Good data about where the coronavirus is spreading, whom it’s affecting, and the capacity of health systems is essential for government officials to make better decisions on, for example, lockdown orders and where to send supplies. Epidemiologists and infectious disease experts can use the information to understand COVID-19 better and, more importantly, to contain it.

But over the past year, experts have watched with bewilderment at how the Trump administration has collected, distributed and used public health data. “We’re still flying blind,” said Ryan Panchadsaram, co-founder of the the United States Digital Response, an all-volunteer effort to collect and publish COVID-19 data that the federal government hasn’t.

An early controversy erupted in May, when White House economist Kevin Hassett produced a set of charts on COVID-19 spread that appeared to predict a rapid decline in cases and deaths. In them, the country appeared to be nearly virus-free by summer. Hassett distributed the charts internally—much to the delight of President Donald Trump, according to The Washington Post. When Hassett published the charts on an official White House Twitter feed, experts reacted in horror and said reopening prematurely was likely to cause a new surge in cases, which is precisely what happened.

One month later, the administration abruptly decided to make a major change in how hospitals report their inpatient COVID-19 data. No longer would they submit information to the Centers for Disease Control and Prevention. Instead, hospitals would file data directly (or indirectly through state governments) to the Department of Health and Human Services.

Administration officials said the change would allow them to collect more complete and standardized data. A chorus of angry public health officials, scientists and hospital representatives warned in a public letter that the transition was likely to lead to less information, not more, precisely when the information was needed most.

Sure enough, data on the HHS website was incomplete and lagging for several weeks, in part because the change came with almost no advance warning.

“Usually you’ve got months and months, sometimes years, to prepare for big changes like this,” said Chip Kahn, president of the Federation of American Hospitals. “Here, you had days.”

Larger hospitals and those that were already using the HHS portal were mostly able to adapt. Smaller hospitals frequently struggled because they hadn’t been using the new system and because they didn’t have the dedicated IT staff of larger hospital networks. “The majority of our hospitals in Nebraska are rural, and the staff at these hospitals are often responsible for many roles, which can stretch resources very thin,” said Margaret Woeppel, vice president for quality and data at the Nebraska Hospital Association. “The nurse who is taking care of patients may also be responsible for quality reporting and infection control.”

Panchadsaram, who served as deputy chief technology officer during the Obama administration, said the quality of the data is slowly improving. But he identified another problem: The Trump administration isn’t sharing it with the public.

Every week, the White House coronavirus task force puts together a detailed status report on the outbreak and the response for all 50 states. It includes everything from test positivity rates to a breakdown of which communities have rising caseloads, along with policy recommendations. The task force gives each governor a copy of their state’s report, but otherwise doesn’t make them public. (The few copies that have surfaced have been obtained by reporters and the Center for Public Integrity, which is posting the reports it obtains on its webpage.)

The White House has not responded to questions about why the reports remain private, leaving critics to wonder whether administration officials are reluctant to publish information that might contradict the optimistic outlook of, say, the president or some Reublican governors. “They have a really good analytics team creating these beautiful reports, those reports have really important recommendations on them, and they're not making them public,” Panchadsaram said. “Maybe that is because the reports clearly contradict what governors are saying, that it’s under control when it’s not.”

For over 40 years, the White House has published economic forecasts, including on growth and unemployment, as part of its midyear budget update to Congress. But this year, the White House decided to omit such projections. Officials told The Washington Post that the volatility caused by the pandemic made it difficult to model economic trends and that “there is no statutory requirement to release this information, just precedent.”

Douglas Holtz-Eakin, a former director of the Congressional Budget Office and a Republican adviser, criticized the move for getting the White House “off the hook for having to say what the economic outlook looks like.”

Trump has repeatedly predicted a rapid rebound for the economy ahead of November’s election, but projections by the CBO, Federal Reserve and outside economists suggest that a full recovery may take years. As of early October, more than 12 million Americans were formally unemployed—double the number before the pandemic struck—and another 7 million Americans wanted a job but were not actively seeking employment. Congress and the White House have been unable to agree on the size and direction of a second stimulus bill that would aid workers and businesses.

The Trump administration refuses to publicly disclose where and how it distributed medical supplies through Project Air Bridge, a federally funded, multimillion-dollar initiative that flew personal protective equipment from overseas and delivered it to front-line health care workers across the U.S. The administration also has yet to disclose details related to the program’s cost and the pricing arrangements the government made with the handful of private companies involved.

