Frank Knight on the Self-Destruction of Liberalism

I recently had occasion to read Frank Knight’s wonderful 1934 essay, “Economic Theory and Nationalism,” first presented at the annual American Economic Association meetings, and published in a collection of Knight's essays titled The Ethics of Competition. The essay is wide-ranging, provocative, and surprising. The sort of thing that is unimaginable coming out of today’s academy, let alone an economics department.

Knight is considered one of the chief intellectual influences behind the Chicago School—a reputation bolstered by the editors of this essay collection: Milton Friedman, Homer Jones, George Stigler, and Allen Wallis. But Knight did not just shape a generation of militant free-marketeers. His work—and this essay in particular—had an impact on the greatest Anglophone left-liberal thinker of the twentieth century, John Rawls. Indeed, core Rawlsian arguments and themes can be found scattered throughout the essay.

One line of thinking made famous by Rawls and characteristic of "luck-egalitarian" theorists is the claim that both material and genetic endowments are arbitrary from a moral point of view. Knight makes a version of this argument, though not in service of the redistributive implications that Rawls et al. took it to have: “no defensible clear distinction can be drawn from the point of view of either ideal ethics or practical politics between external wealth and personal powers as a source of income” (Knight 313). Knight moreover buries a version of Rawls’ difference principle in a footnote. He writes there that the main challenges of a liberal society are:

to prevent continued concentration of power in the hands of individuals or organizations; to assure to all a really equal start, or at least one as fair as possible, through an ‘equitable’ sharing of the material and cultural inheritance; to arrange such ‘handicaps’ as would give everyone a ‘chance’; and to provide the best distribution of prizes for making the contest interesting to participants and spectators. In this view of social life as a contest, enforced equalization would be absurd; a game is not bad or unfair because some win and others lose, but on the contrary its interest and value depend in part on the fact (303-4).

Rawls more fully democratizes this principle. He argues that these inequalities should be distributed not simply to keep participants invested in the game of liberal politics/market economics, but in service of the maximal benefit of the worst-off. 

But the pattern is the same: Inequalities, be they from natural or social origins, are presumptively unjust. They can only be justified if they serve some larger social good. For Rawls, that good is the benefit of the worst off. For Knight, that good is buy-in into the system. If anything, I take Knight's more capacious view of human motivation to be richer than the narrowed egalitarian character of Rawls’ mechanistic maximin mantra.

(The first chapter of Katrina Forrester’s new book emphasizes Rawls’ debt to this essay in formulating his difference principle)

The language of the “game” of politics recurs throughout the essay. Knight's master argument is that liberal politics and market economics tend toward self-destruction. Liberalism of the nineteenth century—which Knight celebrates and fears has disappeared—was at its heart a commitment to individual freedom:

The political ideal of the age was negative, individualistic. Its ideal was individual liberty; i.e. there should be no ideal of individual life enforced by society. It looked toward making society a mutual aid organization of a mechanical sort, by which each individual would achieve maximum efficiency in using his own means to his own ends. In the main, this ideal still stands, though it is visibly crumbling, in America and other countries which have not yet repudiated democracy outright for a dictatorial regime (286).

In an interesting restatement of Turner’s famous thesis, Knight suggests that the liberal ethic was only stable in the age of the open frontier. For Knight, humans are characterized by an insatiable desire to exert power over one another. The possibility of moving West and building an independent life for oneself allowed man to psychologically transpose his natural love of domination into an urge to build for oneself. (The draw of the frontier and the empire perhaps had a little more to do with the desire to exert power over men than Knight acknowledges).

With the closing of the frontier, however, Knight warns that we are returning to the pre-liberal world:

Getting ahead in the world would come again to mean, as it had before, a more direct struggle with other people, than with nature, and a struggle for power over other people, which is overwhelmingly the human meaning, immediate or ultimate, of power over nature itself (290-291).

Here we return to Knight’s emphasis on the “game” metaphor for society. Liberal politics and market economics are like games. They only succeed if the participants play by the rules. But in both cases, it is exceptionally difficult to construct rules powerful enough to compel obedience. Why should the losers accept the terms? And why should the winners make it possible that they might lose in the future?

This is clear in the tendency of market economies to dissolve into monopolization. Why would a temporarily dominant firm submit to market rules that may end up elevating a new, rival market-titan tomorrow? All the incentives point the way of changing the rules of the game once you’ve won:

interest in winning and the interest in the game tend to run into conflict; too much interest in winning first spoils a game and then breaks it up altogether, converting it into a quarrel, or beyond this into a fight. Unless people are more interested in having the game go on than they are in winning it, no game is possible. … The ‘natural’ tendency is for a game to deteriorate, if the participants follow their primitive impulses without conscious exercise of moral restraint. No game is possible unless the players have the attitudes and interests to which the term ‘sportsmanship’ is understood to refer (302).

This is commonly understood with the market, Knight suggests. But the problem is just as acute in liberal democratic politics: “democracy is competitive politics, somewhat as free enterprise is competitive economics, … and it shows the same weaknesses as the latter. In ideal theory, neither is competitive in the psychological sense” (295).

