As noted here a couple of weeks ago, it’s debateable whether Ireland complies with the requirement in EU law that MEPs are to be elected using a system of proportional representation (PR), be that a list system or PR-STV. At present 25 countries use some kind of list system, while Ireland and Malta employ PR-STV.

But for a system of election to be a genuine case of proportional representation, two things are required. First, the formula: both lists and PR-STV meet this criterion. Second, an adequate district magnitude (the number of seats per constituency). It’s on this second criterion that Ireland’s current method of electing MEPs is open to question.

In 23 of the 27 EU member states, MEPs are elected from one nationwide constituency, and in one of the exceptions (Italy), average district magnitude, at 15, is greater than Ireland’s entire delegation. Only Belgium (average district magnitude 7.3) and Poland (average 4.1) join Ireland in dividing the country into small-magnitude constituencies.

Everyone can see that if a country were divided into constituencies each returning only 2 MEPs, or in the extreme case only 1 MEP, then the overall outcome is likely to fall a long way short of proportionality, whatever method is used to fill the seat(s) within each constituency. Those methods would not be genuine cases of proportional representation. Ireland’s district magnitude is unusually small for a PR system at both Dáil elections (around 4 seats per constituency on average) and at EP elections (average 4.67 seats per constituency).

At Dáil elections, with around 40 constituencies, there is at least a reasonable chance that ‘swings and roundabouts’ will operate: parties will be over-represented in some constituencies and under-represented in others, usually emerging with roughly their ‘fair’ share of seats (though the abolition of 3-seaters in particular would be a positive development from this point of view). At EP elections, though, there are only three constituencies, and thus a much greater likelihood that anomalies in one constituency will be amplified rather than rectified by the result in the others.

If we look at overall disproportionality at recent elections, it’s clear that this is what happens. Ireland’s EP elections produce outcomes that are pretty much as disproportional as elections in the UK held under first-past-the-post, which is definitely not a system of proportional representation, and much more disproportional than Ireland’s general elections. The average figures across the last five elections (details at https://www.tcd.ie/Political_Science/about/people/michael_gallagher/EPElection2024.php) are:

    Ireland general elections 2002–20: 5.80

    UK general elections 2005–19: 13.03

    Ireland EP elections 2004–24: 12.39

At the recent EP election, FG and Ff combined received 41 per cent of the votes but took 57 per cent of the seats, and while the parties with the fifth, sixth and seventh most votes didn’t win a seat, the party with the eighth highest number did.

The counter-factual is hard to be certain about since some people may have voted differently under a different electoral system, and parties / candidates that stood in only one or two constituencies would have been able to win votes everywhere if the entire country was one nationwide constituency. What we can say is that if seats had been allocated according to first preference votes by the Sainte-Laguë method (a very fair PR formula since it’s even-handed between large parties and small parties), the outcome would have been FG 3, FF 3, SF 2, II 1, Green 1, I4C 1, Aontú 1, Labour 1, independents 1. So FF, FG and independents would each have received 1 seat fewer than they actually did, while the Greens, I4C and Aontú would each have received 1 more.

How serious is this? For the EU, it’s clearly not as big a problem as when the UK was a member and, for the first four EP elections, used first-past-the-post to elect its MEPs, so in 1979 the Conservatives won 60 seats and in 1994 Labour won 62, giving those parties a strength in the EP out of all proportion to their voting support. The incoming Blair government prudently moved to a PR system prior to the 1999 elections. In Ireland’s case the adjustments would be marginal: some parties up by a seat or at most two, others down by the same amount.

However, even a small number of seats can make a difference. Ursula von der Leyen was elected Commission President in 2019 with only 9 votes to spare, and votes on the Nature Restoration Law in 2023 and the migration pact earlier this year were very close.

And there’s an obvious loss to Ireland in being unrepresented in EU groupings: no Irish Green MEP 2024–29, no Labour MEP in the Socialist group 2014–19, even though both parties would have qualified for a seat under nationwide PR. So, even if the Commission isn’t motivated to act on its own initiative, there’s always the prospect that one of Ireland’s under-represented parties might take a case to the effect that the current method of election does not meet the proportionality requirement.

The merits of PR-STV are not to be dismissed: it maximises choice for the voter and the counting process has an educative value in itself as well as delivering fair and transparent outcomes within each constituency. But given the profusion of candidates, and the reasonable normative objections to taking a more restrictive approach to candidacy requirements, it’s hard to see it as feasible in, say, two 7-member constituencies, let alone a nationwide 14-seat constituency. The widespread use of open-list PR across the EU to elect MEPs suggests that this approach is hardly a violation of democratic rights, and it’s an option that should be given serious consideration in Ireland.