The White House and the Federal Emergency Management Agency have repeatedly touted Project Air Bridge as a major success. But the secrecy around the program makes such claims difficult to assess. A Washington Post review of available data found that White House officials, including Vice President Mike Pence, have on numerous occasions overstated the volume of supplies Project Air Bridge actually delivered. Reporting from the Los Angeles Times has raised questions about the project’s lack of transparency on pricing and preferential treatment for big firms.

In April, Senators Elizabeth Warren and Richard Blumenthal opened an investigation into Project Air Bridge, citing worries about price-gouging from third-party sellers, reports that federal officials had seized supplies, and concerns that the process for distribution appeared to be “driven by politics rather than public health.”

Federal officials have claimed that they consider details such as the names of hospitals, health care providers and local agencies that received the supplies to be proprietary information and therefore not subject to disclosure. The Post reports that some state officials said FEMA told them they could not share information about supplies they received through the program.

One of the Obama administration’s most ambitious environmental policies was an unprecedented drive to reduce emissions of methane, the main component in natural gas and a major contributor to climate change. Two days after Donald Trump was elected president, Barack Obama's Environmental Protection Agency began the process of gathering wide-ranging data from about 15,000 owners and operators of oil and natural gas facilities. Earlier that year, the EPA had introduced regulations on methane emissions at new or modified oil and gas operations. This data was intended to provide the basis for regulating emissions at existing operations, which produce far more methane.

In March, new EPA Administrator Scott Pruitt withdrew the requests—a move that came as no surprise. As attorney general of Oklahoma, Pruitt had joined a group of state AGs to sue the EPA in 2016 over the first round of rules on methane emissions. (The EPA weakened those regulations in August.) Pruitt’s decision to cancel the data collection signaled that, under Trump, the EPA would axe the plans to regulate existing facilities.

Methane dissipates much more rapidly than carbon dioxide, but the gas is at least 86 times more potent during its first 20 years in the atmosphere. The EPA estimates that the oil and natural gas industry is responsible for nearly 30 percent of U.S. methane emissions. A 2018 study found that the EPA may underestimate the industry's methane emissions by as much as 60 percent. Trump’s EPA has launched its own programs to collect information from the industry, but companies are not required to participate, and if they do, they only need to report successful efforts to reduce methane emissions.

In 2012, the Obama administration embarked on a plan to nearly halve planet-warming emissions from new cars and light trucks by 2025. The joint rules from the Environmental Protection Agency and the Department of Transportation progressively tightened both fuel economy and emissions standards, a program the EPA described as “the most significant federal action ever taken to reduce greenhouse gas emissions and improve fuel economy.”

Donald Trump, who had campaigned on promises to revive the auto industry, declared just two months into his presidency that he intended to roll back that rule. In 2018, his EPA and DOT issued a proposal to freeze the standards at 2020 levels through 2026. That would have meant keeping the fuel economy standard at an average of about 37 miles per gallon, rather than increasing it to 54 miles per gallon by 2025.

The administration claimed this would serve both the economy and public safety—and it appeared to have the numbers to prove it. Its proposal said that Trump's rule would have a net economic benefit of up to $201 billion and would prevent nearly 12,700 deaths in car accidents.

There was just one problem: The calculations were chock-full of errors that skewed the outcome to Trump's benefit. A group of prominent economists described the analysis as being "at odds with basic economic theory and empirical studies." Together with DOT officials, Trump appointees at the EPA had blocked the agency's engineers from contributing to the study, The Atlantic later reported. When EPA staffers warned the White House about problems with the analysis, their criticisms were ignored.

The Trump administration's final rule, which was released in March 2020, does not entirely freeze the fuel economy and emissions standards. Instead, it requires automakers to increase fleet fuel economy by 1.5 percent annually—still a dramatic cut from the Obama standards, which would have been about 5 percent.

The accompanying analysis was radically different as well. Rather than the massive benefit claimed in the draft, the final version projects that Trump's rollback could actually cost the economy as much as $22 billion, as compared to the Obama-era standards. Some critics find even this far less favorable conclusion dubious. In June, a coalition of advocacy groups petitioned the EPA to stay the final rule, writing that it is "premised on an analysis riddled with errors and inconsistencies."

Dave Cooke, a senior vehicles analyst at the Union of Concerned Scientists, said some of the errors show evidence of the administration "putting its thumb on the scale towards a pre-ordained outcome." For instance, the petitioners claim that the new analysis added $27 billion in savings under the Trump rule by inflating the "cost of congestion"—an estimate of traffic's negative economic impact—under the Obama standards.

As with the draft, the administration was accused of sidelining EPA staff and ignoring their criticisms. In July, the EPA's inspector general announced that it would investigate "potential irregularities" in the agency’s work.

