Social choice theory, it seems, has always been developing independently of other research pursuits in the study of electoral systems. Although the Black-Arrow theorem has virtually deprived it of the very possibility of searching for the Holy Grail of universally best electoral procedure, and it was untouched by the Duvergerian revolution, the whole discipline nevertheless did not die out after the Second World War. Its day-to-day developments have consisted in constant innovations both in theory and in the criteria by which existing electoral procedures were evaluated, and in designing new procedures based on these evaluations. A few exceptions notwithstanding, these designs were not adopted in the real world. To some extent, this has led to a situation where employing the data social choice theorists produced meant making inferences only about the epistemological community of the theorists of social choice.

Peter Emerson, director of the de Borda Institute in Belfast, has written a book called Designing an All-Inclusive Democracy. Consensual Voting Procedures For Use in Parliaments, Councils and Committees which is a decisive attempt to change the situation. Emerson shows that electoral procedures have a crucial impact on the shape of democracy and, very often, on its survival. This problem is especially poignant in divided societies (whether the divisive issue is ethnicity, or something else), where great victories, but especially great losses, achieved through the effects of majoritarian voting procedures, have very grave consequences. They lead the losers, but inevitably also the winner, towards leaving the electoral arena and entering a battlefield, and only there, amongst all the atrocities and catastrophes, do they realise the inevitability of their common fate – to use the words of the great Czech theoretician of democracy Vladimír Čermák.

So far, so good – Emerson certainly is not the only one to point out the perils of majoritarianism in divided societies and the necessity to establish those societies’ political elites in a more consensual manner than through the exclusively majoritarian procedures. Emerson, however, influenced by the North Ireland experience, differs in one major aspect from the tradition established by A. Lijphart and further developed by, among others, A. Reynolds, D. Horowitz, T. Sisk or B. Reilly. In his conception, the consensual procedures of instituting the political elites are a necessary, but certainly not sufficient, condition of the consensual political process. If the politicians elected in this way continue to vote in the simple majority “aye-nay” fashion, the previous efforts might be completely wasted.

A clear adherent of a Rousseauian conception of democracy, Emerson suggests a set of electoral procedures which should facilitate the selection of the most consensual politicians and policies. According to Emerson, the will of the people can be best ascertained and realised in such an election where the electorate chooses from the least restricted set of alternatives possible, by choosing which candidates or political solutions they prefer more or less. From the family of electoral procedures, he selects three variants in his specific recommendations, based on Borda count (BC), modified Borda Count (MBC), quota Borda count (QBC) and matrix vote (MV), variants he discovered or co-discovered.

MBC is an electoral system that allows one to express not only preference for, but also indifference towards the choices presented, which is an important practical condition in cases where the choice is to be made among a large number of alternatives. The winning alternative is the one with the highest average preference score. QBC is a system employing BC in a multi-member constituency; the candidates who fulfil the quota are elected. Matrix vote is a special electoral system which further increases the electorate’s influence over the elected body, as where MBC is deployed, they decide not only who is elected, but also what function he or she will serve in the body. The three suggested electoral systems are accompanied by a range of examples which should not only document the mechanics of these systems, but also provide a foundation for a discussion of their effects, including comparisons with other electoral procedures.

In the second part of the book, M. Salles, H. Nurmi, C. Bell and P. Kearney with A. Tierney discuss the ability of these designs to fulfil the normative criteria of the social choice school. Unsurprisingly, and given the normative criteria set out by the designs’ author, they find both virtues and drawbacks (H. Nurmi, for example, points out the relatively majoritarian nature of QBC). However, they definitely agree with their editor in his preference for ordinal voting procedures over plurality voting. This is an opinion shared by this reviewer. I believe it is advantageous to employ Emerson’s designs wherever the voters’ best interest lies in ascertaining what support the alternatives suggested have in their midst. Unfortunately, even the comparatively rare practical experiences with ordinal type of voting (Nauru, Kiribati and, for the alternative vote, Fiji) show that they can be easily manipulated and are highly vulnerable to tactical voting. This is caused by the fact that BC and its modifications fail to satisfy the criterion of “independence of irrelevant alternatives” with the consequence that the consensus becomes a result that is uncertain at best. The possibility of tactical voting is even stronger when voting is held on political decisions, as the key operation consists in creating a set of alternatives presented to the voters. Is it reasonable to expect that the agenda setters will be sincere and honest in their activities and that the principle “garbage in – garbage out” will not be fulfilled through backwards induction? I am afraid not. In cases where matrix voting is employed, will the voters aim at a honest expression of their preferences for the issue being decided, or will they attempt to generate effects in another political arena through strategic manipulation? Would that not make the politics even less predictable than it has hitherto been? Only were the preferred electoral alternatives of the social choice theory to penetrate into the real world more often would we see whether the reality is as honest as Emerson’s designs.

In evaluating social theory as a collective endeavour, one characteristic cannot be overlooked: it requires a scholar’s passion and whole-heartedness for the object of his study. Whether we are talking about the work of 12^th century scholar Ramon Lull, whose discovery a few years ago was a true revelation, the dark history of the conflict between the gatekeeper Condorcet and the challenger Borda in 19^th century, or the mutual influences between the work of Duncan Black and Kenneth Arrow, which still remain to be entirely explained – the sense of urgency these scientists have invested into their designs is remarkable. The same can be said of the book Designing an All-Inclusive Democracy, which is commendable not only to those who believe there is a strong connection between particular electoral procedures and particular results.

 

Roman Chytílek

 

(P. Emerson (ed.): Designing an All-Inclusive Democracy. Consensual Voting Procedures For Use in Parliaments, Councils and Committees, Springer Verlag 2007)