Tactical voting isn’t over in Wales, it’s just different now

Eben Myrddin Muse

First past the post is dead in Wales. It was officially struck down – for national elections in Wales – by the Senedd Cymru (Members and Elections) Act, which received Royal Assent on June 24th 2024.

We are now living with a brand new, sixteen-super-constituency, Senedd of ninety-six members, with seats divvied up proportionally-ish via something called the D’Hondt method. 

The new system was established in an effort to make the Senedd more representative, without handing over too much power to smaller parties, or doing away with constituency based representation altogether. It’s a compromise model that ends up pleasing nobody; the Expert Panel on electoral reform recommended in 2017 that a Single Transferable Vote or Flexible List Proportional Representation system be introduced, these were effectively vetoed by Wales’ larger parties who preferred a closed list system for reasons that will become obvious.

Party control is given priority over voter choice with parties selecting a closed list of candidates for election, and larger parties hold a clear advantage; there is a de facto vote threshold (of around 12-15% depending on the size of the field) below which a smaller party will win no seats at all. In a parliament of 96 members this is decidedly unproportional. It is ironic that this sets up a shrunken Labour to be hoisted by its own petard, as they fall dangerously close to this cliff themselves, alongside the Welsh Conservatives.

The 7th of May will be the first definitive test of our new system; will it cut the mustard and help achieve a Senedd with better representation and more effective scrutiny? And what about the impact of tactical voting on all of this?

Tactical Voting Persists?

Tactical voting – defined by the electoral reform society as voting for a candidate or party you wouldn’t normally support, in order to prevent another from winning – is a hallmark of first past the post elections. A YouGov poll in 2024 showed that 22% of UK voters planned to cast their vote this way. When confronted by a wildly divisive party, as we currently are by the rise of Reform UK, tactical voting generally kicks into overdrive – as seen at the Caerphilly byelection in October last year, where huge numbers of voters saw Plaid as the only viable route to keeping Reform out. They duly saw an unheard of 27.4% swing in their favour, bringing them victory and denying Reform a dramatic victory. 

The new system was designed to spare people the need to hold their nose while voting – so that people could vote for the party they want to vote for, that best represents their views. Voting through gritted teeth is something many voters are intimately familiar with, myself included. It’s demoralising, depressing, undermines the democratic process, and drives low voter turnout.

Under true proportional representation there’s no need for tactical voting because in each constituency there are several winners – multiple seats distributed proportionally. But we don’t have true PR. 

Not quite. Tactical voting lives on, although it is transformed.

The D’Hondt Conundrum

The way the D’Hondt system works is that parties that win seats get their vote share progressively divided by a larger number for every seat they win. Six seats for every constituency means that the top few seats go to the largest parties and polls say that these will usually be Plaid or Reform – there is little debate about that. But the final make-up of the Senedd as a whole is actually incredibly difficult to accurately predict. Why?

The next Senedd and Welsh Government will be defined in its character and final composition by the winners of the final few seats on each list; as the larger parties’ vote shares are divided by greater numbers, in many constituencies their count will approach that of the smaller parties. If you have voted for Plaid, who may have won two seats, your vote when it comes to later rounds is now effectively weighed as 1/3rd of a vote against someone who voted Green or Labour in the first place when it comes to deciding who wins final seats. The same goes if you cast your vote for Reform UK.  As the rounds go on, votes for smaller parties retain more of their impact than votes for parties that have already secured seats.

In such cases, smaller parties are not just viable, they are often where a vote carries the most weight. This means that strangely, tactical voting under the new system is occasionally inverted; to gain the biggest advantage for a coalition in Government (either centre-left or conservative), it will, in many circumstances, make more sense to vote for a smaller party who could snatch the final seat, than it would to give your vote to be counted for less toward a larger party who are already out of the running for the sixth or fifth seat. This will also determine the kind of coalition government that is ultimately formed.

To show you how this might play out in practice, allow me to illustrate a hypothetical election scenario based in the beautiful and entirely fictional model constituency of Llanbedinodyn Abergofiant, full of made up election candidates who hold no intended resemblance whatsoever to real-life figures. This green and pleasant place is a bellwether; the way Llanbedinodyn Abergofiant goes, so goes the nation, in that its vote breakdown is identical to the recent national poll carried out by More in Common between February 15 – March 3 this year. Its population is 190,000, which just so happens to be the average population of a Senedd constituency. There was a slight uptick in turnout, to a respectable 55%, and I’ve just received word that the final count is in: 

• Plaid Cymru: 27,245
• Reform UK: 27,068
• Labour: 20,824
• Welsh Conservatives: 10,435
• Green Party: 10,310
• Liberal Democrats: 7,504
• Other: 1,113

There’s a pretty good chance that this ends up being a good approximation of how a few of Wales’ new superconstituencies will look – a dead heat between Plaid and Reform, a beleaguered Labour party lagging well behind, and a few smaller parties duking it out for the scraps. Swap one and two, and five and six, and it’s equally plausible. As a voter, how can I cast my vote in the way that is most likely to secure my preferred electoral outcome?

