The D'Hondt system of proportional representation is used in the European elections in England, Scotland and Wales, but not in Northern Ireland. It is also used in elections in many other countries. D'Hondt is described very well in Wikipedia, along with its variants and related systems,
My particular interest in the way D'Hondt is used in the UK is its impact on the elections to the European Paliament in 2019 and in 2014. On both occasions there was an over-representation of parties that were more interested in damaging the European Union than improving it.
The Electoral Reform Society has, as one would hope, the clearest description of the way the voting system works for elections to the European Parliament, and suggests that "The way we elect our MEPs in the UK is now much better than the system in place before 1999. But the closed list system is still far from perfect and limits voter choice."
"Closed list" is the system in which we, the voters, choose a party, and the party chooses the MEPs from a defined list. The Electoral Reform Society argues that a Single Transferable Vote ("STV") system would be better.
The Electoral Reform Society also notes that "Some countries [...] limit the size of their constituencies to create an effective threshold" to "stop parties with very low support getting a seat". This is a particular point that piqued my interest, especially since its (possibly unintended) side-effect was to give anti-European-Union parties an advantage.
The aim of the D'Hondt system is to distribute votes across a number of parties so that the number of candidates elected for each party represents, as closely as possible, the proportion of the votes for each party. In the extreme case where there is only one seat available (only one candidate can be elected) then D'Hondt is exactly the same as the "first-past-the-post" system that is used for UK Members of Parliament for each constituency, and and for US Representatives for each district. The candidate with the most votes wins the seat.
The more seats are available, the better the D'Hondt system distributes the votes among candidates. The UK has 73 MEPs (Members of the European Parliament). But the UK is divided into 12 Constituencies and the number of MEPs elected by each constituency varies from 3 to 10. The Northern Ireland constituency uses a Single Transferable Vote (STV) system and D'Hondt is used in all the others. In the constituencies with only 3 MEPs the allocation of votes is very limited and only the most popular parties are represented; in the constituency with 10 seats, significant minorities get a more fair representation.
When all the constituency MEPs are added up, the numbers of MEPs differs quite significantly from a truly proportional distribution. The charts below show the share of the vote and the number of MEPs for all the parties that won seats in the May 2019 European Parliament election. On the left are the actual results, using the constituency approach; on the right are the results we would have seen if the UK (except Northern Ireland) were treated as one constituency. The chart on the right shows that the number of MEPs matches the vote distribution much more faithfully.
| Actual Distribution of MEP Seats Applying D'Hondt to Each Constituency and Then Adding Up The Results | Distribution If The UK Were Treated As One Constituency and D'Hondt Applied To That |
|
|
|---|
In the 2014 European Parliament election it was a similar story. Again, the anti-European-Union party (UKIP then) was over-represented in the allocation of MEPs. Perhaps if the representation had been more accurate, David Cameron would not have felt so threatened and would not have been stupid enough to call the 2016 referendum. Who knows?
| Actual Distribution Of MEP Seats Applying D'Hondt to Each Constituency and Then Adding Up The Results | Distribution If The UK Were Treated As One Constituency and D'Hondt Applied To That |
|
|
|---|
_________________________________________________________________ More to follow.