Leoni on Voting and the Market
Source: Bruno Leoni, Freedom and the Law, expanded 3rd edition, foreword by Arthur Kemp (Indianapolis: Liberty Fund 1991). Chapter: 4: Voting Versus the Market
Copyright: This work is copyrighted by the Institute for Humane Studies, George Mason University, Fairfax, Virginia, and is put online with their permission.
Fair Use: This material is put online to further the educational goals of Liberty Fund, Inc. Unless otherwise stated in the Copyright Information section above, this material may be used freely for educational and academic purposes. It may not be used in any way for profit.
Voting Versus the Market
We have seen in the preceding lecture that notwithstanding many similarities that may exist between voters on the one hand and market operators on the other, the actions of the two are far from actually being similar. No procedural rule seems able to allow voters to act in the same flexible, independent, consistent, and efficient way as operators employing individual choice in the market. While it is true that both voting and operating in the market are individual actions, we are compelled, however, to conclude that voting is a kind of individual action that almost inevitably undergoes a kind of distortion in its use.
Legislation considered as a result of a collective decision by a group—even if consisting of all citizens concerned as in the direct democracies of ancient times or in some small democratic communities in medieval and modern times—appears to be a law-making process that is far from being identifiable with the market process. Only voters ranking in winning majorities (if for instance the voting rule is by majority) are comparable to people who operate on the market. Those people ranking in losing minorities are not comparable with even the weakest operators on the market, who at least under the divisibility of goods (which is the most frequent case) can always find something to choose and to get, provided that they pay its price. Legislation is a result of an all-or-none decision. Either you win and get exactly what you want, or you lose and get exactly nothing. Even worse, you get something that you do not want and you have to pay for it just as if you had wanted it. In this sense winners and losers in voting are like winners and losers in the field. Voting appears to be not so much a reproduction of the market operation as a symbolization of a battle in the field. If we consider it well, there is nothing “rational” in voting that can be compared with rationality in the market. Of course voting may be preceded by argument and bargaining, which may be rational in the same sense as any operation on the market. But whenever you finally come to vote, you don’t argue or bargain any longer. You are on another plane. You accumulate ballots as you would accumulate stones or shells—the implication being that you do not win because you have more reasons than others, but merely because you have more ballots to pile up. In this operation you have neither partners nor interlocutors but only allies and enemies. Of course your own action may still be considered rational as well as that of your allies and enemies, but the final result is not something that can be simply explained as a scrutiny or a combination of your reasons and of those of people who vote against them. The political language reflects quite naturally this aspect of voting: Politicians speak willingly of campaigns to be started, of battles to be won, of enemies to be fought, and so on. This language does not usually occur in the market. There is an obvious reason for that: While in the market supply and demand are not only compatible but also complementary, in the political field, in which legislation belongs, the choice of winners on the one hand and losers on the other are neither complementary nor even compatible. It is surprising to see how this simple—and I would say obvious—consideration of the nature of group decisions (and particularly of voting, which is the usual procedural device used to make them) is overlooked by both the theorists and the man in the street. Voting, and particularly voting by majority rule, is often considered a rational procedure not only in the sense that it renders it possible to reach decisions when the members of the group are not unanimous, but also in the sense that it seems to be the most logical one under the circumstances.
It is true that people usually admit that a unanimous decision would be ideal. But owing to the fact that unanimity in group decisions is rare, they feel entitled to conclude that the second best is making decisions by majority vote—the implication being that these decisions are not only more expedient but also more logical than any others.
On another occasion I have dealt with a defense of this position by Dr. Anthony Downs.1 I think it is worthwhile to reconsider Downs’s argument, which has the merit of summarizing in a short way all of the main reasons presented in favor of majority rule in the political literature that I know.
