to autonomy. Indeed, the fact that one leads a life of obedience and self-denial does not mean that one has lost the capacity to make choices. One might object, however, that such persons cannot choose other than to lead a religious life because they are constrained simply by belonging to a particular group, that their culture does not allow them to choose otherwise. Let us turn now to this idea.

AUTONOMY AND CULTURAL BELONGING
In my view, the idea that some people cannot choose other than to lead a religious life because they are constrained by belonging to a particular group, rests on a false conception of the nature of cultural membership. It presupposes that belonging culturally to a particular group implies exclusivity (that one is a member of only one group) and inexorability (that one cannot be a member of another group). However, to begin with, the idea that a culture is composed of a group of consistent and stable characteristics that determine the identity of its members is clearly doubtful. In fact, as noted by Tariq Modood, 9 among others, hardly anyone can be found who still defends an essentialist conception of culture. Thus, cultural belonging does not imply inexorable conformity. Also, the idea that the cultural options available in a liberal community, for example, cannot be conceived of as accessible to members of a religious group is doubtful as well. In fact, it is clear, for example, that the members of the congregation of Grey Nuns of Montreal are aware of what life is like outside their convent. It cannot be said, therefore, that this choice is inaccessible to them, strictly speaking. Indeed, some of them leave their congregations, even if such choices are frowned upon by members of their community. It must also be noted that, if one considers the mental health of those in religious groups, the relationship between autonomy and health does not appear to be as direct as one might think. In fact, some studies have demonstrated that depression rates among the Amish in the United States are half as high as those recorded for the rest of Americans. 10 As noted by Schwartz, American citizens have never in their history had so much choice and, paradoxically, they do not seem to benefit psychologically, but rather the contrary. 11 This is another facet of the idea put forth by Jeff Spinner-Halev, who notes that conservative religious communities may offer fewer choices to their members, but that these choices may be perceived as being personally more significant than those available in a liberal community. 12 It should not be understood by this that we should all join closed religious communities to improve our mental health -after all, the data on the prevalence of depression among the Amish do not include, for example, those who had to leave the community -but rather that internalization of choices is a complex phenomenon that should be taken into account when considering health promotion broadly defined. In fact, these observations allow us to underscore a fundamental point regarding autonomy and cultural membership: a difference between a religious group and a liberal community lies in the level of support and solidarity they usually provide. Indeed, tightknit communities often offer more support than the kind we usually find in our modern liberal societies. This intuition is confirmed by studies on self-determination, which have demonstrated that it is not so much the source and the objective quality of a person's choices as the support received while these choices are being made that positively affects a person's wellbeing. 13 Since supportive close-knit subgroups (such as families) are not restricted to conservative religious communities, we should not conclude that meaningful support towards health-influencing choices is absent from liberal societies, but simply that there does not seem to be a clear cut path from substantive autonomy to health. So, in essence, this means not only that there are distinct healthy lifestyles that value autonomy quite dif-ferently, but that the link from autonomy to health is more complex than what is sometimes conveyed through health promotion literature.
At the very least, this observation leads us to observe that the quality of the choices people make (their substantive autonomy) is of secondary importance to their health relative to their capacity to make choices (their procedural autonomy) 14 . Consequently, any theory that seeks to clarify the relationship between autonomy and health should underscore the precedence of procedural over substantive autonomy. This precedence is of major significance to any valid conception of autonomy promotion, and I will now consider such a conception. 15

WHAT IS AUTONOMY PROMOTION?
What form should a conception of autonomy promotion take, particularly if it must be justifiable to those who do not consider autonomy to be important? The main issue at stake here is that of paternalism. Determination of the level of paternalism considered acceptable in a state's interventions in the lives of its citizens is a classic normative problem in the public health field. 16 What is generally implied by paternalism is the idea that the state is justified in forcing individuals to act in their own interests when it is apparent that people are going to behave in a manner that is harmful to themselves. Paternalism stems from the idea that it is sometimes necessary to intervene to protect individuals from themselves. Clearly, situations exist in which such interventions are justifiable -when, for example, individuals do not have the capacity to make informed choices (as in the case of children), when the restrictions placed on individuals are, broadly speaking, negligible (as in the case of mandatory seat belt laws), or when the consequences of one's harmful actions extend to other people (as in the case of cigarette smoking). However, in J.S. Mill's classical work on the subject 17 , paternalism, as briefly defined here, is rejected for the simple reason that its proponents are not able to morally justify the proposed constraints on the freedom of the individuals concerned, since these constraints are coercive and when the targeted individuals are capable of autonomous choice.
