Floyd was killed during an interaction with Minneapolis police on May 25 2020, which was 20 days before the post above was shared. No citation appeared alongside the Facebook post.
On Twitter, versions of the same claim received similar engagement and contained similar parameters, some but were posted on different dates. A June 16 2020 tweet had identical wording:
Police in the US have killed 120 people since George Floyd.
20 days. 120 lives.
— Uppity (@blackartivist) June 16, 2020
One from June 15 2020 was slightly longer, but involved the same parameters:
Since George Floyd died cops have killed 120 ppl in 20 days. Black people are being hung. Women and children are being abducted and killed. The list goes on. Fuck sports and entertainment man. It’s time to really tackle all this shit. Sports a distraction right now.
— Godfrey Durham Jr. (@Godfrey2_) June 15, 2020
On June 17 2020, Twitter users were sharing the same claim:
Although the “police in the U.S. have killed 120 people in the 20 days since George Floyd’s death” phrasing didn’t really vary, it did appear on several different dates in June 2020, and none of the versions with six-figure share or like counts seemed to include a citation for the claim.
Locating a Source for 120 Deaths in the 20 Days Following George Floyd’s Death
While the claims were not accompanied by any specific citation in their formats above, comments often referred to a site — FatalEncounters.org. An “About” page on that site was empty, but a dropdown “About” menu included a page titled “Why [Fatal Encounters] Exists,” which explained:
This project to create a comprehensive national database of people who are killed through interactions with police started with a simple question: How often does that happen?
May 18, 2012: I was on my way home from work when I noticed a bunch of cop cars down by the Truckee River. As the editor of a newspaper, the Reno News & Review, I was curious. We’re a weekly, so we don’t much cover the police beat—not the day-to-day stuff anyway—but it’s my nature to satisfy my curiosity. So when I got home, I turned on the scanner app on my cell phone, fired up my laptop, and poured a glass of red wine.
It turned out the police had pulled over a stolen car, and they’d shot and killed the driver. (Jace Herndon, 41, we found out later.) Honestly—and not because I’m one of those hard-boiled, cynical types—I wasn’t particularly surprised or offended. Criminals often come to a bad end.
But again, I’m an editor, so I noticed when a gaping hole appeared in every single news story I read about the incident. There was no context. I kept looking for a sentence that said something like “This was x person killed by police in Washoe County this year.”
But it was never there. I searched the web for a few minutes, came up short, and started doing something productive. I simply considered the missing information a failing of the local news media, and I moved on. Still, its absence bugged me. I felt as though I’d accidentally left my wallet on my nightstand; while I knew I could retrieve it if I needed to, not having it was bothersome.
And then a few months later, an 18-year-old, naked and unarmed college student, Gil Collar, was killed by University of South Alabama police on Dec. 6, 2012. Early reports said the officer never got within five feet of the kid, and no non-lethal methods were tried. “Wow, how often does that happen?”
A separate page indicated the site was owned and operated by journalist D. Brian Burghart. When we visited the site, there was no immediately obvious citation for the claim police had killed 120 civilians between the death of George Floyd on May 25 2020 and June 14 2020 (the first iteration of the claim we found, and one posted 20 days after Floyd’s death).
Under the third item on FatalEncounters.org’s navigation bar, “Tools [for Journalists],” was a dropdown menu and an option labeled “Spreadsheets.” One of the spreadsheets was viewable on Google Docs, titled “FATAL ENCOUNTERS DOT ORG SPREADSHEET.” In its initially accessible format, entries were listed descending by date. Floyd’s name was not immediately viewable.
We made a copy of the spreadsheet in order to sort, and sorted entries by “Date of injury resulting in death (month/day/year)” or “column I.” That appeared to be the source of the claim, because when sorting by date, the spreadsheet had one header row for line one, and a floating row in line two — indicating entries after a specific date had “not been fact-checked.”
Toggling back to the original copy of the sheet indicated the entries which had “not been fact-checked” line appeared at the end of the sheet’s entries, and no entries appeared below it.
When sorted by date, descending, Floyd’s name appeared on line 122. Subtracting line one (header rows) and line two (new entries not yet “fact-checked”), Floyd was number 120 as of June 17 2020:
Data from FatalEncounters.org
In January 2020, Michael Harriot authored an article for The Root, titled “Here’s How Many People Police Killed in 2019…We Think.”
Harriot began by describing two incidents that would not be included in statistics involving deaths at the hands of police:
On Oct. 14, 2019, 12-year-old Akeelah Jackson was returning to her St. Louis home from a Family Dollar store.
As she crossed the street, an unidentified St. Louis County police officer was pursuing a traffic violation at 59 mph on a street zoned for 30 mph, according to the St. Louis American. When the car smashed into the seventh-grade student, the collision broke every bone in her body except for the ones in her feet … Akeelah Jackson died from injuries sustained in the accident.
Even though the police officer was speeding; even though the cop didn’t turn on his police siren or his police lights, he still hasn’t been charged with Akeelah’s death.
When most organizations tally the number of civilian deaths caused by police in 2019, Akeelah Jackson will not be counted.
No one actually knows how many people police kill every year.
Harriot went on to point to frequently-cited, ongoing tallies of civilian deaths and police, explaining that specific and commonly-referenced resources had individual blind spots:
For instance, the Washington Post’s database only lists people who were shot and killed by officers, leaving out victims like Akeelah Jackson. They don’t count deaths in police custody; they don’t count people who have died as a result of beatings or tasers or vehicular deaths. Most outlets also don’t include people who died at the hands of off-duty officers, so casualties like [45-year-old Johnathan] Liddell doesn’t appear in the databases.
