After Hurricane Irma hit three months ago in Orlando, Florida, the local police got a desperate 911 call from a 12-year-old boy reporting that his mother and siblings were unconscious. Fumes overcame the first deputy who rushed to the scene. After the police arrived at the property, they found Jan Lebron Diaz, age 13, Jan’s older sister Kiara, 16, and their mother Desiree, 34, lying dead, poisoned from carbon monoxide emitted by their portable generator. Four others in the house went to the hospital. If 12-year-old Louis hadn’t made that call, they might have died, too.

Portable generators release more carbon monoxide — which is particularly dangerous because it is odorless and invisible — than most cars. As a result, the devices can kill efficiently and quickly, though accidentally. The Diaz family usually placed the generator properly, outside the house, a neighbor told local reporters. But for some reason, they had brought it into their garage. From there, the generator’s murderous byproduct spread silently through the house.

During hurricanes, floods, and nor’easters, portable generators save lives — except when they take them. Irma, Harvey, and Maria all left thousands without power and reliant on their portable generators. The government has not yet done its official count, but 11 people using these generators died just from Irma, according to preliminary government estimates. Many more died from Harvey and Maria, experts say, especially in Puerto Rico, which has been without a functioning power grid for months.

These deaths rarely merit more than short stories on local news sites. Civil servants then accumulate the statistics into dry reports that end up buried somewhere on .gov websites. The latest of these shows that portable generators have killed on average 70 people a year since 2005. That’s a small fraction of the toll from car accidents. Still, generators rank as one of the deadliest consumer products on the market. A further 2,800 people a year suffer from carbon monoxide poisoning caused by the equipment.

Portable generator deaths are preventable, and for the past 16-plus years, the United States government has tried to do just that. The job has fallen to the Consumer Product Safety Commission, which, with its $126 million budget and 520 employees, oversees almost every product Americans use in their home, office, or out in the yard, save for food, drugs, and cars. The CPSC, based in Bethesda, Maryland, is tiny, especially compared to many of the corporations it regulates, and hamstrung by Congressional rules that require it to seek voluntary standards before attempting to impose mandatory ones.

But the problem of portable generators was so obvious that the little agency felt it needed to make a stand. They were one of the “persistent deadly hazards we felt we had to address,” says Elliot Kaye, who was the chairman of the CPSC from 2014 to February of this year and remains a commissioner. (The CPSC has five commissioners who vote on agency measures; the party that controls the White House tends to have the majority.)

Throughout the 16 years the CPSC has been pushing the issue, the portable-generator industry fended off regulations that would have required it to reduce the carbon monoxide emissions of its devices. The companies argued such changes would be too costly, and that they lacked the technology to make the machines safer. The industry lobbied hard, and also wielded an arsenal of delaying measures and misdirection, not to mention occasional strong-arm tactics to enforce industry discipline, its critics say.

But in early November 2016, during the final months of the Obama administration, the CPSC took one of the most significant steps it can take: The commission voted in favor of a rule to force manufacturers to lower their generators’ carbon monoxide emissions. The vote was 4 to 1, with one Republican joining the majority of Democrats.

Donald Trump was elected a week later. In January, he elevated the only commissioner to vote against the rule — Ann Marie Buerkle, a 66-year-old former Republican congresswoman from upstate New York — to be the acting chair of the CPSC, and she took on the role in February. (She awaits Senate confirmation to become the chair.) The administration has nominated, as a potential commissioner, a lawyer from the corporate firm Jones Day who specializes in defending companies from product-liability cases; one other vacancy remains. So Buerkle will likely soon have new allies.

Self-effacing and warm, Buerkle wins universal praise from agency employees and fellow commissioners as a pleasant colleague. Even the Democratic appointees feel she listens to them. Indeed, the CPSC took pride that it wasn’t riven by the partisan rancor that infects so much of Washington. Kaye, a Democrat, brought his family to visit fellow commissioner Buerkle and her family when they vacationed in upstate New York.

Buerkle’s gentle personality, however, belies hardline views on regulation. Buerkle has never, in her fellow commissioners’ recollection, advocated for the agency to regulate a product that the CPSC staff thinks is unsafe. She is a government regulator who doesn’t appear to believe in government regulation.

Voluntary standards are “a better way to go,” Buerkle told ProPublica. “They are quick to complete. There’s much more efficiency in implementation. And there’s much more buy-in from stakeholders.” Never mind that in this instance it took more than a decade, and ongoing government prodding, for companies to get close to adopting a voluntary standard of its own. Pressed on which product hazards are her priorities, Buerkle says: “Fidget spinners are a big deal.” (They should not be placed in the mouth, the commission warns.) She also mentions children’s products and toys, pool safety, and portable generators.

