What the New York Times Democratic Primary Endorsement Got Wrong

The New York Times editorial board came under criticism for its dual endorsement of senators Elizabeth Warren and Amy Klobuchar. What bothers us about the editorial has nothing to do with their choice, but with an introductory statement that suggests the board misunderstands public opinion research and its value. Amber's letter to the editor breaks it down.

To the editor:

As a public opinion researcher, I was taken aback at the editorial board’s glib claim that “traditional polling [is] in tatters,” underscoring a challenge to “foretell what voters want.”

The Times has demonstrated a commendable level of transparency and diligence in determining which Democratic candidate would earn the paper’s endorsement. That makes this unsupported – and false – statement all the more surprising.

An evaluation conducted by the American Association of Public Opinion Research (AAPOR) found that national polls leading up to the 2016 election “were among the most accurate in estimating the popular vote since 1936.” True, shifts in public behavior make capturing attitudes more difficult (e.g., transitions from landline phones, unwillingness to answer calls), but survey methodologists continuously refine their approaches to address these issues.

But more importantly, predicting election results is one of the most difficult and, in my view, least valuable services pollsters can provide. Well-designed surveys explore the values underpinning policy preferences and the importance of leadership traits. Taken together, these measures can indeed indicate what voters want in their next president.

The Times does its readers a disservice when it casts doubt on one of the best tools democratic societies have for understanding public attitudes.