on question wording and push polls

For the past few years I have been placing questions on some statewide (North Carolina) polls on various areas, including human rights and collective bargaining for NC state employees. The last two were my own polls, carried out by Public Policy Polling under contract to me. As someone who cares about organized labor and has been active in related areas for a while, North Carolina’s General Statute §95-98, which prohibits collective bargaining by any public-sector employees groups, strikes me as draconian and backward. Some groups I like have worked to put HB1583 on the agenda, which would remove the ban on public employee collective bargaining.

Last week the John W. Pope Civitas Institute, which used to be a fairly interesting, principled Libertarian-oriented think tank on North Carolina Politics, but has recently veered toward toeing the Republican Party line, released a press release via their Red Clay Citizen blog claiming widespread opposition to allowing public employees their collective bargaining rights. Sadly, the question they used verges on push-polling:


Happily, I have been running my polls with a much more neutrally-worded question:

As you may know, the State of North Carolina, towns, cities, and counties and other public workplaces in North Carolina are not allowed to negotiate or sign contracts with labor unions representing their employees. Would you support or oppose a law that would allow them to do so?

…and, perhaps not surprisingly, my results are dramatically different.

The Progressive Pulse, a North Carolina progressive blog run by NC Policy Watch, picked up my results. So far nothing on either of the polling results from the state’s mainstream press, which is pretty much par for the course; collective bargaining has been essentially off their radar screens from the beginning.

I strongly believe that my polling is better for whatever definition of “better” you prefer, but the sociological geek in me is also impressed at how very different the results can be when you ask questions in different ways. Actually in this case it’s really a priming effect. The Civitas poll framed the issue as one of government corruption–a hot topic in North Carolina, but not one that anybody has seriously raised about collective bargaining–and so evoked respondents’ concerns about corruption. Mine asked first and foremost about collective bargaining.

Author: andrewperrin

University of North Carolina, Chapel Hill

5 thoughts on “on question wording and push polls”

  1. So, if the question had “the State” instead of “elected officials they might endorse in elections,” would that have seemed fair?


  2. Well it’s not necessarily the State, so that wouldn’t have been technically correct. But “their employers” or “state and local governments that employ them” or something like that, I suppose, would have worked fine.


  3. Wow – that was fast! A week on Scatterplot and you’re getting called out on another blog (and Scatterplot is being called your blog)! I am delighted to see you commanding a goodly share of the bloggerly attention market.


Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: