An Investigation of War Crimes

Just a note that the topic of this essay requires some difficult-to-read narrative, and that may not be everyone’s cup of tea.

Over the last year, I worked with a team from the New Yorker as they created an online database of possible war crimes in Afghanistan and Iraq. The database and accompanying description can be found here and a podcast series about the endeavor can be found here. My role was to help the investigative team analyze a database that took years to assemble. The goal of the database, as lead author Parker Yesko reports, was to document “allegations of violence perpetrated by U.S. service members or deaths in U.S. custody that happened in Iraq and Afghanistan after September 11, 2001.”

The database makes it possible, for the first time, to see hundreds of allegations of war crimes—the kinds that stain a nation—in one place, along with the findings of investigations and the results of prosecutions. The picture that emerges is disheartening. The majority of allegations listed in the database were simply dismissed by investigators. Those which weren’t were usually dealt with later, by commanders, in a justice system that can be deferential to defendants and disbelieving of victims.

To say that war is a complicated business is an understatement of immense proportion. My interest in this topic began long ago, and unexpectedly. Right out of college, I worked for Nebraska Senator Bob Kerrey on his 1992 presidential campaign. In those days, campaigns were brief and intense, albeit less furious than today. Kerrey announced his candidacy on the last day of September 1991, with the Iowa caucus and New Hampshire primary barely more than four months away.

When I think of that time now, it is always the same image that comes to mind, a moment from a Veterans Day campaign event on a deeply grey November day. Early that morning we picked up Bob in downtown DC and drove around two beltways to the Garrison Forest Veterans Cemetery north of Baltimore, almost to Reisterstown. There were four of us in a small sedan and I remember nothing of our conversation, only that Bob treated us to McDonald’s on the way home.

Before the event, where the Senator offered some remarks, and before the press arrived, Bob drifted into the cemetery to pay his respects, alone in the gloom. We were early and almost no one else was there, this was a man alone with his reflections. It was a profound moment for me, because at that time, and I suppose until today, Bob Kerrey was the only person I ever met who was genuinely larger than life. Kerrey was many things in 1991—senator, presidential candidate, business executive, celebrity (he was dating Debra Winger)—but above all else, he was a genuine war hero, a winner of the Medal of Honor (you can read his citation here).

His walk into the gloom was no small thing, his walking at all was no small thing, having left part of his leg and his foot behind in the war. When the campaign staff previewed his first television commercial, Bob walking across a New Hampshire ice rink announcing his tariff plan, we cheered. We didn’t cheer for the tariffs—we cheered for an amputee walking unassisted across the ice. We cheered for that ever-present courage.

Lt. (j.g.) Kerrey's courageous and inspiring leadership, valiant fighting spirit, and tenacious devotion to duty in the face of almost overwhelming opposition sustain and enhance the finest traditions of the U.S. Naval Service.

Years later, another story from Bob’s time in Vietnam emerged. During a swift boat raid to a remote village, Thanh Phong, in a highly volatile region, Kerrey’s SEAL team left 21 civilians dead, including 13 unarmed women and children. The killing was a slaughter. Some members of the SEAL team recollect only a firefight with the villagers caught in the middle. But according to other reports, the killings were methodical, including three of the children killed individually by soldiers with knives. The facts are in dispute about how the tragedy unfolded, but the SEAL team’s responsibility for the dead is not. The New York Times magazine published a full account of the massacre and the story is haunting.  

Bob Kerrey wrote later that he learned that “the greatest danger of war is not losing your life but the taking of others', and that human savagery is a very slippery slope.”

And, to the point of the New Yorker data collection, Bob Kerrey said this to ROTC cadets at the Virginia Military Institute:

It was not a military victory; it was a tragedy, and I had ordered it. How, I have anguished ever since, could I have made such a mistake? Though it could be justified militarily, I could never make my own peace with what happened that night.

“Though it could be justified militarily.”

Though killing 13 women and children could be justified militarily.

