I recently wrote the following to friend and journalistic superhero Josh Marshall, in the context of his discussions of a recent ethical kerfuffle in the world of political polling:
Have you ever considered moving toward an argument that the obsession with polling itself--a novel development of the last several decades--is, itself, politically, morally, and civically damaging? That people of civic good will shold make them as small a part of their practice as is practicable? Not sure if you saw this piece of mine from 2024 in which I document atht polling in presidential general elections is coutnerproduct tive in prediction the outcome about half the time. So why does it persist? In my judgement, it is (1) therapeutic, stanching existentailly paintful doubt about the future that is central to the human condition; (2) profitable (I need not say much), and (3) a useful source of grist for those seeking to manipulate public perception more generally (reaching its apotheosis with the present-day practice--though with its roots in figures like the execrable Mark Penn—of commissioning polls, and, sometimes, "polls," precisely to do that.)
I also pointed him to a piece I did in this regard in 2024. Also check out the piece on Mark Penn linked above, from 2007, and note that, soon after, Hillary Clinton picked him as her top strategist for er first presidential campaign.
Anyhoot, Josh responded today. Let’s discuss!
I am more with you than with him on not quite the same grounds.
The share of reportage and discussion and attention devoted to results of under-specified polls is enormous, and takes attention away from more reporting and analysis of more substantial matters, like indeed policy. I don’t think what people are interested in is independent of what they have easily accessible information on— nor of what they seem to be taught they should be interested in.
When it comes to politics, in particular, I think that a lot of USians come to reporting to be Responsible Adults or to be The Smart Guy— and that what gets reported teaches them what constitutes those identities.
I have asked myself what/whom polls are good for, apart from whatever entertainment value they may have. And I have two answers.
1. Politicians. A politician in office or on the campaign trail can benefit from a poll that showcases constituent and national concerns, to know where to spend more resources on making contact and/or explaining a policy concern and stance. Or, you know, changing their minds about policy in response. I suppose that with our gangstery campaign finance systems poll results also *have to* play a role in contribution-seeking— as obvious they’re doing.
2. People in the gambling industry. Which has hypertrophied.
I'm with RP on this one. Polling (and sampling) has never caught up to the internet/cell phone age and is easier than ever to manipulate. The questions never allow nuance or out of the box reasoning. The world is still analog. Polling attempts to reduce it to digital streams of ones and zeros. Prediction markets feed off polls and are even more subject to fakery.