I have been writing for some time about the huge discrepancy between the number of wiretaps reported annually by the Administrative Office (AO) of the US Courts and the numbers reported by phone companies and online service providers in their transparency reports. It never occurred to me that the AO might be at fault for some of the apparent under-reporting of wiretaps. But it is true -- for all the detail, the Wiretap Reports are not so transparent and apparently not so accurate due in part to AO reporting rules.
To refresh the magnitude of the under-reporting, the AO reported 3,554 wiretaps in 2014 while the four major U.S. telephone carriers alone reported 10,712 wiretaps for the same period -- a threefold discrepancy. For 2015, the AO reported 4,148 wiretaps in its Wiretap Report, but the gap widened with service providers reporting at least 11,944 orders during 2015. As I previously wrote here and here, there are no good explanations for the gap, and the AO has offered no explanation despite being asked. It appears that some of the problem may rest with the AO’s own reporting rules and the practice of prosecutors to exempt themselves from providing complete numbers despite the statutory mandate to do so.
Section 2519 of Title 18 requires any judge who issued, denied or extended a wiretap order during the preceding year to make a report to the AO each year. There are no exceptions in the statute for reporting the actual number of wiretaps authorized by the courts. None.
Federal and state prosecutors likewise are required to report wiretap information each year. It is mandatory and there are no exceptions to reporting here either. In June of each year, the AO is required to provide Congress “a full and complete report concerning the number of applications for orders authorizing or approving wiretaps … and the number of orders and extensions granted or denied during the preceding calendar year.” In sum, all orders issued, denied or extended in the prior year should be reported to the AO and ultimately to Congress. No exceptions.
Here’s what’s actually happening. Let’s use the 2015 Wiretap Report as illustrative. On its face, it purports to cover intercepts concluded between January 1, 2015, and December 31, 2015. But in the summary section, the AO says that “[f]ederal judges and state judges authorized 958 wiretaps and 411 wiretaps, respectively, for which the AO received no corresponding data from prosecuting officials.” In other words, prosecutors failed to meet their legal reporting obligations under Section 2519 for 25% of the wiretaps they conducted the prior year.
However, the AO actually has permitted this late-reporting practice. The Wiretap Report explains: “Some of the prosecutors’ reports were received too late to include in this document, and some prosecutors may have delayed filing reports to avoid jeopardizing ongoing investigations. Information about these wiretaps should appear in future reports.” As noted above, there is nothing in the statute that authorizes a prosecutor to delay reporting wiretap data let alone based on his or her own determination that doing so will jeopardize an ongoing investigation. So how do prosecutors get away with it?
Remarkably, the AO’s Wiretap Reporting Instructions state that no report is required “when... a federal wiretap case is sealed for security purposes (i.e., terrorist cases) or as part of an ongoing investigation.” Section 2519(c) certainly permits the AO to promulgate regulations “dealing with the content and form of the reports required to be filed.” But again, the statute provides no basis for delayed or non-reporting for security reasons or otherwise. The AO rules should follow the statute, not amend it, especially because non-reporting skews the actual reporting to Congress and analysis significantly. How significant is it? Table 8 of the 2015 Wiretap Report shows that “additional costs of $121,869,918” arose from wiretaps between 2004 and 2015 that were reported late. How late? Who knows.
The AO says that the withheld data “should” be supplemented by prosecutors later, but there are no rules or requirements for doing so. One might surmise that reporting should occur when the investigation is complete, or a crime charged, or trial had, but the AO rules are silent and there is nothing in the data to indicate when the data was supplemented.
Moreover, the use of so-called “Supplemental Reports” by prosecutors to report wiretaps that should have been reported in prior years as required by Section 2519, is at best a suspect practice and mostly misleading. The Wiretap Report and the AO rules imply that supplementation is for previously reported wiretaps where costs or outcome of trials wasn’t available. Neither the Wiretap Report nor AO rules state that some wiretaps will first be reported in a Supplemental Report. Indeed, contrary to being required as the AO states, Section 2519(2) does not even address or authorize Supplemental Reports.
The data on Supplemental Reports in Table 8 of the Wiretap Report is even more troubling because it undermines the reliability of the AO wiretap totals. For example, it reveals that 926 wiretaps for 2014 were not reported by prosecuting attorneys, presumably because prosecutors decided that to do so would jeopardize an investigation. But the data does not reveal whether there are more that prosecutors have failed to report, or when they might be reported if so.
It is not even clear at this point whether these 926 wiretaps were included in the reports filed by authorizing judges. The number of wiretaps for 2014 could be as reported at 3,554, or as high as 4,480 if you add these “supplemental” wiretaps to the reported figures, or somewhere in between if some of the 926 wiretaps were in fact reported but new details as to cost or arrests/convictions became available after the reports were filed, or worse, greater than 3,554 but by how much we don’t know.
Part of reason for distrust is that the AO Guidelines actually permit the prosecutors to prepare the judges’ reports for them. As noted above, no report is required from prosecutors themselves in cases where the investigation might be endangered. So do prosecutors not include those numbers in the reports they prepare for the judges? Who knows. I asked the AO, but no response so far.
What is clear is that the AO authorized prosecutors to not report for certain cases despite the clear requirements of the statute, and prosecutors now fail to report in a significant percentage of cases. The AO instructions contain no clear direction on when delayed reporting must be cured, and the supplemental report summary by the AO gives no indication as when the supplemental data was received or how many reports remain outstanding.
And what clearly remains is a significant chasm between what the AO reports and what providers have revealed in terms of total numbers of wiretaps implemented in any given year. At this point, there can be no confidence in the AO’s reports, and the AO seems uninterested in getting to the bottom of the discrepancies or determining how many wiretap orders prosecutors have held back from reporting over the years.
It is time for the AO to clean up the discrepancies and require prosecutors to file timely and complete reports so that Congress and the public are provided full and complete information about the cost, effectiveness and scope of wiretapping. As it stands, the Wiretap Report data can’t be trusted and, worse, actually may be misleading.