As America emerged from the pandemic, communities continued to experience a rising tide of gun violence. School shootings and the rate of children and teens killed by gunfire both reached all-time highs since at least 1999. ProPublica’s coverage of gun violence reveals how first responders, policymakers and those directly affected are coping with the bloodshed.

This article is produced in collaboration with The Texas Tribune and the PBS series FRONTLINE. Sign up for newsletters from The Texas Tribune and from FRONTLINE.

During an October conference to prepare law enforcement for dealing with an active shooter, Nevada State Police department Lt. Jacob Fisher played body camera footage that showed what he believed was a key misstep during the country’s deadliest mass shooting.

The footage from the 2017 massacre at the Mandalay Bay Resort and Casino showed a veteran Las Vegas Metropolitan Police officer and his trainee waiting on the floor below the shooter instead of confronting him as he spewed gunfire into the crowd of concertgoers outside. Six years later, Fisher told the room full of law enforcement officers and firefighters at the conference in Grapevine, Texas, that he had trained the veteran officer and felt like he had failed.

Sixty people died, including Fisher’s best friend, and more than 850 others were injured.

“I failed, and because of that I had to carry my best friend’s casket nine days later,” Fisher said. “I failed his kids. I failed his wife. I failed that organization because, for whatever reason, I couldn’t break through and find a way to convey a message to that cop to where he would run 800 yards toward active gunfire.”

While stressing the importance of training and leadership, Fisher made a stunning admission. He said he sends his children to school with bulletproof backpacks and tourniquet kits because he doesn’t trust that law enforcement officers would save them in the event of a mass shooting.

“Why do I as a parent have to arm my children at the ages of 8 and 10 and teach them those things?” said Fisher, who spoke with a reporter at the conference but could not later be reached for comment. “Because the cops in my jurisdiction, I don’t trust to go save them.”

The Nevada Supreme Court ordered the release of the video Fisher played six months after the October 2017 massacre, siding with news organizations who had sued the Las Vegas police department to make the records public. The veteran officer was fired after its release but was later reinstated, according to what the union president told newsrooms at the time.

The concerns that Fisher raised about the officer’s inaction were not mentioned in the after-action report, nor was the fact that the officer was fired for his behavior that day. The nearly 200-page accounting of the police response simply stated that 10 minutes after two officers arrived on the hotel floor below the shooter, they moved to the same level as his locked room, where they waited in the stairwell. They then remained there because their radios didn’t work and they were “unsure of what was taking place,” the report stated.

ProPublica, The Texas Tribune and FRONTLINE found that active shooter training varies widely across the country and that law enforcement officers make similar mistakes in mass shooting after mass shooting. Yet those failures are not always clearly identified in reports dissecting the incidents, adding to the difficulty of learning from past missteps.

Communities often rely on after-action reviews of mass shootings for a comprehensive and independent assessment of what happened. In the Las Vegas case, the body camera footage allowed the public to see what the report did not address, but the failure to release records, video and other evidence after mass shootings can leave many in the dark even when these analyses are issued.

Despite the U.S. facing more than 120 mass shootings in the past two-and-a-half decades, ProPublica, the Tribune and FRONTLINE found that there is no agreed-upon national standard for who conducts after-action assessments of law enforcement’s response, what they should examine or whether the resulting findings should be released.

Reports were never publicly issued in many cases, such as the 2018 shooting in Santa Fe, Texas, that resulted in the deaths of 10 people and the 2019 El Paso, Texas, shooting in which 23 died. And when they were made public, there was wide variability in what they contained. The news organizations analyzed more than three dozen publicly available after-action reports, finding that some excluded key details about officers’ actions or failed to fully explore other missteps, including individual delays in engaging the shooter.

Some reviews have “really important chunks that are missing,” said Louis Klarevas, a mass shooting expert and research professor at Columbia University’s Teachers College who argues that more records should be made public. “That raises a red flag.”

“Why did they focus on just this one aspect, or these two aspects? And what about everything else?” he added.

While the after-action reviews often focused on identifying broader issues, such as breakdowns in communication, incident command and multi-agency coordination, the newsrooms found that some did not evaluate whether significant delays in confronting shooters were justified.