It is useful to distinguish four types of paternalism. Strong paternalism is the view that the state can legitimately prevent persons from acting in a manner that is contrary to their own interests; weak paternalism is the view that the state should seek the informed consent of persons to accept such practices, rather than attempting to oblige them against their will to act in to their own interest; narrow paternalism is limited to the use of coercive mechanisms; and broad paternalism includes other forms of intervention typical of public health, for example, empowerment, collaboration, information and so on. The position most compatible with the conception of autonomy and its promotion defended in this text is that of weak and broad paternalism. Weak paternalism is congruent with the conception of autonomy promotion defended here because, by seeking above all to ensure the consent of individuals, it assigns priority to the informed exercise of the capacity to make choices, and not to the quality of the choices thus made. Weak paternalism depends, above all, on the capacity to give one's consent, on the absence of unacceptable coercion (extortion, threats, physical coercion), and on assimilation of the information relevant to the exercise of a choice. 18 Weak paternalism is thus very useful in cases involving religious or cultural practices that large numbers of people who do not belong to the groups concerned may consider in some way or other reprehensible or contrary to one's well-being. Indeed, we tolerate a large number of practices when they respect the conditions established under weak paternalism (think, for example, of certain sexual practices such as sadomasochism and bondage, or partner swapping). If the conditions are not met for freely chosen harmful behaviour -in the absence of valid consent 19 -then weak paternalism reverts to strong paternalism. But, what is important to remember here is that weak paternalism is compatible with the idea that individuals may autonomously choose to limit their own autonomy.
Broad paternalism opens the door to the use of an interesting range of interventions. Nuritt Guttman 20 has identified strategies that are generally used in the field of public health: apart from coercion, we find environmental strategies, incentives, persuasion, collaboration, and facilitation. In the case being examined here, it must be emphasized that if weak and broad paternalism is the approach adopted, then coercion must be removed from the list of strategies; the other strategies are very much legitimate. 21 Thus, we can envision a variety of strategies involving incentives, collaboration or facilitation (these are in fact the strategies preferred for empowerment initiatives) aimed at promoting autonomy.

AUTONOMY, PATERNALISM AND ADAPTIVE PREFERENCES
One objection to broad and weak paternalism can be expressed through reference to the problem of adaptive preferences. The concept of adaptive preferences refers to the idea that we all have a tendency to choose what is relatively accessible to us and not to choose what is not so accessible. The relative inaccessibility of certain choices may itself "constrain" some people to accept lives that are contrary to their wellbeing. The classic example of the problem of adaptive preferences is that of the status of women living in countries where their autonomy is not valued, who themselves internalize this value system and apparently consent to lead a life that is not autonomous. 22 Part of the solution to this problem has already been evoked. Such constraint cannot be attributed to the very culture of individuals because the idea that we are acted upon by social regularities is meaningless. It is rather that these roles are often enforced through coercive measures.
Part of the answer to the problem of adaptive preferences can be found in the conceptions of autonomy and weak paternalism that have been presented here. Firstly, the idea that autonomy can be promoted in isolation does not make sense. The solution to the problem of adaptive preferences necessitates a conception of autonomy promotion that includes several accompanying measures aimed at ensuring its success. Martha Nussbaum, emphasizing this point, states that autonomy promotion must be reinforced by physical and psychological integrity, the absence of coercion (which is consistent with the absence of coercion demanded by weak paternalism), the possibility of maintaining significant social relationships (which is consistent with the idea that the quality of social relations is closely tied to the exercise of autonomy), the possibility of escaping from unfavourable environments, and access to quality information. Weak paternalism should ensure that if these conditions are met and the persons concerned persist in leading a life that might be considered contrary to their wellbeing, their choice must be respected.
