Monday, March 07, 2016

AG overstated number of Texas police shootings, failed to promptly post reports; most black folks shot by police were in Houston

Grits contributor and UT-Austin postdoc Amanda Woog last week issued a new analysis of police shootings reported to the Texas Attorney General in the first four months after a new law, HB 1036 by Rep. Eric Johnson, caused agencies to begin reporting in September. Among her findings:
Of the 65 people shot by peace officers, 10 were reportedly unarmed and 55 were armed. Of the 24 people killed, two were reportedly unarmed and 22 were armed. Officers with the Texas Department of Public Safety (“DPS”) were involved in both deaths of reportedly unarmed people. DPS also saw disproportionately more deaths than other law enforcement agencies. Sixty-seven percent (four out of six) of the incidents that involved DPS officers resulted in death, compared to 33% (20 out of 61) of the incidents not involving DPS officers.
So overall, 44 percent of people shot by law enforcement during this period were killed, and 15 percent of people shot were unarmed. Here's a bit more detail regarding the circumstances surrounding police shootings:
Twenty-seven incidents occurred as a result of an emergency call or request for assistance; five as a result of a traffic stop; four as a result of a hostage, barricade or other emergency situation; two as a result of the execution of a warrant; and 27 were as a result of an “other” type of call. Four of the incidents listed as “other” were described as accidents, such as an accidental discharge or ricochet fragments from a training exercise resulting in injury.
Law enforcement interests should appreciate Woog's work to disaggregate and analyze this new data, since, in its statutorily mandated report,  the Office of the Attorney General (OAG) overstated the number of people killed by Texas cops by more than 20 percent.  From Ms. Woog:
Under HB 1036, the OAG is required to publish an annual report by February 1 summarizing the prior year’s data and appending the reports that were submitted that year. On February 1, 2016, the OAG released the 2015 report, which was posted on the OAG website and submitted to the governor and the standing legislative committees with relevant jurisdiction.

The OAG reported, “there were seventy (70) separate incidents statewide involving peace officer shootings. … Those incidents resulted in twenty-nine deaths and forty-one injuries to individuals; additionally, four (4) peace officers were injured, none were killed.” The OAG reported five more incidents of officer-involved shootings than this report but both reports are based on the same 90 reports that were submitted to the OAG.

This discrepancy is likely due in large part to redundancy in current reporting procedures, which require that law enforcement agencies file separate reports for each peace officer or civilian involved in a shooting incident. On the face of the reports, it is not always clear whether similar reports concern the same or separate incidents, or the number of individuals involved in a given shooting. For example, on October 30, 2015, a 29-year-old man was shot and killed by police outside of Mesquite. 
Three reports were submitted documenting the incident: two were filed by the Mesquite Police Department on November 2; and one by DPS on November 6. The DPS report was filed on a different date and reports a different incident location. The Mesquite Police Department recorded the location as “18600 IH 635,” while DPS recorded it as “18400 LBJ Fwy.”

Although the inconsistencies across these reports appear to indicate that they concern separate incidents, further examination reveals that they likely refer to the same event. A lieutenant with the Mesquite Police Department told media that two Mesquite Police Department officers and a DPS officer fired their weapons. In addition, the area of the road where the Mesquite Police Department reported the incident took place is known locally as the “Lyndon B. Johnson Freeway.” It is possible that the OAG report counted these reports as two or even three separate incidents, while they likely refer to the same event. In this study, these reports were counted as a single incident.

There were three incidents in 2015 in which multiple reports were filed by different agencies on different dates, but the reports likely refer to the same incident. One incident saw five separate reports filed by four separate agencies and the other two incidents were each reported by two different agencies. The OAG reported 29 deaths, but this report found 24; the difference may be explained by these incidents having been counted multiple times.
This could be fixed easily. For example, the form the OAG produced for agencies to fill out does not include the name of the victim, which would easily allow them to tell when they're counting multiple reports for the same shooting.

On the flip side, there were at least two police shootings resulting in death which were not reported by local agencies to the OAG:
the Washington Post database of people who have been shot and killed by police identifies two officer-involved deaths in 2015 that should have been, but were not reported to the OAG under HB 1036. The first incident was in West Odessa, where, according to news reports, a man was shot and killed by three officers after he allegedly charged at them with a knife. The second unreported incident was in Houston, where according to news reports, a man was shot and killed by a Harris County Sheriff’s Office deputy after the man allegedly lunged at the deputy with a knife.
Also, it appears the OAG is not always posting shooting reports online within five days of receiving them as required by HB 1036. From a footnote: "There are eight reports in the OAG annual report that are not posted on the OAG website as of the date of this report. The reports document incidents in September in Freeport, Plano, Houston, Alvin, Balch Springs, El Paso, and Dallas." They clearly had the reports, why not post them online?

Notably, people shot by the Houston Police Department were younger and more likely to be black than those shot in other jurisdictions:
Officers with the Houston Police Department (“HPD”) shot a significantly higher percentage of black people than officers in the rest of Texas. Of the 12 people reportedly shot by HPD officers, 75% (9) were black, compared to 12% of people shot by officers in the rest of Texas. Further, HPD officers were involved in all three reported deaths of black men. People shot by HPD officers were also younger when compared with the rest of Texas; the median age of people shot by HPD was 22, compared with 32 in the rest of the state.
Granted, this dataset is still relatively small (though far larger than one would wish). If those disparities continue throughout 2016, they may require more serious explanation.

While the popular assumption may be that most people shot by police are black, during these four months, 45 percent of people shot by Texas law enforcement were white, 23 percent were black (with most of those shot in Houston), and 31 percent were Latino.

Finally, readers may recall Ms. Woog had suggested revisions to the Attorney General's rules governing this data collection to fix some of the problems which produced these errors and to document whether reports are turned in within the legal timelines. (See her column from the Dallas News discussing the proposed amendments.) In this latest report, she offered this update:
During the public comment period, three individuals/organizations submitted comment on the proposed rules. Suggestions for improvement included creating supplemental forms to be used when multiple officers have fired their weapons or multiple people have been shot; including an entry on the forms for a deadly weapon description if one was reportedly used; providing guidance on how an “incident” is defined; and creating a single publicly-available source aggregating the information in the reports.
Thank heavens Amanda took on this project or, despite HB 1036, we STILL wouldn't have an accurate picture of the number of people shot and killed by law enforcement in Texas.

1 comment:

Anonymous said...

It should be noted that double amputees killed while in a wheelchair holding a permanent marker are counted under their definition of "armed"...