According to Dr. Downs,
the basic arguments in favor of simple majority rule rest upon the premise that every voter should have equal weight with every other voter. Hence, if disagreement occurs but action cannot be postponed until unanimity is reached, it is better for more voters to tell fewer what to do than vice versa. The only practical arrangement to accomplish this is simple majority rule. Any rule requiring more than a simple majority for a passage of an act allows a minority to prevent action by the majority thus giving the vote of each member of the minority more weight than the vote of each member of the majority.2
To continue our favorite comparison between voting and operating on the market: This argument seems to be the same as saying that we must give a one dollar bill to everybody in order to give each one the same purchasing power. But when we consider the analogy at closer quarters, we realize that in assuming that 51 voters out of 100 are “politically” equal to 100 voters, and that the remaining 49 (contrary) voters are “politically” equal to zero (which is exactly what happens when a group decision is made according to majority rule) we give much more “weight” to each voter ranking on the side of the winning 51 than to each voter ranking on the side of the losing 49. It would be more appropriate to compare this situation with that resulting in the market if 51 people having one dollar each combine in buying a gadget which costs 51 dollars, while another 49 people with 1 dollar each have to do without it because there is only one gadget for sale. The fact that we cannot possibly foresee who will belong to the majority does not change the picture much.
Some historical reasons obviously played a very important role in preventing people from reflecting on the contradictions of a doctrine that claimed to support equality of opportunities for everybody in politics and simultaneously denied that very equality through the application of the majority rule. The supporters of majority rule used to conceive of it as the only possible means of opposing the unrestricted power of oligarchies or tyrants over the large masses of people. The “weight” given to the will or to the “ideal vote” of tyrants in the political societies that they dominated appeared so disproportionately overwhelming in comparison to the weight left to the will of all of the other individuals in those societies that the application of majority rule seemed to be the only suitable way to restore the equality of “weights” for the wills of all the individuals concerned. Very few people bothered to inquire whether the political scale was not going to become unbalanced on the opposite side. That common attitude is poignantly expressed, for instance, in a letter that Thomas Jefferson wrote to Alexander von Humbolt on June 13, 1817:
The first principle of republicanism is that the lex majoris partie is the fundamental law of every society of individuals of equal rights: to consider the will of the society enounced by the majority of a single vote as sacred as if unanimous is the first of all lessons in importance, yet the last which is thoroughly learnt. This law once disregarded, no other remains but that of force, which ends necessarily in military despotism. This has been the history of the French revolution; and I wish [so Jefferson added in some prophetic mood] the understanding of our Southern brethern may be sufficiently enlarged and firm to see that their fate depends on its sacred observance.3
Only later in the last century did several prominent scholars and statesmen start to realize that there was no more magic in the number 51 than in the number 49. For instance the French garantistes as well as some famous English thinkers had no hesitancy in declaring their dislike for the unconditional application of majority rule in political decisions, and for the underlying assumption that Herbert Spencer was to brand in 1884 as the superstition of “the divine right of majorities.”4
But let us repeat Dr. Downs’s summary of the main arguments in favor of majority rule. “If disagreement occurs but action cannot be postponed until unanimity is reached, it is better for more voters to tell fewer what to do than vice versa. The only practical arrangement to accomplish this is simply majority rule.”
We may admit that the circumstances hypothesized by Downs, urgency of a decision and lack of unanimity, may occur more or less frequently in all political societies. The fact, however, is that both urgency of decision and lack of unanimity may be, so to speak, artificially created by people who are in a position to compel all the other members of a political community to make any group decision whatsoever instead of making none. I propose to go back to this question. I only wish to point out here that even if we assume that both urgency and lack of unanimity are the existing conditions of the decision concerned, to state flatly, as Dr. Downs does, that therefore “it is better for more voters to tell fewer what to do than vice versa” is a simple nonsequitur. Indeed, we can easily imagine situations in which only a few people have the necessary amount of knowledge required to make the corresponding decision, and that therefore it would be much less reasonable in these cases for more voters to tell fewer what to do than vice versa.
