Kant: Liberal, Illiberal, or Both?

Since it’s possible for thinkers to make mistakes in drawing practical implications from their own principles, the question “how good a classical liberal is Kant?” must be distinguished from the question “how good a foundation for classical liberalism does Kant’s philosophy provide?”

It might turn out, for example, that while Kant himself defends classical liberal policies, the more basic principles to which he appeals in doing so actually have far more illiberal implications. (This is essentially the Randian critique to which Mark White alludes.) Or it might be the other way around: that Kant’s principles actually support a far more liberal approach to politics than the one he ends up defending.

My own answers to those questions are complicated. I think Kant’s political writings have strong classical liberal, even libertarian strands, and that these stem in large part from his more basic principles. I also think his political writings contain deeply illiberal elements. In many cases, I take these to be the result of Kant’s misapplication of his own principles, and would expect a more consistent application of those principles to result in a still more thoroughgoingly libertarian set of policies. To that extent, I take the side of Kant’s defenders against critiques like Rand’s.[1] All the same, I also think that some of the illiberal aspects of Kant’s politics do have to be laid, at least in part, at the door of flaws in his basic philosophy; and here I find a kernel of truth in Rand’s criticisms.

The classical liberal components of Kant’s political philosophy are undeniable. His account of a system of rights as defining the “fully reciprocal use of coercion that is consistent with everyone’s freedom”[2] is an exemplary libertarian formulation, and his account of property rights[3] lies solidly within a Lockean tradition that Kant shares with such libertarian thinkers as Lysander Spooner,[4] Émile Levasseur,[5] Murray Rothbard,[6] and Robert Nozick[7] (though also, more complicatedly, Johann Gottlieb Fichte[8] and G.W.F. Hegel).[9] Indeed, I think Kant’s account of the justification of property rights is in some respects more subtle and profound than that found in many libertarian writers. He also applies his theory of property in some admirably libertarian ways – for example, in his defense of native land rights against colonialist expropriation.[10]

But however libertarian Kant’s theory of property may be in its basic outlines, it is far less so in application. And here I cannot agree with Dr. White’s statement that “Kant did not support forced beneficence,” or that Kant’s political theory “rules out any state welfare system.” As part of the social contract (a fictional social contract – more on that anon), Kantian citizens are understood to cede to the ruler a degree of ownership over their private holdings; hence Kant explicitly defends

the right of the supreme commander … as supreme proprietor (lord of the land), to tax private owners of land, that is, to require payment of taxes on land, excise taxes and import duties, or to require the performance of services (such as providing troops for military service) …. On this supreme proprietorship also rests the right to administer the state’s economy, finances, and police.[11]

Notice how the right to conscript citizens into military service is casually included here as part of the taxing function. The Kantian state’s claim on its citizens goes well beyond a mere claim to their external property.

In any case, the taxing power also includes “the right to impose taxes on the people … to support organizations providing for the poor, foundling homes, and church organizations,” which sounds like a state welfare system to me. Kant even stresses that such public support is to occur “not merely by voluntary contributions” but “by way of coercion,” explaining that taxpayers “have acquired an obligation to the commonwealth, since they owe their existence to an act of submitting to its protection and care, which they need in order to live.”[12] Kant goes on to maintain that taxation is morally superior both to lotteries (because they’re regressive) and to begging (the latter supposedly being “closely akin to robbery,” while taxation allegedly isn’t).

Kant’s (classically) illiberal streak goes much farther than mere support for state welfare. He holds that illegitimate children have “stolen into the commonwealth … like contraband merchandise” and so stand outside the protection of civil law;[13] he denies citizens a right of self-defense against the state by declaring the ruler to be above the law and not to be rebelled against;[14] he denies wage workers the right of self-defense against the wealthy classes by depriving them of the vote;[15] and he is so enamored of capital punishment as to insist that “if a civil society were to be dissolved by the consent of all its members … the last murderer remaining in prison would first have to be executed.”[16]

One of the most problematic portions of Kant’s theory of justice is his category of “rights to persons akin to rights to things.” These are rights to treat certain other people to some degree as one’s property, a right allegedly consistent with the categorical imperative not to treat others as mere means. Thus Kant justifies the coercive authority of husbands over their wives, since “this dominance is based only on the natural superiority of the husband to the wife”;[17] he likewise justifies a coercive authority of employers over employees, maintaining that a head of household is entitled to “fetch servants back and demand them from anyone in possession of them, as what is externally his.”[18]

