Go To Search
Click to Home
Issues with Police Oversight
By Andrew Keegan

We are all aware of the intense local and national focus on interactions between police officers and members of the public. Unfortunately, this is often because those interactions were negative. The worst possible outcome in these situations is the death or injury of someone involved. Other less serious outcomes (at least from a government entity’s perspective) can include lawsuits, damage to public trust, and critical media coverage. These risks highlight the importance of a process to identify and address these negative interactions, and an audit to evaluate whether that process is effective. The City of Austin did just that in 2016, with the Police Department’s Handling of Complaints Audit. Our audit found problems with all phases of the complaint process. We noted barriers to filing complaints, policies and practices that made it difficult to ensure consistent investigations and discipline decisions, and factors that limited the Office of the Police Monitor’s (Police Monitor) ability to provide effective oversight. In this article, I will provide some background on police oversight, describe how the process was supposed to work in Austin, and get into the details of some of the issues we discovered. I’ll also discuss some things that happened after our audit, and conclude with some lessons we learned that might help others who are considering doing a similar audit.)

Note: Obviously, an audit of the complaint process does not directly address the risk of death or injury, since it evaluates reactions to an event that has already occurred. We are currently doing an audit that more directly addresses that risk, which may be complete at the time of this publication.


Any discussion of police oversight must come with a disclaimer that there is no set best practice for how things should look. According to a report from the National Association for Civilian Oversight of Law Enforcement (NACOLE), “jurisdictions should focus on ‘Best-Fit’ rather than ‘Best Practices’ when considering how to structure civilian oversight.”1  The NACOLE report listed things organizations should consider when designing an oversight structure. These include the political issues of the jurisdiction, available resources, and the culture of the law enforcement organization.


Even other cities in Texas had vastly different oversight models. This is due to a State law that established civil service for police officers.2  Among other things, this law sets a time limit for imposing discipline on officers and restricts access to the personnel files of officers. However, the law allows labor agreements between cities and police unions to supersede the law. For example, under the law, a Police Chief cannot fire an officer for an incident that occurred more than 180 days ago. However, a city and the police union could agree to extend this deadline, shorten it, or get rid of it altogether. As a result, cities in Texas can have vastly different complaint processes. In fact, the restrictions on an officer’s personnel file established by the law are so strict that an effective civilian oversight structure can ONLY exist through a labor agreement.


At the time of the audit, people who had a negative interaction with an Austin Police Department (APD) officer could file a complaint in one of three ways. They could contact the officer's supervisor, APD's Internal Affairs Division, or the Police Monitor. The Police Monitor is the civilian agency tasked with oversight of the APD. APD officers could also file complaints about other officers through the same channels, and APD management could order an investigation into an officer's action even if there was no formal complaint.

APD’s Internal Affairs Division was responsible for investigating complaints related to possible violations of APD policy.3 An officer's chain of command was responsible for deciding if a violation had occurred, and, if so, determining an appropriate discipline. The Police Monitor provided oversight throughout this process, including reviewing evidence and commenting on discipline decisions.


As I mentioned earlier, we found issues at every stage of this process and with the Police Monitor's ability to provide oversight.

Barriers Discouraged People from Making Complaints
To test what information was available to the public, I visited City libraries and other City facilities “undercover.” I would tell the employees at these locations that my girlfriend was upset about a recent negative interaction with a police officer, and ask them what I could do. The most frequent advice I got was to contact APD. Although this wasn’t wrong, we recognized that people who had recently had a negative interaction with a police officer might be unwilling to engage with APD again. Very few of the City employees I talked to mentioned the Police Monitor, even though most were sitting at computers and it was the first result in a Google search for "Austin police complaint."

In addition to the lack of information about how to file a complaint, we noted several other barriers that could discourage someone from making a complaint. For one, the Police Monitor’s office was far from downtown, which is where most police interactions occurred. Another issue was that complaint forms were only available in English and Spanish. To make it worse, there was an error with the Spanish translation. Lastly, there was conflicting information about whether people could file complaints anonymously.

A major issue we noted was that the culture around complaints might have deterred people from making a complaint. A few years before our audit, the official City Twitter account tweeted the contact information for the Police Monitor prior to a well-known Austin festival. The message advised people to contact the Police Monitor if they had a negative interaction with an officer while they were in town. After swift backlash from the Police Chief and police union, the City tweeted an apology for giving the wrong impression by providing residents and visitors with that information.


