Skip to content

Growing Pains for Hate Crime Statistics

Kai Wiggins

October 19, 2021

Introduction

When the Federal Bureau of Investigation (FBI) released its 2020 Hate Crime Statistics report on August 30, 2021,1Press Release, Fed. Bureau of Investigation, FBI Releases 2020 Hate Crime Statistics (Aug. 30, 2020), https://​www.​fbi.gov/news/pressrel/press-releases/fbi-releases-2020-hate-crime-statistics. many were caught off guard. Journalists, researchers, and civil rights organizations have come to expect the preceding year’s report each November. And because the FBI typically shares the data with some parties in advance, a wave of news articles, in-depth analyses, and press releases often floods the internet the moment the report goes live. This year, the FBI published its report almost three months ahead of schedule.2See Fed. Bureau of Investigation, UCR Program Quarterly 8 (Feb. 2021), https://www.fbi.gov/file-repository/ucr/ucr-program-quarterly-february-2021.pdf/view. The data received some attention, but the wave never crashed.

Some might question the Biden Administration’s approach to the 2020 release, which effectively preempted the rush of bad coverage that often accompanies the publication of hate crime data.3See, e.g., Arjun Singh Sethi, Opinion, The FBI Recorded a Surge of Hate Crimes Last Year. But It Undercounted – By a Lot, Wash. Post (Nov. 14, 2018), https://www.washingtonpost.com/​outlook/​2018/11/14/fbi-recorded-surge-hate-crimes-last-year-it-undercounted-by-lot. If we have learned anything about the nature and extent of hate crime in the United States, it is that hate crime statistics do not tell us much, and recent years have seen a push to reform, if not transform, the national reporting system.4Recently, congressional hearings have addressed hate crime underreporting, see Press Release, Rep. Jamie Raskin, Subcommittee Chairman Raskin Seeks Update on FBI Process to Collect and Report Hate Crimes (Feb. 12, 2020), https://raskin.house.gov/2020/2/​subcommittee-chairman-raskin-seeks​-update-fbi-process-collect-and-report-hate, and national civil rights organizations have prioritized hate crime statistics reform, see, e.g., Letter from The Leadership Conf. on Civ. & Hum. Rts. to U.S. Sens. & Reps. 12 (Jan. 9, 2019), http://civilrightsdocs.info/pdf/​policy/task-force-priorities/​2019-Leadership-Conference-Legislative-priorities-FINAL.pdf. In May 2021, for example, Congress passed the Covid-19 Hate Crimes Act, which includes provisions that encourage law enforcement agencies to submit hate crime data and adopt standardized reporting practices.5Barbara Sprunt, Congress Passes Bill to Counter the Rise in Anti-Asian Hate Crimes, NPR (May 18, 2021), https://​www.npr.org/2021/05/18/997847571/congress-passes-bill-to-counter-the-rise-in-anti-asian-hate-crimes. And for its part, the Biden Administration is working to improve hate crime statistics through increased coordination with state, local, and tribal partners.6See Memorandum from Merrick Garland, Att’y Gen., to Dep’t of Just. Emps. (May 27, 2021), https://​www.​justice.​gov/ag/page/file/1399221/download.

Aside from these developments, the FBI has quietly made three adjustments to the national reporting system in recent years. First, in addition to collecting hate crime data from state, local, and tribal law enforcement agencies, the FBI has started to collect data from federal law enforcement agencies as well. Second, the FBI began publishing interactive hate crime statistics instead of providing annual data in static, online reports. And third, all law enforcement agencies that submit hate crime data to the FBI are now required to do so using the National Incident-Based Reporting System (NIBRS), which captures more comprehensive and detailed information than its predecessor. These reforms predate the inauguration of President Joe Biden and, compared to more recent efforts in Congress and the Biden Administration to improve hate crime statistics, have received little attention. That is a problem, because even though each reform promises more accurate, reliable hate crime statistics in the long term, growing pains abound.

I. Federal Law Enforcement Reporting

The FBI publishes hate crime statistics pursuant to the Hate Crime Statistics Act (HCSA) of 1990, which, as amended, requires the Attorney General to “acquire data . . . about crimes that manifest evidence of prejudice based on race, gender and gender identity, religion, disability, sexual orientation, or ethnicity.”734 U.S.C. § 41305. The FBI carries out these responsibilities through its administration of the Uniform Crime Reporting (UCR) program, which publishes statistics based on crime data voluntarily submitted by the nation’s law enforcement agencies.8Uniform Crime Reporting Program, Fed. Bureau of Investigation, https://www.fbi.gov/​services/​cjis/ucr (last visited Oct. 4, 2021).

