November 16, 2020

2019 Joint Report: Analysis of APD Racial Profiling Data

This report by the Office of Police Oversight, Equity Office, and Office of Innovation examines Austin Police Department (APD) motor vehicle stop data from 2015-2019 to understand how various racial/ethnic groups in Austin experience motor vehicle stops. This report offers recommendations to address areas where disproportionality exists and to improve data collection.  

View and interact with our data on Tableauopen_in_new.

PDF Content

Disclaimer: The following text was extracted from the PDF file to make this document more accessible. This machine-generated content may contain styling errors due to redactions. In some instances, text may not load if the original file is a scanned image or has not been made searchable. For the full version of the document, please view the PDF.

THE OFFICE OF POLICE OVERSIGHT, THE EQUITY OFFICE, AND THE OFFICE OF INNOVATION 2019 JOINT REPORT: ANALYSIS OF APD RACIAL PROFILING DATA NOVEMBER 2020 TABLE OF CONTENTS Terminology............................................................................ 4 Methodology........................................................................... 5 Data Sources........................................................................... 6 Proportions of Stop by Race................................................. 8 Race Known Compared to Race Not Known in Motor Reason for Motor Vehicle Stops........................................... 10 Austin-Round Rock Metropolitan Area Population Compared Proportion of Searches, Hits, and High Compared to Proportion of Warnings or Field Observations by Race.....16 Proportion of Arrests by Race.............................................. 17 Proportion of Citations by Race............................................ 19 Geographic Disparity of Warnings, Field Observations, Gender and Race..................................................................... 25 Motor Vehicle Stop Outcome Totals for the Four Outcome Percentages............................................................ 27 Comparison of 2018 and 2019 Outcome Data................... 31 Executive Summary............................................................... 1 Introduction .......................................................................... 3 Background............................................................................ 4 Data Analysis.......................................................................... 8 Vehicle Stops .......................................................................... 9 to City of Austin Demographics Analysis of Stops...............12 Low-Discretion Searches, by Race........................................ 13 and Arrests.............................................................................. 21 Most Populous Races/Ethnicities.......................................... 26 Discussion............................................................................... 33 Recommendations................................................................. 35 Data Recommendations........................................................ 37 Conclusion............................................................................... 40 Appendices.............................................................................. 41 "Austin City Hall - South Facade & Plaza" by Diorama Sky is licensed under CC BY-NC-ND 2.0 1 EXECUTIVE SUMMARY This report by the Office of Police Oversight, Equity Office, and Office of Innovation examines Austin Police Department (APD) motor vehicle stop data from 2015-2019 to understand how various racial/ethnic groups in Austin experience motor vehicle stops. This report offers recommendations to address areas where disproportionality exists and to improve data collection. In summary: • Between 2015 and 2019, racial disparity persisted and, in many cases, grew worse. • Data from 2019 reveals that racial disparity in motor vehicle stops is still a pervasive problem, with Black/African Americans being the most overrepresented of all racial/ethnic groups in Austin. • Black/African Americans made up approximately 8% of Austin’s voting age population, but experienced 14% of motor vehicle stops, 25% of stops resulting in searches, and 25% of stops resulting in arrests. • In 2019, Black/African Americans and Hispanic/Latinos were overrepresented in motor vehicle stops by 6% and 2%, respectively, while White/Caucasians were underrepresented by 6%, and Asians were underrepresented by 3%. • The Black/African American driving population had two times more motor vehicle stops per driving population than the White/Caucasian driving population. • From 2018 to 2019, the overrepresentation of Black/African Americans in motor vehicle stops changed by 1%, from 7% to 6%. In this same period, the underrepresentation for White/Caucasians in motor vehicle stops changed by 1%, from -7% to -6%. There was no change for Latino/Hispanic or Asians in over or underrepresentation in motor vehicle stops. • Black/African Americans were the most overrepresented group across all categories except citations, for which Hispanic/Latinos were the most overrepresented. On the other hand, White/Caucasians were the most underrepresented across all categories, and Asians were slightly underrepresented across all categories. • Hispanic/Latinos were overrepresented across all categories except warnings/field observations. In this category, Hispanic/Latinos, White/Caucasians, and Asians were all underrepresented. • Once pulled over, Black/African Americans were three times more likely to be searched than White/Caucasians and were the only racial/ethnic group to receive more high-discretion searches than low-discretion searches. • Black/African Americans received 58% high-discretion searches versus 42% low-discretion searches. This disparity has grown by an absolute percentage of 7.7% since 2018. • Black/African Americans were overrepresented in cases when their race was known by officers before being stopped. • Black/African American drivers were 6% overrepresented when their race was not known before a stop and were 10% overrepresented when their race was known before a stop. • In the Austin-Round Rock metro area, racial disparity in stops was the same, or worse, than in the City of Austin. • In the Metro Area, Black/African Americans and Hispanic/Latinos were overrepresented by 7% and 5%, respectively. In other words, Black/African Americans, experienced 1% more 2 overrepresentation in the Metro Area compared to Austin, and Hispanic/Latinos experienced 3% more overrepresentation. • On the other hand, White/Caucasians and Asians were underrepresented by 12% and 1%, respectively. In other words, White/Caucasians experienced 6% more underrepresentation in the Metro Area compared to Austin, and Asians experienced 2% less underrepresentation. • APD vehicle stop data from 2019 also revealed a geographic disparity in warnings, field observations, and arrests. • Warnings and field observations were most concentrated on the west side of the city, while arrests were most concentrated on the east side of the city. • A race and gender analysis also revealed disparity. • Among Black/African American drivers, Black/African American males represented 64% of APD motor vehicle stops in 2019. Meanwhile, Black/African American females represented 36% of APD motor vehicle stops for the same period. The same was true for Hispanic/Latino males and Hispanic/Latino females. • The gender gap further widens when focusing on motor vehicle stops that resulted in an arrest. In 2019, there was a 60% difference between Hispanic/Latino men and women and a 56% difference between Black/African American men and women for motor vehicle stops that resulted in arrests. In other words, both Black/African American and Hispanic/Latino men were not only stopped more, but also arrested more after a motor vehicle stop compared to women of the same race/ethnicity. • An analysis of outcome data (i.e., what happens after a stop), not taking proportionality into account, also revealed racial disparities negatively impacting Black/African Americans and Hispanic/Latinos. • Black/African Americans and Hispanic/Latinos received higher percentages of searches and arrests, and Hispanic/Latinos received the highest percentage of citations. On the other hand, White/Caucasians and Asians received higher percentages of warnings/field observations. • Once stopped, Black/African Americans were three times more likely to be searched and approximately three times more likely to be arrested than White/Caucasians. 