This, in turn, underlines the fact that the conception of autonomy promotion defended here respects the endorsement constraint described by Ronald Dworkin, which stipulates that a life cannot be fully appreciated unless it is endorsed. As the latter states: "Threats of criminal punishment corrupt rather than enhance critical judgment, and even if the conversions they induce are sincere, these conversions cannot be counted as genuine in deciding whether the threats have improved someone's life" 23 The conception I have defended here offers a solution to the apparent paradox of autonomy promotion referred to earlier in this text. It in fact proposes a way of reconciling the responsibility of institutions to protect and promote autonomy and the basic freedoms of persons who do not consider autonomy to be important. This conception may have relevance not only for policies aimed specifically at empowering individuals or communities, but also for policies that may relate to the issue of reasonable accommodation. 24 How then can we shed light on public policies addressing diversity issues using the distinction between weak and strong paternalism? Table 1 classifies various policies that address the issue of cultural pluralism in one way or another. The policies are classified according to whether or not they are based on weak paternalism and according to the level of importance they assign to autonomy. Case 1 includes policies for which autonomy promotion is one of the main objectives and which employ coercive means, usually legislative measures that criminalize illiberal practices that are judged contrary to individual autonomy. These policies are motivated by the conviction that we must protect certain vulnerable sub-groups within given groups by forbidding certain practices; thus, the example used in the table is the eradication of excision, but any other example of a targeted policy coercively protecting the autonomy of a "minority within a minority" 25 would have served equally well, such as, for example, the ban on wearing the veil in public schools in France. Case 4 also includes policies for which autonomy promotion is central, but only those which do not use coercive methods. To return to the case of excision, instead of using a coercive approach, a number of initiatives in many countries have sought to reduce the abuses related to this practice and to change the conditions that lead to its being widespread in the first place; and in fact, these initiatives have been generally more effective that coercive eradication policies. 26 Cases 2 and 5 refer to policies for which autonomy promotion is a secondary objective. This category includes, among others, social or family policies whose implementation would render illegal any cultural practice judged contrary to autonomy promotion, as defined by the policy. For example, Canada does not recognize polygamous marriages but nevertheless Table 1. Examples of public policies characterized by their degree of paternalism and the level of importance assigned to autonomy promotion awards rights to women who have contracted such marriages, in case, for example, of separation. This situation contrasts with that in France, where not only are polygamous marriages illegal, but following adoption of the Pasqua law in 1993, women who had contracted such marriages and were living in France were expelled from the country. 27 This law would be a good candidate for case 2. A policy that decriminalizes polygamy and actively seeks to award rights to women in polygamous unions would be a good example of case 5.
Cases 3 and 6 refer to policies that may encroach, for potentially justifiable reasons, on the autonomy of certain cultural or religious sub-groups by prohibiting a practice that is a central to the group's identity. These two categories obviously have less to do with autonomy promotion than the others, and yet, given its importance, must still be taken into consideration. As an example, we have only to think of safety policies that impinge on the clothing customs of certain religious groups. Such a policy would be coercive (case 3) if it invariably gave clothing-related safety demands precedence over cultural practices with which they entered into conflict; for example, the mandatory wearing of a safety hat on job sites and orthodox Sikh employees; or the wearing of pre-determined clothing in an operating room and Muslim employees. On the other hand, if the policy was reasonably accommodating, i.e. it openly recognized the damaging effects of imposing clothing obligations on certain persons due to their inclusion in a given group, and recognized that the accommodation of such persons did not necessarily constitute an excessive burden for the organizations having to apply the policy, it would be considered non-coercive (case 6).
These few considerations are aimed at illustrating types of policies in which autonomy promotion in a multiethnic setting plays a role. If the conception of autonomy promotion presented here is a valid one, then any policy for which it is an important goal (or even a significant concern) should normally opt for non-coercive methods or, as I have termed this approach, for weak and broad paternalism, when the conditions for weak paternalism are met. The central role played by autonomy in the health of individuals has been explicitly recognized in the field of public health, but, if the conception presented here is valid, then 1) its promotion should be included within a broader framework of policies aimed at improving people's health and 2) its promotion cannot rely on coercive methods, without inval-idating itself, when the behaviour or custom in question is freely chosen. As a final note, it can be anticipated that policies inspired by the conception presented here would be more likely to achieve their goals than policies inspired by strong paternalism. 28 137 L E S A T E L I E R S D E L ' É T H I Q U E V . 5 N . 1 P R I N T E M P S / S P R I N G 2 0 1 0