In 2019, the Donald Trump-appointed director of the U.S. Geological Survey ordered agency staff to only use climate models that project the impacts of emissions through 2040 in their scientific assessments, according to The New York Times. Models with projections through 2100 had typically been used previously.

Scientists told the Times that the new assessments will give a misleadingly optimistic projection of the future—barring changes in current global emissions, models suggest the most severe impacts of climate change will start to be felt after the 2040 cutoff, around 2050.

Decision-makers such as resource managers, infrastructure and city planners, emergency response professionals, elected officials, farmers and the public rely on USGS, the sole science agency for the Department of the Interior, to produce information that provides the scientific foundation for planning around environmental threats to lives and livelihoods.

The U.S. government has a calculation for the “social cost of carbon,” which measures the environmental, economic and public health damages associated with emitting carbon dioxide into the atmosphere. Federal agencies use this figure to evaluate the costs and benefits of new climate regulations. At the end of President Barack Obama’s tenure in office, that figure was $40 per metric ton (a number environmental groups argued was too low). The number is expected to increase as the planet experiences rapid global warming.

But shortly after taking office, President Donald Trump signed an executive order that scrapped the Obama-era calculation, putting the economic harm of emitting climate-warming carbon dioxide far lower, at between $1 and $7 per ton. A July report by the Government Accountability Office determined that Trump’s calculations of the social cost of carbon are about seven times lower than previous federal estimates. And while Obama-era estimates placed future climate damages at $82 per ton of carbon by 2050, Trump’s calculation put it at just $11, the report found.

The lower figure allows the administration to argue that the upfront costs of reining in greenhouse gas emissions outweigh any future benefit to property, agriculture, coastal ecosystems and human health. Trump's EPA has used the number to justify weakening environmental safeguards and to roll back Obama-era regulations that slashed carbon pollution from power plants and new vehicles. For example, the EPA claims that rescinding the Clean Power Plan would save the U.S. $33 billion in “avoided costs” by 2030, a calculation it came to by looking only at the domestic costs of carbon pollution and by downplaying the impact that emissions will have on future generations.

Meanwhile, the Trump administration has led a campaign to undermine the validity of climate modeling so it can ignore possible environmental threats when evaluating new infrastructure, such as pipelines, power plants, highways and airports. When more than a dozen federal agencies published a sobering 1,600-page report in November 2018 that warned of the dire threat of global climate change, then-White House press secretary Sarah Huckabee Sanders claimed the assessment was “based on the most extreme model scenario” and “not data-driven.” Trump flatly denied the science: “I don’t believe it,” he told reporters.

Dramatically reducing immigration to the United States has been one of the Trump administration’s most cherished goals. In March 2017, as the White House prepared to announce the maximum number of refugees the president would accept during the next year, he ordered a report on the long-term costs of offering a new home to victims of persecution and conflict abroad.

The Department of Health and Human Services produced a comprehensive 55-page draft. But far from showing that refugees were a drain on government finances, as immigration hardliners contended, it found that refugees had contributed $63 billion more in government revenue than they had received in benefits between 2005 and 2014.

That study never reached the White House. Instead, HHS handed in a three-page report that ignored refugees' fiscal contributions entirely. According to The New York Times, Stephen Miller, President Donald Trump's senior policy adviser and ringleader of the anti-immigration contingent, sought to keep the subject out of deliberations about the refugee cap. Miller reportedly told people at HHS: “The president believes refugees cost more, and the results of this study shouldn’t embarrass the president."

In the end, the administration announced that it would admit a maximum of 45,000 refugees in the 2018 fiscal year—then a historic low. In 2021, the U.S. will accept no more than 15,000.

Four days after President Donald Trump’s inauguration, the Department of Health and Human Services posted a draft revision of an annual survey it conducts to improve its services for disabled Americans. One month later, when the agency released new proposed revisions to the survey, advocates discovered that a single edit had quietly been made: The proposed question asking for the respondent’s sexual orientation had been deleted.

As a group of senators led by Republican Susan Collins pointed out, removing this question significantly limits HHS’s understanding of the specific challenges faced by LGBTQ people with disabilities—including financial insecurity and discrimination that prevents them from accessing necessary services. "The only way to have the evidence you need to prioritize and spend wisely to address disparities is to have data about those disparities," a Johns Hopkins researcher told Politico.

An administration official told The Associated Press that Trump political appointees at HHS had reviewed the survey and targeted only the LGBTQ question for removal.

