I shall try to respond to some of the points made in three interesting papers by way of direct comments on my essay “Government, Bound Or Unbound” or tangent to its subject.
By entitling his lively paper “Can An Omnipotent Government Make A Rock Bigger Than It Can Lift?” Michael Munger very properly places the debate on the level of logic, where it mostly belongs. The question he poses is not a true paradox, a no-exit logic trap, for it has a straightforward issue. A rock too big to lift by an omnipotent government cannot be made any more than an object that both flies and rests on the ground, or one that is both black and white at the same time can be made. Such objects are defined by contradictory premises, and cannot possibly exist. They are not feasible and lifting them is consequently also infeasible. Making, changing, or breaching rules is at least feasible, if not always simple. Government and its allies in what I call the “winning coalition” can undo rules that stop them from doing what they wish to do or feel they must do to retain their supremacy over the “losing coalition.” The basic reason is that the rule system is ineluctably self-referring: it defends certain limits upon government action by providing for the government to enforce the limits.
Though I do not think it is as primordial as theorists steeped in American political traditions and culture regard it, I must at this early point in the argument try to come to an understanding on the separation of “powers.” Michael Munger illustrates its workings by envisaging three persons of diverse inclinations and interests, each holding a different key. For a momentous decision, all three keys are needed but not many decisions will be acceptable to all three.
Munger omits to say, though, as does everybody else who represents the separation of powers as the mightiest bulwark against excessive government, that only one person of the three carries a gun, the sole “legitimate” gun in existence (as Max Weber’s question-begging definition has it). In the blessed United States, this may have no more than theoretical significance. But the threefold separation of “powers,” which I insist on calling “functions,” is being aped all over the world, from Pakistan to Venezuela and across Africa. In those places, the monopoly of the gun (politely called the “security services”) does have strong practical significance. As men are merely human, it would be strange if it did not, and it may one day have such significance even in mature civilizations if and when push comes to shove. Let that day never come.
Let us now revert to the central point. Kenneth Arrow called the “social” (or to use a less value-laden word, the “collective”) choice rule a constitution. This equivalence should be a reminder that politics is the rule-bound practice of dividing a population in two halves, one whose preferences prevail and one who must forgo theirs. Unanimity needs no rules, while a collective choice rule is a standing invitation to potential winners to coalesce and impose their will on the prospective losers.
The question, then, reduces to this: given that a prospective government must bid for enough support to form a winning coalition, and an incumbent must do the same to prevent rival bidders from unseating it, can a collective choice rule that would thwart the auction and put the maximizing solution out of reach, survive for long?
Michael Munger clearly answers “yes”: limits on government action are perfectly “incentive-compatible” (both Ulysses and the ship fare better if he stays tied to the mast). The contrary temptation is rendered powerless by a commitment to resist it, and that includes a complex commitment not to sanction those outside the government who may impose the resistance upon it. Now, commitment in the rigorous sense is rare. It is difficult to make credible and effective, and is vulnerable to abuse and exploitation. Nevertheless, if government could have commitment power, we should sing Hallelujah. The thesis that they do have commitment power would revolutionize political theory. Remember that Thomas Schelling equated a constitution to a vow. Unlike a true commitment, a vow is kept as long as you choose, for whatever reason, to keep it. Unlike a promise to another (which is stronger than a vow but may fall short of a commitment) the breach of a vow does not violate the right the promise has conferred upon another party and hence involves little risk of retaliation.
However, Munger wisely does not long insist on commitment, but falls back on self-restraint by prudential interest. He cites the example of 17th and 18th century England that never used its power to default on its debts and was therefore always able to raise money at reasonable rates. The French king, on the other hand, was too powerful for his own good, so lenders would not accept the default risk and thus he was left chronically short of cash.
I should be among the last to complain of subjective readings of history, but this particular one certainly needs some amplification. The French royal administration was not unaware that if it defaulted on a loan, Florentine bankers would get the message and future borrowing would become difficult. The problem was that the judgments, or probably misjudgments, made about the requirements of French national security had led to almost permanent warfare, first against the Spanish Habsburgs and then against England, culminating in the Second Hundred Years’ War ending at Waterloo. By the end of Louis XIV’s reign, France had been bled white and was quite poor. Things got no better during the Regency and the reign of the amiable Louis XV (who, far from being too powerful, was unable to prevail over the established corporations and his excellent ministers had to abandon their efforts at reform), and so the hapless Louis XVI inherited a country teetering on the edge of bankruptcy. In short, France could not borrow because it was insolvent, England could because it was more and more solvent. The French kings being too powerful, as Munger tells it, may have had something to do with France becoming insolvent, but the causation is second- or third-order.
Be that as it may, it would be foolish to deny that self-imposed restraint may bring benefits to the government and its winning coalition in certain areas. Such restraints lead the winners to forgo certain benefits that could be had from imposing losses on the losers. The real point at issue is not which is bigger, nor the distributive incidence of each. The real issue is whether there is a set of rules that can actually force a government not to do what it desires to do, or believes it needs to do, and would do if the rules were different. It is having to live with such a rule system that would be “incentive-incompatible.”
