ProPublica

by Megan Rose

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up to receive our biggest stories as soon as they’re published.

Update, Sept. 30, 2022: A military judge found Seaman Recruit Ryan Mays not guilty on Friday of setting fire to the USS Bonhomme Richard. Mays, 21, had stood trial on charges of aggravated arson and willfully hazarding a vessel for the four-day blaze that destroyed the amphibious assault ship in 2020.

On the morning of July 12, 2020, the first orange flickers of destruction took hold in the bowels of the hulking USS Bonhomme Richard as it sat moored at a San Diego naval base.

Unimpeded, the fire gathered force, surging upward, conquering one level of the 844-foot ship and then the next, while the crew — the ship’s critical firefighting force — fled to the pier. There, the captain and his sailors stood by as the Bonhomme Richard burned, in cruel irony of its motto “I have not yet begun to fight.”

Not until the San Diego fire department went aboard did anyone spray water on the fire — nearly two hours after it had started. By then it was too late. Gas cylinders were exploding and shooting through the air, and firefighters didn’t have a map or even a sailor to guide them through the smoky maze of the ship. A firefighter’s warning that a compartment was “about to blast” forced firefighters off the Bonhomme Richard just minutes before an explosion so powerful it was heard 13 miles away and hurled debris onto a nearby destroyer.

That afternoon, the flames, hot enough to warp steel beams, danced along the flight deck and engulfed the ship’s outer structure. As the inferno raged, it melted the inside of the 300-ton control center on top of the ship, spewing molten aluminum onto the decks below.

Before nightfall, the Bonhomme Richard was a salvage heap. Sailors later watched as the forward mast, where the American flag flies while the ship is at sea, collapsed.

In this handout released by the U.S. Navy, a fire engulfs the USS Bonhomme Richard at Naval Base San Diego on July 12, 2020. (Mass Communication Specialist 2nd Class Austin Haist/U.S. Navy via Getty Images)

Four days later, when the fire was officially out, the U.S. Navy faced the sickening loss of a $1.2 billion warship, not to war, or even at sea. But to a wholly preventable fire while moored in a stateside port. For the Navy, whose reputation as the world’s finest had been battered by recent collisions at sea and allegations of shoddy equipment and training, the loss of the Bonhomme Richard was an embarrassing — and painfully public — blow.

The service immediately launched two parallel investigations into what went wrong and why.

The command investigation, led by a three-star admiral, sent a team of investigators on a prodigious and methodical examination of the fire. As the months passed, the investigators uncovered in exhaustive detail an astonishing array of failures — broken or missing fire hoses, poorly trained sailors, improperly stored hazardous material — that had primed the ship for a calamitous fire.

A separate investigation by the Naval Criminal Investigative Service, for its part, focused on whether anyone was criminally responsible. As the months passed, NCIS investigators appeared to operate in isolation, discounting the damning findings of the command investigation to pursue a case of arson, despite scant evidence.

Six weeks into both inquiries, the Navy told the command investigation to accept at face value what NCIS and federal fire investigators judged to be the fire’s origin. Both investigations concluded in 2021.

The command investigation traced the problems back to when the Bonhomme Richard docked for maintenance and Navy leaders throughout the ranks abandoned responsibility for the ship’s safety. Risks mounted, and nobody paid attention. All told, investigators determined that the actions of 17 sailors and officers directly led to the loss of the ship, and those of 17 more, including five admirals, contributed. The long list was a staggering indictment of everyone from sailors to top admirals who had failed in their jobs.

The NCIS investigation, however, laid the blame at the feet of a single young sailor. The true culprit, the one who bore responsibility for the billion-dollar loss, the Navy said, was then-20-year-old Ryan Mays. And for that, he should face life in prison.

The Navy continued its pursuit of Mays, even as a military judge recommended against it, bluntly calling out the lack of evidence and citing the findings of the Navy’s own command investigation.

Starting this week, Mays is being court-martialed in a military trial in San Diego for aggravated arson and willfully hazarding a vessel.

Ryan Mays, in long sleeves, and his lawyer Lt. Cmdr. Jordi Torres after a court hearing in August (Devin Yalkin, special to ProPublica)

A Navy spokesperson said the service couldn’t comment on ongoing litigation, but noted that the admiral who ordered the court-martial carefully reviewed the recommendation before deciding to move forward. “Everyone is entitled to a fair trial and the Navy is committed to upholding that principle,” Cmdr. Sean Robertson said.

In court, the Navy has tried to stymie questions about its motivations — and even to quash evidence its own officers gathered about dangerous conditions aboard the Bonhomme Richard. Last month, the prosecutor, Capt. Jason Jones, asked the judge to forbid Mays’ counsel from presenting the conclusions of the command investigation. Mays, who now holds the lowest military rank of E1, should not be able to ask the court to view him as the fall guy, Jones said.

The Navy wasn’t the problem, Jones said in court, and the trial argument shouldn’t in any way imply that “the Navy needs a scapegoat and therefore we picked an E1.”

Firefighters prepare to board the Bonhomme Richard two days after the fire started. (Mass Communications Specialist 3rd Class Jason Waite/U.S. Navy)

On Naval Base San Diego’s Pier 2, investigators cordoned off the charred wreckage of the Bonhomme Richard with yellow police tape.

Within days of the fire being extinguished, Bureau of Alcohol, Tobacco, Firearms and Explosives investigators, who co-led the criminal inquiry with NCIS, stepped into the watery shell of the ship, looking for what ignited the blaze.

Because the amphibious assault ship had been “in the yards” at the time of the fire, its flight deck, which stretched longer than two football fields, had been empty of the helicopters and F-35 fighter jets it was designed to carry. One of eight such ships in the Navy, the Bonhomme Richard is like a small aircraft carrier, made to ferry thousands of Marines into combat. It had been out of commission for a year and half during a $249 million overhaul. The morning of the fire, the upgrade was nearly complete.

The agents determined that the fire had started in an area of the ship known as the “lower V,” which normally stowed dozens of Marine Corps tanks and other vehicles, but during the overhaul was being used as a catchall, according to testimony and reports. On the day of the fire, the lower V had been packed with two fueled forklifts, a man lift, pallets of hand sanitizer, lithium batteries and other combustibles, wood beams, scaffolding, rope and thick, tall cardboard crates, some stacked two high.

A warship is an unusual scene for ATF investigators. They turned to the ship’s damage control assistant, Lt. Cmdr. Felix Perez, for a tour. Perez was the officer directly in charge of the firefighting hoses and systems aboard the ships, training sailors to fight fires and ensuring the ship followed fire prevention precautions.

Perez guided the agents through the ship, stopping at the fire stations closest to where the fire began. At three, hoses were missing, cut or otherwise unusable. Perez told the agents he or his staff had walked the ship two days before the fire, and it was nearly impossible they had overlooked the fire stations, according to an NCIS affidavit about the case.

The stations, Perez told the agents, must have been tampered with.

Sailors stand at attention during morning colors as the Bonhomme Richard burns in the background. (Mass Communications Specialist 3rd Class Jason Waite/U.S. Navy)

A few weeks later, in a building catty-corner from the Bonhomme Richard, some 77 naval officers, enlisted experts and civilians set up shop for the command investigation.

Early on, investigators realized Perez had not done his job well, according to a person close to the investigation who spoke to ProPublica on the condition their name wouldn’t be used so they could speak freely about sensitive matters. The fire stations were inoperable from broad neglect — and Perez and other leaders had failed to recognize the disintegration of the ship’s condition.

From the start of the command investigation, NCIS immediately curtailed the team’s efforts, forbidding its experts and officers from interviewing anybody from the ship, command investigators wrote in their report. Backed by Navy policy, NCIS’ criminal inquiry took precedence over the systemic investigation. So the command investigators turned instead to 26,000 pages of records, downloading databases and piecing together logs of the ship’s equipment, maintenance and training.

The investigators soon discovered an astonishing list of ways the ship was at risk, so many that cataloging the bad decisions day after day became depressing, the person involved said. For long stretches, all the ship’s heat sensors, sprinklers and other emergency systems were turned off, investigators wrote in their report. On the day of the fire, just 29 of the ship’s 216 fire stations and 15 of 807 portable fire extinguishers were in standard working order.

Perez was the ship’s principal representative on its mandated fire safety council, which investigators found met ad hoc and seemed to exist simply to waive safety requirements. Investigators scoured the meeting minutes and logs looking for ways Perez and others had considered mitigating each risk created by waiving those requirements and found almost none. They concluded that Perez, as well as his direct boss and other senior leaders, had abdicated their responsibilities for addressing fire prevention on the ship. Perez declined to comment on the investigation’s findings but said he thought his naval record showed he was a good officer.

Investigators also learned that Bonhomme Richard sailors had been living nearby on a berthing barge, basically a floating dormitory, until shortly before the fire. But that week they’d started to move back onto the ship while it was still in disarray. The supplies of both the contractor and the ship were shoved everywhere. The disorder had become normalized enough, the investigation found, that sailors of all ranks routinely walked by improperly stored oil drums, gas cylinders and other combustible items without much thought to the danger. To investigators, the condition of the ship was an “ideal environment for the fire to develop and spread.”

Just days before the fire, Mays had angrily texted his division officer, complaining about having to live among contractors who were doing work that was “hazardous as fuck.” A worker was welding near his bunk as he slept, and Mays said he was burned by a stray spark. In 2015, a major fire started on another warship in a shipyard with similar conditions: sailors moving aboard while “hot work” was being done.

The command investigators hung posters of ship drawings all over the walls, each one tracking a different potential problem. While NCIS’ early impressions of the case included a theory of sabotage, another picture altogether was becoming clear to command investigators: The Bonhomme Richard had been a tinderbox.

Smoke rises from a fire onboard the U.S. Navy amphibious assault ship USS Bonhomme Richard at Naval Base San Diego on July 12, 2020. (Bing Guan/Reuters)

On July 24, 2020, less than one week after beginning the criminal investigation, the ATF preliminarily ruled the Bonhomme Richard fire to be arson.

The lead ATF agent, Matthew Beals, and his team of investigators had found no physical evidence anyone purposefully set the fire. Beals later testified that he’d ruled out accidental causes, such as electrical and mechanical, as well as natural ones. With those causes eliminated, along with his assessment of how the fire grew and witnesses' statements, he concluded it must have been arson.

The National Fire Protection Association’s 921 guide, essentially the fire investigation bible, requires investigators to use the scientific method to determine cause. “You can’t in the absence of everything else rule it was arson,” Robert Duval, a director with the association, told ProPublica. ATF said it could not comment on pending litigation. Beals testified that his methods followed the manual and his conclusions were based on a variety of evidence.

In the area of the lower V that was most heavily scorched by the fire, Beals focused on the large cardboard crates he’d later call “Amazon boxes on steroids.” He theorized based on field tests that someone used an accelerant to ignite them, but said in court that he couldn’t find any fire data to corroborate.

He would testify at Mays’ probable cause hearing that he believed the fire was started by someone purposefully using an open flame and possibly an accelerant on the boxes.

NCIS began interviewing Bonhomme Richard sailors. Earlier in the year, NCIS publicly acknowledged the field office in San Diego had mishandled a high-profile investigation into Navy SEAL Special Operations Chief Edward Gallagher for war crimes; Gallagher was acquitted. The Navy took corrective action against seven agents. As the NCIS office developed its case on the fire, its leadership was still under a cloud.

The Bonhomme Richard’s top enlisted sailor identified Mays as someone who disdained the Navy, according to an NCIS affidavit. To go with their finding of arson, agents had a name of a possible suspect.

Ryan Mays (Devin Yalkin, special to ProPublica)

Mays, just out of high school and still baby faced, was known as something of a loudmouth on the ship. He’d enlisted early in Kentucky with his mother’s permission as a 17-year-old senior who liked science, swimming and running across the football field shirtless with the flag before Friday night games. His Navy contract guaranteed him a chance at BUD/S, the notoriously grueling training program for the SEALs.

Mays appreciated the camaraderie of orientation, but wasn’t mature enough to last when training got intense. He quit on the fifth day. That left him without a Navy career path, and he was in limbo for months afterwards, awaiting a new assignment.

Mays stole a pair of headphones from the base store, thinking naively that it would be “a fast track out of the military,” he told ProPublica. To his surprise, he was only reprimanded, and the commander encouraged him to embrace his potential and his time in the Navy.

Mays said that restoked his desire to become a SEAL. The Navy gave sailors three chances. When he was sent to the Bonhomme Richard in March 2020, Mays was single-minded. He could stand on the flight deck and, on a clear day, see across San Diego Bay to the Coronado beaches where the SEALs train. That was where he wanted to be. Or at least in a special warfare program, like explosive ordnance disposal.

He buddied up with another sailor with similar goals, and they tested their physical limits with overnight workouts. Mays was small but strong and his ego surfaced in the usual teenage way: social media preening. Around the ship, he earned eyerolls from fellow sailors when boasting that he belonged among elite sailors. Mays was in reality a low-ranking deck sailor, mopping, painting and carrying out other menial custodial duties aboard the Bonhomme Richard.

Records and interviews showed agents zeroed in on Mays’ discontent and the fact that he dropped out of SEAL training.

Command investigators, meanwhile, were finding that the basic principle of firefighting as survival had withered during the Bonhomme Richard’s extensive time in port. Just as every Marine is a rifleman, every sailor is a firefighter. No matter what a sailor’s job is, knowing how to contain and extinguish a fire is paramount. When a ship is at sea, there’s nowhere to escape.

The investigative team broke into smaller groups to examine what should have happened during the fire and what actually did.

Shortly after 8 a.m., sailors first reported spotting smoke. Investigators were dumbfounded at the lack of urgency after that. Navy policy, they wrote in their report, dictates that sailors must douse flames with water as soon as possible but at most within 12 minutes. On the ship that day, more than 10 minutes elapsed before anyone even announced the fire over the ship’s loudspeaker. The slow response, they found, was typical for the Bonhomme Richard. For 14 drills in a row leading up to the fire, the crew failed to respond in time — a lack of proficiency that neither the ship’s leadership nor higher commands took steps to address.

That critical gap between the sign of smoke and the sounding of the alarm, investigators found, was the first in a cascading set of failures, by both the crew and leadership on the pier. Once the sailors realized many of the hoses nearest the fire weren’t operable, investigators learned, none of them moved to another important shipboard strategy to contain the fire: slamming shut the heavy steel hatches and watertight doors between compartments. And the sailors revealed that at first no one thought to use the ship’s sprinklers to distribute thick, white foam that can help extinguish the fire. Even if they had, they would have been unable to easily turn on the system: A maintenance report had been falsified in April, saying the system worked when it didn’t.

Ryan Mays (Devin Yalkin, special to ProPublica)

As the command investigation took a wide view of what led up to the fire, NCIS, records show, sharpened its focus on Mays. On the morning of the fire, Seaman Kenji Velasco, who had just come on duty along with Mays, had been standing watch at the top of the ramp to the lower V. He told NCIS that he saw someone go down the ramp shortly before the fire.

Velasco told nobody about this person for days, according to testimony — not even during the fire, when that person could have been in danger of being killed. When Velasco sat down with agents, he told them that someone had walked behind him, “but I’ve never seen him before.”

Agents went back to Velasco the day after the interview. This time Velasco said he was “fairly sure” the person was Mays, according to NCIS documents. Velasco told the agents that Mays was cocky and talked too much.

The agents then brought Velasco back again: How certain was he that he saw Mays? “90%,” Velasco told them.

The next month, agents scooped up Mays, surprising him as he checked in for work. The ATF’s Beals and an NCIS agent questioned Mays in a recorded interview for nearly 10 hours. He told them more than 150 times that he didn’t set the ship on fire.

The morning of the fire, Mays should have had a broom and dustpan in his hand, cleaning the back of the ship. Mays told agents he was instead just hanging out there, scrolling through his phone. With 24 hours of duty and not much to do, he wasn’t in a hurry right after roll call, and besides, he told ProPublica later, the general culture of the ship on its second extension in the yards was lackadaisical.

On a recording of the interview, Mays, wearing a brown uniform T-shirt and occasionally sweeping his hair off his forehead, vacillated between confrontation and distress.

In a confident voice, he asked eight times in two minutes if he was being detained.

“I’m not answering your questions, Ryan,” Beals said.

During his interview, Mays crudely put on display his disregard for the fleet Navy, and spoke of his desire to be a SEAL. In the beginning he even asked NCIS agent Albert Porter, a former SEAL, for a recommendation. Porter told Mays he’d never have another shot at the training program: “You’re not going back, dude. It’s not happening.”

Beals pressed him to “just admit to what you’ve done.” At one point, he told Mays they had him on video.

“You’re a liar,” Mays said.

“You’re a liar,” Beals replied.

Several times throughout the day, Mays asked to call his mom. He tried to think of anything he could tell the agents that would show them he was innocent. He begged them to take his DNA, search his phone and use GPS to track his whereabouts at the time.

At one point when the agents left him alone, he exclaimed to the empty room: “I didn’t do it. Let me go.”

Then he laid his head on the table and sobbed.

When Mays learned close to midnight he was going to the brig, a sailor who had been preparing to transport him said she heard Mays say something like “I’m guilty. I did it, I guess,” according to records and testimony.

Agents took the alleged remark as a confession. Mays said he was being sarcastic, expressing disbelief he was being arrested for a crime he didn’t commit.

The Navy booked Mays into the brig on Aug. 20, 2020.

A page from the Navy’s command investigation report shows aluminum that melted during the fire and flowed throughout the ship, top, and the gaping hole the fire left in the interior of the ship, below. (U.S. Navy command investigation)

In the months after the fire, NCIS and ATF agents were slow to interview some of the sailors who had been aboard the ship during the fire, NCIS records of interviews show. One of those sailors was among the first to see smoke that morning.

About a month after Mays was arrested, the witness told agents she had seen a sailor sprinting from the lower V about the time she saw a “white fog” wafting up. She identified the sailor as Seaman Recruit Elijah McGovern.

A month and a half later, records show, NCIS and ATF agents asked McGovern where he was when the fire started. McGovern denied setting the fire. He gave a series of stories over several interviews about where he was at the time that were later contradicted, witnesses testified. ProPublica could not reach McGovern for comment.

Beals and NCIS agent Maya Kamat investigated McGovern for months but ultimately set a low bar to clear him. They found grainy, distant video of a base exit near the Bonhomme Richard that showed a person leaving about 25 minutes after the fire started. Beals and Kamat testified they could not identify McGovern’s face on the video. But Beals said he could tell by the person’s “gait and walk and general build” that it was McGovern. Kamat said she thought the nondescript clothes matched McGovern’s. The agents decided the video helped provide McGovern an alibi, they testified at a preliminary hearing. NCIS said it could not comment on pending litigation.

The Navy was also quick to dump its own expert. After Mays was arrested, someone scrawled on a port-a-potty near the ship, “I did it. I set the ship on fire,” among other things, including a crude drawing of the ship in flames. The military’s handwriting examiner said he matched the script to McGovern’s.

Handwriting analysis is controversial, but the government often presents it as trustworthy evidence. Here, though, the graffiti didn’t identify the sailor the Navy ended up accusing of the crime, so prosecutors wanted it excluded from Mays’ court-martial, arguing at a preliminary hearing it was not strongly conclusive and irrelevant.

McGovern was kicked out of the Navy for misconduct the week of his last interview with investigators, records show.

A ship moored at Naval Base San Diego, 2022 (Devin Yalkin, special to ProPublica)

After the Navy lost a submarine, the USS Miami, to a fire in a Maine shipyard in 2012, the service adopted a “never again” mentality and issued a new 129-page fire safety manual. Command investigators concluded it was merely a paper fix. Three major fires in shipyards after the Miami had similar troubling patterns. And investigators found that some admirals in charge of maintenance weren’t following the manual.

NCIS had kept investigators from interviewing 150 Bonhomme Richard crew members and others until the first week of December, five months after the fire, according to the command report. By the time investigators could sit down with the ship’s leadership, they had already pieced together the ship’s poor condition. As officers came in one by one, investigators were surprised to find out how little they knew about the state of their own ship.

Capt. Gregory Thoroman, the Bonhomme Richard captain, broke down several times throughout the interview. As a Navy pilot put in charge of a large ship, he was somewhat out of his depth of expertise. For example, Perez hadn’t always told the captain about the safety requirements he was waiving, investigators wrote, and Thoroman didn’t know enough to ask.

The Navy’s backstop system for when an aviator such as Thoromon is in charge is to require the No. 2 be a surface warfare officer. That also failed: Capt. D. Michael Ray, investigators learned, wasn’t paying attention either.

Investigators were startled to find that even though the ship had recently loaded 900,000 gallons of fuel, none of the ship’s leadership knew which emergency response systems were working. Thoroman hadn’t read the fire safety manual, investigators found. Ray and other key officers on the ship, including Perez, didn’t understand it. Neither Thoroman nor Ray responded to requests for comment.

Command investigators also found that the admirals charged with overseeing ships in maintenance hadn’t noticed the rising risks on the Bonhomme Richard. Other admirals and captains responsible for fire response didn’t ensure even foundational precautions, such as having large fire pipes on the piers and the distribution of ship maps to local fire departments.

The Navy was at risk for mishandling even a minor fire, investigators found.

Mays and Torres enter legal offices on Naval Base San Diego in August. (Devin Yalkin, special to ProPublica)

A year after the fire, as top Navy leadership was endorsing the command investigation findings, the service was also charging Mays.

To some, the Navy’s actions were reminiscent of an ugly piece of its history. In 1989 an explosion on a turret of the USS Iowa killed 47 sailors, and the Navy tried to pin it on one of the dead sailors, who leaders suspected was gay. Only after Congress stepped in did the Navy acknowledge there was no evidence for its assertions.

More recently, after back-to-back collisions in the Pacific in 2017, ProPublica exposed how the Navy downplayed systemic culpability and fired those who had raised alarms.

After a dayslong probable cause hearing for Mays in December 2021, the judge said she wasn’t persuaded of Mays’ guilt. A nuclear-trained surface warfare officer who later became a Navy lawyer and then judge, Capt. Angela Tang is known for being thorough.

“Given the state of the evidence presented to me, I do not believe there is a reasonable likelihood of conviction at trial. Therefore I do not recommend referral of these charges even though there is probable cause to support them,” Tang wrote in her findings.

ProPublica reviewed the conclusions of her 43-page report, which the Navy has withheld from the public, as it has with almost all other court records in the case. Probable cause is a low bar, Tang cautioned. In explaining her recommendation to drop the charges against Mays, she repeatedly refers to “if” the fire was arson.

Her report also notes that defense expert witnesses testified about finding two other possible causes: lithium batteries that had leaked and exploded and arcing from an engine wire on a forklift. Given that evidence, the experts testified, the only reasonable conclusion was “undetermined.”

None of the evidence proved the fire was deliberately set, Tang wrote, and ATF’s conclusion could be doubted because agents missed possible causes found by the defense in the four hours they had at the scene. Tang noted Velasco was the prosecution’s lone eyewitness and — even if believable — wasn’t enough to hang the case on. She wrote that Velasco’s credibility was undercut by how long he took to report seeing someone and by his inconsistent statements, as well as by other witnesses who contradicted what Velasco said Mays was wearing. Furthermore, even if Mays was seen, it did not prove he set the fire or that it was arson, she wrote.

She also said investigators had valid reasons to suspect McGovern and wrote that jurors would probably view Mays’ “I’m guilty” remark as sarcastic.

Tang referenced, too, the Navy’s own conclusions that if not for systemic failures, the fire would have been extinguished long before the ship was lost. The Navy charged Mays with hazarding a vessel, but Tang wrote the fire, if arson, seemed intended to cause a distraction, not destroy the ship.

Vice Adm. Stephen T. Koehler, the military commander with the final say about whether to prosecute Mays, disregarded Tang’s recommendation. In February, the Navy announced Koelher had decided to send Mays to court-martial.

Much of the Navy’s case against Mays hinges on his alleged motive. Prosecutors are claiming he was a disgruntled sailor who hated the Navy so much he torched the ship out of revenge.

The Navy believes Mays was consumed by the realization he’d never don the SEAL trident, and having to move to a new berthing aboard the ship “sent him over the edge,” one of the prosecutors, Lt. Cmdr. Shannon Gearhart, said at a preliminary hearing.

Mays’ civilian lawyer at the time, Gary Barthel, argued at the probable cause hearing that Mays’ hubris proved he was an unlikable braggart but not an arsonist.