How much more “Kantian” in spirit is Herbert Spencer, who rejects the husband’s authority over his wife as “utterly repugnant,”[19] and who describes the “master-and-workman type of industrial organization” as a “remnant of the régime of status” by which an employee is “temporarily in the position of a slave,” and so calls for the replacement of the wage system by a system of workers’ cooperatives.[20]

Now many of these cases can fairly be diagnosed, at least in part, as instances in which Kant has simply failed to apply his own principles correctly. In particular, he understands contract right as a right to specific performance – a “possession of another’s choice”[21] – whereas a more authentically “Kantian” approach, insistent on rejecting all forms of interpersonal subordination, would see personal services as inalienable, and thus would treat contractual obligation as rooted in the transfer of title to external resources only.[22] Likewise, using a greater degree of compulsion against a person than is necessary to restrain her from injuring others arguably treats that person as a mere means, a consideration that invalidates the institution of coercive punishment.[23]

But Kant’s theory of the social contract, which is the basis both for his claim that a political ruler is the supreme proprietor of citizens’ land, and for his rejection of the right of rebellion, is a bit more deeply rooted. Regrettably, Kant’s version of social-contract theory owes as much to Hobbes and Rousseau as to Locke.

To begin with, Kant’s theory that rights in the state of nature are simultaneously genuine and merely provisional is an unconvincing attempt to fuse a) the Lockean insight that rights exist prior to the state with the Hobbesian claims that b) we have no reason to respect others’ rights in the absence of law, and that c) law cannot exist without the state.[24]

Both of the Hobbesian claims, (b) and (c), seem incoherent. On the one hand, the view that in the state of nature one should withhold cooperation until other parties show themselves to be cooperative first is evidently not universalizable, which should be enough to condemn it by Kantian standards.[25] On the other, by confusing law with legislation,[26] Kant, in his very attempt to defend the rule of law, makes the rule of law impossible by exempting the ruler from its application.

If this Hobbesian muddle weren’t bad enough, Kant maintains that victims of state coercion are really free, since the state represents their true will via the social contract – so that, e.g., when the state taxes its citizens it’s really exercising their property rights on their behalf. Kant thereby adds Rousseau to the Locke/Hobbes mix – with the added twist that the contract need not even be genuine.

I’ve written elsewhere of the conflict between vicarious and immanent forms of liberalism.[27] Immanent liberalism is concerned to protect relations of mutual consent in people’s everyday lives, while vicarious liberalism displaces the chief locus of consent somewhere else – to the social framework as a whole, or to a merely hypothetical starting point, or both. Kant’s liberalism contains both immanent and vicarious strands; but the amount of emphasis placed on the vicarious strand threatens to poison the system, since actual curtailments of everyday freedom are all too easily justified as part of a system that one has tacitly consented to, or would hypothetically consent to.

And here I do think we find a set of problems that can be traced back to defects in Kant’s basic ethical principles. Kant’s theory of the categorical imperative draws a sharp distinction between duty on the one hand and happiness on the other, where happiness is seen primarily as a matter of following our subjective inclinations. Dr. White is right to point out that Kant is not nearly as hostile to individual happiness as he is often thought to be; nevertheless, Kant insists that there is no necessary connection between duty and happiness, so that they always can come apart even if they usually don’t.

On the one hand, the duty/inclination divide fuels Kant’s Hobbesian side; since inclinations are essentially nonmoral and always potentially antimoral, we can conclude a priori that people cannot be trusted without state compulsion. But on the other hand, the duty/inclination divide also fuels Kant’s Rousseauvian side, since the Kantian doctrine that we are autonomous only when choosing duty over inclination lends itself all too easily to vicarious liberalism and the privileging of indirect, tacit, or hypothetical consent over actual consent. This is what is right in the Randian critique of Kant.                                       

Where Rand goes wrong, I think, is in taking Kant’s theory of a duty/inclination divide as motivated by a pathological hatred of human happiness. The duty/inclination divide represents Kant’s attempt to solve a genuine philosophical problem; I think he gives the wrong solution to it, through grasping one half of the truth while discarding the other. And while Rand is right to identify Aristotle’s approach to ethics as offering the best way out,[28] I don’t think Rand quite manages to follow where Aristotle leads; instead, she ends up seizing the piece of the truth that Kant has discarded while discarding the piece he has seized. But that’s a point I’ll elaborate on in the rest of the discussion.