Election handicapper Nate Silver’s forecasting model is swinging in favor of President Trump in the post-convention pre-Labor Day stretch of the campaign.

Formerly the head of FiveThirtyEight, Mr. Silver’s model is highly respected in the election analysis world and this week it is showing Trump with an advantage over Vice President Harris, giving Trump a 52 percent chance to win in November today, while it had favored Ms. Harris as recently as earlier this week.

Mr. Silver identified Pennsylvania as the culprit for the swing. In the past few days, a few relatively unfavorable polls of Pennsylvania for Ms. Harris have dropped. This includes an Emerson College survey out Thursday showing the two candidates tied, and a recent InsiderAdvantage survey showing Trump ahead by about half a point.

The model is also being swung by recent Republican pollsters, including Trump’s own pollster, Fabrizio, Lee and Associates, which shows Trump up by half a point in Pennsylvania, and a Cygnal survey that shows Trump up by a tenth of a point in the state. A Rasmussen Reports survey from mid-August also shows the candidates tied.

Mr. Silver, who now works at the betting site Polymarket, notes in his analysis that “The model is relying more heavily than I’d prefer on the Emerson College poll as it’s the only fully post-DNC/RFK data point.”

Recently, attorney Robert Kennedy Jr. suspended his presidential campaign and endorsed Trump. Recent surveys have suggested that most of Mr. Kennedy’s supporters were more likely to support Trump over Ms. Harris.

“Now, all of this could change quickly with one or two high-quality polls showing Harris ahead in the Keystone State,” Mr. Silver writes.

The recent swings in the forecast highlight the importance of Pennsylvania for the electoral college this cycle. If Ms. Harris does not win Pennsylvania, she will need to win Georgia, Wisconsin, Arizona, and Nevada as well as either Michigan or North Carolina to win.

On the other hand, if Trump wins in Pennsylvania, he will likely only need to win in North Carolina and either Michigan or Arizona to get to 270 electoral college votes.

Mr. Silver suggested that it’s possible that the recent endorsements Trump has accrued might be helping him more in Pennsylvania than in other states where they don’t seem to have made a major impact on the race.

He also cautioned that “it’s also possible that all of this is noise and/or that the model is overdoing the convention bounce adjustment.”

The “convention bounce adjustment” Mr. Silver is referring to is a measure used in many forecasting models that assumes post convention polling will be more favorable to a candidate than polling conducted a few weeks laters. 

In effect, Mr. Silver explained in a tweet, “It’s basically subtracting a net 2 points from Harris for polls during/immediately after the DNC. So it treats the Emerson PA poll today showing a tie as Trump plus two, and that gets a lot of weight as the only fully post-DNC/RFK PA poll.”