There is nothing methodologically that means my Superman poll should be treated no more or less seriously than the one below, from C-SPAN.
Nothing about the C-SPAN survey makes it more legitimate than mine, including the number of respondents. After all, anyone could have sent a link to the poll to a large group of supporters of either of the participants in last week’s presidential debate, encouraging one vote or another. In fact, C-SPAN’s may be less legitimate, since it’s far more likely that someone who voted in mine was familiar with Superman than it was that someone voting in C-SPAN’s actually watched the debate.
And, yet, the C-SPAN tweet was treated seriously. On Oct. 1, President Trump shared the results with his tens of millions of followers.
His point was clear: He hoped to reinforce the idea that he was the victor in the first matchup between himself and former vice president Joe Biden. But actual polls — polls conducted by actual pollsters using actual statistical methods to measure actual views of the candidates’ debate performances — consistently show that it was Biden, not Trump, who prevailed.
On Sunday, another poll was added to that mix. A poll conducted by NBC News and the Wall Street Journal found that about half of respondents thought Biden did better than Trump, with about a quarter giving the advantage to the president. Biden was seen as the winner by a nearly 2-to-1 spread, similar to the results he saw in a CNN-SSRS poll conducted shortly after the debate and in a CNBC-Change Research poll released this week. Only in a CBS News-YouGov poll conducted shortly after the debate was the spread even close — but even there, Biden was viewed as the winner.
It’s worth noting that the results of CNN’s poll appear to reflect an actual assessment of the contest. The pollsters asked the same people both before and after the debate who they thought would win and, then, who actually did. Before the debate, most Republicans thought Trump would win and most Democrats thought Biden would. After, the percentage of Democrats giving the debate to Biden increased, while the number of Republicans who saw Trump as the winner plunged.
A poll conducted by FiveThirtyEight and Ipsos didn’t ask a direct “who won” question but nevertheless reflected the broader trend.
Six in 10, for example, thought that Biden had a good performance, while only a third said the same of Trump. Asked both before and after the debate how likely they were to support either candidate, the level of support for Biden rose — albeit slightly — after the contest, while support for Trump fell.
A poll from the New York Times, conducted by Siena College, looked at views of the debate in two key swing states, Pennsylvania and Florida. There, once again, Biden was seen as winning the debate by a 16-point margin.
The pollsters also evaluated views of how the candidates did directly, with nearly two-thirds of respondents saying that they disapproved of how Trump handled himself during the event and nearly half saying that they were less supportive of him after it was over.
Again, all of these are actual polls of the sort we use to measure public opinion and gauge the likelihood of certain outcomes. The kind of polls that, in 2016, suggested that Hillary Clinton had a narrow national lead — an assessment proved accurate by the results of the national popular vote.
Incidentally, the average of national polls in the 2020 race has Biden with an eight-point lead. One week ago, before the debate, his lead was just over seven points.
To put a fine point on it, there is no more evidence that Trump won the first presidential debate than there is that you can relax in the event that your car drives off a cliff because Superman will catch it in midair.