Recent press accounts have the nation's big cities in the throes of a “new crime wave.”
Police chiefs and mayors have lobbied the federal government for assistance in combatting violent crime. In response, Attorney General Loretta Lynch held a “Summit on Violent Crime” in Washington on October 7th that brought together chiefs, mayors and federal officials to address the crime increase and discuss policy options.
I attended that meeting and had the chance to ask FBI Director James Comey a question that has concerned me throughout the news coverage of the crime increase: Why are we relying on press accounts instead of the FBI's Uniform Crime Reports (UCR) for crime statistics?
The UCR, after all, is the nation's official source of crime data from the police. If we were facing an influenza outbreak, we would be getting weekly case counts from the Centers for Disease Control. If the unemployment rate had jumped, we would turn to the Bureau of Labor Statistics for last month's job figures. So, why doesn't the FBI release timely statistics to track monthly crime rates so that policymakers and the public can determine whether the country is, in fact, in the midst of a significant crime increase?
In answer to my question, Director Comey said that, as far as he knew, the FBI could not release up-to-date crime figures because it doesn't have them. He also said he would check with the UCR program to see what might be possible in the future.
The Uniform Crime Reports
It is useful to recount how the UCR program compiles crime data from the approximately 18,000 law enforcement agencies in the United States. In most instances, the program no longer receives the data directly from local law enforcement agencies. Instead, as the Data Quality Guidelines for the FBI's reports make clear, most agencies report monthly crime and related UCR data to their state-level UCR program or comparable Statistical Analysis Center (SAC).
The UCR's state partners check the data for completeness and accuracy and then send it on to the FBI, where final checks are done and the data are readied for release in the annual report Crime in the United States.
At any given time, therefore, the FBI has on hand monthly crime information that has gone through an initial data quality evaluation by its state partners. State UCR programs and SACs vary in how rapidly they submit their data to the national UCR. Some do so monthly, others wait until after the collection year.
If many state programs do not currently submit timely crime data to the national UCR, nothing prevents the FBI from requesting more rapid reporting. Even if the national UCR program received timely crime data covering just five percent of the nation's law enforcement jurisdictions, we would have a picture of emerging crime problems in 900 areas across the country, rather than the few dozen cities that the most diligent news reporters have managed to survey—which has led to what one writer has characterized as “scare headlines.”
Editors Note: The following two sections are drawn in part from “The Big Picture: 2010 Presidential Address to the American Society of Criminology,” by Richard Rosenfeld, published in Criminology (v. 49, pp. 1-26).
Back to the Future
Crime in the United States is released annually about nine-to-ten months after the end of the collection year. The FBI published yearend data for 2014, for example, in late September of this year.
It has not always been this way. The FBI has actually taken several steps backward over time in promulgating the UCR data in the slender time increments needed for rapid policy evaluation and response.
When the UCR program began in 1930, the data were released on a monthly basis. A quarterly release schedule was adopted in 1932. True, there were fewer law enforcement agencies back then, but the data were entered in pen and ink, or on typewriters, and sent by the post office to the FBI.
As a cost-saving measure during World War II, the FBI switched to semi-annual reports, which lasted until 1958 with the publication of the first annual volume of Crime in the United States. So, here we are, well into the twenty-first century when most law enforcement agencies compile crime data in electronic form, looking back in envy to the good old days of 1932—when the nation's crime statistics were released on a schedule that made them useful for tracking crime problems as they arose.
It is difficult to believe that the FBI is incapable of reviving this 80-year-old best practice in an age of electronic data entry and instantaneous submission.
The Need for Comparative Crime Data
We also know that it is possible to disseminate the UCR data more rapidly than 21 months after the beginning of the collection year, because a large and growing number of local police departments already post monthly and, in some cases weekly, crime counts on their public websites. And, as noted, many state UCR programs obtain and in some cases publish crime statistics well before they are made available by the FBI.
But the move to real-time data availability at the local and state level prompts a question: If the more rapid release of UCR crime data is tenable, why is it necessary? After all, local police departments, especially the larger ones, are now capable of producing real-time data that can be used to address emerging crime problems.
Rapid data collection and retrieval are the digital foundation for the hot-spots and problem-oriented policing strategies now so much in vogue. If the St. Louis or San Francisco police department has adequate data for monitoring and responding to local crime problems, what use would they have for comparable data from Cleveland or Chicago?
The same question might be asked about the policy relevance of public-health or economic data. If we already know the local heart-disease or unemployment rate, for example, why do we care about the prevalence of heart disease or unemployment elsewhere? The answer of course is that the origins of heart disease and unemployment transcend local boundaries. The local conditions are manifestations of broader patterns and trends. The same is true of crime, a point made repeatedly by the local representatives at the Attorney General's crime summit.
If local crime rates represent variations on a common theme, then the first task of the policy maker is to place the local changes in the context of general trends. Knowing whether local crime rates correspond with or diverge from broader patterns is essential for developing effective local responses. If local conditions are part of a general pattern, then interventions proven effective elsewhere should be high on the list of candidates for local adoption.
If the local rates are diverging from common trends, it is a good bet they are being driven by idiosyncratic local conditions (e.g., a flare up in gang conflicts), which would call for interventions specifically tailored to the local scene. Either way, comparative crime data are needed to determine the nature of local crime problems and to respond appropriately.
What Should Be Done?
FBI staffers are currently working with the Bureau of Justice Statistics to upgrade national statistics on crime and police shootings. Director Comey urged the mayors and police chiefs at the October 7th crime summit to adopt the National Incident Based Reporting System (NIBRS) in their jurisdictions, which has been in the works since the 1980s but still covers only about a third of the US population. These are welcome steps, but they will not help national and local policymakers address current crime increases. We cannot respond knowledgably to this year's crime problems, determine their size and scope, or figure out what may be causing them with last year's crime statistics.
Reporters have done due diligence in informing us that crime rates are up in some cities. They have filled an information void that is almost unimaginable in any other policy arena well into the 21st century.
The FBI should close the crime information gap now, or hand off this urgent policy priority to another federal agency that will.
Richard Rosenfeld is the Founders Professor of Criminology and Criminal Justice at the University of Missouri – St. Louis. He welcomes your comments.