In 2009, officers in Binghamton, New York, responded within three minutes to calls of an active shooter at the American Civic Association, a nonprofit that provides immigration services. Not hearing gunfire, officers waited 40 minutes for specialized teams to arrive and enter the building. They found the shooter had killed himself, according to an after-action report conducted by the county. The review of law enforcement’s actions during the shooting, in which 13 people died, called the police response “timely.”

In 2018, police in Thousand Oaks, California, initially approached the restaurant where the gunman shot patrons but retreated after he fired at them. Law enforcement did not reenter for more than 40 minutes, and ultimately did so only after SWAT arrived. A dozen people died in the shooting, including a responding officer who was accidentally struck by bullets from another police officer’s gun. The after-action report, conducted by the sheriff’s department, mentioned the delay but did not draw a conclusion about whether it was justified, stating, “While a request for the SWAT team was appropriate, the question of a second entry by patrol, prior to SWAT’s arrival, remains.”

In another case in 2016, a shooter unleashed hundreds of rounds of gunfire into a crowd at the Pulse Nightclub in Orlando, Florida, while the first officer on scene waited outside for backup. Officers then attempted to negotiate with the shooter after he barricaded himself in a bathroom despite 911 calls indicating hostages had been shot. It took three hours to take him down. An after-action report determined that law enforcement had followed best practices and agency protocol, but stated that those policies and training needed to be reexamined.

Authors of the Las Vegas and Binghamton reports did not respond to the newsrooms’ questions. Neither did the officer Fisher mentioned or the union that represented him.

The Ventura County Sheriff’s Office, which conducted the after-action review of the shooting in Thousand Oaks, said its report was “an accurate and transparent assessment of what went right, and what could be improved upon.”

Jim Burch, president of the National Policing Institute, a nonprofit that conducted several such reviews, including the one for the Pulse Nightclub shooting, said after-action reports are intended to identify areas of improvement and should not be viewed as the final arbiter of officers’ individual decisions.

Burch said because the reviewers are not conducting a criminal or disciplinary investigation, they must balance transparently describing what went wrong against avoiding assigning individual blame.

He and other experts who conduct such reviews told the newsrooms that their work often unfolds concurrently with litigation, criminal investigations, and internal disciplinary reviews. As a result, they can face limitations that include key personnel declining to speak and an unwillingness or inability, legal or otherwise, to share records. In other cases, the scope of the review may be restricted to a specific aspect of the shooting, such as the medical response, or confined to the role of the agency that requested the report, even if multiple departments responded.

“Some places just aren’t friendly to these things,” Burch said. “Some jurisdictions just don’t buy in. Some leadership doesn’t buy in. So, look, that’s just something we have to work around.”

That’s what happened after the 2021 shooting at Oxford High School in Michigan. A nearly 600-page after-action report delivered about two years after the shooting largely condemned a slew of missteps and missed opportunities by school leaders. The report, however, did not reach a conclusion about a critical action by an armed school employee because reviewers could not interview her. The monitor opened the door to the boys’ bathroom where the gunman had cornered two students during his shooting spree, but she did not enter. After she left, the shooter killed one of the students, according to the report. The other, Keegan Gregory, escaped.

The monitor’s reasons for leaving “are unclear and confusing and she is the only one who could provide the clarity that the community desires,” the report found. Without her cooperation, the report said, “we have only her repeated statements that she does not know why she did what she did. Unfortunately, we cannot provide any further insight.”

The employee and the report authors did not respond to a request for comment from the news organizations.

Chad Gregory, Keegan’s father, is frustrated that the report did not provide all of the answers parents were seeking because of a lack of participation from school personnel.

Gregory said that, ideally, there would be a national organization or agency in charge of setting standards for how police respond to mass shootings and how schools should conduct lockdown drills, as well as ensuring that communities receive needed services. After a shooting, such an agency could also put pressure on departments that were involved to ensure accountability and corrective action. Otherwise, Gregory said, the burden falls on parents and survivors who are trying to cope with “war-level” trauma to continue demanding resources and accountability.

“There is no authority. There is no dedicated resources,” Gregory said. “It’s everyone fend for themselves and figure it out.”