Among her first actions as chair, Buerkle did two things. She sent a letter in August to Scott Pruitt, the head of the Environmental Protection Agency, agreeing with his assertion that the CPSC does not have the legal authority to make a rule about carbon monoxide emissions from portable generators.

In a second move, Buerkle appointed Patricia Hanz to be her general counsel. Hanz comes from Briggs & Stratton, a manufacturer of engines in Wauwautosa, Wisconsin, where she was the assistant general counsel. Briggs & Stratton, which brings in $1.8 billion a year, happens to be one of the biggest portable-generator manufacturers in the world. Hanz also served as the vice president of the portable-generator trade group.

The rise of Hanz and Buerkle — and the issue of portable generators — has injected a new contentiousness into the CPSC. In an impassioned speech in October, decrying the future of her agency, CPSC commissioner Marietta Robinson called Hanz “the one person who fought the hardest against any safety measures” for portable generators.

Hanz declined to comment on a detailed list of questions. In a statement from the CPSC’s public-affairs office attributed to Hanz, she said she has recused herself. “Under the Trump Administration Ethics Pledge I cannot have any involvement with my former employer, nor with PGMA (Portable Generator Manufacturers Association) for two years, including contracts and regulations. In addition, to avoid any appearance of partiality, I will have no involvement in matters related to any PGMA members,” she wrote in an email. Briggs & Stratton also declined to respond to detailed questions. The company said in a statement it “is and has always been committed to safe operation of generators.”

While Trump has achieved few of his legislative priorities, his administration is succeeding in broadly eroding federal regulation. Having fought the long war, the portable generator industry is now poised to benefit from the president’s success. The government’s portable-generator rule has not been finalized — and now with Buerkle at the helm, it probably will never be.

---

The Consumer Products Safety Commission’s efforts to fix portable generators owe their origins to what now seems like a silly panic just before the turn of the century. In the lead-up to the year 2000, people worried that computers wouldn’t be able to process, in their databases, the transition from the year 1999, a theoretical calamity referred to as the Y2K problem. If they went on the fritz, some predicted, the country might face catastrophes like mass blackouts. Regulators worried that people would rush out to buy portable generators, leading to a spike in carbon monoxide deaths.

Y2K passed without incident, and those who purchased generators to protect against the apocalypse sheepishly stored them away unused. But the CPSC realized that it should do something about needless carbon monoxide deaths. Portable generators were killing more people than all heating systems combined.

When the CPSC thinks about regulating products, it considers what it calls a “safety hierarchy.” The best solution is to design a product that is safe. Not all products, of course, can be perfectly safe — cars must go fast and table saws must be sharp. So the second-best remedy in the hierarchy is to mitigate the risk with, say, airbags or hand-guards. The third option is a warning label, but the CPSC staff typically views that as least desirable because people often don’t abide by the warnings.

In 2002, the commission contacted Underwriters Laboratories, a private company that tests product safety and helps develop industry standards, to see if it would help companies make safer generators. Seeing the government moving, the manufacturers began, with reluctance, to place warning labels on the machines. At the time, the companies claimed, a label that cautioned people not to operate the machines indoors was as far as they were able to go: The technology, they asserted, did not exist to make a generator engine that emitted less carbon monoxide.

The CPSC staff was happy enough at the time because third-best is better than nothing. Still, in the case of portable generators, the instructions can produce confusion. They tell consumers not to operate them indoors, but also not to let the machines get wet, which would seem to rule out placing them outside in many cases. Given that people tend to use their generators precisely when weather conditions are snowy or torrential, this is hard advice to follow. Some users solve this conundrum by doing things like putting the generator on their porch or in their garage. Sometimes people have not put their generators far enough outside: More than a quarter of portable generator deaths occur from these sorts of placements.

The companies were not sympathetic, says CPSC commissioner Kaye. His impression of the industry view was: “If consumers are too stupid to read the label and they die, that’s their fault.”

In the short term, adopting the labels had one positive effect — at least for the manufacturers: It staved off stricter regulation for several years. Unfortunately, it was a different matter for consumers. The labels had no apparent effect. The number of carbon monoxide deaths and injuries caused by generators did not decline.

The continuing deaths were disturbing enough that in 2006, under the George W. Bush administration — no fan of regulation — the CPSC decided to take more significant steps. The CPSC began work on a mandatory rule requiring manufacturers to make their machines safer. How the industry got there would be up to them. Companies could develop engines with lower emissions or install switches that automatically shut off the engine when carbon monoxide levels got too high. The agency produced what it calls an “Advanced Notice of Proposed Rulemaking,” announcing its intentions and allowing the public to comment.

Comments flowed in, especially from the industry. In February 2007, manufacturers wrote to the CPSC urging it not to impose a mandatory standard. They said it would be costly. They also raised questions of whether the CPSC even had the power to require engines with low emissions. As for shut-off switches, the companies warned that “the presence of such [carbon monoxide] detections capability may create a false sense of security,” lulling consumers into thinking they don’t need to take other precautions. One of the three signatories was Patricia Hanz, then working for Briggs & Stratton.

