A popular check on the Commons (3)

(b) The direct-democratic check

A principle driving the encroachment on representative institutions by political parties seems to be that representative democracy breaks down when the voters want to make decisions directly, as they nowadays do in the case of the choice of one or other party to form the government. The people will then elect not representatives of independent views, but delegates pledged to particular parties; and efforts to thwart this electoral strategy tend to be not only futile but illegitimate.

As suggested in the previous post, it is plausible that severing the electoral accountability of representatives, by limiting them to single terms of office, would produce a degree of independence from both parties and constituents on the part of the representatives. These might pledge to support specific parties only so long as they retained the vaguely defined support of the local constituents; and such a pledge would be sufficiently nebulous that it would leave much to the personal discretion of the representative. This may seem like a somewhat underhanded way of circumventing the tendency of voters to prefer delegates over representatives.

The other possible approach is not to elect representatives at all, but to embrace (in this branch of government) some form of direct democracy, resembling that of the ancient city-states. The popular will is ascertained either by a referendum in which the whole electorate can participate, or by establishing a randomly selected political jury as a representative body. Both the ordinary voter and the political juror, being unelected, will have no occasion to pledge themselves to any party as a condition of their election, but will remain free to change their party allegiances in accordance with their personal assessments of the changing relative merits of the parties. (Again, it may also be necessary to take measures to prevent parties from rewarding individual voters or jurors in return for their support. The simplest such measure would be to use secret balloting in both referenda and jury votes.)

The most obvious way of implementing a popular check on the Commons would be, on the executive side, to change the constitution so that a new general election had to be called whenever a certain supermajority of the electorate so decided in a referendum. As a legislative check, one might additionally establish a Swiss-style system of popular referenda on controversial items of legislation. As already discussed, periodic general elections should continue to be held as well, but perhaps at somewhat longer intervals. Since we presumably do not want to hold referenda on the continuation of the government very often, it would be reasonable, at any given time, first to gauge whether there was a significant chance that a referendum would in fact remove the government. This would ideally be done through frequent opinion polling conducted by some impartial agency. Alternatively, one might require a certain number of voters' signatures on a petition to hold a general election; this would resemble the referendum or "initiative" procedures of Switzerland and some American states.

However, the referendum seems to be an unnecessarily radical method for instituting a popular check; political juries would represent a less radical approach, despite their greater novelty. The establishment of political juries would be an intermediate measure, between reliance on an elected representative assembly and on the entire mass of voters, which should preserve some of the advantages of representation while sacrificing very little in faithfulness to the popular interest. This is because a jury can be made relatively compact, and thereby to resemble the elected assembly in its concentration of responsibility. The members of a relatively small jury, unlike those of a multi-million-person electorate, each have some significance in determining the jury's decision. In return for the exceptional power they have been delegated, it is reasonable to expect the jurors—much like jurors in criminal trials—to devote considerable time and effort to arriving at their verdict; for example, the members of the jury could reasonably be required to attend its relevant debates or hearings before voting on a question. The jurors are also likely to be prepared to devote this energy willingly, not only out of a sense of public responsibility but also because their votes will no longer seem to have only infinitesimal significance.

Because a political jury would become somewhat more closely familiar than the electorate at large with the matters on which it was voting, its opinions would tend to differ from those of the public. But this should in no way discredit the decision of the jury in the public's eye. The public should be aware that the jury consisted of members who differed from themselves solely in their greater familiarity with the matters at hand. Any unexpected decision by the jury should therefore be accepted by the public as the result of that greater familiarity. In other words, the jury would be a representative body to which the public should be prepared to defer. Somewhat like the Commons in its original form, the political jury should thus be able to serve as a mediator between the government and the casually informed population, justifying government positions to the people at the same time as it subjected those positions to a well-informed popular check. Perhaps no other approach could accomplish this, in a society that had ceased to defer to any specific social group.

The basic drawback a jury has by comparison with a mass electorate is that statistical fluctuations associated with the random sampling procedure mean that a jury's interests will not precisely reflect those of the parent population (the electorate as a whole). This effect becomes more marked as the jury is made smaller; thus, there is a tradeoff between the greater responsibility of the small jury and its statistical unrepresentativeness. In a majoritarian political decision, it seems unlikely that a jury of much smaller than, say, a couple of thousand members, for which typical statistical fluctuations are around 2% of the votes, would be acceptable. (This is also a typical sampling size used in opinion polling, where a comparable tradeoff between accuracy and expense of the poll applies.) Such a large body might not be markedly more responsible than the electorate voting in a referendum—though its jurors should be more responsible than the respondents to an opinion poll, since the jury's opinion, unlike that of poll respondents, would take direct political effect. Perhaps the problem of the tradeoff between responsibility and representativeness could to some extent be addressed by adopting a two-stage procedure for political jury decisions: a preliminary decision might be made by a compact but statistically unrepresentative jury, but require ratification by a larger, more casually interested but more statistically representative one. I believe that something broadly resembling this two-stage approach was adopted in the ancient Greek democracies, though not necessarily for the same reasons.