Today brings some important new polling data on the issue of the Employee Free Choice Act (EFCA) now making its way through Congress. The release from Gallup is getting the most attention, but new findings from Rasmussen Reports and from the pollster for the AFL-CIO are also worth a review. Collectively, they show that the challenge of polling on this issue is in trying to predicting the substance and reach of the coming debate.
Gallup released results from two questions. Read their summary for the relevant details, but there are four key findings:
- Very few Americans are paying attention - Only 12% say they are "very closely following" news about "a bill in Congress that would change the rules governing how unions can organize workers," a level that is "exceptionally low relative to public attention to other news issues Gallup has measured over the last two decades."
- A majority (53%) favor and 39% oppose "a law that would make it easier for labor unions to organize workers."
- Democrats are much more supportive of such a law (70%) than Republicans (34%).
- Those paying very close attention oppose such a law by a wide margin (40% to 58%). "The bill enjoys its highest support -- 58% -- among those not following the bill at all."
Rasmussen Reports also released new results (article, toplines) including questions that attempt to measure current attitudes (they also tested questions that gauged reactions to specific provisions in the bill - more on that to follow);
- A majority of Americans (57%) say it is at least somewhat difficult "for workers to form a union" if they "do not have a labor union at their company" (only 19% say it is not very or not at all difficult).
- Contrary to the Gallup finding, 33% agree and 40% disagree that Congress should "change the law to make it easier for workers to join a labor union."
Gallup's headline finding about how "receptive" Americans are to union organizing is driving the discussion today, but findings #1 and #4 above are probably the most important. Here is why.
The biggest challenge in polling on an issue like EFCA is that so few Americans are aware of the actual bill or the ongoing conflicts between organized labor and management. Pollsters rarely ask the sorts of questions that measure knowledge (like those used in this report), but I would wager that the number of Americans genuinely familiar with EFCA is no greater than the 12% who say they are "closely following" the bill. That lack of awareness can create a lot of inconsistency in poll results.
Take the apparent contradiction between Gallup (54-39% favor/oppose) and Rasmussen (33-40%) on bills that would make it easier, respectively, for unions to organize or for workers to join a union. We could do an entire post on the potential reasons. It may be that Rasmussen's screening for likely voters, and perhaps their use of the automated methodology, reach a sample that is more informed and interested in the EFCA bill and thus, (given the Gallup finding) more prone to express opposition. The difference in the "don't know" category -- 8% on the Gallup question compared to 27% on Rasmussen's version -- is also important.
But consider that while two questions ask about the same bill, they may sound very different to Americans not following the debate. In other words, a law that makes it "easier for labor unions to organize workers" may sound like something different than a changing a law to "make it easier for workers to join a labor union."
If subtle changes in wording can produce such different results, then we can assume that many respondents are forming opinions on the spot rather than sharing pre-existing views on the actual legislation. "Public opinion" in this sense isn't so much "fluid" (a favorite pollster cliche) as non-existent.
That conclusion raises an important but difficult to answer question: What will ultimately be more influential with members of Congress, the opinions of those currently familiar with EFCA or the opinions that may develop as Congress debates this bill? The answer is probably the latter, but only if the debate merits sufficient news media coverage to heighten interest and solidify or change opinion.
The legendary political scientist V.O. Key had this sort of issue in mind when he defined public opinion as the "opinions held by private citizens which governments find it prudent to heed" (a passage I quoted last month). In Public Opinion and American Democracy, Key also wrote about the notion of "latent" opinions:
[I]n the practice of politics and government, latent opinion is really about the only type of opinion that generates much anxiety. What opinion will develop about this prospective candidate? What opinions will be stirred by this legislative proposal? What opinions. anxieties, and moods will be generated by this event or by that action?
Key also explored the problem of whose opinions might matter most:
[E]stimation of latent opinion is not simply estimation of the direction of majority sentiment. The problem extends to the estimation of what sectors of society, influential or non-influential, will have opinions. It involves the question of whether an issue will command the attention of a large public. On such estimates from time to time may turn decisions that could scarcely withstand the uproar that would be created if they came to universal attention.
And this brings us to the questions asked by Rasmussen and by Peter Hart Research for the AFL-CIO (as blogged today by my colleague Marc Ambinder) that explicitly inform respondents about various aspects of the legislation and test reactions (I have copied the text and results from each after the jump). Not surprisingly, the questions show very different results. While I am sure that both pollsters consider their questions "fair and balanced," partisans on both sides will see evidence of bias or leading language.
Whatever your judgement about the most accurate and truthful rendering of the arguments for and against EFCA, the huge challenge in testing these "latent" opinions is in predicting two hard-to-predict things: (1) the substance of the messages that will be communicated by the news media and paid advertising and (2) the degree to which any of this debate ultimately reaches ordinary Americans. Remember that any "message test" question ultimately assumes, unrealistically, that 100% of Americans (or likely voters) will hear and process the information tested.
Consider the example of the 2007 immigration bill. In the spring of 2007, many surveys asked questions that tested reactions and found "broad support" for the provisions of the bill. Yet the day before the key Senate vote, Gallup released a survey showing that only 18% were closely following news about the "bill to deal with the issue of illegal immigration" (not much more than are following EFCA now). They also followed up with a question asking of respondents favored or opposed "this proposed bill," and found that the majority (58%) chose the third option, "or don't you know enough to say." Among those with an opinion, Gallup found nearly three-to-one opposition (30% to 11%). Which constituents were most likely to reach out to their representatives? Which opinions did members of Congress find it "most prudent to heed?"
So the most critical question regarding EFCA may not be which question best captures the "truth" of the substance of the bill. It may be about how much the substance of the debate ultimately reaches ordinary Americans. Supporters of EFCA must do more than win the message war. They must also reach and mobilize opinion among Americans that are ordinarily inattentive to politics.
The text of the "reaction" questions follows after the jump
From Rasmussen Reports:
3* Under current law, if enough workers express interest in forming a union, a secret ballot is held. Is it fair to require a secret ballot to determine if workers want to form a union?
22% Not sure
4* Some people believe that a secret ballot vote is not necessary and that a union should be formed whenever a majority of workers sign a card saying they want one. If a majority of a company’s workers sign a card saying they want to form a union, is it fair to form a union without having a vote?
16% Not sure
5* Suppose a company and their employees union cannot reach an agreement on a contract within 90 days. Should the government be allowed to mandate an agreement defining pay and benefits for that company’s employees?
24% Not sure
From Peter Hart Research/AFL-CIO (via Marc Ambinder):
I am going to read you two statements about this legislation, and please tell me which one you agree with more.
Supporters say that the system is broken and working people are struggling to make ends meet today, and the middle class is being squeezed. One way to help average people get their fair share is to let them bargain with their employers for better wages and benefits. Workers in unions earn twenty-eight percent higher wages on average, are sixty-two percent more likely to have employer health coverage, and four times as likely to have a pension. It's time our economy worked for everyone again.
Opponents say that this legislation is a bad idea because it would abolish the secret ballot elections now held to determine union representation. This legislation would force more workers into unions, because union bosses can use coercion or deception to collect authorization cards. And with our economy already weak, we don't need laws that give more power to the unions that wrecked the American auto industry.
54% support card check