But with the warning label clearly no longer placating the government, the industry gave some ground, agreeing to begin developing its own voluntary standard in the hopes of avoiding a stricter mandatory standard from the government.

For the next three years, little happened. Underwriters Laboratories, today known as UL, again tried to help the industry come up with its voluntary standard, but in 2009, as UL worked toward guidelines manufacturers considered too stringent, they decided to go their own way.

The companies created their own trade group, the Portable Generator Manufacturers Association. The PGMA launched its own initiative to create a voluntary standard. But the CPSC remained skeptical. One problem was that the PGMA’s members don’t sell all the generators purchased in the U.S. Other makers wouldn’t be bound by anything that body came up with. (Today, the PGMA says it represents 80 percent of the market, which manufacturers I spoke to estimate to be anywhere from $700 million to $1 billion.)

The following year — a decade after the CPSC began trying to prevent deaths from carbon monoxide poisoning — the PGMA held its first technical committee meeting to discuss the safety of portable generators. Among the attendees was Michael Gardner, the vice president of new product development at Techtronic Industries, a maker of generators and other products (including Hoover and Dirt Devil vacuum cleaners). He waited for a discussion of carbon monoxide emissions. And waited.

Nobody ever brought them up. “The one topic the technical committee was not talking about — the elephant in the room — was the roughly 70 people dying [each year],” Gardner says. “The technical committee was established to write the safety standard but it did not include carbon monoxide in that standard.”

Carbon monoxide poisoning is far and away the chief safety concern from portable generators. Nothing else comes close. But Hanz, then on the PGMA board of directors, argued that the industry should figure out all the easy things first, rather than tackle the most contentious issues right away.

With the voluntary standard going nowhere, the CPSC decided it needed to conduct the manufacturers’ research and development for them. It sent out a proposal asking for engineers to try to solve the problem. The University of Alabama answered, and came up with a prototype for a safer generator. In October 2012, the CPSC gave a technology demonstration to the industry showing the new engine could lower emissions by more than 90 percent.

Many of the companies scoffed. They argued the prototype was unproven and unreliable, and that the University of Alabama results were obtained in unrealistic conditions. They even suggested the new engine might be dangerous, giving users the sense that they needn’t worry. One company, however, embraced the new ideas: Techtronic. “That was a tipping point for us to say it can be done,” Gardner says.

Not until September 2014 did the PGMA release an initial voluntary technical safety standard for manufacturers. The detailed list included provisions for durability (generators had to survive being dropped from a height of eight inches onto a concrete surface); temperature tolerance (wind speed less than 6.7 miles per hour during testing), and rain resistance (a generator must be soaked, wiped off, and then run for 15 minutes).

But the standard was mute about the emissions that could kill customers. The CPSC sent a politely worded letter in January 2015 that the “staff notes with concern” that the proposal’s only mention of carbon monoxide was in the context of warning labels and external carbon monoxide monitors. On average in the three years through 2012 (the most recent year for which official data is publicly available — the government’s death-count lags as the tally only becomes official years later), carbon monoxide from portable generators killed 63 people per year. The CPSC’s displeasure carried a threat to the industry: If it didn’t get moving, the commission would move forward on its own mandatory standards effort.

To confront this threat, the industry countered with one of its own. In March 2014, the PGMA sent a letter to the commission insisting that it was “not appropriate” for the CPSC “to establish a working group” on the issue, statements that carried the implication it might sue the government if the commission tried to implement a mandatory rule. At that point, the PGMA finally formed two groups of its own to examine two engineering solutions: a low-emissions engine based on the University of Alabama prototype and an automatic shut-off valve. Most of the manufacturers, if pressed, preferred adding shut-off valves; they’re much cheaper. After two meetings, the PGMA picked that option and discontinued the group aimed at looking at lowering emissions.

A year later, the PGMA was ready to show off its progress. In March 2016, the trade group hosted a technical summit. Representatives from manufacturers flew in from all over the country. But only one PGMA member made a presentation regarding a new low-emission generator: Techtronic’s Gardner. In front of a crowd that included manufacturing employees, lobbyists, CPSC staff, and one CPSC commissioner, Republican appointee Joseph Mohorovic, Gardner demonstrated that the low-emissions technology worked. Techtronic’s engine produced 90 percent less carbon monoxide than a similar machine.

When Gardner finished, he looked around the room and heard nothing but silence. Then, in a scene out of “Citizen Kane,” he heard one lone person begin to clap: Mohorovic. After the talk, the commissioner stood up, came over, and shook Gardner’s hand.