The young sailor was also making efforts to reapply to SEAL training, both through his punishing workouts and by asking for recommendations and taking steps to earn special qualifications to bolster his application, such as search and rescue swimming, according to Mays and his lawyers. His mom, Christy Hall, told ProPublica that her often-stubborn son “was bound and determined to go back” to the training program. He’d only had one strike, and had two more tries, she said he told her.

Ryan Mays (Devin Yalkin, special to ProPublica)

Two years into the ordeal, at a hilltop park overlooking San Diego, Mays sat with a military bearing, speaking with a newly reflective self-awareness. He said it’s fair for the Navy to say he had a bad attitude and didn’t want to be on the Bonhomme Richard.

“I don’t know how I feel about the word ‘disgruntled’ that they’re using right now. That doesn’t really make a lot of sense to me,” he said. “I was just like every other sailor in the department that didn’t appreciate, you know, cleaning shit. And so I don’t know if that makes me disgruntled or not.”

When Tang recommended against a court-martial, Mays said, he briefly felt relief. He thought that the anxiety that had him throwing up nearly every morning would finally subside. Then the Navy announced it would still prosecute him.

The experience, especially his time in the brig, has been “soul crushing,” he said. “A piece of me died in there and I don’t know if I’ll ever get it back.”

The Navy decommissioned the Bonhomme Richard after estimating it would have cost $3 billion and at least five years to fix. The service said that it made changes to its fire prevention policies, including instituting random safety checks and clarifying the chain of command. More than 20 people, including three admirals, were disciplined. The captain, executive officer and top enlisted officer on the ship all received punitive letters of reprimand, which is typically a career ender.

Mays’ defense has noted that on the same day as the Bonhomme Richard fire, another big deck ship at the San Diego base caught fire. NCIS concluded the USS Essex fire was arson, but that blaze was extinguished before it caused any damage. In this case, like Mays’, there was a single eyewitness who identified someone at the scene shortly before the fire was discovered. But the case was closed without charges after a yearlong investigation. The agent in charge testified that the suspect was “ultimately eliminated because nothing more linked him to the fire aboard the Essex.”

One of Mays’ defense attorneys pointed out at a preliminary hearing that “the Essex facts are eerily similar.” Except that ship didn’t burn down, so the investigation “just went away,” Lt. Pete Link said.

In the Bonhomme Richard case, there was a costly loss of an entire vessel, Link said, “and now here we are in court.”

by Emily Hopkins

This article was produced for ProPublica’s Local Reporting Network in partnership with Outlier Media. Sign up for Dispatches to get stories like this one as soon as they are published.

What typically is routine procedure — a utility requesting and then receiving approval for a consumer rate increase — has turned unusually contentious for DTE Energy, the largest provider of its kind in Michigan.

Critics are both highlighting the financial impact on Detroit-area consumers and drawing attention to other issues affecting local communities, including widespread outages and the company’s treatment of customers who can’t afford to pay their bills.

A March investigation by ProPublica and Outlier Media revealed that DTE had cut service for nonpayment more than 200,000 times during the pandemic. In an August story, the news organizations showed how DTE had sold off old customer debt, an unusual financial maneuver by a Midwest utility. Reporters found that DTE had received just pennies for every dollar of debt it sold to a collections company owned by a private equity firm. The consequences were severe for thousands of Detroiters who were sued and in some cases had their wages garnished.

Last week, three members of Congress, including Rep. Rashida Tlaib, a Democrat whose district covers much of Detroit, introduced a resolution recognizing access to utility services such as electricity, water and broadband as a human right. That resolution calls for a ban on the sale of household debt, the creation of a federal database to track disconnections, and holding a congressional hearing on utility issues, among other things. Tlaib also testified at a public hearing on the rate increase, joining in a chorus of protest that included dozens of customers.

Following our article last month about debt sales, a spokesperson for Michigan’s office of the attorney general said it is “more closely examining” DTE’s debt sale practice. The spokesperson added that the attorney general plans to raise the issue in negotiations over DTE’s current and future regulatory cases, which are decided by the Michigan Public Service Commission.

In addition, Detroit council member Angela Whitfield Calloway, who represents part of the city’s northwest side, told ProPublica that she intends to ask DTE officials to appear before the City Council to face questions about the company’s shut-offs, outages and debt sales.

“Selling the debt in my opinion is egregious,” said Whitfield Calloway, who, following the March investigation, co-sponsored a resolution calling for DTE to put a one-year pause on electricity and gas shut-offs.

“What’s in it for DTE?” she asked. “You’re causing harm to your customers.”

The DTE rate proposal would bring in an additional $388 million in annual revenue from residential, commercial and industrial customers combined. About 60% of that revenue increase would come from residential customers, who would see rates increase by 8.8%. On Monday, an administrative law judge issued a proposed decision that would cut the revenue increase to $145.7 million. The commission has until Nov. 21 to issue a final order.

Brynn Guster, spokesperson for DTE, said in an email that when the commission approves new rates, it will be the first base rate increase in nearly three years.

In public filings, the company said its rate increase proposal is driven by investments in infrastructure improvements that would prevent power outages and improve worker safety.

Guster also said that the company plans to invest in “a grid of the future that supports our fast-evolving lifestyles, businesses, and economy.”

The utility has defended its shut-off policies; its rate of shut-offs was higher than that of the six other investor-owned electric utility companies in the state, reporters found. A DTE spokesperson told reporters that the company works with customers to arrange an affordable payment plan or find financial assistance through programs for low-income communities.

In response to questions about DTE’s debt sales, Guster previously said the sales lowered the financial burden on other customers and that DTE only sold debt from “closed” accounts, where customers’ utilities had been shut off or residents had moved away from DTE’s service area.

An estimated 200 people attended an August hearing on the proposed rate increase. It’s the first time the Michigan Public Service Commission, which regulates utility rates, has held a public hearing dedicated to taking testimony on a rate increase.

“We understand that people are frustrated,” said Matt Helms, spokesperson for the commission. “This is why the Commission decided to hold the hearing in Detroit, so that commissioners could hear directly from customers concerned about the costs and reliability of their electric service.”

Annie Beaubien of Detroit testified at the hearing about two outages in as many days in July that left her home without air conditioning or fans during 90-degree weather. Getting information from DTE about what was wrong or when power would be restored was difficult, she said.

“It’s just completely ridiculous, the amount of money we pay versus the quality of service we get,” she said in an interview.

Tlaib, meanwhile, took aim at DTE’s shut-off policies at the hearing. “You know what’s outrageous, and what should be the biggest outrage for all of you as members of the commission: In 2020, during the worst of the pandemic, DTE shut off power to customers more than 80,000 times,” she said. Democratic State Reps. Laurie Pohutsky and Yousef Rabhi also testified against the rate increase.

On Aug. 29, just a week after the hearing, severe storms left some 265,000 customers without power. The outages led to closures at 24 Detroit public schools, and some homes were left without power for days. More than 43,000 customers were still without power as of the afternoon of Sept. 2.

Guster said the company apologized to its customers who were affected by the late-August storm and said that the extensive damage from high winds and the complexity of the repairs delayed DTE’s efforts to restore power to some customers. She added that 99% of customers who had experienced outages were reconnected by the evening of Sept. 2.

Siedah Spencer-Ardis, a marriage and family therapist in Detroit, said she went four days without power after the Aug. 29 outage — a situation she described as “hell.”

Her two kids missed three days of school. She and other therapists had to juggle appointments as outages affected both them and their clients. She said her family had to discard food from two refrigerators in her six-member household. And this isn’t the first time: She estimated that she’d lost groceries during DTE power outages four times over the last three years.

“This is like a recurring thing where we lose power,” she said. “They need to do better.”

by ProPublica

by Karim Doumar and Cynthia Gordy Giwa

Sign up for ProPublica’s User’s Guide to Democracy, a series of personalized emails that help you understand the upcoming election, from who’s on your ballot to how to cast your vote.

Welcome to ProPublica’s User’s Guide to Democracy! This series will give you the information and resources you need to vote in the 2022 elections.

Want a personalized voting guide for the 2022 midterms? Sign up for the User’s Guide to Democracy email series to see detailed information about voting in your congressional district.

Who You’ll Be Voting for in the 2022 Midterm Elections

We’re going to start off with some basics. (You probably learned some of this in elementary school, but it’s easy to forget, and we need to start somewhere!)

Members of Congress

Made up of the Senate and the House of Representatives, which together are theoretically coequal to the presidency, Congress is tasked with making laws on our behalf.

Each member of the Senate represents their entire state, with two senators per state. Unless filling a vacancy, senators are elected to six-year terms, and every two years about one-third of them are up for election. That means a lot of places don’t have a Senate race this year. You can see which senators’ terms are ending here.

No matter what state you live in, your congressional district is voting for a House representative in this year’s election. Each of the 435 House members represents a portion of their state, known as a congressional district, averaging 760,000 people. This is the person in the federal government closest to you, working in your district’s name. You can find your representative and their voting history in ProPublica’s Represent database.

The 2022 midterms have an added complication: redistricting. Every 10 years, states go through a political process of redrawing their congressional districts. Sometimes it’s because the state has gained or lost house seats after the census. For example after the 2020 census, Texas gained two house seats while California and West Virginia lost one each. In fact, seven states lost a seat, and six gained seats.

But even without gaining or losing seats, states change their district maps. Using increasingly sophisticated mapping techniques, many states try to look at voting history to construct districts they know will be safe seats for either party. This is called gerrymandering.

A CNN analysis shows that after the most recent round of redistricting, there are 17 fewer competitive districts across the country.

State Officials and Down-Ballot Contests

Which state officials are up for election in 2022 depends on where you vote. You might see candidates for governor, lieutenant governor, attorney general, secretary of state and judges from various levels of courts. Further down the ballot, you might also find offices such as treasurer and school superintendent.

Your state legislators are probably also on the ballot. Across the 50 states are 99 legislative chambers — all states have their own lower (larger) and upper (smaller) chambers except for Nebraska, which works as one big assembly. Of those 99 chambers, 88 have seats up for election this year.

You’ve heard this before, but it bears repeating: The decisions made by your local elected officials are the ones that most directly affect your daily life and can set the tone for your community’s values. But they are also the hardest to keep track of and understand, and voting patterns show they get the fewest votes. We’ve all shown up at the ballot box before without a full understanding of the job description for an official we’re charged with electing.

ProPublica can’t be everywhere at once, but here are just a few stories from the last year about how hyperlocal officials like school boards impact the communities they serve.

• White Parents Rallied to Chase a Black Educator Out of Town. Then, They Followed Her to the Next One.
• “We’re Going to Be Conservative.” Official Orders Books Removed From Schools, Targeting Titles About Transgender People.
• A Push to Remove LGBTQ Books in One County Could Signal Rising Partisanship on School Boards

Local governments also take the lead on crucial issues such as policing and holding cops accountable in your community. Their work affects everything from school funding and services, to rent, affordable housing, the environment and public transit. Your local government has tremendous influence over how your community is run.

That’s one reason ProPublica started our Local Reporting Network. Here’s some of what our partners doing watchdog work in communities around the country have found:

• The City Where Investigations of Police Take So Long, Officers Kill Again Before Reviews Are Done
• $53.3 Million. 33 Jobs. No Plan. That’s How Mississippi Lawmakers Are Spending BP Oil Spill Money.
• New York Let Residences for Kids With Serious Mental Health Problems Vanish. Desperate Families Call the Cops Instead.
• Casinos Pled Poverty to Get a Huge Tax Break. Atlantic City Is Paying the Price.

The User’s Guide to Democracy will be primarily focused on federal offices like the House and the Senate. For information on what else is on your ballot, check out Ballotpedia. Enter your address into this online resource to see a sample ballot of what you’ll be voting on in November.

Are You Registered to Vote for the 2022 Midterms?

Even if you’re pretty sure you are, take a moment to get 100% certain.

Sometimes, election officials clean up voter rolls to remove people who are inactive and those who may have moved or died. These efforts, unfortunately, sometimes mean that active, registered voters are swept from the rolls without their knowledge.

Take a moment now to double check your voter registration status. Vote.org, a nonprofit working to increase political engagement, has a look-up tool that lets you verify your voter registration in seconds.

If you’re not registered, or if you need to update your information, there may still be time — but deadlines are different in every state. Get all the forms and facts you need at your secretary of state’s website.

by Mick Dumke

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up for Dispatches, a newsletter that spotlights wrongdoing around the country, to receive our stories in your inbox every week.

Citing years of broken promises to build affordable homes, a Chicago City Council committee rejected a plan to lease public housing land to a professional soccer team owned by a billionaire ally of Mayor Lori Lightfoot.

That was on Tuesday. Less than a day later, allies of the mayor called a do-over and reversed the vote.

The full City Council then voted Wednesday to approve a zoning change needed to let the Chicago Fire soccer team build a practice facility on the 26-acre site.

A June story by ProPublica detailed how the land was once part of the ABLA Homes, a public housing development on the Near West Side where 3,600 families lived. After demolishing most of the ABLA buildings and displacing thousands of people, the Chicago Housing Authority promised to build more than 2,400 new homes in the area. So far, it has finished fewer than a third of them.

Lightfoot offered the ABLA site to the Fire late last year, and the CHA board signed off on the plan this spring. It is one of a series of deals the CHA has made to sell, lease or give away its land for nonhousing uses, including Target stores, a private tennis facility and a school running track and turf field. The Fire are owned by Joe Mansueto, founder of the investment research firm Morningstar.

The zoning change was considered one of the easiest steps needed to finalize the Fire lease, since Lightfoot’s allies control the city commission and the council committee that had to sign off. But it turned out that distrust of the CHA led to an embarrassing, if short-lived, setback for the mayor and her team, which had to resort to an unusual ploy for more time.

It’s hardly unprecedented for Chicago mayors to advance their agendas with strong-arm tactics and calculated legislative maneuvers. For example, former Mayor Richard M. Daley once waited until the council was about to adjourn before ramming through the repeal of a law he didn’t like in just a few minutes. Rahm Emanuel mastered a long tradition of letting loyal aldermen use committee budgets for patronage and perks. And Lightfoot fought to keep meetings with aldermen out of public view.

On many occasions, aldermen have been known to vote against ordinances they sponsored or to embrace measures they criticized after getting calls from the mayor. The do-over is another reminder of how Chicago mayors almost always get their way.

In defending the deal, the CHA has said that turning over land — the largest open plot at ABLA — to the Fire will not impact its plans for more housing. The agency has also said residents will benefit from recreational and job opportunities when the Fire facility is built. The team has emphasized similar points in describing the deal as an investment in the West Side.

But after years of watching CHA fail to deliver on its commitments across the city, many housing advocates and aldermen are skeptical. From the beginning of Tuesday’s zoning committee meeting, CHA and city officials were playing defense. “The CHA is well and far behind its goals,” Alderwoman Maria Hadden said at the meeting.

With rising homelessness and thousands of people on CHA waiting lists, she said, “It’s concerning to see such sluggishness.”

“Housing is what we do,” responded Ann McKenzie, chief development officer for the CHA. She said the agency hoped to start building 220 more units at ABLA this fall, split between low-income housing and market-rate buyers.

But McKenzie acknowledged that Hadden was right about the agency falling short of its commitments. Even when the new units are done, the CHA will have delivered less than half the homes it is obligated to under court agreements.

McKenzie vowed more housing will be built. To adjust for the land leased to the soccer team, she said, the CHA would have to concentrate more housing on nearby blocks. But she argued the Fire agreement could bring money and momentum.

“We actually think this is an improved plan,” she said.

McKenzie said under the current proposal the CHA would get $8 million up front plus about $750,000 to $800,000 a year for as long as 40 years.

A group of housing attorneys suggested in a letter to the committee that the plan violated civil rights laws and court orders. McKenzie and a city lawyer said that those issues would be part of a review by the U.S. Department of Housing and Urban Development, which must sign off any plan to dispose of public housing land.

Alderman Tom Tunney, Lightfoot’s hand-picked chair of the zoning committee, noted: “This is a zoning process. This is not about HUD agreeing with the use of the land.”

But aldermen continued to express frustration with the CHA.

“I came from public housing,” said Alderman David Moore, who grew up in the CHA’s Robert Taylor Homes and now represents a South Side ward. He wanted to see letters of support for the Fire deal from resident leaders.

McKenzie said she didn’t have letters but assured him the CHA had been talking with resident groups for months.

That wasn’t good enough for Moore. “I cannot support this item without a letter of support,” he said.

Alderman Jason Ervin, whose ward includes the Fire site, urged his colleagues to support the zoning change, arguing it would provide a boost to the community. Still, he said, “The concerns that have been raised are valid given what’s transpired with the CHA over the last 20 years.”

When Tunney began the roll call on Tuesday, it was quickly clear the measure was in trouble. Seven aldermen — including Hadden and Moore — voted no. Seven members of the committee were no longer present. And only four joined Tunney in voting yes. The item would not advance out of the committee.

But then, minutes later, Tunney announced that the committee would reconvene Wednesday morning to “reconsider the vote.”

With the help of two aldermen who changed their votes, the do-over on the CHA-Fire deal only took a few minutes.

Tunney started the meeting Wednesday by announcing the committee now had letters from resident leaders backing the Fire agreement. He asked for a motion to reconsider the first vote.

Under council rules, a vote can be reconsidered only if someone from the winning side moves to do so. Alderman Felix Cardona Jr., who voted against the Fire deal on Tuesday, was ready to make the motion. Tunney then called for a vote to reconsider the first vote. It passed 9 to 5 with the help of Cardona and Moore, who flipped from the day before.

Alderman Anthony Beale, a Lightfoot critic who cast one of the no votes, accused Tunney of “skirting our rules” by arranging for a second vote on the proposal.

Tunney brushed off the criticism, then announced that Cardona had moved for a vote on advancing the Fire plan. Cardona hadn’t said anything.

But the committee went ahead with the new vote to advance the zoning change to the full council. This time it approved the Fire deal, 9-5.

Afterward, Moore said the letters from resident leaders swayed him, while Cardona said Ervin persuaded him that ABLA residents were in favor of the proposal. Still, Cardona said he would make sure the council calls CHA officials to a public hearing to explain how they will build more housing.

“As you heard today, all my colleagues have an issue with the CHA,” Cardona said. “So the best thing for us is to bring them to the carpet and have a serious conversation with them.”

A few hours after the committee advanced the Fire deal, the full council approved it by a 37-11 vote. The CHA now has to submit the proposed agreement to HUD for review.

At a press conference later, Lightfoot said the CHA had done “an extensive amount of community engagement” in the ABLA neighborhood “to make sure the development really reflected what the community said they need.” Then she turned to praising Mansueto.

“Joe Mansueto has been very intentional about answering my call for business people to invest in areas of our city that have seen either little to no investment or, frankly, have been disinvested in,” Lightfoot said. “So I’m grateful to Joe Mansueto.”

by Perla Trevizo, ProPublica and The Texas Tribune, and Mike Hixenbaugh, NBC News

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up to receive our biggest stories as soon as they’re published.

This article is co-published with The Texas Tribune, a nonprofit, nonpartisan local newsroom that informs and engages with Texans. Sign up for The Brief Weekly to get up to speed on their essential coverage of Texas issues.

This article was also produced in partnership with NBC News.

The generator industry’s promised fix for deadly carbon monoxide poisoning was put to the test last year on a narrow patio outside Demetrice Johnson’s home after Hurricane Ida plunged much of Louisiana into darkness.

Johnson’s brand-new generator — equipped with a safety mechanism that manufacturers have said prevents “more than 99%” of carbon monoxide poisoning deaths — hummed into the night, inches from her family’s back door on Sept. 1, 2021, powering an air conditioner and a refrigerator.

If carbon monoxide levels got too high, the generator was designed to automatically sense the danger and trigger a shut-off switch.

But by the time emergency responders entered the three-bedroom brick house in Jefferson Parish the next morning, Johnson and her children, 17-year-old Craig Curley Jr. and 23-year-old Dasjonay Curley, were dead. They had been poisoned by exhaust fumes that flowed from the generator into their home, according to a sheriff’s office report, exposing a safety deficiency that federal officials and consumer advocates have warned about.

The safety switch’s failure to save Johnson and her children is detailed in an April report by the Consumer Product Safety Commission (CPSC) that was obtained this month by ProPublica, The Texas Tribune and NBC News through an open records request. The federal report followed an investigation by the news organizations that detailed the family’s deaths and found that attempts to make portable generators safer have been stymied by an oversight process that empowers manufacturers to regulate themselves, resulting in limited safety upgrades.

CPSC investigators couldn’t say whether the shut-off sensor on Johnson’s 6,250-watt Briggs & Stratton Storm Responder had activated at any point during the night, but when emergency responders arrived the next morning, the generator was in the “on” position with an empty fuel tank.

“This tragic incident exemplifies one of the limitations” of voluntary safety upgrades that have been championed by generator makers in recent years, one of the agency’s engineers wrote in a letter to the industry that accompanied the report.

Briggs & Stratton did not respond to messages requesting comment.

The Portable Generator Manufacturers’ Association has repeatedly said that automatic shut-off sensors would prevent more than 99% of deaths associated with what they call “misuse.” Joseph Harding, the group’s technical director, said in an email that the association stands by that claim in the wake of the CPSC investigation into the Louisiana case. Harding said no safety feature can prevent 100% of deaths.

“Unfortunately, the incident in Louisiana was a perfect storm of misuse operating in an outdoor location,” he said. “This tragic situation was in the ‘less than 1%’ category.”

A police report photo shows the position of a generator with the exhaust pointed toward the back door of Demetrice Johnson’s home. (Jefferson Parish Sheriff’s Office)

The findings add to growing scrutiny of an industry under pressure to make its products safer. In February, the CPSC announced that it intended to propose new mandatory regulations in its 2023 fiscal year to force stricter generator safety upgrades. And in June, a congressional committee launched an investigation, which remains open, into whether portable generator manufacturers have done enough to protect the public from carbon monoxide poisoning.

Portable generators are one of the deadliest consumer products on the market, killing about 80 people in the U.S. each year and poisoning thousands of others. The machines, used to power medical equipment and appliances during electrical outages, emit toxic levels of carbon monoxide fumes that can turn deadly when allowed to build up inside homes.

Carbon monoxide deaths caused by generators predictably follow nearly every major power outage caused by extreme weather, which scientists say is becoming more common with climate change. Generators played a role in at least 10 deaths in Texas during the February 2021 winter storm and electric grid failure, according to medical examiner investigations and incident reports. The Louisiana Department of Health reported that at least six people, including Johnson’s family, died of carbon monoxide poisoning after Hurricane Ida.

Federal regulators have known about these dangers for more than two decades, but the CPSC has not implemented mandatory safety standards that would require manufacturers to vastly reduce carbon monoxide emissions. Instead, the agency allowed the industry in 2018 to develop its own less costly solution: letting manufacturers voluntarily outfit generators with sensors that are supposed to automatically turn off the engines when carbon monoxide builds up to an unsafe level around them.

Harding, the generator industry representative, emphasized that generators should only be operated outside with the exhaust pointed away from windows and doors. He directed reporters to the industry’s public awareness campaign, which instructs users, “To protect yourself from these carbon monoxide emissions, all you have to do is take it outside.”

Photos included in the CPSC’s investigation show that the generator that killed Johnson and her children was placed inches from a back door, with the exhaust pointed toward the house. Because the generator was outside, the carbon monoxide safety sensor was unable to measure the amount of gas flowing through the back door and building up inside, a flaw in the safety mechanism that the CPSC and consumer advocates have highlighted.

Marietta Robinson, a commissioner with the CPSC from 2013-18, said the Louisiana incident “demonstrates perfectly that incorporating a shut-off switch in a portable generator instead of lowering emissions simply is not a way to protect consumers from this hidden hazard.” She noted that Johnson’s side yard was so small — only a few feet wide, according to photos — that it would have been impossible to follow generator makers’ instructions to keep the machine about 20 feet from the house.

As part of the CPSC’s announcement in February that it planned to propose new mandatory regulations, the agency released a report that studied the effectiveness of the generator industry’s voluntary safety measures. It concluded that too few manufacturers had adopted safety upgrades and that, based on a series of simulations run by the agency, carbon monoxide shut-off switches were not effective in preventing poisonings and deaths in some scenarios. Notably, the agency found, automatic shut-off sensors don’t work when users set up the machines outside with the exhaust pointed toward windows or doors — the same scenario that killed Johnson and her children.

Based on its own simulations, the agency found that automatic shut-off sensors could prevent 87% of deaths caused by generators — lower than the 99% figure promoted by the industry — while still leaving some consumers exposed to carbon monoxide levels toxic enough to require hospitalization.