Of course the enthusiastic supporters of unconditional majority rule may counter that they derive their conclusion not so much from the hypotheses of urgency and lack of unanimity as from the implicit hypotheses of equal knowledge, or even of equal ignorance, on the part of the voters of the issues at stake. This last hypothesis, however, is rather unrealistic, particularly in contemporary highly differentiated societies. The same author admits in another connection, “specialization creates minority groups with objective interests [and, I would add, with corresponding kinds of knowledge] which differ widely from each other.” Thus, the real basis of the conclusion is still the famous concept of “equal weight” of voters or, to say it more properly, the amphibious use of this slippery concept.
There is one more point of the summation presented by Downs to be examined. “Any rule,” he says, “requiring more than a simple majority for passage of an act allows a minority to prevent action by the majority, thus giving the vote of each member of the minority more weight than the vote of each member of the majority.”
Let us concentrate first on the last part of this statement. It seems to be indubitable that if the rule adopted is a qualified majority rule, the number comprising the winning majority out of 100 voters would be, say, 60 or 70, instead of 51, and the corresponding number comprising the losing minority would be 40 or 30 instead of 49. But this does not mean that the vote of each member of the minority is now given “more weight” than the vote of each member of the majority. The fact is that once again out of a total or set of 100 voters, the winning voters ranking in the subset of 60 or 70 are still each given, according to the new rule, more weight than is given to each of the voters ranking in the subset whose sum is 40 or 30. The 60 or 70 voters are considered in this new example as “politically” equal to 100, while the other 40 or 30 are considered as “politically” equal to 0.
The only difference we can notice in this example is that each voter, when the “magic” number is 60 or 70, has—in abstracto—a lesser probability of ranking in the losing subset than he had in the previous example where the magic number was 51. But it would be wrong to derive from this statement the conclusion that “therefore” each voter ranking in the losing subset in the latter example is given more weight than each voter ranking in the winning one.
We must now examine the first part of Mr. Downs’s statement whose last part we have just considered.
As we saw, Downs refers to any rule requiring more than a simple majority for making a political decision as one allowing a minority to prevent action by a majority, and he seems to imply that this possible prevention should always be rejected according to his principle of the equal weight of the voters.
It is obvious, however, that there are several kinds of “prevention” and a more careful analysis of this concept is probably indispensable before drawing conclusions.
It may be useful to remember in this connection an example presented at the beginning of this century in this country by a distinguished scholar (whose name is probably unduly forgotten in our time: Lawrence Lowell) in his stimulating book, Public Opinion and Popular Government.5 I have quoted this example on another occasion but it seems to me so good that I would like to repeat it. Bands of robbers, said Lawrence Lowell, do not constitute a “majority” when after having waited for a foot traveler in a lonely place they deprive him of his purse. Nor can the latter be called a “minority.” There are constitutional protections and, of course, criminal legislation in the United States as well as in other countries, tending to prevent the formation of such “majorities.” I must admit that several “majorities” in our times have often much in common with the peculiar “majority” described by Lawrence Lowell. Notwithstanding this, it is still possible and, I would say, very important to distinguish between the paradoxical “majorities” of the Lowell type and “majorities” in a more orthodox sense. Majorities of the Lowell type are not allowed in any efficiently organized society of this world for the simple reason that practically every member of these societies wants to be given the possibility of preventing at least some actions by any majority. Downs’s argument that in those cases each voter of the minority is given more weight than each voter of the majority would not be considered very impressive by anybody, even if we assumed that argument were correct.
Would we then maintain that those cases in which each individual wants to preserve his power to prevent majorities, regardless of their size, from taking actions like robbing or murdering are perfectly similar to other cases in which a capricious or wicked dissenter would be able to prevent his fellow citizens from reaching some innocent and useful ends of their own? It seems obvious that the word prevention has a different meaning in each case, and that it would be advisable to distinguish between these cases before assuming any general conclusion about the applicability of the majority rule. In other words, even if we admitted the correctness of the “equal weight” argument, we should recognize the necessity of some important qualifications that would reveal the insurmountable limits of the argument itself.