Before 2018, the FBI did not collect data from federal law enforcement agencies, relying instead on data collected solely from state, local, and tribal governments.9 About Hate Crime Statistics, Fed. Bureau of Investigation, Hate Crime Statistics, 2018, 4 (2019), https://ucr.fbi.gov/hate-crime/2018/resource-pages/about-hate-crime.pdf. Additionally, in collaborative investigations involving multiple agencies, such as the state police, sheriff’s office, and local police department, the FBI’s “most local reporting rule” required incident reports to come from the “most local” jurisdiction.10Fed. Bureau of Investigation, Uniform Crime Reporting Handbook 9 (2004), https://​ucr.​fbi.gov/additional-ucr-publications/ucr_handbook.pdf. This meant that cases prosecuted under federal hate crime statutes were reported by nonfederal, and most often local, law enforcement agencies. Ironically, the same factors that might prompt a federal hate crime prosecution—insufficient training or resources within state or local justice systems,11Most states do not have laws requiring police training on identifying, reporting, and responding to hate crime, see State Hate Crimes Statutes, Brennan Ctr. for Just. (July 2, 2020), https://​www.brennancenter.org/our-work/research-reports/state-hate-crimes-statutes, and in response to data requests from reporters, some law enforcement officials have expressed misgivings about the quality of hate crime statistics due to inadequate training, resources, or reporting procedures, see Ken Schwencke, Why America Fails at Gathering Hate Crime Statistics, ProPublica (Dec. 4, 2017), https://www.propublica.org/article/why-america-fails-at-gathering-hate-crime-statistics; see also Maya Berry & Kai Wiggins, Opinion, FBI Hate Crime Stats Are Scary. So Is What’s Missing, CNN (Nov. 14, 2018), https://www.cnn.com/2018/11/14/opinions/​fbi-hate-crimes-data-whats-missing-berry-wiggins/index.html (noting that, in 2017, Alabama, Arkansas, Georgia, and Mississippi, states with “dark histories of racial and ethnic violence [that] have not enacted sufficient legislation to address hate crime,” each had fewer than ten agencies report hate crimes to the FBI). the lack of an applicable hate crime statute or a restrictive state-law definition of hate crime,12Few states have enacted hate crime statutes covering the range of conduct, regarding both the underlying offense (e.g., assault, homicide, vandalism) and the bias motivation (e.g., race, sexual orientation, disability), deemed reportable under the HCSA. See Brennan Ctr. for Just., supra note 11. Furthermore, among the states with laws requiring police reporting of hate crimes, most limit the required reporting categories to the bias motivations enumerated in the corresponding hate crime statute. See Id. or the deprioritization of hate crime enforcement within a certain jurisdiction13See James J. Nolan & Yoshio Akiyama, An Analysis of Factors that Affect Law Enforcement Participation in Hate Crime Reporting, 15 J. Contemp. Crim. Just. 111, 123–24 (1999) (identifying multiple agency-level factors, including organizational attitudes and concerns about distributing resources to priority areas, that affect law enforcement participation in hate crime reporting).—might also decrease the chance that law enforcement will report an incident as a hate crime.

In the last decade, several cases prosecuted under federal hate crime statutes have not been reported in hate crime statistics. In 2011, police in Jackson, Mississippi, did not report the racially motivated murder of a Black man named James Craig Anderson,14See Table 13: Mississippi: Hate Crime Incidents per Bias Motivation and Quarter, by State and Agency, 2011, Fed. Bureau of Investigation, Hate Crime Statistics, 2011 (2012), https://​ucr.​fbi.gov/hate-crime/2011/tables/table-13-1/table_13_mississippi_hate_-crime_incidents_​per_​bias_​motivation_and_quarter_by_state_and_agency_2011.xls (showing the Gulfport Police Department was the only law enforcement agency in the state to report a hate crime in 2011). which was the first homicide prosecuted under the Matthew Shepard and James Byrd, Jr., Hate Crimes Prevention Act.15Scott Bronstein, Three Men Plead Guilty to Federal Hate Crimes in Mississippi Killing, CNN (Mar. 12, 2012), https://www.cnn.com/2012/03/22/justice/mississippi-hate-crime/index.html. And in 2017, the year before the FBI introduced federal reporting, state and local agencies failed to report two of the highest profile hate crimes committed that year, both of which were prosecuted under federal hate crime statutes.16Emanuella Grinberg, 4 Days, 5 Reports of Hate Crimes, and a Disturbing Trend Developing in America, CNN (Dec. 12, 2018), https://www.cnn.com/2018/12/11/us/hate-crimes/index.html. As a result, the murders of Srinivas Kuchibhotla in Olathe, Kansas, and Heather Heyer in Charlottesville, Virginia, went unreported in official hate crime data.17 Id. State and local law enforcement agencies have also misreported incidents charged under federal hate crime statutes. In 2019, the El Paso Police Department reported a fatal mass shooting at a local Walmart, in which the alleged perpetrator “targeted Mexican shoppers,” as an “Anti-Other Race/Ethnicity/Ancestry” hate crime and not an “Anti-Hispanic or Latino” hate crime. See Hannah Allam, FBI Report: Bias-Motivated Killings at Record High Amid Nationwide Rise in Hate Crime, NPR (Nov. 16, 2020), https://www.npr.org/2020/11/16/935439777/fbi-report-bias-motivated-killings-at-record-high-amid-nationwide-rise-in-hate-c; see also Erin Coulehan et al., Federal Hate Crime Charges Filed in El Paso Shooting that Targeted Latinos, N.Y. Times (Feb. 6, 2020), https://​www.nytimes.com/2020/02/06/us/politics/el-paso-shooting-federal-hate-crimes.​html.

