December 11, 2024

Judge Declares Some PACER Fees Illegal but Does Not Go Far Enough

Five years ago, in a post called “Making Excuses for Fees on Electronic Public Records,” I described my attempts to persuade the federal Judiciary to stop charging for access to their web-based system, PACER (“Public Access to Court Electronic Records”). Nearly every search, page view, and PDF download from the system incurs a fee ranging from 10 cents to $3 (or, in some cases, much more). I chronicled the many excuses that the courts have provided for charging what amounts to $150 million in fees every year for something that should—by all reasonable accounts—not cost much to provide.

I thought the courts were violating the law. I suggested that someone file suit. Two years later, the good folks at Gupta/Wessler did (in partnership with Motley Rice). Yesterday, Judge Huvelle of the US District Court for the District of Columbia agreed—in part. You can read her opinion here, and see all documents in the case here. Under her ruling, approximately $200 million will likely be returned to people who paid PACER fees from 2010 to 2016. This is good, but not good enough.

It also does not address the larger constitutional issues that I raise in my forthcoming paper, “The Price of Ignorance: The Constitutional Cost of Fees for Access to Electronic Public Court Records.”

Judge Huvelle is a good and fair judge. She rejected the reasoning of both the plaintiffs and the defendants (the Judiciary). Instead, she substituted her own analysis. Unfortunately, her analysis was both legally and technically flawed. Under her ruling, PACER fee-payers will not recover another $750 million (or so) of fees that I think are unlawful. The rest of this post explains why, and what might be next.

The law says that the Judiciary “may, only to the extent necessary, prescribe reasonable fees… to reimburse expenses incurred in providing these services.” The lawsuit centered on the meaning of terms like “only to the extent necessary” and “these services.”

During the litigation, the Judiciary provided a spreadsheet showing how PACER fees were spent across different categories (categories invented by the Judiciary). About a quarter of these fees were spent on things that related to public access tangentially at best (for example, “courtroom technologies”). The judge decided that these were illegal. About 15% of fees were under a different heading: “Public Access Services.”  The plaintiffs did not allege that these were illegal even though ~$25m per year seems an awful lot for serving digital documents to the public. Only the middle set of categories—about $100m per year—was seriously in dispute. Pages 18 to 20 show the breakdown for 2016, which fall into the following categories (items in bold were seriously in dispute).

2016 Example:

  • “Program Requirements” (all deemed legal by Judge Huvelle)
    • Public Access Services (~$24m)
    • Case Management/Electronic Case Files System (CM/ECF) (~$40m)
    • Communications Infrastructure, Services, and Security (~$46m)
    • Court Allotments (~$7m)
    • Electronic Bankruptcy Noticing (~$7m)
  • “Congressional Priorities” (all deemed illegal by Judge Huvelle)
    • Victim Notification (~$100k)
    • Web-based Juror Services (~$2m)
    • Courtroom Technology (~$25m)

The practical question in the case was whether the line between legal and illegal uses of PACER funds fell above, below, or somewhere in-between the items in bold. At oral argument, plaintiffs’ counsel was clearly surprised that the judge seemed ready to rule on the merits of that question at this phase of the litigation. The parties had stipulated to limited discovery during this phase, so there is little in the record about what those categories mean.

Aesthetically, it would be clean to make the split between the two major headings. Judge Huvelle did not base her reasoning on spreadsheet headings, but her ruling was consistent with them. Everything under “Program Requirements” was allowed.

On Twitter, I’ve seen people describe this decision as “splitting the baby.” If this means that the decision chooses a somewhat arbitrary point between two parties’ positions, that is true. Of course, the origin of the phrase is King Solomon’s wise adjudication of a child custody dispute. When he threatened to literally cut a baby in two, the true mother revealed herself and justice was served. To actually split the baby does not serve justice.

CM/ECF is the system for attorneys to file documents electronically and for the courts to manage almost all case materials that they used to manage on paper. CM/ECF is an essential tool for operations of the courts, and has generated tremendous cost savings for the Judiciary. Attorneys do not pay for CM/ECF, and the Judiciary has never asked for appropriations to fund it. Instead, it has been funded entirely by PACER fees. In the words of Judge Huvelle, this is justified because “PACER cannot be divorced from CM/ECF.” True, PACER users would not have much to access without electronic filings. But CM/ECF, on the other hand, can be divorced from PACER. The judge has simply analyzed the “but for” causation in reverse. CM/ECF is as essential to the functioning of the modern courts as the physical clerk’s office. Regardless of whether PACER exists, CM/ECF must exist. (And if PACER is “merely the portal to the millions of electronically-filed documents that are housed by the judiciary on CM/ECF” then why does it cost $24m per year to operate?)

