By Bryce Yazdiya Photo Credit: AFP; Getty Images
Following Donald Trump’s recent GOP primary victories, it is becoming more likely by the day that the former President will be the Republican nominee that will battle against Joe Biden in a race for the White House. However, with the 2024 election coming up fast, who holds the edge between these two candidates?
According to a February poll conducted by YouGov and the Economist, Donald Trump leads Joe Biden 45% to 44% nationally with the margin of error being 4%. Another poll conducted by the Morning Consult shows that Donald Trump possesses a 43% to 42% lead over the current President, with Biden having closed the gap since their previous poll which had Donald Trump ahead 45% to 41%.
Several states that Joe Biden won in the 2020 election are now showing Donald Trump with the advantage going into 2024 as well with the former President leading Joe Biden among swing voters according to recent polling data in Arizona, Georgia, Michigan, Nevada, North Carolina, Pennsylvania, and Wisconsin. Donald Trump’s poll leads are within the margin of error in all these states except in Georgia and North Carolina where he has a lead that extends beyond the margin of error.
While these results reveal a lot of information, are these polls a reliable predictor of what’s to come on election night in 2024? Polling predictions in previous presidential elections could provide some context in regards to the accuracy of these polls in 2024.
According to FiveThirtyEight’s 2016 election forecast which simulated the 2016 election 20,000 times after analyzing all the polling results and trends, Hillary Clinton had a 71.4% chance of winning the election and was predicted to take home 302 electoral votes compared to Trump’s 235. The actual election results, Donald Trump 304 electoral votes and Hillary Clinton 227 with Donald Trump winning Florida, Wisconsin, Michigan, Pennsylvania, and North Carolina, all states predicted to vote for Hillary.
In regards to the 2020 election between Joe Biden and Donald Trump, Joe Biden was given an 89% chance to win the presidency and was predicted to garner 348 electoral votes to Donald Trump’s 190. In actuality, Joe Biden ended up winning 306 electoral votes to Donald Trump’s 232. While still a victory for Biden, the results were much closer than the polls/predictions suggested. What could then be the cause of the stark contrast between these predictions and the actual results of the election?
First and foremost it should be noted that every poll includes a margin of error which encapsulates the full range that statisticians believe could be the actual result of an election within a 90 or 95% confidence level. Polls are often unable to predict the actual result vote for vote. For instance, in a hypothetical poll in which Joe Biden led Donald Trump 48% to 41% with a margin of error of 3%, within a 90-95% level of confidence, the actual result could actually be as close as 45% to 44% in Joe Biden’s favor as Donald Trump could have as much as 3% more than the poll suggests and Joe Biden could have as much as 3% less than the poll suggests. The result would be a stark contrast to what the poll originally showcased.
The 2016 election showed us a prime example of this phenomenon when, just days before the election, a poll came out in which Hillary Clinton led Donald Trump 49% to 45% with a margin of error of 4%. However, Donald Trump would go on to win the state 49% to 47%, a result still within the poll’s margin of error, thus handing Trump 29 electoral votes. So while unlikely, a distinct contrast between a poll and the actual election result is always well within the range of possibility.
Another potential explanation for the difference between polling results and actual vote count, according to some experts, is nonresponse bias. Nonresponse bias is when people systematically don’t respond to polls. As shown in a Politico article conducted in 2019, phone polling has dramatically decreased over the last several decades with the response rate to phone polls dropping from roughly 36% in 1997 to 6% in 2018, further showcasing that the polling industry may need to expand its reach to a larger share of the populace.
Another issue in the previous two elections could have been the “shy Trumper” situation in which voters answered differently when polled compared to who they ended up voting for in the actual election. According to data conducted by the Cato Institute in 2020, roughly “88% of Trump-voting graduates compared to just 44% of Clinton-supporting graduates agreed that ‘The political climate these days prevents me from saying things I believe because others might find them offensive.’
Is nonresponse bias affecting these 2024 election polls? Are people afraid to reveal their true opinions on these candidates? Or, are these results actually not as surprising as people would think given polls’ margins of errors?
As stated earlier, Donald Trump appears to hold a slight lead over Joe Biden both nationally and in many swing states, but, as previous election cycles have shown, these results should not be taken as truly representative of American citizens’ opinion. Will the former president end up taking back the White House? We won’t know for certain until election night itself.