CPSC staff members also tested a more stringent approach of equipping the machines with both shut-off sensors and engines that emit far less carbon monoxide and found that the combination would eliminate “nearly 100%” of generator deaths and the vast majority of hospitalizations.

In comments to the CPSC, industry officials have argued that requiring generators to emit less carbon monoxide in addition to shut-off switches “would only further exacerbate the burden on manufacturers, add unnecessary cost, and not provide any significant increase in benefit over the shutoff approach alone.”

Do You Have a Tip for ProPublica? Help Us Do Journalism.

by Stefanie Dodt, ARD German TV

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up to receive our biggest stories as soon as they’re published.

This story was co-published with ARD German TV.

A federal district judge in Washington, D.C., has ordered immigration officials to allow Mexican citizens with visas to sell their blood plasma in the U.S.

U.S. District Judge Tanya Chutkan granted a preliminary injunction overturning a policy announced last year by U.S. Customs and Border Protection officials that barred Mexican visitors from participating in what had become a multibillion-dollar business along the border.

Judge Chutkan ruled that CBP officials had “failed to consider” the extent to which blood plasma companies were relying on Mexican donors and that they had failed to adequately justify the policy.

In issuing the preliminary injunction, the judge found that the companies had shown they had a “likelihood of success” to overturn the ban if the case went to trial. She noted that the costs of opening collection centers in other regions to make up the shortfall — $2.5 million to $4 million per center — would be “substantial.’’

A spokesperson for CBP declined to say whether the agency planned to appeal the ruling, saying “this matter is still under litigation.”

U.S. officials had long acknowledged that the role of Mexican citizens in the blood plasma business along the border was a “gray area”in immigration law, with some Border Patrol agents refusing to let people enter the U.S. for donations while others allowed it.

Many people living on the Mexican side of the border hold visas that allow them to come to the United States to shop or visit relatives. After the CBP imposed its ban, they risked losing those visas if they were caught selling blood plasma.

ProPublica, ARD German TV and Searchlight New Mexico reported in 2019 that thousands of Mexicans were crossing the border to donate blood as often as twice a week, lured by bonus payments and hefty cash rewards, an economic lifeline for many of them. Mexican law bars people from selling plasma.

In Facebook posts, the Spain-based pharmaceutical company Grifols, which had teamed up with its Australia-based rival CSL Plasma to sue CBP over the ban, is already inviting Mexican donors back to the U.S. plasma centers. Some plasma donors are still warning others to be careful at the border to avoid having their visas confiscated. “As if you would tell them on the bridge where you go,” a comment in Spanish says.

Some donors interviewed by ProPublica said they would welcome the legal clarity provided by the court decision. “Finally, we can cross without fear,” said Gamaliel, a resident of Ciudad Juárez who previously told his story of being a donor. “It was always so risky.” Nevertheless, he said, he would still only go back to donate in case of a financial emergency, because of the health risks associated with frequent plasma donations. As ProPublica and ARD found, frequent plasma donation was harming the health of some Mexican citizens who relied on the system for money. Some frequent donors were underweight and showed low levels of antibodies.

In her ruling, Chutkan, who was named by then-President Barack Obama to the bench, wrote that she had considered health risks for the donors, quoting studies analyzing potential negative long-term effects of plasma donations.

She rejected the companies’ argument that Mexican blood donors were no different from couriers ferrying laboratory samples across the border, since both were “carrying a substance that ‘originates in Mexico.’”

“The comparison is unpersuasive,” Chutkan responded in her ruling. “A person is more than just a shopping cart of biological products to be bought and sold at a later date.” She said that her decision to grant a preliminary injunction reflected the crucial need for blood plasma in manufacturing lifesaving medications.

“The decision recognizes the critical importance of the need for plasma in the manufacture of life-saving therapies for hundreds of thousands of people. We are excited to welcome back Mexican donors,” a spokesperson with CSL Behring, CSL’s U.S. subsidiary, said in a statement provided to ProPublica. A Grifols spokesperson said: “The decision is good news for the patients in the U.S. and worldwide who depend on the lifesaving medicines made from plasma.”  

The pharmaceutical companies are now scrambling to reorganize their centers along the border after the ban on Mexican donors had prompted layoffs of employees in a reduction in operating hours. Grifols announced it would provide buses beginning today to transport donors from the U.S.-Mexican border to its plasma centers. The Mexican newspaper Diario Frontera posted a picture of people sitting in line in the dark in front of a plasma center in Brownsville, Texas. 

People identified by Diario Frontera as Mexican donors line up in front of a plasma center in Brownsville, Texas, on Monday. (Facebook)

In the suit challenging the ban, the companies had acknowledged for the first time the extent to which Mexicans contribute to the world’s supply of blood plasma: Up to 10% of the blood plasma collected in the U.S. — millions of liters a year — came from Mexicans who crossed the border with visas that allow brief visits for business and tourism. The U.S. accounts for approximately 60% of plasma collected worldwide. It is the only country that allows people to donate plasma as many as 104 times a year.

Do You Have a Tip for ProPublica? Help Us Do Journalism.

Correction

Sept. 27, 2022: This story originally incorrectly stated that the U.S. is the only country where donors are allowed to give plasma twice a week. At least one clinic in Canada also allows donations at that frequency.

by Vianna Davila, ProPublica and The Texas Tribune, and Davis Winkie, Military Times

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up to receive our biggest stories as soon as they’re published.

This article is co-published with The Texas Tribune, a nonprofit, nonpartisan local newsroom that informs and engages with Texans, and with Military Times, an independent news organization reporting on issues important to the U.S. military. Sign up for newsletters from The Texas Tribune and Military Times.

A Texas representative whose district includes one of the nation’s largest Army posts is calling for hearings to examine the military’s pretrial confinement system, which gives commanders the discretion to detain service members facing criminal charges ahead of trial.

Rep. Veronica Escobar, a Democrat who represents El Paso and sits on the House Armed Services Committee, said this month that an August investigation by ProPublica and The Texas Tribune raises serious questions about the use of pretrial confinement in the military. The news organizations’ first-of-its-kind analysis of nearly 8,400 Army courts-martial cases over the past decade revealed that soldiers accused of sexual assault are less than half as likely to be placed in pretrial confinement than those accused of offenses like drug use and distribution, disobeying an officer or burglary.

“Pretrial confinement had not been on my radar, so it was really eye-opening,” Escobar said of the investigation, which highlighted the case of Christian Alvarado, a private first class who avoided detention for months despite facing sexual assault accusations from multiple women. Alvarado, who has since been convicted of sexually assaulting two of those women, was a soldier at Fort Bliss. Escobar’s district includes the post.

Escobar, who is vice chair of the House Subcommittee on Military Personnel, said that during the congressional hearings she would like to explore ways to ensure all cases across the military are held to the same standard. She also said she reached out to the new commanding general of Fort Bliss to ask more about Alvarado’s case. Fort Bliss officials have declined to discuss with the news organizations why Alvarado was not initially put in pretrial confinement, saying they would not comment on internal deliberations.

The congresswoman has discussed the issue with Rep. Jackie Speier, D-Calif., who chairs the subcommittee. Speier shares Escobar’s concerns and believes pretrial confinement should be “part of the broader work of military justice reform,” her office said. A staffer in Speier’s office said the chair looks forward to learning more from military officials about pretrial confinement at a planned subcommittee hearing on Wednesday that will focus on the implementation of military justice reforms.

Army spokesperson Matt Leonard told Military Times that the rules governing pretrial confinement are “currently under revision.” He declined to provide details, referring questions to the Department of Defense. A DOD spokesperson said the military doesn’t have any “updates to announce at this time.”

News of the proposed revisions comes after publication of the ProPublica-Tribune investigation and of an article by Military Times that detailed another case in which a soldier was not placed in pretrial confinement despite multiple domestic violence allegations. The news organizations are partnering to cover aspects of the military justice system.

Under the current rules that are outlined in the Manual for Courts-Martial, commanders must determine if there’s good reason to believe a service member committed a crime and weigh whether the person is likely to flee before trial or engage in serious criminal misconduct. They must also first consider if less stringent restrictions are enough to keep service members out of trouble. Unlike in the civilian justice system, there is no bail in the military.

Rachel E. VanLandingham, a former Air Force judge advocate, said commanders should not control pretrial confinement because they are not trained as military attorneys. Commanders can override the advice of legal advisers in many situations.

“Because these commanders are F-16 pilots, these commanders are infantry officers, this is not their full-time job,” said VanLandingham, now a professor at Southwestern Law School in Los Angeles.

Congress last year overhauled much of the military justice system, stripping commanders of the power to decide whether or not to prosecute some felony-level crimes, including sexual assault. Amid pushback from military leaders, lawmakers allowed commanders to retain numerous powers, including the ability to place a service member in pretrial confinement.

Escobar had hoped to see broader reforms that would have removed commanders’ prosecutorial authority over all serious crimes, including robbery, assault and distribution of controlled substances. She said that taking away commanders’ control of pretrial confinement is “a possibility we absolutely should consider.”

“I completely understand the need for commanders to have the control that they want and have traditionally had,” Escobar said. “But I believe that we do need highly trained legal professionals who are making some of these decisions, because in my view, the status quo hasn’t worked. And in fact, it has failed soldiers.”

Despite facing sexual assault accusations from three women in 2020, Alvarado was not placed in pretrial confinement, according to the news organizations’ investigation. While being interviewed by an Army investigator about the first two allegations, Alvarado admitted that he continued to have sex with one of the women after she passed out. A month later, he sexually assaulted a third woman, which he admitted in text messages to the woman.

But Alvarado’s commanders did not place him in pretrial confinement until more women accused him of assault. A military judge later found him guilty of sexually assaulting two women and strangling one of those two women. He was sentenced to 18 years in a military prison and a dishonorable discharge. Alvarado’s case is currently under appeal. His appeals attorneys did not respond to a request for comment.

“That was shocking to me,” Escobar said. “When you have clear evidence of someone who is a repeat offender, that in and of itself, to me, merits pretrial confinement, and one of the things I do want to know about is why didn’t it happen?”

Help ProPublica and The Texas Tribune Report on the Military Justice System

by Cezary Podkul

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up to receive our biggest stories as soon as they’re published.

If you’re like most people, you’ve received a text or chat message in recent months from a stranger with an attractive profile photograph. It might open with a simple “Hi” or what seems like good-natured confusion about why your phone number seems to be in the person’s address book. But these messages are often far from accidental: They’re the first step in a process intended to steer you from a friendly chat to an online investment to, ultimately, watching your money disappear into the account of a fraudster.

“Pig butchering,” as the technique is known — the phrase alludes to the practice of fattening a hog before slaughter — originated in China, then went global during the pandemic. Today criminal syndicates target people around the world, often by forcing human trafficking victims in Southeast Asia to perpetrate the schemes against their will. ProPublica recently published an in-depth investigation of pig butchering, based on months of interviews with dozens of scam victims, former scam sweatshop workers, advocates, rescue workers, law enforcement and investigators, along with extensive documentary evidence including training manuals for scammers, chat transcripts between scammers and their targets and complaints filed with the Federal Trade Commission.

“We’ve had people from all walks of life that have been victimized in these cases and the paydays have been huge,” said Andrew Frey, a financial investigator for the Secret Service, the federal agency that is taking a lead role in combating online crime and trying to help victims recover their stolen funds.

These swindles are not only highly organized but also systematized. Here’s how the fraudsters typically go about it, including photographs, excerpts from text exchanges between scammers and targets, advice from training guides for fraudsters and police reports from pig butchering cases:

1. Create a fake identity

Pig butchers most often begin by creating a phony online persona, typically accompanied by an alluring photo (which itself might have been stolen) and images that convey a glamorous lifestyle.

This Instagram profile was reported to the Federal Trade Commission by a Florida resident who complained of losing $89,000 to a pig butchering scam. (Meta, which owns Instagram, said it’s investigating the account, whose owner didn’t respond to a request for comment.) (Screenshot blurred by ProPublica) 2. Initiate contact

Once they’ve got an online profile, fraudsters begin sending messages to people on dating or social networking sites. Alternatively, they may use WhatsApp or another messaging service and pretend to have stumbled on a “wrong number” as they contact you. (A spokesperson for Meta, which owns WhatsApp, previously told ProPublica that the company is investing “significant resources” into keeping pig-butchering scammers off its platforms.)

In December 2020, a Connecticut man received these messages on WhatsApp from a seemingly friendly stranger. He responded and eventually ended up getting tricked into two scams that cost him a total of $180,000.

[12/28/20, 12:06 AM] SCAMMER J: Long time no see, how are you recently

[12/28/20, 10:10 AM] SCAMMER J: 🙈Are you not Kevin? Sorry, I guess I added the wrong person, sorry

[12/28/20, 10:16 AM] TARGET C: Not Kevin.

[12/28/20, 10:16 AM] SCAMMER J: Sorry, I made the wrong call. Since I have many business partners, my assistant saved the wrong number, please forgive me

[12/28/20, 10:17 AM] TARGET C: No prob. What country are you calling from?

[12/28/20, 10:17 AM] SCAMMER J: I come from Hong Kong. Hong Kong is a metropolis with technology, finance and food. Have you ever been here

[12/28/20, 10:18 AM] SCAMMER J: Acquaintance is fate, where are you from

[12/28/20, 10:18 AM] TARGET C: I’m from NYC originally

[12/28/20, 10:19 AM] SCAMMER J: Your place is a very beautiful place, I went there years ago

3. Win the trust of the target

The next step is starting a conversation with a potential victim to gain their trust. The scammers often initiate benign chats about life, family and work with an eye toward mining their targets for information about their lives that they can later use to manipulate them. They’ll fabricate details about their own life that make them seem similar to you. After all, people like people who are like them.

When a Houston woman revealed that her brother was born with cerebral palsy, a crook countered with a similar-sounding tale:

[2/25/21, 6:32:38 PM] TARGET P: I have one brother that is handicapped and lives with my parents. Of course he’s coming with them for the weekend

[2/25/21, 6:35:36 PM] SCAMMER C: I see. My parents are taking care of my brother. I hope he will live well

...

[2/25/21, 6:38:49 PM] TARGET P: My brother was born with cerebral palsy

[2/25/21, 6:39:19 PM] SCAMMER C: Sorry things, but also hard for your parents

[2/25/21, 6:39:29 PM] TARGET P: He’s healthy over all but you have to do everything for him. He can’t talk, dress, or feed himself

4. Sign them up

Before long, the swindlers will pivot to a discussion of investing. They’ll make claims about their own purported investing successes, perhaps sharing screenshots of a brokerage account with gaudy numbers in it. They’ll try to convince targets to open an account at their online brokerage. Unbeknownst to the target, the brokerage is a sham, and any money deposited will go straight to the scammer. Most victims don’t figure out that last part until it’s too late.

Guides for scammers recommend touting the reliability of MetaTrader, a trading app that fraudsters use for nefarious purposes, by pointing out that the app is available in Apple’s App Store, so it must be safe. (MetaTrader did not respond to requests for comment. An Apple spokesperson said the company has shared complaints with MetaTrader’s parent company, and asserted that the parent has taken steps to respond to the complaints.)

5. Get them to put real money into the fake account

Once marks agree to learn investing tricks, the scammers will “help” them with the investment process. The fraudsters will explain how to wire money from their bank account to a crypto wallet and eventually to the fake brokerage. Typically the fraudster will ease the process by recommending a modest initial investment — which will inevitably show a gain.

A woman in Michigan became intrigued by her online boyfriend’s references to making money trading gold and offered to become his student. Two days later, he was teaching her how to get started investing in a fake brokerage accessible through MetaTrader:

[3/16/21, 4:40:00 PM] TARGET T: What are you up to right now?

[3/16/21, 5:11:42 PM] SCAMMER L: I’m reading a book

[3/16/21, 5:14:39 PM] TARGET T: What book are you reading?

[3/16/21, 5:17:02 PM] SCAMMER L: A book about investing in gold

[3/16/21, 5:19:04 PM] TARGET T: Nice. You should teach me. Make me your student.

[3/16/21, 5:20:37 PM] SCAMMER L: 🥰I don't want you to be my student. I want you to be my wife.🥰

6. “Prove” that it’s legitimate

Scammers often allay initial doubts by letting targets withdraw money once or twice to convince them the process is trustworthy. For example, fraudsters allowed a Canadian man named Sajid Ikram to withdraw 33,000 Canadian dollars, according to a statement he filed with the Royal Canadian Mounted Police. That returned money helped convince him that his investment was real. He reported ultimately losing nearly $400,000, including money borrowed from several friends.

7. Manipulate them into investing more

That’s only the beginning. Pig butchering guides offer insights on how to exploit marks’ emotional and financial vulnerabilities to manipulate them into depositing more and more funds. It starts with assurances that the investments are risk-free, then escalates into pressure to take out loans, liquidate retirement savings, even mortgage a house.

Over a period of nine days, one scammer (who called herself Jessica) escalated her pressure, pushing the target, a California man, first to use his cash on hand, then to tap his retirement savings, then to borrow money.

[11/3/21, 8:03:13 PM] TARGET Y: I just don’t want to risk

[11/3/21, 8:03:42 PM] SCAMMER J: When you need money, you can ask for it at any time

[11/3/21, 8:04:08 PM] SCAMMER J: This is not a risk, it is called maximizing profit

…

[11/8/21, 5:31:53 PM] TARGET Y: what can I do

[11/8/21, 5:32:27 PM] SCAMMER J: Your 401K can’t move?

[11/8/21, 5:32:50 PM] TARGET Y: You know how that works. Heavy penalties plus double taxation

[11/8/21, 5:32:53 PM] SCAMMER J: Then you can earn it back with a fine.

…

[11/11/21, 6:20:05 PM] SCAMMER J: Borrowing money from the bank is not a big deal, I often do

[11/11/21, 6:22:11 PM] TARGET Y: I am not ignoring you. I am trying to think

[11/11/21, 6:22:45 PM] SCAMMER J: You are a wise man, this is borrowing a chicken to lay eggs

[11/11/21, 6:23:23 PM] SCAMMER J: Really rich people use bank money to invest

8. Cut them off

Once targets reach a limit and become unwilling to deposit more funds, their seeming investment success comes to a sudden stop. Withdrawals become impossible, or they suffer a big “loss” that wipes out their entire investment.

The California man was aghast when he discovered $440,000 he’d deposited was gone. Ultimately, the swindler persuaded him to invest another $600,000, which also disappeared into the swindler’s account.

[11/18/21, 11:59:16 AM] TARGET Y: I lost all my money

[11/18/21, 11:59:18 AM] SCAMMER J: If the principal is not enough, it cannot be supported to the profit point.

[11/18/21, 11:59:34 AM] SCAMMER J: Don’t worry,

[11/18/21, 11:59:46 AM] TARGET Y: I am negative $480k

[11/18/21, 12:00:01 PM] SCAMMER J: Prepare the funds and earn them back.

[11/18/21, 12:00:12 PM] TARGET Y: I don’t have any money or funds to prepare

[11/18/21, 12:00:20 PM] TARGET Y: That’s all I have!!!!!!!!!!!!

9. Use their desperation to your advantage

Scammers will then turn the screws of manipulation tighter by telling victims there’s a potential solution: If they deposit more cash into the brokerage, they can regain what they lost. Sometimes, the claim is that the investment is successful — but there’s a “tax problem” that requires paying additional funds equal to, say, 20% of their total account value. If the victim pays, the scammer will claim that new obstacles have arisen that require paying new fees.

No matter how much targets pay, it’s never enough, as detailed in the FTC complaint excerpted below, which was filed by a pig butchering victim in Maryland. This person lost almost $1.4 million, in part because the person kept meeting scammers’ demands to pay taxes and various fees to get their money back:

“Once the trading has ended, I applied to withdrawal my money and profit from the website. The broker asked me to pay a tax on the profit of 88,587.90 usd on 8162021, this amount was wired again through Bank of America into a foreign account in Hong Kong. Another request for me to pay security deposit on my profits which was 83,950.00 usd wired out to a different foreign account in Hong Kong once again. The broker asked for a bank and withdrawal processing fee of 27,983.34 usd again was wired out to a different foreign account in Hong Kong. The very last wire was for expediting the withdrawal and the platform asked for 55,966.60 usd wired out to Hong Kong. At this point I already had to much money in the platform so I kept giving in.”

10. Taunt and depart

Once the targets are aware that they’ve been swindled, the fraudsters often insult or taunt them. They soon go silent, and the websites of their phony brokerages stop working. Then they relaunch a new website under a different URL and restart the process with other targets.

After nearly four months of chatting and $30,000 in losses for the Michigan victim, her scammer seemed to revel in unveiling the financial — and emotional — deception:

[7/1/21, 3:25:31 PM] SCAMMER L: I’m a liar, too.

[7/1/21, 3:25:42 PM] TARGET T: What do you mean you are a liar?

[7/1/21, 3:25:58 PM] SCAMMER L: But I am very kind, I only cheated you out of 30K, thank you for 30K

[7/1/21, 3:26:16 PM] TARGET T: Wow.

[7/1/21, 3:26:55 PM] TARGET T: ‎You deleted this message.

[7/1/21, 3:27:23 PM] TARGET T: You don’t really love me? We are not getting married?

[7/1/21, 3:27:44 PM] SCAMMER L: Surprise or surprise. I’m not surprised.

What to Do If You’ve Been Scammed

If you’ve been victimized, report the crime to your bank and law enforcement — the FBI, the Secret Service and local police — as quickly as possible. The longer you wait, the harder it is for your bank to reverse any fraudulent transactions and for law enforcement to trace, freeze or seize stolen funds. “We are definitely going to be more successful if you immediately report,” said Erin West, deputy district attorney at the Santa Clara County District Attorney’s office, which has had some success seizing assets linked to pig butchering scams.

by Mike Smith, The Times-Picayune | The Advocate

This article was produced by The Times-Picayune | The Advocate, a member of ProPublica’s Local Reporting Network. Sign up for Dispatches to get stories like this one as soon as they are published.

LAKE CHARLES, La. — Hilda Brown ambled down a wooden walkway with the help of a cane and gingerly took a seat at a table between her badly damaged house and a FEMA trailer. It had been more than a year and a half since Hurricane Laura, but the then-64-year-old widow still didn’t know where to turn.

She didn’t have the money to finish repairs on the home, which she couldn’t afford to insure. The Federal Emergency Management Agency had lent her the trailer as a temporary solution, but the deadline to return it was approaching, and she was facing the possibility of being forced to pay steep rent to keep living in it.

“It’s been hard,” Brown, who lives on Social Security benefits, told The Times-Picayune | The Advocate in April. “Everything is just bam, bam, one thing after another.”

Since Laura struck in August 2020, Lake Charles, a small city in Louisiana’s industrial southwestern corner, has served to spotlight the shortcomings in America’s system for helping communities rebuild in the wake of catastrophe.

The process of distributing key components of long-term aid — meant to help not in the first frantic days after a storm but to rebuild housing and infrastructure — is unpredictable, often caught up in battles over federal spending. As a result, the country essentially forgot one of its cities.

Lake Charles, with a population of 85,000, was Louisiana’s fifth-largest city before it was hit by a series of storms. The region’s economy was booming when Laura hit, thanks to the billions being poured into petrochemical plants and gas-export terminals, as well as a steady stream of Texans crossing the border to plunk down cash at three casinos.

But it took more than a year and a half after Laura for the federal government to approve what local officials deemed adequate long-term recovery assistance for the region. And more than two years after the storm, the first federal check from the Department of Housing and Urban Development has still not arrived.

The money has been entangled in the government’s byzantine process of doling out such long-term recovery funds, a process that can drag on for many months and get buffeted by the political priorities of congressional leadership and the White House.

In addition to the wait, it also means that communities in the wake of disaster do not know how much Congress will eventually provide, making recovery planning difficult. Appropriations can vary widely from disaster to disaster.

“[T]here’s really an unevenness and almost an unfairness in these recovery dollars,” said Andy Winkler of the Bipartisan Policy Center, a Washington think tank that has examined disaster assistance. The kinds of delays that Lake Charles experienced are nearly certain to be repeated — perhaps in Louisiana, perhaps somewhere else.

In a time of climate change and intensifying storms, Lake Charles’ struggles have been particularly acute. The city faced four natural disasters in less than a year, beginning with Laura, one of the most powerful hurricanes to ever hit Louisiana. Hurricane Delta followed six weeks later, then a severe winter storm in February 2021 and heavy flooding three months after that — all during the pandemic. One study determined that around half of all homes in Calcasieu Parish, where Lake Charles is located, were damaged by the two hurricanes.

Over the last two years, houses, apartments, public housing complexes and businesses have sat in ruins, exacerbating an affordable housing shortage. The region, which had one of the country’s fastest-growing economies before 2020, has since seen one of the nation’s largest drops in population.