Policies and Practices Did Not Ensure Consistent Investigations or Discipline
We noted several issues with APD policies and practices that cast doubt on the ability of the department to consistently investigate complaints and discipline officers. One of the most critical inconsistencies we noted related to discipline. APD’s policy included a discipline matrix to provide guidance for determining an appropriate discipline. However, this matrix did not include what all possible policy violations were; it listed a range of potential discipline for some violations, and it did not address situations were an officer violated multiple policies in a single incident. We noted several incidents where different officers who violated the same policy received different disciplines, as shown in the table below.


Another issue is that APD supervisors may not have always forwarded complaints they received to Internal Affairs. One reason was the APD policy that outlined this requirement was unclear. It required different actions depending on the seriousness of the allegation, but did not define what a “serious” allegation was. For example, a survey of supervisors indicated that some felt “rudeness” was a serious allegation while others felt it was minor. Even the concept of rudeness was inconsistent, with some supervisors making a distinction between belittling someone and being rude.

Police Monitor’s Ability to Provide Oversight Was Limited
Lastly, we noted several issues that limited the Police Monitor’s ability to provide effective oversight. For one, their access to electronic investigation records was limited. This was despite an agreement with the police union giving the Police Monitor “unfettered access” to investigations. That agreement did limit their oversight ability in other instances though. (I should note again that that agreement was the only reason that civilian oversight existed, in any form, in Austin.) Police Monitor staff were not allowed to solicit complaints, gather evidence of policy violations, or participate in discussions about discipline. The Police Monitor could make recommendations to the Police Chief if they disagreed with a discipline decision, but we saw no evidence that recommendations had any effect on discipline decisions.


Despite the sensitive nature of the topic and the issues we found, the report was well received. Our work resulted in 11 recommendations, which is a lot for us. We spread the recommendations between APD (8), the Police Monitor (2), and the City Manager (1). Some of these recommendations were actually implemented prior to us issuing the final report. For instance, the Police Monitor met with branch librarians and distributed information (in multiple languages) on filing complaints. Additionally, the Police Chief sent a memo to the department re-iterating that all complaints should be forwarded to Internal Affairs. He also included “failure to forward complaints” in the discipline matrix.

Due to the importance of this topic, we did a follow-up audit this year. While we noted progress on the recommendations, a significant hurdle to full implementation was that the City’s contract with the police union expired in 2017, and the City Council rejected a new contract. As a result, State law once again was the law of the land, forcing the City to disband most of its civilian oversight structure. This meant that unfortunately, implementation of some of the recommendations was put on hold until after the labor agreement negotiating process concluded.


If your organization includes a police function and you haven’t audited how they handle complaints against their officers, I strongly recommend you consider it. Not only are the risks high-impact, it’s also a relevant topic that is likely to strongly resonate with your stakeholders. There are many lessons we learned doing this audit and I’m happy to discuss them in detail in another setting, but I wanted to share two that particularly stood out.

  1. Get buy-in from officials (Or at least a legal opinion) - One of the challenges we expected in the audit was getting access to records. Although City Code gives us access to all City records, we knew State law had heavy protections in place regarding access to police officers’ personnel records. The Police Chief's support of the audit went a long way towards overcoming any resistance to us reviewing those records though. We also had a legal opinion from the City Attorney acknowledging our right to see the records. Although we never had to play this card, it warmed the team's heart knowing it was there just in case. 
  2. Expect current events to affect the messages – Shortly after starting our audit, an Austin police officer shot and killed a naked, unarmed 17-year-old. During the fieldwork phase, a video emerged showing an Austin police officer pulling a female motorist out of her car and slamming her to the ground after stopping her for speeding. These issues gave us a real-time look at the process, but were technically outside of our scope period. That didn’t mean that they weren’t very much on the forefront of our stakeholder’s minds as we presented our results though.


Andrew Keegan is an Assistant City Auditor for the City of Austin, Texas, where he has worked for 5.828 years (approximately). During that time, he has led audits of public utilities, public safety, and construction management. According to his manager, he needs more friends.


1 Civilian Oversight of Law Enforcement: A Review of the Strengths and Weaknesses of Various Models, September 2016, https://d3n8a8pro7vhmx.cloudfront.net/nacole/pages/161/attachments/original/1481727977/NACOLE_short_doc_FINAL.pdf?1481727977.

2 The law also applies to firefighters, but this article will only focus on the applications of the law to police officers.

3 Another APD unit was responsible for investigating complaints related to possible criminal acts. Our audit did not review these investigations.