Monday’s Pew poll showing that a majority of Americans back the NSA’s phone-record harvesting drew 300+ comments so I figured you’d want to see this one too.

CBS’s numbers look … different:

Seventy-five percent of Americans approve of federal agencies collecting the phone records of people the government suspects of terrorist activity, but a 58 percent majority disapproves of this type of data collection in the case of ordinary Americans.

Majorities of Republicans and independents oppose the government collecting phone records of ordinary Americans; Democrats are divided.

Democrats, who were heavily in favor of tracking domestic calls to fight terror in Pew’s poll, split 48/48 here on collecting the phone records of “ordinary Americans.” Republicans oppose it, 33/66, and independents oppose it, 34/62 — also sharply different from Pew’s poll, where they split 52/47 and 55/43 in favor, respectively. Interesting footnote: Even though Americans disapprove of phone-record collecting overall here, 38/58, a small majority say that collecting those records is nonetheless necessary to fighting terror. Lindsey Graham seems to think that if something is necessary for counterterrorism, it’s automatically worth doing. CBS’s numbers suggest that Americans disagree.

Why are these numbers so different from Pew’s? Two possibilities. One is the timing of the polls. Pew’s was conducted between June 6th and 9th. CBS’s was conducted on the 9th and 10th. The news about PRISM broke on the 6th, so maybe some chunk of Pew’s sample hadn’t yet formed an opinion about it. A few days of hostile media coverage later, with news shows filled with chatter about the surveillance state, and now maybe O’s got a backlash on his hands. The second possibility, as always in polls, is differences in how the pollsters word their questions. Pew framed its question about data-mining phone records by asking if it’s acceptable or not for NSA to get “secret court orders to track calls of millions of Americans to investigate terrorism.” CBS framed its own question by asking if people approve of federal agencies “collecting phone records of ordinary Americans.” No surprise that Pew saw stronger support by emphasizing that this is all being done with court supervision, and no surprise that respondents would grimace a bit more at the thought of all this when CBS emphasized that “ordinary” Americans are being targeted. Polling data on this subject is, I suspect, unusually sensitive to tweaks in the wording of questions, partly because people are so torn about the privacy/security interests involved and partly because the subject is technically and legally complex. My hunch is that Pew’s results are closer to the truth than CBS’s, but we need more data — and more precise descriptions of what the NSA’s doing from poilsters when they ask about this — to know for sure.

As for why I think Pew’s closer to the truth, chew on these two tidbits from CBS’s poll:

concernedbalanceaboutright

When you ask people if they’re concerned about losing some of their privacy due to counterterrorism, 59 percent say they’re “very” or “somewhat” concerned. But as you see above, when you ask them if they’re concerned about the government collecting their own personal phone records, you get just 38 percent who say “very” or “somewhat” and 62 percent who say “not much” or “not at all.” Check the crosstabs at CBS and you’ll see that there’s no strong partisan split on that question either: 61 percent of Republicans and 67 percent of Democrats say they’re “not very” or “not at all” concerned. Likewise, the second data set above shows that despite their disapproval of NSA record-collecting, there are nonetheless majorities among Democrats, Republicans, and independents who say the current balance in infringing on privacy to fight terror is either “about right” or doesn’t go far enough(!). What we’re really looking at when we examine polls like this is whether a broad, sustained backlash among the population that would force O and Congress to scale back NSA surveillance is likely. Do those results suggest that it’s likely?

One more bit of data for you, this time from a separate YouGov poll released this morning:

bo

You can take that first question either way. More people think O’s doing a bad job protecting constitutional rights than a good job, but if you toss in the people who think he’s doing okay/fair, you’ve got 49 percent who aren’t too troubled versus just 45 percent who are. (There’s no partisan split at the link but you can guess what it looks like.) The second question is less ambiguous: Just one-third think he’s gone too far in prioritizing counterterrorism over civil liberties. Forty-five percent think he’s either struck the right balance or hasn’t gone far enough. The YouGov poll was taken on June 7th and 8th so, again, maybe the full impact of the PRISM revelations hadn’t been priced into the data yet. More polls will tell us for sure, but this eeyore is skeptical that they’ll look much different.