3 INTRODUCTION In January 2020, the Office of Police Oversight, Equity Office, and Office of Innovation released the inaugural Joint Report: Analysis of APD Racial Profiling Data (Joint Report). The Joint Report analyzed racial disparity in Austin Police Department (APD) motor vehicle stop data from 2015-2018. Additionally, the report provided recommendations to address the disparities indicated by the analysis.1 Texas Code of Criminal Procedure Article 2.134 requires APD to collect and report racial profiling data annually. The inaugural Joint Report found that APD stopped Black/African Americans and Hispanic/Latinos at disparate rates, and that these rates worsened from 2015-2018. Additionally, the analysis found that APD stopped White/Caucasians less than their proportion of the population. Racial disparities were also found when looking at the outcomes from those stops. The City of Austin has committed to addressing racial and institutional inequities, improving safety outcomes, and implementing policy and cultural changes to address the racial disparities in motor vehicle stops and related citations, warnings, searches, and arrests. First, the City’s multifaceted strategy, Reimagining Public Safety, seeks to improve safety outcomes, address racial and institutional inequities, and increase accountability and transparency in policing. Additionally, in June 2020, Austin City Council passed Resolution 50, which formally adopted a goal of zero racial disparity in policing. To track this goal, Council established the following policy goals for the Safety Outcome of Strategic Direction 2023: ● Zero racial disparity in motor vehicle stops by 2023; ● Zero racial disparity in citations and arrests resulting from motor vehicle stops by 2023; ● Zero use-of-force incidents per year by 2023; and ● Zero deaths at the hands of Austin Police Department (APD) officers per year by 2023. These goals demonstrate the City’s commitment to increasing accountability and transparency within the Austin Police Department, and they will be measured against the baseline disparities identified within the January 2020 Joint Report: Analysis of APD Racial Profiling Data by the Office of Police Oversight, Equity Office, and Office of Innovation and the findings in this report. The City will update the Strategic Performance Dashboard with these performance metrics. The purpose of this report is three-fold: 1. Measure disparities in the 2019 APD motor vehicle stop data, as well as trends in this data over the past five years; 2. Understand how various racial/ethnic groups in Austin experience APD motor vehicle stops; 3. Offer recommendations to address areas of concern. The data and recommendations contained in this report will be key to our efforts to reimagine public safety, eliminate racial disparities, and achieve the goals of equitable policing and the fair administration of justice. In order to begin rebuilding trust with the communities that have been negatively impacted by inequitable policing practices, the Austin Police Department must make every effort to address the racial and ethnic disparities identified in this report and fulfill the accompanying recommendations. 1 See Appendix 1 for a list of these recommendations and APD’s response to each. 4 BACKGROUND Terminology Absolute Percentage: The difference in a number over two periods in time or between two population types. For instance, if one observed a percent of 6% in 2018 and 7% in 2019, the absolute percentage difference would be 1%. American Community Survey (ACS): An ongoing survey conducted by the U.S. Census Bureau that provides vital information on a yearly basis about the nation and its people. Every year, the U.S. Census Bureau contacts over 3.5 million households (approximately 3%) across the country to participate in the ACS, and population data is extrapolated based on the responses of a random sample of households selected to fill out the survey. Data Cleaning: The process of manipulating data from its originally published form into a format that allows for additional analysis. Disparate Impact: A situation in which an outcome or adverse effect falls disproportionately on a protected group. Disparity: “Unequal treatment or outcomes for different groups in the same circumstance or at the same decision point.”2 Disproportionality: The ratio between the percentage of persons in a particular group at a particular decision point or experiencing an event (e.g. maltreatment, incarceration, traffic stop) compared to the percentage of the same group in the overall population. This ratio could suggest underrepresentation, proportional representation, or overrepresentation of a population experiencing a particular phenomenon.3 Metropolitan Statistical Area (MSA): An area that consists of one or more counties that contain a city of 50,000 or more inhabitants or contain a Census Bureau-defined urbanized area (UA) and have a total population of at least 100,000. Counties containing the principal concentration of population—the largest city and surrounding densely settled area—are components of an MSA. Racial Parity: Racial parity exists when a particular racial/ethnic group experiences an outcome equal to the percentage of their respective portion of the population. Racial parity is achieved when there is no underrepresentation or overrepresentation between racial/ethnic groups. Racial Profiling: The discriminatory practice by law enforcement officials of targeting individuals for suspicion of crime based on the individual's race, ethnicity, religion or national origin. Criminal profiling, generally, as practiced by police, is the reliance on a group of characteristics they believe to be associated with crime. Examples of racial profiling are the use of race to determine which drivers to stop for minor traffic violations (colloquially referred to as "driving while Black or Brown"), or the use of race to determine which pedestrians to search for illegal contraband.4 Spread: The difference between the over or underrepresentation of two races/ethnicities’ disproportionality. 2 “Disproportionality and Disparities,” Rowena Fong, Ruth G. McRoy and Alan Dettlaff; Encyclopedia of Social Work 3 “Disproportionality and Disparities,” Rowena Fong, Ruth G. McRoy and Alan Dettlaff; Encyclopedia of Social Work 4 ACLU Racial Profiling Definition 5 Texas Commission on Law Enforcement (TCOLE): The state agency responsible for the licensure of peace officers in the state of Texas. Zero Disparity: A goal set by Austin City Council that seeks to achieve racial parity and equity. In essence, motor vehicle stops of members of a particular racial/ethnic group should equal the percentage of their respective portion of the population. Zero disparate impact means all groups would experience outcomes at the same percentages. Methodology The goals of this analysis are as follows: 1. Examine how often specific groups of people, identified by race/ethnicity, are stopped by APD compared to their portion of the population; and 2. Explore whether disparate impact exists in the decisions following a traffic stop (e.g. search, arrest, citation, or warning). Relevant Population Base While this report explores data related to motor vehicle stops, driving age is not a readily available dataset. As a result, this report uses voting age data from the 2010 census as the relevant population base.5 Disproportionality and Racial Disparity Racial disparity exists when a particular racial/ethnic group experiences an unequal outcome compared to other groups despite being confronted with the same circumstances. When investigating racial disparity, it is important to not only analyze outcome data, but also disproportionality data. Disproportionality data provides insight as to how the outcomes experienced by a racial/ethnic group compare to that group’s share of the population, and whether the group is underrepresented, overrepresented, or proportionally represented. To determine underrepresentation and overrepresentation in traffic stops, one must calculate the difference between a racial/ethnic group’s percentage of the population and that group’s percentage of stops or outcomes. A racial/ethnic group that experiences a percentage of vehicle stops that is less than its share of the population is underrepresented. On the other hand, a racial/ethnic group that experiences a percentage of vehicle stops that is more than its share of the population is overrepresented. Calculations for Racial Disparity Disparity is calculated in this report using two decimal points for motor vehicle stop percentages and two decimal points for percent of population. When reported in charts and graphs, the disparity is rounded to the nearest whole number percentage point (e.g. –2.51% is rounded up to –3%, or –2.49% is rounded down to –2%). This report rounds to the nearest whole number percentage point when referring to percentages in the text. 5This report uses the following population data as defined by the Census Bureau in the 2010 Decennial Census question P10: Black/African American = Black or African American alone and Asian = Asian alone. Census Bureau 2010 Census question P11: White/Caucasian = White alone, not Hispanic or Latino and Hispanic/Latino = Hispanic or Latino 6 Table 1: Model Calculation of Motor Vehicle Stop Proportionality6 KEY Formula for Calculating Proportionality: A/B = C, D/E = F, and C-F = G. G = Racial disparity Use of General Population Data to Analyze Racial Disparity The use of general population data to analyze racial disparity issues is widely accepted. For example, Title VI of the Civil Rights Act prohibits discrimination based on race, color, and national origin in any program or activity receiving federal financial assistance. In giving guidance to investigators, the Department of Justice Civil Rights Division Title VI Legal Manual states, “In certain types of cases involving whole areas, like cities, counties, or states, the investigating agency may use general population data where everyone in that population may be affected.”7 When analyzing disparate impact of the decision points after a traffic stop, we follow legally established analytical