Research Methods

 

How cases were selected for inclusion in the Center’s diagrams

Deason Center researchers collected data from misdemeanor courts in four Texas counties. To preserve the anonymity of counties, they are referred to using the pseudonyms Bluebonnet, Cactus, Longhorn, and Tumbleweed.

Researchers collected data on a total of 2,618 closed cases across the four counties. Those cases were first identified from lists of cases closed provided by each County Court Clerk’s office. Totals from each county are shown in the table below.

 

County

Total cases collected

Date range of case closure

Bluebonnet

938

April 2023 to June 2023

Cactus

701

January 2021 to September 2022

Longhorn

483

January 2016 to October 2022

Tumbleweed

496

January 2016 to March 2023

 

When selecting data for inclusion on the diagrams, cases were excluded either because they were missing key documentation, or because the magistration procedures in the case were substantially different to the ones described. Exact numbers of cases excluded, and the reasons for exclusion, can be found below each diagram.

Exclusion for missing documentation

Because the Center’s diagrams track representation status in cases over time, any case that was missing relevant information had to be excluded. These included cases where records did not indicate whether the defendant requested counsel, cases where the date of magistration was missing, and cases where no magistration record was discovered at all.

Exclusion for inconsistent procedures

Researchers spoke with county officials and reviewed records in each county to understand magistration procedures and procedures for appointing counsel. Often, those procedures had evolved over time, and researchers sought to identify the scope and timing of such changes.

Based on these conversations, Center researchers focused the diagrams on cases where magistration procedures and processes for appointing counsel were known and consistent. Researchers excluded cases when counties changes from paper to electronic application systems, when judges who lacked appointing authority stepped in to magistrate in the absence of judges who had it, and where counties experimented temporarily with different approaches to appointing counsel, for example. Researchers also excluded cases where key information that could be used to verify magistration procedures (such as the date of the magistration itself) were missing.

Identifying the type of attorney at arraignment

Researchers spoke with county officials and reviewed records in each county to understand magistration procedures and procedures for appointing counsel. Often, those procedures had evolved over time, and researchers sought to identify the scope and timing of such changes.

Researchers inferred the identity of a defendant’s attorney at arraignment from documents dating from that arraignment. Researchers coded cases as ‘court-appointed’ if the attorney’s name matched that on an approved application – or on a subsequent order for substitution of appointed counsel – dated on or before the arraignment. Researchers coded cases as ‘Retained’ if the attorney’s name matched that on a ‘letter of representation’ dated on or before the date of the arraignment. Researchers also coded cases as ‘retained’ when other documents (such as a notice of hearing or motion) indicated an attorney was retained on the case. Researchers coded cases as ‘unrepresented’ where they found none of the above indications of an attorney’s presence among the documents they retrieved.

County-specific methods

Bluebonnet County: Researchers retrieved files from a single case management system in the County Clerk’s office. The system collated records from magistration alongside records of cases from proceedings in County Court. For each case, researchers downloaded and reviewed all available documentation.  

Cactus County: Researchers first retrieved documents from the County Court Clerk’s office and a public, online jail data portal. County officials did not grant researchers access to magistration records in Cactus County, though researchers were able to obtain needed information with the assistance of local attorneys. Without showing Center researchers the records, attorneys looked up records one-by-one in a county magistration database and dictated key information which researchers transcribed. Researchers matched magistration database records to County Court Clerk office and jail portal records using the defendant’s name, their charge, and offense date.

Longhorn County: Researchers collected files from the Longhorn County Court Clerk’s office. Because those records did not consistently include magistration records, Center researchers also searched for and obtained magistration records from case files held in the County Attorney’s office, matched using the defendant’s name, charge, and key dates in each case. Researchers also obtained small number of additional magistration records from the office of the Justice of the Peace that magistrated the large majority of cases in the county.

Tumbleweed County: Researchers retrieved an initial set of records from the Tumbleweed County Clerk’s office. Researchers then combined those data with data from documentary and database records kept in the county jail, where a large number of magistration records were located. Researchers matched records using the person’s name, charge, and key dates in their arrest, incarceration, and court case. (Researchers also inspected records in Justice of the Peace and county archives yielded only duplicate information to that obtained elsewhere.)