Overview of Changes in Election Polling Techniques and Their Impact on Accuracy

The accuracy of election polls has been a matter of considerable debate, especially following the 2016 US presidential election. While these polls were generally off, the errors were particularly pronounced in certain regions, notably the Midwest and particularly the Rust Belt states. In these states, the polls significantly underestimated Donald Trump's chances, leading to a perceived upset victory for Hillary Clinton.

The 2016 Election Polling Mistakes

The primary reason for the errors in the Midwest states, including Iowa, Wisconsin, Michigan, Ohio, Minnesota, and Pennsylvania, was the failure to accurately identify the likely voter pool. While it was easy to determine party affiliation, it was far more challenging to predict which voters would actually cast their ballots. Henderson (2017) argues that the polls relied too heavily on registered voters, who were more likely to be Democrats, skewing the results.

What made the 2016 election particularly challenging for pollsters was that not as many members of the traditional Democratic coalition (the "Obama coalition") supported Clinton. In contrast, Trump's support was unexpectedly strong in these critical states.

The Role of Likely Voters in Future Polls

Going forward, it is anticipated that the polls will become more accurate as they transition from using registered voters to likely voters. According to the analysis by Alford (2023), likely voter polls should provide a more accurate representation of the electorate. This shift is crucial because Republican voters are more likely to vote than Democrats.

For Trump, the likely voter pool will be known, which addresses one of the major challenges of the 2016 polls. However, identifying the likely voter pool for Joe Biden remains a significant challenge. The baseline will be adjusted to reflect the change in the voter pool from Hillary Clinton to Joe Biden, but the specific adjustments required are complex and difficult to determine accurately.

Conclusion and Future Expectations

In summary, it is expected that while polls will be less accurate early in the election cycle, they will improve in accuracy as the election draws closer and likely voter polls become the norm. Pollsters acknowledge that a certain degree of uncertainty is inherent in the polling process, even with an incumbent candidate involved.

It is crucial to understand that while these changes will help improve the accuracy of polls, they do not eliminate all uncertainties. However, with better data and a focus on likely voters, the polling should become more reliable in the weeks leading up to the election.