Open Austin logo

Democratizing Criminal Defense Data

Our organization, Open Austin, in partnership with Texas Fair Defense Project, began creating a database for criminal court case records that were previously siloed within their respective county websites.   The goal of this work is to visualize this data to help policymakers, advocates, and everyday people understand the current state of public defense representation in their communities and, where appropriate, advocate for improvements.

How are we determining evidence of representation?

A defense attorney that is providing good representation generally files motions into the court case file to advocate for better results for their clients. Some of these motions include motions for production of discovery, to reduce their client's bond, to request a speedy trial, or to suppress evidence at trial (sometimes called a motion to suppress or a motion in limine).

Keep in mind that we are only looking at records from a court file, which will never be able to tell the whole story about the attorney-client relationship. It is possible for a person to get great representation without their lawyer filing any of these motions, or for a lawyer to neglect their client's case even if one of these motions is filed. We are using motions as a proxy because most of the time, filing at least one of these motions is a good indicator that a lawyer is putting at least some effort into their client's case.

Note: We are in the process of scraping case outcomes, but we do not have this data available yet.
In the future, we intend to compare outcomes with attorney type.

Why does evidence of representation matter?

Typically, poorer people with appointed counsel go through the legal system without the zealous advocacy that a retained attorney or well-funded institutional public defender might provide because their appointed attorneys do not have the time to investigate a variety of legal strategies in each case.

Try it out yourself

Try out the dashboard below to compare what kind of representation people get from court appointed lawyers with that of retained lawyers. You can compare different types of motions they file across different types of cases.

How does this disparity look over time?

The graph below also displays the change in evidence of representation over time.

Differences in charge category based on attorney type

We explored whether people who are charged with certain types of offenses were more likely to have an appointed or a retained attorney.

We found that people charged with DWIs were more likely to have retained attorneys and that people charged with property offenses were more likely to have court-appointed attorneys. Overall, there were lots of similarities between people charged with different offenses.

What type of attorneys are represented more per charge category?

Whether someone accused of a crime hires an attorney or is appointed one depends on a lot of factors. The accused may have more money and resources to afford an attorney. For instance, if more people arrested for driving while intoxicated charges tend to have a higher income compared to people charged with other types of crimes, then we would see a higher proportion of folks accused of DWIs retaining their own attorney.

The nature of the charge may also be associated with someone's ability to retain their own attorney. If someone who is experiencing homelessness must sleep in a private place to avoid the elements (criminal trespass) or must take food in order to survive (theft), then people with lower incomes or who may be experiencing homelessness may be more likely to be charged with those types of crimes.


What's next for this project?

  • We currently only have data from Hays County, but intend to expand that to all Texas counties.

  • Because of the difficulty in formatting for outcomes among all case records, we are still working on scraping final verdicts.

  • Race and gender data is limited in our current dataset. In the future, we plan to publish this aggregate demographic information.

Have any feedback for us?

We're still learning and figuring out how we can better scrape and analyze this data, and we would love any feedback or suggestions you have to improve!