Don’t Ask, Don’t Tell and the Problems with Polling
Congress may not be getting much done surrounding “Don’t Ask, Don’t Tell,” but at least the commentary has been good. Over the last few weeks, politicians and scholars have said some surprisingly smart things about the DADT poll given to our military and the problems with polling in general.
The DADT poll given to military personnel has been mired in controversy for months. Senator John McCain in particular has been extremely crabby about the fact that the poll asked troops to reveal their feelings about a potential repeal of DADT without asking for their opinion on whether DADT should be repealed. When the results of the poll were released last month, Defense Secretary Robert Gates once again defended this choice:
As was made clear at the time and is worth repeating today, this outreach was not a matter of taking a poll of the military to determine whether the law should be changed. The very idea of asking the force to in effect vote on such a matter is antithetical to our system of government, and would have been without precedent in the long history of our civilian-led military… Nonetheless, I thought it critically important to engage our troops and their families on this issue, as ultimately it will be they who will determine whether or not such a transition is successful. I believe that we had to learn the attitudes, obstacles and concerns that would need to be addressed should the law be changed. We could do this only by reaching out and listening to our men and women in uniform and their families.
Whatever your feelings about the DADT poll, it’s worth noticing what Gates is acknowledging in this quote. Gates knows what Gerard Hauser has been telling us for a while: polls don’t reveal what a group actually thinks; they “measure attitude, which often influences what [those conducting a poll] will say to satisfy their constituents’ needs.”
In other words, polls are an exercise in group psychology, not a method of political decision-making. Should the troops have a say in whether or not to repeal DADT? I’m not sure, although I do know that Truman didn’t poll military personnel before he desegregated the military, and that seemed to go ok. In any case, Gates’ remark reflects a clear understanding of what polls can and cannot do. They can reveal attitudes, but they cannot—and should not—be used to dictate the decisions of our leaders.
Another commentary on the DADT poll worth noting came from historian Nathaniel Frank in an interview with Terry Gross. Frank discusses surveys conducted by other nations seeking to end bans similar to DADT conducted in Canada and Britain:
…in both cases in Britain and Canada—and these are very large surveys—majorities said that they would not support serving with gays and lesbians. Some actually said… large percentages said that they would refuse to serve, that they would leave, that they wouldn’t work with gays and lesbians. And yet, when the transitions were made, almost no one left. So there’s a big gap between what people say and what actually happens.
Frank points out another limitation of polls: they often measure the hypothetical, and people are not always good at assessing what they will actually do in a given situation. Researchers in and outside of rhetoric have long been aware of the problems of asking people what they would do or how they would feel in a hypothetical situation. Moreover, Canada’s and Britain’s polls provide further evidence that asking soldiers how they will react to repeal may not provide any reliable predictions about the future. Those who want to keep DADT might well use this discrepancy to argue that the reaction to repeal would be more severe than the poll predicts, but historical evidence seems to indicate the opposite.
If you want to know more:
- I hesitated to attribute “crabbiness” to a respected politician, but there’s really no better way to describe Sen. McCain’s behavior whenever he is asked about DADT.
- My Gerard Hauser quote is taken from Vernacular Voices, page 27.
- The quote from Gross’ interview of Frank is taken from the audio version of the interview, not the limited transcript I linked to above.
- If you’re interested in the problems that come with using hypothetical questions and thought experiments in research, see the work of Daniel Dennett.
- As I wrote this piece, I kept thinking about all those people who said they would move to Canada if George W. Bush were elected or re-elected. I don’t remember seeing many moving vans the next day, do you?