Local officials resorted to pleading with Congress and the nation for help as months passed without action. Lake Charles Mayor Nic Hunter called it “a humanitarian crisis right here on American soil.”

Both President Donald Trump and President Joe Biden visited, but still southwest Louisiana has waited.

There have been proposals to fix the system, but the nation has been caught up in other challenges and priorities, and the two political parties rarely see eye to eye.

Lake Charles Mayor Nic Hunter stands outside City Hall in May 2022. “We were let down at every turn,” he said of long-term disaster relief. (Leslie Westbrook/The Times-Picayune | The Advocate)

Hunter, who rode out Laura in Lake Charles, soon realized that the hurricane’s destruction “was basically our worst nightmare unfolding.”

“Throughout the process, I kept getting pats on the back saying: ‘Don’t worry, it’s coming. Don’t worry, it’s coming,’” Hunter said in April of assurances of long-term disaster relief. “We were let down at every turn.”

He spoke from his 10th-floor City Hall office, finally repaired a year and a half after Laura. Nearby, the housing authority was still operating out of a modular building, and Lake Charles’ biggest office tower remained boarded up and fenced off.

Recently, Hilda Brown finally received more help for her house repairs from volunteers and the disaster rebuilding nonprofit SBP, as well as a reduction on her trailer rent to $50 per month. She expected to move back into the house soon, and hoped that the federal money set to arrive in the near future could help with the remaining relatively minor work.

“It could be any day,” she said Wednesday of moving back into her house. “I’m so excited.”

“Difficult to Predict”

In the aftermath of a natural disaster, immediate federal aid is designed to flow in quickly. Federal agencies mobilize to help with search and rescue, food and water distribution, temporary housing and other tasks.

Since the two hurricanes hit southwest Louisiana in 2020, the state has drawn more than $2 billion in help from FEMA and in low-interest loans from the U.S. Small Business Administration, which are among the early disaster responders.

But that money falls short of addressing the region’s long-term recovery needs.

That’s where longer-term federal aid — normally distributed through what bureaucrats call CDBG-DR, or Community Development Block Grants-Disaster Recovery — comes into play. It is meant to address serious housing, infrastructure and economic development needs not covered by FEMA and insurance. Typically 70% of the funds must benefit low- to moderate-income residents, and families’ homes must be severely damaged to qualify.

But reporting by The Times-Picayune | The Advocate over more than a year in Lake Charles found significant flaws in that system.

To start with, there is no set time frame for when, or whether, the aid will be approved, leaving it to Congress and the White House. Money can be appropriated for communities in mere days, or they can wait for a year or more.

Consider: It took just 15 days for Congress to appropriate long-term aid intended for victims after Hurricane Harvey; it took 26 times as long for the federal government to agree on an aid package that included Laura.

How Long They Waited

The amount of time between a disaster declaration and the appropriation of long-term relief dollars varies wildly. Here is a sampling of those time frames:

• Hurricane Harvey (2017): 15 days
• Hurricane Ida (2021): 32 days
• Hurricane Sandy (2012): 93 days
• Houston ice storm (2021): 223 days
• Hurricane Sally (2020): 375 days
• Hurricane Laura (2020): 398 days
• California wildfires (2020): 404 days

Source: Bipartisan Policy Center and The Times-Picayune | The Advocate

Then, once Congress does appropriate money, it goes through a complex process to actually distribute it to communities. First, HUD must calculate how much each state, local government or Native American tribe will receive. Once those allocations are announced, HUD publishes rules for how the money can be spent. This prompts states to come up with their own spending plans, solicit and evaluate public comments, and ultimately submit those plans to HUD for approval. Only then can the money begin to flow.

This process can get caught up in budget politics and allegations of political interference. Puerto Rico, for example, faced delays in the aid disbursement process after hurricanes Irma and Maria in 2017 because of unusual requirements put in place after the Trump White House accused the territory’s government of being corrupt, as The Washington Post reported.

Southwest Louisiana’s congressman, Clay Higgins, a far-right Trump supporter, faced accusations of failing to do enough to have the money approved, particularly after he didn’t show up for Biden’s visit to Lake Charles in May 2021. He has defended his record.

“It always gets caught up in budget politics,” said Stan Gimont, who once oversaw HUD’s CDBG programs and is now with Hagerty Consulting. “And it just is very difficult to predict what event or what critical mass of events will move that needle for Congress to act.”

A HUD-commissioned study from 2019 found that disaster recovery grants took an average of 4.7 years to be spent, measured from the date of the disaster declaration. The study looked at disasters between 2005 and 2015.

With Laura, it took Congress 13 months after the storm to approve funding — an unusually slow process for post-hurricane long-term aid. A month later, in October 2021, HUD announced an initial allocation of around $600 million for southwest Louisiana. It then took three more months to release the rules for spending it.

It would take until March 2022 for the region to be approved for an amount Hunter said he considered equitable, $1.06 billion.

The risks of these delays are obvious. Blight worsens. Affordable housing is in short supply. Those who were already struggling to get by and who have temporarily relocated elsewhere may never return. A community survey done in the Houston area after 2017’s Hurricane Harvey showed 18% of the unsheltered homeless people who responded said they were homeless because of the storm.

Additionally, FEMA’s post-disaster programs providing trailers end after 18 months, and delays mean that people can be cut off from that help well before long-term housing aid begins to flow. That deadline can be extended in certain circumstances — as happened in southwest Louisiana — but those living in the trailers can also be asked to begin paying rent after the 18 months are up.

“People have enough trouble accessing the FEMA program,” said Sarah Saadian of the National Low Income Housing Coalition, which has closely studied post-disaster programs. But when those short-term programs end, she said, “we end up seeing people who are the lowest-income, the most marginalized, lose their housing again. They face displacement, and then in the worst cases are pushed into homelessness.”

U.S. Rep. Garret Graves, R-Baton Rouge, whose district has been repeatedly hit by disaster, has focused his criticism on the convoluted process of disbursing the aid.

“You’ve got a process that is so ridiculous,” said Graves. “You can either spend all this time and money trying to get it perfect, or you can lean a little bit more towards getting the money out the door quickly.”

No Cash for Repairs

Allison and Charles Ned were among those who fell through the cracks.

During Hurricane Laura, they had squeezed into a friend’s house with three other families near Beaumont, Texas, about an hour away. They drove back to Lake Charles the next day, navigating trees and downed power lines — an obstacle course laid down by 150 mph winds.

For 20 years, they had rented a house in Lake Charles, where Louisiana Cajun and Creole traditions mix with Texas cowboy culture.

Allison and Charles Ned stand at the entrance to their FEMA trailer. (Leslie Westbrook/The Times-Picayune | The Advocate)

Allison, 57, had a cleaning service. Charles, 58, worked for a house-moving company and is now a maintenance worker for Calcasieu Parish schools. He prides himself on self-sufficiency and sports a belt buckle he won in a calf-roping competition.

When they made it home after Laura, the couple found their roof had been lifted up and set back down by the storm. It was leaking everywhere. The Neds had raised their two children there and kept two horses in the backyard along with their Akita, Champ. Charles, normally stoic, cried as he took in the damage.

For six months, they stayed at a friend’s house in Jennings, 40 minutes away. FEMA eventually provided a trailer, which they set up in the yard next to their damaged house.

Their landlord battled his insurer and decided to sell, Allison said. The Neds bought the house as is, still unrepaired and unlivable — and they didn’t have the cash to fix it.

Allison Ned walks into her house while it was still under repair in May 2022. (Leslie Westbrook/The Times-Picayune | The Advocate)

But each time they asked for help in rebuilding, they were rejected because the system was not set up to help people in their unique situation. They did not qualify because they didn’t own the home at the time of the storm, even though they’d lived there for two decades. Allison said she wrote to the governor hoping an exception could be made, but received no response.

Charles said: “All you could do was pray to God.” He did what he could, using skills from his previous job to make the house level again.

Around 18 months after Laura, the Neds were told they would soon have to start paying rent for their FEMA trailer this spring. Under federal rules, they would have to move out of it by the end of October.

It took until early 2022 for the Neds to get a hand to rebuild, and when they did, it didn’t come from the government. Instead, the SBP nonprofit, a recovery organization founded after Hurricane Katrina, along with volunteers from organizations including the Fuller Center Disaster ReBuilders, came to their aid.

But as trying as that period was, they were on their own land, unlike others caught up in the affordable housing shortage, which long-term disaster relief can help solve.

Michael Blaney, 69, had been living in a damaged and decrepit public-housing complex for the elderly. The housing authority did not want people to live in those conditions and threatened some residents, including Blaney, with eviction, although it also attempted to find alternative housing for them. The eviction notice was eventually lifted, and Blaney remained in the complex despite the authority’s concerns; he didn’t find a new place to live until almost two years after Laura. He is now living in a privately run building for the elderly and disabled in downtown Lake Charles that didn’t reopen until April.

Many others ended up leaving the city entirely. Estimates from July 2021 show the metro area, which includes Calcasieu and Cameron parishes, had the biggest population drop in the nation that year. The two parishes combined lost 5.3% of their residents — around 11,700 people.

Michael Blaney sits outside the public housing complex where he lives in Lake Charles in March 2022. (Leslie Westbrook/The Times-Picayune | The Advocate) Finding a Permanent Solution

Criticism of the unpredictability of long-term disaster relief has come from a range of politicians and advocates seeking to change the system.

Proposed solutions often revolve around one key component: a permanent authorization of funding.

“Because it’s not a permanently authorized program, you really don’t know if it’s going to be provided for any given event, and even if it is provided, what its timing might be,” said Gimont, the former HUD official who also serves on a disaster response reform task force with the Bipartisan Policy Center.

In one proposal with bipartisan support, a fund would be made permanently available for federal block grants, eliminating the need for congressional authorization. That money could serve as a down payment, and if more is needed, Congress could step in.

The Reforming Disaster Recovery Act, sponsored by Sen. Brian Schatz, D-Hawaii, includes a provision for funding the block grants. Sen. Bill Cassidy, R-Louisiana, is among the co-sponsors.

Graves, the Louisiana congressman, has offered his own proposal, co-sponsored with U.S. Delegate Stacey Plaskett, D-Virgin Islands. He, too, wants a pre-authorized set of disaster funds to be available, but favors having FEMA handle it rather than using the current block grant system.

Asked about the delays, HUD cites the lack of a readily available source of funding.

“The absence of a permanently authorized CDBG-DR program contributes to much of the delay as does the continued absence of a standing source of CDBG-DR funds, which are now only made available as supplemental appropriations rather than as an annual part of the HUD budget,” a spokesperson said in response to questions.

After such a long struggle, the community seems to have turned a corner. A number of popular restaurants and other businesses have managed to reopen. Even the assurance that long-term federal relief was finally coming seemed to provide a jolt of optimism.

Still, it’s not hard to find destruction and blue tarps in the city. Hunter, who has lived in Lake Charles his entire life, believes a bigger city with more political influence would have seen a faster response.

“Had Hurricane Laura hit another major city with more congressional support, with more population, I think things would have been very different,” he said. “And that is what I hope changes in the future.”

There should be a formula for disaster funding, he said, “versus simply the whims and the wills of Congress or the president. It shouldn’t be such a political process.”

by Moe Clark, photography by Eli Imadali

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up for Dispatches, a newsletter that spotlights wrongdoing around the country, to receive our stories in your inbox every week.

As Andrew Montano’s parole date neared, he had two options: spend another year in prison or finish his sentence at a state-funded halfway house.

Fellow inmates cautioned him against entering Colorado’s community corrections system, saying it was overly punitive and too often a ticket back to prison. But after nearly 13 years behind bars — his entire adult life — Montano’s desire to embrace his long-term partner and start a career overshadowed those warnings.

In September 2019, he was accepted into a halfway house outside Denver, then owned by the private-prison operator CoreCivic. The rules were strict. But Montano, 35, who was sent to prison in 2007 for assault with a deadly weapon, a crime he committed as a juvenile, was determined to stay on track.

Four months later, during a routine meeting, Montano’s case manager asked if he’d ever been to a location without prior approval from the facility, a violation of the rules. He answered truthfully: On three occasions, while waiting to catch a bus for the hourlong commute back to the halfway house, he’d used a gas station bathroom.

He left thinking the exchange didn’t have much significance. His prior write-ups had been for minor infractions: failing to make his bed, having a box of raspberries in his room and forgetting to sign off on mandatory chores. But later that day, he was told he couldn’t leave the facility, even for work or mandatory classes. For more than a week, he waited.

Montano near his home. He was released directly onto parole from prison.

“Without even a hearing from the halfway house, without being able to even talk to anybody about it, the cops just came and picked me up,” said Montano, who was sent back to prison after a parole hearing. He was released directly onto parole nearly a year later, in December 2020.

“They want us to change, they want us to grow, they want us to learn, they want us to have integrity and to be honest and truthful and a member of society, but we don’t have the chance to be able to (do) that.”

CoreCivic declined to make a representative available for an interview and did not respond to questions sent by email. In a written statement, a spokesperson said, “The staff and leadership teams at our residential reentry centers in Colorado strictly adhere to the policies and standards established by” the Colorado Division of Criminal Justice. “We see our purpose as making sure those in our care are better prepared to succeed no matter where life may take them next. We’re proud of our longstanding track record of delivering high-quality and meaningful residential reentry programs.”

When Colorado formed its community corrections system in 1974, it intended to address prison overcrowding and rehabilitate people in the justice system by providing addiction treatment, job training and other services. Officials saw it as sound fiscal and criminal justice policy: Halfway houses are cheaper to run than prisons, and more assistance would reduce recidivism, meaning fewer people would land back behind bars.

But the reality is more of the people who pass through Colorado’s halfway houses end up incarcerated than rehabilitated. Of every 100 people in a halfway house, only two will be reincarcerated for a new crime, while 26 will fail and likely end up behind bars for technical violations and 14 for running away from a facility.

Among the 57 who do graduate, 22 will be back in the criminal justice system within two years, data since 2009 shows. (Due to rounding, these scenarios add up to 99 people, not 100.)

Colorado’s overall recidivism rate — defined by the Department of Corrections as returning to prison within three years — is 50%, one of the worst in the country, according to a nationwide analysis of recidivism published in 2018 by the Virginia Department of Corrections.

Colorado Halfway Houses’ High Failure Rate

For every 100 people in a halfway house...

Source: Colorado Division of Criminal Justice. Notes: Includes fiscal years 2009-2019. Later years are excluded because they don’t have two years of follow-up data. Figures are rounded.

“Is that really the results that we are OK with?” asked Terri Hurst, health justice coordinator for the Colorado Criminal Justice Reform Coalition, a nonprofit group that advocates for policies that decrease reliance on the criminal justice system. “These programs have been operating for years without any sort of transparency or accountability or scrutiny.”

Katie Ruske, manager of Colorado’s Office of Community Corrections within the Division of Criminal Justice, which sets standards for halfway houses and audits for compliance, said her office is wary of requiring certain program completion rates for fear it would come at a cost to public safety. “At times it is in the best interest of other individuals in the program, victims and the community that a person be returned to custody.”

Montano at home with his newborn boy

ProPublica spoke with nearly 50 current and former halfway house residents, staff members and experts who attributed Colorado halfway houses’ high rate of failure to numerous and often pointless rules, which can be more punitive than those in prison; a scarcity of employment training; addiction treatment programs that are rudimentary or too brief; and financial costs imposed by the facilities that can sink residents into debt.

As a result, Coloradans are, in effect, paying twice: first to fund residents’ time in halfway houses and again when they end up behind bars. From 2019 to 2021, more than 2,000 people were sent from a halfway house to prison, where the cost to incarcerate them more than doubled, to $56,000 a year, according to a Department of Corrections spokesperson.

“By and large,” said state Rep. Leslie Herod, a Denver Democrat, halfway houses “have become just another place to warehouse people, to profit off of people’s misfortune.”

Michael Anthony Martinez attends church in March after being released from prison, where he was sent from a halfway house. Since his release, he has been sent to another halfway house to complete an addiction treatment program and then transferred back to the facility that sent him to prison. John Sherman greets his sister’s dog at home in Denver. The rules of the halfway house he graduated from in January 2021 gave him “intense dread” he’d be sent back to prison. “The Insanity of This System”

There are three ways people typically end up in Colorado’s 26 state-funded halfway houses. Diversion clients are sentenced by a judge to community corrections in lieu of jail or prison. Transition clients apply to finish their prison sentence there. And some former prisoners are required to complete a halfway house program as a condition of parole.

The time it takes to complete a program varies, but it averages around eight months. And afterward, many diversion clients remain on “non-residential status” for an additional nine months on average, according to a 2018 state report, meaning they no longer live at the facility but report to their case manager and are subject to drug and alcohol testing. (The state’s data excludes those who participate in lengthier specialty treatment programs.)

Over the past three years, Colorado averaged about 6,000 halfway house stays annually. A majority of Colorado’s halfway houses are owned by companies specializing in detention and community-based supervision. Three — CoreCivic, Advantage Treatment Center Inc. and Intervention Inc. — operate 15 of the 26 state-funded facilities.

This fiscal year, community corrections will receive $87.7 million in state funding — or nearly 16% of the state’s public safety budget — to cover operational costs and treatment services. The facilities aren’t required to report in detail to the state oversight agency or lawmakers what they do with the funding.

“There are, at times, various things that we require them to show us, but there isn’t one standard,” Ruske said. “We’re not a financial auditing entity.” Every three years operators are required to hire a public accountant and submit a financial report to the state, but a majority haven’t filed a report since 2017.

A halfway house run by Intervention Community Corrections Services, under the umbrella organization Intervention Inc., in Lakewood, Colorado. Residents at ICCS facilities are prohibited from possessing candles, gift cards, debit cards or cash, according to a client manual.

Until July, facilities also charged rent to residents, about $510 per month — $17 a day — for a room shared with three to 24 people. During fiscal 2020, facilities collected approximately $15 million in rent.

Although state taxpayers provide most of their funding, the state delegates much of the oversight to local community corrections boards that vary in their makeup, protocols and oversight practices. Board members are typically appointed by county commissioners and include mayors, parole officers, law enforcement, prosecutors and judges.

This local control gives facilities broad authority to establish rules and program requirements for individual halfway houses, creating a patchwork of policies across the state.

A special government audit in 2021 spurred by citizen complaints found that a Colorado Springs facility run by ComCor Inc. had an “excessive” number of rules that residents had to abide by, some “disrespectful in nature.”

Mark Wester, executive director of ComCor, said the facility has “moved to a more supportive therapeutic approach to client care” in response to the audit. He said it also established a committee to review rules, as well as the sanctions it imposes.

Residents at halfway houses owned by the nonprofit Intervention Community Corrections Services are prohibited from possessing candles, gift cards, debit cards or cash, according to a client manual. And in interviews, former residents said they’re barred from wearing tank tops, having cellphones or driving. Failing to “follow a direct order by staff” and returning to the facility a few minutes late are also violations, according to the manual.

The current and former residents of some Colorado halfway houses who spoke to ProPublica acknowledged they need to account for their whereabouts but said such requirements are sometimes impossible to comply with, especially when relying on public transportation.

“If you’re a minute late, they’ll try to write you up. But sometimes due to weather or the bus or something like that, you might not be able to make it if those are unforeseen circumstances,” Montano said. “Their response is, ‘Well, you better start running.’”

Punishment for violating the rules also varies, from added chores to a return to prison.

Sign up for Dispatches, a ProPublica newsletter about wrongdoing in America.

A person can be expelled from a program for “any reason or for no reason at all,” a 2012 law says. If a diversion client pleads not guilty to a violation, a hearing is held to determine their punishment. But, as noted by numerous state audits, documentation of such hearings — and the outcomes — are inconsistent and lack detail.

ProPublica obtained through a public records request incident reports from all Colorado halfway houses in operation during a three-year period. They showed residents facing serious discipline for expressing suicidal thoughts, not reporting a traffic ticket and driving without permission. If they’re kicked out of a program, most end up incarcerated. One resident was terminated and sent to jail because he was suspected of stashing a gun in the woods even though police didn’t have enough evidence to arrest him.

John Sherman went 34 years in prison without a write-up. But that changed within days of being released to a for-profit halfway house in Denver.

Sherman

On his second day at CoreCivic Dahlia, the 66-year-old was given permission to travel to Walmart to buy clothing. When he entered the store, more than a dozen family members were waiting to surprise him. Loved ones pulled him in for hugs, stuffing money and gifts into his hands. His sister interrupted the reunion, frantically reminding Sherman that he needed to report his arrival at the store. But the customer service desk wouldn’t let him use its landline, and he missed his deadline by a few minutes.

Strict protocols track residents’ movements. Their route must be approved with a deadline to reach their destination. In some halfway houses, residents aren’t allowed to have a cellphone, forcing them to find a landline to report back to the facility.

In the weeks that followed, Sherman, whose sentence was shortened by former Gov. John Hickenlooper, received seven more write-ups for failing to call the facility. He said he lived with “intense dread that I’m going back to prison.”

Research shows someone with Sherman’s track record is positioned for success after prison: He had family support, didn’t struggle with substance abuse and had a place to live. He would ultimately succeed, but he said it was only because he stayed away from the facility and its rules as much as he was allowed.

“The halfway house doesn’t care if you leave or succeed,” said Sherman, who completed his program in January 2021. “Somebody’s gonna fill that bed no matter what.”

Roger Jarjoura, a researcher and senior adviser with the American Institutes for Research’s National Reentry Resources Center, said structure and supervision are key aspects of any reentry program, but arbitrary rules coupled with excessive supervision increase the likelihood a person will fail. The NRRC, which is funded by the U.S. Department of Justice’s Bureau of Justice Assistance, provides technical assistance to reentry organizations and local and state governments to improve recidivism rates.

“That’s the insanity of this system: People who have already shown that they are not great with structure or rules have to follow way more rules than a typical person would,” Jarjoura said. ​​“Supervision can be effective when it’s connected to the right kinds of treatment and the right kinds of support.”

Sherman, an artist who became known for murals in prison, works on a piece at home. Sherman spends time with family on Easter Sunday.

Kathryn Otten, who until 2016 directly oversaw two halfway houses as division director of Justice Services in Jefferson County, said the facilities’ punitive policies set residents up for failure.

The Jefferson County facilities — one of which recently closed — are run by Intervention Community Corrections Services, under the umbrella organization Intervention Inc., which currently operates four other halfway houses in Colorado.

Otten recalled that after a bathroom was defaced at one facility, “instead of getting to the bottom of it to figure out who did it,” staff put all residents on lockdown. “They made everyone stay in, miss work, miss everything,” she said. The facility, which like all Colorado halfway houses controls residents’ finances, then took money from everyone for repairs.

“We stepped in at that point and said, ‘No, you can’t do that,’ and ensured that residents got their money back,” said Otten, who in 2016 became the county’s director of housing, homelessness and integration.

ICCS’ executive director, Brian Hulse, declined to be interviewed or respond to questions sent by email. In a written statement, he said: “We use evidence-based and evidence informed practices to maximize desired outcomes and positively impact the lives of our clients. We are proud to serve our community in this capacity.”

Wendy Sawyer, research director for the Prison Policy Initiative, a research and advocacy organization, said it’s not just the number of rules or how arbitrary they can be, “but the way that the person who’s supervising them chooses to apply those rules.” She added, “There’s a ton of discretion.”

Broderrick Rimes, a former ComCor security manager who left the company in July 2021, said the enforcement of rules was often dictated by what was financially beneficial to the facility. When clients broke a rule, staff contacted the billing department to see if they had paid their rent.

“If they’re up to date on their rent, they’ll never get sent back to jail or prison,” he said. But if they owed money they would be kicked out and replaced by someone who could pay. “That is the mentality of community corrections.”

A 2017 audit of ComCor by the Office of Community Corrections found “a clear pattern of inappropriate application of serious sanctions to minor behavioral violations, especially those related to financial matters.”

Wester, who noted the audit occurred before he joined ComCor in 2019, denied that staff check whether clients have paid their rent before disciplining them.

A halfway house, operated by ComCor Inc., in Colorado Springs, Colorado Staying Clean

One in five residents of Colorado’s halfway houses is completing a sentence related to substance abuse, according to the state’s three most recent annual reports.

Nearly a dozen current and former residents and staff members from various facilities told ProPublica that the treatment they received was rudimentary, redundant and sometimes overseen by inexperienced staffers. Residents receive treatment of varying length and intensity, from drug and alcohol classes sometimes taught by halfway house staff to 90-day intensive residential treatment.