frameworks affirming that the relevant population base for adverse disparate impact is the subset of the population that is affected by the decision.8 This disparate impact analysis avoids sample bias.9 The number of outcomes per race/ethnicity is divided by the number of motor vehicle stops experienced by that specific race/ethnicity to demonstrate what percentage of motor vehicle stops result in a warning/field observation, citation, search, and arrest by race/ethnicity. Data Sources Traffic stop and race/ethnicity data: APD produces an annual report on racial profiling that includes the number of vehicle stops it conducts, categorized by the race/ethnicity of the driver. This joint analysis focuses on the four most populous races/ethnicities in Austin: White/Caucasians, Black/African Americans, Hispanic/Latinos, and Asians. Traffic stop outcome data: APD released data on 2019 vehicle stops categorized as warnings, field observations, citations, and arrests. Information about motor vehicle stops is collected by police officers on Texas Commission on Law Enforcement (TCOLE) forms and then added to the APD Reports web 6 Total motor vehicle stops includes all motor vehicle stops, not just those of the four racial/ethnic groups analyzed in this report. Total population also uses the total population of people living the city of Austin, not just the population of the four racial/ethnic groups analyzed in this report. Disparity is reported at the whole number level. Motor vehicle stop percentages and percent of population uses two decimal points. 7 Section VII: Proving Discrimination – Disparate Impact, Pages 22-24, Title VI Legal Manual 8 Section VII: Proving Discrimination – Disparate Impact, Page 21, Title VI Legal Manual 9 Sample bias occurs when some members of a population are systematically more likely to be selected in a sample than others. Sampling bias limits the generalizability of findings. 7 page.10 The report team accessed this vehicle stop data through the City of Austin Open Data Portal. For information related to reasons for arrests, the report team requested and received data from APD directly. Population data: This report utilizes voting-age population data from the 2010 Census instead of the American Community Survey (ACS) for several reasons. First, the Census Bureau says the 2010 Census is 99.99% accurate, while the margins of error for annual ACS population data for those 18 and older in Austin are, on average, more than 30% for Black/African Americans and 9% for White/Caucasians per age category. While the population of the 4 most populous races/ethnicities grew by 25% from the 2010 Census to the 2018 ACS data, the share of the population for each of the 4 races/ethnicities changed an average of .5% in that time when comparing the 2010 Census data to the 2018 ACS data. The average difference between using the 2018 ACS data and the 2010 Census data in 2019 total motor vehicle stop disparities for the 4 most populous races/ethnicities in Austin is 1% per race/ethnicity. 10 Austin Police Department Reports webpage. 8 DATA ANALYSIS Proportion of Stops by Race Since 2015, racial disparity in motor vehicle stops has persisted. While some modest gains have been made, they do not begin to close the gap between the most recent data and racial parity. To illuminate how far conditions are from racial parity, one can analyze historical trends in the spread in stops among different races.11 In 2019, the spread between White/Caucasians and Black/African Americans in motor vehicle stops was 12%. In 2015, the spread between these two groups was 8%, with Black/African Americans overrepresented by 5% and White/Caucasians underrepresented by 3%. While the gap between motor vehicle stops for White/Caucasians and Black/African Americans improved by 2% from 2018 to 2019, this does not represent significant progress toward racial parity. Table 2: Proportionality by Race/Ethnicity of all Motor Vehicle Stops in 201912 From 2018 to 2019, the total number of motor vehicle stops increased by 54,064. In 2019, APD data showed that White/Caucasians were stopped 65,704 times, representing 47% of all motor vehicle stops of that year. White/Caucasians represented 54% of Austin’s population, meaning White/Caucasians were underrepresented in motor vehicle stops by 6%.13 In contrast, Black/African Americans were stopped 19,520 times and represented 8% of Austin’s population. In 2019, Black/African Americans were stopped at 14% and thus were overrepresented by 6%. Hispanic/Latinos were overrepresented by 2% in 2019, and Asians were underrepresented by 3%. 11 “Spread” is the difference between over and under representation. See Terminology section for full definition. Any difference in previous years‘ values between this and previous reports are due to adjustments in rounding. 12 The ”Motor Vehicle Stops % of Total“ column in this chart do not add up to 100% but rather 98% because this report compares each racial/ethnic group to the total motor vehicle stops (139,445), including those of races/ethnic groups not included in the report. The ”City of Austin Over 18 % of Population“ column in this chart similarly does not add up to 100% but rather 99% because this report uses the total city of Austin over 18 population (614,925 in the 2010 Census), including people not in the four races/ethnic groups of this report. 13This report uses data for its calculations at the decimal level. The % of motor vehicle stops for White/Caucasian drivers is 47.10%, and the White/Caucasian share of the Austin population in the 2010 Census is 53.58%. When subtracting these numbers, the difference/disparity is -6.48%, which rounds to –6%. 9 Chart 1: Proportionality of Race/Ethnicity of All Motor Vehicle Stops from 2015-2019 In short, data from both 2018 and 2019 consistently indicates that the largest disparity between stops and proportion of the population within any racial/ethnic group continues to be amongst Black/African American and White/Caucasian motorists: Black/African American motorists are overrepresented by 6% and White/Caucasian motorists are underrepresented by 6%. Race Known Compared to Race Not Known in Motor Vehicle Stops APD collects data for each motor vehicle stop and indicates “whether the subject's race was known to the officer before the stop.”14 This data is reported by officers and is separated into Race Known versus Race Not Known categories. Between 2018 and 2019, there were increases in the proportion of motorists pulled over when their race was known before the stop: a 2% increase in stops involving Hispanic/Latinos and a 1% increase in stops involving Black/African Americans. Conversely, there was a 1% decrease in stops involving White/Caucasians when their race was known before the stop. 14 As defined in the Austin Police Department’s 2019 Racial Profiling Report Guide, published in the City of Austin’s Open Data Portal. 10 Further, the proportion of motorists pulled over when their race or ethnicity was not known before the stop did not change for any race or ethnicity during the same timeframe. The data indicates that Hispanic/Latino and Black/African American motorists were pulled over more often when their race was known before the stop than when their race was not known before the stop. Upon further examination of cases when race was known before the stop, Black/African Americans were substantially overrepresented compared to other races. In 2019, Black/African Americans comprised 18% of overall stops when race was known before the stop, a 10% overrepresentation compared to their share of the voting-age population. In contrast, White/Caucasian motorists comprised 50% of overall stops when race was known before the stop. While this is a 3% increase for this group compared to when race was not known before the stop, it is still a -4% underrepresentation when compared to their share of the voting-age population. Further, Asians and Hispanic/Latinos were pulled over more when their race was not known before the stop compared to when it was known. In 2019, APD officers reported that race was known before the stop in 4% of motor vehicle stops overall, a 1% decrease from 2018. Table 3: Proportion of Motorists Pulled Over if Race is Known before the Stop in 2019 When race was not known before the stop, Black/African American motorists comprised 14% of stops, Hispanic/Latino motorists comprised 33% of stops, White/Caucasian motorists comprised 47% of stops, and Asian motorists comprised 4% of stops. Black/African Americans were overrepresented in the Race Not Known category by 6% and Hispanic/Latinos were overrepresented by 2% when compared to their share of the voting-age population (see Table 2 for respective share of the voting-age population). In contrast, White/Caucasians and Asians were underrepresented by 7% and 2%, respectively. Reason for Motor Vehicle Stops As stated previously, APD officers document motor vehicle stops on forms provided by the Texas Commission on Law Enforcement (TCOLE). The form requires that officers enter a reason for the stop. APD data reflects the following 4 reasons for vehicle stops: • moving traffic violation; • pre-existing knowledge (i.e. warrant); • vehicle traffic violation (equipment, inspection, or registration); and • violation of law other than traffic. In 2019, moving traffic violations were the reason for three-quarters of all motor vehicle stops. Only 0.24% of a total of 139,445 stops were due to pre-existing knowledge (i.e., warrant). 