The dominant account of welfare in medicine is an objective one; welfare consists of certain favoured health states, or in having needs satisfied, or in certain capabilities and functionings. By contrast, I present a subjective account of welfare, suggested initially by LW Sumner and called “authentic happiness”. The adoption of such an account of welfare within medicine offers several advantages over other subjective and objective accounts, and systematises several intuitions about patient-centredness and autonomy. Subjective accounts of welfare are unpopular because of their implications for justice and the autonomy of the healthcare professional. This account of welfare, however, seems to have the resources to resist these criticisms.
Statistics from Altmetric.com
The concept of welfare is a familiar and important one in medical ethics. However, many medical ethicists, while writing about welfare, are vague about what they mean by this word. Harris, for example, sees welfare as one of the two essential elements of respect for persons, but defines welfare simply as “the state or condition of doing or being well”,1 which is almost tautologous. Similarly, Beauchamp and Childress open their chapter Beneficence by asserting that “Morality requires that we contribute to (persons) welfare”—without defining welfare at all.2 This seems an important omission; if welfare occupies this central position in medical ethics, we (arguably) need to have a fairly clear notion of what welfare is.
On the few occasions when welfare is defined in a medical ethics context, it is usually understood in a rather objective sense, relating to health, to functioning or to the absence of disease. For example, Norman Daniels3 sees welfare in terms of satisfaction of needs and characterises needs as “objectively ascribable; we can ascribe them to a person even if he does not realise he has them and even if he denies he has them because his preferences run contrary to the ascribed needs.”i More recently, Anand (following a lead given by Amartya Sen) has argued that medical welfare is best seen in terms of objective functionings and capabilities.4
Similarly, when non-medical ethicists are trying to find examples of objective notions of welfare, they often turn to medicine to provide illustrations. For example, Nussbaum produces a list of 10 objective capabilities, of which five (normal life span, bodily health, bodily integrity, senses, imagination and thought) are at least partially related to medical goals.5
In this paper, by way of contrast, I want to defend one version of a subjective account of welfare. This is Sumner’s account of authentic happiness. This particular account of medical welfare offers several theoretical and practical advantages over more objective versions, and has major implications for how healthcare professionals should treat their patients.
As a starting point, welfare is the condition of doing or faring well. Synonyms include “wellbeing” and “best interests”. Antonyms include “the condition of doing or faring badly” and such clumsy terms as “illbeing” and “ill-fare”.
Subjective theories of wellbeing make our well being logically dependent on our attitudes of favour or disfavour. Objective theories deny this dependency. So on an objective account of wellbeing, an individual’s life can be going well even if he has no positive attitude at all towards it, even, indeed, if he has a negative attitude towards it. The two categories, objective and subjective, are mutually exclusive—if a theory of welfare happens to include both subjective and objective components (as is often the case), then this hybrid theory would be classed as subjective, as it contains a positive (or negative) attitude as an essential component of welfare.
In medicine, it is more natural to think of illbeing rather than wellbeing. An objective theory of illbeing would therefore include reference to certain objective states of ill health such as illness or impaired functionings or unmet needs. A subjective theory on the contrary would simply insist that illbeing contained a necessary reference to the subject’s having some kind of negative feeling about their life.
SUMNER’S THEORY OF AUTHENTIC HAPPINESS
There are two important objections to objective theories of welfare.6
First, objective theories of wellbeing fail to meet the criterion of formality. That is, they can provide us with a list of which items contribute to an individual’s welfare, but they fail to provide us with an explanation of what gains each item a place on this list. (This clearly matters a great deal in medicine; before committing time and resources to improving the welfare of an individual, we need to know that the area that we are targeting is really a part of that individual’s welfare.)
Sumner considers and rejects several formal theories of objective welfareii and finds them all wanting for the same reason—they all fail to explain why items on the list are present or absent.
Second, objective theories fail to recognise the intuitive subjectivity of welfare—in the words of Mark Lebar:
Your attitude towards your own life matters deeply because you are the one who experiences that life. It affects you as it affects no other. It is your life.12
Objective theories of welfare cannot (easily) incorporate this attractive subjective notion of welfare.iii
So what would an adequate subjective theory of welfare look like? Two much-discussed (though ultimately unsatisfactory) subjective theories of welfare are those of narrow hedonism and preference satisfaction.
