On January 17, 2017, journalism nonprofit ProPublica launched an initiative called Documenting Hate to crowdsource reliable data about hate crimes so that journalists—and the public—could be more informed about the prevalence of these crimes. The goal of Documenting Hate is to compile an authoritative database of hate crimes in the United States, aggregating data from user-submitted reports, civil rights groups, law enforcement agencies, and other sources, into a single, accessible location to make this data more readily available to all. On the surface, Documenting Hate should be heartening—civic technologists are employing data to better understand and tackle a serious social problem—however its existence raises a very pressing concern: why isn’t there already an authoritative source for this data?
When it comes to hate crimes, the main sources for official crime statistics are woefully lacking. There are a few ways federal criminal justice agencies measure crime. The Federal Bureau of Investigation (FBI) oversees the Uniform Crime Reporting (UCR) program, in which 18,000 participating law enforcement agencies around the country voluntarily submit data about crimes in their jurisdictions, which the FBI analyzes and publicly reports. Additionally, the Department of Justice administers the National Crime Victimization Survey (NCVS) biannually to measure incidents of criminal victimization, which includes both reported and unreported crimes. In 2012 (the most recent year for which data is available), NCVS reported 293,790 hate crime victimizations across the country, while UCR program reported just 7,164 victims. Given that NCVS also records unreported crimes in its surveys, some discrepancy would be understandable. However the Bureau of Justice Statistics (BJS) found that just over half of violent crimes go unreported per year, with some variation based on the type and seriousness of the crime, so a discrepancy of this magnitude is cause for concern.
This large discrepancy indicates that law enforcement agencies are dramatically underreporting the actual number of hate crimes that occur within their jurisdictions. There are two main reasons why the FBI’s official statistics are likely faulty. First, since the UCR program is voluntary, not every law enforcement agency shares data with the FBI—in 2014, 15,494 of the 17,985 law enforcement agencies in the country shared hate crime data. Second, many of the agencies that did report hate crimes likely reported data they knew to be false. Just 1,666 of those 15,494 agencies reported any incidence of hate crimes within their jurisdictions, the rest reported none. When examining the breakdown of hate crime reports by jurisdiction, it quickly becomes clear that these numbers are almost definitely not accurate. For example, despite Mississippi having 75 participating agencies covering 1.3 million people, there is only one reported hate crime in the state for all of 2014. The 63 participating agencies in Wyoming reported zero. Several other states, including Arkansas, Georgia, and Louisiana also reported unrealistically low numbers, establishing a clear and suspicious trend of misleading hate crime reporting. When other states, such as Massachusetts, report hundreds of hate crimes per year, it is simply implausible that some states report experienced a scant handful or even none at all.
Such bad statistics can have serious implications for criminal justice issues. For example, the opaque ways the federal government measures sexaul assault makes researching the issue unnecessarily challenging and dangerously warps public perception about the prevalence of the issue. Accurate hate crime statistics are crucial for law enforcement officials and policymakers to make well-informed decisions, and misleading data will inappropriately influence how these decision-makers allocate resources to address this problem.
Without action from local, state, and federal policymakers, official hate crime data will not get better. First, all law enforcement agencies that do not already do so should participate in the UCR program. While agencies should do this for its own sake, they should also recognize that UCR participation qualifies them for federal grant funding. Additionally, offices of state inspectors general in states with suspiciously low hate crime reporting rates should investigate the crime reporting practices of local law enforcement agencies to ensure hate crimes are accurately recorded, processed, and reported. For example, the Baltimore Police Department was able to dramatically underreport the number of sexual assault cases in its jurisdiction for years by deliberately misreporting the nature of these crimes and the status of their investigations in their reporting. On the federal level, though the FBI provides guidance on hate crime data collection, it should do more to ensure that law enforcement agencies are recording and reporting data about hate crimes accurately, such as by offering supplementary training or investigating agencies likely failing to report this data correctly.
ProPublica should be applauded for taking the initiative to address the glaring problem of faulty hate crime statistics. However, leaving this problem in the hands of civil society is neither sustainable nor responsible. Without continuous government funding to support the collection and maintenance of this data, there is no guarantee that it will be around forever, and the fact that there is need for an initiative like Documenting Hate indicates that the UCR program is not succeeding in its mission to gather reliable and accurate crime statistics.