Elaborating Eidelberg's election scheme

As far as I am aware Eidelberg has not spelled out the details of his proposed combination of senatorial nomination and popular election of presidential candidates, other than to suggest that the senate nominate no more than three candidates (in addition, perhaps, to the incumbent, if he is eligible to run again). It is interesting to try to work out what these details might reasonably be.

One important question would be what level of support within the senate would be needed to win such a nomination. For example, if each nomination was determined by a separate vote, with a simple majority of senators necessary to win a nomination, it seems plausible that the senate would frequently nominate only one candidate (including any eligible incumbent)—the candidate most favoured by a majority senate coalition. Obviously, in this event the people would have to have the option of rejecting that nominee: the popular aspect of the election would be a farce otherwise. Equally clearly, following such a rejection the senate would have to be required to nominate someone else, in order to avoid the prospect of a vacant presidency; and this process of senate nomination followed by popular rejection might conceivably be repeated several times. Thus the potential of a kind of "bargaining" between the senate and the people seems to be implicit in Eidelberg's scheme. The senate cannot nominate without regard for popular opinion, because the electorate can reject the senate's choice (at least if the senate confines itself to a single nominee).

Another plausible nomination procedure would be to permit minority senate coalitions to nominate presidential candidates. Intuitively speaking, this would tend to encourage the nomination of a larger number of candidates. For instance, suppose a nominee required the support of at least (25% plus 1) senators, each of whom would thereafter be prohibited from supporting any other nominee. The senate could then be divided into, at most, three groups, each with the right to nominate one candidate. Even if the groups are equal-sized, there are not quite enough senators to make up four such groups; if the groups are of markedly unequal sizes, it may be possible only to form one or two. This rule would therefore be the most obvious way of ensuring that the senate nominated no more than three candidates. (Incidentally, I think something resembling this method is currently used within at least one major British political party for nominating its leader by a vote of the party's Members of Parliament; the leader is then chosen from among the nominees by a ballot of ordinary party members. But in this case the 25% threshold for nominees is, if I remember correctly, replaced by a 5% one.)

Even if there are two or more nominees, it actually seems reasonable to continue to give the ordinary voters the right to reject all of them, just as in the case of a single nominee. For one thing, if this prerogative were not granted, the senate might nominate some utterly unsuitable and unpopular candidate, "B", along with a more realistic but still unpopular one, "A", as a means of forcing the people to approve of "A", who would be the only bona fide nominee. Perhaps this prospect is somewhat far-fetched; but in any event, giving the people the right to reject all nominees seems desirable from the point of view of constitutional legitimacy.

A presidential election with as many as three of four candidates, in which the voters retained the option of rejecting all the nominees, would be rather different from a familiar election where the idea was to identify the collectively preferred candidate regardless of whether the voters think any candidate is actually worth electing. Probably one should here ask each voter not only to put the candidates in rank order (as seems advisable in multi-candidate electoral procedures), but to state at what point, in his ranked list of candidates, the candidates become "acceptable" to him. All candidates might be "acceptable", or none might be; this would be entirely up to the voter. Then once the leading candidate had been identified in some conventional way, on the basis of the voters' ordered lists of candidates, it would be necessary further to determine whether that leading candidate was "acceptable" to a majority of the voters. If not, the senate would have to come up with a new set of nominees and the election be held again.

A simplified version of this procedure might be to ask each voter to state which candidates were "acceptable", without bothering with more detailed rankings. Again, the voter would be free to choose how many candidates were "acceptable" to him. The winner among the candidates would then be the one who was "acceptable" to the largest number of voters. (This is known as the "approval voting" procedure.) But for that winner actually to be elected, that number of voters would also have to exceed 50%.

One cannot help thinking here of the existing widespread dissatisfaction of voters with all political parties. If our electoral procedures encouraged voters to state whether they found the candidates acceptable in an "absolute" sense, rather than merely to rate the candidates against one another, one might see increasingly discouraging results. If our rules permitted the voters to reject all candidates, some elections might fail to elect anyone. As this would paralyse the existing political system, it would be a less than wise idea to permit the voters to indicate this sort of opinion; but the result is that the political system may remain illegitimate in the eyes of the voters for an indefinitely long time. On the other hand, where the nomination of candidates is in the hands of a compact body such as a senate, it is not unrealistic to allow voters to "reject the system" in this way. The senate would then quickly have to revise its criteria for nominating candidates, so that these were more acceptable to the public; but this should not be beyond the flexibility of such a body.