11 Table 4: Reason for Motor Vehicle Stops in 2019 Similarly, most stops that resulted in arrests were also due to moving traffic violations. In 2019, moving traffic violations were the reason for 68% of stops that resulted in arrest. Meanwhile, only 3% of 2019 motor vehicle stops that resulted in arrests were due to pre-existing knowledge (i.e., warrant).15 Table 5: Reason for Stop that Resulted in Arrest in 2019 Chart 2: Reason for Stop that Resulted in Arrest in 2019 15 See the next section for more context on the role of warrants in arrests. 12 Across all categories of stops that resulted in arrest in 2019, Black/African Americans and Hispanic/Latinos were disproportionately represented. The data demonstrates that moving traffic violations comprised most of the reasons for motor vehicle stops and stops resulting in arrests. Table 6: Reason for Stop that Resulted in Arrest by Race in 2019 Austin-Round Rock Metropolitan Area Population Compared to City of Austin Demographics Analysis of Stops When compared against the broader Austin-Round Rock Metropolitan Statistical Area (MSA) voting-age population, racial disproportionality in vehicle stops is largely consistent.16 While there was a 1% decrease in overrepresentation of Black/African Americans between 2018 and 2019, this group remained overrepresented. Table 7: Racial Disparities in Motor Vehicle Stops in the Austin Metropolitan Area in 2019 Comparing the Austin-Round Rock MSA racial composition to APD racial profiling data for motor vehicle stops reveals further disproportionality. White/Caucasians were further underrepresented in motor vehicle stops at -12% in the Austin-Round Rock MSA, versus a -6% underrepresentation solely within the City of Austin. Black/African Americans were 7% overrepresented when compared with the population in the Austin- Round Rock MSA, versus 6% overrepresentation within the City of Austin. Hispanic/Latinos were overrepresented by 5% when compared with the population in the Austin-Round Rock MSA. Conversely, Asians were underrepresented by 1% when examining data from the Austin-Round Rock MSA. 16 Last year’s report utilized the 2018 Census Bureau ACS data for population by race/ethnicity. This year’s report uses the 2010 Decennial Census for consistency with the rest of the report, which uses the 2010 Decennial Census. 13 One suggestion for the overrepresentation in stops of Black/African Americans in Austin is a pattern of Black/African Americans residing in surrounding communities and commuting to Austin for work and entertainment. However, the data refutes this as the racial disparity is worse in the Austin-Round Rock MSA. Proportion of Searches, Hits, and High Compared to Low-Discretion Searches, by Race Searches Racial disparity in searches remained largely the same from 2018 to 2019, with the exception that White/Caucasian motorists were further underrepresented by 1% and Black/African Americans were further overrepresented by 1% in 2019. Chart 3: Proportionality by Race/Ethnicity of Motor Vehicle Stops Resulting in a Search from 2015-201917 17 2015 data is from the 2015 OPM report and 2015 APD Racial Profiling report. The 2016 APD Racial Profiling Report provides overall motor vehicle stop and search numbers for 2015 which are different from the other 2015 reports. According to the 2016 APD report “although the state requires the reporting of motor vehicle stops that result in a citation or arrest, we [APD] have modified this year’s report to include all motor vehicle stops.” This report uses the originally reported 2015 numbers when possible in order to maintain continuity with previous OPM report figures. The ratios remain very similar. There are slight discrepancies between data on the City of Austin Open Data Portal and official APD Racial Profiling report data. This report uses the APD figures when possible, however in the case of 2019 data, this report uses data from the City of Austin Open Data Portal because the 2019 APD Racial Profiling report does not separate various motor vehicle stop types by race. 14 In 2019, there were 9,454 searches initiated after APD motor vehicle stops. For the number of vehicle stops that resulted in searches, Black/African Americans were overrepresented by 18% and Hispanic/Latinos by 13%. In contrast, Asians and White/Caucasians were underrepresented by 5% and 25%, respectively. As with the overall stop data, the disproportionality in stops that resulted in searches highlights the challenge of achieving racial parity. Racial parity is reached when there is no underrepresentation or overrepresentation between racial/ethnic groups. In 2019, the spread between searches involving Black/African Americans and White/Caucasians was 43%. The spread between searches involving White/Caucasians and Hispanic/Latinos was 38%. In short, the disproportionality of searches of White/Caucasians compared to Black/African American and Hispanic/Latino populations underscores the gap between the existing circumstances and racial parity. Hit Rates According to the APD 2018 Racial Profiling Report, “productive searches or ‘hits’ are searches where contraband is found (e.g., drugs or weapons).”18 The search “hit rate” describes the percentage of searches that result in finding contraband. Between 2018 and 2019, there was a 2% decrease in hit rate for White/Caucasians and a 3% decrease in hit rate for Hispanic/Latinos. In contrast, there was a 1% increase in hit rate for Black/African Americans and a 1% increase in hit rate for Asians during the same time period Table 8: Search Hit Rates by Race/Ethnicity in 2019 In 2019, White/Caucasians had a 25% search hit rate, or 681 hits among 2,723 searches, while Black/African Americans had a 32% search hit rate, or 772 hits among 2,397 searches. There is a 7% difference in hit rate between White/Caucasians and Black/African Americans. However, based on their proportion of the population, White/Caucasians were underrepresented in motor vehicle searches by 25% and Black/African Americans were overrepresented by 18% (see Chart 3). The difference in hit rate may further be explained by the disproportionality of high and low-discretion searches. High versus Low-Discretion Searches Low-discretion searches are situations in which policy requires an officer to conduct a search, such as due to an arrest or a vehicle being towed. High-discretion searches can only be conducted when there is consent, probable cause, or contraband.19 In both 2018 and 2019, Black/African Americans were the only demographic more likely to receive a high- discretion search than a low-discretion search. Black/African Americans were 7.7% more likely (absolute percent increase) to have received a high-discretion search in 2019 than 2018. In the 2018 data, searches 18 APD 2018 Racial Profiling Report 19 The Science of Policing Equity: Measuring Fairness in the Austin Police Department While the search of a motor vehicle is normally exempted from the search warrant requirement, police do need a basis for the search. The most common reasons cited are consent, incident to arrest, probable cause, contraband in plain view, frisk for safety, and subject to towing; these are reported here. Many factors contribute to the existence of probable cause, but the basic premise is that probable cause requires facts or evidence that would lead a reasonable person to believe the vehicle contains contraband or evidence. 15 for Black/African Americans were almost equally high-discretion and low-discretion. In 2019, however, APD officers used their discretion to search Black/African Americans 58% of the time. Table 9: Discretion Classification for Search Types APD data from 2019 shows that Black/African Americans receive proportionally more high-discretion searches than other races/ethnicities. Since low-discretion searches are required by law and high- discretion searches are not, data shows that Black/African Americans received disproportionally more stops and optional searches than other races/ethnicities. Table 10: Racial Disparities between High and Low-Discretion Searches in 2019 In 2019, 32% of all high-discretion searches resulted in hits, and high-discretion searches yielded more hits than low-discretion searches. Black/African Americans received a larger total number of high-discretion searches compared to White/Caucasians (1,124 versus 755) despite Black/African Americans having a lower share of the total population. 16 Chart 4: Search Hit Rates for High-Discretion Searches by Race in 2019 Proportion of Warnings or Field Observations by Race According to the APD General Orders, a field observation is “documentation of a subject stop when there is not a corresponding incident report, supplement or citation for the stop.” A warning issued by an officer is a statement that the motorist has committed some offense but is being spared the actual citation. Officers use their discretion in deciding whether to issue a citation or warning. While Black/African Americans continue to be overrepresented in motor vehicle stops resulting in warnings or field observations, there has been a gradual shift toward racial parity since 2017.20 20 Warnings and Field Observations are reported in the same City of Austin Open Data Portal dataset by APD and are grouped in this analysis. 