For the first time yesterday, I briefly came to the conclusion that the balance of probability slightly favoured a Leave victory. I was very surprised by the complete turnaround shown by Ipsos-Mori, which until now had tended to be the most Remain-friendly of telephone pollsters. I would have expected the Remain lead to be sharply down in their new poll, but a very clear Leave lead was a shock. (That was partly due to methodological changes, but even so.) A Survation phone poll also suggested a small Leave lead, and with some online firms reporting a Leave lead of as much as seven points, it seemed to me the likelihood was that Leave had built up just enough of a cushion to have a decent chance of holding on.
But then came the tragic murder of Labour MP Jo Cox yesterday afternoon. That's dominated the news for more than 24 hours now, and is bound to have some kind of impact on public opinion, even if it's entirely unclear so far what that impact will be. It may be direct or indirect (or both), and could be any of the following -
* Floating voters who had begun to swing towards Leave may take a step back and think "what is this country becoming?", and revert to Remain.
* There may be a backlash against the attempts by some on the Remain side to (with varying degrees of subtlety) politicise the tragedy. That could help Leave.
* There could be an effect on turnout. The 2004 Madrid bombings - which occurred days before a national election - didn't seem to directly change many people's voting intentions, but it did lead to a bigger turnout than anticipated, which helped the socialists to an unexpected victory. If younger people turn out to vote in solidarity with the Cox family, it could help Remain. But if the turnout is depressed by the suspension of campaigning, it could boost Leave.
* The public may conclude that the tragedy has no relevance at all to the referendum, but will still find themselves indirectly influenced by the suspension of campaigning. There have been some suggestions that the Leave camp will lose the momentum they had built up, but I suspect Remain might have the greatest problem in this scenario. The advocates of the status quo tend to have the most cards to play at the close of a referendum campaign, and Remain have effectively lost two days that they presumably would otherwise have spent ramping up apocalyptic fears about the economy. (Given the public mood after Ms Cox's death, it's also going to be hard to resume a relentlessly negative, fear-based campaign when hostilities resume.)
If I was going to guess, I would say that Remain are more likely to benefit in one way or another, but we'll just have to wait for polling evidence. The information will probably come gradually - first we'll get polls that were only partly conducted after Ms Cox's death, and only after that will we get definitive evidence of any shift in opinion. As always, it's also worth remembering that a substantial minority of votes have already been cast by post - and there are strong rumours that Leave have done very well so far.