I think the rubber stamp reputation of the FISA courts undermines this assumption. Are 99.7% of cases ~really~ ironclad? Building a rock solid case takes a lot of work and if the FISA courts aren't actually going to scrutinize the evidence, I expect that the people bringing these cases will take advantage of that.
12 denials from 33942 requests. 1 out of roughly 2828 requests instead of 282 requests.
Previously I had assumed it was a lot of rubber-stamping, however the folks I worked with frequently bemoaned about how incredibly difficult and time consuming it was to get a FISA warrant.
That being said, the approval rate for the court is pretty shocking. If this isn't a rubber stamp court, I don't know what is. Their approval rate on 41,000+ warrants in 40+ years is 99.97%.
I don't think the notion that each of those 41,000+ warrants were air tight, or wouldn't be rejected by other normal (read: public, transparent, democratic) courts, passes the smell test.
I would be curious to know what the approval rate for regular search/arrest warrants is too.
>> I agree that this is too low and is a sign of rubber stamp-iness, but what would be the right rejection rate to make you feel more comfortable about the FISA process?