17 Chart 5: Proportionality by Race/Ethnicity of Motor Vehicle Stops Resulting in a Warning or Field Observation from 2015-2019 White/Caucasians were underrepresented by 3% in 2015 and were underrepresented by 1% in 2019 when compared to their share of the voting age population. Similarly, Hispanic/Latinos and Asians were also consistently underrepresented between 2015-2019. Proportion of Arrests by Race Since 2015, racial disparities in arrests resulting from motor vehicle stops have persisted and worsened. In 2019, Black/African Americans were the most overrepresented racial/ethnic group in this category, experiencing 17% overrepresentation in arrests resulting from motor vehicle stops. Hispanic/Latinos were the second most overrepresented at 13%. Chart 6: Proportionality by Race/Ethnicity of Motor Vehicle 18 Stops Resulting in Arrest from 2015-2019 Reason for Arrest This data was not made available for the inaugural Joint Report and represents a new depth of examination. According to the data, there are four scenarios in which people are arrested during motor vehicle stops. The scenarios include the following: • Outstanding warrant: indicates the driver was found to have a pre-existing warrant for arrest. • Violation of city ordinance: indicates the driver was found to have violated a local ordinance, such as playing amplified music from a vehicle. • Violation of penal code: indicates the driver was found to have committed a crime, such as possession of controlled substance, driving while intoxicated, or other violations of criminal law. • Violation of traffic law: indicates the driver was found to have violated a traffic law, such as driving with a suspended or invalid license. Table 11: Reason for Arrest in 2019 19 Chart 7: Reason for Arrest in 2019 In 2019, 67% of all arrests were made based on a violation of penal code. The data shows 27% of arrests from traffic stops were the result of an outstanding warrant. Six percent of drivers were arrested solely based on violations of traffic laws. Reason for Arrest by Race Black/African Americans and Hispanic/Latinos were overrepresented across every category of reason for arrest, while White/Caucasians and Asians were underrepresented. Table 12: Reason for Arrest by Race in 2019 Proportion of Citations by Race Racial disparities for Hispanic/Latino and Black/African American motorists worsened in terms of the proportions of citations by race/ethnicity. While Hispanic/Latinos were proportionately represented in 2015 20 APD motor vehicle citations, by 2019, they were overrepresented by 9%. Additionally, Black/African Americans remained consistently overrepresented by 5% in 2018 and 2019. Conversely, White/Caucasians grew increasingly underrepresented from -2% in 2015 to -12% in 2019. The trend over the last five years indicates a shift away from racial parity for every racial/ethnic group, except Asians, who were consistently underrepresented. Implications of the racial disproportionality of citations are explained in the Discussion section. Chart 8: Proportionality by Race/Ethnicity of Motor Vehicle Stops Resulting in Citation from 2015-2019 21 Geographic Disparity of Warnings, Field Observations and Arrests As shown in the map below, APD organizes the city into 10 sectors. Map 1: APD Sectors The geographic concentration of arrests and geographic concentration of warnings and field observations remained largely unchanged between 2018 and 2019. The maps that follow display occurrences of warnings, field observations, and arrests in 2019 by sector. The maps do not display occurrences of citations as APD is currently unable to map citations.21 Warnings and field observations occurred at a 21 See Data Recommendations section. 22 higher rate on the west side of the city, with the highest numbers taking place in Adam Sector, followed by Baker and David, respectively. Map 2: Number of Warnings and Field Observations by Sector in 2019 23 Chart 9: Number of Warnings/Field Observations by Sector in 2019 More arrests occurred on the east side of the city, with the highest numbers taking place in Charlie Sector, followed by Edward and Henry, respectively. 24 Map 3: Number of Arrests by Sector in 2019 25 Chart 10: Number of Arrests by Sector in 2019 The geographic disparity in warnings and field observations, as well as for arrests, illuminates how far existing conditions are from racial parity.22 Additional visualizations of the data used in the analysis of geographic disparities are publicly available on this Tableau dashboard, including visuals of motor vehicle stops by race, time, zip code, and intersection. Gender and Race In 2019, regardless of race, male motorists in Austin were more likely to be pulled over or arrested than female motorists. In terms of likelihood of being pulled over, the widest gap between the genders was experienced by Black/African American and Hispanic/Latino motorists. Among Black/African American drivers, Black/African American males represented 64% of APD motor vehicle stops in 2019. Meanwhile Black/African American females represented 36% of APD motor vehicle stops for the same period. Among Hispanic/Latino drivers, Hispanic/Latino males also represented 64% of APD motor vehicle stops in 2019. In contrast, Hispanic/Latino females represented 36% of APD motor vehicle stops for the same period. Table 13: Percentage of APD Motor Vehicle Stops by Race and Gender in 201923 The gender gap further widens when focusing on motor vehicle stops that resulted in an arrest. In 2019, there was a 60% difference between Hispanic/Latino men and women and a 56% difference between Black/African American men and women for motor vehicle stops that resulted in an arrest. 22 See Discussion section for further explanation. 23 The contrast between the difference totals for Black/African Americans (28%) and Hispanic/Latinos (29%) is due to rounding. 26 Table 14: Percentage of APD Motor Vehicle Stops that Resulted in Arrest by Race and Gender in 201924 The gender and race data highlights additional disparities in stops and arrests. The data indicates that in 2019, among Black/African American motorists and Hispanic/Latino motorists, males in both groups were not only the most stopped, but also most likely to be arrested after a motor vehicle stop. Motor Vehicle Stop Outcome Totals for the Four Most Populous Races/Ethnicities In 2019, APD conducted a total of 136,496 motor vehicle stops for the four most populous races/ethnicities.25 From 2015 to 2019, the total number of stops for the four most populous races/ethnicities decreased by 16,219. The overall number of citations for the four most populous races/ethnicities decreased by 55,217 and the overall number of arrests for the four most populous races/ethnicities decreased by 1,439. However, warnings increased by 108%, or 40,437, for the four most populous races/ethnicities between 2015 and 2019. In response to questions about increased stops, increased warnings, decreased citations, and decreased arrests in 2019, APD responded that a few variables may have contributed to these changes. However, APD reported that there was not a readily identifiable or specific cause, such as a policy change or formal directive. 24 The calculated difference (35%) between White/Caucasian males and White/Caucasian females is due to rounding. 25 The four most populous races/ethnicities in this report (Asian, Black/African American, Hispanic/Latino, and White/Caucasian), represents on average more than 97% of APD motor vehicle stops per year since 2015. 27 Chart 11: APD Motor Vehicle Stops by Outcome by Year from 2015-2019 Despite the changes in total stops, citations, warnings, and arrests, racial disproportionality in these areas has continued and has generally worsened since 2015. This is concerning, as it conveys a persistent problem of racial bias in policing. Outcome Percentages Outcome percentages are calculated by the number of outcomes per race/ethnicity divided by the number of motor vehicle stops of that race/ethnicity. The results of this calculation demonstrate what percentage of motor vehicle stops result in a warning/field observation, citation, search, and arrest by race/ethnicity. Table 15: Model Calculation of Post-Stop Outcome Percentages A B C Outcome Type Motor Vehicle Stops per Demographic Total of Outcome Type per Demographic Percent Outcome Field Observation/Warnings 5,517 3,555 64% Citations 5,517 1,875 34% Arrests 5,517 87 2% KEY Formula for calculating outcome percentages: B/A = C C = Percent of an outcome per race/ethnicity and is compared to other outcome percentages per race/ethnicity. Comparing racial/ethnic groups in this way illustrates any differences in outcomes for each group without factoring in citywide population proportionality. As a result, patterns and trends are more clearly highlighted. Going forward, the Joint Report will use this method to analyze whether APD stop data reveals racial disparity and, as the Joint Report matures, this analysis will become the principal measure to identify potential disparity. 