Political Polling 101

With an election year underway, we reprint an updated piece by David Preston of New Harbor Group, the Society’s government relations and communications counsel, on political polling during the campaign season that appeared on his company’s blog. To read more, click here.


Election year is upon us, and candidates are beginning to gear up for the November elections. And with campaign season upon us, that also means it’s time for more and more political “polls.” But what do those polls really mean – if anything?

There are two main reasons for a campaign to conduct a poll:

1. To find out information that you need to know

2. To get some degree of validation for information that you want others to know

Reason #1 is the political equivalent of market research. Some may find this idea unsavory because of worries that it could lead to pandering by candidates. Well, maybe. Certainly poll-driven candidates are out there, but I give voters more credit for seeing through them. In reality, it’s crucial for candidates to have good information about what voters care about in order to stay in touch with the electorate’s top priorities. If a candidate wants to talk about apples, but the electorate wants to know about oranges, the campaign’s in trouble. A poll can help a candidate know what voters want to know about – and address those concerns.

Unfortunately, polls with reliable numbers are hard to find and expensive to conduct. What makes it hard – and expensive - to conduct a good poll?

  • You need good callers who understand the questions and ask them in the right way;
  • You need callers who can be trained to pronounce candidate names and place names correctly;
  • Informative polls are long, and it’s difficult to get people to stay on the line all the way through to the end;
  • It’s harder to find voters in the era of the cell phone, which are off-limits to pollsters by law;
  • It takes time and real expertise to draft a sound, useful questionnaire;
  • It takes time and expertise to interpret the data, weighing it properly so it’s reflective of area demographics, and gives an accurate, useful view of public opinion.

In short, the kind of good data you need for the first kind of poll is expensive and difficult to come by – and becoming even more so.

Which brings us to the second kind of poll. These tend to be notoriously unscientific and rigged to score points for whoever is releasing it. Your radar should really go up if poll numbers are released by a campaign, a political party, an organization affiliated with one, or a special interest group.

In reality, all public polls should be greeted with healthy skepticism. Media outlets used to do them with the necessary rigor, but few can afford it anymore, with some national exceptions (New York Times, national networks, etc.). Local media outlets are, for the most part, unwilling to spend the money it takes to get good, sound data. They usually settle for on-line polls or automated telephone polls. What you end up with is very cheap data that barely passes the accuracy laugh test – but is reported by the media as near-gospel. The best information is usually found in the hands of well-funded candidates, and jealously guarded like the precious commodity that it is.

If you want to know who’s really up and who’s down, the best indicators are not public polls but the activities of the candidates. If they’re spending precious time and resources campaigning in neighborhoods, towns, counties or states they are expected to win, they’re in trouble. If they’re launching a desperate attack on an opponent, they’re in trouble. But if they’ve kept a consistent message, are counter-attacking from the high ground, and campaigning in areas that are considered a toss-up, they’re looking at good numbers.