Nieman Foundation at Harvard
The classic crime beat, dating from at least the mid 19th century, is evolving. It has to evolve. There are good reasons to believe that the routines of “traditional” crime coverage produce a journalism that just isn’t as good as it needs to be. We need to try something new.
These changes start with the spot story: the routine, straightforward “this just happened” report. The police and the courts have always been the primary sources for such stories, but the authorities are increasingly publishing their reports and documents directly. Meanwhile, there’s a lot of evidence that existing coverage has failed to give an accurate picture of the state of crime and punishment in the public mind — nor do traditional story forms provide a good place to talk about crime, its aftermath, and possible solutions.
Fortunately, a number of people and organizations are pioneering new approaches to crime reporting. Their techniques are based on comprehensiveness, extensive use of data, new story forms, and — in the most drastic departure from the traditional routine — deciding not to visit the crime scene.
The police have always been the key source for the crime beat, a.k.a the cops beat. You can see this in much crime coverage: Take a look at this homicide spot story versus this police press release. (No disrespect to the story’s author — don’t hate the player, hate the game.) But in recent years, there has been a massive increase in police resources devoted to communicating with the public. In many places, officials talk to the public using social media and put crime data, press releases, and court documents online — a prime example of sources going direct.
The routines of professional journalism evolved long before the Internet, when one could assume that it was hard for people to find out about a crime without the journalist (and there was no way to link to online resources). If that’s no longer the case, shouldn’t these routines change? Put another way, what does a reporter bring to crime reporting, beyond simply telling people that something happened? This was my question on Twitter last week, which touched off an intense discussion among my journalist peers, captured here (and included at the bottom of this piece). It produced a number of good answers.
One of the participants in that discussion was former crime reporter (and current Boston Globe newsroom developer) Andy Boyle. We talked afterward about what can be gained by visiting the crime scene yourself. “I’ve been able to get information that wasn’t in the press release,” he said. “You can actually get a richer, more developed story.”
A journalist can discover things that the cops missed. Boyle recalls a specific incident where Tampa police initially labelled the shooting of Derek Anderson “drug-related,” but neighbors insisted that Anderson wasn’t involved with drugs. Boyle called in this tip and his editors incorporated it into the online story. While he was still at the crime scene, the detectives returned. “I could hear them talking to other people and they’re like, ‘Yeah, there’s a story that went up online, so they sent us back out here to talk to more people,'” said Boyle. “It appeared to them, initially, this was a drug-related crime, when it turns out this guy had beef with Derek Anderson on the basketball court, so he shot him.”
Beyond individual incidents, several people on Twitter pointed out that “trend” or interpretive stories are important. The statistics may be public, but do we really want to leave it to the police to speak frankly about longer term patterns in crime? There are also issues with the crime data itself, in how crimes are counted or reported. Our online discussion surfaced numerous examples of journalists discovering that official statistics were flawed or cooked, as in Los Angeles, Milwaukee, and New York. More generally, it’s the journalist’s job to hold the police accountable in light of their sweeping powers, as in this eloquent story about fighting for public acces to Baltimore police records.
But trend stories and police department accountability are not the bread and butter of crime reporting. Exceptional cases like Boyle’s aside, I’d argue that paying professional journalists to produce the classic “this just happened” crime spot story is a lot less valuable than it used to be. The question becomes even sharper when one considers it in terms of opportunity cost. Clearly, a reporter can produce a better crime story if they visit the crime scene themselves and write the story in their own words. But is it still worth assigning someone to do this, instead of any other story they could be reporting?
Moreover, even individually accurate stories can, over time, add up to the wrong impression.
Polling shows that most Americans believe that violent crime is increasing, and have believed this for several decades, yet nationwide violent crime rates have been decreasing steadily since the mid-1990s. This singular fact, more than any other, makes me believe there is something very wrong with the way we cover crime. Here are the charts, from Gallup:
Why is this? What makes a person believe that crime is going up or down? One answer is that — as experimental psychology research has consistently shown — people tend to reason from examples. When thinking about crime (or most other things), we naturally call to mind specific notable incidents. We don’t think about overall rates, or compare the statistics to last year or to other life risks. By this standard, journalism naturally produces a highly skewed report of crime because journalism reports on the most heinous crimes.
In Chicago in 1987, fewer than a third of 684 homicides were reported in the daily papers. In Los Angeles County from 1990 to 1994, it was 2782 out of 9442 (29 percent). In national papers in England and Wales 1993 to 1996, it was 1,066 of 2,685 (40 percent). Smaller places often have a higher coverage rate, such as 163 out of 223 (73 percent) in the Baltimore Sun in 2010. But even then, there is a severe skew: generally, these studies have found that a crime is more likely to be covered, or more likely to be covered prominently, if the victim is young, female, white, and/or rich, or if the killing is particularly gruesome or involves multiple victims or sex.
Moreover, homicides command a disproportionate amount of coverage. Something in the range of 70-90 percent of crimes are property crimes, but coverage is dominated by violent crime. White-collar crime is also drastically undercovered, especially on TV news. Coverage is also much more likely if the victim was a random stranger, even though in actuality the majority of victims know the perpetrator.
This is all rather dismal. One group of researchers concludes:
Collectively, the findings indicate that news reporting follows the law of opposites — the characteristics of crimes, criminals, and victims represented in the media are in most respects the polar opposite of the pattern suggested by official crime statistics.
To some extent, this is to be expected, because the most sensational crimes — the ones everybody wants to talk about — are by definition rare. As the adage has it, “Man Bites Dog” is news, “Dog Bites Man” isn’t. Unfortunately, because we tend to think in examples rather than statistics, the novelty criterion for news gives precisely the wrong impression in the case of crime reporting.
One simple possible answer to this selective misrepresentation problem is simply to report all crimes. That’s not possible in the traditional model because of time constraints (and, for print, space constraints.) But as NYU journalism professor Jay Rosen has suggested, trying to cover all of something is a great way to force innovation in news.
So, data: All police departments collect crime statistics, and many now put them online. Sometimes these are inconvenient PDFs, but increasingly police departments are producing machine-readable feeds that can power maps and other types of analyses. The standout example here is the Los Angeles Times’ crime map, a project spearheaded by newsroom developer Ben Welsh. The Times’ map displays crimes for your neighborhood, and even generates automated alerts for unusual spikes in a particular type of incident.
It wasn’t a trivial matter to produce this map, because we’re still in the early stages of open government data — different agencies use different systems and formats, and data quality varies widely. The Times had to work with both the Los Angeles Police Department and Los Angeles County Sheriff’s Department to obtain all of the necessary data in suitably formatted feeds. Even then, after the launch of the site, the Times discovered that the official data contained major omissions and errors (since fixed).
The result is a comprehensive view that individual stories cannot deliver. No one has to tell a user looking at these maps that property crime is much more common than more sensational violent crimes. With color-coded incidents, anyone can see this at a glance.
Crime data can be used for maps, personalized alerts, and trend analysis. It’s even possible to have a computer automatically write text stories from data, as Narrative Science does for sports and business. This may not make for great stories (yet?) but that’s not the point here. Algorithmic techniques “can take a little bit of the burden off, so you actually can dive deeper and do bigger stories, while still trying to do some of the bread and butter stuff that you’re doing,” said Boyle.
Homicide Watch D.C. also takes the “cover 100 percent” mantra, but relies as heavily on Internet-era story forms as it does on data. Just two (very busy) people cover every homicide in the metropolitan D.C. area “from crime to conviction.” (The site’s motto: “Mark every death. Remember every victim. Follow every case.”) There were 105 homicide incidents in D.C. in 2011.
Site editor Laura Amico was a “very traditional” police reporter before moving to D.C. and growing frustrated with existing crime coverage. She calls the Homicide Watch process “structured, data-driven beat reporting…it was very publicly available who homicides were happening to, where, when, and what the outcomes of those homicides were through the courts process.” But there was no place where all of this information was tied together. She asked: If a police reporter made all of the information that they had gathered in the course of their reporting available to the public, what would that look like?
In a big departure from her previous reporting work, neither Laura Amico nor husband Chris Amico ever visit a crime scene. Instead, they rely on official releases, social media traffic (often faster than official channels, and always carefully verified), remote interviews, documents, and data feeds. They’ve even been able to determine a victim’s name by looking at the search terms that led people to their site. All of this information is compiled into a single page for each victim and for each suspect.
Collecting such complete data allows all kinds of journalistically interesting analyses. Aside from their exhaustive year-in-review coverage of homicide and criminal justice trends (how they did it) the Amicos are able to hold the police department accountable with data. “When the chief of police says ‘I solve most murders in under a week,’ I can check it,” said Chris Amico. He can also check statements like “most homicides are drug-related” and challenge stereotypes such as the idea that most murders happen in Southeast DC (not true; Northeast had more in 2011). “This idea of having data that backs your beat is somewhat foreign to a lot of journalists, but at this point, I think it’s the only way to do it,” he said. The model seems effective and replicable, and Homicide Watch is looking to license their reporting platform. (Laura Amico will be coming to Harvard this fall as a Nieman-Berkman Fellow in Journalism Innovation.)
But more than data, these pages have become a community resource, with friends and relatives discussing cases and remembering victims long after the crime. We have a human need to talk about tragedy, a need that traditional crime reporting does not always fill.
I want to end by grasping for the limits of what journalism can do for society. In a fascinating piece of experimental research, Renita Coleman and Esther Thorson explore a “public health” model for crime reporting. Rather than throwing up our hands and saying that criminals are irresponsible or crazy, can we isolate and address societal “risk factors”?
By categorizing violence with other public health problems and applying the same scientific tools used to control other epidemics, public health proponents believe they can convince Americans that violence is predictable and potentially preventable. They see their task as no different from the one public health experts faced in the 1960s when they advised that adding safety features to cars, wearing seat belts, and not drinking and driving would reduce automobile deaths and injuries.
Until the 1960s, traffic accidents were blamed on “the nut behind the wheel.” Prevention strategies were limited to advising people to drive more safely. When researchers began identifying the role of societal and environmental risk factors in auto crashes, public health advocates took the findings to the media and sought to change the way these events were covered. The media began including the type of cars involved, road and weather conditions, and whether people were driving drunk or wearing seatbelts. Soon, perceptions of the causes of auto injuries and deaths changed, and more social policies were enacted to discourage drunk driving, build safer roads, and force car manufacturers to design safety features into cars. The rate of automobile deaths and injuries slowed.
Crime and violence are not so different, epidemiologists say. Some of the risk factors associated with high levels of many kinds of violence include poverty, racial segregation and discrimination, unemployment, alcohol, firearms, the portrayal of violence in the media, lack of education, child abuse, childhood exposure to violence, and the belief in male dominance.
Coleman and Thorson suggest a very specific change to crime coverage: add “contextual and statistical information about the type of crime portrayed in the story.” In their experiments, this change succeeded in getting people to “become more critical of society’s role in crime and violence” and less likely to adopt “fatalist” attitudes such as “there’s nothing that anyone could have done to prevent this.”
Shifts in attitude are not a decrease in violence, and journalism alone will not solve the problem of crime. But it might be able to get people to believe that something can be done — and, perhaps, in the model of solution journalism, point to promising directions.