(As an aside, Knight suggests that Adam Smith overlooked the role of the passions in driving human behavior. Here he echoes (or anticipates, rather) themes from Albert Hirschman's classic The Passions and the Interests. But both Knight and Hirschman neglect Smith’s discussion of the role the libido dominandi plays in sustaining feudal and slave societies. Smith too was acutely aware of man's love of domination, and the mechanics of moral motivation underlying capitalism are not that different from those underlying feudalism and slavery. On this point, see Daniel Luban’s terrific article on Smith and vanity. See also my earlier blog post on Smith/Rousseau on vanity and trinkets).

Liberal economics and democratic politics have the same weakness. They cannot sustain themselves. The raw love of power makes them incapable of being self-sustaining. Their weakness:

lies in the natural, cumulative tendency toward inequality in status, through the use of power to get more power. The main error on the political side, in the theory of liberalism as expounded by the advocates of political cures for economic ills, is that competitive politics is not better than economics in this regard, but definitely worse (296).

I’m reminded here of a comment G.K. Chesterton made in his introduction to an edition of Thomas Carlyle’s Past and Present:

[Carlyle] is already the first prophet of the Socialists and the great voice against the social wrong. He has, indeed, almost all the qualities of the Socialists, their strenuousness, their steady protest, their single eye, also something of their Puritanism and their unconscious but instinctive dislike of democracy. Carlyle was the first who called in political inequality to remedy economic inequality, but he will not be the last.

It is easy enough for socialists and other economic critics of capitalism to notice the market’s internal tendency toward monopolization and inequality. The natural response, Carlyle and his followers suggest, is to combat economic inequality with political inequality.

The mistake of this approach, Knight insists, is that political inequality is even more liable to the dangers of monopolization and domination. Social reformers insist that the state will somehow be immune to these dangers so long as they remain democratic. Knight thinks this is ridiculous. He mocks idealists like John Dewey for naively insisting on the psychological stability of democracy. As he repeats over and over again, there is no reason to think that a free democratic people won’t ultimately opt for dictatorship. The political events of Knight's day seem to suggest this is the most likely course of action.

The democratic politician—not the captain of industry—has the greatest chance of exploiting our psychological weaknesses:

Neither abstract reasoning nor the evidence of experience affords ground for belief that, given the moral drive toward such values as the dominant motive in society, democratic political processes could fail to distribute them even more unequally still than does competitive business (308-9).

At least there is something objective about the material wealth produced by competitive firms. To be electorally successful requires no such objective achievement. It requires only the capacity to captivate an audience: “it goes without saying that competence to persuade is only accidentally and improbably associated with competence to counsel and to lead” (305). Democracy rewards those who can get elected. But there is little reason to believe the skill of winning an election has anything to do with the skill of governing well.

As Plato has Socrates put it in the Gorgias: “evidently oratory produces the persuasion that comes from being convinced, and not the persuasion that comes from teaching, concerning what’s just and unjust” (454e). For Socrates, such oratory is of no value at all if it teaches injustice. The orator wins by playing to the crowd. Socrates continues: “If you say anything in the Assembly and the Athenian demos denies it, you shift your ground and say what they want to hear” (481e).

Knight says the same. To be successful in a democracy you must balance flattery with manipulation. This whole exercise is about as rational as having the patients choose the doctor.

(There’s force to this point. Then again, far better for the patient to choose the doctor than for the doctor to choose the patient).

Democracy, Knight continues, is just “government by discussion.” And there is nothing more psychologically unstable than discussion. Another crucial aspect of that instability is human pride. It is difficult for us to part with our view, even after we should have been convinced of its errancy. Knight notes that it is rare to see professional academics model this form of intellectual humility. I recall being told once that the leading scientific critics of the big bang theory were never convinced. They went to their grave believing in a steady-state account of the universe. 

How could we expect more from mass democracy? 

Democracy is government by discussion, and there is little as reliably truth-agnostic as discussion. As he remarks in a wonderful paragraph:

It [dictatorship] will perhaps be most democratic in its policy of suppressing freedom of discussion. In fact, ‘discussion’ needs little formal suppression. There is little evidence that any large mass of people ever wanted to discuss or attend to discussion, of serious issues, involving real intellectual effort. Real discussion is rare even among professional intellectuals, and their ‘argumentation’ commonly illustrates the tendency of a contest to deteriorate. Debate, and the preaching of unorthodox views, which are very different from discussion, may or may not be popular, depending on the aesthetic character of the performance. The successful dictator will have to provide suitable entertainment along these lines without permitting anything really dangerous. It is of course, a delicate problem to stage verbal gladiatorial combats and supply the craving to be shocked without running some risk of arousing loyalties competing menacingly with loyalty to a particular leader or party (323)

And again: “Once a man’s mind is liberated and set thinking, he becomes an inveterate ‘theorizer’ and is as partial to his own ideas as he is to his own children, or to any individual interest whatever” (355).