Narrow-hedonism theories were proposed by Mill and Bentham as theories of welfare. A life is going well, said Bentham, if the individual has a positive balance of pleasure over pain, and it is going badly if the reverse is true. But this is too narrow—there is much more to life than pleasurable sensations. For example, what if an individual prefers to dedicate himself to art or to the welfare of others, or even to some personal project? In that case, his life might be said to be going well, even if these projects do not necessarily lead to pleasure and the absence of pain. (Freud famously preferred not to take morphine when he was dying of cancer, so that he could finish writing his book.)
One obvious move at this point might be to explain the previous intuition and concentrate on the satisfaction of preferences as the measure of welfare—the so-called preference-satisfaction theory of welfare. Under this account, a life is going well for an individual if she has her (main) preferences fulfilled. This is still a subjective theory, because it depends on a kind of positive attitude, but, like narrow hedonism, it is also flawed. Satisfying a preference does not necessarily lead to a better life. Many desires are profoundly disappointing when they are satisfied. (After a busy week in the surgery I form a preference to take a day off work to do some shopping; but I miss the interest of medicine and count the day a failure.) Some desires can be satisfied without our even being aware of it, so how can they possibly cause our life to go better? (I desire that a patient should be cured. The patient moves abroad and subsequently is cured, but I never hear about it.) Furthermore, much welfare comes from serendipity, where no desires are being satisfied at all. (I hear some Baroque music in a café and listening to it adds considerably to my welfare, but I never had a desire to listen to that sort of music.)
It seems therefore that satisfaction of desire is neither necessary nor sufficient for welfare, and desire-satisfaction is therefore an unsatisfactory theory of welfare.iv
It is against these difficulties with traditional concepts of welfare that Sumner suggests that welfare is “authentic happiness”. What does he mean by this? Sumner does not understand happiness in the classical hedonistic way, as a positive balance of pleasure over pain. Rather, he sees happiness as “life satisfaction” or “personal satisfaction”. Being happy, for Sumner, is “having a certain kind of positive attitude towards your life, which in its fullest form has both a cognitive and an affective component”.6
What does this actually amount to? The cognitive aspect of happiness consists in a positive evaluation of the conditions of your life, a judgment that at least on balance it measures up against your standards or expectation: “taking everything in to account, your life is going well for you” (p 145).6
The affective side of happiness consists in what we commonly call a sense of wellbeing: “finding your life enriching or rewarding, or feeling satisfied of fulfilled with it” (p145).6
Clearly there is much overlap between cognitive and affective happiness. But sometimes they can separate. Some organisms can only be happy in the affective sense, as they lack the competency to review and assess their life (but still are able to feel happy about their life). Many non-human animals, for example, can only be happy in the affective sense, as can small children and adults with severer forms of mental impairment. But the majority of adults are capable of being happy in both the affective and cognitive senses of the word.
Such a formulation is an improvement on the previous accounts of welfare. It retains the subjective nature of welfare, yet avoids the problems inherent in narrow hedonism or preference satisfaction. But such a formulation cannot be complete, for two reasons.
First, the subjective assessment of a life going well (or badly) must be based on full information—otherwise the assessment is no guide to how the life is going. Imagine a woman (let us call her Josie) who assesses her life as going well (is happy with it), mainly because she believes her husband is a dedicated, loving husband. In reality, however, he is a complete cad, and has affairs with many of his work colleagues. Is her life really going well? Many outsiders might think Josie’s life was going badly. Indeed, once Josie discovered the deception she might well agree. But she might not agree—she has the option of saying, “well he was a complete bounder, but it was fun while it lasted,” or, “although feckless, he was good for me in many ways”. But for Josie to be able to accurately assess her welfare she needs full information—otherwise her assessment is flawed and we cannot rely on it as an assessment of her welfare.