To know the answer, we can take these vote counts and run them through the new voting system machine to see what comes out. Who wins, who loses, and who was close enough for a few votes to have made a crucial difference?

Let’s get counting!

Round 1 (Votes)

Plaid Cymru: 27,245 (winner – 1 seat)
Reform UK: 27,068
Labour: 20,824
Welsh Conservatives: 10,435
Green Party: 10,310
Liberal Democrats: 7,504

The first round of counting feels nostalgic, as it’s just a first past the post election. The spoils simply go to the party with the most votes; Plaid Cymru, whose number-one candidate, hill-farmer ‘Iddon Ap Rheinallt’ is elected. Their vote count is therefore divided by their number of seats plus one, which is two. From here on out, we aren’t working with raw vote numbers – rather, rounds are won by the party with the highest quotient; think of it as a score that’s the result of dividing the vote numbers according to the D’Hondt Formula.

Round 2 (Quotients)

Plaid Cymru: 13,622.5 (1 seat)
Reform UK: 27,068 (winner – 1 seat)
Labour: 20,824
Welsh Conservatives: 10,435
Green Party: 10,310
Liberal Democrats: 7,504

Clear winners Reform UK are therefore awarded seat number two of six by a massive margin, and they follow Plaid in having their quotient halved for the next round. Congratulations to crypto trader ‘Jason Smythe’ for his election as Reform’s top dog.

Round 3 (Quotients)

Plaid Cymru: 13,622.5  (1 seat)
Reform UK: 13,534 (1 seat)
Labour: 20,824 (winner – 1 seat)
Welsh Conservatives: 10,435
Green Party: 10,310
Liberal Democrats: 7,504

Congratulations to Eluned Morgan’s Labour who manage to snatch the third seat relatively comfortably, seeing their quotient halved for the next round, as you will have probably come to expect by now. Former special advisor ‘Abigail Gridroe’ is elected as part of Welsh Labour’s fabled new generation.

Rounds 4–5 (Quotients)

Plaid Cymru: 13,622.5 (winner – now 2 seats)
Reform UK: 13,534 (winner – now 2 seats)
Labour: 10,412 (1 seat)
Welsh Conservatives: 10,435
Green Party: 10,310
Liberal Democrats: 7,504

Since the halved quotients of both Reform and Plaid are still comfortably ahead of Labour, the next seats are awarded to Plaid Cymru and Reform UK in that order, with the parties now seeing their original vote counts divided by a third in time for the sixth, and final round. Congratulations to ‘Lleucu Wyn Llwyd’ for Plaid, and ‘Ritchie Gruyère’ for Reform, who get a seat each.

Round 6 (Final Quotients)

Let’s check in on what we’ve observed. As clear frontrunners, Plaid and Reform were sure to win two seats apiece by a very wide margin indeed, and Labour, being well clear of the minnows, were able to secure one for themselves.

Plaid Cymru: 9,081.7 (2 seats)
Reform UK: 9,022.7 (2 seats)
Labour: 10,412 (1 seat)
Welsh Conservatives: 10,435 (winner – 1 seat)
Green Party: 10,310
Liberal Democrats: 7,504

Conservative councillor ‘Fliss Parker-Brownstone’ who definitely never considered defecting to Reform, is surprised to be elected as our final Senedd member.

As we can see, due to their relative success with the first few seats up for grabs, Plaid Cymru and Reform have seen their quotients gravely diminished by machinations of the D’Hondt formula, leaving them effectively out of the frame for the last seat.

With a quotient divided by three, additional votes they might have attained would only add one third to their quotient for the final round. That means that in order to beat out the Tories for the final seat, Plaid Cymru or Reform needed more than three thousand additional votes, which is quite a lot to ask.

Meanwhile, for Labour to snatch that second seat, they’d have required more than 23 additional quotient points, which equates to 47 extra votes. Quite achievable. The Greens, for whom one vote is still one ‘quotient point’ (since they have no seats), would have required only 125 extra votes to win their historic first ever seat in Llanbedinodyn Abergofiant.