Federal reporting can preempt such embarrassing mistakes. If there is a federal hate crime prosecution, or even a federal hate crime investigation, then perhaps federal law enforcement, and not state, local, or tribal agencies with potentially less training, fewer resources, and different priorities, should report the incident to the FBI. On balance, this development will likely improve the accuracy and protect the credibility of hate crime statistics.

The problem, however, is that after three years of federal hate crime reporting, the FBI has published zero guidance on how it works. This leaves the public uninformed about the factors that determine whether an incident is to be reported by federal, state, local, or tribal law enforcement as a hate crime to the FBI. In some cases, to be sure, we can assume the reporting determination is straightforward. On one end of the spectrum, cases unlikely to merit federal involvement, such as vandalism and simple assault, are probably less likely to be reported by federal law enforcement. And on the other end of the spectrum, cases likely to result in a federal hate crime prosecution, for example, acts of mass violence that attract significant media attention, are probably more likely to be reported by federal law enforcement.

To some extent, this bears out. In 2020, federal agencies accounted for less than two percent of the vandalisms and less than one percent of the simple assaults reported in hate crime statistics, but most of the homicides.18 See Table 11: Offenses: Offense Type by Participating State and Federal, 2020, Fed. Bureau of Investigation, Hate Crime Statistics, 2020 (2021), https://​crime-data-explorer.​fr.​cloud.​​gov/​pages/downloads (download 2020 “Hate Crime Statistics Annual Reports,” and open Table 11). The federally reported homicides include the February 23 shooting death of Ahmaud Arbery,19[2] See Additional Datasets, Hate Crime, Fed. Bureau of Investigation Crime Data Explorer, https://crime-data-explorer.fr.cloud.gov/pages/downloads (last visited Sept. 14, 2021) (download “Hate Crime Statistics dataset,” filter column B for “2020,” filter column V for “Murder and Nonnegligent Manslaughter”). which is being prosecuted under federal hate crime statutes,20Press Release, U.S. Dep’t of Just., Three Georgia Men Charged with Federal Hate Crimes and Attempted Kidnapping in Connection with the Death of Ahmaud Arbery (Apr. 28, 2021), https://​www.justice.gov/opa/pr/three-georgia-men-charged-federal-hate-crimes-and-attempted-kidnapping-connection-death. and seven anti-transgender murders,21See Fed. Bureau of Investigation Crime Data Explorer, supra note 19 (download “Hate Crime Statistics dataset,” filter column B for “2020,” filter column V for “Murder and Nonnegligent Manslaughter”). at least three of which can be linked to press releases and news articles announcing federal hate crime charges.22See Press Release, U.S. Dep’t of Just., Two Individuals Charged with Carjacking, Murder, Firearms Offenses, and Destruction of Property (May 13, 2020), https://www.justice.gov/usao-pr/​pr/​two-individuals-charged-carjacking-murder-firearms-offenses-and-destruction-property; Press Release, U.S. Dep’t of Just., Puerto Rico Men Charged with Hate Crimes for Shooting Transgender Woman with a Paintball Gun (Aug. 6, 2021), https://www.justice.gov/opa/pr/puerto-rico-men-charged-hate-crimes-shooting-transgender-woman-paintball-gun.

