What presidential campaign polls can – and can’t – tell you

WASHINGTON (AP) — The presidential race is hotly contested.

That's about what the national polls can tell us right now, even if it looks like Democrat Kamala Harris is down in one poll or Republican Donald Trump is up in another.

And that's just okay.

Although polls are sometimes thought of as predictions, they are not intended to tell you who is likely to win.

Surveys are better for some things than others. Tracking changes in voter intent is difficult with a survey, especially when the number of truly persuasive voters is relatively small. Voters' opinions can change before Election Day, and that often happens. Horse racing polls can only capture people's viewpoints at a particular point in time. Even then, a margin that looks like it could decide an election—say, one candidate has 48% support and the other has 45% support—might make no real difference at all.

When reporters from The Associated Press report the choiceThe poll numbers on horse racing are not the focus. That's because the AP believes focusing on pre-election polls may overstate the significance or reliability of these numbers.

Election-year polls are still useful, especially when assessing how the public feels about the candidates or the state of the country. For example, they told us very clearly that many Americans wanted Democratic President Joe Biden to drop out of the 2024 race. But they are not the same as an election result, and even a poll conducted close to Election Day still reflects opinion before all the ballots have been cast.

Even with high-quality surveys, every result is just an estimate

Surveys are useful tools, but it's important not to overstate their accuracy. After all, a polling company cannot speak to every single person in the country. They instead rely on a sample to produce a statistically valid estimate of the views of all adults. While surveys can provide a reasonable approximation of the views of the larger group, the question is how much individual results may vary.

The margin of error that all quality pollsters report with their results helps capture some of this uncertainty. This means that in a poll with a margin of error of plus or minus 3 percentage points, finding that 47% of voters say they will support a particular candidate actually means that there is a very good chance that somewhere between 50% and 44% of voters support this candidate. If the other candidate has 45% support, which could actually be between 42% and 48%, the 2 percentage point difference is not statistically meaningful.

For this reason, the AP only says a candidate is in the lead if that candidate is ahead by more than twice the margin of error.

If you look at a subgroup rather than a national sample, the potential error is even greater. The fewer people are surveyed, the greater the error rate. This means that state-level polls, or polls that measure the views of a subgroup such as women, men, Hispanic Americans, or black Americans, are subject to even more error than a national poll.

The sampling error rate is not the only source of error in surveys. It is simply the only one that can be quantified using established statistical methods. But there are other factors too. The wording and order of questions can affect how people answer. An interviewer's skills can have an impact. Even with high-quality surveys, some respondents are less likely to respond, meaning their views may be underrepresented.

Don't forget the Electoral College

National polls measure how voters across the country feel about the election. But that's not how we elect presidents.

The Electoral College system means that presidential elections are functionally decided by a small number of states. So in some ways, a better way to gauge the state of the race is to look at the polls in these states.

But state-level polling comes with its own challenges. They are not conducted as frequently as national surveys, and some states conduct surveys more frequently than others. Additionally, the number of people surveyed in national surveys is often smaller than in national surveys, meaning the margin of error is larger.

What about the poll averages?

Some media outlets or organizations publish poll averages or aggregates that combine the results of multiple polls into a single estimate. There are some organizations that create polling averages or models during elections to find out which candidate is leading in the overall polls.

However, averaging survey results does not eliminate survey errors and may lead to additional problems. Survey averages contain their own methodological decisions, such as which surveys are included or given higher weight. Some of them also take into account other factors such as the economic situation to convert these estimates into forecasts.

In election polls, poll averages can provide a general sense of the state of a race. But it is also important not to overestimate the accuracy of an average or expect it to have an impact on the election outcome. Sometimes the individual results of several different surveys can give a better idea of ​​the possible range of results than an average reduced to a single number.

____

Read more about how US elections work Declare election 2024a series from The Associated Press designed to help understand American democracy. The AP receives support from several private foundations to improve its explanatory coverage of elections and democracy. Learn more about AP's Democracy Initiative Here. The AP is solely responsible for all content.