A number of polls from highly respected outlets are disclosing a historically tight race between Vice President Harris and President Trump, with Pennsylvania as — unsurprisingly — the most contested state of 2024. Some Democrats think that Ms. Harris is much stronger than she appears, however. 

Just on Wednesday, two highly rated pollsters released their survey results for Pennsylvania. Monmouth University polling center showed Ms. Harris and Trump tied with each taking 48 percent. 

“It’s important to note that any movement we’ve seen since the last Monmouth poll is well within the margin of error. What we said last month still applies,” says the director of the poll, Patrick Murray. “Percentage point shifts are too small to be statistically precise in a poll, but they could be consequential if real. The bottom line is this was an incredibly close race in September and remains so today.”

An in-state Pennsylvania pollster, Susquehanna Polling and Research, found Ms. Harris leading Trump by just two-tenths of a percentage — 46 percent to 45.8 percent, with 6 percent saying they were undecided. A CNN poll also released on Wednesday found Ms. Harris and Trump tied again, with each taking 48 percent of the vote. 

That CNN poll did find that Ms. Harris is leading Trump by five points in Michigan and six points in Wisconsin, however. 

Downstream of those polls are the polling averages that show Ms. Harris’s best path to victory is through the so-called “Blue Wall” of Pennsylvania, Michigan, and Wisconsin. The RealClear Politics average currently has her losing only Pennsylvania, with Trump taking the lead by less than 1 percent in the average. 

It isn’t just polling averages that are finding such a tight race now, but the Electoral College models as well. Some Democrats on X found it amusing on Wednesday when the Economist’s 2024 election model found that the most likely scenario, based on where polls are, is that Ms. Harris and Trump each take 269 Electoral College votes, resulting in a tie. FiveThirtyEight’s Electoral College model found that Trump has a 51 percent chance of winning the election, compared to Ms. Harris’s 49 percent. 

Democrats on X have accused pollsters of “herding” or “hedging” their polls in the lead-up to the 2024 election because those who are deemed the most “accurate” of the election this year will likely get two or four years of adulation and additional polling contracts from campaigns. Some of those complaints point to pollsters’ underestimating of Trump’s support in both 2016 and 2020, leading to closer than expected results, which then allegedly drew some of those same pollsters to oversample Republicans and white working class voters in order to get a more accurate picture of the electorate. 

Many of those say that 2022 was a repudiation of that method of sampling voters, because pollsters missed the mark on some vital races for the House and Senate, leading Republicans to a historic underperformance at a time when a Democratic Congress and Democratic White House were deeply unpopular. 

The RealClearPolitics averages in the Georgia Senate race showed the Republican winning by two points even though the Democrat won by three points. The same average showed the Republicans picking up a Senate seat in Nevada with a healthy four-point margin, though Democrats narrowly held on. Similarly, in Pennsylvania, the average showed Senator Fetterman losing his race by just half a percentage point, even though he would go on to win the race by five points. 

Some electoral analysts agree that the pollsters are “herding” to hedge their bets. “There are too many polls in the swing states that show the race exactly Harris +1, TIE, Trump +1. Should be more variance than that. Everyone’s herding,” said the famous pollster Nate Silver on X. 

A professor of economics at the University of Queensland, Lionel Page, says herding is natural this close to an election because pollsters have a vested interest in protecting their own brands. “Poll herding is a real practice among pollsters. They hedge their bets by tilting their estimates towards the average of other polls,” Mr. Page said on X. “This means that all pollsters are looking at each other, and there is not as much information in the average of polls as we might think.”