In 2016, after a protracted political fight, the Obama administration scored what looked like a win in its fight for equal pay for women and people of color: The Equal Employment Opportunity Commission finally got the go-ahead to start collecting pay data from large companies that was organized by gender, race and ethnicity.

The EEOC is supposed to enforce federal antidiscrimination laws, and has long collected demographic data from companies that helps the agency detect discriminatory hiring patterns. Pay was a logical addition: A similar effort in the United Kingdom revealed significant differences in how much men and women earn in a surprising number of fields. Simply forcing companies to collect and report this kind of information might even push some to address disparities, experts reasoned.

Alas, in the summer of 2017, the Trump administration nixed the whole thing. In a short memo, the Office of Management and Budget claimed with no evidence that collecting pay data would be too burdensome for companies and was of dubious utility. Ivanka Trump, a supposed proponent of equal pay, put out a statement supporting the move.

In March 2019, a judge ruled that the administration had acted arbitrarily and that companies needed to turn over two years worth of pay data. Incredibly, most companies produced the numbers by the summer of 2020. And yet, the ruling doesn’t compel the EEOC to actually do anything with this trove of never-before-collected information.

And so it hasn't. In what seems like an obvious stalling tactic, all the agency has done so far is announce that it will convene a panel of experts—merely to examine whether the data is of any use. Meanwhile, in the U.S., women on average continue to make 82 percent of what men earn and Black women make only 62 percent.

The Trump administration has proven itself an ally of chemical companies, at times sidelining science to preserve the industry’s ability to use toxic chemicals.

President Donald Trump’s Environmental Protection Agency refused to ban the agricultural use of chlorpyrifos, an insecticide linked to brain damage in children that has been used on a variety of crops since the 1960s. It’s a move that defied the EPA’s own scientists, who in 2015 recommended banning the pesticide due to its potential health risks. In September 2020, the agency released a new assessment of chlorpyrifos that excluded a number of epidemiological studies, including one showing exposure to the chemical in the womb is associated with abnormal brain development.

The chlorpyrifos example is part of the Trump EPA’s broader push to crack down on so-called “secret science.” It’s limiting the types of scientific studies that can be used to craft new policy in the name of "transparency."

The Trump EPA also blocked the release of a yearslong study of the health risks associated with environmental exposures to formaldehyde, a widely used chemical known to cause cancer in humans. And it cut off funding for more than a dozen research centers around the country that study, among other things, childhood leukemia and how exposure to chemicals and industrial pollution impact child development.

If there's no data about how these exposures might hurt Americans, well—then what we don't know can truly hurt us.

Even in cases where the Trump EPA does have the data, moves to hide it have often left the public in the dark. Take, for example, the efforts the agency often touts to address per- and polyfluoroalkyl substances, or PFAS, the cancer-causing “forever chemicals” that contaminate the drinking water of millions of Americans. In 2018, the EPA and White House attempted to block publication of a study into widespread PFAS contamination after a Trump aide warned it would be "potential public relations nightmare," according to an explosive Politico report.

That study, later released in June 2018, found that the chemicals threaten human health, even at levels seven to 10 times lower than EPA’s recommendations. Meanwhile, the Trump EPA has continued to approve new varieties of PFAS with little to no testing.

Since the early days of Trump’s administration, his Environmental Protection Agency has been on a mission to neuter Obama’s landmark air pollution restrictions. One tactic has proved especially consequential: changing the way that agencies evaluate the economics of environmental regulations. If, on paper, the cost of a regulation suddenly outweighs its benefits, the administration can argue that the rule is too expensive to keep.

Changing these calculations is at the center of the EPA’s move to weaken a 2011 regulation curbing power plants’ mercury emissions, which can cause lasting brain damage in children and fetuses. After the rule was introduced, utility companies spent billions to meet the standards or switched to natural gas or renewable energy. Over the next six years, mercury emissions from power plants fell by more than 80 percent.

But the rule had a vulnerability. To withstand legal challenges, Obama's EPA had to establish that its new emissions restrictions were "appropriate and necessary." Its most powerful arguments were about the huge public health gains: Curbing mercury pollution would slash other forms of power plant pollution as well, potentially preventing 130,000 asthma attacks, 4,500 heart attacks, and up to 11,000 premature deaths annually. The agency projected that the rule could save society up to $80 billion each year.

To reach the opposite conclusion, Trump’s EPA simply decided to ignore almost all of the rule’s health benefits. EPA Administrator Andrew Wheeler argued that the agency should only measure the value of reducing mercury pollution, not any other indirect benefits that the rule would produce. After this mathematical sleight of hand, the rule was only projected to save $6 million annually—too little, the agency now said, to justify regulating mercury emissions. For now, the regulation remains intact, but the new assessment leaves it wide open to lawsuits from coal companies.