Shawn Brndiar, a licensed clinical social worker and addiction counselor in Centennial, Colorado, said the three-month addiction programs are not long enough. Effective programs are typically longer than a year, he said.

“Addiction doesn’t start in a vacuum. It’s not like just one day, somebody woke up and said, ‘I’m going to start shooting heroin,’” said Brndiar, who has been an addiction counselor for nine years. “There’s a whole history of things that happened to this person.”

An analysis by the MacArthur Foundation and the Pew Charitable Trusts found that the intensive residential treatment programs in Colorado’s community corrections system were an inefficient use of taxpayer money and “not very effective at reducing recidivism.” They received the lowest score among 21 criminal justice programs that were evaluated, including those offered in jail or prison.

Over the past five years, 66% of residents enrolled in intensive residential treatment programs graduated.

Michael Anthony Martinez, 29, said he benefited from skills he was taught to manage anger and codependent relationships while in intensive residential treatment in 2020 and 2021. But he said drugs were easily accessible at the Sterling facility owned by Advantage Treatment Centers and he failed out.

ATC did not respond to requests for comment.

The most recent audit of the facility, in 2019, found it was out of compliance for controlling contraband entering the facility.

After failing multiple drug tests, Martinez, who was a diversion client, was sent to prison.

Martinez and other residents of a sober living home relax after a game of football at a park. Martinez shows a tattoo that reads, “Family: we may not have it all together, but together we have it all.”

He spent 10 months in prison before being released on parole. After a few months at a sober living home and another 90-day intensive drug treatment program — his third — at a different halfway house, he graduated in July 2022 but was transferred back to the Sterling facility, according to state documents.

“I’m ready to be (a) successful man and show everybody that I can do something right,” he said. “Because this is just sickening. In and out, in and out.”

Requiring people to repeat ineffective programs is a way for companies to boost their profits, Brndiar said.

“It’s a business,” said Brndiar, who is a professor at Metropolitan State University of Denver. “There’s an incentive to keep people in this cycle over and over and over and over again because either way they’re going to get money.”

While some programs provide financial assistance to help with program fees, many residents are required to reimburse the facility once they obtain employment. In fiscal year 2021, 2,521 residents graduated owing on average $1,076, according to the state’s annual report.

At some ICCS facilities, residents receive treatment for mental health and addiction provided by a related company, Behavioral Treatment Services, and are sometimes required to pay out of pocket for those services, according to Otten. The CEO of ICCS’ parent company, Intervention Inc., Kelly Sengenberger, is on the BTS board of directors and was listed as BTS’ owner in state documents from 2017 to 2019. The companies listed the same business address with the state.

Ruske, of the state’s oversight agency, told ProPublica that the financial relationship was not disclosed to the state. ICCS’ contract with the state requires disclosure of potential conflicts of interest. Neither Sengenberger nor BTS responded to emailed questions.

Alycia Samuelson completed a three-year residential halfway house program in 2014 at ICCS. She said part of her treatment was overseen by interns doing outpatient rounds who were ill-equipped to help people who, in her case, had experienced decades of substance abuse and trauma.

“All I’d ever known was addiction,” she told ProPublica. “My mom had given me my first line of cocaine when I was 13. So I didn’t know a lot of the skills for how to be a successful person.”

Alycia Samuelson gets ready for work in her Denver apartment, which is covered in affirmations.

Samuelson ultimately graduated from the residential program and, after being homeless for more than a decade, moved into her own apartment. And for the first time in her adult life, she wasn’t using meth. But less than a year later, she relapsed and failed a drug test. The facility brought her back into the program then sent her to prison.

“Addiction is a relapsing disease,” Brndiar said. “People relapse, and to make your first move the most severe punishment? It only re-traumatizes the person going through this process.”

Recovery from drug addiction is a long-term process and frequently requires rounds of treatment, according to the National Institute on Drug Abuse. As with other chronic illnesses, relapses can occur and signal a need for additional treatment, research shows.

When Samuelson was released from prison, she was homeless, fell back into drug use and ended up in another halfway house in 2018. Six months into her program she said she was suicidal. In a matter of weeks, her husband left her, her son was diagnosed with colon cancer and a friend overdosed in the facility and died. Samuelson said staff ignored her struggles so she ran away.

“I said I felt like I’m gonna take my life and I thought they were going to call the ambulance and they didn’t,” said Samuelson, 47, who now lives in Denver. “I literally ran from the halfway house to save myself.”

A 2020 law reduced the charge for escaping from a halfway house to a misdemeanor, instead of a felony, for those serving time for nonviolent crimes, meaning Samuelson wasn’t returned to prison. In 2021, 910 halfway house stays, or 24%, ended when the person ran away from a facility, a sharp increase from 2012 when the rate was 12%, according to state data.

Those with a diagnosis of mental health issues are significantly more likely to fail out of community corrections, according to a 2018 analysis by the Colorado Division of Criminal Justice’s Office of Research and Statistics. Only 48% graduated between 2014 and 2016, compared with 61% of those without such a diagnosis. But a diagnosis had little to no effect on whether the person committed a new crime in the future, the report found.

“If all they have are other people who are part of the criminal justice system and underpaid, underqualified people providing treatment, they’re not going to (get what) they need,” Brndiar said.

Finding a Job

Samuelson receives job training at a 1950s diner. State audits have shown that few halfway houses are meeting quality standards for employment services despite the benefits to the long-term success of residents.

Employment is key in determining whether someone stays out of prison, according to academic studies and reports by Colorado’s Division of Criminal Justice. Yet state audits show employment services are some of the least-available resources in state halfway houses.

People entering the facilities face challenges getting a job. Some haven’t worked in decades. Others were juveniles when they entered the corrections system and may never have held a job. A criminal record, lack of transportation, addiction and mental health issues can further complicate finding employment.

“If they don’t have a means to support themselves, they’re going to figure out how to do that,” said Jarjoura, the recidivism expert. “People don’t choose crime. That’s how they learn to survive.”

Limited scope audits by the Office of Community Corrections show that of the facilities where employment services were assessed from 2015 to 2018, only two were in full compliance with state standards. The agency stopped evaluating the quality of employment services in 2018, after a new manager took over OCC.

“The future strategic goals of the office include developing processes for reviewing compliance with the additional Colorado Community Corrections Standards not currently reviewed,” OCC said in a written statement.

There’s pressure to start working immediately, residents said. They must pay for rent, hygiene products, clothes and transportation. On top of that, they pay restitution, court fees, child support, state and federal taxes and set aside savings in order to graduate from the program.

The urgency makes the job search more about the facilities getting their money than residents’ long-term success, former staff and residents said.

“You threaten the client, you say: ‘You better get a job, you better work, you’ve got another week. If you’re not working in another week, you’re going to go back,’” said Rimes, the former ComCor security manager.

Otten said residents at the two Jefferson County facilities were pressured to accept the first job they found in order to start paying rent. The jobs were often low paying, ultimately slowing their reentry process.

After getting home from work at a grocery store, Sherman waters his front lawn and talks with a childhood friend he was incarcerated with.

She said among the biggest barriers to employment is offenders’ inability to spend money without prior approval from the facility, which can sometimes take weeks. Facilities have power of attorney over resident’s finances, and getting caught with a debit card or cash can lead to discipline, “even if that money is to buy steel-toed boots in order to get a job,” she said.

When residents found jobs, they were sometimes denied by the facility, Otten said.

One resident wasn’t allowed to work a high-paying construction job because it was out of cellphone range, Otten said. Another was prohibited from working at a law firm because it was located in a house, not an office building. “They sent her to work at the diner, waitressing,” she said.

Before becoming director of justice services, Otten ran workforce training programs for formerly incarcerated people at the Colorado Department of Labor and Employment.

“Community corrections folks were always more difficult to work with because of the rules and regulations that didn’t allow (residents) to work in higher-wage jobs in demand industries,” she said.

More Money, Same Outcome

In response to recommendations from the Office of Community Corrections, lawmakers this year approved increased funding for Colorado’s halfway houses.

This fiscal year, operators will receive $67 per day for each occupied bed, up from $47.96. The increase requires facilities to stop charging rent, per their state contracts, but state law doesn’t prohibit it.

Lawmakers also approved funding that for the first time will be tied to performance, including graduation and recidivism rates. Meanwhile, the state’s oversight agency has made it easier to qualify for the extra dollars by changing its definition of recidivism.

Historically, recidivism was defined as a criminal filing within two years of graduating from a halfway house.

The new definition: a felony conviction within two years of entering a facility.

For example, ComCor’s recidivism rate under the old definition was 41%, according to the state’s data dashboard; under the new definition, it’s 3%.

Only three facilities that were eligible did not qualify for the additional funding.

An Intervention Inc. halfway house in Henderson

“They moved the bar even lower than it was before,” said Hurst, the Colorado Criminal Justice Reform Coalition policy expert. It’s unlikely someone charged with a felony will get through the court process within the two-year timeframe, she said.

Ruske, of the Office of Community Corrections, told ProPublica that if facilities are not trying to improve, the state has the authority to cut their funding to “encourage participation.” OCC last exercised that authority in 2016.

Although ComCor qualified for the new incentive funding tied to recidivism, Wester, the executive director, doesn’t think a facility should be held accountable for its former clients’ success or failure after they leave a community corrections program.

“We do everything we can to work with clients, but once (they) get out, I mean, they control much of their lives and they start making decisions,” he said.

Hurst said she’s heard facility operators make that argument many times but believes it ignores the purpose of halfway houses to provide treatment and support so people stay out of the criminal justice system.

“That’s the whole premise of community corrections,” Hurst said. “So if the outcome is that people are ultimately ending up in prison, in jail, anyway, that’s concerning.”

Do You Have a Tip for ProPublica? Help Us Do Journalism.

by ProPublica

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up for Dispatches, a newsletter that spotlights wrongdoing around the country, to receive our stories in your inbox every week.

Applications are now open for five spots in ProPublica’s Local Reporting Network. We’re seeking to work with local journalists in Mississippi, New Mexico and New Orleans who are interested in investigating wrongdoing and abuses of power in their communities.

Our new partners will begin work on Jan. 2, 2023, and will continue for one year. Journalists from local and regional publications covering those three locations are eligible to apply.

ProPublica will pay each full-time reporter’s salary (up to $75,000), plus an allowance for benefits. Local reporters will work from and report to their home newsrooms while receiving extensive support and guidance for their work from ProPublica, including collaboration with a senior editor and access to ProPublica’s expertise with data, research, engagement, video and design. The work will be published or broadcast by your newsroom and simultaneously by ProPublica.

Applications are due Nov. 1, 2022, at 11:59 p.m. Pacific time. Here are more details for those interested in applying.

ProPublica launched the Local Reporting Network at the beginning of 2018 to boost investigative journalism in local newsrooms. Since then, we have worked with nearly 60 news organizations. The network is part of ProPublica’s local initiative, which includes offices in the Midwest, South and Southwest, plus an investigative unit in partnership with the Texas Tribune.

Reporting by the Local Reporting Network and other local partners has had significant impact.

MLK50, a nonprofit news organization in Memphis, Tennessee, reported on how the area’s largest hospital system sued and garnished the wages of thousands of poor patients, including its own employees, for unpaid medical debts. The hospital subsequently curtailed its lawsuits against patients, erased $11.9 million in unpaid medical debts, dramatically expanded its financial assistance policy for hospital care and raised the minimum wage it pays employees. The stories won the Selden Ring Award for Investigative Reporting.

Our partnership with the Miami Herald looked at the deeply troubled Florida program intended to provide services and a financial cushion for the families of children born with devastating brain injuries. The “Birth Rights” series found that the program protected doctors at the expense of suffering families and that it had amassed $1.5 billion in assets while families waited for help. The reporting pushed the state legislature to quickly enact long-needed reforms and spurred the program’s executive director to roll out further benefits for the families before she ultimately resigned.

And our collaboration with Nashville Public Radio (WPLN) went deep into one county in Tennessee that was arresting and locking up children at extraordinary rates. The series about Rutherford County was read more than 3.5 million times and spurred demands for reform. Eleven members of Congress called for the U.S. Department of Justice to open a civil rights investigation. Tennessee’s governor called for a review of Rutherford County’s juvenile court judge. In January 2022, legislators introduced a bill to remove the judge, citing an “appalling abuse of power.” An hour after ProPublica wrote about that bill, the judge announced that she would retire this year rather than run for election.

Applications to join the Local Reporting Network should be submitted by newsroom leaders proposing a particular project and a specific reporter. If you lead a newsroom and are interested in working with us, we’d like to hear from you about:

• An investigative project. The proposed coverage can take any number of forms: a few long stories, an ongoing series of shorter stories, text, audio, video or something else. We are looking for some specific qualities in a proposal: what makes this story unique to your community (Why here?); how your project is different from prior coverage; how it points the finger not only at harm but at wrongdoing committed by a person, policy, law or entity; why this project has particular urgency now; and how your newsroom plans to execute the work.
• The reporter whom you ideally envision spearheading the work and the market salary you would need to pay them from Jan. 1, 2023, through Dec. 31, 2023. This could be someone already on staff or someone else — for example, a freelancer with whom you hope to work. Please include a personal statement by the reporter explaining their interest, at least three clips and, of course, a resume.

Freelancers are also welcome to apply, but must submit a joint application with an eligible news organization willing to publish their work.

• Have an idea? You can find more details on how to apply, what we’re looking for and how the program works on our website. Proposals need to be submitted using this form.
• Want feedback on an idea you’re developing? You can send a written draft of your proposals to Local.Reporting@propublica.org no later than Oct. 18 and we will get back to you with written feedback within a few days.
• Anything else you’d like to ask? Feel free to email us at Local.Reporting@propublica.org.

Please submit your proposal by Nov. 1, 2022, at 11:59 p.m. Pacific time. Entries will be judged principally by ProPublica editors. Selected proposals will be announced by early December.

by Jodi S. Cohen, ProPublica, and Jennifer Smith Richards, Chicago Tribune

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up for Dispatches, a newsletter that spotlights wrongdoing around the country, to receive our stories in your inbox every week.

This story was co-published with the Chicago Tribune.

Join us Sept. 20 for a live virtual event, “Ticketed at School.”

At Illinois’ largest high school district, Black and Latino students were suspended more often than white students, disciplined more often for subjective reasons like dress code violations, and referred more frequently to the local police, who in many cases then issued costly tickets for misbehavior, data submitted as part of a state investigation shows.

The information provided to state officials by Township High School District 211 reveals widespread disparities involving not just students of color but also those with disabilities. The Illinois attorney general’s office is in the early stages of a civil rights investigation that aims to determine if some groups of District 211 students have been unfairly disciplined at school.

The data, obtained through a public records request, shows that Black and Latino students together received about 65% of the roughly 470 tickets police have issued to high school students since the start of the 2018-19 school year. Those groups make up just 32% of district enrollment. White students, meanwhile, make up nearly 43% of enrollment but received about 28% of the tickets.

More than 22% of the tickets were issued to students whose disabilities made them eligible for an Individualized Education Plan, though those students make up only about 9% of the enrollment. The district, with 11,800 students, operates five high schools and two alternative schools in the large suburbs of Palatine, Hoffman Estates and Schaumburg.

The attorney general’s office began its investigation following publication of “The Price Kids Pay,” an investigation by the Chicago Tribune and ProPublica that revealed schools across Illinois have been working with police to punish students with costly tickets for violating municipal ordinances.

The news organizations’ investigation included a first-of-its-kind database of tickets issued at Illinois public schools over three school years. The attorney general’s office has said District 211 stood out both for the number of tickets issued and for racial disparities in the ticketing. (The news database includes tickets provided by local police departments, not the school district, from the school years ending in 2019, 2020 and 2021. The data from the state investigation includes the school year ending this summer and comes from the district).

District officials deny that race factors into their disciplinary decisions. In a statement, District 211 Superintendent Lisa Small said the data on disparities reduces “individual students to simplistic statistics.”

“They are not a statistic to us, but a developing young adult,” she wrote. “Race is not and has never been a factor in our student discipline. Decisions are highly individualized and based on the specific behavior and are not well-suited to a simple numerical analysis.”

Small also said the district recently stopped alerting law enforcement when students are truant. A 2019 Illinois law made it illegal for schools to notify police when students are truant so that officers can ticket them, but the Tribune and ProPublica identified dozens of school districts that routinely failed to follow this law, including District 211.

Amy Meek, chief of the Civil Rights Bureau in the Illinois attorney general’s office, said in an interview last week that an initial review of the District 211 data is revealing disparities related to students’ race and disability status, and that “those are the kinds of things that we would typically want to look more closely at.”

Meek said the next step is to try to understand, possibly through additional records and interviews, the reasons for the disparities. “You have to not just identify the disparity but also assess whether this disparity is the result of a particular policy or set of decisions and is there an alternative that would have a less disparate impact,” Meek said.

One Black parent whose son received a ticket at Palatine High School said the disparities in the data don’t surprise her. Her son’s ticket, issued by a school-based police officer, was dismissed at a hearing. Racial bias exists everywhere, she said, including at school.

“It is already inside their minds like, ‘These people are the bad people so we have to keep an extra eye out on them,’” said the woman, who attended the hearing at the village’s police station. “Of course they should be doing something about it.”

Some District 211 records show that school officials were aware of concerns about racial disparities in discipline. The district commissioned a study on race and equity at its schools that began in 2020 and became the basis for a districtwide plan to address racial equity.

“Parents and students believe that there are unfair disciplinary practices in our schools” that result in a “harsher discipline consequence” for Black and Latino students, according to early findings based on focus groups of students and parents.

The district’s data confirms that disparities exist, including in enforcement of minor disciplinary matters such as insubordination, dress code violations and student ID violations. In addition to the ticketing disparities, district records show that Black and Latino students together received 65% of the suspensions that schools imposed. About 36% of the suspensions went to students with disabilities.

But in a letter to the attorney general’s office accompanying the data, the district’s lawyer denied any bias.

“The District does not believe the data contained below reflects a disparate impact of disciplinary consequences,” the letter states. The district also told the attorney general’s office that it has prioritized training and professional development on “educational equity and trauma informed care.” It provided hundreds of pages of materials, including training for school employees on how to respond to students in crisis and on understanding equity.

The district also distanced itself from actions police take at school, stating in the letter that police, not high school employees, write the tickets. Workers also don’t call police to come to the school, according to the letter. Instead, employees refer students to “police consultants,” or police officers who are already stationed at the schools, and those officers decide what action to take.

The district also said that when it sends students to police, its “intent is not that students are issued fines or fees, but to add a layer of support.” Yet district training and handbook materials warn that misbehavior can result in fines.

Tickets issued by Palatine police carry fines and, in some cases, a $55 fee for an administrative hearing, where tickets can be disputed or penalties reduced. The hearing officer in Palatine can approve community service instead of a fine, but he also can impose up to a $750 fine if the student doesn’t show up to the hearing or complete the ordered community service.

The attorney general’s office is also examining the village of Palatine and its Police Department, which has officers stationed in three district buildings.

District 211 and Palatine sent records to the attorney general’s office in July. The school district then provided those records — more than 1,500 pages’ worth — to the Tribune and ProPublica in response to a Freedom of Information Act request. The Palatine Police Department has refused to turn over its records to the news organizations.

The district’s records show that when school officials involve the police, a process known as a police referral, Latino students are more likely than other groups to receive a ticket.

The federal government considers all tickets written on campus as stemming from a police referral. Districtwide, Latino students were referred to police 400 times over the time period the attorney general is examining, and about 64% of those referrals resulted in a ticket. When white students were referred to police, tickets were written about 52% of the time. Black students referred to police received tickets about 31% of the time.

The data does not include tickets issued in the 2020-2021 school year, when buildings were closed for much of the year because of the pandemic. And though the data shows tickets were issued at all District 211 high schools, the district acknowledged that it may not have records on all tickets issued and that the three police departments serving the district could have records concerning additional tickets.

The Tribune and ProPublica previously documented the tickets issued on school property or at school events by police from Palatine, Schaumburg and Hoffman Estates and reported that the reason for the tickets was often possession of e-cigarettes, cannabis or drug paraphernalia; truancy; or disorderly conduct. So far, the attorney general’s office has requested records only from the Palatine police, but Meek said it’s typical for her office to issue subsequent requests for information and conduct interviews in the course of civil rights investigations.

Police chiefs in Palatine and Hoffman Estates, where officers issued most of the school-based tickets, which often carried hundreds of dollars in fines, did not respond to requests for comment. In Schaumburg, where students received fewer tickets, police Chief Bill Wolf said decisions to arrest or issue citations to students depend on “the seriousness of the crime, previous history, and the ability or desire of the parents or guardians to correct the behavior.”

“The race or disability status of the juvenile has absolutely no part in the determination to make an arrest or pursue a citation,” Wolf said in an email. The Schaumburg tickets often do not result in a fine, he said.

In June and July, reporters attended hearings at the Palatine police station where high school students who had received tickets appeared before a hearing officer. Students had been ticketed for truancy, possession of a vaping device and vandalism.

One student, a Palatine High School sophomore, was ticketed in April after he returned to school from a lunch break. The police officer stationed at the school accused the 16-year-old student, who is Black, of damaging a chain-link fence that students often pass through as a shortcut at lunchtime.

The school is listed as the complainant on the ticket. The student said in an interview that he told the school-based police officer he didn’t damage the fence.

Palatine High School students regularly walk through the gap in this fence that separates the school from an apartment complex. A school-based police officer issued a sophomore a ticket for vandalism earlier this year. The student denied damaging the fence. (Armando L. Sanchez/Chicago Tribune)

The ticket, which accused the student of violating Palatine’s vandalism ordinance, carried a $200 fine. At the student’s June hearing, the hearing officer suggested that the student do 16 hours of community service, which the village allows as an alternative to paying the fine. When the student started to talk about his fast-food jobs, including one at Smoothie King, the hearing officer said he thought he recognized the teen from there. “I remember your hair,” he told the student. The student was confused; he had not yet started working at that job.

A police officer based at Palatine High School issued this $200 citation to a sophomore he accused of damaging a fence. The school is listed as the complainant. At a later hearing, the ticket was dismissed after another student took responsibility for the hole in the fence. (Redactions by ProPublica. Ticket obtained by ProPublica)

The student’s mom then interjected, asking her son whether he had damaged the fence. The student said he hadn’t, that it was already damaged when he went through. “It wasn’t me,” he said. “I am saying I didn’t rip the fence. I didn’t rip the fence open.” The hearing officer responded by telling the student the police officer could have charged him with a crime instead of writing a ticket.

The hearing officer ultimately dismissed the student’s case when another student who received a similar ticket acknowledged he had damaged the fence.

But the hearing officer didn’t seem to recall the dismissal when he saw the first student at Smoothie King a couple weeks later, according to the student.

“The judge came to my job and mentioned it in front of my coworkers,” the student said. “He asked in front of them, ‘Did you get those [community service] hours done?’”

Attempts to reach the hearing officer for comment were unsuccessful.

“It was uncomfortable,” said the student, who felt forced to explain the situation to the other employees. “It’s embarrassing to do it in front of my co-workers because it could have ruined my job.”

by Andrea Bernstein

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up to receive our biggest stories as soon as they’re published.

Ever since the FBI came out of Mar-a-Lago last month with box after box of documents, some of them highly sensitive and classified, questions have wafted over the criminal investigation: Why did former President Donald Trump sneak off with the stash to begin with? Why did he keep it when he was asked to return it? And what, if anything, did he plan to do with it?

It’s true that Trump likes to collect shiny objects, like the framed Time magazine cover that was stowed, according to the U.S. Justice Department, alongside documents marked top secret. It’s true, as The Associated Press reported, that Trump has a “penchant for collecting” items that demonstrate his connection to famous people, like Shaquille O’Neal’s giant shoe, which he kept in his office in New York’s Trump Tower.

But I’ve covered Trump and his business for decades, and there’s something else people around him have told me over and over again: Trump knows the value of hoarding sensitive, secret information and wielding it regularly and precisely for his own ends. The 76-year-old former host of “The Apprentice” came up in the world of New York tabloids, where trading gossip was the coin of the realm. Certainly sometimes he just wanted to show off that he knew things about important people. But he also has used compromising information to pressure elected officials, seek a commercial advantage or blunt accountability and oversight.

A Trump spokesperson did not respond to a request for comment.

Take a little-known episode where Trump tried to pressure former Republican New Jersey Gov. Christine Todd Whitman.

In 1997, Trump, then a major casino owner in Atlantic City, was furious with New Jersey elected officials for supporting a $330 million tunnel project. The tunnel would run from the Atlantic City Expressway almost to the doorstep of a casino run by then-rival Steve Wynn. “Trump didn’t want Wynn in Atlantic City,” Whitman recently told me. Trump “wanted to control the gambling there.”

