On September 27, 2021, the FBI released much-anticipated crime data on that most unusual year 2020. The statistics revealed a continued steady decline in major crimes overall—apart from one unfortunate outlier: homicides. Despite homicides being at historic lows, especially when compared to the 1980s and 1990s, the murder rate last year rose by 30 percent compared to the previous year. This rise has left journalists and analysts seeking explanations. Yet the notoriously volatile nature of short-term crime data renders such efforts futile. Ascribing a short-term fluctuation to any particular cause—even a global pandemic—is impossible.

While police and allies have attempted to use the data to tie “bail reform” and racial justice protests to this past year’s rise in murders, those claims are contradicted by the geography of the rise in homicides, which occurred across the country: in red and blue states, in jurisdictions that have seen some measured wins for criminal and civil justice and those that haven’t, in jurisdictions that saw protests against police violence, and those that haven’t—and all despite massive police budgets.

Let’s take a step back. My admittedly dry account above of the newsworthiness of the new FBI data and subsequent efforts to twist it is how the story could and should have been reported by journalists. No sensationalism. No speculation. At least some context and nuance. And what we can actually determine based on the data.

But if you were to read the coverage of the data’s release by news sources like The New York Times and NPR, you would now likely believe that the only news from the FBI data was that there was an unprecedented spike in homicides—and that this unprecedented spike, against all evidence to the contrary and the FBI data itself, could very well have been caused by bail reform and protests for racial justice following the police killing of George Floyd.

This kind of “justice” reporting is not just false; it is dangerous, widespread, and long-standing. I write this not to attack the Times or NPR or the reporters of these stories, nor to take away or distract from the very real and disturbing tragedy of every single one of these murders, but to call attention to an insidious and historically rooted contributor to the system of policing and prison in our country: a pro-police worldview deeply ingrained in journalism.

When I talk about journalism as one of the most pressing racial and social justice issues today, people—even close colleagues of mine—often look at me quizzically. But after serving for nearly a decade as a public defender, I know well that every cruel and irrational policy of the mass incarceration era—policies that I saw devastate predominately Black and brown people in Brooklyn criminal court every day—was propped up by harmful journalistic biases and practices just like the ones on display this week from some of the most prominent media outlets in our nation.

At a time when the impact, cost, and failure of our system of mass criminalization is more visible and widely acknowledged than ever before, we have the opportunity, finally, to end our over-reliance on policies that have failed to produce public health or safety despite our spending more on these “solutions” than any other society in the history of the world.

Today, we know, both from experience and overwhelming research, that releasing people from jail prior to trial reduces crime for years in the future—and saves tens of millions of dollars in each major city. We also know, again based on experience and also the most robust criminogenic analysis in history—a meta-analysis of 116 studies just released this month—that long sentences have zero effect on crime.

Yet journalism today continues to ignore these “criminological fact[s]” while instead following the familiar and dangerous patterns from the 1980s and ’90s that helped drive mass criminalization itself: overly simplistic stories with alarmist headlines and dehumanizing language that rely predominantly on police as sources, neglect nuance, provoke fear in the public, speculate about short term crime data—and posit police, prosecution, and prison as the solutions to crime. Such journalism distracts the public from other solutions like reducing poverty, investing in mental health and substance treatment, providing quality education and affordable housing, and leaning into violence prevention and restorative justice approaches that have been shown to prevent and reduce violence.

Recent events provide abundant examples of journalism shaping criminal legal policy. In New York, after bail reform was passed in 2019, police and prosecutors immediately started cherry-picking and weaponizing—and the media started publishing—sensational cases and short-term statistics that drove the untrue narrative that releasing people before trial fueled a rise in crime. That fearmongering worked. Bail reform was rolled back in 2020—mere weeks after it was enacted. We are already seeing the same scare tactics in attempts to recall forward-thinking prosecutors in California and in Illinois, where recently passed and overwhelmingly bipartisan pretrial justice reform legislation is also now at risk. Bad journalism has consequences.