Former EPA Administrator Carol Browner warned that the move was a "sinister Trojan horse." Sure enough, in June, Wheeler proposed a new rule that would give little or no consideration to the indirect benefits of all pollution regulations under the Clean Air Act. “These economic cost-benefit analyses have been an important driver of Clean Air Act regulations for 40 years,” former EPA official Richard Morgenstern told The New York Times. “What this rule is doing is altering the math in such a way to potentially downplay the economic benefit to public health, so they are justified in writing weaker rules in the future.”

Black people are dying from the coronavirus at a rate more than twice that of white Americans. One likely reason for this inequity is the very different air we breathe. People of color are more likely to live near highways and other sources of soot pollution—also known as fine particulate matter—which can cause health conditions that make people more vulnerable to the virus.

And yet on April 14, EPA Administrator Andrew Wheeler announced a preliminary decision not to issue tougher standards governing how much soot can be released into the air—a move that was cheered by the fossil fuel industry. Wheeler insisted that he was simply following the science. In December 2019, a majority of the Clean Air Scientific Advisory Committee—a group of outside experts that advises the EPA—recommended leaving the standards for fine particulate matter unchanged. However, in the draft report that CASAC had reviewed to reach its recommendation, EPA scientists had advised tightening them, on the logic that long-term exposure to soot pollution was killing up to 52,100 people in the U.S. annually. As Gretchen Goldman, a research director at the Union of Concerned Scientists, explained to The Washington Post, “It’s not surprising they would retain the standards, because they broke the process.”

There are hundreds of advisory committees throughout the U.S. government, and they are part of a time-honored system that helps ensure that policy decisions are based on sound research. President Donald Trump’s administration has systematically sidelined, frozen and eliminated many of these committees.

The attack was particularly fierce at the EPA. Trump's first administrator, Scott Pruitt, and his successor, Wheeler, have replaced scientists on influential EPA committees with members sympathetic to the industries the agency was supposed to be regulating. Pruitt’s hand-picked chair for CASAC had worked as a consultant for groups such as the American Petroleum Institute and openly questioned whether air pollution causes serious health problems. For decades, CASAC had also convened panels of additional experts to help review the science on specific pollutants. Wheeler disbanded the panel that was assisting with the particulate matter standards. Six months later, CASAC said it lacked the expertise to carry out its review. But rather than reconvening the panel, Wheeler appointed a dozen new consultants. One of only two academic scientists on the committee pointed out that not a single expert in air pollution epidemiology was involved in the review. Inside Climate News reported that the recommendation to retain the soot standards partly rested on a widely criticized study based on unverified data.

Meanwhile, in May 2017, then-Interior Secretary Ryan Zinke temporarily halted the work of his department’s 200-plus committees. That year, committees dedicated to scientific and technical issues met the fewest times in any year since record-keeping began in 1997, according to an analysis by the Union of Concerned Scientists. At both the Department of the Interior and the National Oceanic and Atmospheric Administration, panels on climate science were shut down entirely, as was a food science panel at the Food and Drug Administration and a forensic science committee at the Department of Justice.

In June 2019, Trump took things a step further. The president instructed federal agencies to eliminate at least one-third of their committees and capped the total number across the government at 350. So far, the EPA, the Department of Energy and the Department of Commerce have each eliminated a handful. The Centers for Disease Control and Prevention shut down three.

A look at two of the former CDC panels reveals the vital functions that advisory committees perform behind the scenes. For example: In recent years, evidence has mounted that glyphosate, an herbicide used in weed killers, may cause cancer. The CDC does not collect countrywide data on Americans' exposure to glyphosate, despite having a system—the National Biomonitoring Program—that does just that for more than 300 other dangerous chemicals. So a now-disbanded committee that advised both the Agency for Toxic Substances and Disease Registry and the CDC's National Center for Environmental Health was working to have glyphosate and other herbicides added to the list, according to former member Dr. Aaron Bernstein, a pediatrician and interim director of the Center for Climate, Health, and the Global Environment at Harvard University’s school of public health.

The CDC also shut down its Advisory Committee to the Director, which consulted with agency leadership on big-picture strategy: Notably, it had advised CDC leaders during the Zika and Ebola outbreaks. “I'm looking at what's happening right now with COVID-19," said Dr. Lynn Goldman, a former committee member and dean of George Washington University's school of public health, "and I can't believe that the CDC director does not have a council of outside advisers that's able to serve as a sounding board for him.”

