Choosing and Electoral System: The Primacy of the Particular

A little more than three generations ago, the eminent British political scientist Sir Henry Maine discussed what he referred to as "that extreme form of popular government which is called democracy." He wrote not as an opponent of the form but as an inquirer into its success, which seemed to him "to have arisen rather from skillfully applying the curb to popular impulses than from giving them the rein."¹ This view, which today may appear unfashionably paternalistic, can be redeemed by stressing the word "impulses." Few contemporary governments would regard it as either sensible or necessary to respond to each and every fluctuation of public opinion in between the settled contests that now determine and legitimate periods of rule by competing sets of politicians. But the question of how these contests should be staged, in terms of the method for translating votes into legislative seats, still leaves much room for argument.

That it is a topical argument is as evident in long-established democratic polities like France and Britain as in the emerging democracies of East and Central Europe, or indeed in the moves toward multiparty politics that have recently occurred in many states which date their independence to the post-1945 collapse of European colonialism.

In the recent debate in the Journal of Democracy among Arend Lijphart, Guy Lardeyret, and Quentin L. Quade that followed Lijphart's original article, two opposing positions were presented.² Lijphart advocated strict proportionality in the allocation of popular representation. His two critics catalogued the perils of such close attention to the unmediated arithmetic of party support and endorsed plurality elections. The arguments of the protagonists, however, implied an uncomfortable absolutism, for all rested on the presupposition that a particular electoral system can be advanced as universally superior to all alternatives. This notion had already been challenged in an earlier essay in the Journal of Democracy by Larry Diamond. Writing on "Three Paradoxes of Democracy," Diamond had drawn attention to the problem of "representativeness versus governability" and contended that "each country must find its own way of resolving this universal tension."³

My purpose is to expand upon that sage contention, which I shall seek to do in three ways: first, by questioning the starkness with which the alternatives were earlier presented; second, by resisting the amputation of electoral systems from the whole body politic; and third, by considering some examples that may suggest greater subtlety in value judgments than attachment to general prescriptions easily allows.

A Range of Formulas

In the first place, the choice does not lie tout court between plurality and proportional representation (PR). There is a range of formulas that is so elaborate as to make each set of national arrangements virtually sui generis.⁴ When, for example, we consider systems that are not based upon mere plurality (and note that plurality systems differ in important ways), we confront provisions as distinct as the additional-member system in Germany, the alternative vote in Australia, and the single transferable vote in Ireland.⁵ Furthermore, even where proportionality is embraced as a goal, actual practices extend from Portugal's regional system, with its large variations in the size of constituencies, to the nationwide PR system found in the Netherlands. There is therefore a real danger of setting up a debate that is grossly reductionist in its essential terms. But that is only part of the problem. Lijphart can certainly be commended for his broad-brush search for objective measurements of the various outcomes of different systems. This is a vastly different undertaking, however, from endeavoring to decide which of the two highly generalized alternatives is the ultimate answer.

This is very clearly brought out by the efforts of each of the three debaters to press his case. Two difficulties here seem insuperable. The first—and this is something that the interlocutors themselves largely concede—is the impossibility of separating out the influence of electoral systems from all the other forces that can affect a polity over time. The second is that electoral systems cannot simply be pulled out of a drawer in the way that one...