This is the nature of the D’Hondt system – it leads to a closer fight as the rounds go on. It means that any party with the ability to draw between 10-15% has a fighting chance when it comes down to the wire. Knowing whether there’s a genuine chance for a smaller party to win the final seat(s) is key to knowing how to cast your vote most effectively.

Coalition? What coalition?

Why does this matter? Can’t I just vote for the best contender and hope for the best?

Well, sure. But it’s worth noting that no credible poll has predicted that any single party will win an outright majority (49 seats). The system was designed with coalition government in mind, which means that any party of government will need to govern in partnership, and the nature of that partnership will be decided by who wins the last few seats in each of the 16 constituencies. If the Greens, Labour, or Lib Dems pick up more seats on balance, then it will be more likely that a Plaid led partnership of some sort takes power. If the Tories manage to hoover enough up, we could be looking at some kind of Reform-Conservatives coalition.

For voters primarily motivated (as many Welsh voters are!) by stopping one party from dominating, it creates a counterintuitive reality: the most effective way to do so may be to hope for a strong performance from smaller parties who will take seats from the opposition and shape the coalition – not simply to back the biggest party.

Given that Plaid are frontrunners, it seems sensible to wonder about the flavour of coalition they will govern through? Will they be pushed to the left by a bloc of Greens? Will Labour retain some semblance of power? Will Jane Dodds somehow hold the balance of power once again? These are major questions with major implications.

This idea might sound strange at first – that voting a certain way grants you an outsized influence on electoral politics, but as an idea it’s not so different from the American notion of swing states, which is where every single US election is won or lost. Voters there are intimately familiar with the power of casting a vote in a close contest.

Wasted Votes? A Careful Choice.

For all the talk of casting a vote for smaller parties – there is, unfortunately, still such a thing as a wasted vote in the new system. Wales is no Llanbedinodyn Abergofiant, it will not follow the law of averages, and each constituency will be different.

Although smaller parties gain more vote efficiency as rounds go on, short of that magic 10-12% that puts them in the running for a seat, a vote that yields no representation can still be a wasted vote, as the system’s imperfections come into play. This could be true of the Conservatives in some constituencies as it could be for the Liberal Democrats, Labour, or Greens in others. In these circumstances, tactical voting in its traditional form does not disappear—it returns. How can we know which way to go?

Wales, being underpolled and completely unpolled at a constituency level, poses a challenge. In the absence of good data we have to rely on statistical projections based on previous elections and known demographics. It’s not an exact science but Cavendish Cymru have created a helpful map which shows you projected winners of the final seat along the lines of recent polls, but crucially also who is predicted to be the close loser. This is one way to see the choice that is available to you if you wish to cast your vote along the lines of the sixth seat in your constituency.

The sixth seat in Gwynedd Maldwyn (according to the January YouGov poll), for example, is fought between the Greens and Reform, with the Greens set to take it with a 1.84% margin. According to More in Common (Feb 2026) the final seat in Islwyn Maldwyn is likely to be fought between Reform and the Conservatives, with Reform taking it with a possible margin of 0.06%. Suffice to say that these results are well within the margin of error and that they’ll be determined by an astoundingly small number of votes.

‘Only we can win here’

Tactical voting isn’t over in Wales – it is simply transformed, flipped on its head. It is more conditional, more local, and harder to work out. That doesn’t mean you can’t do it, shouldn’t do it. When a smaller party is within striking distance, electoral arithmetic means that it makes a lot of sense to vote for them – even if your preferred outcome is for one of the main parties to win. Tactical voting is inverted. But where they are way behind, traditional tactical voting makes a sorry return.

Wales’ biggest parties will try and tell you otherwise – that only they can challenge each other, keep each other in check, as they chase an outright majority that is a near statistical impossibility. They’ll tell you this where it is true, and they’ll tell you this where it is false.

The truth is not so convenient, and neither, in many places, will be the choice. Laying claim to someone’s vote because you’re already in the lead or have the best chance to win, aside from anything else, is incredibly lazy. It’s the juvenile and deceitful politics of the ‘Only Lib Dems can win here’ bad bar-chart election flyers. Let’s bin it.

Wales, however imperfectly we go, is headed away from first-past-the-post politics, and we are headed away from one party Government too. Don’t get suckered into first past the post thinking – let’s make this next Senedd a good one.