In 1996, an enterprising tobacco industry lawyer named Chris Horner sent a memo to a client detailing a plan to stymie the Environmental Protection Agency’s crackdown on second-hand smoke. The idea was to “construct explicit procedural hurdles that the Agency must follow in issuing scientific reports”—in particular, requiring that agency data be transparent and independently reproducible. Sounds totally reasonable, right? In reality, it would disqualify a lot of peer-reviewed research. Steve Milloy, a lawyer with deep ties to tobacco and fossil fuel companies, became one of the biggest proponents of this war on so-called “secret science.”

Fast forward more than two decades, and this industry playbook is set to become standard operating procedure at the EPA. In April 2018, then-Administrator Scott Pruitt unveiled a proposed “transparency” rule to limit the types of scientific studies that the agency can use to craft new policy. It allows the EPA to ignore studies that don’t make their raw data public, effectively giving more weight to industry-backed research.

Most significantly, the latest version of the rule would disqualify public health studies that rely on keeping certain personal and medical information confidential. (This includes, for example, research on respiratory health that could be key to combating COVID-19.) The goal, experts say, is to make it harder for the government to enact more stringent protections for our air, water and environment.

Horner and Milloy, who served on Donald Trump’s EPA transition team, were both on hand for Pruitt’s announcement; Milloy has called it “one of my proudest achievements.” (The Interior Department later consulted him on a similar initiative.) The agency intends to finalize the rule this year—and although it is all but certain to face legal challenges, it would likely take a future Democratic administration months, if not years, to undo.

Every five years, the U.S. Department of Agriculture and the Department of Health and Human Services ask a panel of experts to survey the most recent science and recommend what foods Americans should eat and drink to promote health and avoid chronic disease.

The panel offers sober, scientific advice, but it can cause political uproar as the food industry clashes with nutrition advocates in hopes of shaping the final recommendations. Such was the case in 2015, when the federal panel called for eating less red meat.

For the latest round, the agencies dictated the topics the expert panel would address. The 80 questions the panel received omitted issues of red and processed meats, which mounting evidence links to diseases like colorectal cancer, and salt consumption, which studies show can affect cardiovascular health. They also excluded the effects of ultraprocessed foods, consumption of which is increasing.

Columbia University legal experts called these issues “the most critical questions that need to be addressed in the guidelines"—and, coincidentally, said they “represent the issues that large food companies find most objectionable."

The White House also limited the research that could be considered in studies the agency appointees had vetted, preventing the panel from weighing the full body of recent findings on health and food. During the last guidelines review in 2015, science from outside the agencies accounted for 50 percent of what the panel considered.

The administration also stacked the panel with experts from industry groups like the National Potato Council, National Cattlemen's Beef Association and the American Beverage Association. Roughly half had ties to the International Life Sciences Institute, a notoriously bombastic industry group funded by agribusiness and pharmaceutical corporations.

The result was a narrow set of recommendations, such as telling Americans to limit their added sugar intake to 6 percent of daily calories, down from 10 percent. Yet even that is drawing a backlash. In September, the conservative Heritage Foundation urged the Trump administration to “certainly not reduce the recommended limit,” because doing so would “defy logic.”

When the U.S. Department of Agriculture announced in 2019 that it was relocating key research staff to Missouri, it may have seemed like a real estate problem. The agency claimed moving would help attract and retain qualified staff, put USDA resources closer to stakeholders, and save taxpayers money. But the move has blown a crater in divisions that produce essential data.

Rather than uproot their families across the country, 75 percent of the staff at the National Institute of Food and Agriculture and the Economic Research Service simply quit or retired early, according to a Congressional Research Service report from May.

NIFA's main charge is annually awarding $1.5 billion in external grants to support agricultural research and education across the country. ERS is the USDA's main internal research agency, and it has an $85 million annual budget to support economic and statistical analyses on issues from crop prices to food safety and nutrition. The relocation has delayed the delivery of funds and the publication of new research.

A group of 19 Democratic senators raised alarms in a June 2019 letter to the USDA's Inspector General, noting that this data informs "everything from the way our safety net programs work to the planting decisions made by our nation’s farmers." The letter also called for an IG investigation, warning that without one, the actions could be seen as "suppressing, undermining, discounting, and wholesale ignoring scientific data produced by their own qualified scientists."

"If you eat, you're involved in agriculture," Laura Dodson, an agricultural economist at USDA and union leader at the American Federation of Government Employees, told NPR. "And this is the government making itself dumber about agriculture."