There is one more argument that we have to consider among those presented in favor of simple majority rule. As we saw, Downs not only objected to the application of other rules as contrasted to the principle of “equal weights,” but stated rather flatly that “the only practical arrangement” in order to let more voters tell fewer what to do was “simply majority rule.” The idea is that if we adopt any other (qualified) majority rule, a minority would be able to prevent a majority from telling the former what to do. Or, to put it in another way, a minority would tell a majority what not to tell. Unfortunately for the supporters of unconditional majority rule, this argument in favor of simple majority rule is no more correct than the preceding ones.
The adoption of simple majority rule does not prevent strongly organized minorities from telling all the other members of the political community what to do. The Italian theory of elites presented by Mosca, Pareto, and to a certain extent by Roberto Michels has always strongly emphasized this possibility. In their recent essay, The Calculus of Consent, John Buchanan and Gordon Tullock, while trying to reject the elite point of view, have rather unconsciously adopted it in their analysis of vote trading as a real phenomenon taking place in representative democracies of our times.
Unless I am wrong, Buchanan and Tullock managed to demonstrate in a probably irrefutable way that whenever a minority is well organized and determined to bribe as many voters as necessary in order to have a majority ready to pass a desired decision, the majority rule works much more in favor of such minorities than is commonly supposed. If for instance only ten voters out of a hundred gain the whole benefit of 100 dollars from a group decision whose costs of 100 dollars are to be charged equally to each member of the group, those ten voters may be interested in bribing 41 more people—at least by refunding to each of them his individual cost for the decision, that is, 1 dollar each. In the end, 41 people belonging to the majority will then break even, 40 belonging to the official minority will pay 1 dollar each without getting any benefit from the decision, and each of the real gainers will get 10 dollars benefit against 5.10 dollars cost from the decision taken by the group.
Of course the rule may also work in an opposite way when the losing 40 manage to organize themselves the next time and bribe at least two members of the majority in order to transform the latter into a minority and leave the ten shrewd fellows with empty hands. But it is obvious that simple majority rule may actually work in favor of a minority in both cases.
True, one cannot say that simple majority rule is the only one besides minority rules susceptible to working in favor of minorities. All other majority rules may work in favor of some minorities when the benefits resulting from the group decision are concentrated and its costs are distributed enough to encourage shrewd minorities to bribe as many voters as necessary to form the majority prescribed by the existing rule. The costs of that bribing, however, increase according to the number of voters prescribed for the formation of a valid majority. It may well happen that the costs involved in the process of bribing an increasing number of voters will discourage any shrewd minority from attempting to maximize their utilities at the expense of their fellow voters or of a part of them. The conclusion seems to be that simple majority rule is to be less discouraging for said minorities than other possible qualified majority rules. This sounds rather different from the statement we have been considering so far. That is, that simple majority rule would be the “only practical arrangement” in order to let “more voters tell fewer what to do.”
One of the interesting possibilities also pointed out in the analysis made by Buchanan and Tullock of the way in which simple majority rule may work is the continuous attempt on the part of new minorities of maximizers to bribe other voters originally indifferent to the issue concerned in order to create new ephemeral majorities at the expense of the less informed, less shrewd, or less careful minorities. Another interesting possible result pointed out by the same authors is that the disproportion between benefits and costs for the maximizing minorities induces them to neglect the possibility of minimizing the total costs of the group decision they are promoting in their own behalf.
The general situation may be termed, as I termed it in another connection, a legal war of all against all or, to adopt the famous expression used by the eminent French economist and political scientist Frederic Bastiat, the great fiction of the state “by which everyone seeks to live at the expense of everyone else.”6
A continuous overinvestment through group decisions tends to take place in a political community whenever the decision-making rules are such as to encourage minorities of shrewd maximizers to get something for nothing by letting minorities of less shrewd victims foot the bill. Dr. Downs has tried, however, to defend simple majority rule against the accusation of favoring the initiatives of maximizing minorities at the expense of other members of the group. In his essay devoted to the analysis of an article of Tullock’s that has been substantially reproduced in the new book by Tullock and Buchanan, Downs makes the already mentioned remark that all kinds of qualified majority rules may work in approximately the same way as simple majority rule in order to encourage shrewd minorities to maximize their own profits at the expense of other voters. In fact, Downs is forced to admit that simple majority rule is not exempt from the above-mentioned criticism. But he seems to completely neglect the fact, now stressed by Tullock and Buchanan, that qualified majority rules appear to be more discouraging than simple majority rule for all kinds of shrewd maximizing minorities.
