If I may, Mr. Chair, I'd point the member to paragraph 2.21. We actually targeted 33 files. We didn't do a random sample, so we couldn't tell you that it was the whole population to try to get exactly what you're saying—why and where it slowed down. We listed some areas, but you'll see that sometimes it was that something was sitting in the wrong queue.... There are a whole bunch of reasons.
I point the member to that paragraph.