At the meeting, someone asked whether companies were going to introduce a low-emissions engine, and if so when. Briggs & Stratton’s Hanz said it would be many years, perhaps five or six. Not so, Techtronic said. It was planning to introduce one in 2017.

---

Finally, in October of last year the time had come. American society makes bargains with its machines. The country is willing to pay, sometimes in lives, for less expensive and more convenient products. But this deal seemed, to the government, too costly.

To make any rule, the CPSC is required to conduct a cost-benefit analysis. In this case, the commission estimated that new portable generator restrictions would save $145 million annually, accounting for the government-computed value of the lives saved. (The industry took issue with those figures.) The commission rarely votes on mandatory rules. Indeed, Congress requires the agency to try to get industries to implement voluntary standards first. But the portable generator makers had not done so.

The commissioners concluded they had no choice. They voted on a Notice of Proposed Rulemaking (NPR), a big — but not final — step toward a mandatory rule. It required manufacturers to build low-emissions machines. The CPSC staff said it was the best alternative. After investigating four different shut-off technologies, they determined that option was “not feasible.” With Mohorovic joining the Democrats, the vote was 4 to 1.

“Our staff engineers spent well over a decade trying to get manufacturers to make portable generators safer and all but a couple steadfastly refused to do so,” says Marietta Robinson, a commissioner appointed by President Obama. “This NPR was absolutely essential in forcing manufacturers to do the right thing to save lives.”

Meanwhile, the manufacturers were still dithering on a voluntary standard. They couldn’t even agree on how to measure how much carbon monoxide their engines emitted. In April of this year, UL tried to take a preliminary step; it proposed a uniform method for measuring emissions. Techtronic lobbied for it. Briggs & Stratton pushed against it. Things got heated: At one point, a Briggs & Stratton employee labeled Techtronic a “rogue company” in an email circulated to PGMA members. During the voting, some manufacturers changed their position, from being for the standard to being against. Briggs won and the measure was narrowly defeated.

Today, with Buerkle as chair, the fate of the mandatory standard seems doomed. She has already voted against it once, and stated that the CPSC has no power to regulate generators’ carbon monoxide emissions. Installing Hanz as general counsel has only made some CPSC officials more pessimistic about regulating portable generators. Buerkle says Hanz is qualified and that her appointment was fully vetted by ethics officials.

Nearly two decades after carbon monoxide concerns first arose, the industry says it is close to a voluntary standard. Even this move has twists and turns. PGMA is working on a less stringent guideline, only requiring a shutoff sensor. It is aiming to have it completed by this year.

At the same time, UL circulated a competing, stricter measure, requiring both a shutoff sensor and a low-emission engine. Again, members lobbied intensely. The nay votes had a surprising supporter: Mohorovic. The Republican commissioner has left the CPSC and taken a job at the law firm Dentons. And he has flipped from the position he held as a public servant. He now is against a low-emissions engine, sending a series of emails attacking UL’s standard and urging members to vote against. “There is no evidence, data or modeling” that UL’s standard will save lives, he wrote recently in an email to one voter. (He did not respond to questions about why or who his client was.)

Buerkle, too, advocates for a shutoff sensor. She contends they are the safer alternative. “I’m told by CPSC staff,” she says, that “shutoff technology will be far more effective” in saving lives. She says the staff told her that shutoff switches would reduce the figure by somewhere around 99 percent.

But in fact that 99 percent figure is an industry estimate. The CPSC acknowledges that it has neither calculated how many deaths the shutoff valves will save, nor concluded that they’re more effective at saving lives than low-emission technology. The staff at the commission has worried that carbon monoxide can migrate away from a machine such that it kills without ever triggering the shutoff switches. And staffers worry that the switches might trigger bothersome shut-offs, leading consumers to disable the switches. The industry says it has resolved those concerns, but Techtronic takes issue with the figure. Calling PGMA’s 99 percent claim “misleading,” the company says it has not been peer-reviewed and doesn’t account for many scenarios when portable generators are used.

Buerkle is satisfied with the industry’s progress: “I’m just happy where we are right now. We are on the verge with technology that will save lives.”

Others at the CPSC remain skeptical. “I am extremely concerned that backing off now, as our chair and general counsel have made clear they intend to do, will, at a minimum, delay these life-saving efforts and, perhaps stop them altogether,” says Robinson, the Democratic commissioner whose term has expired. (She will leave when her replacement is confirmed.) But with Buerkle’s ascendancy, there’s not much Democratic commissioners can do.

The industry says some manufacturers aim to start marketing machines with shutoff switches next year. “This should be a good news story that industry stepped up to do the right thing,” says Edward Krenik, a lobbyist who represents the PGMA. “Though it took a while.”

For its part, Techtronic kept its promise to put a low-emissions generator on the market. They’re available now — and they sell for less than many other competing products. But to the industry, that makes Techtronic an outcast.