Philosophical Notes 90, Libertarian Paternalism’s Red Herring (2014), by Lamont Rodgers

Share this




‘Nudging’: Libertarian Paternalism’s Red Herring
Lamont Rodgers

Philosophical Notes No. 90

ISBN: 9781856376662
ISSN 0267-7091 (print)
ISSN 2042-2768 (online)

An occasional publication of the Libertarian Alliance,
Suite 35, 2 Lansdowne Row, Mayfair, London W1J 6HL

© 2014: Libertarian Alliance; Lamont Rodgers

Lamont Rodgers is professor of philosophy at San Jacinto College North Campus, in Houston, TX.  He has published in Reason Papers, Libertarian Papers, and Southwest Philosophical Studies.  He specializes in theories of justice and political legitimacy.


The error of seeking a foundation or justification

Cass Sunstein and Richard Thaler endorse a version of paternalism that is alleged to be compatible with libertarian principles.1  Their version of paternalism aims to nudge people toward good choices without violating those people’s rights.  Accordingly, they call their version of paternalism ‘libertarian paternalism’.  This form of paternalism has been widely criticized on conceptual grounds.2  Libertarian paternalism might aim to square two incompatible notions.  This paper shows that the conceptual compatibility between libertarian principles and the nudges Sunstein and Thaler endorse is irrelevant to whether libertarians may oppose the political implementation of libertarian paternalism.  The problem is that Sunstein and Thaler focus on showing the conceptual compatibility between their goals and individual rights.  Goals must be pursued via through action.  Sunstein and Thaler ignore the fact that a principled opposition to the initiation of threats and violence, which this paper takes to be characteristic of libertarianism, will preclude the funding and enforcement of the apparatus necessary for the government to engage in libertarian paternalism.3  So even if Sunstein and Thaler’s goals are themselves compatible with libertarian principles, the real debate over the political implementation of libertarian paternalism violates libertarian principles.

Libertarian Paternalism

For this sake of this discussion, understand an action to be paternalistic if two conditions hold.  First, one agent obstructs the activities of another agent.  Second, the goal of that interference is to benefit the obstructed individual.4  Libertarians are at least a little uneasy with paternalism.  They are uneasy with it because, among other reasons, paternalistic interference might involve preventing others from acting on their informed desires.5  Call paternalism that prevents agents from acting on their informed desires “hard paternalism.”  Prohibition of drugs and alcohol, seat belt laws and so on, all have hard paternalistic justifications, at least in part.  Part of the rationale for these laws is that they are good for the individuals on whom they are imposed; and it does not matter if those individuals know very well the dangers of engaging in the activities the laws proscribe. 

Paternalism is not always objectionable to libertarians though.  Sometimes it is welcome.  Paternalistic interference does not always thwart the informed desires of others.6  In fact, the interference may be welcome, as judged by the obstructed individual’s lights.  If agent A knows agent B is about unknowingly to drink poison, and A prevents B from drinking it, the paternalistic interference is welcome.  Call paternalism that does not thwart the agent’s informed desires “soft paternalism.”  For the purposes of this paper, I will follow Steve Wall in regarding this as a necessary condition of soft paternalism.  To illustrate the appeal of distinguishing between hard and soft paternalism, Wall presents the following two scenarios:7

  • Dangerous Drug: An adult man, while of sound mind, desires to consume a recreational drug known to cause serious health problems.  The man is aware of these dangers, but desires to consume the drug anyway.  Another person forcibly prevents him from doing so for his own good.
  • Rotten Bridge: An adult woman is about to cross a bridge over a ravine.  The bridge contains rotten wooden planks.  An onlooker sees her, but does not have time to warn her of the dangers of crossing the bridge.  He intercepts her.

Dangerous Drugcase illustrates hard paternalism and the RottenBridge illustrates soft paternalism.  The man taking the drug is fully aware of the dangers of taking the drug; while the woman presumably does not know the dangers of crossing the bridge.  Among liberal theorists, the common reaction to these cases seems to be that the interference in Dangerous Drug objectionable, while the interference in the Rotten Bridge is not.8

Libertarian paternalism is to be even more palatable to libertarians than the interference in Rotten Bridge.  This form of “paternalism is a relatively weak, soft, and nonintrusive type of paternalism because choices are not blocked, fenced off, or significantly burdened.”9  It does not seem to involve seizing or obstructing anyone, as the Rotten Bridge case does.  Instead, it requires simply employing various tactics to get individuals to make decisions that are good according to their own lights.  Sunstein and Thaler call these tactics ‘nudges’.10  The nudges merely guide create default policies that aim at getting people to make good decisions.  These decisions are to be good not merely from the vantage of the all-knowing regulator, but from the perspective of the nudged individual.  For example, the government might make individuals automatically enrolled in 401(k) programs11 (a defined-contribution pension account).  Many individuals do not enroll in these programs, even though they programs are, for the sake of argument, in accordance with their informed desires.  If people do not wish to be enrolled in these programs, they simply opt out.  There is no violence or coercion involved. 

