Tag Archives: elections

Hot take: polls are bad for democracy

Why do we vote?

It’s a pretty interesting question, with a long answer. The history of democracy is a good read, but what you need to know is that – like all good things – it was invented in Greece. The city-state of Athens around 5 BC is generally regarded to be the first implemented model of one-person-one-vote democracy as we would recognise it today.

It was a direct democracy in the sense that the people of Athens themselves were the voters on each decision, without further representation. (A footnote to say that ‘people of Athens’ referred to citizens – a somewhat non-inclusive group, excluding women, slaves, young people and so on). And for a while that was good.

Today, in the UK, we have a representative democracy. We don’t vote on each and every issue like the Greeks did – we vote for people who make those decisions on our behalf. Further still, our model is one of parliamentary democracy where we vote for representatives to form a government and the head of government is appointed by the head of state (Queenie!) to run things – that being whoever is the leader of the most popular party, as voted for by that party’s members. Yeah, it’s a bit complicated.

But the basics of it is that in the UK we vote for parties. We divide the country up into about 650 constituencies (voting areas), and the parties field candidates in each constituency. At a general election, whichever party wins the most seats gets to be in charge, since we have a dumb and stupid non-representational voting system.

So far so good. It’s fair, open, easy to understand, and about as democratic as you can hope for in the 21st century. The old Athenian system simply wouldn’t work in 2018. We can’t be expected to understand or have an opinion on every issue. We don’t even have referendums that often (imagine this current Brexit mess, but for EVERY SINGLE DECISION).

There’s some debate about whether MPs are meant to be ‘delegates’ or ‘trustees’. As delegates, they’d be the absolute instruments of their constituents will – perfectly channeling the voice of the people. But as members of a political party that isn’t really feasible. Yet having them as ‘trustees’ means letting them make decisions that might even be against the will of their constituents, if it’s actually in their best interests. For example: most constituents will oppose an increase in tax, but support better public services. A trustee can make the difficult decision to oppose the ‘will’ for the overall good.


When describing how democracy works above, we haven’t mentioned polls at all. That’s because they’re simply not part of the concept. In fact, you could say that there’d an inherent idea core to democracy that is incompatible with political polling: the idea of the secret ballot.

The ‘secret ballot’ is the idea that everyone’s vote is a secret. Everyone has a voice, but nobody knows what everyone else has voted for. This is a good thing for democracy: it prevents voter intimidation or squeamishness. We always have a desire to fit in, and our votes being public might stop people voting from how they really want to. Yes, it helps extreme groups win more votes, but overall it’s a good thing.

Do polls violate the idea of a secret ballot? Clearly not. The secret ballot idea applies at the individual vote level, not overall. After all, we have to know the outcome! But there’s another idea at play here: that the outcome of a vote is uncertain.

Democracy works best when it’s a disinterested process. By this I mean that the result of a vote is up for grabs. After all, what’s the point in voting if it’s a foregone conclusion? Brexit and Trump were ‘good’ votes in that the outcome was surprising (what they demonstrate about the democratic process is another conversation). The will of the people needs to be a potent force, and that’s not possible when decisions are robbed of their jeopardy.

And yet this is precisely what polling does. By ‘polling’ I mean the forecasting and publishing of election results. Bodies like YouGov, ICM, and Ipsos Mori do this, attempting to predict the outcome of elections before they happen. And this is bad, for a couple of reasons.

Firstly, as said above, it robs elections of their jeopardy. If the outcome of an election is a foregone conclusion – which for polling to have any authority, it must consider it to be – then the process is likely not truly democratic. Sure, in a perfect direct democracy there would be issues with a 99% consensus already. But even still, publishing the result of an election, even a ‘predicted’ one, before the vote seems to betray the concept of uncertain outcomes that makes democracy effective.

And don’t think this doesn’t have an impact on turnout. Ask people why they don’t vote and you’ll likely hear “because it wouldn’t make a difference” anyway. I live in the Hackney North and Stoke Newington constituency – which elected Labour MP Diane Abbott to Parliament in the 2017 election with 75% of the vote. The Conservative candidate got 12% of the vote. So if I was a Tory voter, would I feel an empowered democratic citizen having this repeated again and again in the polls running up to an election? Clearly not.

Something that drives me crazy is the use graphs like the above in election pamphlets. “Can’t win here” is the catchphrase they use, and it’s awful. Every election is a blank slate, anyone could win. If every voter decided to vote Green in your constituency, you’d get a Green MP. And yet polls set an alternative narrative: that of the status quo. That the incumbent state of affairs will more or less be maintained. It’s no wonder that young people and other groups feel disenfranchised and powerless.

Polls also shift the way we think about voting, in bad ways. ‘Tactical voting’ is a symptom of a broken democracy. That’s when you vote, not for the party you actually want to win, but in order to help prevent someone else from winning. Cleisthenes must be turning in his grave. Vote-swapping systems and things like that arise because polls turn an election into a scientific game rather than a legitimate decision-making exercise.

And it encourages bad behaviour from our representatives too. During the election cycle, incredible levels of attention are paid to the polls. And the campaign strategy is informed by the poll results. Sure, it’s the sensible thing to do given that polls exist – but it’s a reflection that elections have become a vote-winning version of marketing. And that’s bad for democracy too.

Can we beat this? Can we claw back democracy from the damage polling has done to it? I don’t know. As with all technology and science, once the genie’s out of the bottle, it won’t go back in. On the TV debates they’ve started doing recently, they’ve now got those live ‘worm’ graphs, where support for each speaker is displayed live in real time along with the program. I think that’s too far – encouraging a safe, boring approach to politics where every word of every sentence is rehearsed to media-trained perfection.

Keep polls, sure. But for the sake of democracy let’s stop obsessing over them. Bring back the uncertainty to politics. It’s an open field, anyone can win. Trying to predict the future just sucks the joy out of it entirely.