d’Hondt system

Named after a Belgian lawyer and mathematician, the d’Hondt system is a form of proportional representa­tion widely used in continental Europe. Victor d’Hondt (1841-1901), professor of law at Ghent University, published his Système pratique et raisonné de la représen­tation proportionnelle in 1882. The system was originally devised to ensure an equitable distribution of parliamentary seats among Catholics and Liberals and between the different language communities in Belgium.

Today the d’Hondt system is used by 17 EU member states for their elections (by regional or national lists) to the European Parliament, including in the United Kingdom (except in Northern Ireland), France, Italy, Spain and Poland. It also applies to national elections in at least 13 member states, as well as in Japan and Israel. Equally, the d’Hondt system is employed within the European Parliament as a formula for distributing a fixed number of official posts among political groups of different numerical strengths. In some cases, it is also used to assign those posts between various delegations within the political groups.

The d’Hondt system uses a ‘highest average’ method of calculation: it requires the total number votes received by each party to be divided first by one, then by two, then by three, and so on as necessary. The resulting quotients are then ranked by size, with the order determining entitlement to the seats available. In practice, the calculation is normally used to establish not only the number of seats to which each party is entitled, but also the order in which they are assigned.

The example below shows how the d’Hondt system would work in a hypothetical election involving seven parties contesting eight seats and receiving a total of 1,000 votes:

A practical example of the d’Hondt system in operation is the election of the eight Members of the European Parliament (MEPs) for the London region in the June 2009 European elections.

The overall result in the region was as follows:

Conservative Party: 3 Members on 27.4 per cent of the vote;

Labour Party: 2 Members on 21.3 per cent of the vote;

Liberal Democrats: 1 Member on 13.7 per cent of the vote;

Green Party: 1 Member on 10.9 per cent of the vote;

UK Independence Party (UKIP): 1 Member on 10.8 per cent of the vote;

Others: No Members on (collectively) 12.5 per cent of the vote.

Like most proportional electoral systems, d’Hondt is still disproportional to the extent that it favours large political parties or groups over small ones: in London in 2009, the Conservative and Labour parties together received less than half the votes cast (48.7 per cent), but won five-eights (62.5 per cent) of the seats. This is a consequence both of the method of calculation and the limited number of elective posts available. The bigger the number of posts available for assignment, the more proportional the outcome. So in the larger South East region, where there were ten seats for election, the two biggest parties, this time the Conservatives and UKIP, together scored 53.6 of the vote and obtained 60 per cent of the seats.

To attenuate the problem of this continued, if modest, disproprtionality, some variants of d’Hondt allow political parties or groups to engage in apparentement (the formation of technical alliances) to pool their weight in the counting process, so that between them they get slightly more positions than otherwise. Another variant of the d’Hondt system, known as Sainte-Laguë, attempts to minimise the mathematical disproportionality by only dividing totals by odd numbers, rather than both odd and even numbers. Sainte-Laguë is used for European elections in Germany, Latvia and Sweden.

Whatever the remaining disproportionality of d’Hondt, the contrast with the operation of a majoritarian electoral system is striking. Had the outcome in London in 2009 (above) been determined on a ‘first past the post’ basis, the two main parties would almost certainly have won all eight seats, rather than five. Or consider the implications for the election of office-holders within parliaments, for example. If the allocation of internal positions in the European Parliament, such as committee chairmanships, were simply put to a straight ‘up-down’ vote, on a winner-take-all basis, most of the political groups would come out substantially worse off. The largest two, or possibly three, groups could simply com­bine to deny any positions to their smaller rivals. Such a winner-take-all allocation applies in effect in the assignment of committee chairmanships in both houses of the US Congress. Of course, the United States has a two-party, rather than multi-party system, but one factor preventing the fracturing of the US party system is precisely that it has winner-take-all arrangements both for elections to Congress and for the allocation of spoils therein.

Some commentators, notably the British political scientist, Simon Hix, have suggested that the European Parliament should replace the d’Hondt system with such a winner-take-all method for assigning its key internal posts (What’s Wrong with the European Union and How to Fix It, 2008). This would promote the bipolarisation of political competition within the institution, by encouraging the larger political groups to forge winning-coalitions with a minimum number of partners for this purpose. Such a move could in fact be undertaken without a rule change: the use of d’Hondt for the assignment of the Parliament’s 14 Vice-Presidents, and the chairmen and vice-chairmen of its 22 committees and sub-committees and 40 inter-parliamentary delegations – a total of 286 positions in all – is simply an informal convention between the groups and has no official standing.

The actual election of the Parliament’s Vice-Presidents, for example, normally takes place on the basis of a single slate put forward by all the groups together, with the same number of candidates as there are posts to be filled and the internal assignment between the groups made using d’Hondt. When the vote occurs in the plenary, the only choice being made is to determine the seniority of those elected. However, if an individual chooses to run as an independent, he or she can break the system, provided enough of their colleagues support them – as Edward McMillan-Scott proved in July 2009, when he displaced the official candidate of the new Conservative group, Michál Kiminski. The assignment of committee chairmanships is also made collectively by the groups, using the d’Hondt system: once a group receives such a post, it names a candidate who is then, by convention, elected unopposed by the committee at its first meeting.

Further reading: A. M. Carstairs, A Short History of Electoral Systems in Western Europe, 1980; David Farrell, Electoral Systems: A Comparative Introduction, 2011.

September 2012

Copyright: Anthony Teasdale, 2012

Citation: The Penguin Companion to European Union (2012), additional website entry