A new national poll from American Strategies finds that Joe Biden is one point ahead of Donald Trump in a head-to-head election (39% to 38%). Third-party candidates Robert F. Kennedy Jr. (7 percent), Jill Stein (1 percent), and Chase Oliver (1 percent) receive a combined 9 percent of the vote. Robert F. Kennedy Jr. voters are split in their preference for Biden and Trump; 36 percent say they would vote for Biden if Kennedy was not running, and 39 percent would vote for Trump. Twelve percent are not sure of their vote for president. In terms of intensity, 33 percent very certain of their vote for Biden, and 30 percent very certain of their support for Trump. Thirty-six percent are either not certain of their vote, undecided, or favor a third-party candidate.

Overall, this study interviewed 1,151 respondents for a margin of error of +/- 2.9%. Analysis at smaller, subgroup level, such as age groups or partisan groups will have a higher margin of error. As a note, the number of political polls will increase as the presidential contest and other races heat up in the weeks and months ahead. We encourage readers to take extra caution when interpreting results based on small subgroup samples, such as younger voters or African Americans. Many polls do not include the margin of error for the subsample, which will be much larger depending on the number of interviews in the subsample. For example, in the analysis below, Biden has 54 percent of the Black vote, but there are only 116 interviews in the subsample, with a corresponding margin of error of +/- 9%. Broken down, this means at a confidence level of 95% support for Biden among Black voters falls in the 45%-63% range. With the same confidence level support for Trump among Black voters falls within the range of 12%-30%. Interpretation of polling results should always take sample size into consideration.


With that being said, below are some key findings at the subgroup level. When broken out on party lines data suggests a similar story for both main candidates:

  • Self-identified Democrats (Margin of Error (MoE) +/- 4.9%), support Biden with 80 percent of the vote; 5 percent vote for Donald Trump, 7 percent for a third-party candidate, and 7 percent are not sure.
  • Self-identified Republicans (MoE +/- 5%), support Trump with 81 percent of the vote as well; 4 percent favor Biden, 5 percent vote for a third-party candidate, and 10 percent are not sure who they will vote for.
  • Independent voters (MoE +/- 5%) are statistically split between Biden and Trump (32 to 29 percent). Independent voters are more likely to support a third-party candidate (17 percent), or be undecided (19 percent)

Breaking respondents out by age data shows Trump with a modest lead among respondents aged 18 to 29 (37 to 30 percent). Younger voters are about as likely to support a third-party candidate, remain undecided, or likely not vote (34 percent). Biden performs strongest among seniors. He gets 49 percent of the age 65+ vote, compared to 36 percent for Trump. Trump’s support is strongest among middle-age voters, 40-49, with a 43 to 34 percent lead over Biden.

When it comes to race, White voters are split between Biden and Trump with each getting 40 percent of the vote each. Biden’s support is strongest among African American voters, with 54 percent of the vote. While Trump has a 6-point advantage over Biden with Hispanic and Asian voters, the large margin of error for these smaller subgroups suggests a close and fluid race among these voters.



Survey Methodology

American Strategies and Statara designed and administered this multi-modal survey. Data were collected through a randomly selected online non-probability panel sample provided by Dynata Research. The survey reached 1,151 adults, age 18 or older. The survey was conducted April 11-23, 2024.

Quotas were assigned to reflect the demographic distribution registered voters in the United States. The overall margin of error is +/- 2.9%. The margin of error for subgroups is larger and varies. Percentage totals may not add up precisely due to rounding.