28 Set side by side, comparing data across the four most populous races/ethnicities included in this report highlights differences in the frequency of outcomes within each group. Motor vehicle stops with White/Caucasians and Asians resulted in a higher percentage of warnings/field observations. Motor vehicle stops with Black/African Americans and Hispanic/Latinos resulted in a higher percentage of searches and arrests. Motor vehicle stops with Hispanic/Latinos resulted in the highest percentage of citations. Table 16: Percent of Stops Resulting in a Warning/Field Observation by Race/Ethnicity in 2019 In 2019, White/Caucasians and Asians had a higher percentage of motor vehicle stops that resulted in warning/field observations at 63% and 64%, respectively. On the other hand, 55% of motor vehicle stops with Black/African Americans and 48% of motor vehicle stops with Hispanic/Latinos resulted in a warning/field observation. Table 17: Percent of Stops Resulting in a Citation by Race/Ethnicity in 2019 At 44%, Hispanic/Latinos had the highest percentage of motor vehicle stops that resulted in a citation. White/Caucasians, Asians, and Black/African Americans had similar percentages of vehicle stops that resulted in a citation at 33%, 34%, and 35%, respectively. Table 18: Percent of Stops Resulting in an Arrest by Race/Ethnicity in 2019 At 10%, Black/African Americans had the highest percentage of motor vehicle stops that resulted in an arrest. At 7%, Hispanic/Latinos had the second highest percentage of vehicle stops that resulted in an arrest. White/Caucasians and Asians had vehicle stops that resulted in an arrest at 4% and 2%, respectively. Once stopped, Black/African Americans were approximately three times more likely to be arrested than White/Caucasians.26 26 After being stopped, 9.9% of stops with Black/African Americans resulted in arrest. After being stopped, 3.5% of stops with White/Caucasians resulted in arrest. Black/African Americans were arrested 2.8 times more than White/Caucasians. These numbers appear differently in Table 17 because of rounding. 29 Table 19: Percent of Stops Resulting in a Search by Race/Ethnicity in 2019 At 12%, Black/African Americans experienced the highest percentage of motor vehicle stops that resulted in a search. At 9%, Hispanic/Latinos experienced the second highest percentage of vehicle stops that resulted in a search. White/Caucasians and Asians experienced vehicle stops that resulted in a search at a rate of 4% and 2%, respectively. Once stopped, Black/African Americans were three times more likely to be searched than White/Caucasians. Chart 12: Percentages of Outcomes of Motor Vehicle Stops Involving Asian Adult Population in 2019 30 Chart 13: Percentages of Outcomes of Motor Vehicle Stops Involving Black/African American Adult Population in 2019 Chart 14: Percentages of Outcomes of Motor Vehicle Stops Involving Hispanic/Latino Adult Population in 2019 31 Chart 15: Percentages of Outcomes of Motor Vehicle Stops Involving White/Caucasian Adult Population in 2019 Comparison of 2018 and 2019 Outcome Data By comparing data from 2018 to 2019, this analysis examined changes in rate of outcome by race. The percentage of stops resulting in a warning/field observation increased for all races/ethnicities. In contrast, the percentage of stops resulting in a citation, arrest, and search decreased for all races/ethnicities. Table 20: Percent Change in Stops Resulting in a Warning/Field Observation by Race/Ethnicity from 2018-2019 Between 2018 and 2019, the total number of warnings/field observations increased by 29,585, an absolute percentage increase of 59%.27 All racial/ethnic groups experienced an increased percentage of motor vehicle stops that resulted in a warning/field observation. White/Caucasians and Asians experienced the highest percentage increase at 16%. The percentage of motor vehicle stops with Black/African Americans and Hispanic/Latinos that resulted in a warning/field observation increased by 13% and 14%, respectively. 27 See Appendix 3 for 2018 outcome totals and percentages. Absolute percentage is defined as the difference in a number over two periods in time or between two population types. For instance, if one observed a percent of 6% in 2018 and 7% in 2019, the absolute percentage difference would be 1%. 32 Table 21: Percent Change in Stops Resulting in a Citation by Race/Ethnicity from 2018-201928 Between 2018 and 2019, the total number of citations decreased by 5,972, an absolute percentage decrease of 10%. All racial/ethnic groups experienced a decreased percentage of motor vehicle stops that resulted in a citation. White/Caucasians and Asians experienced the largest decrease of 14% for both groups. For Black/African Americans and Hispanic/Latinos, the percentage of motor vehicle stops that resulted in a citation decreased by 7% and 9%, respectively. Table 22: Percent Change in Stops Resulting in an Arrest by Race from 2018-2019 Between 2018 and 2019, the total number of arrests decreased by 3,381, an absolute percentage decrease of 30%. All racial/ethnic groups experienced a decreased percentage of motor vehicle stops that resulted in arrest. Black/African Americans and Hispanic/Latinos experienced the largest decrease, marked by percentages that went down by 6% and 5%, respectively. For White/Caucasians, the percentage of motor vehicle stops that resulted in an arrest decreased by 2%. For Asians, the number decreased by 0.8%. Table 23: Percent Change in Stops Resulting in a Search by Race from 2018-2019 Between 2018 and 2019, the total number of stops resulting in a search decreased by 3,100, an absolute percentage decrease of 25%. All racial/ethnic groups experienced a decreased percentage of motor vehicle stops that resulted in a search. For Black/African Americans and Hispanic/Latinos, the percentages decreased by 5%, the largest decrease. For White/Caucasians and Asians, the percentage of motor vehicle stops that resulted in a citation decreased by 2% and 1%, respectively. 28 For Hispanic/Latinos, the calculated difference in citations received in 2019 compared to 2018 (-9%) is due to rounding. The number of 2018 citations per motor vehicle stops for Hispanic/Latinos was 53.7%, in 2019, that same number was 44.5% for a difference of -9.3%. 33 DISCUSSION Black/African Americans and Hispanic/Latinos are Repeatedly Overrepresented An analysis of traffic stops alone shows disparities among race, geography, and gender. Once disaggregated by race, the data shows disparity in the number of searches, use of discretion within searches, warnings, citations, and arrests. Across all these areas, Black/African Americans and Hispanic/Latinos were overrepresented, while White/Caucasians and Asians were underrepresented. An analysis looking back across five years of data reveals that racial disparity has persisted and, in many cases, is worsening. In every year since 2015, Black/African Americans have been the most overrepresented in every category, which includes motor vehicle stops, warnings, field observations, searches, citations, and arrests. In the same time frame, White/Caucasians have been the most underrepresented in every category. Racial Disparity in Searches In 2018 and 2019, Black/African Americans were the only demographic to receive more high-discretion searches than low-discretion searches. Compared to 2018, 2019 saw police conducting an even higher proportion of high-discretion searches versus low-discretion searches with Black/African Americans. Racial Disparity in Citations The growing racial disparity in citations poses complex equity concerns. Research from the Brennan Center for Justice at New York University School of Law examined the efficacy and equity of fines and fees associated with the criminal justice system in ten states, including Texas. Findings from the research reveal that citations are more frequently levied on low-income communities of color and are ultimately more burdensome on less-affluent communities.29 Further, the report states that fines and fees are an inefficient source of government revenue. In 2017, Travis County spent almost $4.8 million in administering misdemeanor and traffic courts, as well as $4.6 million jailing those who failed to pay fees and fines.30 Geographic Disparity Geographic disparity was consistent from 2018 to 2019, with more arrests occurring on the east side of the city, and more warnings on the west side of the city. Citation data was not usable for this geographic analysis because, unlike warnings/field observations, it is collected in a write-in format, rather than geolocation (see Data Recommendations section). Geographic disparity highlights issues around gentrification and economic equity that often intersect with race. 5-Year Analysis Reveals Racial Disparity in Policing is a Persistent Problem Over the last five years, the total number of warnings, field observations, citations, and arrests shifted. Even with these changing numbers, racial disparity continued and generally worsened. This trend indicates a persistent problem. 29 Menendez, Matthew, et al. The Steep Costs of Criminal Justice Fees and Fines: A Fiscal Analysis of Three States and Ten Counties. Brennan Center for Justice at New York University School of Law, 21 Nov. 2019, www.brennancenter.org/sites/default/files/2019-11/2019_10_Fees%26Fines_Final5.pdf. 30 Menendez et al. 34 Outcome Percentages Show Disparate Impact on Black/African Americans and Hispanic/Latinos In analyzing the decision to search, arrest, cite, or give a warning, the data showed that stops involving Black/African Americans and Hispanic/Latinos resulted in higher percentages of searches and arrests. Once stopped, Black/African Americans were three times more likely to be searched and were approximately three times more likely to be arrested than White/Caucasians. The percentage of stops involving Asians and White/Caucasians resulted in a higher percentage of warnings/field observations. Stops involving Hispanic/Latinos resulted in the highest percentage of citations. Outcome percentages are not equal among the four most populous races/ethnicities in Austin. 