Introduction The deficiencies of a "pure democracy" are well known: decisions may be made on the spur of the moment in the heat of passion, there is no real limit to the power of the majority, ordinary people with no intrinsic interest in politics, are burdened with the necessity of spending excessive amounts of time to informing themselves about a variety of issues and to actually participating in the meetings of the assembly. It is often argued that republics are better in several of these respects to pure democracy, and that a constitutional republic is better yet. But these claims for the superiority of republics over democracies either imply a doctrine that is untrue (i.e., that representative governments actually reflects the considered opinions of the electorate) or conceal a doctrine that is acceptable to few persons these days (e.g., that the "good of the society" as defined by some elite, not the opinions of the electorate, should be the measure of "good government").
The following is a brief sketch of some well known doctrines and analyses of "political theory" and "public choice" theory developed over the last 150 years. Some of these analyses (e.g., "Proudhon's Paradox") were well known as early as the mid 19th Century, and shaped the selection of electorial rules in the countries that were writing constitutions in that period. Other of these conclusions are fairly new, dating to the 1960s through the 1980s.
The detailed conclusions of this survey are left to the concluding section of this sketch, but the overall conclusion is that there is no necessary or even likely relationship between the "collective choices" generated by a representative system and the collective opinions of the electorate. Thus, those who have traditionally drawn a strong distinction between direct democracy and republican or representative systems have been right to do so, but for reasons not generally appreciated.
What Does "Political Representation" Mean?
Before considering the question of whether, or to what extent, a representative government represents the "will of the people" we must first determine an "ordinary" and acceptable meaning for the terms "represents" or "representation."
Legally, the terms "represents" or "representation" have been associated with the concept of agency. An agent represents his principal within a prespecified scope of delegated responsibilities or duties. The agent has a fiduciary duty [a duty of loyalty and care] to carry out the wishes of his principal, and the agent can be sued if he fails to scrupulously observe and use his best efforts to effectuate his principals instructions. The agency relationship is a "personal" relationship between the agent and his principal, and the agent, thus, cannot assign his duties to represent his principal to another person.
The following sections will examine why the political representative is not and cannot be an agent of his constitutents. Indeed, the term "representative" has only honorific significance, since there is no reason to believe that a representative typically represents the wishes and "instructions" of any of his constituents, and there is certainly no legal obligation for him to do so. The differences between political representation and the law of private agency are, of course, clearly recoginized in prevailing law. Although the private agent may be sued for not scrupulously following the instructions of his principal, there is no legally enforceable duty for the political representative to in fact carry out the instructions or effectuate the wishes of any of his constituents. While there have been occasional suggestions that politicians can be transformed into quasi-agents by requiring them to post "preformance bonds," in fact, no such scheme has ever been adopted anywhere, and, as a practical matter, would be very difficult, if not impossible, to implement.
Who Should Representatives Represent? Presumably, the "electorial rule" in any representative system must be something less than unanimity. If a representative can only be elected by a 100% supermajority vote of the electors in a district, then there would appear to be little "gain" from having a representative system. If unanimity can be achieved, the voting district can as well be "represented" by sending to the legislature its proxy on a specified set of issues. In any kind of political representative system, known or imaginable, a representative will not be elected by all of the electors, but by a plurality, a majority or a supermajority of voters [less than unanimity].
If the representative is the agent of those of his constitutents who voted for him, he cannot "represent" those constituents who voted against him (for they have explicitly chosen that he shall not be their agent). In an
"absolute district system" [see definition and discussion below] the representative is not elected by the minority, and, hence, cannot (or should not) be their agent.
All representative systems must, of necessity, be systems in which the representative will vote on more than one issue during his term of office. Otherwise, the voting district could simply vote on the one issue in question, and send the results of this vote to the "legislature" rather than voting on a "representative". But in a representative system where the electors elect a representative on his stance on a an agenda of issues, rather than just one issue, the agent status of the representative is inherently in question, for the victorious majority as well as for the defeated minority. Since there are a number of issues to be voted on during a candidate's term of office, the typical elector votes for or against a candidate on his announced or supposed stance on more than one issue.
Since there are only 2 or 3 candidates for each office in a typical election, it is unlikely that there will be a perfect match between the announced positions [if any] of any candidate and the preferences of any given elector on the agenda of issues that will be voted on during the legislator's term of office. Each elector, therefore, ends up voting for a candidate who, even if legally bound to carry out his campaign promises, would not vote according to the elector's wishes on every issue [even assuming that the elector's preferred candidate is elected].
The above analysis assumes, of course, that candidates and electors know the issues that will be voted on in the legislature during the legislator's term of office and that the candidates in each election openly announce the "positions" that they will vote in the legislature. It also assumes that a candidate's announced "positions" will not change if the candidate is elected. However, since most candidates in an absolute district system loose elections not because of the unpopularity of their announced positions, but because some "interest group" has waged a campaign of vilification against him, the sophisticated candidate will take as few clear positions as possible, so as to offend as few interest groups as possible.
Unrepresentative Voting Systems
The typical electorial system in the United States is an "absolute district representative system." In an absolute district system the candidate with a majority or plurality of votes in an election represents the district in question, and those candidates with a lesser number of votes [and those that voted for them] "take nothing." In the mid-1800s Pierre Joseph Proudhon presented a "paradox of voting" that illustrates the "unrepresentative character" of absolute district representative systems. Suppose we have a "simple case" where electors elect representatives on only one issue, representatives fully disclose their "positions" on that one issue and the representative is legally bound to vote as he has "advertised" he will vote. Assume that there are three electorial districts, with 100 electors in each district. In District Number 1 and District 2 55 voters vote for the "yes" position and 45 voters vote for the "no" position. In District Number 3 30 voters vote for the "yes" position and 70 voters vote for the "no" position. In society as a whole, 140 voters have voted for the "yes" position and 160 voters have voted for the "no" position, yet in the legislature two representatives vote "yes" and one representative votes "no".