What “Is Limited Government Possible?” seeks to prove is that if such a rule system were put on the statute books, it would be amended, corrupted, bent, got round of, or swept away. One enabling cause of this fate is that the self-referring enforcement mechanism upon which constitutions depend is intrinsically defective. Government will not punish itself. It strikes me as odd that none of the three respondents to my paper has paid any attention to this point. Perhaps it is less important than I believe.
Munger concedes, with a regret I share, that the zone of collective choice has been massively encroaching on the zone of individual choice in much the same way as is explained and predicted by the solution of the maximization problem faced by the government and its winning coalition. I should like to feel that he and I see eye to eye on why this happens.
I do not relish a Methodenstreit, but Gerald Gaus raises a vital point that is drawing me into one. He deplores that my analysis runs in terms of the narrow interests of government and its allies. He believes that moral sentiments are the really important elements of political motivation, and would include them in people’s “utility functions” where they might be the dominant variable. Having done this, he plants us back in the tautological and obsolete world of utility and its analytical extension, utility maximization. Moral sentiments are commingled with any and all other reasons that may motivate choice, and disappear in the impenetrable darkness of the mythical entity “utility.” People make the choices that have the greatest utility, for otherwise they would have made different choices. The statement that they maximize it is neither verifiable nor falsifiable. One is moved to say, with Wittgenstein: “Whereof we cannot speak, thereof we must be silent.” The trouble is that even in these modern days this fastidious advice is so lightly disregarded.
The saving grace of using interest as the motive of choice is that for all its narrowness, it deals in matter that is identifiable, ascertainable and with a bit of luck even quantifiable. It rests on the robust assumption that people would rather have more advantages than less, and will choose alternatives accordingly. Since the advantages can be largely discerned without getting inside people’s heads, the statement is falsifiable, meaningful. It predicts broad outcomes not too badly, though not faultlessly. Imputing outcomes to our moral sense explains outcomes absolutely faultlessly. It cannot be disproven.
One can feel dissatisfied with Gaus’s reversion to the utility-maximizing paradigm on a less exalted level, too. The moral sentiment most relevant to political choice is fairness. People claim fairness mostly for themselves, but they undoubtedly want it for everybody else as well. However, the notion of fairness is an extraordinarily messy, disorderly one. Theorizing about it is an enterprise rather like herding cats. (The attempt to account for the emergence of fairness norms in Cristina Bicchieri’s highly-rated “social grammar” conjures up in my mind a squadron of over-equipped, late-model fighter planes sent to chase a desert Arab on a donkey). Experimental game theory finds that players often deviate from the solution that would maximize money payoffs, and these deviations may be imputed to fairness motives. But it is one thing to interpret the proposers’ tendency in an Ultimatum Game to aim not too far short of half-and-half sharing as a mark of their sense of fairness. It is quite another to translate this finding, derived in highly artificial conditions, to real-life political choices. It is done nonetheless, and we have no means for saying that it is wrong, but we are certainly not justified in claiming that it is right. Egalitarian policies look like confused products of some fairness notion. But they also look suspiciously like the product of one group using politics to make itself better off at the expense of others.
The final point of my rejoinder to Gaus concerns remedies. He proposes that “we should use our intelligence in determining where and how to employ our moral sentiments.” But we have already determined how to employ them, and we can only wish good luck to whoever sets out to persuade us to employ them some other way. Appealing instead to people’s interests and persuading them that egalitarian policies are ultimately harmful even to the apparent gainers among them, looks to me a slightly less unpromising endeavor.
I regret that Randy Barnett chose not to refer to my paper, for I should certainly have benefited from his criticism. He provided instead a compact and persuasive account of the erosion of the U.S. Constitution. I am neither competent nor called upon to comment on it.
He also puts forward two remedies to which I should like to append two remarks.
One concerns the proposal to reverse the course of constitutional development and steer it back toward the framers’ original intent. To achieve this, Barnett suggests, America must elect the right people. However, for the last two hundred years America elected the wrong ones, and American voters look unlikely to shed this bad habit merely to recover the constitution of original intent. So far, they have behaved as if they wanted to get ever further away from it. One would welcome the discovery of an incentive that would induce them to act differently, and durably so.
My other remark relates to the libertarian idea supported by Barnett of a polycentric system of competing providers of justice and law enforcement among which people could freely choose. One is reminded of Nozick in his libertarian phase who envisaged the emergence of competing “protective agencies” that would defend their clients’ “rights.” Nozick saw this as an industry working to increasing returns. Hence a single dominant protective agency would ineluctably emerge. Oddly enough, Nozick did not move on to the further ineluctable consequence that the single agency would protect “rights” against everybody except itself, nor that it would offer more and more free or subsidized public goods and services to the disproportionate benefit of its supporters and at the disproportionate cost of its potential opposition.
The conclusion that tentatively suggests itself is that no solution to the problem of “unbound” government is likely to be found in formal constitutional reform, restoration, or ingenious new design. Formal constitutional constraints are not good or bad, strict or loose, but ultimately irrelevant.