Second, the assessment of happiness must be autonomous. Any self-assessment of happiness will be suspect if it has been influenced by any process which undermines the individual’s autonomy. Such processes might include indoctrination, bullying, coercion and brainwashing. Imagine that Josie’s partner over many many years has verbally abused her and systematically diminished her self-esteem. One day he is slightly less obnoxious towards Josie. At this point, Josie might assess her life as going well, because she has been conditioned to accept very low standards. But it would be wrong to accept Josie’s assessment of her wellbeing as in any way reliablev—because of her conditioning, it is not really Josie’s assessment of her wellbeing. In general we assume that an assessment of happiness is made autonomously provided there is no evidence that the autonomy of the individual has been subverted.
For us to use self-assessment of happiness as a guide to welfare, it must therefore be informed and autonomous—or, to use Sumner’s term, it should be authentic. A life is going well if the individual whose life it is assesses it to be going well and this assessment is made authentically. This formulation is a subjective theory of welfare (because it depends on a subject’s positive/negative assessment of their life), but it is also a hybrid theory, as it contains the objective “authenticity” constraint. Unlike other objective theories of welfare, however, the objective components are explained and motivated by the underlying subjective theory—so it passes the formality test.
SUBJECTIVE AND OBJECTIVE WELFARE AND MEDICINE
Is Sumner’s account of authentic happiness a plausible theory of wellbeing in medicine? There is no reason to think that the general reasons advanced by Sumner against narrow hedonism and preference satisfaction do not equally apply to the practice of medicine. (Our patients’ welfare is more than the balance of pleasure and pain, and preference satisfaction is neither necessary nor sufficient for our patients’ welfare.)
However, objective accounts of welfare seem much more attractive in the medical context. Remember that Sumner criticised objective, list accounts of welfare for lacking formality: proponents of an objective list were unable to explain why the individual items appeared on the list. This seems less of a problem in medicine. Surely all that matters for a medical view of welfare is whether the individual is healthy.vi So when we assess a patient’s welfare, what matters is whether the patient is functioning well, or is physically fit, or pain-free, or disease-free. Aren’t these objective descriptions of welfare much more plausible than any subjective account?
The first response to this claim is to simply hold the line and assert that welfare (even medical welfare) just has to have a subjective element—or else it is not welfare. So a life assessed as going well by any objective health criterion might still be going badly because that individual (authentically) assesses it as going badly. Conversely, take any objective ill health criterion—chronic pain, blindness or immobility, for example—and it is still an open question whether that person’s life is going badly or not. Agreed, the presence of any of these conditions would usually make it harder for a life to go well, but in the final determination of wellbeing, the subject’s authentic assessment of how their life is going must be paramount. And this seems to fit in with our experience of our patients—within wide parameters, the wellbeing of our patients can be independent of their state of illness or health.
At this point, the objectivist about medical wellbeing can concede the centrality of a subjective component to wellbeing but insist that there must be independent objective components as well—in other words, they could endorse a hybrid account of medical welfare where subjective and objective elements are jointly necessary and sufficient. On this account, a life would be going well only if the subject of the life endorsed it as going well and certain objective features such as health and normal functioning were present.vii
But the subjectivist can reply by pointing out that subjectivists are able to offer the bestviii explanation of why objective notions such as “health” usually matter so much for welfare (and, conversely, why sometimes they don’t matter so much). Good health, flourishing, normal species functioning and the like derive their importance from the fact that objective good health (usually) contributes to a happy life, and objective bad health (usually) contributes to an unhappy one. But this is a contingent fact, and sometimes health and welfare can come apart. The example of the dying Freud quoted earlier is a good illustration—Freud’s life may well have been going well (if he had completed an important book, for example) even if, objectively, things looked pretty bad (he was in pain, and dying, so with regard to his health, things could not have been much worse.) Sportsmen and sportswomen undergo great pain, and risk later disability, to make their lives go well. A patient of mine once travelled 40 miles on public transport with a fractured neck of femur so that she could make adequate arrangements for her handicapped daughter—despite the enormous pain she experienced, she had no doubts that had she not done this, her wellbeing would have been much the poorer. (“I could not have lived with myself.”) And so on.