There are other means by which the government might nudge individuals.  Part of the national health care law could require seniors who do not sign up for Medicaid to have a program assigned to them.  They may opt out of the policy, of course, but they will be automatically enrolled.  Having to opt out of a program, assuming the process is not too arduous, is not obviously a rights violation.  Similarly, it could be the law that individuals who receive driver’s licenses must stipulate whether they are going to donate organs.12  This may well increase the stock of available organs and thus benefit the driver.  It is this focus on benefit to the driver that is to make the policy both paternalistic and palatable to libertarians.  If the rationale for having individuals decide whether to be organ donors were the benefit to others, the policy would not be paternalistic, and it would not be libertarian.  Libertarians tend to oppose forcing individuals to aid others.13

A great deal of the rationale Sunstein and Thaler employ aims to show that individuals may not be good judges of what is in their own interest.  Their argument appears to aim to show both that individuals are not good at making choices that will advance their own welfare, and they are not even good at figuring out what their own preferences are.  It is permissible to nudge individuals to make choices that will improve their welfare; and those nudges do not require coercion.  So the libertarian can neither object to the policies on the grounds that they undermine individuals’ welfare, nor on the grounds that they are coercive.  This is the essence of Sunstein and Thaler’s argument.

Several authors have observed errors in the empirical science Sunstein and Thaler adduce.14  The idea seems to be that Sunstein and Thaler cannot adduce a policy that will be in the interest of all individuals.  Thus, some may be swayed to pursue policies that do not enhance their welfare.  Because is beyond the purposes of this paper, I propose to ignore this concern in the following section.  What I will argue instead is that libertarian opposition to coercion poses a serious obstacle to any program of nudge paternalism. 

The Real Questions

Debates between libertarians and proponents of more extensive states tend to focus on what rights protect.  There is a long debate between libertarians and their opponents over whether taxation seizes something – the fruit of their labor, as it is often called – to which people have rights.  However, this habitual focus on what rights protect suggests that the whole debate between libertarians and their opponents is over which extra-personal objects rights protect.  Looking at the debate this way loses sight of the fact that political programs can be objectionable for more than one reason.  Political programs may be objectionable because the programs themselves violate rights to extra-personal objects.  But political programs may also be objectionable because they force people to do things they have rights not to be forced to do.

To set the stage for the present argument, think of Robert Nozick’s claim that taxation is on par with forced labor.  It turns out that Nozick also does not believe taxation appropriates labor time.  After Nozick claims “taxation of earnings of labor is on par with forced labor,” he writes the following: “Some persons find this claim obviously true: taking the earnings of n hours labor is like taking n hours from the person; it is like forcing the person to work n hours for another’s purpose.”15  Then Nozick explicitly rejects this explanation of his remark.16  Nozick explains how taxation is akin to forced labor in the following passage: “The fact that others intentionally intervene in violate of side constraints against aggression to threaten force to limit the alternatives in this case to paying taxes or (presumably the worse alternative) bare subsistence, makes the taxation system one of forced labor.”17

Nozick’s claim is that the threat of aggression to force people to pay taxes is what makes taxation relevantly like a system of forced labor.  With this in mind, suppose there is some activity that does not violate individual rights and benefits people.  Suppose we all benefit from watching a few minutes of comedy each night.  If someone asks whether a government program providing comedic relief would violate rights, it would surely be shortsighted to observe that comedy does not violate rights.18  It would also miss the point to observe that nobody will be required to watch the program.  The whole debate would be over whether it is permissible to force people to fund the program.  Nobody would miss that point. 

Assume there is nothing wrong with a government program to encourage people to make choices that actually satisfy their own preferences.  We cannot move from that assumption to the conclusion that any means of enforcing the program is permissible.  This becomes clear when we notice that the government can only nudge people toward good choices by threatening to employ force against at least some people.  If we look a little closer at the means by which the government delivers its nudges, those nudges start to look less innocuous.

