If you eat at Arby's, you are probably not a liberal. If you own a dog, a flashlight, and glass bakeware, you are almost certainly white.
Our consumer habits — as well as media consumption, social attitudes, and the way we use our time — correlate strongly with demographic identifiers of gender, income, race, ideology, and education. So strongly do they correlate, in fact, that two University of Chicago economists, Marianne Bertrand and Emir Kamenica, were able to create an algorithm that can predict an individual's demographic details if given information on what they buy, watch, think, and do.
As explained in a working paper published with the National Bureau of Economic Research and reported at The Washington Post, American liberals in 2016 (the most recent data year) are most easily identified by what they don't buy:
And white people are really into pets:
The predictions are based on a binary analysis: how each item correlates with being high- or low-income, white or nonwhite, male or female, and so on. Thus, the Post explains, "'doesn't own a pet' predicts that someone isn't white just as strongly as 'owns a pet' predicts that someone is white."
The study used data from 1992 through 2016 and found some interesting changes along the way. In 1992, for example, use of Grey Poupon mustard was the single best predictor of high income. Now, it's iPhone ownership. Read more at The Washington Post here. Bonnie Kristian