FROM DESCRIPTIVE TO NORMATIVE
Whichever ethical theory one subscribes to, it seems inevitable that there will be at least some duty towards the welfare of other agents. (The importance of this duty, however, varies between rival ethical theories. In addition, some ethical theories would not acknowledge a duty towards all welfare, perhaps excluding welfare which comes from sadistic pleasure, or where the welfare is in some way not deserved.)
As Robert Shaver puts it: “[T]he attraction of welfare is that its appeal is uncontested. … [B]ecause the value of welfare is uncontested, no case for it needs to be made”.16 This is particularly true in healthcareand is reflected in the over-arching principle in the General Medical Council’s Good medical practice: “Always make the well-being of your patients your first concern.”17
So, if one accepts the moral demand of welfare promotion and couples this with a subjective version of welfare, it seems that healthcare professionals have at least some responsibility to assess and be concerned about the patient’s authentic happiness.ix And, in general, they should be concerned with improving the subjective welfare of their patient. Normally, of course, this means improving the health of their patients, because, contingently, health and subjective wellbeing are closely linked. But in some circumstances (as at the end of life, or when treatment may frustrate the life goals of an individual), health and subjective wellbeing can come apart. In these situations, faced with a choice of enhancing either subjective wellbeing or health, the health professional should choose to enhance subjective wellbeing. More specifically, healthcare professionals do not have a duty to assess or improve objective welfare unless this contributes to assessing or improving authentic happiness, assessed cognitively, or, if this is not possible, affectively.
Is there any evidence that healthcare professionals do actually have normative beliefs relating to a subjective rather than an objective view of welfare? Most evidence in this area comes from a consideration of what doctors do, rather than what they believe to be right, but nevertheless there are some interesting arguments from a consideration of patient assessment, treatment and autonomy.
ASSESSMENT OF PATIENTS
Many of the strategies that we routinely use in medical interviews relate to subjective welfare (some of these strategies can be classified under the umbrella term of “patient-centredness”). So we ask for information about a patient’s ideas and concerns, partly because it will increase objective welfare (it may guide us to the right diagnosis), but also, and perhaps chiefly, because we are interested in a wider notion of wellbeing than that expressed in a narrow objective health model (health, functioning, and so on). When taking a history, we ask about occupational and social factors because (in part) we want to assess the impact of the problem on subjective welfare. We make diagnoses in physical, psychological and social terms, because understanding the wider social impact helps us to comprehend the alteration in subjective welfare. When making management plans, we do so in cooperation with the patient because we need to know if the treatment we propose will really increase the patient’s welfare (and the patient is probably the best guide to this). We do all of these things because we want to increase or promote subjective welfare, and they are (usually) not necessary if we are working with a simply objective account of welfare (such as “health”).
TREATMENT OF PATIENTS
As well as assessing patients in a subjectivist way, we also treat them in that way. Here are some examples. We prescribe oral contraceptives so that women can avoid having periods on holiday. This does not increase welfare in the narrow objective sense, but certainly increases subjective wellbeing. (Indeed, the impetus to offer contraception and sterilisation is much easier to understand in subjective than objective terms.) We might agree with and support a patient who wants to delay treatment (even if that means reduction of the survival rate) in order to take a special holiday—because this will increase his subjective welfare. Many palliative care decisions are based on an account of welfare that is subjective, and decisions not to have treatment at all (even when there is a small chance of success) will also fit into this concept of wellbeing.
Respect for autonomy is widely considered to be an important normative concept in modern bioethics. If we work with an entirely objective account of medical welfare, this can lead to a false dichotomy between welfare and autonomy, which in turn can cause uncertainty and vacillation in ethical decision making. If, on the other hand, we adopt a subjective account of medical welfare, much (not all) of the conflict between autonomy and welfare disappears. Autonomy has a close conceptual link with subjective welfare, and so the desire to respect autonomy can be seen as a (subjective) welfare-enhancing activity rather than a strategy that will undermine welfare. (Of course, not all autonomy can be subsumed into welfare, however subjective.)
