There is an election coming up in the UK and I am registered to vote. Which party should I vote for?

Expressive Voting

Simon Wren-Lewis lays out the case for tactical voting here. In that article he distinguishes “expressive voting”, where you honestly and straighforwardly give your opinion, from “instrumental voting”, where voting is a way to manipulate the system to produce your preferred outcome from those that are possible. I agree with him, that voting should be instrumental.

But first, imagine that I voted expressively. The most important issue of our time is global heating. My vote should go to a party that puts that issue at the forefront of their message, and commits first to radical decarbonisation, and second to humanely dealing with the global catastrophe to come.

If there was any wriggle-room left after that, I would prefer a party that would shift the economy in a broadly “left” direction. I want more economic equality, achieved by progressive taxation and investment in disadvantaged communities. And I want essential infrastructure subject to natural monopolies to be taken out of private hands and put under democratic control.

The problem here is that none of the party manifestos really align with those priorities. The best option is obviously Green, but their proposals are frustratingly tepid compared to the situation at hand. They certainly would never propose opening the UK as a safe haven for refugees fleeing climate change in the tropics. That would be electoral suicide. Their electorate is English, and the overwhelmingly majority of English people are awful.

With all the parties clustered together in a corner, all using similar language to condemn my actual opinion as “unpatriotic” or “economically unrealistic” or whatever, I could very easily just give up. In practice there are two possible choices: voting, which shores up the system as it exists, or not voting, which brings forward the day we can tear it all down and start again. And you can hear that reasoning from lots of people, both from tired mainstream voters and from engaged radicals.

Tactical Voting

But I’m not an expressive voter, I’m an instrumental voter. And so we come to “tactical voting”. In particular, organised tactical voting. The goal is either to keep the Tories out, or ensure Labour wins, because however bad Labour are they are not nearly as bad as the Tories.

Organisation is important here because the correct choice depends on local factors. In the UK, many seats swing between Liberal Democrat and Conservative rather than between Labour and Conservative. In those seats, voting for the Liberal Democrat is better for Labour than voting Labour, because it potentially denies a seat to the Conservatives. To clarify this complexity there are websites that crunch the numbers. For example, here is one recommendation for my constituency. It recommends that I vote Labour because they have the best chance of winning.

Looking in more detail at polling for my constituency, we can see that it’s actually a safe Conservative seat, and will likely remain in Tory hands despite the landslide against them. It’s also clear that if all the Green voters voted tactically for Labour, Labour would win. Mind you that swings both ways: if Reform voters voted tactically for the Conservatives as well, the Conservatives would win after all. So even though it’s unlikely that the Conservatives will lose this seat, it’s an excellent candidate for tactical voting.

Paradox of Voting

A big flaw in the argument for instrumental voting is that any individual’s choice is astronomically unlikely to affect the outcome at all. Simon Wren-Lewis gives the example of someone convinced their vote wouldn’t make a difference, but in fact the candidate won with 10 votes – thus proving that indeed their vote didn’t count after all.

There are several workarounds for this, but none as powerful as simply saying that an election is an assessment of the opinion of the population, and an accurate result is the most important outcome of all. So that returns us to expressive voting.

The alternative proposed by Simon Wren-Lewis is to just think about your vote as being a part of a wider social movement. I don’t find this very convincing. If there are enough people in society to express the society’s interests collaboratively, then they don’t need me. And if there aren’t enough people, then I’m not going to be able to change that on my own. We just circle back to the same paradox.

Perhaps it works better if you force the system to work like that by formally forming a group. You can imagine someone starting the Union Of Suffolk Computer Programmers. (There are more of us out there than you think!) We would internally choose our preferred candidate, and then all vote as a unified bloc at the election. In the alternate timeline where computer programmers are capable of cooperation at all, that could be very powerful, and blow the Paradox of Voting out of the water. But no such entity exists, in practice it’s just me here deciding on my own.

I think I have a better argument.


“Spad” is British political jargon for “special advisor”. They are the analysts who weigh up policy proposals against opinion polling and make recommendations to politicians. They are, in a sense, worryingly powerful, because they effectively decide everything that happens in government. But in another sense they’re no more powerful than anyone else: they are political engineers, reporting on hard reality. If their predictions are not accurate, either this will be noticed and they will be replaced, or their political patrons themselves will be replaced.

By one way of looking at it, it doesn’t matter who actually gets elected. All parties have spads working from the same data, who will draw the same conclusions and propose the same policies. If you are a citizen who wants to change how the country works, your target is the spads, not the politicians. It’s true that they only care about the election outcome, and your vote is the only thing you can control. But that’s just a relay through which you communicate with the people who actually hold power.


Thinking about it this way makes voting an incredibly powerful tool. Government is a constant stream of decisions. Every one of these decisions is weighed up for its political impact. Votes from the last election are broken down into categories by age, education, race, gender, and hundreds of other factors. The effects of similar decisions in the past are analysed. From this they can predict whether the votes gained by the policy will outweight the votes it will lose. In effect the election is run over and over again every day. Your vote doesn’t count just once, on election day, but thousands of times over the whole of the next term. You don’t even need to leave the house.