Think first of employers.19  In order to deliver nudges, governments will need to engage in activities libertarians tend to oppose.  Suppose an employer wishes not to tell the government about the nature of her agreement with new employees.  Suppose the employer wishes not to ask for a social security number, let alone implement a retirement program.  The only way the government can nudge the employee is by threatening to employ force against the employer.  Should those employers engage in activities that make it harder to deliver the desired nudges, coercion is required to prevent them from doing so.  So employers will be forced to inform the government about whom they have hired and so on. 

If the rationale for forcing the employer to aid in the construction of a retirement plan for employees is the employees’ benefit, we have a clear case of forcing one person to aid another.  Thus, the question at this point shifts from whether the libertarian paternalism is conceptually coherent to whether it is legitimate to force people, under the threat of fines and imprisonment, to pay for projects designed to benefit others.  That debate is not a debate about paternalism at all.  Instead, it is an old debate about whether it is legitimate to force people to participate in a program aimed at benefitting others.  Libertarians have long denied that individuals can be forced to aid others, so even if Sunstein and Thaler are right to suggest that nudge paternalism is not conceptually objectionable to libertarians, their observation settles nothing. 

The implementation problem does not focus only on employers.  Notice also that government employees tend not to work for free.  The government will need to pay officials who will be required to construct the nudges, make sure they function properly, and enforce the fines and penalties against those who fail to cooperate.  Here, it seems that employees will be required to fund the implementation and management of the program.  The taxes the employees pay will support the nudges. 

Once this is understood, the debate shifts from whether the nudges are compatible with libertarian principles to whether it is permissible to force people, under the threat of fines and imprisonment, to pay for projects designed for their own benefit.  This is, at best, a form of soft paternalism.  It is a form of paternalism like that found in the publicly funded comedy program described above.  So the debate is now over whether this form of paternalism is permissible.  And the issue is not resolved by pointing out that libertarian paternalism is coherent. 

One might dig in and argue that if the desired end is permissible and compatible with libertarian rights, there are no good reasons to believe we cannot force people to fund the means necessary to achieve the end.  Whatever the merits of this response, three points need to be kept in mind.  First, the response needs to be considered in light of the goals of libertarian paternalism.  The whole point of libertarian paternalism is that it is to be incontrovertible that the nudges do not violate rights.  But the controversy returns when we think about fining people, and possibly putting them in prison, if they refuse to comply with the program.  Second, the rationale for the nudges seems to undermine the claim that we may force people to fund the nudges.  One reason the nudges are not to violate rights is that people can opt out of the programs in question.  So the claim that we can force people to fund the programs is baffling, if we keep in mind that the individuals then have the right to opt out of the project they were forced to fund.  Finally, as noted above, this debate is not solved by pointing out that libertarian paternalism is coherent. 

The discussion this far has focused on nudges involving retirement plans and the like.  Other sorts of nudges may be deliverable without violating principles shared by all libertarians.  Assume the default for a driver’s license is that all newly licensed drivers will be donors unless they choose not to be.  Some libertarians may not be opposed to the government’s involvement in roads and highways.  Those libertarians may not be averse to such a policy.  Those libertarians hold that the government may rightly claim a monopoly on licensing procedures; and that monopoly may be protected by the use of force.  However, it is clear many libertarians do not see how the government could have an enforceable right to monopolistic control of roads and highways.20  Many libertarians see roads as another service that private individuals should be allowed to provide.

Because at least some libertarians oppose government control of roads – and the licenses people need to drive on those roads – libertarians would be averse to a blanket requirement that newly licensed drivers meet any particular standards.  If the owner of a road wishes not to require a license or to have organ donation as the default, these libertarians would say they may do so.  The government’s attempt to establish standards would need to be enforced via coercion in much the same fashion as the retirement and health programs.  Owners of roads would need to be monitored and threatened.  Patrons of the roads might need to pay taxes to support the enforcement of that monitoring and threatening.  This lesson may be broadened to show that libertarians, with their interest in privatization, could oppose nearly any form of nudge paternalism.  Again, the question of whether libertarians are right to have this opposition is distinct from whether libertarian paternalism is coherent. 

A final concern needs to be addressed before closing this discussion.  One might hold that governments do not need to force anyone to do anything.  Instead, they can enforce libertarian paternalism via tax breaks.  Companies – even those that might run private roads – in compliance with the nudges can simply pay less than those that are not.  Nobody will be fined or imprisoned for failure to comply; and this method of ‘enforcement’ is simply a means of incentivizing participation in the nudges.  It might just be seen as another nudge. 