PROBLEMS WITH A SUBJECTIVE ACCOUNT OF WELFARE
Despite the many attractions of this subjective account of welfare, there are several problems with this approach. In this final section I will sketch two possible serious problems and then outline possible responses to these problems
The first worry is that if we rely on a patient’s subjective assessment of their welfare, this would, by extension, commit us to unreasonable or bizarre methods of treatment. Imagine a patient who assesses that her life is going badly because she is tired. Furthermore, she believes that she needs a zinc supplement to correct this tiredness and hence make her life go better. Does this not impose a duty on the healthcare professional to provide zinc treatment for her?
Here we need to be clear that assessment of welfare is a starting point for therapeutic decisions, and not the final word. An individual’s authentic assessment of how their life is going is the best guide to their welfare. If their life is going badly (and that is usually why patients visit the doctor), their assessment of why it is going badly and what needs to be done about it is usually reliable. (Hence the emphasis given to individualistic autonomy discussed above) But both parts of this assessment can go seriously wrong, and it is no part of the doctor’s duty to blindly agree with the patient’s assessment of diagnosis and management (though it would be exceedingly foolish to ignore such valuable information). Perhaps (objectively) the patient is depressed. Perhaps zinc supplements would be useless. The healthcare professional’s duty is to accept the patient’s assessment of her welfare as a starting point (providing it is authentic) and then work with the patient towards the best available diagnosis and treatment of the problem.
Imagine two patients, A and B. Both assess their wellbeing authentically—that is, both are acting with full information and autonomy. A perceives any slight physiological abnormality (tiredness, cough, loose stool) as something that has a big impact on her welfare and attends the surgery weekly for medical attention in the form of treatment or investigations. For A, small deviations from what she perceives as perfect health cause her to assess her life as going badly. B, on the other hand, copes with severe symptoms and rarely attends the doctor’s. She remains cheerful throughout and assesses her life as going well. The subjective welfare of A therefore seems lower than that of B, and, all things being equal, A should receive a greater proportion of medical resources than B. Indeed, if resources are limited, B might get no treatment at all, while A gets all her problems treated. This seems to get things completely wrong.
There are four responses that might avoid this unattractive conclusion.
First, one could say that A’s assessment of her welfare is not really authentic. Perhaps the fact that A keeps coming week after week with “trivial” problems implies that she is only autonomous in a limited way. Or maybe B has some deep-seated fear of doctors that compels her to systematically downplay her welfare. So both their assessments of welfare are faulty and should not be the basis for any obligation. Rather, the good professional will try to correct this inauthentic thinking by exploration and challenge, rather than committing resources on the basis of the patient’s assessment of welfare. But sometimes A and B are assessing their welfare authentically; in which case exploration and challenge is of no help at all.
A more promising response is to accept that the welfare of A is worse than the welfare of B, but to assess the cause of A’s impaired welfare differently from the cause of B’s impaired welfare. A might (despite presenting with physical problems) actually have psychological ill health and need psychological help; B (let us say) has a physical problem and needs some kind of physical help.
A third strategy might be to dissociate welfare considerations from duty considerations. So although A’s welfare level is low, this may not be due to psychological illness but to some aspect of A’s personality (or to some untreatable conditions) which is not amenable (at all) to medical treatment. This therefore does not (cannot) generate a duty to use resources. Similarly, B’s condition might attract lots of resources because it is eminently treatable.
One more response is possible. Both of the preceding two strategies succeed by characterising individuals with extremely low assessments of their welfare as in some way psychologically impaired. Obviously, this will not always be the case; psychologically “normal” individuals will also assess their own welfare as quite different despite similar “objective” life conditions. However, it may still be better overall if resource allocation is based on a list of objective health problems rather than on a case-by-case assessment of individual subjective welfare. In any complex healthcare system, the best option for resource allocation may be to have a simple system based on what usually is the source of individual welfare (which might turn out to be a list of needs or a list of functionings and capabilities). To individually assess the subjective welfare of all patients might be so time- and resource-consuming, or so prone to bias, that it would be counterproductive. And thus, by default, an objective method of resource allocation might have to be used, even though this does not accurately represent welfare.x
Sumner’s account of subjective welfare represents a radical challenge to the dominance of objective accounts of patient welfare. It has several theoretical advantages over other accounts of welfare, and, overall, it adequately systematises widespread intuitions about the importance of autonomy, patient centredness and shared decision making. Arguably, all patient-professional consultations should include a question asking,
Arguably all medical professionals should regularly ask their patients—“How is life going for you?”
