[DTV]

The thing is that at 10pm on election night the only thing that matters is who's won the election. Opinions are for the previous four weeks, they aren't relevant once voting stops.

Why people voted the way they did matters just as much as how they voted - not just in terms of understanding the result itself, but it has repercussions for politics over the next parliament as MPs and the government try to respond to the public. It is fundamentally odd that UK election coverage doesn't really go into much data-based analysis of this. Instead, you get politicians and journalists writing the narrative of the election - during the programme - based on either motivated reasoning or assumptions that may or may not be correct.

And it impacts the real world. During the 2017 election night programme, the BBC's political editor repeatedly made the claim that, because UKIP's vote was down by around 10 points and Labour's was up by around the same amount, Labour's surge was driven by UKIP voters. But this wasn't true - although there were some UKIP to Labour switchers, they were a small minority (~10%) of the new Labour voters. And you didn't need post-election studies to tell you this, opinion poll data at the time said this. Yet because the BBC programme lacked the data, the assumption wasn't quashed and became one of the stories of the election - I've heard it repeated for years, with the BBC regularly cited as a source, and it arguably impacted Labour strategy in 2019.