As a casino owner, Trump wasn’t able to make donations in New Jersey legislative races, contributions being one of his go-to methods of attempting to exert control over government decisions. But Trump could run caustic ads and file lawsuits, which he did. When none of that worked, and the tunnel was in the final stages of approval, Whitman said, Trump called her up.

A few years before the tunnel vote, Whitman’s son, Taylor, who was in high school at the time, had gotten falling-down drunk at a private dance at Trump’s Plaza Hotel off Central Park in New York City and had to be taken to the hospital. This is something that high school students stupidly do, and Whitman said to me she was happy for Taylor to be sick “to teach him a lesson.” But in the call, Trump suddenly brought the episode up. He said it would be “too bad” if the press found out about her son’s drunken antics.

“He made the threat during the deliberations over the tunnel,” Whitman said, and it “blindsided” her because the high school dance was private and Taylor's behavior had been a family concern. She had no idea how Trump found out about it, she said, but the episode made it clear to her that people collected and delivered sensitive information to Trump about what happened in his properties. She did not buckle to Trump, and he never made good on his threat.

Many people who have found themselves, for better or worse, in Trump’s orbit over the decades — people with far less power than Whitman — told me it was obvious that Trump collected information on people, delighted in it, even. And he was not shy about deploying it. “There was always someone watching,” one former high-level Trump Organization employee told me. “What Donald would do is he would let the person know he knows, in his around-the-corner way. He let the person know he was all-imposing and he knew what was going on.” Like most other former employees, this person did not want to speak on the record for fear that Trump would still come after him all these years later.

It was helpful for Trump that people knew he collected information on others’ less glowing moments as potential ammunition down the road. One top former New Jersey lawmaker told me that he’d been warned to be on his best behavior when he traveled to Atlantic City because Trump kept an eye on important people. Even as a rumor, it furnished Trump with power.

In one infamous case involving a journalist, Trump wielded his knowledge about behavior in the casino town.

In 1990, Neil Barsky, then a Wall Street Journal reporter, came upon a scoop. He was told by a banker that “Donald Trump is driving 100 miles per hour toward a brick wall, and he has no brakes” in Atlantic City. Four large banks had hundreds of millions of dollars of debt on the line. Trump was divorcing his first wife, Ivana, and trying desperately to keep his finances from her and out of the tabloids. Unfortunately for him, Barsky kept writing about Trump’s financial difficulties.

In early 1991, one of Trump’s senior executives offered Barsky comp tickets to a company-sponsored boxing match in Atlantic City. His editor encouraged him to accept a ticket for himself to cultivate Trump Organization sources. In what he later called “an act of bad judgment,” Barsky also accepted tickets for his father and brother. Writing about the episode in 2016, Barsky said he later learned that after the match, Trump called the New York Post, asking, “How would you like to destroy the career of a Wall Street Journal reporter?” The story that ensued conjured a picture of a malevolent Barsky, extorting the tickets in exchange for keeping bad stories out of his paper.

After it appeared, the editors moved Barsky off the beat and Trump no longer had to deal with his tough financial scrutiny.

A decade later, Trump tried the same thing with another journalist, New York Times real estate reporter Charles V. Bagli. For years, Trump had offered Bagli tickets to the U.S. Open. One year, Bagli finally accepted to advance his reporting on a story. Trump had been trying to ingratiate himself with an important beat writer — but now he had a piece of potentially compromising information.

Finally the moment came. After Bagli wrote a story fact-checking the opening credits to “The Apprentice,” writing that Trump “is not the largest developer in New York, nor does he own Trump International Hotel and Tower,” Trump pounced. His lawyer sent a letter to the Times threatening a lawsuit and stating that Bagli had tried to shake Trump down for the tickets and wrote the piece when Trump refused. The accusation was false, and the Times backed its reporter.

If people’s gambling and hotel habits can be valuable, top secret intelligence has the potential to be even more so. As it was back in his casino heyday, just the knowledge that Trump may have compromising secrets, and could use them, confers continued power.

The New Jersey tunnel Trump fought so hard against was ultimately approved, though Wynn, and then Trump, left Atlantic City. But Trump and Whitman never reconciled. In 2016 she declined to support him in the Republican primary for president. Displeased, Trump forwarded a letter to her, Whitman recalled, that again referred to her son’s drunken incident at the school dance. By this time her son, who now works in health care finance, was an adult. As Whitman remembered, on the letter were these words scrawled with a Sharpie: “Too bad you don’t remember the good old days.”

Help ProPublica Investigate Threats to U.S. Democracy

by Mauricio Rodríguez Pons, Aliyya Swaby and Annie Waldman

Sign up for ProPublica’s User’s Guide to Democracy, a series of personalized emails that help you understand the upcoming election, from who’s on your ballot to how to cast your vote.

This story was co-published with Gray TV/Investigate TV.

Before 1965, many Southern states forced voters to prove they could read before casting a ballot, a requirement primarily designed to keep Black people from voting. The Voting Rights Act put an end to those polling site exams. But a ProPublica investigation found that the efforts to block people who have difficulty reading from casting a ballot continue, especially in the South. In fact, today's election system remains a modern-day literacy test.

It didn’t get that way by accident. For decades, conservative politicians have passed laws to make it harder for these voters to cast a ballot and discourage anyone trying to help them.

Olivia Coley-Pearson knows this better than most. She has been criminally charged twice in the past decade for her attempts to help people navigate their ballots; she has never been found guilty of any wrongdoing. Now 60, she serves as a city commissioner in Douglas, the majority-Black seat of Coffee County, where a third of the population struggles to read. On the day of Georgia’s primary elections in May, she woke up early to rally voters and volunteer. ProPublica followed her to capture what it takes to ensure that voters who need help can get it.

To learn more, check out ProPublica’s investigation of Coley-Pearson’s fight and the persistent suppression of low-literacy voters, read our story about successful voting reforms, and see our guide on how to get help with voting.

by Kavitha Surana and Adriana Gallardo

In the wake of the Supreme Court’s decision to overturn Roe vs. Wade, strict new abortion laws went into effect in more than a dozen states. Since then, women have reported being denied care for miscarriages, ectopic pregnancies, fatal fetal anomalies and unforseen crises, like premature rupture of membranes.

We are a nonprofit, nonpartisan, investigative news organization that wants to better understand how these laws are affecting the most intimate of health care decisions between patients and providers. Lawmakers who support the restrictions say the measures include exceptions to address life-threatening emergencies, and, in some cases, rape and incest. But many medical providers say the laws are not clear enough to account for all of the dangers that could arise during pregnancy.

These are significant policy changes and, as reporters, we are interested in learning about how they are experienced by real people. We know there are a lot of strong feelings about this issue, but we’re not looking for opinions. We’re looking for examples and insights. We are especially interested in hearing from caregivers or lawyers who work in the continuum of medical care.

If you’re in a state that tightly restricts abortion, we’d like to know more about your experiences and observations. If you are pregnant or are planning to become pregnant, what are your questions about how your state’s new laws affect your options or care? Have you had a medical conversation about what falls under the definition of “emergency” or a health threat under your state’s law? Contact us using the form below.

by Caresse Jackman, Gray TV/Investigate TV, and Aliyya Swaby and Annie Waldman, ProPublica

Sign up for ProPublica’s User’s Guide to Democracy, a series of personalized emails that help you understand the upcoming election, from who’s on your ballot to how to cast your vote.

This video is the result of a partnership between ProPublica and Gray TV/InvestigateTV.

For all the recent focus on voting rights, little attention has been paid to one of the most sustained and brazen suppression campaigns in America: the effort to block help at the voting booth for people who struggle to read — a group that now amounts to about 48 million Americans, or more than a fifth of the adult population.

Across the country, from California to Georgia, people like Olivia Coley-Pearson and Faye Combs are working to help citizens with low literacy skills exercise their constitutional right to vote, but doing so requires fighting through stigma and increased restrictions on accessibility.

Watch the Investigation (Investigate TV)

While new voting restrictions in states like Florida, Texas and Georgia do not all target voters who struggle to read, they make it especially challenging for these voters to get help casting ballots. ProPublica analyzed the voter turnout in 3,000 counties and found that places with lower estimated literacy rates tended to also have lower turnout.

See for yourself: For the launch of its Right to Read series, ProPublica partnered with Gray TV’s Investigate TV team, which produced the segment above.

Read the full ProPublica investigation.

by Cezary Podkul, with Cindy Liu for ProPublica

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up to receive our biggest stories as soon as they’re published.

The ads on the Telegram messaging service’s White Shark Channel this summer had the matter-of-fact tone and clipped phrasing you might find on a Craigslist posting. But this Chinese-language forum, which had some 5,700 users, wasn’t selling used Pelotons or cleaning services. It was selling human beings — in particular, human beings in Sihanoukville, Cambodia, and other cities in southeast Asia.

“Selling a Chinese man in Sihanoukville just smuggled from China. 22 years old with ID card, typing very slow,” one ad read, listing $10,000 as the price. Another began: “Cambodia, Sihanoukville, six Bangladeshis, can type and speak English.” Like handbills in the days of American slavery, the channel also included offers of bounties for people who had run away. (After an inquiry from ProPublica, Telegram closed the White Shark Channel for “distributing the private information of individuals without consent.” But similar forums still operate freely.)

.image-gallery ul { display: flex; flex-flow: row wrap; list-style: none; margin: 0px auto; padding: 0px; } .image-gallery li { margin: 0 10px; flex-basis: calc(33% - 20px); } .image-gallery img { width: 100%; height: auto; } @media screen and (max-width: 36em) { .image-gallery li { margin: 10px; flex-basis: calc(100% - 20px); } }

• Cambodia, Sihanoukville, six Bangladeshis, can type and speak English, for companies in Sihanoukville targeting foreigners, people who can make decisions come chat, can deliver.
• Sihanoukville, a few Chinese born in the 90s. Cannot share pictures, no passports. Have ID cards, can interview directly. Price relatively high.
• Selling a Chinese man in Sihanoukville just smuggled from China. 22 years old with ID card, typing very slow. Only third day here. Companies who can accept slow typers come chat. 10,000 USD.

Listings from the White Shark Channel on Telegram (Screenshots and translations by ProPublica)

Fan, a 22-year-old from China who was taken captive in 2021, was sold twice within the past year, he said. He doesn’t know if he was listed on Telegram. All he knows is that each time he was sold, his new captors raised the amount he’d have to pay to buy his freedom. In that way, his debt more than doubled from $7,000 to $15,500 in a country where the annual per capita income is about $1,600.

Fan, photographed in Phnom Penh, Cambodia (Cindy Liu for ProPublica)

Fan’s descent into forced labor began, as human trafficking often does, with what seemed like a bona fide opportunity. He had been a prep cook at his sister’s restaurant in China’s Fujian province until it closed, then he delivered meals for an app-based service. In March 2021, Fan was offered a marketing position with what purported to be a well-known food delivery company in Cambodia. The proposed salary, $1,000 a month, was enticing by local standards, and the company offered to fly him in. Fan was so excited that he told his older brother, who already worked in Cambodia, about the opportunity. Fan’s brother quit his job and joined him. By the time they realized the offer was a sham, it was too late. Their new bosses wouldn’t let them leave the compound where they had been put to work.

Unlike the countless people trafficked before them who were forced to perform sex work or labor for commercial shrimping operations, the two brothers ended up in a new occupation for trafficking victims: playing roles in financial scams that have swindled people across the globe, including in the United States.

Tens of thousands of people from China, Taiwan, Thailand, Vietnam and elsewhere in the region have been similarly tricked. Phony job ads lure them into working in Cambodia, Laos and Myanmar, where Chinese criminal syndicates have set up cyberfraud operations, according to interviews with human rights advocates, law enforcement personnel, rescuers and a dozen victims of this new form of human trafficking. The victims are then coerced into defrauding people all around the world. If they resist, they face beatings, food deprivation or electric shocks. Some jump from balconies to escape. Others accept their lot and become paid participants in cybercrime.

Fan and his brother eventually ended up in Sihanoukville in a compound surrounded by a barbed wire fence. They were made to lure people in Germany into depositing funds with a phony online brokerage controlled by their operation, which also targeted English speakers in Australia and elsewhere.

“This idea of combining two crimes, scamming and human trafficking, is a very new phenomenon,” said Matt Friedman, chief executive of the Mekong Club, a Hong Kong-based nonprofit that combats what it calls modern slavery. Calling it a “double hurt,” Friedman said it’s unlike anything he’s ever seen in his 35-year career. The phenomenon has only just begun to come to light in the U.S., including in a Vice article published in July.

The most widely used technique among these operations is known as pig butchering, an allusion to the practice of fattening up a hog before slaughtering it. The approach combines some time-tested elements of fraud — such as gaining trust, in the manner of a Ponzi scheme, by making it easy for marks to extract cash at first — with elements unique to the internet era. It relies on the effectiveness of relationships nurtured on social media and the ease with which currencies can be moved electronically.

Typically, fraudsters ingratiate themselves into online friendships or romantic relationships and then manipulate their targets into depositing larger and larger sums in investment platforms that are controlled by the fraudsters. Once the targets can’t or won’t deposit more, they lose access to their original funds. They’re then informed that the only way to retrieve their cash is by depositing even more money or paying a hefty fee. Needless to say, any additional funds disappear in similar fashion.

Some Americans have lost huge sums. An entrepreneur in California said she was swindled out of $2 million and unwittingly facilitated an additional $1 million in losses by convincing her friends to join her in what seemed like a surefire investment. A hospital technician in Houston enticed her friends and colleagues to join her in a similar scheme, costing the group $110,000. An accountant in Connecticut is no longer preparing for retirement after watching $180,000 vanish in two separate swindles. They were among more than two dozen scam victims from seven countries interviewed by ProPublica.

Out of fear or shame, most pig butchering victims do not report their losses. That’s one reason that the limited data available likely understates the scale of the damage. According to the Global Anti-Scam Organization, a nonprofit founded last year to combat the new form of fraud, at least 1,838 people in 46 countries have lost an average of about $169,000 each to pig butchering since June 2021. Many still seem stunned by the effectiveness of the trickery. “I have to say, it’s brilliant,” said a Silicon Valley CEO who tallied her loss at $800,000 and asked not to be named out of embarrassment. For many victims, the betrayal by a seeming friend only compounds the devastation.

Fan’s ordeal began with a burst of optimism. He flew to Cambodia’s capital, Phnom Penh — it was his first time leaving China — and then waited out two weeks of COVID-19 quarantine in a hotel. He was then driven to a walled-off condominium complex in the city to begin his training. It was only then, in April 2021, that he realized something was off. Instead of learning about food delivery or working in a kitchen, he and his brother were placed in front of computers and told to study materials about how to defraud people online.

Fan, who is serious and reserved, with a crew cut and a round face that betrays little emotion, was able to document parts of his account, including the offer letter that drew him to Cambodia. (Fan is his family nickname; he asked that ProPublica not include his full name out of fear of his captors.) His experiences resembled those of other trafficking victims ProPublica interviewed and aligned with descriptions provided by experts and others.

Job ads, like this one on Facebook, are often used by human traffickers to lure young people into scam sweatshops in Sihanoukville. Facebook removed the post after ProPublica asked about the ad. (Screenshot by ProPublica)

Fan and his brother spent six months engaging in pig butchering schemes before their bosses decided to relocate the operation to Sihanoukville. The bosses presented them with a choice: They could pay the equivalent of $7,000 each to leave, or they could move along with the company. The brothers, who were paid negligible wages for their work, couldn’t afford the fee. So they relocated to Sihanoukville, in the upper floors of a hotel and casino called the White Sand Palace located in the center of the city.

The job could be terrifying. Fan said he witnessed a worker “half-beaten to death” by guards. “People were saying: ‘Help him! Help him!’” he recalled. “But nobody went up to help him. Nobody dared to.”

Only weeks after Fan and his brother arrived at White Sand, they experienced a brief moment of hope, Fan said. A person approached them and offered to get them out. With his help, they managed to leave — only to realize that the seeming savior had sold them to another criminal organization. This one was located in a fortified complex of beige dormitories on the edge of Sihanoukville with the grandiose name Arc de Triomphe. The $7,000 each owed for his freedom had risen to $11,700. And the price would go higher still.

Cyberfraud operations in Asia, including the ones Fan worked for, are highly organized. Some have gone so far as to draft detailed, psychologically astute training materials on how to dupe strangers. ProPublica obtained more than 200 such documents from an activist who helps involuntary workers escape.

Pig butchering scammers often claim that a well-placed aunt or uncle is feeding them inside information. Some training guides include talking points to teach scammers how to utilize this script effectively. (Screenshot by ProPublica)

Step one in the fraud process for Fan and others was to create an attractive online persona. In his case, he was expected to pose as a woman when wooing targets online. His operation bought photos and videos from websites that cater to such operations. For example, bundles of hundreds of photos of good-looking women and men are available for less than the cost of a cup of coffee from a shop called YouTaoTu. Another website markets a “pig butchering scam” package: For the equivalent of $12, it offers a “handsome guy set” of images of a man with perfectly chiseled abs. (Neither online store responded to requests for comment.) Such photos are frequently lifted from the online accounts of unsuspecting people; ProPublica found that images used by one fraudster were stolen from the Instagram profile of a Chinese social media influencer.

An image of a man showing off his chiseled abs is part of a “pig butchering scam” package of 197 photographs (including one of the man driving a Porsche) on sale for $12 on an online marketplace. (Screenshot by ProPublica)

Scamming guides obtained by ProPublica recommend using such photos to set up social media accounts and then bolster them with the simulacrum of an affluent lifestyle by posting photos of luxury cars, along with descriptions of relevant hobbies such as investing. Stressing your belief in the importance of family, one guide adds, is the sort of touch that helps foster trust.

The resulting profiles can seem so real that one Canadian man met his future scammer after Facebook’s algorithm suggested the person to him as a friend. The chance encounter cost him and his friends nearly $400,000, according to a police report he later filed. Other victims told ProPublica they met their scammers on LinkedIn, OkCupid, Tinder, Instagram or WhatsApp. (Meta, which owns Facebook, WhatsApp and Instagram, said it has “long prohibited this content” and is investing “significant resources” into blocking it. Match Group, owner of Tinder and OkCupid, said it’s using machine learning and content moderators to fight fraud. LinkedIn didn’t respond to emails seeking comment.)

The next step for Fan was to contact as many victims as possible. He recalled working on a team of eight under a group leader, who gave each of them 10 phones to make it easy to keep multiple chats going, along with lists of phone numbers to contact. Fan’s job was to try to initiate conversations on WhatsApp. He would do it by pretending he had reached a wrong number, a common ruse. Others would open with a simple “Hi.”

One former scam worker holds the phones he used to contact potential victims. (Photograph provided to ProPublica)

Some tiny percentage of people responded favorably. When they did, Fan’s job was to handle the crucial initial part of the conversation. That’s when scammers are instructed to get to know their victims and discover what one training guide calls “pain points” that can be exploited. It’s also an opportunity to do what another document calls “customer mapping,” screening potential marks to glean information on their wealth and their vulnerability to being “cut,” slang for convincing them to fall for the scheme.

Fraudsters often initiate conversations on WhatsApp using wrong number-type messages. (Screenshot provided to ProPublica by the Global Anti-Scam Organization)

Using WhatsApp offered other practical advantages. Initially, Fan said, his team was aiming its efforts at Germans. Fan doesn’t speak a word of German, but it didn’t matter. All his chats were filtered through language translation software. Later, his team shifted to marks who spoke English. If any of the potential victims wanted to hear the voice of the attractive woman he was pretending to be, Fan said, there was a woman on staff who spoke fluent English and could record voice memos for him.

Because he was a rookie, Fan’s job was mostly limited to enticing marks to download an app called MetaTrader that would provide access to a brokerage where, he told his new “friends,” they could make fortunes trading cryptocurrencies. Fan would try to convince them to buy cryptocurrencies such as ethereum or bitcoin and deposit them in a brokerage controlled by the scam operation. The brokerage would then post phony numbers, including ones that represented supposed gains in their accounts.

If customers complied and began depositing significant sums, Fan said, he would typically hand the phone to his boss, who would take over and begin prospecting for a major strike. The strategy squares with what several scam victims told ProPublica: They sensed they were talking with multiple people. Indeed, they often were.

Why MetaTrader Is a Favorite Tool for Pig Butchering

Consumers who file complaints about pig butchering with the Federal Trade Commission routinely mention MetaTrader as a conduit for fraud. Among 716 such complaints filed since June 2021, consumers reported losing $87 million, FTC data shows. Separately, ProPublica identified 60 fake brokerages that have used MetaTrader for pig butchering.

Why has the app become such a staple of these scams?

MetaTrader isn’t a brokerage. It’s a platform. It’s analogous to using Amazon’s website to buy products from other retailers. Only in the case of MetaTrader, customers use the platform, typically via its phone apps, to access online brokerages where they can trade foreign currencies or other financial instruments. Both Apple and Google distribute MetaTrader in their app stores, giving it broad availability and a patina of legitimacy. (One training manual advises pig butchering fraudsters to cite its distribution by Apple as proof that MetaTrader can be trusted.)

However, MetaQuotes, the Cyprus-based company behind MetaTrader, allows brokerages that it contracts with to sublicense MetaTrader software to other brokerages with few checks to ensure the legitimacy of the sublicensed operations. This has allowed scammers to use MetaTrader as a front for fraudulent websites. Victims who are bilked via MetaTrader see records of trades and account balances, seemingly allowing them to control their money, when in fact that money is already in the swindlers’ possession.

ProPublica shared the list of fake brokerages and the FTC complaints with MetaQuotes CEO Renat Fatkhullin, along with a detailed list of questions. He did not respond. A lawyer for MetaQuotes told one victim that it is “solely a software development company” and has “nothing to do with any complaints of traders against companies that use the software of our clients.” An Apple spokesperson said the company has shared complaints with MetaQuotes, claiming that MetaQuotes has taken steps to respond to the complaints. The spokesperson provided no examples. Google did not respond to a request for comment. The FTC declined to comment.

About 8,000 miles from Cambodia, an American who lives near San Francisco got a WhatsApp message on Oct. 7, 2021, from a stranger calling herself Jessica. She seemed to have reached him by mistake. Jessica asked the man, whose middle name is Yuen, if they knew each other; she said she had found his number on her phone and didn’t know why. Yuen responded that he didn’t know her. But Jessica was chatty and friendly, and her photo was alluring, so they kept talking.

Yuen agreed to tell his story on the condition that ProPublica identify him only by his middle name and omit certain details that could identify him. He saved his chat history with Jessica, which would run to 129,000 words over several months, and later shared it with ProPublica. (Yuen also shared his chat history with Forbes.)

At the moment Jessica initiated contact, Yuen was vulnerable. His father was in a hospital, dying from a lung disease. He had entrusted Yuen, the youngest of four siblings, with the power to decide whether to cut off his life support. It would also be up to Yuen to plan his father’s funeral and distribute his estate.

The family had immigrated to the U.S. from Hong Kong decades earlier. Yuen, who is in his early 50s and works as an accountant for a major university, was more affluent than his siblings, who are all older than him. He felt it was his duty to take care of them in old age, much as he was caring for his father and had cared for his late mother. Jessica told him she admired his commitment to his family. She shared her own tale of having a grandfather in the hospital.

Yuen began trading messages with a stranger named Jessica on WhatsApp in October 2021. (Brian Frank, special to ProPublica)

Jessica was, by all appearances, a savvy and talented woman. She claimed to be a Chinese immigrant herself, a private banker at J.P. Morgan Chase in New York City. (A Chase spokesperson said the bank has no current employee with her purported Chinese name, Wang Xinyi.) Jessica’s photographs showed her spending weekends on Long Island playing with her rich friend’s toddler. She seemed fashionable, loved shopping and found time for yoga nearly every day, and she would flirt with Yuen. When Jessica posted photos of herself at a luxurious beach property, he wrote, “Wish I was there now.” She replied, “We can go play together.”

Text exchanges between Jessica (gray) and Yuen (green) (Screenshots provided to ProPublica)

One Monday in late October, Jessica told Yuen she had just made $100,000 trading gold contracts. She let him in on a secret: She had a rich uncle in Hong Kong who had his own team of analysts who fed her inside quotes about where the price of gold would move. Every time “Uncle,” as she referred to him, called with news of where the market would go, she could make a guaranteed 10% profit by trading on his directions.