35 RECOMMENDATIONS In order to address racially disproportionate impacts in policing and meet the City’s equity goals, the Office of Police Oversight, Equity Office, and Office of Innovation recommend that APD take the following steps: Acknowledge 1. Acknowledge that racial disparity has persisted and, in many cases, worsened over the last 5 years. 2. Acknowledge that the Department’s efforts to address racial disparity have not worked, and that APD is not on track to meet the vision of Fair Administration of Justice identified by City Council and the community within Strategic Direction 2023. Commit 3. Commit in writing to implementing the recommendations from the inaugural Joint Report: Analysis of APD Racial Profiling Data in order to align with the goals of Resolution 50 and Strategic Direction 2023, Fair Administration of Justice. 4. Commit in writing to the goal of zero racial disparity and to publicly sharing the Department’s efforts toward eliminating racial disparity, as well as the results of such efforts, whether successful or not. • Such discussion should appear annually in the APD Racial Profiling Report. • Specific changes and their measured impacts must be included in APD’s public reports. Oversight of this process will be conducted by the Office of Police Oversight, the Equity Office, and the Office of Innovation to ensure the analyses and reports are done properly and with proportionality based on race/ethnicity and other demographic information. Proportionality is essential to this discussion, as City Council used the inaugural Joint Report as a baseline measurement of racial disparity in citations, field observations, warnings, and arrests resulting from vehicle stops for Strategic Direction 2023 metrics. The Joint Report utilizes proportional analysis to gauge racial disparity; for transparency and consistency, APD should do the same. 5. Commit in writing to meeting and collaboratively engaging with the community to address racial disparity in policing. Engage with the Community 6. Work with the community to create a strategic plan to eliminate racial disparity in policing. • The plan should include benchmarks to measure the progress of Department changes within APD General Orders, training, procedures, community interactions, reporting, and any other related areas. 7. Meet and collaboratively engage with the community on a quarterly basis. • These meetings should center communities that are disparately impacted by APD in relation to searches, arrests, and uses of force. 8. Work with the community to establish progress benchmarks and report back to City Council and the City-Community Reimagine Public Safety Task Force. • Currently, APD does not have its own strategic plan to eliminate racial disparity in policing. As stated in the inaugural Joint Report, any racial disparity more than zero is unacceptable. • APD needs benchmarks in order to both comply with City Council’s Resolution 50 and track progress toward the goal of zero disparity. APD should commit in writing to asking the community for acceptable benchmarks and take responsibility for meeting them. 9. Meet the benchmarks, recommendations, and directives put forth by the community, City Community Reimagine Public Safety Task Force, and City Council. 36 • APD has not stated its responsibility and progress toward achieving goals outlined in Resolution 50 and Strategic Direction 2023, Fair Administration of Justice. • In order to meet these goals, the Department should acknowledge its standing in achieving these goals through quarterly meetings with the community. Train and Intervene 10. Acknowledge the role of officer discretion in racial disparity and implement training and intervention measures aimed at addressing it. • The inaugural Joint Report recommended an examination of the role that race plays in police officer decisions and discretion. APD responded by saying “[a]dditional data and analysis is necessary to determine how officer discretion, Departmental procedures, and societal factors contribute to these disproportionalities.”31 • This Joint Report expands the analysis of searches, and the data reveals more about the disproportionate use of police officer discretion in searches with Black/African Americans. • To be clear, officers’ discretion to conduct searches yields a clear disparate impact on Black/African American drivers. The Department should acknowledge the persistently unequal outcomes that result from officer discretion and take both proactive and remedial steps to correct this. • In addressing the role of officer discretion, APD should utilize recommendations from the City-Community Reimagining Public Safety Task Force and the third-party contractor audit of APD communications findings related to the cultural and educational factors contributing to racial disparity. Additionally, the Office of Police Oversight, Equity Office, and Office of Innovation renew all recommendations set forth in the inaugural Joint Report on which APD has yet to taken action.32 31 See Appendix 1 for more information. 32 See Appendix 1 for more information. 37 DATA RECOMMENDATIONS Data analysis is a key component in increasing transparency and accountability. The work of data analysis is less inhibitive when the owners of data ensure that it is properly collected on the front end. Currently, not all APD data is collected or reported uniformly. Additionally, there are unavailable data sets because of APD’s method of collecting and/or reporting. In order to best track progress towards zero racial disparity in policing, data must be accurately collected, publicly available, and usable for analysis. The following recommendations were informed by the data cleaning process conducted by this report team. Collect and Publish Additional Data ● Arrest data cannot currently be attributed to one officer, which poses challenges in understanding officer behavior on an individual level. This data should be collected. APD has access to APD officer data that can be traced from year to year. Tracking officer-level data annually could serve to better observe and understand behavioral trends amongst officers. Additionally, use of force and response to resistance data associated with motor vehicle stops should be linked and shared. ● Furthermore, in order to create a clear picture of policing across the city, ongoing officer activity data should be collected, linked to initiation and outcomes, and shared. On August 21, 2020, the City of Austin’s Office of Innovation released the Calls for Service Trends: January 2017-June 2020 Dispatched vs. Officer Initiated. The APD data set used for this analysis detailed activities such as directed patrol and checking area by county but was not linked to officer activity outcomes. Additional information, such as how many officers were deployed doing these activities, individual officer coded activities, associated geographic coordinates, demographic information of the citizens involved in officer interactions, and linked outcomes can provide clarity surrounding the differences among officers and geographic areas. Engage in Uniform Data Collection and Entry ● Citations are not geographically tracked in the same way as field observations, warnings, or arrests. For citations, the geographic data collected is significantly incomplete, and thus APD omits this data. Geographic data for citations should be collected in the same manner as field observations, warnings, and arrests so that the data is complete and usable. Data containing the X and Y coordinates and ZIP codes were collected for warnings and field observations and arrests, but not for citations. A geographic analysis was not possible for citations because of the missing data. ● Column names were also not uniform. For example, data regarding the county of arrest was under the column titled “County_description” but for warnings the column was labeled “County_desc”. 38 ● APD data is entered utilizing different labels. For example, here is how race is stored in the Arrests table: Here is how race is stored in the Warnings and Field Observations table: ● The lack of uniformity across different data topics makes it more difficult to analyze the data, thereby undermining the utility of its collection. For the purposes of this report’s analyses, data was cleaned both manually and utilizing Python script. Uniform data entry is recommended. Use Data Entry That Is Recognizable to Analytical Software ● Geographic data was collected and shared on the City of Austin Open Data Portal as spherical coordinates rather than standard latitude and longitude coordinates. Most analytical software cannot use spherical coordinates. Longitude and latitude coordinates should be used going forward because they are the most common geographic coordinate system. 