Less reliable than luck of the draw

The New York Times announced their endorsement of the two candidates in the Democratic primaries. Some of my more traditional news sources such as the NYT itself and NPR tried to carry off the endorsement as if it were important and momentous. As if it carried consequence. It is easy to look at them as declining hacks but it is a fair question.

Does the NYT endorsement carry weight? Do people change their voting intentions based on it? As a skeptic, it is a knee-jerk response to say no, but I wonder.

Iowa is a caucus state and their caucus attendees may be more center left and therefore pay more attention to the NYT. I can't really imagine it has much sway in primary states. I wonder if anyone has ever done a study of the effect size, if any, of a NYT endorsement.

But thinking about their declining influence made me wonder about the NYT's history of endorsements. Two questions - 1) When was the last time they endorsed a Republican candidate in a presidential campaign?, and 2) How do NYT endorsements foreshadow a win?

Fortunately there is a Wikipedia page with the data.

In the 29 presidential elections since 1900, the New York Times has endorsed a Republican candidate six times (McKinley, Taft, Wilkie, Dewey and Eisenhower (twice)). Across those twenty-nine elections the candidate endorsed by the NYT won 66% of the time.

The NYT has not endorsed a Republican candidate for two generations, not since 1960, sixty years. In the fifteen elections since 1960 when the NYT has solely endorsed Democrats, what percentage of the elections were won by their endorsed candidates? 47% of the time. They endorsed Democrats fifteen times in a row but Democrats only won 7 elections.

Endorsing a candidate is a different function than forecasting a winner, but they are related activities. 47% is less of an achievement than simply random chance. A chimp picking from a white or black pebble from a covered basket would do as well.

The consistency and length of duration suggests that there is little room for real news, they are bad at forecasting, they are a committed partisan organization, or that they are out of touch with the American voter. Or some combination.

Not anything especially surprising but still kind of arresting.