With a deregulatory agenda, the Trump administration has cut corners and undermined science in an unwavering quest to encourage private industry's exploitation of public lands.

One of the most blatant examples of this is the administration’s rush to open Alaska’s pristine Arctic National Wildlife Refuge to oil drilling. Last year, investigative journalist Adam Federman obtained leaked documents that show how top Interior Department officials ignored, altered and scrubbed career scientists’ work in an effort to downplay the potential environmental impacts of future drilling in the refuge’s 1.56 million-acre coastal plain. Changes included modifying one biologist’s conclusions about the impact of seismic exploration surveys on polar bears, downgrading them from “uncertain or potentially harmful” to “less than significant,” Federman reported.

Interior political appointees also cherry-picked data about carbon emissions from the 2018 California wildfires to justify increased logging on federal lands, The Guardian reported in January, citing internal agency emails. Bureaucrats downplayed fossil fuel emissions and exaggerated the pollution released by the California blazes, in what one forest ecologist described to The Guardian as “blatant political manipulation of science.” In December 2018, President Donald Trump signed an executive order to boost logging on more than 4 million acres of federal lands to combat wildfires—an approach forest experts say won’t work.

In 2017, not long after overturning an Obama-era moratorium on new coal leases on federal land, then-Interior Secretary Ryan Zinke abruptly canceled a government-funded study into the potential health effects of mountaintop removal coal mining. And in 2018, the Forest Service nixed a federal environmental review of mining near Minnesota’s Boundary Waters Canoe Area Wilderness, thus reviving a Chilean firm’s stalled plans to build a $1.7 billion underground copper-nickel mine just a few miles from the most visited wilderness area in the country. The Trump administration subsequently renewed the company’s leases in May 2019, more than two years after the Obama administration revoked them over concerns that the mine could spoil the wilderness and crush tourism in the area.

Republicans have tried for years to weaken the Endangered Species Act, and disdain for the 1973 conservation law runs deep at President Donald Trump’s Interior Department. Agency officials have previously called it “a sword to tear down the American economy” and likened species listings to “incoming Scud missiles.” In 2017, the director of the southeastern region of the U.S. Fish and Wildlife Service issued a directive that set a target of delisting, downlisting or preventing the listing of 30 species per year, Pacific Standard reported.

The Trump administration has frequently used flawed or incomplete data to strip away or deny species protections, gradually chipping away at the landmark conservation law. FWS denied the Pacific walrus protections in 2017, suggesting that the species would adapt to the rapid loss of sea ice it depends on for resting, feeding and giving birth. The mole skink, a small lizard found in coastal areas of the Florida Keys, was also denied protection that year—even though FWS research shows it could lose 44 percent of its habitat to sea level rise by 2060, and as much as 74 percent by 2100. In both cases, the agency determined that the species were unlikely to become endangered in the “foreseeable future,” which it limited to 2060 for the walrus and to a 30- to 40-year timeline for the skink.

In 2018, the Trump administration delisted grizzly bears in and around Yellowstone National Park, arguing that multiple factors indicated the population “is healthy and will be sustained into the future.” The final rule claimed “the effects of climate change do not constitute a threat to the [Yellowstone] grizzly bear [population] now, nor are they anticipated to in the foreseeable future.”

Environmental groups sued. They cited, among other evidence, that disease, insects and climate change have all severely affected white bark pine trees, the seeds of which are an important food source for the bears. A federal judge restored protections for Yellowstone grizzlies in September 2018, ruling that FWS failed to consider how delisting the bears would affect their recovery in other parts of the country.

ESA designations can have potentially far-reaching impacts. For example, conservationists have lobbied the government to protect the greater sage grouse, a ground-nesting bird imperiled by habitat loss, for years. Classification would restrict drilling and mining across millions of acres in the American West. Under President Barack Obama, the Interior Department reached a compromise with states and industry parties in 2015 that kept the bird off the federal endangered species list, which then-Interior Secretary Sally Jewell called a “milestone for conservation in America.” But in late 2019, the Trump administration reversed course, slashing protections across some 9 million acres of greater sage grouse habitat in 10 Western states and opening the door for more energy and mineral development.

The architect of that plan, which a federal judge temporarily blocked last October, was David Bernhardt, a former oil industry lobbyist who took over as Trump’s interior secretary in April 2019.

Bernhardt also played a key role in the administration’s radical overhaul of the ESA later that year, limiting regulators’ ability to account for the impacts of future climate change and allowing economic factors to be considered in whether to grant species protection—something the law previously prohibited.