Though problematic reporting by the Times and NPR has resulted in millions receiving harmful misinformation, it also offers a critical teaching moment for reporters and consumers of media. These news outlets’ missteps are concrete and avoidable. By identifying them, we may be able to provide a road map by which to “call in” other media outlets and journalists to ensure that these same mistakes are not replicated.

First, in both the Times and NPR coverage of the FBI data, the reporters focused on the sensational. Outlets prioritize clicks over nuance. Because many readers don’t look beyond the big print, these kinds of trade-offs can be consequential.

The Times headline read, “Murders Spiked in 2020 in Cities Across the United States.” NPR made a similar editorial choice, publishing as part of their report’s title for the day, “A Surge in U.S. Murder Rates.” These outlets were far from alone: The Washington Post, NBC News, The Hill, and The Guardian, among countless others, focused on the homicide spike not only in their headline but also in the reporting, while burying the relevant fact that there was also a decline in all other major crimes.

This same kind of trade-off between sensationalism and context shows up more routinely in the explicit story bias of outlets. We often read about outlier crimes—the individual who committed a crime while out on bail—yet never about the far more common reality of the tens of thousands who haven’t been caged pre-trial and thrive in their freedom because of bail reform. These choices shape the public’s intuitions about the criminal legal system and, consequently, drive policy.

Second, both the Times and NPR pieces are sourced exclusively from police or criminologists with pro-police bias and/or consulting contracts. The Times article, for example, features the perspectives of two chiefs of police, a former police department supervisor, and a consultant to police departments. The NPR podcast relies solely on the opinion of Richard Rosenfeld, a criminologist with a history of supporting the s0-called—and much criticized—“Ferguson effect” theory, drawing connections between protests against police brutality and an increase in crime and murder.

This routine source bias in criminal justice reporting must stop. It leads to the false conception that police are the only experts on crime and ignores the critical perspectives such as those of public defenders, social workers, and individuals and communities directly impacted by the criminal legal system. It also presents a singular, skewed interpretation of the news, imbuing coverage with a pro-police bias—an assumption that police solve crime and make the public safer.

Every time there is a “rise in homicides,” instead of journalists using the occasion to question the efficacy of policing, police are allowed to use their failures to demand more resources, more funding, more support.

Finally, media outlets need to resist the understandable, but ultimately futile, urge to find potential causes for shorter-term changes in crime data. Indeed, both the Times and NPR acknowledged the limitations of such attempts. The Times story noted, “There is no simple explanation for the steep rise,” yet then provided a platform for a police chief and a police department consultant to “explain” it. Meanwhile, the NPR reporter acknowledged, “It’s difficult to attribute causality because, often, local factors drive crime,” and even admitted, “All these [causes] could’ve factored in. Their exact connection is unknown”—but then asked a criminologist to speculate on possible drivers of the rise in homicides. Given the known links between media, fear, and policy, this common journalistic practice is irresponsible.

Media outlets, editors, and reporters need to improve their practices by turning to sources beyond just police and prosecutors, critically analyzing police sources when used, conveying genuine nuance in their reporting and headlines, and stopping the use of dehumanizing language. But it is also on us—as consumers of media—to hold reporters accountable and be critical readers. We must question any conclusions based on short-term fluctuations in the crime rate, pay attention to which sources that are being held up as credible, and recognize how deeply embedded pro-police bias is and has been in the current media establishment. We can hold news outlets to account by taking to social media, authoring letters to the editor, and sharing the glaring flaws in such news coverage with our colleagues and friends.

The recent momentum and support for systemic change to the criminal legal system is widespread and unlike anything in recent history. But it is also fragile—as recent news reports and elections show. If we are to have any hope at ending this costly, ineffective, and inhumane era of mass criminalization, we cannot overlook or underestimate the role of the media in perpetuating it.