Electoral system for European Parliament elections
For European Parliament elections in Scotland, Wales and England a form of list proportional representation is used, called the d’Hondt system. (STV is used in Northern Ireland.)
Voters get a ballot paper with the names and logos of different parties and get to put one cross against one of those parties. There is no preferential (1, 2, 3 etc.) voting.
Next to each party name on the European election ballot papers is also the list of their candidates to be a Member of the European Parliament (MEP), in order from number one downwards. If the party gets one seat selected, their top name is elected, if two then their top two names, and so on.
Voters, therefore, get to see which candidates will be elected but they can’t change the order of those candidates. For example, if someone really dislikes the top name on a party’s list but likes the second name then they are still forced to choose between voting for that party or not. They can’t say ‘I want to vote for party X but I don’t want the top name and instead want to prefer someone lower on that list’.
This makes the Euro elections a ‘closed list’. Some versions of list PR allow voters to pick their preferred candidate from a party’s list too; this is called an ‘open list’ election but isn’t the system used in England, Scotland and Wales.
This matters as, for example, Labour’s candidates for the European Parliament election include a mix of Remain and Leave supporters. Voters only get to decide whether or not to vote Labour; they can’t choose between Leavers and Remainers in Labour’s lists of names.
(Independents can also stand on their own. Think of them as special lists of just one name.)
Constituencies for the European Parliament election
Those candidate lists are not simply one list for all of England, Scotland and Wales. Rather, there are separate lists for Scotland, Wales and each of nine regions in England. Votes are counted within each of those eleven areas and used to decide how many MEPs each list wins in each of those areas.
The Brave New Not Too Ideological Dawn Party, for example, might run only in those nine England regions and get two MEPs in one region, one in another and none at all in the rest.
Turning votes into seats: d’Hondt
The conversion of votes into seats in each of those eleven areas is done using the d’Hondt system. The idea is that the number of votes a list (party) gets in that area should determine how many seats it gets. If you get half the votes, you should get half the seats, etc.
This is very different from first past the post where, as the 2015 general election in Scotland starkly demonstrated, you can get nearly all the seats with less than half the votes.
The complication is that there are not enough seats to go around to make all the allocation that proportional. For example, Wales has four MEPs. Should a party that gets 20% of the vote get one MEP (25% of the total) or zero seats (0%)? Instinctively, you may think that one seat is better than zero given how close 20% is to 25%. But for elections, precise rules are needed.
Which is where d’Hondt comes in. Named after the Belgian lawyer who invented the system in the 19th century (pause to marvel that the electoral system under which Ukip has done best in the UK is one from a Belgian…), d’Hondt uses the ‘highest average’ method.
It works like this:
- See which party in the area has the largest number of voters. They get the first seat.
- Divide that party’s vote by one more than the number of seats it has won (two, in this case).
- Now look at all the party vote totals and whoever has the largest total gets the next seat.
- Divide that party’s vote by one more than the number of seats.
- Carry on doing this – give a seat to the largest, then divide it, then look again at who now has the largest – until all the seats are allocated.
For a numerical example with vote totals and all the calculations, see this one from the BBC.
I also quite like this rather different explanation from mathematician Helen J. Wilson:
Suppose you are in charge of allocating seats to parties. Once the votes have all been cast and counted, you are faced with a group of parties each of which has a certain number of votes.
You have a set of seats to allocate. The D’Hondt method is, in principle, very simple. You simply “sell” each seat to a party. Each seat “costs” the same number of votes; and each party buys as many seats as it can (and its leftover votes that are worth less than a seat, are discarded).
What is the practical implication of d’Hondt?
Because d’Hondt is used separately in each of Scotland, Wales and the nine English regions, and therefore to dole relatively small numbers of seats in each case (between four and ten), the results can end up quite a way off being proportional.
In particular, it is possible for quite a lot of votes to be ‘wasted’ by going to parties who do not quite reach the threshold for getting one seat. As a result, tactical voting to pump up the support of another party can make sense just as it does in first past the post. With d’Hondt there is, as it were, more than one post. The logic though that tactical voting can help one party past the post still applies. It’s just a bit more complicated to work out what tactical voting option might work best.
Or, to adapt Helen Wilson’s terminology from above, each seat may be sold for the same number of votes – and so parties who run up some votes but not enough to afford the price of even one seat end up with their votes wasted.
That means, for example, that Remain voters face the risk of their votes being split between different new or not so new parties – and so electing fewer Remain MEPs than if Remain voters vote tactically to maximise their impact.
Crucial also, though, is that point about the lists being closed lists. For parties putting up candidate lists containing people with a mix of views on Britain’s place in the European Union, such as Labour who have both those who want Britain to leave the EU and those who want it to remain on their candidate lists, voters are forced either to accept voting for a mixed group of candidates or to vote for another party/list instead.