How Much Should PACER Cost?

by on April 13, 2009 · 143 comments

I wrote a piece about PACER last week, which Katherine Mangu-Ward at Reason was kind enough to link to from Hit and Run. In the comments to her post, a reader asked a reasonable question about the fees you pay to access PACER: “Are you buying the data or paying the court’s bandwidth costs?”

Now I’m usually pretty sympathetic to the idea that the people who benefit from government services should pay for those services through user fees. But there are two reasons this doesn’t apply in a case like this.

First, there’s the math. As it happens, I’m working on a project that will involve hosting large amounts of content, so I’ve been researching hosting costs. One of the most popular managed hosting services is Amazon’s EC2/S3. You can see the pricing for that system here. It provides a good point of comparison for PACER’s fees.

To make the math easy, let’s assume you’ve got a lawyer who downloads 20 documents per week, each of which is 10 pages long and 1 MB in size. Over the course of a year, this lawyer will download around 1000 documents, and he’ll be charged 1000 * 10 * $0.08 = $800 for those documents. (This is actually an underestimate because the lawyer also has to pay for search results) And because each document is about 1 MB in size, the total quantity of data transferred from PACER will be around 1 GB.

Now, if you click over to Amazon’s S3 pricing page, you’ll see that the going rate for a GB of data transfer in the private market is… 17 cents. In fairness, Amazon also charges for CPU time on the EC2 cluster, so if the courts actually built their system on EC2/S3, the marginal cost of a GB of data delivery might be more like 50 cents. But charging $800 is three orders of magnitude too much.

Moreover, as the cost of bandwidth drops, the billing infrastructure is going to be a larger and larger fraction of the total cost of providing PACER. Credit card fees alone will eat up a significant chunk of a 50 cent charge. Hence, making PACER free might actually be cheaper for taxpayers than charging amounts so trivial that they can’t even cover the costs of collecting them.

Now, the courts would probably point out that 17 cents (or even 50 cents) per user wouldn’t come close to covering the full costs of running PACER. They have to run the CM/ECF system that lawyers use to file documents. They have to hire court clerks to organize, file, redact, and (in increasingly rare cases) scan the documents. And they have to hire employees to build and maintain their IT infrastructure.

But these are costs that the courts would have to bear whether or not they made public documents available on the web. Electronic filing of documents improves the courts’ efficiency whether or not those documents are put on the web. And the courts need to protect peoples’ privacy regardless of how the documents made available. It makes no sense to regard these as costs of providing PACER.

There’s also a more fundamental problem with this line of reasoning: the idea that the benefits of PACER flow primarily to the people requesting the documents. If I download some documents from PACER and use them to write a story about some court case, I clearly derive some benefit from that. I might get paid, and I’ll get the various benefits that flow from having people read my work. But I like to think that covering court cases has broader benefits too. Greater media coverage of the courts leads to better-informed voters, and it creates more incentives for judges and other participants in the legal system to behave themselves. In the long run, the largest benefits of journalists having easy access to court records flow to society at large, not to individual reporters.

The same point can be made about many other PACER users. Academics who study the legal system help to promote public understanding of the system and propose improvements. Public interest legal organizations like the Institute for Justice and the Electronic Frontier Foundation use PACER in their litigation, which helps to push the law in more beneficial directions. We would all benefit from the things Google or Microsoft could do to improve these documents with better search and organizational tools. Finally, public access to court record improves the ability of ordinary Americans to read and understand court documents themselves, which leads to greater confidence in the legal system’s reliability and a higher likelihood that malfeasance will be detected and publicized.

Public access to court records is a genuine public good. Given that taxpayers already spend several billion dollars a year to run the federal court system, it’s extraordinarily short-sighted to limit public access to these documents to save taxpayers a few million dollars. That would be true whether providing those documents to the public were cheap or expensive. Fortunately, if done right, public access can be extremely cheap.

Previous post:

Next post: