Announcing My New Column in The Washington Post

Gift link included

I am so excited to announce that I will be writing an opinion column for the Washington Post over the course of the next year. I will be posting previews of each column with a gift link here.

Here is my first column, posted yesterday!

Opinion

Trump’s Black voter bump misses the big picture

The data is real, but the conventional analysis is misleading.

By Annie Duke

October 30, 2024 at 7:30 a.m. EDT

Annie Duke, a cognitive scientist and former professional poker player, is the author of “Thinking in Bets” and "Quit: The Power of Knowing When to Walk Away.” She will spend the next year writing a column on risk and forecasting.

All the polling supports the buzziest statistical claim of the 2024 election: Black voters are moving toward Donald Trump. An October 2020 Pew Research survey found that Black voters favored Joe Biden by a margin of 81 points, 89 percent to 8 percent. But four years later, Vice President Kamala Harris’s margin of support among this group is just 65 points (79 percent for her vs. 14 percent for Trump). In an extremely tight election, a 16-point drop seems like a very big deal.

My expertise isn’t in predicting election outcomes, so I won’t attempt to do so using this data. I’m a decision scientist and a retired professional poker player. I study how we make decisions and how we can train ourselves to make better ones. The conversation around Black voters is an example of one theme I plan to explore in this year-long column for The Post: how our instincts about data can lead us to draw the wrong conclusions.

For all the talk about misinformation these days, misinterpretation of factually correct information is what keeps me up at night. The research backs up my concern. A May study by researchers at MIT and the University of Pennsylvania found that information about coronavirus vaccines that passed a fact check but was misleading was much more consequential than misinformation in driving a potentially bad, high-stakes decision. How much more consequential? Try 46-fold.

Click here to read the rest.