But the reporting determination does not appear to be solely a question of whether the incident was prosecuted under federal hate crime statutes. For one, federal law enforcement reported over two hundred hate crimes in 2020 alone,23See Table 12: Agency Hate Crime Reporting by State and Federal, 2020, Fed. Bureau of Investigation, Hate Crime Statistics, 2020 (2021), https://​crime-data-explorer.​fr.​cloud.​​gov/​pages/downloads (download 2020 “Hate Crime Statistics Annual Reports,” and open Table 12). which exceeds the average number of federal hate crime prosecutions filed annually in recent years.24The reported total also exceeds the average number of annual referrals in the five most recent years for which data are available. See Few Federal Hate Crime Referrals Result in Prosecution, Transactional Recs. Access Clearinghouse (Aug. 12, 2019), https://trac.syr.edu/​tracreports/​crim/569. Moreover, since 2018, some homicides prosecuted under federal hate crime statutes have been reported by local law enforcement,25Local law enforcement reported the 2019 El Paso shooting despite a federal hate crime prosecution. See Press Release, Arab Am. Inst., AAI Statement on FBI’s 2019 Hate Crime Data Release (Nov. 20, 2020), https://www.aaiusa.org/library/aai-statement-on-fbis-2019-hate-crime-data-release; Coulehan et al., supra note 17. while others prosecuted under state hate crime statutes have been reported by federal law enforcement (despite seemingly no federal involvement in the investigations).26State prosecutors have filed hate crime enhancements against the man accused of killing Blaze Bernstein because of his sexual orientation in Orange County, California. See Hannah Fry, Blaze Bernstein Slaying Suspect Faces Hate Crime Sentencing Enhancement, L.A. Times (Aug. 2, 2018), https://www.latimes.com/local/lanow/la-me-blaze-bernstein-20180802-story.html. Bernstein was last seen alive on January 2, 2018, id., and although news reports make no mention of federal officials assisting with the investigation, federal law enforcement—not state or local agencies—reported the incident to the FBI, see Fed. Bureau of Investigation Crime Data Explorer, supra note 19 (download “Hate Crime Statistics dataset,” filter column B for “2018,” filter column V for “Murder and Nonnegligent Manslaughter”).

Without clarification from the FBI, these apparent inconsistencies might cast the impression that federal hate crime reporting is nothing more than a backstop designed to protect the FBI from bad press. Of course, the explanation could be simple. Perhaps there is a bright line rule that escapes public view, or reporting decisions are made case by case, or it varies state to state. But the questions remain. And while some might see little significance in one jurisdiction reporting a hate crime over another, there are policymaking implications at all levels of government.

One obvious implication is the fate of the “most local” reporting rule, while others involve questions of state law and police administration. For example, if reporting decisions depend on the circumstances of the case or the state in which the case occurred, then we might expect to see more federal reporting in states that do not prioritize hate crime enforcement. This would be a welcome development in some respects. In jurisdictions without reliable hate crime statistics, federal law enforcement could fill the gaps. But to the extent that federal reporting lets nonfederal jurisdictions off the hook, efforts to reform and improve hate crime enforcement within those jurisdictions could drift further out of reach.

II. Interactive Statistics

The HCSA does not require the FBI to collect hate crime data specifically through the UCR program.27In fact, the Act does not even mention the FBI. See 34 U.S.C. § 41305. But that is what Congress expected, if not intended.28Earlier versions of the HCSA specifically required reporting through the UCR program, see Hate Crime Statistics Act, H.R. 1171, 99th Cong. (1985), and the congressional sponsors of later versions said that “utilizing the [UCR] system ma[de] the most sense,” as participating agencies were “certainly capable of indicating whether an offense was motivated by prejudice,” Hate Crime Statistics Act of 1988: Hearing on S. 702, S. 797, and S. 2000 Before the Subcomm. on the Const., S. Comm. on the Judiciary, 100th Cong. (1988) (statement of Rep. John Conyers, Jr.). When Congress passed the HCSA in 1990, the UCR had been around for decades, and most U.S. law enforcement agencies were already submitting data for the FBI’s annual Crime in the United States reports.29See Fed. Bureau of Investigation, Crime in the United States 1985, 1 (July 27, 1986), https:​​//​​www.ojp.gov/pdffiles1/Digitization/102130NCJRS.pdf. Because “hate crimes are not separate, distinct crimes,” but rather “traditional offenses” with an underlying bias motivation, the FBI determined that it could fulfill the requirements of the HCSA without burdening law enforcement by adding bias motivation to the information requested in UCR submissions.30See Fed. Bureau of Investigation, supra note 9, at 1.

