President Trump wants the federal government to start publishing weekly lists of crimes committed by undocumented immigrants. An executive order on “Enhancing Public Safety in the Interior of the United States” that Trump signed last month calls for the reports, which would also identify “sanctuary jurisdictions,” cities and counties where local law enforcement authorities don’t report immigration status violations to the federal government.
But from my own work building a national database of crimes by police officers, I’ve learned that collecting and distributing reliable stats in real time may be much harder than Trump thinks. Especially for an administration that seems to have little regard for facts, good data or scientific integrity.
Trump regularly claimed in last year’s campaign that crime was worse than it really is, and he has continued to do so as president. “The murder rate in our country is the highest it’s been in 47 years,” Trump said Tuesday in a meeting with county sheriffs at the White House. “ … I’d say that in a speech and everybody was surprised, because the press doesn’t tell it like it is. It wasn’t to their advantage to say that.” By the FBI’s count, the murder rate is actually near its lowest point of the past five decades. The same day, White House aides insisted that the media had ignored or underreported 78 terrorist attacks, some of which had been among the biggest news stories of the past few years. A few weeks earlier, Trump claimed that “in Philadelphia, the murder rate has been steady — I mean just terribly increasing.” That, too, was false.
Presidents from both parties have used scary rhetoric for political purposes to tell us about some “most important problem” facing the nation during the war on crime and the war on drugs. But Trump’s White House seems determined to create confusion and uncertainty about crime and public safety, insisting that there’s far more danger than the data really shows. With the weekly report of crimes by undocumented immigrants, for instance, Trump seems to be assuming that such crimes are frequent and mostly take place in sanctuary jurisdictions. In fact, research has consistently shown that immigration is often associated with lower crime rates in some of the most disadvantaged urban neighborhoods across the country. Besides, an arrested person’s immigration status is often not known at the time of arrest. It is not clear what trustworthy statistics are readily available that would provide the information required by the order, or even how it would be collected.
Collecting reliable crime data from state and local governments is complicated. I have studied police crime for the past 13 years. The government’s inability to collect data prompted me to set up my own database of police misconduct, and some of my research has been funded by the National Institute of Justice. I also worked with The Washington Post on its own investigation of police shootings. My database includes news reports of more than 10,600 state and local police officers who have been arrested since the beginning of 2005. The backbone of my efforts are 48 Google alerts that I created more than a decade ago, when I realized that the Justice Department did not collect and aggregate reliable nationwide figures on police misconduct. The federal government cannot tell us how many officers during that time period have been arrested on murder or manslaughter charges resulting from on-duty shootings. But I can: In the past 12 years, 79 on-duty police officers across the country have been arrested for shooting and killing someone. Only 27 of those officers have been convicted (14 by jury trials and 13 by guilty pleas). The process that starts with data collection in my research lab also involves tedious triangulation of sources over several years for each arrest case to verify the validity and reliability of my data before full analysis.
Federal crime data collection relies on voluntary self-reporting to the Justice Department by state and local law enforcement agencies. And Washington has long demonstrated an inability to compile it, even when the law mandates such collection. For example, the Violent Crime Control and Law Enforcement Act of 1994 required the attorney general to publish annual reports on the use of excessive force by police. Those efforts failed largely because compliance was voluntary, and many state and local law enforcement agencies did not participate. Within just a few years and with no consequences, the Justice Department abandoned any efforts to comply with the law. The type of data required by the statute on excessive force simply isn’t collected in any usable form by most of the more than 18,000 state and local law enforcement agencies across the United States. And where it is collected, agencies are often reluctant to report it because they don’t want the feds to see them as a problem.
That isn’t the only time the government failed to collect data that the law seemed to require. People convicted of certain crimes of domestic violence are prohibited by the federal Gun Control Act from owning or possessing firearms and ammunition. Yet there is no nationwide program to standardize the reporting of domestic violence misdemeanor convictions to the databases used to conduct gun background checks. Local court conviction records include many cases of misdemeanor assault, but it is often impossible to determine whether the cases involved domestic violence without reviewing court charging documents, prosecutor files and police reports on a case-by-case basis. The result, according to a Government Accountability Office study last year, has been that many people with domestic violence criminal records have escaped scrutiny in background checks and purchased firearms illegally.
Until 2015, when the FBI announced that it would begin its own database of police shootings, the federal government didn’t even know how many people were killed by police officers in any given month or year. FBI Director James B. Comey admitted then that that situation was “embarrassing and ridiculous.” Only 3 percent of state and local law enforcement agencies comply with efforts to collect data on people killed by police, he said. Federal law requires states to gather data and report to the Justice Department on all deaths that occur during interactions with law enforcement personnel. But many states routinely underreport the total, partly because some local agencies simply do not complete the paperwork; a Bureau of Justice Statistics study in December estimated that as many as 25 percent of agencies didn’t respond to surveys or provide full answers. About seven years ago, Maryland went so far as to develop alternative data collection techniques, searching local newspapers to find in-custody deaths that had not been reported by police. More recently, in 2015, the Bureau of Justice Statistics redesigned data collection efforts for its arrest-related deaths program to include news articles and media reports. Preliminary findings indicate what we already knew: Traditional data collection efforts from state and local law enforcement agencies result in gross underreporting of crime data.
The government already does such a bad job of collecting vital data that starting a new weekly report — involving statistics that no agency is likely to be keeping, to prove a political point — is the last thing it should be attempting. Focusing on the areas where past data collection efforts required by statute have failed should be the priority. All too often, criminal justice policy is driven by outliers, to the detriment of research and data collection policies that would provide far more substantive knowledge about illegal activities.
Trump’s order does get one thing right: It is important that Americans be properly informed about public safety threats. But he seems to be looking for a scapegoat more than a solution. “I want you to turn in the bad ones,” he told police chiefs Wednesday, speaking of undocumented immigrants, whom he called “the illegals.” “We’ll get them out of our country and bring them back where they came from, and we’ll do it fast. You have to call up the federal government, Homeland Security, because so much of the problems … are caused by gang members, many of whom are not even legally in our country.”
Available crime data does not support many of Trump’s claims, and his continual misuse of crime statistics is reckless and bizarre. In the end, it’s simply bad crime policy.