Based on 3+ years of employment with a large federal agency, what you are saying ins't correct. I obviously cannot speak for every federal agency. However, at m

Based on 3+ years of employment with a large federal agency, what you are saying ins't correct. I obviously cannot speak for every federal agency. However, at my organization FOIAs are routinely distributed to individuals/divisions that are likely to have pertinent records. The main thing that we are required to search are e-mail inboxes (to include archived e-mails) and local paper/digital files (e.g., paper documents and electronic folders). If there are other databases/systems that an organization owns the records for, then those must be searched as well. There are three main problems with the way FOIA requests are fulfilled. 1) They are incredibly time consuming to do properly, 2) Very few additional resources are given in order to alleviate the time consuming process, and 3) the process for tasking and responding to requests is grossly inefficient. The requests are literally sent out via e-mail and responses back are sent via e-mail attachment. To paint the picture a little more clearly, let's say someone submits a FOIA request for all records related to the "Boston Bombing". For a broad search like that, hundreds if not thousands of people will be searching through their records. Those hundreds/thousands of people each find five documents which they have to send back to the agency's FOIA office via e-mail attachments. Most of the records are duplicates (e.g., an e-mail sent out to a distribution list), but everyone still has to send their record in. Let's just say it is a very frustrating process for everyone involved. I hope it doesn't sound as though I am making excuses - quite the opposite. Merely pointing out that while there are definitely issues that affect accuracy, they might not be the issues everyone assumes. Perhaps this is a sector that is in need of some innovation/disruption =).