The metaphor above of "verbal gladiatorial combats" reminds me of Hobbes' wonderful complaint about democratic deliberations in De Cive:

perhaps someone will say that the popular state is preferable to Monarchy, because in that state, in which of course everyone manages public business, everyone has been given leave to publicly display his prudence knowledge and eloquence in deliberations about matters of the greatest difficulty and importance; and because the love of praise is innate in human anture, this is the most attractive of all things to all those who surpass others in such talents or seem to themselves to do so; but in Monarchy that road to winning praise and rank is blocked for most of the citizens. What is a disadvantage, if this is not? I will tell you. To see the proposal of a man whom we despise preferred to our own; to see our wisdom ignored before our eyes; to incur certain enmity in an uncertain struggle for empty glory; to hate and be hated because of differences of opinion (which cannot be avoided, whether we win or lose); to reveal our plans and wishes when there is no need to and to get nothing by it; to neglect our public affairs. These, I say are disadvantages [of democratic deliberation]. But to lose the opportunity to pit your wits against another man, however enjoyable such contests may be to clever debaters, is not such a disadvantage for them, unless we shall say that it is a disadvantage for brave men to be forbidden to fight, for the simple reason that they enjoy it (X.9).

And regarding this dangerous connection between reason, opinion, and vanity, see Publius in Federalist 10:

As long as the reason of man continues fallible, and he is at liberty to exercise it, different opinions will be formed. As long as the connection subsists between his reason and his self-love, his opinions and his passions will have a reciprocal influence on each other; and the former will be objects to which the latter will attach themselves.

That is the crucial danger: The dominance of self-love over reason. This is why we are witnessing the collapse of democracy: “Fascist-nationalism, then, seems to be clearly indicated as the next stage in the political evolution of the liberal democracies” (319).

Voters react against the moral emptiness of democracy, which has revealed itself to be little more than sloganeering:

The individual reacts from the notion of reaching validity by general discussion—which he has seen degenerate into a contest in ‘selling’—to a faith in ‘strong’ individual leadership, which also represents a reaction from moral and intellectual equalitarianism to hero worship (322).

(Cf. here my earlier blogpost on the Protagoras).

It has been a hope of political theorists for some time now that refining or elevating or promoting deliberation will somehow save or improve democracy. But for Knight and Hobbes, deliberation is the central reason why democracy is naturally self-destructive. (On this topic, see Bryan Garsten’s wonderful defense of rhetoric, which deals with Hobbesian, Rousseauian, and Kantian critiques of rhetoric that rhyme with Knight’s).

What then is to be done? Knight’s prognosis is bleak. He warns that modern social scientific methods align with ever-firmer domination, with a new “technique of control” (341). (Another aside: I find it striking that Knight sounds at various points in this essay alternatively Hayekian and Foucauldian).

Knight warns us against embracing social scientific technocracy as a means of preserving liberal society. If technocracy can succeed, it will do so by monotonizing human behavior with a new, terrible form of tyranny. The power to predict social behavior is the power to control it.

So Knight leaves us with two conclusions.

First, we must be dispositionally conservative. We must do our best to preserve the rules of the game, because we must know that when we try to change the rules, we will almost always make things worse:

One of the evils which has resulted from carrying the natural science conceptions over into the field of social discussion is the common delusion that by the happy discovery of some formula, it may be possible to change the character and constitution of society in a way comparable to the modern development of technology through science. … Political society is a game which must go on under rules, or very quickly collapse into a war of each against all (347-8).

And later: “The notion that the general mass of mankind, taken on the scale of a modern national state, can quickly and reliably think out and apply important constitutional changes, is tragic nonsense” (350).

The second solution is to create a new aristocracy. We need an impartial leadership class (like the idealized clerical class of the Middle Ages) that can protect the realm of freedom while renouncing political ambition. Technocratic social science alone promises a cure worse than the disease. What we need instead is for science to provide “intellectual-moral leadership alongside the moral-religious” (359). The emphasis on the "moral-religious" is key. The appeal of dictatorship has proved decisively that we naturally long for leaders who appeal to our passions and emotions. Deadened, rationalist scientism will not cut it. Who wants to be ruled by Dr. Fauci?

Leadership on a religious-emotional basis is an indefinitely more natural, an easier, and less costly system of order than any other. The weakness of such a system is its excessive strength, its tendency toward rigidity. If freedom is to be maintained, the rate of change must be limited—with perhaps some provision for temporary recourse to authoritarian rule in times of crisis (352).

Here Knight sounds like Coleridge or the early JS Mill. The Mill of the “Spirit of the Age” insists on finding a new leadership class that can guide modern liberal society. He argues there that modern society is in an “age of transition,” and can only return to a natural state once “the opinions and feelings of the people are, with their voluntary acquiescence, formed for them, by the most cultivated minds which the intelligence and morality of the times call into existence.”

Knight and the young Mill agree: Liberal democracies require a liberal aristocracy. My own suspicion is that such an aristocracy will be impossible to find and even harder to civilize and control.

There's lots of other interesting stuff in this essay. In particular, Knight offers an impressive account of the philosophy of social science and of the various approaches to interpreting/predicting social behavior. Much of that goes over my head, unfortunately. But it's worth emphasizing that it's there.