Are Black voters abandoning the Democratic party? | Steve Phillips

Are Black voters gravitating towards Donald Trump in meaningful numbers? Recent polls and articles, most notably the New York Times/Siena College poll, have set off alarm bells across the country with their survey results finding the former president garnering historic levels of support among African American voters.

The short answer is no, there is little credible evidence showing a meaningful shift in the levels of support Black voters give to Democrats. The longer answer is that the current chorus of media concern illuminates serious shortcomings in polling methodology, political interpretation of polling data, and the responsible communication of information to the public in an ostensible democracy.

Ground zero for the latest round of articles and commentary is the 6 October New York Times Siena College polling that found Kamala Harris attracting support from just 78% of Black voters. By contrast, the 2020 exit polls found that four years ago the Biden-Harris ticket was backed by 87% of African Americans. These recent poll results have spawned lengthy articles, extended podcast discussions and widespread water cooler conversation about just what’s going on with Black voters.

There are four fundamental problems with relying on this single poll’s findings.

First, the poll’s findings are contradicted by other contemporary polls sponsored by major news organizations. In the CBS News/YouGov poll conducted the week after the Times poll, Harris is backed by 87% of Black voters. NBC’s latest poll, concluded 8 October, pegs Harris’s Black number at 84%. The 8 October ABC News/Ipsos poll has Harris’s Black support at 82%.

Second, there is a glaring methodological anomaly buried in the underlying data that raises serious questions about the cultural competence of those conducting the poll. When I was writing my first book Brown is the New White, I was guided by the data scientist Dr Julie Martinez-Ortega, who helped me grapple with the elemental, but actually quite complicated, question of which people in the population you include in the category of African Americans. The Census Bureau has two designations: “Black Alone” and “Black in Combination”. Martinez-Ortega schooled me to ignore the “in combination” grouping as it would yield a distorted picture of the Black population.

Unfortunately, it appears that the New York Times received no such guidance, and their polling data includes truly astounding findings about the differences between Black Alone and Black In Combination. Among the Black Alone population, Harris receives the support of 80% of that group, but the Times finds that just 60% of the Black In Combination grouping supports Harris. Such a dramatic divergence of results from subsets of what are supposedly the same sector of the population raise profound questions about the entire data set.

Which leads to the third problem: the lack of common sense. Good scientists look at the data they gather and then step back and try to make sense of it all. If a subset of data falls outside of the realm of decades of prior empirical findings, then one has to ask some basic questions.

Is there a dramatic difference in the electoral behavior of African Americans and mixed-race Black people? Given that no Democratic presidential nominee has ever received less than 83% of the Black vote, how likely is it that Harris – a Black woman (and also an Asian woman) – is backed by a smaller share of the Black population than Michael Dukakis, John Kerry, and a plethora of more mainstream Democratic presidential nominees? How do you reconcile these findings with the demonstrable and quantifiable Black enthusiasm for Harris in the form of the early Zoom calls that engaged tens of thousands of African Americans in a mere matter of hours; the 10,000 people who turned out to the early rally in heavily-Black Atlanta, Georgia; and the historic fundraising numbers that have recently eclipsed $1bn? Which set of information is the outlier?

Tied to the lack of common sense is the fourth problem: insufficient cultural competence. If someone who had deep and extensive experience with African Americans looked at the polling data, they could quickly spot the glaring anomalies. Black mixed-race people aren’t dramatically more pro-Republican than other African Americans. The enthusiasm and excitement demonstrated in the early Zoom calls flow from enduring a lifetime of disrespect and discrimination, and that excitement of finally having someone who shares your cultural experience in the White House isn’t going to dissipate over a few weeks.

Now, to be sure, there are differences in Black voting behavior along gender lines. The most apt example to Harris’s situation occurred in Georgia when Stacey Abrams, an African American woman, won the Democratic nomination for governor. In both of Abrams’s runs, there was a nine-point difference between her support margins from Black women and Black men (97% from Black women in 2018 and 88% from Black men). And cultural competence and common sense would not find this surprising as sexism is embedded in the fabric of this nation and affects all racial groups.

But even in Abrams’s case the vast, vast majority of African Americans backed the Black woman’s bid, and the pertinent and reliable data suggests that a similar phenomenon is unfolding in the 2024 presidential race. The overwhelming majority of Black Americans voted for Democrats in 2020 and in the 2022 midterms (86% according to the exit polls). That enthusiasm is only amplified for Harris.

• Steve Phillips is the founder of Democracy in Color, and author of Brown Is the New White: How the Demographic Revolution Has Created a New American Majority and How We Win the Civil War: Securing a Multiracial Democracy and Ending White Supremacy for Good