I wish to thank my supervisor on the Keele Doctorate in Medical Ethics course, Eve Garrard, for valuable and challenging suggestions in the writing of this paper.
↵i Daniels goes on to suggest that needs are important because they “maintain normal species functioning”.
↵ii Examples are Braybrooke’s theory of needs7, Thomson’s theory of needs and interests8, Sen’s theory of functioning and capabilities,9and teleological theories5 10 To summarise: Braybrooke argued that an individual’s life was going well if his basic needs were being satisfied. These needs include such items as nourishment, exercise, rest, companionship, social acceptance, personal security and so on. Braybrooke tries to justify these items in two ways. First, he sees basic needs as essential to functioning normally. But this begs the question of what it is to function normally. Some individuals would see their lives as going well without reference to many of the items on Braybrooke’s list. The other way in which Braybrooke tries to justify his list of needs is by arguing that they are “indispensable to mind or body in performing the tasks assigned a given person under a combination of basic social roles” (p 48). But this again is circular: why should these social roles matter more than others in determining needs? Garrett Thomson8 tries to make the concept of a need depend on avoidance of harm (p 50). But Thomson then defines a harm as something that frustrates our interests. Interests, however, have a very subjective feel to them and thus it seems that Thomson’s account of needs is fundamentally a subjective one. Sen11 suggests that an individual’s life is going well if it has a certain combination of functionings (achievements) and capabilities (opportunities). But this, too, begs the question. The functionings and capabilities of (say) a professional philosopher and a Masai warrior will be radically different, and it may be impossible to generate a list of functionings that would meaningfully describe the welfare of both individuals. This is even true within one society. Consider the differing functionings of (say) a bricklayer and a brain surgeon. Finally, Taylor10 and Nussbaum5 argue that human beings have certain goals or excellences and that these objectively define what it is for a life to go well. But, whether these excellences are intellectual, artistic or moral, it is still an open questions whether achieving these goals constitutes a life rich in welfare.
↵iii Brink13 believes that accounts of welfare should not be subjective, because a subjective account might imply that repugnant or trivial lifestyles were lives rich in welfare. But even though such a life might score poorly for moral or perfectionist value, it is still arguably good for the liver of the life and, therefore, high in welfare.
↵iv Bykvist14 argues (p 487) that Sumner’s concerns about preference-satisfaction accounts of welfare can be answered. The main thrust of his complex argument is to deny that preferences have to be future oriented, and he argues that many preferences are “for now”. The risk of pushing this viewpoint is to make preferences indistinguishable from pro-attitudes, so that a preference-satisfaction account of welfare begins to look like a hedonistic account of welfare.
↵v Sen first developed this point, though as an objection to preference-satisfaction accounts of welfare.
↵vi See also in this context Dan Brock,15 quoting Leon Kass: “[M]edicine’s proper end is the much narrower one of health, … and other goals such as happiness and gratifying patient desires are false goals for medicine” (p 393).
↵vii Sumner’s account of welfare is also a hybrid account, but in Sumner’s account the objective element (authenticity) qualifies the subjective element and is not independent of it.
↵viii Objective accounts of medical welfare offer some explanation of what makes an item a constituent of welfare (or not)—for example, whether it leads to satisfaction of needs (Braybrooke7) or contributes to functionings and capabilities (Sen9). But it is an open question whether a certain item actually counts as a need or as a functioning. So these types of theory are actually indeterminate about welfare or (worse still for the objectivist) collapse into some form of subjectivism.
↵ix It should be noted that our duties as doctors to promote subjective welfare are limited to the practice of medicine. We do not have a duty qua doctor to lend our patients money, repair their faulty plumbing or teach them Latin—though all of these acts would (arguably) improve their subjective welfare. (We may have a general duty to do some of these things as (say) citizens or members of the same community.)
↵x There is a clear analogy here with the move from direct utilitarianism to indirect utilitariainism.
Competing interests: None declared
The author is a student on the Doctorate in Medical Ethics (DMedEth) course at Keele University, Keele, Staffordshire.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.