Nominations for election to the European Parliament closed on Thursday. We now know which parties will be contesting the elections (if they happen), and who those parties have selected to stand for them in each region of the UK. The Unit’s Alan Renwick offers a brief guide to how the elections will work and what we can expect to learn from them.
With little sign of progress in the Brexit talks between the government and the Labour Party, UK participation in next month’s European Parliament elections looks increasingly likely. The parties have nominated their candidates and begun to launch their campaigns. Much is being said about how the electoral system will shape the outcome, but not all of it is accurate. This post provides a quick guide to the key points and reaches two main conclusions. First, the system will disadvantage small parties: in particular, the anti-Brexit parties will be punished for their disunity. Second, anyone wanting to read the results as a proxy second Brexit referendum will need to do so with great care.
1. The system is proportional…
The UK uses a system of proportional representation (PR) for European Parliament elections. To be precise, it uses two different systems. England, Scotland, and Wales use a list-based form of PR, which was introduced to replace the old First Past the Post system in 1999. This is based on 11 regions, each electing between three and ten MEPs. Each party puts up a list of candidates and voters choose one party’s list. The seats are allocated to the parties in each region in proportion to the votes that they have won.
Northern Ireland, by contrast, has used the Single Transferable Vote (STV) form of PR ever since the first elections to the European Parliament in 1979. Each party again puts up a slate of candidates. But voters rank individual candidates in order of preference, and these votes are counted and transferred according to the preferences expressed until the three seats available have been filled.
Proportional systems make it easier than under First Past the Post for small parties to secure seats. Last time around, for example, the Green Party won three seats with 7.87% of the vote, whereas in 1989, under First Past the Post, it famously captured 14.5% of the vote but no seats at all.
2. …but not all that proportional
‘Proportional’ systems vary in just how proportional they are. In fact, neither of the systems used in the UK is especially so, for two reasons. First, the number of seats available in each region constrains how far it is possible to allocate seats proportionally. In the North East of England, for example, where there are only three seats, it is clearly impossible for any more than three parties to win representation. Even the largest region – the South East, with ten seats – is quite small in seat terms, making it impossible to reflect the pattern of votes perfectly in the allocation of seats.
Second, the particular formula used to allocate the seats – the d’Hondt formula – tends to favour larger parties. Anyone wanting a mathematical proof of this can go here, but be warned that it is not for the faint-hearted. In essence, d’Hondt sets a higher bar than other formulas for a party to win a seat.
The combined effect of these two features is that, especially in the regions with fewer seats, small parties do not win representation. In the North East, no party has ever won a seat with less than 17% of the vote, and in 2014 the Conservatives were locked out despite reaching almost 18%. In Wales, where there are four seats, the Liberal Democrats have never elected an MEP, despite winning just over 10% of the vote in both 2004 and 2009. The effect can be seen even in the larger regions. In 2009, for example, the Greens won a 7.4% vote share in the North West but secured none of the eight seats.
Figure 1 shows, for each of the 11 electoral regions in Great Britain across the four elections to date using this voting system, the average vote share of the smallest party that did win seats (the upper number and the top of the bar) and of the largest party that did not win a seat (the lower number and the bottom of the bar). As you move from left to right and the number of seats available in the region rises, it gets easier for parties to secure representation. But it remains the case that the smallest parties are excluded.
The most recent YouGov poll puts four UK-wide parties – Change UK, the Greens, the Lib Dems, and UKIP – on between 6 and 10% of the vote. All of them risk falling foul of these effective thresholds in many regions.
The overall effects of these patterns are shown in Figure 2, which plots the nationwide vote shares of the top nine parties in Great Britain in 2014 against their shares of the seats. The diagonal line shows where the parties would have been had they won seats in exact proportion to their votes. Parties above this line won more seats than they were entitled to by strict proportionality, while parties below the line won less seats. The advantage enjoyed by the larger parties is clear. The three large parties – UKIP, Labour, and the Conservatives, each won a larger share of the seats than of votes. Among the smaller parties, this was true only of the SNP and Plaid Cymru, which benefited from the concentration of their votes in Scotland and Wales respectively. The Greens and Liberal Democrats, along with two very small parties, were significantly underrepresented.The oft-heard claim that the system does not discriminate against small parties is therefore false. It does. Given the polling evidence, this is particularly pertinent for the unambiguously anti-Brexit parties, which would have won more seats had they united behind a single list than they will running separately. The exact size of the effect is impossible to calculate. Simply to illustrate the basic maths, however, if we take the same YouGov poll as above and assume the same distribution of votes in every English region, we find that the Greens, Lib Dems, and Change UK win, respectively seven, five, and one of England’s seats running separately – giving them a total of 13 out of the 60 available seats. But they would garner 20 under a joint ticket. The Brexit Party and UKIP would also increase their seat haul if they stood together rather than separately.
3. Small parties cannot enter formal alliances
Many proportional electoral systems compensate for such small-party disadvantages by allowing multiple parties to submit joint lists of candidates. Some of those who have recently called for the UK’s anti-Brexit parties to coordinate their efforts for these elections seem to have had something like this in mind.