Jessica offered to teach Yuen — but only him. “Why just me?” he asked. Jessica said it was because she sympathized with Yuen about his dying father. “The money you earn can better help your father,” she explained. Plus, she knew she could trust him to keep her secret about insider trading. “Of course, I won’t tell anyone,” Yuen told Jessica as he pondered whether to join in.

The exchange marked a key moment in Yuen’s relationship with Jessica. The person behind Jessica’s alter ego was using a manipulation technique called “altercasting,” according to Martina Dove, a psychology researcher and author of “The Psychology of Fraud, Persuasion and Scam Techniques,” who reviewed Yuen’s chat log at ProPublica’s request. It puts the scammer in a position of trusting the target so that the target will reciprocate trust later on. Keeping the trading secret also meant less chance that Yuen’s wife or teenage daughter would learn about his chats with Jessica.

When Yuen agreed to put some money into gold, Jessica told him to download MetaTrader from Apple’s app store. She then told him to use the app to search for a brokerage called S&J Future Limited.

Yuen made it clear he couldn’t afford to lose any money. If he did, he said, he’d have to kill himself. Jessica said there was no need to worry: Uncle was never wrong. Yuen owed it to his father to seize the opportunity.

On Oct. 26, the day he had to go to the hospital to discuss his father’s end-of-life care, Yuen put money on the line for the first time. A conservative investor and lifelong saver, he’d been petrified to put even $2,000 into the brokerage. Jessica convinced him to start with $10,000 and taught him the two-step process to fund his account. First, he wired money from his bank to buy a cryptocurrency called ethereum. Then he could transfer the ethereum to a crypto wallet, whose address she provided.

Jessica insisted that using a cryptocurrency would help Yuen minimize his tax burden. He admitted he had very little idea of what he was doing. No matter. When the transfer was done, his S&J account reflected the deposit. And the next day, when Uncle called Jessica with news, Yuen was ready to buy. His account showed he made $746 after fees.

Jessica claimed she had made $500,000 on the same trade. She told him to get his account up to $50,000 to start earning meaningful sums. Yuen agreed and wired $20,000 the next day and another $20,000 a few days after that. When Jessica saw that he was doing as she’d directed, she praised him — “you’re smart” — and reminded him that the more money he put in, the more he’d earn for his father and siblings.

Little by little, Jessica encouraged Yuen to invest more and more. Yuen liquidated some mutual funds and wired nearly $58,000 on Nov. 2. When Uncle called with news later that night, his MetaTrader account showed an eye-popping gain of $17,000.

In those early weeks, Yuen was thrilled with Jessica. He called her his “true angel” in one message and offered up emojis of joy. Jessica wrote back: “I am not an angel, I am a demon.” She added two smiley-face emojis.

The skyline of Sihanoukville (Cindy Liu for ProPublica)

If Cambodia has a capital of fraud, it may well be Sihanoukville, which is named for the country’s onetime king who was ousted in an American-supported coup during the turmoil that erupted as the U.S. bombed the nation during the Vietnam War. The city has transformed over the past five years from a quiet beach resort to a metropolis of casinos and ghostly towers in various stages of construction or decay. The building boom was funded by Chinese investors, who started pouring millions of dollars into Sihanoukville after 2016, when the Philippines launched a crackdown on illegal online gambling outfits that were aimed at Chinese citizens. Cambodia had looser gaming regulations, and its government welcomed Chinese investment, making it a perfect substitute.

Soon Cambodia experienced the same influx of organized crime that had prompted the crackdown in the Philippines. Cambodia, under pressure from the Chinese government, announced a ban on online gambling in August 2019. Months after that, the COVID-19 pandemic struck and casinos in Cambodia were suddenly emptied of customers and workers.

Criminal syndicates repurposed their emptied real estate and began using it for scamming operations, according to Jason Tower, Myanmar country director for the United States Institute of Peace, and other observers in the region. “They’re criminal businesses, but they’re businesses at the end of the day,” he said. “So what did they do? They adapted.” And thanks to the pandemic, human traffickers found no shortage of job seekers with computer skills.

Forced Labor Compounds Have Spread Across Cities in Cambodia and Elsewhere in Southeast Asia (Map by Lucas Waldron, ProPublica. © OpenStreetMap Contributors. Research by Danielle Keeton-Olsen.)

These facilities, which are housed in everything from office buildings to garish casino complexes, aren’t all tucked away in isolated neighborhoods. Some are prominently situated in the heart of cities. The White Sand Palace, which contains not only a gambling establishment but also multiple floors of fraud operations, according to former workers there, is located diagonally across the street from the summer residence of the Cambodian prime minister. White Sand didn’t respond to a request for comment.

Scam operations are often in central locations. The White Sand Palace in Sihanoukville is only a block or so from the prime minister’s summer residence. (Photo by ProPublica)

Many fraud operations are surrounded by barbed wire fences. It’s routine to see windows and balconies completely enclosed by bars. In the Chinatown area of Sihanoukville, storefronts for a noodle shop and a barbershop look unexceptional, until you walk inside and notice that there are bars inside preventing anyone from exiting the complex of heavily guarded buildings.

The beige stretch of towers, center, and the buildings to the left, located in the Chinatown area of Sihanoukville, house scam operations, according to people who say they were held captive there. (Cindy Liu for ProPublica)

Over the past year, an array of activists, journalists and nongovernmental organizations in Southeast Asia have begun revealing what’s going on behind the bars in these buildings. Ngô Minh Hiếu, a reformed hacker who now works as a cybersecurity analyst for the Vietnamese government, was one of the first to identify the sites. NGOs such as the International Justice Mission, as well as local media outlets, most notably VOD News, have revealed details of the operations. (ProPublica collaborated with three reporters affiliated with VOD to prepare this article.)

Lu Xiangri, a survivor of human trafficking who became a rescuer, wears a red bracelet he received at a pagoda where he prayed for people trapped in scam compounds. (Cindy Liu for ProPublica)

Others, such as Lu Xiangri, who became a volunteer rescuer after escaping a Sihanoukville scam sweatshop, have collected videos depicting abuses in these operations. Lu witnessed severe mistreatment when he was briefly detained inside the Arc de Triomphe last October: He saw a man with a broken leg and a bruised back begging to be sold so that he could avoid further beatings; Lu said the man later died of his injuries. Determined to help others avoid a similar fate, Lu joined a volunteer rescue team, which exposed him to a steady stream of pleas for help that often include graphic images of wounds left by electric shock batons and other corporal punishment. (ProPublica examined scores of similar photos and videos, some of which depict torture — including the use of electronic shock devices on workers’ genitals — but is publishing only a limited number whose authenticity was verified by Lu.)

ProPublica drove up to the gates of three compounds in Sihanoukville where people have alleged being detained and compelled to work as fraudsters. They included the Arc de Triomphe, one complex in Chinatown and another sprawling compound known as White Sand 2. Security guards at the three locations either denied that anything illegal goes on inside or refused to answer questions. “Talk to the boss,” one said, without specifying who the boss was.

Alleged scam compounds in Sihanoukville (Cezary Podkul/ProPublica)

Fan said his life was tightly circumscribed when he worked and lived inside the Arc de Triomphe compound. He could leave his building and enter an adjoining casino and karaoke bar — he said he had no interest, though some workers did gamble or go to the karaoke bar — but the presence of guards would dissuade any hopes of going out into the street. During the four months he was at the Arc de Triomphe, he said, he never set foot outside the compound.

His schedule and routine were regimented. Fan worked on the second floor of a building from 5 p.m. to 9 p.m., then again from 11 p.m. to 5 a.m. He slept in a dorm room with metal bunk beds, with four or five people per room. There was even a small clinic in the compound that provided first aid and rudimentary medical treatment. As Fan put it: “You can’t go anywhere. You’re either eating, sleeping or working.” The days ran into each other, and Fan tried to anesthetize his own feelings, willing himself into emotional torpor. His only pleasure came from playing a fantasy warfare game on his phone each night before going to bed.

The Arc de Triomphe compound in Sihanoukville (Cindy Liu for ProPublica)

Fan hated the work. Cheating people out of money was the last thing he thought he’d be doing when he answered a job ad. But he couldn’t leave the compound, nor could he afford to buy his way out. His bosses at the Arc de Triomphe demanded $23,400 for him and his brother. The two were essentially paid on commission, which meant that the more he wanted his freedom, the more he’d have to bilk.

In part because he would hand over promising targets to his boss, but perhaps also because of his reluctance, Fan never delivered a big score. The most he landed was $30,000. He said he felt so terrible after that “success” that he deleted the victim’s contact information from the organization’s database to make sure the person couldn’t be further stripped of cash. Others on his team, he said, extracted as much $500,000 from a single victim.

On Nov. 3, as Jessica was helping him turn his newest deposit of $70,000 into cryptocurrency, Yuen got a message that his father had been taken to the hospital. Yuen raced to join him, and as he sat in the waiting room, some other news came through. It was Jessica, saying her uncle in Hong Kong had given her another signal to trade.

Yuen explained to Jessica that his father was dehydrated and losing the will to eat. He was back in the hospital two days later, crying as he wiped his father’s hands and face. Shortly after, Jessica messaged to ask if his latest deposit of $20,000 had gone through. Yes, he said. He added that he’d decided to give his father comfort in his dying days by moving him to a hospice.

Jessica didn’t seem to grasp what a hospice was. When Yuen explained that it was a care facility for the terminally ill, she perked up: “You need to make more money.” Jessica told him he should raise his account balance to $500,000 so he could cover the cost more easily.

Over the next nine days, Yuen cashed in a $20,000 CD that his mother had bequeathed him and his siblings and tapped a dormant home equity line of credit for $200,000. Each time he traded with Jessica, his account showed an increase, and soon he surpassed the $500,000 mark she’d set out for him.

As Yuen was moving his father into hospice, Jessica pressured him to increase his deposits. “You need to make more money,” she told him. (Brian Frank, special to ProPublica)

Yuen’s father died in the early morning hours of Nov. 14. Yuen was the only one with him when he breathed his last. He wrote Jessica, seeking sympathy, but got a perfunctory response. This is a common stratagem, said Dove, the psychology researcher. She calls it “scarcity”: withdrawing attention unless the target is doing what the scammer wants. When Yuen wanted to talk about anything related to money, Jessica engaged. When he wanted her attention for anything else, she was distant and tried to steer the conversation back to investing.

The next day, with his father now gone, Jessica gave Yuen another goal. She bragged that she was buying yet another home in New York. The conversation turned to real estate and how Yuen could afford a pied-à-terre there. Why, he asked? “So that we can get very close,” Jessica responded. She explained that Uncle had told her a “big market” was coming soon. “If you want to buy a house in New York, you need to increase your capital,” she said.

In just a few weeks of trading, Yuen had shed much of his previous caution. But now he resisted. He was planning his father’s funeral, he told Jessica, and he was overwhelmed at work. Buying a home in New York would have to wait. Jessica urged him to take out another loan. When Yuen refused, she chided him: “You are a wise man, this is borrowing a chicken to lay eggs.” But Yuen didn’t budge.

By Nov. 18, he had gone six days without depositing more money into his account. That’s when his investing idyll came to an end. That day, his MetaTrader app suddenly closed him out of his positions. By the time it was over, his account showed a balance of minus $480,000.

Yuen panicked. He couldn’t lose any of this money, but he felt he couldn’t turn to anyone for help, either. He’d been keeping his MetaTrader habit secret. He lied to his wife and daughter when they asked who he was messaging so frequently, brushing it off as an endless stream of work requests. His siblings didn’t know either. No one knew. No one except Jessica.

Jessica convinced him it was his fault. He must have exited the MetaTrader app instead of following her directions. But it was OK, she said. The big market was still there and he could make everything back quickly. “Prepare the funds and earn them back,” she said.

Yuen didn’t know it, but he had now entered the final stage of pig butchering. This is when scammers sense that their targets have been squeezed dry and are unlikely to deposit more funds. They then shift to the final manipulation: Making the targets aware that they’ve lost all their money and offering them a seeming lifeline to earn it back. The move aims to heighten the targets’ distress. “We normally don’t make our best decisions when we’re in a state of emotional arousal,” said Marti DeLiema, a gerontologist at the University of Minnesota who researches how older Americans are swindled.

Yuen immediately dialed up the financial institutions that managed his family’s savings and ordered a sale of $500,000 worth of mutual fund shares. As he waited for the money to be transferred, he debated whether to inform his family about the loss. Jessica told him, via an emoji depicting an index finger over closed lips, not to say a thing. If he’d only wait a few more days and deposit more funds, he’d turn his loss into a gain. “Yes, we will earn it back,” Yuen said.

The following week, Yuen borrowed $100,000 from his brother-in-law and resumed trading with Jessica as he made final preparations for his father’s funeral. On the day of the funeral, he messaged Jessica. “When I was crying today,” he wrote, “I wasn’t sure if I was crying because I lost my father or I lost all the money.” She responded, “Money can be earned, but people are gone if they are gone.” Yuen thanked Jessica for her help.

Midway through the following week, Jessica pushed him to borrow even more, but Yuen said he had no one he could turn to. Jessica wasn’t buying it. “I don’t think you’ve reached your limit,” she said, adding that every time she had asked him to gather cash before, he’d been able to do so.

When she pushed him again the next day, Yuen exploded. “Omg.!!!” he wrote. “You don’t understand! I have no more resources to get anymore money!” By that time, he couldn’t sleep or eat or do anything other than worry about how to make back his losses.

On Dec. 3, at 11:31 a.m., Jessica messaged Yuen to get ready to do another trade with Uncle’s news. Three minutes later, Yuen executed his 23rd trade with Jessica. Once again, disaster struck: All of his positions suddenly got closed, and his entire portfolio vanished as he watched.

Yuen convulsed with panic. He spent the next several hours in shock and terror as the consequences of the loss raced through his mind. “Give me a solution,” he begged Jessica. She told him to put more money in. When he told her that all he had left was $105, Jessica answered: “With $105, start from scratch, I believe you, you can do it.”

Later that day, Yuen confessed to his family. He told his wife he’d lost 30 years’ worth of their savings. He later admitted that the money he’d borrowed from her brother and the bank was also gone. In all, he had lost just over $1 million. Yuen asked his brother to call an ambulance to escort him to a psychiatric ward, where he was placed on a suicide watch.

If you or someone you know needs help with suicidal thoughts, you can call or text the National Suicide Prevention Lifeline at 988.

Yuen was released two days later and spent December wondering what had happened. Jessica stopped replying to his messages after a few days, but he kept on asking. “It’s Christmas. Hope you have the heart to help me!!!” he messaged on Dec. 25. (ProPublica got no response to messages it sent to the WhatsApp number used by Jessica.)

Yuen said he didn’t accept that he’d been cheated until after Jan. 1. It was only through the intervention of close friends and relatives that he acknowledged what had happened. He found a support group, the Global Anti-Scam Organization, and began piecing together details of the scam, like a Reddit post warning that S&J Future Limited was a sham brokerage. A fellow victim set up a GoFundMe page to help him, and others began to chip in, including a Massachusetts woman who had lost $2.5 million herself.

In the end, Yuen lost $1 million, more than he and his wife had saved over 30 years. (Brian Frank, special to ProPublica)

But Yuen still struggled to comprehend what kind of person — and where? — would impose such suffering on someone else. He got a partial answer on March 31. That’s when Jessica contacted him again using a different phone number. Yuen was prepared: Another GASO member had taught him a trick to track down a person’s IP address to figure out their location. The chat log shows Jessica fell for the ruse. When the IP information came back, Yuen said, it indicated Cambodia.

Well before then, Fan and his brother had passed the point of desperation. They began seeking ways out of the Arc de Triomphe. In late January, Fan messaged the governor of Preah Sihanouk province via Facebook. The governor’s office responded, asking for Fan’s phone number, he said, and soon after the police called.

But the attempt backfired. Fan’s bosses found out about the call and summoned him and his brother. They berated the two for tarnishing the company’s reputation and threatened legal consequences, according to Fan. The meeting culminated in a videotaped confession in which Fan’s brother read a statement, on behalf of both brothers, prepared by their bosses. A video recording shows Fan’s brother reading a script in which he stated that they had gotten a “personal loan” from the company and had to repay it. He ended by saying, “We would like to apologize to the provincial governor.”

When Fan returned to his desk, his boss was furious. The boss slapped him, Fan said, threw a water bottle in his face and told him to go find the money to pay for his freedom. His boss warned him, according to Fan, that “it doesn’t even matter if you die in here” because it would be so easy to kill him. No one would care. (At least six dead bodies have been discovered in the marshlands or beaches near Sihanoukville’s scam compounds, many of them Chinese men.)

Fan’s police report turned him and his brother into troublemakers in the eyes of their employer. They got sold to another fraud operation, this one back in Phnom Penh, which tacked on further charges to their debt. Each now would have to pay $15,500 for his freedom.

Fan stands on the roof of the guesthouse he stayed in after his escape. (Cindy Liu for ProPublica)

In February, Fan found a way to get out. He noticed that his new bosses were less strict about security than his previous captors. They occasionally allowed workers to venture outside the compound. So Fan came up with an excuse — visiting a friend — and received permission to leave. He suspects that his captors let him go because they believed that, as long as they still held his brother, he would return.

But Fan didn’t return. Meanwhile, his brother called the police, and this time they came through. He was released at the urging of local authorities. But before his brother left, he was forced to confess again, this time in writing. That handwritten letter, which Fan shared with ProPublica, stated that he had borrowed $31,000 from the company, was happy and working voluntarily and had never been kidnapped or beaten.

Fan spent his first months of freedom in the Great Wall Hotel, a modest five-story guesthouse steps away from Phnom Penh’s airport that has become a haven for Chinese scam workers who manage to escape. Life at the Great Wall was safe but monotonous. Most residents were just passing the time as they waited for an opportunity to return to China, which has restricted travel due to its zero-COVID-19 policy. Those limits contributed to rising costs for airline tickets, putting a return nearly out of reach for many of the escapees.

In June, Fan moved out of the Great Wall Hotel. He declined to reveal his exact whereabouts, as he’s still afraid that he will be abducted by bounty hunters. Fan has obtained paperwork that will allow him to return to China without his passport, which a scam compound still holds, and his father recently managed to cobble together enough money to pay for him to fly home. Fan dreams of returning to work on his family’s farm, tending to ducks and chickens while safely under his parents’ roof. “I won’t come out to work again,” Fan said. “There’s not much future working for other people.”

The front desk at the Great Wall Hotel in Phnom Penh (Cindy Liu for ProPublica)

Cambodia’s fraud operations often have links not just to organized crime but also to the country’s political and business elites. The Arc de Triomphe, for instance, is owned by K99, a real estate and casino junket operator led by Rithy Raksmei, brother of the late tycoon Rithy Samnang. Samnang was also son-in-law to ruling party senator Kok An, whose business empire includes properties that have faced allegations of forced scam labor. And the complex in Chinatown has a hotel that is part-owned by Xu Aimin, a Chinese fugitive who was sentenced to 10 years in prison for masterminding an illicit international gambling ring. (None of these individuals or entities have been prosecuted for involvement in Cambodian scam compounds and none responded to ProPublica’s requests for comment.)

In July, the U.S. State Department downgraded Cambodia to the lowest tier on its annual assessment of how well countries are meeting standards for eliminating human trafficking. The department asserted that Cambodian authorities “did not investigate or hold criminally accountable any officials involved in the large majority of credible reports of complicity, in particular with unscrupulous business owners who subjected thousands of men, women, and children throughout the country to human trafficking in entertainment establishments, brick kilns, and online scam operations.” A United Nations special rapporteur on human rights in Cambodia put it in searing terms in an August report: Workers trapped in Cambodian scam compounds are experiencing a “living hell.”

The day the U.N. report appeared, Cambodia’s government reversed months of denials and acknowledged that foreign nationals have been trafficked to the country to work in gambling and scam operations. Cambodian Interior Minister Sar Kheng condemned what he called “inhumane acts” and expressed regret. The statement came only days after a dramatic, widely seen video emerged of some 40 Vietnamese men and women breaking out of a reported fraud compound and, chased by baton-wielding men, frantically jumping into a river that divides Cambodia from Vietnam.

Cambodia’s senior official working to combat human trafficking, Chou Bun Eng, told ProPublica in a July interview that her government was still figuring out how to respond to scam sweatshops. “This is new for us,” she said. Top officials from Cambodian police, immigration and other government agencies met in Phnom Penh in late August to discuss a strategy. They pledged action, then almost immediately, the government’s Ministry of Foreign Affairs and International Cooperation undercut that stance by releasing a statement in September asserting that human trafficking in Cambodia is “not as serious, bad as reported.”

Lacking any form of legally recognized status, escapees from scam compounds are left at the mercy of Cambodian police, who often treat them as illegal immigrants or criminals. Rescued people frequently end up in crowded immigration detention centers, sleeping on the floor in tight quarters without any air conditioning, according to images shared by a detainee.

The police have sometimes pursued the rescuers. Chen Baorong, the former head of a charity group that helped human trafficking victims escape, was arrested in February and charged with incitement. In late August, he was sentenced to two years in prison. Lu Xiangri, the volunteer rescuer, took up Chen’s mantle after his arrest, only to himself flee Cambodia in July out of concern for his safety. In response to questions from ProPublica, Cambodia’s General Commissariat of National Police wrote that ​​“it is not the government policy to collude with any criminal group or facilitate the use of Cambodian soil by criminals as a hotbed for fraudulent activities overseas.”

Lu in Sihanoukville in May, before he fled Cambodia (Cindy Liu for ProPublica)

The governments of China, Indonesia, Pakistan, Thailand and Vietnam have issued warnings in recent months about high-salary job offers emanating from Cambodia. Authorities in Taiwan and Hong Kong have gone so far as to station workers at airports to question people emigrating for work and to warn them about overseas employment scams. Still, even as the governments issue warnings about Cambodia, new operations are gravitating to places like Myanmar, where the violent aftermath of a military coup has created an opportunity for criminal syndicates to expand.

In the U.S., law enforcement and victims are trying, against long odds, to recapture lost money. In May, the Santa Clara County District Attorney’s Office seized $318,000 of stolen crypto funds on behalf of one pig butchering victim. Erin West, the deputy district attorney spearheading the effort, said her team has been able to seize an additional $233,000 since then and has a few more seizures in the works. Still, most funds aren’t recovered, and the chances drop rapidly as time passes.

Yuen is losing hope that he’ll recover his funds. At one point, he turned down an offer from a self-described hacker to introduce him to an FBI agent who would track down his stolen funds if Yuen paid him $5,000. Cautious, Yuen asked to see a photograph of the agent’s FBI identification. The badge looked authentic, as did the photo ID. But under the photograph, Yuen noticed the signature. It read “Fox Mulder,” the name of the fictional detective on “The X-Files.”

Do You Have a Tip for ProPublica? Help Us Do Journalism.

Mech Dara and Danielle Keeton-Olsen contributed reporting from Cambodia, and Salina Li from Hong Kong.

by Aliyya Swaby and Annie Waldman

Sign up for ProPublica’s User’s Guide to Democracy, a series of personalized emails that help you understand the upcoming election, from who’s on your ballot to how to cast your vote.

This story was co-published with Gray TV/Investigate TV.

For nearly 10 hours on Georgia’s primary day, Olivia Coley-Pearson tracked down every potential voter she could find, working two cellphones as she paced the parking lot outside the polls, repeating the same message: “You need to tell all your cousins, your brothers, your sisters, your aunts, your uncles — everybody you know — to come on down here to vote.”

A third of her neighbors in Coffee County struggle to read at a basic level, and she wanted to make sure they had help navigating their ballots. In the late afternoon, she slid behind the sparkly pink steering wheel of her SUV for her final push of the day, heading down a long stretch of road where buildings gave way to fields and thickets of pine. She turned in to the Kinwood Estates mobile home park and stopped at the edge of a familiar dirt driveway just as Shondriana Jones, 30, bounded down the steps of a trailer.

“I can’t find my ID and Mama, she’s still at work,” Jones said.

Coley-Pearson has helped the family vote for years — she’s known them since she and Jones’ mother, Sabrina Fillmore, were young. Now 60, Coley-Pearson serves as a city commissioner in Douglas, the majority-Black county seat. Fillmore, 54, works at the local poultry plant cutting chickens. Neither Fillmore nor her daughter can read beyond a first-grade level, but they rarely miss an election, believing their votes can influence everything from their electricity costs to the way police treat them.

Coley-Pearson urged Jones to track down a utility bill to prove her identity at the election office just as Fillmore returned from a 10-hour shift, exhausted. With the women aboard, Coley-Pearson started the car, anxiety brewing in her mind.

