PACER Officials Give Weak, Nonsensical Excuse For Why PACER Deleted Tons Of Public Court Records With No Notice
from the total-failure dept
Charles Hall, a spokesperson for the Administrative Office, told The Post via e-mail that the change was made on Aug. 11 in preparation for an overhaul of the the PACER architecture, including the implementation of the next generation of the Judiciary's Case Management and Electronic Case Files System. "NextGen replaces the older CM/ECF system and provides improvements for users, including a single sign-on for PACER and NextGen," he wrote.Yes, I can understand how there might be some backwards compatibility problems, but it seems like there's a pretty straightforward way to deal with that: move the old files to the new system. But no one seems to be doing that. The claim that it's okay because the documents are available "directly from the relevant court" is a ridiculous, weak excuse. The whole point of PACER was to make the records more accessible and to get away from having the documents only available to people who have the time and resources to go down to the local court to get the documents.
However, as a result of the changes the locally developed legacy case management systems of some courts were no longer compatible with PACER, he says. Since PACER works as a sort of distributed network of different archives rather than one centralized database, that's a major problem.
However, Hall says, the dockets and documents no longer available through the system could still be obtained directly from the relevant court and "all open cases, as well as any new filings, will continue to be available on PACER."
The Washington Post also spoke to Brian Carver from UC Berkeley's School of Information, who recently took over the management of RECAP (the crowd-sourced effort that many of us rely on and contribute to) to post any documents obtained via PACER to the freely available web via the Internet Archive. He made a pretty good point, that it's absolutely ridiculous that no one at the courts made this information public earlier so that people could try to do something to preserve the availability of those case records:
Carver says their group would be happy to host the files publicly, and are reaching out the courts to see if that is possible. But he was still shocked by the lack of advance warning. "If we had known about it in advance maybe we could have done something to target these documents and archive them publicly," he says. "It was really an announcement of an accomplished feat -- we weren't told until after this deed was already done."Meanwhile, Aaron Greenspan, who runs another site that makes court documents publicly available, PlainSite, (and who first alerted us to this issue), has sent the Administrative Office of the US Courts a response to this, pointing out, as we mentioned above, how ridiculous it is:
The supposed explanation that the dockets and documents were stored in an "incompatible" format is nonsensical and has no technical justification whatsoever. As you well know, database schema changes happen all the time, and it is easily within the realm of possibility to write conversion scripts—especially given how much money the AO has spent on PACER NextGen already. In fact, the technology known as ODBC exists for this very reason.He further has offered to "reverse-engineer the supposedly legacy format(s) ourselves and make the data available to the public for free at no cost to the Government" and even asked them to ship the hard drives to him. Somehow, I doubt that's going to happen, but it seems like working with Carver or Greenspan or anyone else, frankly, would have made more sense than just deleting these public records.
Update: And... Carl Malamud of Public.Resource.Org has sent letters to all of the courts that had cases deleted, requesting copies of the deleted cases. We'll see how that goes.
While it's great that PACER is apparently moving to a new system (the login page is already different, though the underlying pages still remain horrifyingly painful to use), dumping a bunch of cases down the memory hole is a huge problem.