Some of the unrepresentative features of absolute district representative systems can be somewhat ameliorated by adoption of a proportional representative system. In an absolute district representative system "winners take all." The candidate with the most votes wins the office in question and candidates with a lesser number of votes take nothing. In a proportional representative system electors vote for parties not persons. If a party gets any significant proportion of the vote in a district, then that party's "voting strength" in the legislature is accordingly increased. Hence, if Party X gets 40% of the vote in each district it gets no seats in the legislature in an absolute district legislature and 40% of the seats in the legislature in a proportional representative system. [In fact no proportional representative system works quite in this way, but are some variation on this "pure model".] Since "minority views" can more easily gain representation in a proportional representative system, the legislatues in such systems tend to be more "ideological" and "fractionalized" than those in absolute district systems. [Bad characteristics according to those that seek the "national unity" of everyone marching in lockstep.]
Optimal Campaign Strategy
As mentioned above, the decisions made by a representative once elected to the legislature may have
little relationship to the campaign promises made by the representative-as-candidate trying to get elected. This is not because legislators are "born liars." It is, rather, because, in an absolute district representative system, it pays the candidate who is seriously seeking to become a legislator to promise everyone everything when he is running for office. By doing so he is "appealing" to as many people as possible, and because there is no way that the candidate can be legally held to his promises once elected. Even that unfortunate representative who blatently lies during the campaign and is "caught in his lie" through unwanted publicity regarding his "inconsistencies" may gain the votes of many of those who he has "betrayed" in the next election. Many of the "betrayed" electors will have developed other concerns by the next election, and will decide that their betrayer is, once again, the "lesser of evils," despite his past "indiscretions."
Who Does The Legislature Represent ?
Legislators are not angels, nor are they lacking in personal interest. Like most human beings, legislators will do what they can get away with doing to enhance their incomes and make their lives better off. The only difference between the average legislator and the average businessman is the types of acts that are rewarded in politics versus business (the difference in the frameworks in which choices are made).
As mentioned above, the decisions made by representatives in the legislature may have little relationship to the campaign promises made by the representative when he was trying to get elected. To get elected the representative must promise everyone everything. To extract the maximum rents from the office he was elected to the representative will want to prove his influence on those few issues of importance to those who can be expected to increase his "campaign" coffers and/or who may provide him with employment should he loose a future election. Hence, the typical representative will trade off votes on issues that are of little "importance" (in the above sense) for additional support for the "correct position" on those issues that are of greater "importance" (in the above sense).
Setting The Agenda In The Legislature
The above does not consider the unequal distribution of power between legislators that is implicit in the seniority rules and committee system existing in virtually every legislature. The power to "set the agenda" for the legislative session, to table bills or move bills rapidly through committee, or even to set the order in which legislation is considered is often tautomount to the power to make or break legislation.
Are there any "reforms" of the representative system that might ameliorate this unequal distribution of power that is implicit in the committee system? How about a system in which legislators don't merely vote "ya" or "ney" on bills brought to the floor, but express their preferences, though "rank voting," of all bills introduced in a given session. For simplicity, let us also assume for a moment, that legislators don't make up their public policy preferences as they go, but that they have pre-established and unchanging preferences when elected. Does this system assure "rational public decisionmaking," in the seemingly innoxious sense that majority voting in the legislature will result in a consistent public policy agenda? No, it doesn't. As an example: Assume a legislative agenda of three bills or proposals, that are to be "voted on" through ranking by each legislator. Assume a legislature of three legislators. Legislator 1 prefers proposal A to B and he prefers proposal B to C. Legislator 2 prefers B to C and C to A. Legislator 3 prefers C to A and A to B. In a majority voting scheme, A wins over B [since two out of three legislators prefer A to B]and B wins over C, but C also wins over A. Hence, the "social agenda" derived through ranking and majority rule decisionmaking in the legislature is a logically inconsistent social agenda. Conclusion The above points imply that the result of an election of representatives is to elect rulers not agents. The ancient Greeks and Romans understood the concept of the "tyrant" who was elected during an emergency to exercise his unfettered discretion for a discrete period of time. The modern republican systems have transformed this temporary expediency of the ancient world into the customary political institution of the modern world. The representative system is not "corrupt" because it is filled with "bad men" but because it inherently allows wide if not unlimited discretion to those who hold public office. The motto of the representative is "You must have power to do good,." and power is exactly what he is given.
The fundamental fallacy of the contemporary analysis of representative government is that representative
government will, or is suppose to, "represent the people." The only reason that classical liberals generally preferred "representative democracy" or republicanism to monarchy was that it was possible to change rulers without the social disruption of revolution. The notion that the legislature was inherently a repository of liberty, since "the peoples' representatives" would never tyrannize the people was abandoned by all legitimate liberals who were attentive to the history of the French Revolution and the American Civil War well before the close of the early 19th Century. Representative government was the lesser of evils, not a good in itself.
It is time to reevaluate the traditional liberal commitment to "representative democracy" as the best of the available forms of government. The technology existing today, which is progressively more accessible to "the masses" as well as "select elites," makes direct democracy or even unanimity more feasible than it has been since the days of the city state. Since democracy is inherently devisive, it is to be expected that a commitment to democracy will also be associated with a rejection of large scale political institutions such as the nationstate. Wereas representative government makes feasible the largescale tyrannies of the 20th Century,. pure democracy will tear apart such structures in favor of the self rule of groups of people with common values and ideals. Democracy is not freedom, but it is one of the several tools that may enhance the chances for the triumph of freedom. .