If law enforcement wants to maintain the trust of the communities they serve, officials need to be transparent about what transpired, said AJ DeAndrea, one of the SWAT team members who responded to the 1999 Columbine High School shooting.

An after-action report should be “critical about what happened in a way that we can learn from it,” DeAndrea said.

“That is the purpose of an after-action review, to say, ‘This is what we did good. More importantly, this is what we did bad,’” he added.

We reviewed dozens of reports that were issued since the Columbine High School shooting. Of those, only the 2018 shooting at Marjory Stoneman Douglas High School in Parkland, Florida, and the 2022 massacre at Robb Elementary School in Uvalde, Texas, explicitly condemned officers’ inaction to stop a shooter. In both cases, state lawmakers had called for the reviews.

In the Parkland case, a commission established by legislators, which included several Florida police leaders, issued a report that pointed to various missteps by law enforcement, including that the first officer on the scene never entered the building and told others who arrived later to stay away from the school. The report concluded that the officer, Scot Peterson, “failed to act consistently with his training and fled to a position of personal safety” while the shooter killed students and staff.

Several officers were fired after the shooting and Peterson was prosecuted for neglect of a child, culpable negligence and perjury. A jury found him not guilty of the charges in June of this year. Mark Eiglarsh, an attorney who represented Peterson, said the ruling proved that neither the accusations in the report nor the charges were justified. The Florida sheriff who led the commission that produced the report declined an interview with the news organizations.

A report on the Uvalde shooting that was released in July 2022 by a bipartisan committee appointed by Texas House Speaker Dade Phelan concluded that law enforcement at all levels had failed in their response. Nineteen children and two teachers died that May.

Before the report was issued, the only public review of what occurred was a limited assessment published more than a month after the shooting by Texas State University’s Advanced Law Enforcement Rapid Response Training Center. The analysis, which was conducted at the request of the Texas Department of Public Safety, was intended to identify gaps in officers’ training. But the report was criticized by some who believed it reinforced the narrative put forth by DPS that local law enforcement was largely to blame, despite hundreds of local, state and federal officers descending upon the scene. Pete Blair, ALERRT’s executive director, declined to comment on the report. He has previously defended it as reflective of the information that was available at the time.

The 77-page report by the speaker-appointed committee was released almost two weeks later. It offered what was at the time the most complete account of the failed response, although it did not publish any of the raw materials that the committee had obtained.

The committee also faced limitations. Most state and local officers complied with the committee’s interview requests, but it was unable to directly speak with more than 180 federal officers who responded, instead relying on comments they were seen making on other officers’ body camera footage and interviews they provided to DPS investigators, said state Rep. Joe Moody, the only Democrat on the committee. Lawmakers also did not have access to medical examiners’ reports for the victims, making it difficult to definitively determine how many might have survived if there had been a faster response.

Jesse Rizo, an uncle to 9-year-old Jackie Cazares, who died that day, said the state committee’s report was a good first step. But since then, he and other families have been waiting for a final review from state investigators; it has yet to be released 19 months after the shooting. A federal review by the Justice Department is also pending.

ProPublica and the Tribune are part of a coalition of news organizations that have sued DPS for records related to the shooting. A state district judge ruled in the news organizations’ favor this month, but DPS is appealing.

Rizo said some families have relied on media reports to fill in the gaps left by officials.

Since the House committee’s report was released, news organizations, including ProPublica and the Tribune, have revealed multiple missteps in the response and published video and audio that had not been released as part of any previous reports. In an investigation published this month, ProPublica, the Tribune and FRONTLINE showed that while the children in Uvalde followed their training when confronted with a mass shooter, many officers did not. Among multiple flaws in the law enforcement response, many officers said that they did not initially realize there were students in the classrooms because they were so silent. The children were following their training.

The lack of transparency regarding law enforcement’s failings, Rizo said, has some families questioning their faith in the democratic process.

“You ask the questions and you think, ‘Man, I pay taxes, I should be getting these answers,’” Rizo said. “And what do you get? Nothing.”

“You begin to acknowledge, to begin to, I guess, accept the fact that those answers may never come,” he said.

Anna Clark contributed reporting.