That system remains in effect. Agencies that report crimes to the UCR program indicate whether those crimes were bias motivated, the FBI collects the information, and then publishes it in hate crime statistics.31The National Incident-Based Reporting System (NIBRS) lists bias motivation as a mandatory data element in all submissions, but agencies using the defunct Summary Reporting System (SRS) submitted hate crime data separately. See Unif. Crime Reporting Program, Fed. Bureau of Investigation, Hate Crime Data Collection Guidelines and Training Manual 24 (Feb. 27, 2015), https://www.fbi.gov/file-repository/​ucr/ucr-hate-crime-data-collection-guidelines​-training-manual-02272015.pdf/view. Simple enough. But as mentioned in the preceding section, UCR statistics do not provide the most accurate measure of hate crime in U.S. jurisdictions.

Three limitations are obvious. First, most victims do not report hate crimes to law enforcement, and the UCR program is a data collection effort comprised of law enforcement agencies.32See Grace Kena & Alexandra Thompson, Bureau of Just. Stat., U.S. Dep’t of Just. Hate Crime Victimization, 2005-2019, 2 (Sept. 2021), https://bjs.ojp.gov/sites/​g/files/​xyckuh236/​​​files/​media/document/hcv0519_1.pdf (showing that forty-four percent of hate crime victimizations from 2010 to 2019 were reported to police). Second, determining bias motivation can be difficult, and reporting officers might lack adequate training or resources to make those determinations.33See Brennan Ctr. for Just., supra note 11; Schwencke, supra note 11. Third, the UCR relies on standardized offense definitions,34Nathan James & Logan Rishard Council, Cong. Rsch. Serv., RL 34309, How Crime in the United States Is Measured 3 (Jan. 3, 2008), https://sgp.fas.org/crs/misc/RL34309.pdf. and while the FBI has developed reporting guidelines,35See Unif. Crime Reporting Program, supra note 31. variations in the definition of “hate crime” from one jurisdiction to the next still cause confusion.

A fourth limitation is time. The FBI depends on dozens of federal agencies, hundreds of state, tribal, and territorial governments, and thousands of local jurisdictions to produce national hate crime statistics. And all that data must undergo review. The FBI has traditionally required agencies to submit annual data by March of the following year for publication in November.36See, e.g.,Fed. Bureau of Investigation, supra note 2, at 3, 8. Inevitably, some miss the deadline. In 2012, for example, more than eight percent of agencies participating in the UCR program did not submit their data on time for publication in the Hate Crime Statistics report.37Author’s own calculations. See About Hate Crime Addendum 2012, Fed. Bureau of Investigation, Hate Crime Addendum 2012 (2014), https://ucr.fbi.gov/hate-crime/2012-addendum/about-hate-crime-addendum-2012_​final.​pdf; Press Release, Fed. Bureau of Investigation, FBI Releases 2012 Crime Statistics (Sept. 16, 2013), https://archives.fbi.gov/archives/news/pressrel/press-releases/fbi-releases-2012-crime-statistics (noting that “18,290 city, county, state, university and college, tribal, and federal agencies participated in the UCR program in 2012”). And in 2020, Ohio reported only thirty-four hate crimes in time for publication, despite recording almost six hundred.38See David Nakamura, Ohio Submits Updated Hate-Crime Figures to FBI that Would Make 2020 U.S. Tally Highest Since 2001, Wash. Post (Sept. 10, 2021), https://www.washingtonpost.​com/​national-security/ohio-hate-crime-data-us/2021/09/10/8f18fb28-1261-11ec-882f-2dd15a067dc4​_​story.​html.

These undercounts diminish the accuracy and credibility of hate crime statistics. And although agencies have been able to forward overdue incidents for storage on the FBI’s “master data files,” you cannot amend a news cycle that has already passed.39Fed. Bureau of Investigation, supra note 2, at 5. As mentioned in the preceding section, the 2017 car attack that killed Heather Heyer and wounded thirty-six others in Charlottesville, Virginia, was not reported in hate crime statistics. The Charlottesville Police Department has since “amended a report . . . to officially add [the] case to the federal data,” Rachel DePompa, Charlottesville Police Department Adds Car Attack to Federal Hate Crime Data, WWBT (July 25, 2019), https://​www.​nbc12.​com/2019/07/25/charlottesville-police-department-adds-car-attack-federal-hate-crime-data, and the incident is now in the FBI’s master data file, see Fed. Bureau of Investigation Crime Data Explorer, supra note 18 (download “Hate Crime Statistics dataset,” filter column B for “2017,” filter column V for “Murder and Nonnegligent Manslaughter”). But the damage was done. See, e.g., Mallory Simon & Sara Sidner, Heather Heyer’s Not on this FBI List. How Hate Crimes Become Invisible, CNN (Aug. 12, 2019), https://www.cnn.com/2019/08/12/​us/​charlottesville-heather-heyer-death-not-in-fbi-report-soh/index.html. An obvious solution to this issue would be for the FBI to extend the publication deadline. But as the criminologist Jeff Asher has noted, the FBI’s hate crime statistics are already “impossibly delayed.”40Jeff Asher, Why There’s Not Much Data on Anti-Asian Violence, Lawfare (Mar. 23, 2021), https://www.lawfareblog.com/why-theres-not-much-data-anti-asian-violence.This leaves the FBI in a difficult position. It must give reporting agencies enough time to submit their data, but not too much as to render annual hate crime statistics effectively useless. The answer, it appears, is the 2018 introduction of the Crime Data Explorer (CDE).

