One of the conclusions of an AAPOR paper on the 2016 US polls contains this detail:

As this report documents, the national polls in 2016 were quite accurate, while polls in key battleground states showed some large, problematic errors. It is a persistent frustration within polling and the larger survey research community that the profession is judged based on how these often under-budgeted state polls perform relative to the election outcome.

The paper does support a lot of its conclusion with statistics, but there's no actual data about (poll) funding in it. So, is there some other data (in the form of say funding comparisons [which would have to take electorate sizes into account, so not exactly trivial to make fairly], or [better] funding vs accuracy comparisons) supporting this idea that under-funding is a substantial factor affecting the accuracy of US state-level polls?

(N.B. I've also asked this on politics SE, but it doesn't seem to be an interesting question for that audience... and the gist of the question seems suitable here as well, even if it's not a definitive claim as to the [main] cause. So a correlational answer would be quite acceptable.)

  • 3
    I'm not sure what the reason is to doubt this claim. Of course if you throw enough money at the problem, you can improve poll accuracy (hire more people, poll more people, process data better). – user5341 Aug 6 '18 at 12:31
  • @user5341: sure, more money buys better polls. The interesting claim here is that there wasn't enough money spent (in some places). – Fizz Aug 6 '18 at 12:44
  • 2
    State polls are not like State health agencies, they are not run by the state. The polls are conducted by the media companies, the fourth estate that "wields an indirect but significant influence on society". So maybe this paper is saying the business men backing the Democrats didn't spend enough to manipulate people in the fly over states. – daniel Aug 6 '18 at 20:13
  • 1
    This seems like it might be a more interesting question for SE.Politics or something, but as far as SE.Skeptics goes, the issue of whether or not polls are "underfunded" seems like a matter of opinion. I mean like @user5341 said, you can throw more money at the problem to get better results. And if you don't like the current results, you can very reasonably argue that having spent more money on the polls could've led to better polling results. But whether or not the lack of such increased funding constitutes being "underfunded" seems like a matter of policy opinion. – Nat Aug 28 '18 at 5:20
  • 1
    On SE.Politics, a good answer might try to do a societal cost/benefit analysis of the expected improvement in polling results leading to increased societal benefits, then lay out an argument for whether or not the expected benefits would be worth the cost. That'd seem factual enough to constitute a non-opinion-based answer for SE.Politics. – Nat Aug 28 '18 at 5:30

You must log in to answer this question.

Browse other questions tagged .