Gen. Keith B. Alexander, director of the National Security Agency. (J. Scott Applewhite/Associated Press) Gen. Keith B. Alexander, director of the National Security Agency. (J. Scott Applewhite/Associated Press)

The question shouldn’t be whether the National Security Agency’s massive secret effort to log all of our phone calls is useful. Of course it’s useful. The question should be whether the program is necessary – and officials haven’t yet shown that it is.

Useful vs. necessary is an important distinction. I believe it’s also important to distinguish between the NSA’s vacuum-cleaner collection of telephone “metadata” pertaining to U.S. citizens — which still has my jaw on the floor — and the separate PRISM program that apparently selectively harvests the e-mails and other electronic communications of foreigners.

From what we know so far, PRISM seems less likely to infringe on Americans’ privacy. True, it seems much more likely to infringe on the privacy of anyone not in the United States, but that’s something Angela Merkel, Vladimir Putin, Francois Hollande and the others perhaps should have taken up with President Obama at the G-8 meeting.

Phone data collection creates an immense cache of information about our contacts, associations, movements, habits. The program is not in the least “transparent,” as Obama claims, but I will stipulate that there are laws, rules and safeguards in place, and that these are enforced by people of good will, even if all the enforcement takes place in secret. The surrender of privacy is so great, in my opinion, that it is only justified if it is necessary.

The director of the NSA, Gen. Keith Alexander, told Congress Tuesday that the two programs “have protected the U.S. and our allies from terrorist threats across the globe to include helping prevent the terrorist — the potential terrorist events over 50 times since 9/11.” We need a lot more clarity, please.

How many of these “potential terrorist events” were thwarted because of the phone metadata rather than the PRISM program? Of those in which the phone data played a role, how big a role? Were there cases in which phone data gave analysts the first hint of a terrorist plot? Or was it used instead to confirm information that came originally from other sources?

You see what I’m driving at. If compiling a log of everyone’s phone calls is alerting us to genuine terrorist plots we wouldn’t see otherwise, then those who argue this is something the government should not be doing — as I contend — have a much higher burden of proof. But so far, I haven’t heard anyone make this claim in anything resembling plain language.

In his interview with Charlie Rose, Obama said that “at the margins we are increasing our chances of preventing a catastrophe … through these programs.” Something that contributes at the margins is certainly useful. But is it necessary?