Judge Huvelle sidesteps this issue when she dismisses the notion that the appropriate yardstick of permissible PACER fees is “the marginal cost” of operating PACER. This language comes from the congressional report on the E-Government Act, which created some of the statutory language in question. “Marginal cost” refers to the economic principle that the true cost of providing a good or service is only the additional cost above and beyond what would otherwise be spent. Tim Lee explained, in his 2009 post “The Trouble with PACER Fees” here, why even a principled libertarian should be in favor of taxing the general public for the bold items above. It’s not just economics. It’s about maintaining a liberty-enhancing justice system that is transparent and knowable by those that it governs.

Another principle of our democracy is that only Congress may impose taxes, because it is the branch most accountable to the people. If it were otherwise, the Executive and the Judiciary could impose taxes with impunity. When Congress delegates its taxing power, it must do so clearly and unequivocally. That is the lesson of the unanimous Supreme Court case Skinner v. Mid-America Pipeline Company. Judge Huvelle brushes away this line of argument in her opinion by claiming that the statute unambiguously supports her line-drawing.

It bears repeating that I think that Judge Huvelle is a good and fair judge. Her opinion is the best summary of legislative and judicial history in this area. It is solid within its four corners. However, it is wrong.

One need only begin to pull on the dangling threads in the unexplored sub-line-items to see that something is deeply wrong. Two examples from CM/ECF illustrate the point.

First, in 2016, the Judiciary appears to have spent a couple million dollars per year on running conferences for judicial staff to learn how to use CM/ECF (“CSO Combined Forum”). This line-item is described in slightly more detail on page 84 of a document dump from the Judiciary, filed a week before the March 23 hearing. (And that same document contains budget spreadsheets from before 2010, which characterize the bold items as “Congressional Priorities” rather than “Program Requirements”.)

The second example has a more dramatic effect on overall costs.  The CM/ECF headings for 2010 to 2016 have sub-line-items for “NextGen,” the complete overhaul of CM/ECF that began in 2009 and continues to this date. Annual totals frequently top $10m. However, PACER users in district courts that have implemented NextGen (such as Connecticut), will notice essentially no change in functionality in exchange for their years of user fees. Judge Huvelle construes the plaintiffs as having forfeited any argument about NextGen costs because she thinks that the attorneys “rais[ed] no challenge to CM/ECF if the statute authorizes ‘PACER fees to cover all costs necessary for providing PACER access and other public access services.'” After the March 23 hearing in which it became clear that she was about to rule on CM/ECF without further discovery, the plaintiffs’ attorneys submitted a supplemental filing (at p. 3) indicating that even their preliminary review of NextGen raised serious questions about whether it was indeed necessary for PACER.

Other line items—such as the ~$10m/year Electronic Bankruptcy Noticing (EBN) system, which notifies creditors when someone files for bankruptcy—are permitted and will apparently undergo no further scrutiny.

I don’t know what is next for this litigation. It is good that PACER users will likely see some refund of illegally charged fees. But the systemic problems created by PACER fees will likely persist unless something else forces action.

Comments

  1. R.Steinmetz says

    Was any consideration given to the costs avoided by having PACER? It seem the Judiciary is spending a lot less on managing paper submissions than it would “but for” PACER.

    • Stephen Schultze says

      No, although I think you are mixing together PACER and CM/ECF (which is understandable given the way the opinion treats the two). The cost savings you describe would be due to CM/ECF (the filing and case management system). To be sure, CM/ECF has generated tremendous cost savings. Furthermore, it has been funded entirely by PACER fees. In that sense, the cost savings would not exist “but for” PACER. However, I suspect that if the Judiciary had requested an appropriation to fund CM/ECF–rather than funding it out of PACER fees–Congress would have given it to them.

      Whatever the case, the Judiciary has indeed been recognizing significant cost savings–directly and indirectly–as a result of charging PACER users in ways that I think are illegal.