Okay, first some self-praise. When I wrote a column on October 18 stating that Trump would win 312 college votes with Harris getting 226, I was largely derided. Even my own colleagues thought I had lost the plot. Hours before the actual voting began, one of them suggested I should take it down as I would soon look very stupid.
As the final votes are tallied, the result looks certain to be – you guessed it, Trump 312 and Harris 226. Everyone is staggered that Trump has won a near landslide. No one saw it coming. The polls were, yet again, completely wrong.
Actually, they weren’t. Here’s the thing: the polls were right all along, it’s just nobody – absolutely nobody else – managed to read them properly.
Let me explain: the way the US electoral system works, any candidate needs 270 college votes to become president. The college votes are distributed across different states, the number depending on the population sizes. Given the huge leads both Trump and Harris already had in many states (for example Trump was guaranteed to win Texas and Harris California), one thing all experts agreed on was that on election night, Trump would start with 219 votes and Harris 226. Those were in the bag.
So it was going to come down to 7 “swing” states which between them accounted for 93 votes. Whoever got more of those would win.
Now to the polls: for the last 2 months, over 3,000 different opinion polls all came to the same conclusion: neither candidate had a lead of more than 1 per cent in any state. Pretty much a dead heat. Hence the official narrative: the election was too close to call.
But here’s where everyone went wrong: it was never too close to call. The polls, as they always state themselves, have a margin of error of 2-3 per cent. Which meant Trump could win a state by 2-3 per cent, or Harris could win a state by 2-3 per cent.
So, the next question to ask was if this is true, which way would the 2-3 per cent go – could it go different ways in different states?
Well, in the last 2 elections, Trump had outperformed every poll by 4 per cent. So, it was safe to assume the extra 2-3 per cent would all go to Trump. We have over 120 million real votes data to make this assumption from the last eight years.
Given the polls were forecasting a dead heat in every swing state, that meant Trump would almost certainly win every single swing state – and all 93 votes – and so end up on 312.
In fact, if you look at actual results were the polls said it would be a dead heat – Pennsylvania, North Carolina, Georgia – he won each of these by 2-3 per cent. Across the country, with some counting still to do, he is on 72mn votes against Harris is 67mn – a swing of 3.8 per cent. Almost identical to the 4 per cent extra he has got in the last two elections.
In other words, the polls were EXACTLY right. They were always pointing to a Trump landslide. It also means, sadly for me, I am not a genius, I just know how to use a calculator.
Follow us on
