About what it said it was, maybe a point or two higher due to some of his core voters being the types that are incredibly disconnected and hard to reach through traditional surveying methods. Incumbent president reelection numbers don't usually exactly match their approvals, and not all of the potential American electorate votes. He certainly didn't have 47% of the total population's approval.
According to 538, on day 1,382 of their presidencies (around election day), Bush had 48.4% approval, Obama had 49.0% approval, and Trump had 44.6% (a high for him). In their re-election bids, Bush won 50.7% of the vote, Obama won 51.1%, and Trump won 46.9%. For Bush that's a difference of +2.3, for Obama it's a difference of +2.1, and for Trump it's a difference of +2.3. It seems that there is a consistent >2% of support for incumbents that isn't recorded by averaging approval polling.
However, approval polling seems to accurately capture the support of the opposing party's candidate. Bush had 47.5% disapproval, Obama had 47.4%, and Trump had 52.6%. John Kerry won 48.4% of the vote in 2004, Romney won 47.2% in 2012, and Biden won 51.3% in 2020. For Kerry, it's a difference of +0.9, for Romney it's extremely close to Obama's disapproval, only -0.2 points off. However for Biden, it's a larger difference of -1.3 points off of Trump's disapproval.
Very interesting, thank you for that information!
I wonder why approval polls are more accurate than electoral polls. Its odd that people are comfortable saying who they approve/disapprove of, but not who they plan to vote for
Biden currently has 53% approval rating. If we follow the pattern and add two points, he would win 55% of the vote. With that, he would win the 413 nut map
