Proportional (mis-)representation?

"...it’s the beauty of politics that predicted outcomes are often confounded..."

The surge is support for the Liberal Democrats has prompted much talk about proportional representation. One of the received wisdoms about PR, one which has been stressed especially by Cameron’s Conservatives, is that it leads automatically to coalition and therefore weak governments.

Leaving aside the question of whether coalition governments are necessarily weak (there is plenty of evidence from our European partners that they do not have to be), is it true that PR always leads to coalitions?

It might be instructive to look at the example of Spain. There, the architects of the post-Franco democratic constitution deliberately established a PR-based system in the very hope that it would create coalition governments.

Their aim was to ensure there was no repetition of Spain’s previous experience of democracy under the Second Republic in the 1930s, when the President was able to intervene in politics in such a way as to undermine the elected premier.

So, the 1978 Constitution under a restored monarchy gave the prime minister very extensive political powers to ensure political stability, but then sought to temper those powers by designing an electoral system which would ensure the need to build politically inclusive coalitions.

As so often, the best laid plans went awry. What has been Spain’s experience of coalition government? In practice, it has been non-existent: since the return of democracy, there has not been a single national-level coalition government in over thirty years of democracy.

To be sure, there have been several minority administrations, which have depended for survival on a series of deals with minor (usually regional nationalist) parties – but no formal coalitions.

Indeed, four of Spain’s ten administrations since 1977 have enjoyed absolute majorities, and only three have relied on any formal support from other parties. Moreover, there has been remarkable political stability at national level, with just five premiers since 1977: Adolfo Suárez (1977-81), Leopoldo Calvo Sotelo (1981-82), Felipe González (1982-96), José María Aznar (1996-2004) and José Luis Rodríguez Zapatero (since 2004).

How do we explain that? In part, it is down to the particular form of proportional representation chosen. Spain, in common with several European democracies (and as also used in European parliamentary elections), operates the d’Hondt system – a somewhat complex list system named after Victor d’Hondt, the Belgian mathematician and lawyer. Under d’Hondt, seats are awarded one at a time according to the highest average, obtained by dividing the number of votes by the number of seats plus one in a given constituency until all seats are allocated. (There’s an online guide to calculating d’Hondt outcomes).

But the real key to the explanation lies in the way that electoral boundaries were drawn. In Spain, a crucial element was to draw boundaries in such a way as to over-represent rural votes, in the hope of countering the expected left-leaning urban vote and thereby favour the centre-right. In technical terms, there is a high ‘index of disproportionality’ in Spain, reflected in the fact that it takes far more votes to elect a deputy in Madrid than it does, for example, in Soria.

What happened in practice was that the Socialist party increasingly picked up rural votes, leading to their crushing electoral majority in 1982, which ushered in fourteen years in power (and four consecutive elections wins).

The key point here, though, is that proportional systems vary in their degree of proportionality. It’s a question of design: there are many different possible systems out there, and how proportional they are varies widely.

It may be that the particular system favoured by the Liberal Democrats – Single Transferable Vote – would encourage coalitions, but there is no law that says PR necessarily leads to coalitions.

What’s more, once it’s in place, people’s voting behaviour may well change. But it’s the beauty of politics that predicted outcomes are often confounded.

Professor Paul M Heywood