This entry is based on my “Paternalism and the Ethics of Suicide Prevention,” forthcoming in M. Cholbi and P. Stellino, eds., Oxford Handbook of the Philosophy of Suicide.

On its face, suicide prevention is the kind of phenomenon that lends support to paternalism. Suicide often seems prudentially ill-advised, a choice wherein a person ends up depriving themselves of a longer, better life. Furthermore, it is a choice where we seem to have reason to suspect that the person’s judgment or volition is somehow impaired, because (say) they are acting impulsively, are overwhelmed by their present suffering, or suffer from affective disorders such as depression, which prevent them from appreciating reasons they have for continuing to live. If ever we are justified in engaging in paternalism—in subjecting a non-consenting person to measures that aim to improve their well-being or prevent them from suffering self-inflicted harm—preventing suicide looks to be just such a case.

This pro-paternalist stance is common within medical and psychiatric circles. I encapsulate it in what I call the Paternalistic Presumption:

Paternalistic Presumption: For any particular individual S who entertains suicidal thoughts, makes plans for suicide, or engages in suicidal conduct, others have sufficient paternalistic reasons (that S’s dying by suicide would be harmful to them) for preventing S from ending their life due to suicide by using measure M such that the burden of proof falls on the side of refraining from using M.

In a forthcoming article, I cast doubt on this Presumption. Although I have general moral reservations about paternalism, my case against the Presumption does not rest on those reservations. Indeed, I concede arguendo that suicide prevention measures can in principle be justified on paternalistic grounds, i.e., that subjecting a person to suicide prevention measures to which they do not rationally consent cannot be ruled out a priori. My case against the Presumption is instead an empirical one: for suicide prevention to be paternalistically justified, it must meet several criteria, and there is compelling evidence that these criteria are unlikely to be met in many instances of paternalistically motivated suicide prevention.

The first criterion is that we need to be confident that the individual whose suicide we believe we will prevent is, in fact, likely to engage in suicide. But it is easy to underestimate the difficulty of that task. Research has identified many risk factors for suicide: a sense of isolation and lack of social support; conflicts in interpersonal relationships; job or financial losses; chronic pain; a family or personal history of suicidal conduct; alcohol or other substance abuse; and mental disorders, especially depression, bipolar disorder, or borderline personality. The trouble is that not only are these risk factors commonplace, but they also significantly underdetermine actual suicidal behavior. For instance, although those who engage in suicide often have mental illness, the proportion of those with the associated mental illnesses who engage in suicide is minimal. Thinking about suicide, wanting to die, etc., also occur far more often than do suicidal planning, action, or death. In general, then, we cannot identify those individuals within a large risk pool who are likely to be suicidal. As one expert has concluded, “any given warning sign for suicide will be wrong far more often than it is right,” with the result that suicide prevention efforts are rife with ‘false positives,’ i.e., instances where suicide is predicted but does not materialize. One prominent suicide researcher has concluded that even experts are only able to predict future suicidal thoughts with the “same degree of accuracy as someone with no knowledge of the patient who predicted based on a coin flip.” The Paternalistic Presumption assumes that we can make reliable inferences about suicidal behavior from evidence about risk factors, the presence of thoughts of death, etc., when in fact such inferences are largely unreliable.

A second criterion is that our suicide prevention measures should be effective, and in particular, that paternalistic measures, which we implement without the rational consent of their targets, are more effective than measures that proceed with their targets’ consent. Here, too, skepticism is in order. Much of the conventional wisdom regarding how to intervene to prevent suicide is questionable. For example, research by Fox et al. on prevention efforts directed at “self-injurious thoughts and behaviors” (SITBs), including suicidal thought and action, found that the efficacy of virtually all clinical prevention measures targeted at individuals is very modest and that little difference exists among commonly used prevention measures with respect to their efficacy. Their meta-analysis of over 1,100 randomized controlled trials of SITBs’ efficacy concluded that the most used interventions for suicidal individuals (pharmacotherapy, popular varieties of “talk therapy”) do not significantly reduce the likelihood of either suicide attempts or deaths due to suicide. In addition, suicide preventers will often face choices between consensual and non-consensual measures: voluntary versus compelled hospitalization, voluntary versus compelled pharmacotherapy, voluntary versus compelled group therapy, etc. At the same time, evidence suggests that measures that incorporate individuals’ consent (thereby rendering them non-paternalistic) are more likely to be effective. On grounds of effectiveness, consensual measures appear to outperform non-consensual measures. This is unsurprising, given that a person who cooperates in receiving others’ help is more likely to benefit from it. But here too the Presumption is weakened: Many prevention measures are not especially effective, and treating individuals non-paternalistically (that is, providing them only prevention measures to which they consent) increases the likelihood of success.

A third criterion is that the suicide prevention measures must increase a person’s overall lifetime well-being. Some measures that may well prove effective in preventing suicide arguably make individuals worse off across their lifetimes. Involuntary hospitalization, for example, can result in longer-term harms, including decreased employability and depth of social connection, alongside increased trauma, a sense of stigma, and feelings of dehumanization, themselves factors positively associated with suicidal thought and behavior. And less dramatic prevention measures also carry long-term risks. All psychoactive drugs (especially antidepressants) have side effects, and intensive therapy can be time-consuming. That a paternalistically motivated prevention measure extends a person’s life is not sufficient to justify it. The preventive act must actually enable the suicidal individual to have a better life overall as a result, and some measures fail on that score.

I submit that many paternalistically motivated suicide prevention measures fail at least one of these three necessary conditions. Hence, we should oppose the Paternalistic Presumption, not because it rests on false moral principles but because actually achieving paternalistic aims via suicide prevention is significantly harder than we recognize. Clinicians (and others) seeking to prevent the harms of suicide would thus do well to second-guess their paternalistic instincts, preferring consensual to non-consensual measures, as well as showing greater caution in their judgments about the efficacy and beneficence of most suicide prevention efforts.

The post From Paternalism to Suicide Prevention: No Simple Path first appeared on Blog of the APA.