But the legal framework for European Parliament elections in Great Britain does not allow that: only a party that has registered with the Electoral Commission can submit a list. Even had they wanted to, it would never have been practicable for the various anti-Brexit parties to merge themselves in time for these elections – the deadline for nominating lists passed on 25 April, and parties had to have their applications to register processed and accepted before that. UKIP and the Brexit Party are clearly not about to go through a reconciliation, but had they wanted to join forces, they would have faced the same problem.
In this situation the only option for parties that wanted to avoid splitting the vote would therefore have been to agree that just one of them put up a list in each region. Anti-Brexit parties did coordinate in this way in a few constituencies in the 2017 general election: the Greens stood down their candidates in 22 constituencies, and the Lib Dems reciprocated in Brighton Pavilion. But, even with considerable planning, there were few such instances and local parties often resisted them. It would be a much bigger step for a party to stand down across an entire region of the country.
4. There are no transferable votes
British voters have some experience of electoral systems that allow them to express their first preference but also have a back-up option if that candidate is not successful. Mayoral elections in London and other English cities, Police and Crime Commissioner elections in most of the country, and local elections in Scotland all follow this principle. Such systems would allow small parties to coordinate while maintaining their distinct identities. That is the case for the STV system used in European Parliament elections in Northern Ireland.
But the rules for European Parliament elections in Great Britain prevent this. The voter can express just one preference for one party. If you vote for a party that wins no seats, you have no means of influencing the result any more. You cannot, say, switch your vote from Change UK to the Lib Dems if Change UK win no seats; nor can you switch from UKIP to the Brexit Party. Voters are therefore forced into tactical decisions about which parties are likely to be in contention for a seat in their region. In the absence of regional opinion polls, such tactical voting may well prove difficult.
5. You can’t vote for individual candidates
Most list-based PR systems allow voters to express preferences among the candidates of their favoured party (see my book, co-authored with Jean-Benoit Pilet, for details). But the system used for British elections to the European Parliament does not allow this. Rather, the lists are ‘closed’, meaning that candidates are elected in the order determined by their party. If a party wins, say, three seats in a particular region, then the top three candidates on its list are elected.
This means that voters cannot vote for individual candidates whom they particularly agree with or like. The election results will be interpreted in terms of how much support there is for each party and its official policies, irrespective of the positions that might be adopted by individual candidates. That might be particularly relevant for Labour and the Conservatives, whose candidates are most likely to deviate from their party’s Brexit line. Such candidates will not be able to make a separate policy pitch to voters as sometimes happens in other elections.
6. The franchise is not the same as for general elections or referendums
European Parliament elections use a modified version of the Westminster electoral franchise. Using the Westminster franchise means that the minimum voting age is 18 throughout the UK. The main difference from elections to the House of Commons – and from the 2016 referendum – is that EU citizens resident in the UK are entitled to vote (though, to do so, they must print out a separate form from the normal voter registration form and send it to their local electoral registration office). There is no official data on how many EU citizens are registered to vote here. About two million more people are registered for local elections than for Westminster elections – a difference of about 4% – and EU citizens comprise the main difference. But how many of these might register for the European Parliament elections is unknowable. So the electorate is likely to be slightly more pro-Remain than in a referendum, but by an unclear margin.
7. Will the election be a proxy second Brexit referendum?
Few people in the UK will be greatly interested in what the results of these elections mean for the balance of power within the European Parliament. Rather, the outcome will be interpreted primarily in terms of what it says about the balance of opinion on Brexit. Many will tout them as a proxy for another Brexit referendum.
The elections will provide some useful information on this. But it will fall well short of a referendum as an indicator of the public’s view. Among the parties with a clear Brexit position, the seat shares will not be a reliable indicator of Leave/Remain sentiment, for the reasons given above: votes will translate into seats rather haphazardly. Beyond seats, the combined vote shares for the strongly pro- and anti-Brexit parties will indicate voters’ Brexit preferences rather better. Meanwhile, however, inferring Brexit attitudes from support for Labour and the Conservatives would be hazardous. In addition, the franchise will likely be different from that of any actual referendum, and turnout will almost certainly be much lower.
All of this leads to two conclusions. First, despite the system’s broadly proportional character, the small parties will be disadvantaged. Most obviously, the anti-Brexit parties will be punished for their disunity. Many voters will have to make difficult tactical decisions.
Second, it will be important to avoid simplistic interpretations of the result. Which single party ‘wins’ (in terms of either seats or votes) will mean little. Aggregate votes across the blocs will signify more. But the meaning of some votes – particularly Labour and to a lesser extent Conservative votes – will be somewhat opaque. An election is not a referendum. Parties and voters are entitled to treat it as though it were, and may advance their cause by doing so. But those wanting to interpret what the results actually mean will need to do so with great care.
About the author
Alan Renwick is Deputy Director of the Constitution Unit. He is the co-author of Improving Discourse During Election and Referendum Campaigns and The Mechanics of a Further Referendum on Brexit. He also served as Research Director for the Independent Commission on Referendums.