New DOJ FOIA Report an Exercise in “Grade Inflation;” Rewards Some FOIA Denials as “Achieving Efficiencies.”

“Increased efficiencies” allowed the DOJ to claim this astronomical 94.5 FOIA release rate. For the DOJ’s real FOIA release rate, see below.

The Department of Justice Office of Information Policy’s latest report on the status of FOIA, entitled Summary of Agency Chief FOIA Officer Reports for 2012 and Assessment of Agency Progress in Implementing the President’s FOIA Memorandum and the Attorney General’s FOIA Guideline with OIP Guidance for Further Improvement, paints an overly laudatory picture of the status of FOIA in the US.  Its appendix lists 99 agencies that are graded on 17 metrics, such as “high release rate for requests processed for disclosure,” “new material added to website,” and “decrease in backlogged requests”.    The vast majority of these grades –over 80 percent– are marked as green.  The true state of FOIA in the US is markedly more mixed; the DOJ OIP is engaging in grade inflation.

The report is a useful synthesis of information, culled from each agency’s Chief FOIA Officer Report and Year-End FOIA Report.  The section about the steps that agencies have taken to increase proactive disclosure of documents is especially interesting reading.  The addition of the “Did they close their Ten Oldest Requests?” metric is a welcome and beneficial measurement.  Even the DOJ OIP’s 17-point grade scale of agencies could serve as a useful measurement of FOIA performance, if it were graded more accurately.  The chief problem with the report is its conclusions.  Primarily, that the DOJ OIP has, once again, cooked its books to omit huge quantities of unfulfilled FOIA requeststs from its analyses.

The real release rate for *all* DOJ FOIA requests is a much more pedestrian 56.7 percent.

The first paragraph of the report begins with a misleading statistic.   It purports that government-wide, records are released in “more than 92 percent of requests where records are processed for disclosure.”  “Ninety-two percent??  Wow!!  Maybe all the National Security Archive’s FOIA Audits, tales of twenty-year-old FOIA requests, and horror stories of the never-ending referral process are wrong after all!!  Ninety-two percent… that’s not bad!!”  — Except that this sky-high 92 percent release rate does not include the extremely large universe of FOIA requests that are not processed for disclosure.  These non-processed (and apparently non-counted) requests include those that were denied due to: fees (pricing requesters out); referrals (passing the request off to another agency while the requester still waits); “no records”  responses (very frequently the result of inadequate searches); and requests “improper for other reasons” (which ostensibly include the “can neither confirm nor deny” glomar exemption).

In the same paragraph that the DOJ boasts its 92 percent release rate claim, it also boasts that agencies are “achieving efficiencies” to cope with their incoming requests.  But all too often (as the above paragraph shows), “achieving efficiencies” is code for “finding a way to deny requests at the outset so that you don’t actually have to process them.”  This can be done by attempting to price requesters out, referring them to another agency (often improperly), performing only a half-hearted search for documents, or finding some other administrative reason.  (The DOJ recently attempted to change its FOIA regulations so that requests had to be made to the specific DOJ component –often very hard for citizens to discern– or be denied.  In the face of opposition, the DOJ has withdrawn the reg —for now.  Components of the Department of Homeland Security have led the way in “achieving efficiencies” by pricing out requesters; the Electronic Privacy Information Center recently wrote a letter to the Office of Government Information Serves asking them to investigate DHS’s use of this practice.)

Two days after the National Security Archive awarded the Department of Justice the Rosemary Award for worst open government performance by a government agency, the DOJ OIP posted on its website that the Department of Justice had achieved an astounding 94.5 FOIA release rate.  A National Security Archive analysis of these figures debunked this claim and showed that the DOJ’s actual FOIA release rate was a much more pedestrian 56.7 percent.  Despite this, Attorney General Eric Holder gave a speech to kick off Sunshine Week (he did not mention his Rosemary Award), where he rehashed these incorrect figures and told the audience that the DOJ’s work on FOIA was, “nothing short of remarkable.”

Now, the DOJ is attempting to employ this misleading metric government-wide so that it can present a “Potemkin Report” that downplays the problems facing the Freedom of Information Act and encourages agencies to “achieve efficiencies” so that they don’t actually have to process large amounts of requests.

————————————

Other problems with the DOJ OIP’s FOIA report include:

• Not a single mention of one of the most positive FOIA developments: The FOIA Portal being jointly developed by the Office of Information Services, Environmental Protection Agency, and the Department of Commerce.  This may be because the DOJ wants agencies and users to use its own FOIA.gov website.  However, FOIA.gov only provides links where users can find where to submit a FOIA request to any one of 99 agencies.  The FOIA portal –under development by OGIS, EPA, and DOC and open to other government agencies– will allow requesters to submit requests at one central location with one uniform form and then track requests.  Most importantly, the FOIA portal will post all documents online released by FOIA, negating duplicate FOIA requests and providing an extremely robust form of proactive disclosure.  See here for more on the DOJ’s “combative” refusal to join the portal initiative.
• It gave the CIA a grade of “green” for “Adding New Material to its Website.”  The Agency has drawn the ire of intelligence historians for its extremely slow rate of posting documents on its website.  (If it were a member of the forthcoming OGIS/EPA/DOC FOIA portal, its FOIA releases would be automatically available as they were released.)  At one point this year, the CIA did not release a single new document on its website for more than eight months.  As such, it’s online release rate is probably not worthy of  a “green” grade.  I don’t know which other websites this inflation may also apply to.
• The Department of State received a grade of “green” for “offers to ability to make FOIA requests electronically,” but the Department will not provide documents in an electronic format, as required by the Freedom of Information Act.  Neither will the CIA.  I know that the DOD does; as such, claims that the classification system requires that FOIA documents only be printed appear specious and  –at any rate– not worthy of a “green.”
• Receiving a grade of green for “conducted or attended training” does not necessarily mean that all FOIA Officers (or, in an ideal world, all employees because “FOIA is everyone’s responsibility“) have been adequately trained.  According to the report, a grade of “green” appears to mean that some level of unspecified training merely had to be offered by the agency; a high attendance was not necessarily mandatory.
• The admirable FOIA work of smaller agencies may be over-represented.  As the National Security Archive’s FOIA Audits have repeatedly shown, most smaller agencies do good work with FOIA.  This is primarily because they receive a much smaller volume of requests than larger agencies.  However, when a citizen makes a FOIA request, the probability is much greater that he will make it to a large agency.  It’s requests to these large agencies which often take years (or decades) to process, can be subject to the endless referral merry-go-round, and are often victims of “achieved efficiency” denials.  This means that the “sea of green” of good smaller agency FOIA shops may obscure the very dire problems faced by the FOIA offices of larger agencies.
• Think my critiques are dead wrong?  Have a critique I missed?  Tell me in the comments!