39 ● Time is currently stored as a number and not as an actual timestamp. For example, the time 23:19 PM is stored as the number 2319 as seen here: Table 24: APD Time Data Storage 2019 ● To undertake time-based analysis, one would need to break down the numbers into the last two digits for the minutes, and the first one or two digits for the hours, add a colon between them, and then convert this to a timestamp. Data should be stored in timestamp format so it can be more easily used. Provide Retroactive Data-Cleaning for The Public ● The purpose of data collection is to understand patterns and trends. This can most easily be accomplished if the data is stored conveniently on the City of Austin Open Data Portal in a method that can be analyzed. In the meantime, it should be APD’s responsibility to make the data presentable and usable for the public. Our data team has volunteered to contribute all data- cleaning scripts to APD should the Department need to use them. 40 CONCLUSION To begin rebuilding trust with the communities that have been negatively impacted by inequitable policing practices, the Austin Police Department must make every effort to address the racial and ethnic disparities identified in this report and fulfill the accompanying recommendations. The data contained in this report will be crucial to our efforts to reimagine public safety, eliminate racial disparities, and achieve the goals of equitable policing and the fair administration of justice. Zero disparity in motor vehicle stops is a goal that APD must work collaboratively with other stakeholders to achieve, including City management and departments, City Council, and the community. This is the path for APD to fulfill its mission and vision to protect and serve all the communities in Austin effectively. 41 APPENDICES Appendix 1 Summary of Recommendations in the January 2020 Joint Analysis and APD Response: January 2020 Recommendations APD Response Rec 1: Acknowledge that racial disparity exists and is worsening. The Austin Police Department consistently and unequivocally acknowledges that racial disparities exist throughout aspects of our city, including police enforcement actions. Accordingly, the Department has readily taken many steps to address the disparities within APD’s purview over the past five years, as detailed in the January 14, 2020 response. Racial disparities have persisted despite these efforts, and the widening of certain gaps has raised additional concerns that demand further attention and analysis. Rec. 2: Acknowledge that the methodology previously used omitted the context of proportionality and therefore was an incomplete analysis. This resulted in a perception that a trend of disparity did not exist. The primary purpose of APD’s annual racial profiling report is to comply with state legislative mandates that require the reporting of specific data. Proportionality assessments are not compulsory. However, recognizing the importance of such information, APD collaborated with the Center for Policing Equity to conduct a comprehensive analysis of the racial disparities manifested in the Department’s enforcement actions. The report was the first to apply the National Justice Database’s independent analytic framework to police data made available through President Obama’s Police Data Initiative, Measuring Fairness in the Austin Police Department. That report is posted alongside the Department’s racial profiling reports on the City’s website. http://austintexas.gov/sites/default/files/files/Police/Austin_PDI_Report_ 2016_Release.pdf Rec 3: Acknowledge that race plays a major role in who we stop, search, and for whom we use discretion favorably. The Department acknowledges that the outcomes of many police activities result in racial disparities. Additional data and analysis is necessary to determine how officer discretion, Departmental procedures, and societal factors contribute to these disproportionalities. Rec 4: To gain community trust, proportional racial disparity in motor vehicle stops, arrests, searches, field observations, warnings, and citations should be zero. The Department is committed to reducing racial disparities to zero, particularly disparities that are the result of officer discretion. 42 Rec 5: The official comprehensive analysis of racial profiling shall be conducted and released by the City of Austin Office of Police Oversight, although state-mandated reporting may continue under the purview of the Chief. The Department will continue to release its state-mandated racial profiling report on an annual basis and welcomes the Office of Police Oversight’s independent analysis and insight, in the manner the City Manager deems necessary and appropriate. Rec 6: In order to uphold data integrity, accuracy, and transparency, officers should verify the racial and ethnic identity with people they stop. The verified data should be documented in officer reports and be published in the Racial Profiling data sets on the City’s Open Data Portal. In accordance with departmental procedures, Officers are required to document the race and ethnicity of the individuals they stop. The City has contracted with Dr. Alex Del Carmen, an expert on racial profiling and discrimination, to regularly audit the Department’s racial profiling data to ensure accuracy in data collection and reporting. The traffic stop data, which includes race, is published in the racial profiling datasets on the City’s Open Data Portal. Rec 7: Analyze and report on the operational inefficiencies and costs that disproportionate racial disparities create by the second quarter of the fiscal year 2020 and provide to the City Manager and Council. Currently, the Department is not staffed or equipped to quantify and analyze this data but would readily collaborate with the City Auditor’s office or another entity, at the direction of the City Manager. Rec 8: Explore promising practices from Oakland and Nashville that use a scoring mechanism for disproportional behavior to identify at-risk officers and assign appropriate interventions and use in the determination of promotions. The Department agrees the City should invest in sophisticated oversight tools that are more adept at identifying, flagging, and tracking at-risk officers in order to facilitate timely and effective interventions. Rec 9: Include implicit bias testing in the Austin Police Department hiring process. Based on the best available evidence from subject-matter experts on bias, the Department operates with the understanding that every applicant will have implicit biases. Therefore, the Department administers training to ensure all employees are aware of their biases, promulgates explicit policies to set clear expectations that bias-based actions are intolerable, and utilizes oversight mechanisms to identify inappropriate behavior. Rec 10: For current employees, require implicit bias testing and flag high-scoring officers for appropriate intervention. As stated above, the Department has mechanisms in place to identify and rectify inappropriate behavior. Additionally, the Department is open to exploring proven, evidence-based testing methods that are capable of effectively supplementing current training, policies, procedures, and audits. 43 Rec 11: Identify and implement bias- countering policies, practices, methods, processes, and standard operating procedures to mitigate bias. The Department recommends the City contract with a suitable academic institution to conduct an independent, comprehensive, and evidence-informed assessment of the Department’s enforcement practices, cultural norms and customs, training, accountability procedures, and any resulting racial disparities. A similar partnership between the City of Oakland and Stanford University yielded promising results and provided a roadmap for creating community- based strategies aimed at addressing the unique historical and cultural challenges of a particular city: Data for Change & Strategies for Change. Rec 12: Include the comprehensive Racial History of Policing curriculum in the cadet training academy and adapt it into required training for existing officers, at all ranks, annually. The Department intends to incorporate the Racial History of Policing training in future cadet classes and is determining the best approach and frequency for administering the training to existing officers. Rec 13: Follow the guidelines for racial equity training established by the Equity Office. The Equity Office and Office of Police Oversight shall be consulted for final selection of official racial equity training for officers at all ranks. The Department is committed to following the established guidelines for racial equity training and welcomes input from the Equity Office and Office of Police Oversight. Rec 14: Develop a method to provide racial equity training on an ongoing basis (a minimum of 40 hours per year) for all staff, sworn and civilian, in the department, annually, during every year of service. The Department is eager to provide additional racial equity training for all employees in an effective, feasible, and sustainable manner. The Department will consider this recommendation as part of the FY21 budget process. 44 Appendix 2 Chart 16: Motor Vehicle Stops Per Year for the Four Most Populous Races/Ethnicities from 2015-2019 Appendix 3 Table 25: Percent of Stops Resulting in a Warning/Field Observation by Race/Ethnicity in 2018 Table 26: Percent of Stops Resulting in a Citation by Race/Ethnicity in 2018 45 Table 27: Percent of Stops Resulting in an Arrest by Race/Ethnicity in 2018 Table 28: Percent of Stops Resulting in a Search by Race/Ethnicity in 2018
feedback

Did you find what you were looking for on this page?