What do Americans think of government snooping into telephone and Internet data? The answer depends very heavily on precisely what is asked.
The Pew Research Center ran an experiment this month to find out how much the wording of questions would affect public attitudes toward the widely discussed data-collection operations of the
Pollsters asked one group whether they "would favor or oppose the government collecting data, such as date, time and phone numbers, from nearly all phone calls made in the U.S., with court approval as part of anti-terrorism efforts."
A second group got the same basic question, but without the reference to "court approval." Another group got the question, but without the reference to "anti-terrorism efforts."
For another group, the pollsters asked not about collecting "data, such as date, time and phone numbers," but "recordings of nearly all phone calls." A parallel set of questions asked about collecting data or contents of emails, rather than telephone calls.
The results suggest that Americans retain a great deal of faith in their court system. Support for the hypothetical program was significantly higher when the question included the reference to "court approval." The approval of a court mattered even more than the idea that the information dragnet was "part of anti-terrorism efforts," a phrase that also mattered significantly.
Not surprisingly, more Americans supported collection of "data, such as date, time and phone numbers" than supported collecting actual recordings of conversations. But that shift in description only moved the results by six percentage points – half the shift that court approval brought.
The distinction between email and telephone calls made little difference.
Overall, support for the hypothetical program ranged from 41% when it was described as data collection with court approval and as part of an anti-terrorism effort down to 16% when the question asked about recordings of calls and did not mention either terrorism or judicial scrutiny.
One other important word: "would."
These questions all asked people "would you favor" a hypothetical program. At about the same time, Pew was running a separate survey asking people "do you favor" the government's existing collection efforts. In that survey, 50% said yes.
The results suggest that Americans are at least somewhat more likely to approve of something the government is already doing than of a hypothetical thing the government might do in the future -- a finding that may have implications considerably beyond data surveillance.