Lies, Damned Lies, and Polls

Actually, I like polls. You can learn a lot from them. But increasingly, you can’t trust what people–even the pollsters themselves–say about polls. You have to read the raw data for yourself.
A case in point is recent polling by the Pew Research Center, which was called to our attention by reader Gene Allen. As reported by the Associated Press, the Pew data show that “the number of people who think the United States doesn’t have enough international support yet for military action [against Iraq] is growing.” Pew Research Director Andrew Kohut says that “the most important findings of the poll are that the debate with longtime allies in the United Nations, the report by U.N. weapons inspector Hans Blix and overseas peace protests ‘have affected public opinion.'”
Further, “recent events are raising public concerns about the need for international backing. The Pew poll found public sentiment shifting significantly between early February and after the United Nations heard from Blix.”
If you look at the poll’s raw data, however, the support for “taking military action in Iraq to end Saddam Hussein’s rule”–the most straightforward question in the survey relating to Iraq–has held remarkably steady, with the percentage favoring military action currently at 66%.
The most obvious change in Pew’s most recent data is the following:
In late January 2003, 68% said they favored military action. When those 68% were then asked, “Should we attack Iraq only if our major allies agree to join us, or attack Iraq even if allies do not want to join us?” the responses were:
Attack even if allies won’t join: 26%
Attack only if allies agree: 38%
Undecided: 5%
In the most recent survey, 66% favored military action. When asked the same follow-up question, the responses were:
Attack even if allies won’t join: 38%
Attack only if allies agree: 22%
Undecided: 6%
Thus, the net effect of weeks of antiwar demonstrations, the equivocal Blix report, and constant antiwar agitation in the press was that the proportion of respondents favoring an attack without allied support rose by nearly 50% and now represents a clear plurality.
It is a sad commentary that many of our pollsters are so dedicated to liberalism that they will distort their own polls’ findings to serve their political agenda, and that virtually all of our reporters are so dedicated to liberalism (or so lazy) that they would never dream of looking at actual data to determine whether pollsters’ interpretations are accurate. It falls to amateurs like Gene Allen and to bloggers like us to do the job that reporters should be doing.


Books to read from Power Line