Human savagery is indeed a very slippery slope. That is what led to my interest in working on the war crimes story and to help as I could with the data harmonization and analysis. With this history on my mind, my own questions were humble—how does military justice assemble the facts and render justice in the midst of chaos?

What we found, from my perspective, is little appetite by the military to render judgment. You are welcome to peruse the data yourself and arrive at your own conclusions. But the military system seems primarily to be designed to avoid all but the unavoidable. There is justice when there must be—and there is less justice when there should be.

Consider this. When you are in the United States and believe you have been the victim of a crime, you call the police and report the crime. That event is now officially a ‘crime reported to the police’ and the investigation begins. Maybe there is evidence to support your belief that you have been victimized, and maybe there is not. Maybe a suspect is identified and maybe that suspect is arrested. Or not. But regardless of what happens after the call for service, in all but the most unique circumstances, your call for service counts as a crime reported to the police. Crimes reported to police are not a nuanced idea.

This is not how military justice works. In the military justice system, there is no analog to a call for service, no analog for a reported crime. For there to have been a crime—for a crime to be recorded—it is necessary to determine who the suspect is and that the suspect is a member of the military, because until that can be established no military crime can be recorded. So once a crime is committed, for the military to record that crime, the perpetrator must be known.

You see the problem of course. Military crime statistics are not analogous to civilian crime statistics. Civilian crime statistics count victimizations; military crime statistics count perpetrators.

In domestic US criminal matters, the best analog then is not crimes reported to police, but rather the rate at which cases are ‘cleared’. Cleared is a little bit of an ambiguous term—it does not mean justice has been served, only that we are pretty sure we know who did it. A case can be cleared in two ways, generally: it can be cleared by arrest. Or, it can be cleared exceptionally, meaning that the suspect is known but for whatever reason cannot be arrested.

But more often than not, the authorities—whether military or civilian—never learn who committed the crime. Even in homicides, which have high clearance rates relative to assault or robbery, only about half of cases are cleared. That means that the number of crimes reported to the military is far less than the number of victimizations.

But that is not the only bar that must cleared, so to speak, before a case comes before the military justice system. It must also be determined that the event was not justified militarily. Making that determination is a profound challenge. So, to sum it up, a perpetrator must be known and the confusion of firefights and checkpoints and detainees must be sorted out, and thus only a small portion of what would be documented as a crime in the US criminal justice system is even labeled as a crime in the military justice system.

Now, it would be reasonable to expect that given the high bar for a crime to even be determined to be a crime—the identity of the suspect is known, the circumstances are clear enough that a determination can be made that the event was not militarily justified—that defendants would face stiff penalties. This is not the case. Here is Parker Yesko again,

Of the seven hundred and eighty-one cases we found, at least sixty-five percent had been dismissed by investigators who didn’t believe that a crime had even taken place…

In a hundred and fifty-one cases, however, investigators did find probable cause to believe that a crime had occurred, that the rules of engagement had been violated, or that a use of force hadn’t been justified…

We identified five hundred and seventy-two alleged perpetrators associated with these hundred and fifty-one criminal cases. Only a hundred and thirty of them were convicted.

The records show that they rarely received lengthy prison terms.

Again, I would invite you to investigate this yourself and draw your own conclusions. The database, the article, and the podcast combine as a terrific piece of scholarship by an outstanding group of serious investigative journalists. I was honored to spend time with them.

Crime in the United States

It seems to be the consensus of opinion that the FBI will release national crime statistics for 2023 next week and I will try to put pen to paper or fingers to keyboard or voice to Siri and sort through the data. Sometime later in September, we can expect the National Crime Victimization Survey (NCVS) data for 2023. What will be most interesting is comparing the results of the two data series. You may recall that in 2022, while the levels of reported crime from these two sources were similar, the trends were wildly divergent—the FBI showing crime inching down and the NCVS showing crime skyrocketing. This appears to have mainly been due to difficulties in the NCVS survey operations during COVID. Given the primacy of crime and justice issues in the ongoing elections, consistent reports could quiet matters while disparate conclusions could open a new front in the scuffle. My bet is that the two data series will be generally similar, but stay tuned…