Why tactical voting still matters: Welsh polling and the voting system

Richard Wyn Jones

The prospect of change and, specifically, the end of Labour hegemony has already led to an unprecedented amount of polling ahead of the Senedd election.

We can expect more between now and the 7th of May. While some of the pollsters have substantial experience in Wales, others have yet to test their methods and assumptions against the results of a real-world Welsh election. No doubt this is contributing to some of the differences in the current polling.

But beyond the noise, there is in fact substantial consensus on some very important things.

First and foremost, two parties have pulled ahead of the pack. Whether Plaid Cymru and Reform are effectively in a dead heat or Plaid slightly ahead of Nigel Farage’s latest political vehicle, it is clear that they are the only two with a realistic chance of being the largest party in Wales.

Labour, for so long the dominant political force in Wales, is in third place. While some polling suggests that it’s merely very bad for the party, most pollsters concur that its position is calamitous. Indeed, if the latter are correct then Labour may even find itself in a tussle with the Conservatives and the Greens to secure that third position.

Yes indeed, how the mighty have fallen.

Meanwhile, the pollsters agree that the Liberal Democrats – the contemporary inheritors of the once dominant Welsh Liberal tradition – are trailing in sixth place.

Let’s assume that this is broadly the structure of political opinion as we approach polling day. What might this mean when translated through the new electoral system that has been adopted to elect Senedd members?

Let me make two inter-related points.

First, this system – and in particular the use of the d’Hondt system to apportion seats between the different parties – favours the largest parties. This is because in any given constituency it will require progressively fewer votes for a party to win each additional seat after winning the first. So not only does size matter – it is actively rewarded by d’Hondt.

Second, the threshold to win that first seat is high. Exactly how high will depend on the relative performance of each party in any given constituency. My colleague Jac Larner’s various election simulations suggest that it will vary from around 10% to perhaps 16%. But let’s say, on average, around 12.5%.

If the polls are broadly correct this means that, outside a few constituencies where their party has a realistic chance of reaching that threshold, there will be a lot of wasted Conservative, Green and Liberal Democrat votes.

If the more pessimistic prognoses for Labour support are correct there is also every prospect that a substantial number of that party’s votes will also be wasted.

Campaigning

Readers will know that I’m sceptical that campaigning makes more than a marginal difference, at least in devolved elections. But it should nonetheless be obvious from the preceding analysis that there is potentially a large prize at stake for Plaid Cymru and Reform as the current campaign reaches the final stages.

If Dan Thomas and his colleagues can manage to persuade sympathetic Conservative voters in those constituencies where Tory candidates have no chance of being elected to vote for Reform, then their party will maximise the benefits of the d’Hondt bonus for larger parties – potentially increasing its number of seats from 1 to 2, 2 to 3, or even 3 to 4 in each of the 16 Senedd constituencies.

But similarly, if Plaid Cymru can persuade Liberal Democrat, Green or Labour supporters in their parties’ ‘no hope’ constituencies to vote for Plaid if only to ‘stop Reform’ from being the largest party, then d’Hondt will do its thing and the party’s prospects will be strengthened still further. And the potentially great news for Plaid is that the pool of progressive voters is larger than the pool of more conservative voters in Wales.

Yes, this is a more proportional voting system, but given the structure of party support in Wales and the specific details of the operation of the new electoral system, the incentives for tactical voting are still very much in place. That is, if the electorate can recognise them.

Professor Richard Wyn Jones is Director of the Wales Governance Centre at Cardiff University