In January 2020, The Root partnered with FatalEncounters.org’s Burghart to attempt to arrive at a more accurate figure:
Like the Washington Post, and Killed By Police, Fatal Encounters’ data comes primarily from media reports. But unlike the previously mentioned sites, Fatal Encounters also counts citizens who were stabbed, tasered, or beaten to death, or otherwise died during a police encounter. Fatal Encounters also includes suicides and people who were killed by suspects during police encounters, but we removed these numbers from our tally.
After comparing the Fatal Encounters dataset to the Post’s we determined that the Post failed to identify 117 deaths that fit Fatal Force’s criteria. The Post’s list also had 13 duplicate entries; three deaths that occurred prior to 2019; two people who died, but weren’t killed by a cop and one person who did not die at all.
Fatal Encounters’ dataset also had mistakes. The errors include three duplicate entries; two people who were not killed, and one death that they completely missed. But FE also listed 117 shootings that fit the Washington Post criteria but were not included in the Washington Post dataset. The total number of people who died during encounters with police, according to Fatal Encounters, totaled 1,777. When we removed suicides, the number fell to 1,531—forty-four fewer deaths than 2018.
In terms of scope, The Root examined a single year (2019) for their reporting. Harriot cited six entries of FatalEncounters.org’s 1,777 that were labeled as “errors,” or .33 percent of its yearly tally. The Root subtracted suicides (246) from FatalEncounters.org’s 2019 numbers and arrived at a figure of 1,531, of which six was .39 percent.
Harriot’s conclusion lamented the hodgepodge record-keeping when it came to police and deaths of American civilians, adding that media reporting remained the most reliable source to create any sort of tally:
For now, we must depend on the press to report how many people die at the hands of cops every year. All we know is, it’s more than you think. Even though the most-cited source is increasingly flawed, it’s all we have until America finally solves this problem …
On June 8 2020, about a week before the Facebook post and tweets about 120 deaths in 20 days began circulating, the Washington Post published an article that began:
Protests spread over police shootings. Police promised reforms. Every year, they still shoot and kill nearly 1,000 people.
The Washington Post noted that it began compiling records of officer-involved shootings in 2015, the year Michael Brown was killed in Ferguson, Missouri:
That year , The Washington Post began tallying how many people were shot and killed by police.
By the end of 2015, officers had fatally shot nearly 1,000 people, twice as many as ever documented in one year by the federal government.
With the issue flaring in city after city, some officials vowed to reform how police use force.
The next year, however, police nationwide again shot and killed nearly 1,000 people. Then they fatally shot about the same number in 2017 — and have done so for every year after that, according to The Post’s ongoing count. Since 2015, police have shot and killed 5,400 people.
Immediately thereafter, the paper included a bar graph of police shootings each year, from 2015 onward. It appeared 2019 was the deadliest year to date in the Post‘s records with 1,004 fatal police shootings.
As of June 7 2020, 463 fatal police shootings were tallied for 2020:
Additional Figures and Percentages
Per the Washington Post‘s 2019 numbers, an average of 2.75 Americans were killed in fatal police shootings each day — but Floyd died of asphyxiation, and thus would not be included in those figures. June 7 2020 was the 158th day of 2020, in which the Post counter 463 fatal police shootings. Averaged out across 158 days, that represented approximately 2.9 fatal police shootings per day.
Based on the average of 2.9 fatal police shootings per day in 2020 through June 7 2020, a 20-day period would average 58 fatal police shootings alone. However, factors such as ongoing unrest across the country could result in numbers skewed higher due to unusually frequent and tense interactions between civilians and police.
The 2.9 figure based on that article was — as Harriot said — not inclusive of people like George Floyd, where shooting was not the manner of death. Put another way, the error rate Harriot calculated of FatalEncounters.org was .39 percent, which didn’t adjust the 120 deaths since George Floyd’s May 25 2020 death meaningfully.
If we applied the rate to the 1,777 deaths FatalEncounters.org tallied and the 1,531 The Root included in its figures, it slightly complicated the numbers. If we divided 1,777 and 1,531 by the 365 days per year, we arrived at 4.86 and 4.19 respectively.
So an unadjusted average for FatalEncounters.org’s 2019 rates multiplied by 20 would result in 97.2 deaths — short of the 120 tallied on the list. Adjusting down by The Root’s parameters, the twenty-day average would be 83.8 deaths. And if we adjusted the 120 figure down by 13 percent — the percentage change between 1,777 and 1,531 — the figure was approximately 104.5.
In Sum: We Don’t Know
Several viral posts claimed 120 people in the United States died at the hands of police in the 20 days between George Floyd’s May 25 2020 killing and June 14 2020; subsequent posts on subsequent days carried over the figures of the original post. That data was derived from FatalEncounters.org, a site which uses journalism to compile a complete list of people’s deaths at the hands of police. Occasional replies to the posts claimed the figures were “bullshit,” but the actual answer was a bit more sobering — we don’t actually know how many people die each year at the hands of police, and nor do we know how many have died since Floyd was killed. Attempting to replicate the figures going by proportions did not suggest the number was inaccurate or exaggerated. In fact, it appeared based on our calculations that the number was close to accurate — and possibly still under-counting such deaths.