Abstract
People in our liberal pluralistic society have conflicting intuitions about the legitimacy of coercive hard paternalism, though respect for agency provides a common source of objection to it. The hard paternalist must give adequate reasons for her coercion which are acceptable to a free and equal agent. Coercion that fails to meet with an agent’s reasonable evaluative commitments is at least problematic and risks being authoritarian. Even if the coercer claims no normative authority over the coercee, the former still uses coercion to replace the latter’s reasons or will with his own reasons or will. But does every hard paternalistic view have to invite such objection? Throughout I will assume that defenders of what I will call “Neutral Paternalism” (NP) and “Commonsense Paternalism” (CP) aim to offer reasons for coercion all can reasonably endorse despite evaluative diversity, in opposition to more objectionable forms of coercive paternalism, such as those which defend it on religious or perfectionist grounds. I will argue, nonetheless, that Gerald Dworkin’s defense of NP and Danny Scoccia’s defense of CP succumb to the same problems of objectionable imposition that saddle other forms of coercive paternalism. The shortcomings in their views suggest that even modest hard paternalism is nonetheless problematic for liberals.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
I follow Gerald Gaus (2010) in holding that “acceptable” means something like “consistent with a person’s normative commitments.” This is not the only way to understand the concept. For instance, Thomas Scanlon holds “acceptable” to mean “permitted by principles that no one could reasonably reject,” which in turn is a function of the objective relative weight of the objective reasons for people to reject or insist on a principle, where this is not determined by anyone’s actual normative commitments. I lack room to argue for the superiority of the Gausian over the Scanlonian view and so must take it as assumed for present purposes. Thanks to an anonymous referee for helping me to clarify this distinction.
The reader will notice repeated references to “voluntarily performed” actions. It is not necessary to determine a standard of voluntariness here, so the reader may use her preferred notion. I have in mind those irrational actions which nonetheless reflect voluntary failures of reason or will.
Scoccia explicitly says he is ignoring liberal neutralist arguments in his discussion. He is entitled to do this for the sake of discussion, but I will suggest in a later section why he must ultimately confront such arguments, especially given his own appeal to liberal rights in qualifying CP.
Robert Taylor (2004) argues that paternalistic coercion is not a universalizable maxim and so fails the first formulation of Kant’s categorical imperative.
It’s not clear on what grounds Scoccia can elicit our agreement that Frank is better off alive than dead—or whether there is some unambiguous standard of “better off”—but I will assume Frank is mistaken for the purposes of the argument.
To further complicate matters, agents might have MCEM views over some domains but not others. Frank might welcome paternalism in other contexts but not with regard to his ultimate choice of life or death. I mention this only to highlight that the severity/imminence/irreversibility of harm need not lend greater prima facie weight to a case for interference with actions leading to that harm. Frank might welcome a scheme that taxes his junk food because he isn’t that concerned with dietary autonomy and wants incentives to avoid the relatively lesser harms of obesity. On the other hand, Frank might claim the freedom to seek PAS because the decision of when to terminate his life is of utmost importance to him.
Perhaps we need a cooling-down period to assess his reasons. Discussion of the propriety of cooling down periods is beyond the scope of this essay.
Perhaps modular rationality requires Frank at t2 to endorse paternalism retroactively. Consider the Cold War nuclear deterrence strategy of Mutually Assured Destruction (MAD). MAD was premised on a bluff which if called at t1 gave the actor at t2 no reason to follow up on his bluff, since the effectiveness of the bluff was bound up in the reason for giving it in the first place. Why bomb one’s opponent in blind, murderous vengeance if one has already been bombed out—what’s the point, since the state of affairs the bluff was meant to prevent came about anyway? But MCEM does not involve a failure of modular rationality since the freedom from paternalism it demands at t1 is part and parcel of the agent’s ability to choose well at t2 and so prevent catastrophe at t2 from eventuating. One cannot control another agent’s actions in MAD, but one can control one’s own actions in the case of paternalism.
Saying that paternalism “would” benefit the subject on some occasion holds only if he were to choose the harmful action. It doesn’t benefit him in terms of well-being if he were to choose the prudent action anyway, and it might actually detract from his utility if he could realize more by choosing prudently in the absence of a hard paternalistic scheme.
References
Benn S (1988) A theory of freedom. Cambridge, Cambridge
Dworkin G (1972) Paternalism. Monist 1:64–84
Gaus G (2010) The order of public reason. Cambridge, Cambridge
Scoccia D (2008) In defense of hard paternalism. Law Philos 27:351–381
Taylor R (2004) A Kantian defense of self-ownership. J Polit Philos 12:65–78
Acknowledgements
I wish to thank Ben Bryan, Ryan Davis, Gerald Gaus, Eric Mack, Danny Scoccia, Chad Van Schoelandt, and two anonymous referees for helpful discussion or feedback on earlier drafts of this paper. Any errors are mine.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Glod, W. Against Two Modest Conceptions of Hard Paternalism. Ethic Theory Moral Prac 16, 409–422 (2013). https://doi.org/10.1007/s10677-012-9339-6
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10677-012-9339-6