Even more than that. Your vote doesn’t only count in your own electorate. The way you vote has lessons for how people like you will vote in other electorates. Your own electorate may be safe, but you can prove that there are votes to be won, and those votes also exist in marginal constituencies, where that can really change policy. This can even spill over into other countries!

Having the vote count thousands of times over rather than just once certainly weakens the Paradox of Voting. What kills it though, is that unlike the election, many of these decisions are fuzzy. A winning margin of one vote is still a win at the election. But a prediction on a spreadsheet that after a particular choice, an incumbent politician will win by only one vote, is a red alert. Every additional vote can make the issue more salient and help to change the outcome. Some decisions are even about allocation. A pool of funds might be distributed among electorates proportionally to the tightness of the winning margin. In those cases, each vote is cash money, albeit only a small amount on its own.

But all of this calculation is contingent on you actually voting. Groups that don’t vote don’t get counted. Every decision has winners and losers. If you don’t vote, you will always be one of the losers.

Safe Seats

This is all less persuasive if you live in a safe seat. If the same party is reliably going to win election after election, why would they even bother running the analysis?

There is a notable pattern, especially looking at the USA. A minor or local politician will be caught in some blatant abuse of power, or say something considered outrageous by all right-thinking people. And you would surely expect that person to get kicked out, either by the party immediately, or the voters at the next election. But it turns out that they are in a safe seat, and they are allowed to continue without consequences.

In fact, there is value for a party in having an attack dog, someone who will play dirty without caring how they are perceived. This appeals to a certain type of party member, who will be motivated to work harder for the party. Such a politician can also be used for political dirty work, from fundraising to obstructing the work of the legislature when advantageous. So the party will deliberately install such people in safe seats, since they could not possibly win a marginal.

This means there is value in voting even in a safe seat. If you would prefer your foaming-at-the-mouth partisan to be replaced by someone who sincerely cares about their voters, you can reduce the margin and give the local party pause for thought. If, conversely, you would actually prefer a radical extremist pushing the Overton window in your direction, then by all means give them an even bigger winning margin than they otherwise would. Either way, your vote does count.


All of this describes a political process that works fairly smoothly to follow the preferences of the people, the way democracy is supposed to work in theory. But it would be naive to assume that this is always true, or even typically true. In reality, political choices are limited by the institutions in which they are made, and often the design of the institutions was chosen deliberately to prevent certain democratic choices. For example, the American electoral college only makes sense as an attempt to preserve the institution of Black slavery. It was always going to take more than just accumulated election results to change that.

Bad as it is, I wouldn’t compare the current mainstream British political consensus to chattel slavery. But that’s from my perspective as a comfortably middle-class white male professional. If one of the many people permanently excluded from prosperity in the UK told me they chose not to vote because any vote would only perpetuate the current broken system, I wouldn’t have a good answer to that.

I like to compare the UK to Iran. This is partly because constitutionally speaking, Iran is significantly more democratic. There is a formal constitutional procedure for democractically and legally removing the Supreme Leader. No equivalent provision exists in the British constitution. It’s true that Iran’s constitution does dedicate the country to following and promoting one specific religion, but so does Britain’s. Iran holds regular elections with candidates at least as diverse as the interchangeable suits rotating in and out of executive suites who lead the UK. It should be just as possible to effect change in Iran one election at a time as it is in the UK, even if the current crop of politicians doesn’t inspire confidence.

Would Simon Wren-Lewis suggest Iranian voters should dig their hijabs out of the back of the wardrobe so they can go vote for a candidate who would have them beaten for refusing to wear one? No, he doesn’t say that. In fact we all understand that the institutions of Iranian politics are beyond saving, stuck in an anti-democratic feedback loop that only revolutionary change can solve. Until then, boycotting the fake elections is the only reasonable choice.

I don’t believe the UK is at that stage yet. But there is no particular institutional reason it could not be. And if it did complete that slide into authoritarianism, it’s unlikely that I would be the first to notice. It would therefore be wrong of me to reflexively dismiss voices who claim the UK has already passed that threshold.

My Choice

If my electorate was a safe Labour seat, I would be voting Green. I would want to send a signal that while environmentally sustainable policies might lose the votes of short-sighted voters, there are votes to be gained as well. I would want the spads to treat me as a persuadable voter.

However, in reality my electorate is safe Conservative, and I would like to push the Conservative party towards treating it as a marginal constituency. Also, I find the fact that the Labour and Green vote combined would win the electorate to be very persuasive. I would like to send a message to other Green voters, that while I share their views, there is a way for us together to make a real difference. So my choice is to vote Labour.

In practice, unfortunately, I live overseas and my voice doesn’t matter after all. So just forget the whole thing.

Leave a Reply

Your email address will not be published. Required fields are marked *