Olivia Coley-Pearson (Joseph Ross, special to ProPublica)

Even though federal law guaranteed the two women the right to have someone help them vote, Coley-Pearson knew too well that this right was under attack. For all of the recent uproar over voting rights, little attention has been paid to one of the most sustained and brazen suppression campaigns in America: the effort to block help at the voting booth for people who struggle to read — a group that amounts to about 48 million Americans, or more than a fifth of the adult population. ProPublica analyzed the voter turnout in 3,000 counties and found that those with lower estimated literacy rates, on average, had lower turnout.

“How the system is set up, it disenfranchises people,” said Coley-Pearson, who blames Southern political leaders for throwing up hurdles. “It’s by design, I believe, because they want to maintain that power and that control.”

Conservative politicians have long used harsh tactics against voters who can’t read — poor, often Black and Latino Americans who have been failed by the U.S. education system and who conservatives feared would vote for liberal candidates. Some states have required voters who needed help to sign an affidavit explaining why they need assistance; some have prevented voters who couldn’t read from bringing sample ballots to the polls and limited the number of voters that a volunteer could help read a ballot. Time and again, federal courts have struck down such restrictions as illegal and unconstitutional. Inevitably, states just create more.

Over the last two years, the myth of election fraud, supercharged by former President Donald Trump in the wake of his 2020 loss, has fueled a barrage of new restrictions. While they do not all target voters who struggle to read, they make it especially challenging for voters with low literacy skills to get help casting ballots.

Last year, Georgia passed a law limiting who can return or even touch a completed absentee ballot. Florida expanded the radius around election locations in which volunteers are prohibited from asking people if they need help. Texas passed a law prohibiting voters’ assistants from answering questions or paraphrasing complicated language on the ballot; a federal judge struck down several sections of the law in June. But the court left other provisions in place, including ones that increase penalties for helping voters who don’t qualify and require people who assist voters to fill out more paperwork. Texas did not appeal the decision.

To appreciate the impact of voter suppression, consider that recent elections have been determined by a narrow sliver of the electorate:

Despite losing the popular vote, Trump secured the presidency in 2016 by winning Pennsylvania, Wisconsin and Michigan by a margin of just under 80,000 total votes.

President Joe Biden prevailed in 2020 by winning Arizona, Georgia and Wisconsin by just over 40,000 votes combined.

Coley-Pearson recognizes the importance of this moment for Georgia, which is no stranger to close elections. Republican Gov. Brian Kemp faces another challenge from Democrat Stacey Abrams, and Sen. Raphael Warnock is attempting to hold on to his seat in a race that could tip the Senate back to Republican control. But to Coley-Pearson, helping people vote isn’t only about politics or even just about their rights as individuals. It is about the future of democracy at a time when it seems like the views of the majority are being marginalized by the actions of the few.

The Gladys Coley resource center. Second image: A memorial plaque bearing the name of Gladys Coley. (Joseph Ross, special to ProPublica)

As a child in the 1970s, she’d watched as her mother, Gladys Coley — who stood just above 5 feet and had only an eighth grade education — rose to the helm of the local NAACP and challenged the discriminatory school system and police department. Her mother begged her not to return from college in Atlanta, but Coley-Pearson wanted to fight for the people of Coffee County, too. As she headed to the polls on primary day this past May, though, she couldn’t subdue her fear that by helping Jones and Fillmore, she was putting a target on her own back.

Over the course of several years, she’d become tangled in an investigation of supposed voter fraud, which took aim at her attempts to assist voters who requested help. She had pleaded her case to television cameras and at a hearing before the state’s highest election official. She had even wound up in jail.

“Intimidation is real,” Coley-Pearson said. “If we don’t continue to vote, they’re going to have us right back where it used to be.”

(Mauricio Rodríguez Pons/ProPublica)

Watch video ➜

Coley-Pearson was born in an era when Southern states forced convoluted literacy tests on voters to keep Black people out of the polls. In those days, local voting officials often made exceptions for white people who couldn’t read. In 1965, Congress passed the Voting Rights Act prohibiting racial discrimination at the polls. That didn’t stop white conservatives, especially in the South, from continuing to discriminate against voters with low literacy skills, who, due to centuries of oppression, were disproportionately Black.

An excerpt from a Louisiana voter literacy test that was in use around 1963. (Civil Rights Movement Archive)

Conservatives argued that removing barriers for voters who couldn’t read would allow the federal government to overrule states’ decisions on how to run local elections and would hand more votes to liberal candidates. Clearly, they said, voters with low reading skills would be easily swayed by anyone assisting them, leading to rampant fraud.

“Today the bureaucrats are issuing certificates to vote to people who cannot read the ballot nor even the instructions on a ballot or on a voting machine,” segregationist Alabama Gov. George Wallace declared in late 1965. “The left wing liberals need as many illiterates as they can get to vote in order to keep them in power.”

The Rev. Fred C. Bennette Jr., a civil rights movement organizer, right, instructs Black people in Atlanta how to fill out registration forms in 1963. (Horace Cort/AP Photo)

By 1981, voters of color, including those with low literacy levels, still faced “white resistance and hostility,” according to a U.S. Commission on Civil Rights report. “For many minority voters, the kind of assistance that they receive at the polls determines whether they will vote,” the report stated. “If minority voters who do not speak English or who are illiterate receive inadequate assistance, they may become too frustrated and discouraged to vote or they may mark their ballots in such a way that they will not be counted.”

Congress amended the Voting Rights Act in 1982 to affirm that voters who need help due to an “inability to read” could bring someone, other than their employer or union representative, to assist them in the voting booth. A string of subsequent lawsuits shows this federal action again failed to eradicate the discrimination.

In a 2001 case, the federal justice department claimed that white poll managers in Charleston County, South Carolina, were intimidating Black voters who requested assistance. According to testimony given in the case, the poll workers launched a barrage of questions at these voters, such as, “Can’t you read and write? And didn’t you just sign in? And you know how to spell your name, why can’t you just vote by yourself? And do you really need voter assistance?”

A federal judge found that there was “significant evidence of intimidation and harassment,” but said evidence of the mistreatment was too “anecdotal” to take direct action.

In 2012, the chairman of Coffee County’s board of elections filed a complaint against Coley-Pearson and three other residents, alleging that they’d assisted voters who didn’t legally qualify for help. Georgia law only allows voters to receive assistance if they are disabled or cannot read English. The secretary of state’s office, then under Kemp’s leadership, initiated an investigation.

Alvin Williams (Joseph Ross, special to ProPublica)

The following summer, a 52-year-old line cook named Alvin Williams answered his phone to find a state investigator on the other end. The man had questions about the 2012 election. “It looks like you were assisted by Olivia Pearson,” said state investigator Glenn Archie, in a recording obtained by ProPublica. (Archie did not respond to a request for comment.) “It’s not marked why she assisted you and I was wondering why you needed assistance.”

The tone of the man’s voice made Williams nervous. “Because I can’t read. I’m illiterate,” Williams told Archie. He’d dropped out of school at 16 to work full time catching chickens and selling them to the local poultry plant, a job he’d skipped classes for since he was 11 to help support his family.

“I’m sure she read the candidates to you,” Archie said. “Did you get to pick the people you wanted to vote for?”

“Yes, sir,” Williams said. “I can’t read. That’s why she was helping me.”

“That’s no problem,” the investigator assured him. “She can assist you if you have problems reading.”

But the call left Williams humiliated and fearful of how his vote could be used against him or Coley-Pearson. “I don’t fool with the law,” he said in a recent interview. “And I don’t do nothing for them to fool with me.”

Some other voters told investigators that they had requested and received help even though they could read. The investigation found that Coley-Pearson and the other volunteers neglected to verify whether some voters qualified for help and incorrectly filled out forms indicating why voters needed assistance. It also found that election workers failed to include required information on many forms and turned them in without making sure they were accurate.

Testifying at a 2016 hearing chaired by Kemp, Coley-Pearson maintained that she hadn’t broken any laws. In response to a poll worker’s claim that she’d touched the voting machine, Coley-Pearson said she’d merely accompanied voters who had requested her assistance and stood by to answer questions about the process or read names on the ballot. She said she followed the instructions of the poll workers, signing forms when directed.

“If someone asks me for help, I felt an obligation to try to assist if I could,” she testified at the hearing, stressing that she never told anyone who to vote for. Coley-Pearson suspected there was a deeper significance to the investigation and told the board, “Sometimes things are done to try to maybe dis-encourage, or whatever, other people from voting, and I don’t feel like that is fair.”

The state election board chose not to recommend her case for criminal prosecution, but a local district attorney’s office prosecuted her anyway, which made national headlines in BuzzFeed. It charged her with two felonies for improperly assisting a voter and for signing a form that gave a false reason for why a voter needed assistance. The trial ended with a hung jury. One of two Black people on the jury told a local reporter that she was the only holdout; everyone else voted to find Coley-Pearson guilty. She was tried again in a nearby county and, after about 20 minutes of deliberations, the new jury acquitted her of all charges. The district attorney’s office did not respond to ProPublica’s emailed questions.

Watch the Video On the day of Georgia’s primary elections in May, ProPublica followed Olivia Coley-Pearson to capture what it takes to ensure that voters who need help can get it. (Mauricio Rodriguez Pons/ProPublica and Zach Read for ProPublica)

Three other volunteers took plea deals in which they admitted to making false statements on forms indicating the reason that a voter needed assistance; in exchange, they got probation, after which any fines would be waived. One of them, James Curtis Hicks, said that if he had fought his case and lost, he could have faced jail time or a mountain of fines. He didn’t want to take any risks. “Around here, to me, they target the leaders, the people that are standing up for the rights of the minority,” he said in a recent interview. “To shut me and Ms. Pearson down, it would stop a whole lot of people going to the polls.”

For years, the 59-year-old truck driver had kept tabs on Coffee County voters to see if they needed help reading the ballot. But after the settlement, he stopped. “I didn’t want a focus on me to suppress anyone else,” he said. “I really felt intimidated.”

But the charges didn’t deter Coley-Pearson.

(Mauricio Rodríguez Pons/ProPublica)

Watch video ➜

Before Jones could vote that May afternoon, she needed to get temporary identification. Dodging the pouring rain, she and Coley-Pearson scuttled into the elections office shortly before it closed. At nearly 6 feet tall, Coley-Pearson towered over the woman sitting behind a plexiglass barrier.

“She needs a voter ID, sweetie,” Coley-Pearson said, leaning in. The woman handed Jones an application.

“You need me to do it, baby?” Coley-Pearson asked softly.

Jones nodded, “Yes, ma’am.”

The woman at the counter emphasized that Jones had to complete it on her own.

“She has trouble reading and writing,” Coley-Pearson said.

After a tense moment, the woman agreed that Coley-Pearson could fill out the form. She read the questions out loud and filled in Jones’ answers, pointing out which lines to sign and date.

Shondriana Jones (Joseph Ross, special to ProPublica)

Jones is in the third generation of her family that is not able to read. Her grandmother never learned how, and her mother, Fillmore, left high school in her sophomore year, after frequently being disciplined for fighting. As an adult, Fillmore briefly attended an education program to help her learn how to read, but she felt discouraged and left.

Jones graduated from high school in Coffee County but says she reads at the same first-grade level as her mother. She remembers attending special education classes with more field trips than written assignments and says teachers never diagnosed her with a learning disability or gave her one-on-one assistance. School administrators also frequently suspended her for fighting, she said. “They were trying to get rid of me.”

Coffee County has long failed to provide an equal education for students of color. In 1969, federal officials sued its school board for refusing to integrate white and Black schools. Even after the school system was integrated, Black students continued to receive fewer academic resources and harsher punishments than their white peers. A decade ago, the district acknowledged its shortcomings in reading instruction and the need to rectify its problems with literacy, which were more pronounced for Black students.

The county’s lower literacy rate is related to its high poverty rate, and since integration, the district has worked to increase opportunities for students of color, Coffee County School District Superintendent Morris Leis said in an email; he added that the district does not use discipline to “push out” children who have academic challenges, and it has reduced racial disparities in discipline after it initiated a new program in 2014. By that time, Jones had graduated.

She aspires to learn how to read through an adult education program and to eventually work at a child care center, but she cannot do so without steady transportation. She has not applied for a driver’s license; though she could take the written test orally like her mother did, she hasn’t been able to find someone who has time to help her study the examination booklet.

Ordinary tasks are often insurmountable for her. She owns a smartphone, but mining the web for information is daunting. After she fell several months behind on her electric payments, she could not read the notice that warned her lights would be cut off. She likely qualifies for low-cost internet, but she cannot navigate the instructions for accessing it. When she takes her son to the doctor’s office, she prints his first and last name on the forms but asks the staff for help with the rest. Unable to decipher her most valuable documents, like her birth certificate, she entrusts them to her aunt, who can read and helps determine what she needs for appointments and applications.

Jones worries most about keeping up with her 4-year-old son as he grows. She can read beginner books to him, but she knows his knowledge will soon surpass her own.

For Jones, the voting process itself is like a literacy test. If she changes her address, she cannot easily update her registration. If she enters the polling booth alone, she may recognize a few names on the ballot, but any unfamiliar words could confound her, particularly when it comes to the often-confusing constitutional amendments. She prefers voting by mail, which allows her more time to process her choices, but Georgia’s new election law is making that more difficult. The law has banned outside groups from mailing out absentee ballot applications that have the resident’s information already filled in, and it has limited who can submit the applications on voters’ behalf. The law does include exceptions for people helping “illiterate” voters, but experts say its limit on assistance could still discourage those voters from requesting help.

“Any law that limits assistance is going to have an impact on voters with limited literacy,” said Sean Morales-Doyle, acting director of voting rights at the Brennan Center for Justice. “Whether or not that’s the intention of the lawmakers, that’s always a difficult thing to say. But I do think sometimes it may very well be the intention.”

It is impossible to say precisely what role literacy plays in voter turnout. There are many other factors that contribute to lower participation, including some closely intertwined with literacy, such as income and education level. But to put the importance of reading ability in perspective, ProPublica analyzed data on turnout from the three most recent national elections and compared it to average estimated literacy levels for over 3,000 counties. (Read more about our analysis, and the data used, in our methodology. Watch Investigate TV’s segment about this story.) Our analysis found that if low-literacy counties had turnout similar to high-literacy counties, they could have added up to about 7 million votes to the national total for each of those three elections.

Across the country, people like Jones are stumbling through inscrutable election processes fraught with poor ballot design and rigid registration rules. Some are choosing not to vote at all. (Read more about how some states are trying to make voting more accessible.) “We know in general that the more barriers we put in front of people, the lower the participation rate,” said Donald Moynihan, a professor of public policy at Georgetown University. “Even if someone with lower literacy has the same desire to vote as someone reading this article, they have to overcome more barriers.”

In 2014, for example, Ohio legislators began requiring voters to fill out more complicated versions of absentee and provisional ballot forms while at the same time limiting the assistance they could get from poll workers. Minor errors in the paperwork could lead to people’s votes not being counted. In a lawsuit, the Northeast Ohio Coalition for the Homeless claimed that the laws disproportionately harmed poor, nonwhite and low-literacy voters who would be more likely to have their ballots rejected for minor errors.

Data submitted as evidence shows that thousands of forms were tossed in the 2014 and 2015 general elections for simple problems such as incomplete addresses and birthdays. Poll workers refused one form because the street name “Cuthbert” was misspelled as “Cuthberth.” Several others were rejected because birth dates were listed as the current date, an indicator that voters may not have understood the instructions.

In 2016, a federal judge struck down the measures, concluding they disproportionately harmed Black voters. The 6th U.S. Circuit Court of Appeals agreed that state rules requiring perfect completion of absentee ballot forms posed an undue burden to voters. But the panel said the other measures were minimally disruptive and left in place regulations that limited the assistance voters could get from poll workers and the amount of time voters were given to correct errors on absentee and provisional ballots.

“What the case demonstrates is the indifference of officials from one political party, and of unfortunately many federal judges, to voting rights and to the need to make voting not only secure, but relatively unburdensome,” said Subodh Chandra, an attorney for the plaintiffs.

A similar law in Georgia suspended voter registration applications when the information on the form didn’t exactly match a driver’s license or social security record. (If voters didn’t correct the information within 26 months, Georgia could cancel their registrations.) When then-Secretary of State Kemp ran for governor against Stacey Abrams in 2018, his office suspended the applications of an estimated 53,000 voters, most of them Black, due to these discrepancies. Kemp won the election by about 55,000 votes.

A federal judge ordered Georgia to ease the restrictive program, calling it a “severe burden” on some voters. Politicians, academics and advocates have accused Kemp of voter suppression not only for suspending registration applications over minor discrepancies, but also for purging tens of thousands of infrequent voters from the rolls — a more aggressive effort than is made in other states.

Kemp press secretary Katie Bryd disputed the allegations and noted that Kemp had implemented automatic voter registration through the state’s department of motor vehicles in 2016, which added hundreds of thousands of eligible voters to the rolls. “Politically driven, irresponsible accusations of voter suppression alleged at Governor Kemp have been repeatedly found void of basic facts and validity,” Byrd said in an email.

Today, voters flagged for minor discrepancies in their registration paperwork can no longer be removed from the rolls, but they do have to show a photo identification before they vote.

(Zach Read for ProPublica)

Watch video ➜

As Coley-Pearson parked at the polling station, her thoughts flew back to a similar day not long ago when she wound up handcuffed in the back of a police cruiser.

In October 2020 — more than two years after she was cleared of the felony charges — she was standing in a voting booth helping a young woman with low literacy skills read a ballot, as is allowed by law, when the county’s election supervisor, Misty Martin, confronted her. Martin yelled at Coley-Pearson to not touch the machines and told her she was barred from returning to the polls. Coley-Pearson said she wasn’t touching any machines. “We’re done,” she told the young woman after she finished voting. “Let’s go.”

Martin, who also has used the last names Hampton and Hayes, called the police to report that Coley-Pearson was disruptive, and the department issued a trespass warning barring her from the polls indefinitely. Later that morning, when Coley-Pearson returned to drop off another voter, she was arrested in the parking lot and charged with criminal trespassing. The Georgia Bureau of Investigation is looking into election interference claims in Coffee County, including an incident in which Martin allowed several computer experts connected with Trump’s efforts to challenge the 2020 results into her offices, where they may have had access to election systems; Martin resigned from her county post under pressure last year. She did not respond to ProPublica’s questions related to either incident.

The charge hung over Coley-Pearson for nearly two years; this past June, a state judge agreed to drop the case if she signed a consent order agreeing to follow election law. “There was no evidence of any crime here,” Coley-Pearson said. “It feels like you’re fighting a losing battle.”

Her daughters see how the last several years have worn her down. AiyEsha Coley said she would sometimes wake up at 4 a.m. to feed her newborn and would find her mother on Facebook, reading through disparaging comments. Her daughters have long campaigned for her to retire from city commission, scared that the stress might eventually kill her. She’s starting to come around, and she plans to leave her post next year.

Now peering into her back seat, Coley-Pearson worried her presence could interfere with Jones and Fillmore’s ability to vote. “I did not want any type of confrontation, I did not want any kind of accusations, I just didn’t want any hassle,” she said.

She told them she would not be going in with them and instructed two close friends to help them instead. “When you get through, you all come down there to the tent,” she said, motioning to where her volunteers were sitting out of reach of the rain.

Coley-Pearson watched the women shuffle into the building and fretted as she waited, leaning on her mobile walker at the edge of the parking lot with a group of volunteer canvassers. She had reminded her friends of the rules, but she knew that sometimes, following them was not enough. “They might try to look for anything they could use against them,” she said.

After nearly an hour, Jones and her mother emerged, beaming.

Coley-Pearson’s nerves settled, at least for the moment.

One in Five Americans Struggles to Read. We Want to Understand Why.

by Asia Fields

ProPublica is a group of reporters.

We do not tell you who or what to vote for.

We want to give you information.

• Read this guide in English.
• Lea esta guía en español.
• 点击这里阅读简体中文版
• 點擊這裏閱讀繁體中文版
• Đọc bản hướng dẫn này bằng tiếng Việt.
• Basahin ang gabay na ito sa Tagalog.
• .إقرأ هذا الدليل بالعربية
• اس رہنمائی کو اردو میں پڑھیں۔
• इस गाइड को हिंदी में पढ़ें।
• এই গাইডটি বাংলায় পড়ুন।
• 이 안내서를 한국어로 읽으십시오.

Voting is a right.

You do not have to prove you can read or write to vote.

The law says you can get help with voting if you have a disability or cannot read.

We wrote a story about how some states make it hard for people to vote.

You can watch a video about our story.

We hope this guide helps you vote.

You can call the Election Protection hotline if you need help voting.

• English speakers: Call or text 866-687-8683.
• Spanish speakers: Call 888-839-8682.
• Bengali, Cantonese, Hindi, Urdu, Korean, Mandarin, Tagalog or Vietnamese speakers: Call 888-274-8683.
• Arabic speakers: Call 844-925-5287.

The Election Protection hotline will not tell you who to vote for.

Election day is November 8.

You do not have to vote on everything.

You can just vote for what matters to you.

Can I vote?

You need to be 18 years old or older.

You need to be a citizen of the United States.

You do not need to speak English.

You do not need to know how to read.

You must sign up to vote.

You may not be able to vote if a judge convicted you of a crime.

• Check to see if you can vote.

People with intellectual disabilities can vote in most states.

You may not be able to vote if you have a guardian.

In most states, only a judge can tell you that you cannot vote because of a disability.

You can get help to protect your voting rights.

How do I sign up to vote?

You have to register to vote. To register means to sign up.

• Check if you are registered to vote.
• Register to vote.

Can I vote through the mail?

You can vote through the mail if you have a disability.

Many states let all people vote through the mail.

You need to tell your state ahead of time if you want to vote through the mail.

• Sign up to vote through the mail.

Where do I go to vote in person?

• Find your polling place. Your polling place is where you go to vote.
• Check if you need to bring identification (ID) to vote.

An ID is a card or piece of paper that proves who you are. Some examples are:

• Driver’s license.
• Passport.
• Birth certificate.

Can I have someone help me vote?

The Voting Rights Act is a law.

It says some people can get help with voting.

You can have help with voting if you have a disability.

You can have help with voting if you cannot read.

You can have help with voting if you cannot write.

Who can help me?

You can ask almost anyone to help you.

These people can help you vote:

• Your child.
• A family member.
• A friend.
• A person working at your polling place.

These people cannot help you vote:

• Your boss.
• Your union representative.

How can they help me?

They can read you the ballot. The ballot lists all the people and issues you can vote for.

They can answer questions you have about voting.

They cannot tell you who to vote for.

They cannot look at your ballot unless you ask them to.

What if I do not speak or read English?

Some places have ballots in languages other than English.

• Check to see if your area has ballots in languages other than English.

They may have workers who speak languages other than English.

You can bring someone to help you translate.

If you cannot find someone to help you, call the Election Protection hotline.

What if someone tells me I can’t vote or can’t get help?

Call the Election Protection hotline if you have any problems voting.

• English speakers: Call or text 866-687-8683.
• Spanish speakers: Call 888-839-8682.
• Cantonese, Hindi, Urdu, Korean, Mandarin, Tagalog or Vietnamese speakers: Call 888-274-8683.
• Arabic speakers: Call 844-925-5287.

Some workers may not know that you are allowed to have someone help you vote.

Tell them the Voting Rights Act says you can have help.

It is a crime for someone to intimidate you.

Intimidate means they say or do things that scare you to try to stop you from voting.

We want to know if you have problems voting.

You can leave a message for ProPublica at 212-379-5781.

How do I vote in person?

You may have to wait in line to vote.

Your polling place may close while you wait in line.

If you are in line, you can vote after your polling place closes.

Let the workers know if you brought someone to help you.

The workers may ask you to sign a form.

The form says you need help to vote.

You can ask a worker to help you vote.

Some polling places use voting machines.

Some machines will read your ballot to you.

You push buttons to vote.

Ask a worker about this type of voting machine if you need one.

How do I turn in my mailed ballot?

Your ballot says how to turn it in.

You can have someone else turn in your ballot for you in most places.

Some states do not let other people turn in your ballot for you.

Find out if your state lets other people turn in your ballot.

Call the Election Protection hotline for help.

Sign up to learn how to become a more informed, more engaged citizen with ProPublica’s User’s Guide to Democracy.

You can sign up to get emails with more information about the election.

Asia Fields is an engagement reporter at ProPublica. She wrote this guide. You can send her feedback to literacy@propublica.org.

Rebecca Monteleone translated this story to make it easier to read.

Noah Jodice created the drawings in this story.