Downs also tries to defend simple majority rule as well as other (qualified) majority rules from the accusation that they would tend to produce overinvestment through a series of group decisions passed by ephemeral and changing majorities. He admits that “logrolling” takes place in the real world, but he assumes that if the costs actually tended to exceed the benefits in the decisions adopted by political communities based on representatives elected by the people, these representatives would be obliged to reveal the bad results of the vote trading at the end of their legislature. Then the electors would get disgusted and punish them by appointing other representatives.
However, this argument does not seem very convincing. One trait of the political game is that in contemporary political societies logrolling starts earlier than in the Chambers of the representatives. It actually starts in the constituencies when the constituents accept some (for them) disadvantageous items of a political program in order to get some benefit from other items that are advantageous for them. They may consider among the disadvantageous items of the program the possibility that the logrolling their representative is going to start at the Chamber will result in some net losses for them at a given moment. This moment may or may not coincide with the moment at which their representative will present himself to his electors at the end of the legislature. But even if it coincides, their representative will be able to argue that, if he were to be returned to office, he would have new opportunities for improving the present situation through a more beneficial series of logrolling on behalf of his electors. Because the political game is never finished, there is no reason for the electors to fire a representative who can always state that at a given moment, or in a series of given moments, the results of his logrolling were beneficial for his electors, and who can further argue that analogous or even higher benefits would result from his continuation of the same logrolling in the next legislature.
Downs tries, however, to adopt another argument in favor of his thesis. He concedes that if voters are ignorant of some of the costs resulting from a series of group decisions they may vote for a program containing a government budget which is really too large. But, he says, if we admit ignorance into the picture, voters may also be ignorant of some of the benefits they receive, and if this is the situation ignorance may produce a government budget that is too small as well as one that is too large.
It seems to me that this argument is even less convincing than the preceding one. We moved from the rather realistic assumption that shrewd minorities know better than their fellow voters what costs and benefits will emerge for them from a decision that they try to have passed by the group, notwithstanding the possible overinvestment that this decision implies for the group. If such a decision is to pass, it is precisely because other voters are to foot the bill, which means that they are careless or less organized or ignorant of the real consequences for them emerging from the decision. Ignorance may be, therefore, one of the reasons overinvestment occurs, although it is not the only one. But let us assume that ignorance (on the part of the losing voters) is the only reason. Downs counters that ignorance may also produce a government budget which is “too small.” But this is a completely different case which has very little to do with the case of overinvestment that we have just considered. Downs’s argument would be acceptable only if we could assume that merely because of the ignorance of the voters who are footing the bill for a decision promoted by a shrewd minority, the benefits of that decision for the group may not only be less but also greater than the costs. But this would imply that investment is a form of gambling in which rational and well-informed operators are no more likely to succeed than irrational or ignorant ones. If ignorance could randomly produce the same beneficial effects as information, economic activities would obviously be very different from what they are now. In the world in which we live, the assumption that ignorance may pay as well as information seems to be rather inappropriate to explain human action—not only in the economic field but in the other fields as well. On the other hand, it is reasonable to assume that minorities who promote a group decision for their own benefit know what they are doing better than the other voters. They have a precise idea of what they want and of the possible results for them of the corresponding group decision. They may know equally well that the benefits of the decision for the other members of the group will be less than the costs for those same members. But they can disregard this fact. The final result will very probably be an investment that will cost much more than it would have cost under different conditions.