As “part of the FBI’s broader effort to modernize” the UCR program, the CDE “present[s] crime data in a more immediate venue that reflects the constant change in the nation’s crime circumstance.”41About, Fed. Bureau of Investigation Crime Data Explorer, https://​crime-data-explorer.​fr.​cloud.gov/pages/about (last visited Sept. 14, 2021). Whereas the FBI formerly published hate crime data once per year in static, annual reports, statistics are now available to the public in the form of “downloadable files, interactive charts and tables, and an API (Application Programming Interface),” which can be updated regularly.42Id.

The move from static to interactive data not only enables the FBI to correct reporting mistakes and add late submissions, but also dampens their effect. Crime statistics are best understood as frames in a time lapse: not a definitive record of what occurred in a jurisdiction, but snapshots of the data federal statisticians had collected from that jurisdiction at particular moments in time. What differentiates static and interactive data, in part, is the number of frames the public gets to see. Traditionally, if a hate crime was not reported to the UCR program before the annual publication deadline, then, for the purposes of hate crime statistics, it never happened. But the more snapshots the FBI provides of its data, the less deadlines matter, and the closer we get to knowing what happened, where, and when.

Despite the promise of interactive statistics, the CDE appears to be a work in progress. For one, it is not clear how frequently the FBI will update statistics on the CDE. Less than two weeks after the FBI published 2020 hate crime statistics, the Ohio Department of Public Safety announced that it had sent revised numbers to the FBI that would increase the nationwide total by more than seven percent.43Nakamura, supra note 38. A month later, the FBI still had not incorporated the revisions. Additionally, different parts of the CDE provide different figures. On one section of the website, the FBI has published 2020 hate crime statistics in the form of downloadable files and interactive data visualizations, while in another, it has uploaded 2020 statistics onto an existing “hate crime statistics dataset.”44Compare Hate Crime, Fed. Bureau of Investigation Crime Data Explorer, https://crime-data-​explorer.​fr.cloud.gov/pages/explorer/crime/hate-crime (last visited Sept. 14, 2021), with Additional Datasets, Hate Crime, Fed. Bureau of Investigation Crime Data Explorer, supra note 19. According to the former, U.S. law enforcement agencies reported 7,759 hate crimes in 2020.45 Hate Crime, Fed. Bureau of Investigation Crime Data Explorer, supra note 44. According to the latter, 9,795 hate crimes were reported.46See Additional Datasets, Hate Crime, Fed. Bureau of Investigation Crime Data Explorer, supra note 19 (download “Hate Crime Statistics dataset,” filter column B for “2020”). Those 2,036 incidents are the difference between 2020 being the highest total in twelve years and the highest total on record.47See Press Release, Arab Am. Inst., AAI Statement on the FBI’s 2020 Hate Crime Data Release (Aug. 31, 2021), https://www.aaiusa.org/library/aai-statement-on-the-fbis-2020-hate-crime-data-release; Nakamura, supra note 38 (noting that the 9,730 hate crimes reporting in 2021 is the highest total on record).

 As with the federal reporting issues discussed above, the explanation could be simple. Perhaps the additional hate crimes recorded in the dataset did not meet the FBI’s publication standards, and we should therefore view the dataset with suspicion. Or perhaps the dataset contains late submissions and, therefore, provides a more accurate count. Both explanations are plausible. But the discrepancies give pause.

