Alright, I looked at the poll, its DEEPLY flawed, firstly there were only 875 (434 men, 429 women, which only adds up to 863, with no other gender options listed as responding, where are the remaining 12 people?) respondents, that is an absurdly low sample size for all voters across all states, secondly, the responders aren't indicative of the general population, 598 (~71%) of the people who responded to the mentioned question were white while only 241 (~28%) of the respondents were any other race, according to census.gov, the percentage of 'White alone, non-Hispanic' population as of 2020 is 57%.

There are other issues I don't have time to go into, like that they mainly collected data from phone calls and that they aren't listing the distribution of respondents across states, but all in all I would not use this poll to draw any conclusions, and the article being titled like that is at the very least being disingenuous if not actively malicious in order to generate clicks/impressions