We have seen that simple majority rule is not the only rule that may cause these effects. Any other kind of (qualified) majority rule may have similar results. We have also seen, however, that qualified majority rules may work better than simple majority rule to discourage maximizing minorities from imposing their will on the whole group through the well-known procedure of logrolling. It would seem at this point that the more you increase the majority necessary for making a group decision, the more you protect the dissenting minorities from being exploited by well-organized elites of maximizers. However, this is not the case. As Buchanan and Tullock have shown in their work, the costs of reaching an agreement among the voters of a group tend to increase rather sharply when the qualified majority rule approximates the unanimity rule. In other words, any voter tends to consider his consent as very precious when he knows that the number of voters needed for a valid decision is very high. If the other voters want his consent he may be tempted to act in the same way as a discriminating monopolist in order to obtain the full advantage of the bargain.
A situation which is to a certain extent analogous to that occurring with shrewd minorities under simple majority rule tends to emerge under highly qualified majority rules or under unanimity rule. New minorities of maximizers may emerge, not in order to buy other people’s votes at the cheapest price, but merely to sell their own votes at the highest price to those voters who want to have a decision passed under the existing rule, i.e., a highly qualified majority rule. Unanimity rule would simply exacerbate this tendency, and this is the reason unanimity rule is very rarely adopted because of the high or even prohibitive costs that it may imply for all those who want to have a decision passed under that rule.
If we now go back to the concept of “equal weight” of the voters, we must conclude that no rule for decision making is really apt to give equal weights in the sense of equal possibilities to each and all of the voters. However, it may be presumed that some qualified majority rules tend to put all of the voters in a position of a fair equilibrium, while minority rules, simple majority rules, highly qualified majority rules, and finally unanimity rules inevitably result in disequilibrium for the voters concerned.
This conclusion reminds us of some insurmountable differences which we must assume exist between the process of voting and that of trading in the market under conditions of competition.
Political competition appears to be much more restricted by its very nature than economic competition, particularly if the rules of the political game tend to create and maintain disequilibria rather than work in the opposite direction.
We must conclude that there is little sense in praising the simple majority rule as the best possible rule for the political game. There is much more sense in adopting several kinds of rules according to the ends we want to reach, e.g., adopting qualified majority rules when the issues at stake are rather important for each member of the community, or adopting the unanimity rule when the issue is absolutely vital for each of them. I believe almost all of these points have been brilliantly stressed in the recent analyses based on the economic approach.
But we must bear in mind that none of the rules adopted or adoptable in political decisions can produce a situation which is really similar to that of the market under conditions of competition. No vote trading could be sufficient to put each individual in the same situation as the operators who freely buy and sell goods and services in a competitive market.
When we consider law as legislation it can be clearly shown that the law and the market can in no way be considered similar from the point of view of the individual and his decisions.
In fact, the market process and the legislative process are inescapably at variance. While the market allows individuals to make free choices provided only that they are prepared to pay for them, legislation does not allow this.
What we should now ask and try to answer is: Can we make a more successful comparison between the market and nonlegislative forms of law?
[1 ] See Bruno Leoni, “Political Decisions and Majority Rule,” Il Politico, Vol. XXV, No. 4, 1960, pp. 724-733.
[2 ] Anthony Downs, In Defense of Majority Voting (Chicago: University of Chicago, 1960). (A mimeographed essay, it was written as a general critique of a paper by Gordon Tullock, “Some Problems of Majority Voting.” This latter paper was an early version of Chapter 10 in The Calculus of Consent.)
[3 ]The Writings of Thomas Jefferson, vol. 15, Editor in Chief, Andrew A. Lipscomb (Washington, D.C.: The Thomas Jefferson Memorial Association of the United States, 1904), p. 127.
[4 ] See “The Great Political Superstition,” The Man Versus the State, Herbert Spencer (Indianapolis: Liberty Fund, Inc., 1981), p. 129.
[5 ] A. Lawrence Lowell, Public Opinion and Popular Government (New York: Longmans, Green, & Co., 1913).
[6 ] Frederic Bastiat, Selected Essays on Political Economy (New York: D. Van Nostrand Co., 1964), p. 144.