Assessing this suggestion requires understanding exactly what it aims to show.  Does it aim to show that this means of implementing nudges into our current system would not violate libertarian principles?  If so, it seems to fail.  After all, it is hard to find libertarians who think our taxes are as low as they should be.  A tax break is simply a move to the level required by justice.  The break should not come with strings attached.

To illustrate, observe that the rationale for the tax break could be applied to any sort of behavior libertarians oppose.  Imagine taxes are raised to 100%.  Then assume the government allows people tax breaks, provided people do things like pass a yearly drug test, get married, register for selective service and so on.  Each activity might gain the individual a tax break of n%.  Nobody would think that this system is compatible with libertarian principles.  Nobody would think this precisely because libertarians are opposed to the taxes in the first place.

A more interesting interpretation of the proposal is that it aims to show that in an ideal libertarian world, a dominant protection agency could implement libertarian paternalism without violating libertarian principles.  This proposal is rich and requires a much more extensive treatment than I am prepared to offer here.21  But it is important to see that this construal of the proposal has nothing to do with the current political climate and whether libertarians can oppose nudges as things are.

This paper has argued that libertarian paternalism’s goals are a red herring.  The conceptual compatibility between libertarianism and the goals, as defined by Sunstein and Thaler, resolves no political questions.  Libertarians can object to libertarian paternalism even if libertarian paternalism is not an oxymoron. 


(1) Cass Sunstein and Richard Thaler, ‘Libertarian Paternalism is Not an Oxymoron’, American Economic Review 93: 175 (2003); Libertarian Paternalism is Not an Oxymoron, 70 U.  Universityof Chicago Law Review 1159, 1160 (2003); Nudge: Improving Decisions about Health, Wealth, and Happiness (Penguin Books: New York, 2008).

(2) Gregory Mitchell, ‘Libertarian Paternalism is an Oxymoron’, Northwestern University Law Review Vol.  99, No.  3: 1245 - 1277.

(3) This characterization is found in Robert Nozick’s Anarchy, State, and Utopia (New York: Basic Books, 1974) (I refer to this work as ASU henceforth; Murray Rothbard For a New Liberty: The Libertarian Manifesto (Auburn Alabama: Ludwig von Mises Institute: 1973): 27; David Boaz Libertarianism: a Primer (New York, London, Toronto, Sydney, Singapore: The Free Press: 1997).

(4)This characterization of paternalism is meant only to correspond to Wall’s analysis.  I do not argue that this is an accurate conceptual analysis of paternalism, hard or soft.  For a more extensive, albeit non-decisive analysis of paternalism, see Bill Glod’s “Against Two Modest Conceptions of Hard Paternalism” Ethical Theory and Moral Practice 16: (2): 409-422 (2013)

(5) Wall, Steven “Self-ownership and Paternalism” Journal of Political Philosophy, Vol. 17, Issue 4, p. 403 (December 2009).

(6)  Ibid, 403.

(7) Ibid, 403.

(8) Of course, libertarians tend to oppose the legal implementation of policies in the spirit of Dangerous Drug.  See John Hospers “Libertarianism and Legal Paternalism” Journal of Libertarian Studies, Vol.  IV, No. 3 (Summer 1980).

(9) Sunstein and Thaler 2008: 5 - 6.

(10) Ibid, 6.

(11) Sunstein and Thaler, 2003: 1160.

(12) Ibid, 1191.

(13) See ASU ix; and Hospers (1980).

(14)  Mark D. White, The Manipulation of Choice: Ethics and Libertarian Paternalism (New York, New York: Palgrave Mcmillan 2013).

(15) ASU, 169

(16) ASU, 169

(17) ASU, 169; interpolation original.

(18) This thought experiment follows Nozick’s government funded radio program.  See ASU 93 to 94.

(19) The next few paragraphs follow a point made by Eric Mack in “Self-Ownership, Marxism and Egalitarianism, Part II: Challenges to the Self- Ownership Thesis” Politics, Philosophy and Economics 1 (2):  75 - 107 (2000).

(20) The canonical defense of private roads and highways is written by Walter Block.  See The Privatization of Roads and Highways (Auburn, AL: Ludwig von Mises Institute 2009).

(21) For an investigation into the different possible configurations of a dominant protective agency, see David Miller “The Justification of Political Authority” Robert Nozick ed. David Schmidtz (Cambridge, 2002): 10 -23.  In citing this work, I should be taken to endorse Miller’s suggestion.  Many libertarians do not think Nozick successfully justifies even the minimal state.  See, for example, Randy Barnett’s “Wither Anarchy? Has Robert Nozick Justified the State?” Journal of Libertarian Studies, Vol. 1, No. 1, 1977.