When pollsters ask people for their view on whether collective bargaining should be limited for public employees should pollsters call collective bargaining a “right”?
At the Quinnipiac University Poll, this was the question we faced when we were going into the field two weeks ago with an Ohio poll as this issue was the big issue not just in Ohio but nationally as well.
Those who advocate limiting collective bargaining for public employees would argue that it is not a “right”.
Those opposed to such limits would say that collective bargaining is a “right”. This poses a dilemma for public opinion pollsters who are trying to accurately measure public opinion on this issue.
Should we use the word “rights” when framing the question?
Most news reports on the issue have characterized collective bargaining as a “right”. Similarly most polls on the subject have also referred to it as a “right”. Our concern that the inclusion or the lack of inclusion of that word could make a difference in the responses that we received led us to conduct a split simple experiment in our March 2011 Ohio poll.
We asked half the sample the question including the word “rights” and the other half was asked the exact same question without the word “rights”.
As we expected, there was some difference in the responses of the two groups. However, no matter how the question was asked at least a plurality of voters opposed limits on collective bargaining for public employees. By a 48-41 percent margin, voters opposed the limits when the question did not include rights. By an even greater margin of 54-35 percent, voters opposed limits when the word “rights” was included. This is a 12 point swing (from a 7 point margin to a 19 point margin).
What made this experiment unique was not that we did a split sample experiment to test the impact of different question wordings.
It was unique because I haven’t seen any other polls which have done such an experiment on this issue.
Drilling down a little deeper on the numbers, we found that not everyone was affected by the wording difference. Partisans, not surprisingly were largely unaffected by the question wording. This is not surprising because many issues are seen through partisan lenses. Certainly this issue has become partisan with the Democrats in the legislature lined up against the limits and the Republican Governor, John Kasich, and Republican legislators generally for the limits.
The independents were the ones affected by the wording change.
Another difference we found was gender. Women weren’t affected by the wording change but men were.
The voters that were most affected by the question wording were those less than 50 years old.
While 18-49 year olds supported the limits by a 2 point margin when the question didn’t include “rights”, they opposed the limits by a 27 point margin when the wording included “rights”. That is a huge 29 point swing.
To see the size of the swing among different groups see the table below.
Group Size of swing Gender Men 21 Women -4 Party ID Reps 7 Dems 4 Ind 19 Age 18-49 27 50+ -2 Education <Coll 15 College 4 Income <30K 10 30-50 -2 50-100 25 >100 20 Union 14