III. NIBRS Transition

In January 2021, the FBI retired the Summary Reporting System (SRS), which was the older of two programs available to law enforcement for reporting crimes to the FBI.48See National Incident-Based Reporting System (NIBRS), Fed. Bureau of Investigation, https:​​//​​​www.fbi.gov/services/cjis/ucr/nibrs (last visited Sept. 14, 2021). The remaining program, the National Incident-Based Reporting System (NIBRS), captures more comprehensive and detailed information than its predecessor,49See Barack Obama, Commentary, The President’s Role in Advancing Criminal Justice Reform, 130 Harv. L. Rev. 811, 862-63 (2017), https://​harvardlawreview.org/wp-content/​uploads/​2017/​01/​811-866-Online-Rev-vf.pdf; see also Emily J. Hanson, Cong. Rsch. Serv., R46668, The National Incident-Based Reporting System (NIBRS): Benefits and Issues 3–5 (Feb. 1, 2021), https://​crsreports.​congress.gov/product/pdf/R/R46668/2. and requires agencies to indicate whether a reported crime was bias motivated.50See Unif. Crime Reporting Program, supra note 32, at 24. Although the NIBRS transition will improve the accuracy and reliability of hate crime statistics over time, a significant percentage of U.S. law enforcement agencies still have not made the switch.

As of October 2021, thirty-seven percent of state, local, and tribal agencies had not achieved NIBRS compliance,51Explore by Location and Dataset, Fed. Bureau of Investigation Crime Data Explorer, https://crime-data-explorer.f​r.cloud.​gov/​pages/home (last visited Oct. 9, 2021) (showing sixty-three percent of participating agencies contribute NIBRS data). including the entire states of California, Florida, New Jersey, and New York, which accounted for thirty percent of the hate crimes reported nationwide in 2020.52See Fed. Bureau of Investigation, supra note 24. For the time being, these agencies have no means of reporting hate crimes for compilation in national statistics.53Agencies that have not transitioned to NIBRS also face threats to federal funding. See Bureau of Just. Assistance, U.S. Dep’t of Just., Edward Byrne Memorial Justice Assistance Grant (JAG) Program Frequently Asked Questions (FAQs) 4 (May 2021), https://bja.​ojp.​gov/​sites/g/files/xyckuh186/files/media/document/jag-faqs.pdf.

According to the FBI, federal statisticians will provide estimates for jurisdictions that do not submit NIBRS data,54Five Things to Know About NIBRS, Fed. Bureau of Investigation (Nov. 25, 2020), https://​www.​fbi.gov/news/stories/five-things-to-know-about-nibrs-112520. and we should assume this applies to hate crimes as well. But while the FBI often provides estimates for agencies that either do not report or submit incomplete data for crime statistics, it has never done so for hate crime statistics.55Asher, supra note 40.

Two questions arise from this point. First, why hasn’t the FBI provided estimates for hate crime statistics? And second, if the FBI provides hate crime estimates to offset underreporting during the NIBRS transition, why should it stop?

The answer to the first question might rest on the particularities of hate crime reporting, which is less common, more specific, less standardized, and more inclusive than traditional crime reporting. First, law enforcement agencies report far more criminal offenses than hate crimes per year, which means it will be more difficult to produce statistically reliable estimates. In 2019, for example, municipal police departments in California reported over one thousand times more criminal offenses than individual hate crimes,56Compare Table 8: California: Offenses Known to Law Enforcement by City, 2019, Fed. Bureau of Investigation, Crime in the United States, 2019 (2020), https://ucr.fbi.gov/crime-in-the-u.s/​2019/crime-in-the-u.s.-2019/tables/table-8/table-8-state-cuts/california.xls (add totals for “Violent crime” and “Property crime”), with Table 13: California: Hate Crime Incidents per Bias Motivation and Quarter by Agency, 2019, Fed. Bureau of Investigation, Hate Crime Statistics, 2019 (2020), https://​ucr.fbi.gov/hate-crime/​2019/​tables/table-13-state-cuts/​california.​xls (add quarterly totals for “Cities”). and disparities were even greater in other states.57Compare Table 8: Mississippi: Offenses Known to Law Enforcement by City, 2019, Fed. Bureau of Investigation, Crime in the United States, 2019(2020), https://​ucr.fbi.​gov/crime-in-the-u.s/​2019/crime-in-the-u.s.-2019/tables/table-8/table-8-state-cuts/mississippi.xls (add totals for “Violent crime” and “Property crime”), with Table 13: California: Hate Crime Incidents per Bias Motivation and Quarter by Agency, 2019, Fed. Bureau of Investigation, Hate Crime Statistics, 2019 (2020), https://ucr.fbi.gov/hate-crime/2019/tables/table-13-state-cuts/mississippi.​xls (add quarterly totals for “Cities”).

