Public Data that Isn’t (or Wasn’t) Public

–Guest Post: Justin Gallagher, Assistant Professor of Economics, Case Western Reserve University






I recently completed a coauthored working paper, together with Paul J. Fisher, that examines whether electronic monitoring via red light traffic cameras leads to fewer vehicle accidents and injuries.  As part of the project, we encountered the bizarre situation where researchers at The University of Texas at Austin, a flagship public state university, fought tooth and nail against releasing vehicle accident data originally collected by the Texas Department of Transportation (TxDOT).

The accident data are routinely used by academic researchers at the University of Texas at Austin Center for Transportation Research who publish in peer reviewed journals.  Moreover, the same data for later years are available free via the TxDOT website.  We eventually received the accident data via a Public Information Act Request, but only after the Texas State Attorney General’s office ruled in our favor following a petition by UT Austin against release of the data.

When we began work on the project we used the publicly available traffic accident data.  These files are part of the Crash Record Information System (CRIS), exclude personal identifying information, and are posted on the TxDOT website.  The TxDOT website states, and we confirmed, that TxDOT does not retain any CRIS files before 2010.

The post-2010 accident data were first made publicly available by TxDOT following a 2008 Public Information Act Request filed by an investigative journalist.  In a foretelling prequel to our own saga, TxDOT appealed to the Texas Attorney General’s office in an attempt to prevent release of the data.  The Attorney General ruled that the accident data should be available to the public.  Following the ruling, TxDOT spearheaded a bill in the Texas State Legislature that would have made all accident data exempt from any Public Information Act Request.  The TxDOT-sponsored bill with the restrictive language failed to pass.  Next, TxDOT transferred all pre-2010 accident data in CRIS to the University of Texas at Austin Center for Transportation Research and expunged their own records.

We first attempted to access the pre-2010 CRIS data by contacting the authors of peer reviewed research papers that used these data.  All of the researchers had a connection to the Center for Transportation Research.  Our emails and phone calls were met with some form of the response: “I do not have the data” or “The data are confidential.”  In one instance the author would not even admit whether the data existed.  We even inquired about bringing one of the researchers at the Center for Transportation Research onto our project so as to facilitate access to the data.  Finally, after numerous email exchanges we were unequivocally denied access to the data.  The stated reason was that the University of Texas at Austin Center for Transportation Research had made an agreement with TxDOT not to act as a data “pass through” to other parties.  Incredulous, we followed up with the current Director of Crash Data and Analysis at TxDOT who confirmed the no “pass through” policy.

So we filed a Public Information Act Request.  We requested the exact same information that had previously been ruled permissible under the 2008 request, and which was already available for later years on the TxDOT website.  The University of Texas at Austin Center for Transportation Research fought the request.  Legal counsel at the University of Texas wrote a letter to the Texas Attorney General arguing that data should not be released.  At this point, we solicited help from our university’s legal counsel and filed our reply to the letter.  After multiple rounds of attorney letters, the Texas Attorney General’s office ruled in our favor (see Public Information Act Documentation) .

Why would researchers at the University of Texas at Austin Center for Transportation Research go to such great lengths to prevent the release of the data? One possibility is that they genuinely believed, and wished to respect, the legally dubious no “pass through” policy as advocated by TxDOT.

My view, informed by my study of economics, is that opposition to the data release follows from the University of Texas at Austin Center for Transportation Research’s position as a monopolist. A monopolist has market power, can extract additional payment for services, and has a strong incentive to protect their advantageous position.  The University of Texas at Austin Center for Transportation Research has been very successful at winning research grants from TxDOT. In 2016, the Center for Transportation Research was awarded more than $8.9 million in TxDOT research grants, which amounted to more than 50 percent of the Center’s annual budget. The successful acquisition of TxDOT grant money is not too surprising as they are among the only ones with complete pre-2010 Texas accident data!  (The Texas A&M Transportation Institute may also retain a copy).

We have posted complete Texas CRIS accident data from 2003-2009 at Dataverse and on my website.  We are hopeful that there will be no need for other researchers to star in the third installment of this Public Information Act Request.  As is often the case, the plot lines of the sequel were unoriginal and mimicked the first release.  Moreover, the sequel was a colossal waste of time and money for all involved.

One thought on “Public Data that Isn’t (or Wasn’t) Public

  1. In the hierarchy of motives UT will always choose self enrichment. This is counter to the stated goal of the learning institution. Sad.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.