Environmental Protection Agency Administrator Andrew Wheeler vowed on Wednesday to make greater use of the agency’s official board of science advisers. But soon after he spoke, some of those advisers and other outside experts protested that the agency was instead limiting their role.

Wheeler took the unusual step of addressing the first meeting in a year of the Science Advisory Board (SAB), a group of advisers established by law who should be holding six to eight meetings annually, according to its charter. The reason he was there, Wheeler said, was to show his commitment to a better process.

“I’m glad to be here to lay out a new direction for how EPA interacts with and uses the SAB,” he said. “I will be the first to admit we have not utilized you the way that we should. We can and we will do better.”

The Science Advisory Board has become a center of controversy as President Donald Trump‘s administration has sought to remake the federal agencies. Researchers who receive EPA grant money have been barred from serving on the board, while several industry consultants and contrarian scientists have been brought aboard. Although some holdovers remain from previous administrations, Trump-era appointees now make up a large portion of the 44-member body. Some of them are known for their outlying views beyond the prevailing scientific consensus, or for representing the interests of regulated businesses.

Noting the lengthy vetting process for appointees, Wheeler said, “Why should we have talented public servants go through all of that only to use you sparingly? Today we are beginning the transition to a new process. It’s no secret that the process was broken, and was not beneficial to the EPA, the board, or the public we all serve.”

But some members of the board parried with Wheeler and other EPA officials at the meeting over exactly what part of the process is broken—the advisory process, or how the Trump administration is proceeding with a deregulatory agenda that gives less weight to some of the science that has guided the agency for years.

One SAB member who is a holdover from before Trump, Steven Hamburg, chief scientist of the Environmental Defense Fund, asked Wheeler why the Science Advisory Board had not had an opportunity to weigh in on some of the latest science in the agency’s repeal of the Obama administration’s regulations on control of methane emissions from oil and gas operations. The SAB raised concerns over that proposal when it met a year ago, and it had an opportunity to confer with another group of outside EPA experts, the Clean Air Science Advisory Committee, but Hamburg said their focus was more on technology than science. EDF, an environmental group, has been involved in a number of large scientific studies on the extent of methane, a potent greenhouse gas, released by the oil and gas industry.

Wheeler said that in the future, he hoped to involve the Science Advisory Board earlier in the process—soon after proposals are made public, but he couldn’t change what happened in the past.

“The one you cite—I’m sorry, it’s too late in the regulatory process,” Wheeler said. “I can’t go back and reopen some of the regulations we’ve already done.” The agency finished taking public comment on its methane proposal last December, but has not yet finalized a rule.

One public commenter, John Bachmann, a 33-year veteran of the EPA who is now retired, said that under the 1978 law that governs the EPA’s use of its science advisory board, the agency is actually required to provide its regulatory proposals and the underlying science to the Science Advisory Board before a proposal is made public. Bachmann now works with the Environmental Protection Network, a group of former agency scientists and other staffers who have been critical of the EPA’s moves under the Trump administration.

Wheeler’s Controversial ‘Secret Science’ Plan

But much of the debate at the first day of a two-day meeting centered on a controversial policy proposal that Wheeler said he was committed to finalizing this year even though it had provoked an outcry that left him “a little shocked.”

The proposal, called “Strengthening Transparency in Regulatory Science,” would limit the kinds of science the agency can use in its decision-making. The agency received some 600,000 comments on the proposal, including from leading scientific journals, professional organizations and universities. Because many human health studies rely on information gathered under confidentiality agreements, and the raw data has not been made public, the proposal could prevent the agency from using some of the seminal studies underpinning air pollution regulations, for example.

Barbara Morrissey, of Washington State’s Department of Health, sitting on the board on an ex officio basis as chair of another advisory group, the Children’s Health Protection Advisory Committee, told Wheeler she was concerned about how EPA could meet its legal requirement of using the best available science, if that science can’t be replicated in the manner the EPA envisions.

Wheeler said there would be “a safeguard” in the system, giving the EPA Administrator the discretion to include an important study even if it can’t be replicated. “I do fundamentally believe that the more information we put out, the more sound our decisions will be and the more easily understood and accepted by the public,” Wheeler said.

But one public commenter, Veena Singla, associate director of the reproductive and environmental health program at the University of California, San Francisco, said her concerns about the proposal were not assuaged.

“Having a political appointee deciding what science is in or out of policymaking is not a safeguard, but rather a fundamentally biased process,” Singla said.

Before the meeting, the EPA had given the Science Advisory Board a charge to weigh in on two narrow aspects of that rule—how the agency should deal with confidential business information, and how to shield personally identifiable information in the large health studies that in the past have been the basis for much of the agency’s decision-making.

SAB ‘Should Be Giving an Independent Look’

Allison Cullen, associate dean of the school of public policy and governance at the University of Washington, was one of several board members who objected to the fact that the scope of comments that the EPA sought from the SAB was being limited—in contrast, she said, the public had been asked to comment on the proposed regulation broadly.

“The SAB is an independent body that should be giving an independent look at this, not necessarily responding to what the public comments are,” Cullen said.

Hamburg said that the time frame that the SAB had been given to weigh in on the measure was also too compressed for the board to weigh in effectively. “There’s a lack of transparency about this ‘transparency’ proposal,” said Hamburg. “It’s inconsistent with what the administrator says about wanting to fix a broken system.”

New Appointees Side with Wheeler

But several of the board members who were appointed by Wheeler and by his predecessor, former EPA administrator Scott Pruitt, praised the agency for its effort.

New SAB member Robert Phalen, an air pollution researcher at the Irvine campus of the University of California, has long been critical of the air pollution studies that the EPA has relied on, which use raw data not available to the public. “A short cut—if a study cannot be replicated for privacy concerns, probably the alternative is to simply drop that from consideration and rely on the vast amount of other studies. That’s just a raw suggestion from an SAB member.”

Stanley Young, a statistician who did work for the pharmaceutical industry and others before being appointed to the SAB by Pruitt, said that there should be no privacy concerns in epidemiology studies that track mortality. “Once a person is dead, they are not considered a person,” he said. He said he hoped that the agency was prepared for “epidemiological foot dragging and stonewalling” about release of its data.

The SAB will be holding a teleconference with the EPA later this summer to further discuss the transparency rule.