How Trump Rattled Swing Voters, But Biden Won the Day | Analysis

In the last few days, president-elect Joe Biden has completed his transition from the kindly-looking bloke who hung around with Barack Obama back in the day, to the Sole Savior of Western Liberal Democracy. He has never been so popular, as in the post-election period, basking in the Democrats' relief that he got over the line. In public profile terms, he's doing better now than when people were voting.

Court battles aside, and on the assumption they are futile, the Democrats are feeling happy again—apart from a few uncomfortable numbers. Yes, he got over 76 million votes. Trump got over 71 million. For comparison—Obama (with Biden on the ticket, of course) got 69 million in 2009 and 65 million four years later. The demographics are different, even in two decades, but still...

Furthermore, as we in the UK learned the hard way over three and a half tortuous years of Brexit, a victory by 52 percent to 48 percent does not lead to a healing of divisions.

It was all so, so much closer than was assumed polling companies who have, for months, fed us the view that Biden's victory was assured and the margin would be healthy. That's not how it looked for most of last week, as things got so much closer than predicted.

Partly that's an issue with how we all read the polls. If someone tells us that there is a 70 percent chance of a Biden win, we absorb that Biden will win. Not that there's a 30 percent chance that Trump will. And we get angry that we weren't beaten over the head with both those possibilities.

Partly it's also an issue with the idiosyncrasies of the electoral college system where votes in swing states count for so much more than those in stable states.

But, mostly, for the polling industry itself is that they got things wrong. The issues of misreading the white working class (male) vote that were obvious in 2016, were obvious again. The issues of assuming every immigrant group would vote homogeneously (and liberally) were also shown to be wrong. The possibility of Trump voters being as shy as their candidate was bombastic was missed again.

The blunt instrument of polling was exposed - the notion of using tiny sample sizes from demographics of the population they didn't understand and then asking them binary questions they didn't care about, seems limited and, more importantly, inaccurate.

Instead, over the last few weeks, Newsweek has been looking at election data through the prism of the social media sentiment of 40,000 swing voters through the work of Impact Social, who have been following those voters for four and a half years and understanding the sentiment (for each candidate throughout the whole process). Physically reading thousands of posts to align the sentiment and topic to the fortunes of each candidate opens up a far more nuanced view, since it allows the electorate to raise the issues, rather than asking individuals to react to presumed agendas. Sometimes more volatile and always based more on the emotions of the electorate than the logic of the Beltway, it has shown a consistent pattern over the last few months which accurately reflected the outcome. But the numbers have also always shown that Biden has much work to do yet.

While Biden was always ahead on the sentiment numbers, the gap was based more on the polarizing effect of Trump, rather than his own gravitational pull. Few voters have ever been sanguine about Trump - the idea of a floating Trump voter is based more on the conflicts of enjoying the message while being ever unsure whether to trust the messenger. As a result, his ratings have always hit more extreme numbers. Trump always set the agenda - conversations were way more driven by opinions about what he had done and said than anything Biden could muster, so that extremity of reaction was, simply a measurement of activity. If Trump spoke, the Richter scale of social media would tremble violently.

Not for Biden—in all the online conversations, there was little to no talk of his message. His utterances were swamped by Trump's and his rating were driven by the fact those utterances weren't Trump's. There has consistently been little evidence that voters knew what Biden stood for when they voted for him. They just knew he wasn't Donald. That's been enough to get him over the line, but he doesn't start with a defined purpose in the way his old pal Obama did. Biden is fortunate enough to have a blank page on which to write the story of his administration, but it's odd to have reached the President's desk seemingly without having picked up the pen.

And in four years' time, the pollsters can try and measure how that went and whether they learn the lessons this time. Perhaps in a second round with Trump. The outgoing President used social media as a tool to get elected and that's now the established playground for the primary voices of our politicians. It makes sense to measure the debate in that same place.

1. Biden saying very little. Not defined by anything other than "not Trump". Dangerous position to be in. Trump well ahead in discussions on economy and China
2. Woodward book makes little impact on Trump, swing voters talk instead of the bigger issues they are interested in like Russia, corruption, Covid. Trump needs to get the attention off himself and onto the issues that impact people people speak more favorably (economy / China) and his sentiment score will rise. Biden benefits a little from the military fall out. Swing voters say he is "a good guy" but give little reason. There is no explanation as to why they like him. With little to say Biden seems to have hit a ceiling.
3. Biden net sentiment nosedives. Support is likely still there but voters have become bored and Biden needs to give them something to rally around. Meanwhile Trump gets UEA peace deal and uses it to present himself as a man of peace and label Biden as a warmonger. This is as close as it gets between Trump and Biden. This highlights Trump's success when he gets the conversation away from himself and onto an issue of importance. Trump gains support when he shows more delivery and less circus.
4. News of Justice Ruth Ginsberg's passing was seen as a game changer as Trump moved to fill the SCOTUS post. Media predict a firestorm between left and right. But this doesn't happen. Swing voters conclude that Trump is within his rights and they quickly move on. Even the impact on women's rights and abortion are given short shrift. Trump flatlining heading into the first TV debate and needs to do well if he is to bridge the gap to Biden. Yet there is an increase in support for Biden. Again this is not about policy and so is precarious.
5. Trump tax returns leaked and he is largely blamed for a chaotic TV debate. But his sentiment flatlines, suggesting minds are made up. Trump contracts Covid, which brings the issue back into focus.
6. Covid fallout makes Trump look reckless (not tough) and forces the focus. Biden basks in the contrast. He says little, but is spoken of as "responsible" and "presidential". Conversation continues to be more anti Trump than pro Biden. Biden is letting Trump talk his way out of the White House.
7. Trump says less and his sentiment improves. Makes himself less of a target. Floating voters respond positively, but will Trump learn?
8. Hunter Biden emails: Trump attacks but this just highlights his own family's failings. Also serves to bring people out in defence of Biden. So the attack instead backfires on Trump. Biden has momentum
9. Trump rallies see him shouting into the wind. Minds are made up. No one cares outside his base. Biden remains dignified.
10. Biden wins comfortably.