Second, because hate crimes are categorized according to thirty-four distinct bias motivation types (e.g., Anti-Asian, Anti-Jewish, Anti-Transgender),58Unif. Crime Reporting Program, supra note 32, at 5. the dataset is not only smaller, but also more specific. This could make it even more difficult to produce reliable estimates. Third, as discussed in the preceding sections, the amount of historical data from which to produce estimates, the definition of hate crime, and the prioritization of hate crime enforcement vary significantly between jurisdictions.59Some states’ hate crime statutes do not cover all criminal offenses, see, e.g., N.M. Stat. Ann. § 331-18B-3 (West 2021) (excluding capital felonies), or bias motivations, see, e.g., Ohio Rev. Code Ann. § 2927.12 (West 2021) (excluding sexual orientation, gender, disability, and gender identity). Similar variations have prevented the FBI from providing information for certain offenses in traditional crime statistics, such as arson.60See Arson, Fed. Bureau of Investigation, Crime in the United States, 2017 (2018), https://​ucr.fbi.​gov/​crime-in-the-u.s/2017/crime-in-the-u.s.-2017/topic-pages/arson (noting that the FBI does not provide estimates for arson “because the degree of reporting arson offenses varies from agency to agency”).

Finally, hate crime statistics include information about offenses not covered in traditional crime statistics, such as the number of intimidation, vandalism, and destruction or damage of property offenses known to law enforcement, which often account for most hate crimes reported each year.61Compare Offense Definitions, Fed. Bureau of Investigation, Crime in the United States, 2019, 2–3 (2020), https://ucr.fbi.gov/crime-in-the-u.s/2019/crime-in-the-u.s.-2019/topic-pages/​offense​-​definitions.pdf (explaining only arrest data are collected for intimidation and vandalism), with Methodology, Fed. Bureau of Investigation, Hate Crime Statistics, 2019, 3–4 (2020), https://ucr.fbi.gov/hate-crime/2019/resource-pages/methodology.pdf; see also Table 4: Offenses: Offense Type by Bias Motivation, 2019, Fed. Bureau of Investigation, Hate Crime Statistics, 2019 (2020), https://ucr.fbi.gov/hate-crime/2019/topic-pages/tables/table-4.xls (showing that “Intimidation” and “Destruction/damage/vandalism” accounted for most hate crime offenses reported). Producing hate crime estimates would therefore require more than simply applying the FBI’s existing imputation techniques.62For an explanation of the FBI’s imputation techniques, see Nathan James & Logan Rishard Council, Cong. Rsch. Serv., RL34309, How Crime in the United States Is Measured 19-21 (Jan. 3, 2008), https://sgp.fas.org/crs/misc/RL34309.pdf.

These concerns might explain why the FBI has never provided hate crime estimates. Assuming estimates are in store for at least some jurisdictions during the remainder of the NIBRS transition, the FBI will need to address all four concerns. This leads into the second question. If the FBI is prepared to provide estimates for hate crime statistics, then why stop once the NIBRS transition is complete? After all, researchers have argued that estimates would lead to a better understanding of hate crimes in the United States,63See Asher, supra note 40. something for which there is a demonstrated need.

Conclusion

Federal reporting, interactive statistics, and the NIBRS transition each promise more accurate, reliable hate crime statistics in the long term. But all three raise questions that the FBI has not publicly addressed.

First, what are the factors that determine whether a hate crime is reported by federal law enforcement, and how might federal reporting influence efforts to improve the response to hate crime in state, local and tribal jurisdictions? Second, how frequently should we expect updates to the FBI’s interactive hate crime statistics, and what explains the discrepancies between datasets published on the CDE? And third, will the FBI provide hate crime estimates for jurisdictions that have not made the switch to NIBRS, and, if so, should the FBI incorporate estimates into hate crime statistics beyond the NIBRS transition?

The answer to each question could be simple. And getting an answer should not be difficult. In the last year, Congress and the Biden Administration have taken multiple steps to improve hate crime reporting and addressing these questions should keep the momentum going. To start, the FBI and Department of Justice could publish additional information about federal efforts to improve these more technical aspects of hate crime reporting. However, if the administration does not act of its own accord, then Congress should exercise its oversight authorities. Whether these congressional inquiries are done via letter or committee hearing is up to lawmakers. But the public needs answers.


Kai Wiggins, J.D. Candidate, Stanford Law School, 2023

Suggested Citation: Kai Wiggins, Growing Pains for Hate Crime Statistics, N.Y.U. J. Legis. & Pub. Pol’y Quorum (2021).