Honey bees face mounting threats: climate change, the rapid loss of habitat, pesticide exposure and more. So when President Donald Trump’s administration announced in July 2019 that it was suspending data collection for an annual report that tracks declining populations of these vital pollinators, the backlash was swift and fierce.

The decision came on the heels of the worst year on record for honeybee hive loss, with U.S. beekeepers losing 40 percent of colonies. Honeybees pollinate a third of all the crops eaten in the United States, but the number of hives has plummeted from 6 million in the 1940s to about 2.5 million today.

The U.S. Department of Agriculture said the suspension was “necessary” because of budget shortfalls, but many saw it as the latest move in the Trump administration’s broader war on science.

USDA ultimately resumed the survey in October 2019. But the damage had already been done. A full three months of data on each state’s population of honeybees and losses to colony collapse disorder is absent from its most recent Honey Bee Colonies report, a gap that will affect future honeybee research.

"It's like having a GDP report with one quarter's report missing," Dennis vanEngelsdorp, an entomologist and bee expert at the University of Maryland, told CNN.

It sounds like the least partisan government activity imaginable: Every 10 years, the Census Bureau fans out across the country to count the people who inhabit it. And yet the 2020 census has become the target of a relentless sabotage campaign.

First, congressional Republicans limited the agency to the $12.3 billion it spent in 2010, even though a growing population means each census is more complicated than the last. This forced the bureau to scale back its ambitions and roll out new counting methods without adequate testing.

Then, in March, the bureau had to suspend field operations due to the coronavirus pandemic. The following month, the Trump administration asked Congress to extend the deadlines for data collection and processing, but ended up reversing its decision in July, sending agency staff scrambling. A California court ruled in September that data collection could continue through the end of October, but on October 13, the Supreme Court granted the Trump administration’s request to halt the census pending further litigation, effectively ending the count.

Since August, the Government Accountability Office, House Democrats and a group of former Census Bureau directors have all sounded alarms that this truncated timeline will result in significant gaps in population statistics. As of October 13, the bureau claimed to have counted 99.9 percent of U.S. households. The American Statistical Association, however, noted that this figure only takes into account the quantity of data, not the quality. Many Americans have been counted “by proxy,” meaning their information was collected from their neighbors rather than directly—to pick just one of the corners field workers had to cut to finish on time.

The chaos of the 2020 count is unprecedented, said Bruce Spencer, a statistician at Northwestern University.

“This is the first time I’ve seen the government try to damage the counting process,” he said. In previous years, Spencer said, Democrats and Republicans squabbled over statistical analysis after the count was finished, but political actors have never interfered in the basic data-gathering.

The errors in the 2020 census will not be evenly distributed. The population count nearly always overrepresents white Americans while systematically overlooking minorities. The 2010 census overcounted whites by 0.8 percent and undercounted Blacks by 2.1 percent. Native Americans on reservations were undercounted by 10.6 percent in 1990, overcounted by 0.5 percent in 2000 and undercounted by 4.9 percent in 2010.

And the errors get worse with more specific subgroups. The 2010 census undercounted children by roughly 1.7 percent, but Black and Hispanic children were undercounted by 4.6 percent and 7.5 percent, respectively. And while Hispanics were undercounted by 1.5 percent overall, undocumented immigrants were undercounted by up to 14 percent.

These gaps have far-reaching consequences. Companies use population figures to determine where they should locate stores and open factories. Public polling firms use census data to adjust their results. Undercounting hard-to-reach populations creates a fiction of a whiter, richer, more educated America that will echo through the country’s basic statistics for the next decade.

“This is the invisible data infrastructure of the country,” said Jonathan Auerbach, a statistician at the American Statistical Association. “It’s the basis of an entire ecosystem of data products that informs millions of government decisions.”

What matters most, Auerbach said, isn’t the raw error rate throughout the country, but the difference in error rates between states. More than $675 billion in federal funding is tied to census allocations, from the distribution of unemployment insurance to highway planning to Pell Grants.

Census errors could even affect public health. Medicaid spending is allocated according to states’ income data—missing low-income voters will make states look richer than they are, resulting in less federal funding. If the Centers for Disease Control and Prevention manages to produce a COVID-19 vaccine, census counts will likely affect the number of doses allocated to each state.

And then there are the political consequences. A census error rate of just 4 percent could shift as many as 12 House seats from one state to another, according to a 2018 analysis. If states undercount Blacks and Latinos and overcount whites before drawing new district boundaries, they will end up diluting the votes of their poor and minority populations.

“It’s taking gerrymandering one step further,” Spencer said.

Which is, of course, the whole point.

“I see no scientific value to speeding up the census—only political value,” Auerbach said.