In early 2011, after years of study, the Occupational Safety and Health Administration moved to reduce the permissible levels of silica dust wafted into the air by industrial processes like fracking, mining or cement manufacturing. The move came after years of public comment and hearings, and reflected emerging science about the dangers posed by even low levels of dust. OSHA predicted the rule would save 700 lives annually and prevent 1,600 new cases of silicosis, an incurable, life-threatening disease.

The proposal stirred fierce opposition from an array of industries, which argued that the costs of reducing silica levels far outweighed the potential benefits. When OSHA pushed ahead, the lobbyists took their arguments to the Office of Information and Regulatory Affairs, a division of the Office of Management and Budget. Few people have ever heard of OIRA even though it is part of the White House and has broad authority to delay or suggest changes in any draft regulation.

OIRA's deliberations on the silica rule began in February 2011, and lasted two and a half years. During that time, records show, its officials held nine meetings with lobbyists and lawyers for the affected industries, but sat down only once with unions and once with health advocates.

Last August, the office sent a revised version of the rule back to OSHA; the worker protection agency has yet to act.

Labor advocates noted that the lengthy delay appeased House Republicans and pushed a decision opposed by the U.S. Chamber of Commerce out of the 2012 presidential campaign. "During that delay thousands of workers were further exposed to silica,'' said Peg Seminario, director of safety and health at the AFL-CIO. "People have gotten sicker and some will die because of the exposures that have continued to take place."

What happened to the silica rule is no isolated example. A series of executive orders over the past three decades have given OIRA significant authority to reassess rules on every imaginable subject, from health care to the environment to transportation. The office shares early drafts of rules with the president's top advisers as well as other Cabinet-level agencies that might object.

Although some on OIRA's team have degrees in science and engineering, former officials say its leadership and staff are largely drawn from the realms of economics, law and public policy. Regardless, the office does not hesitate to rework agency rules that were years in the making and backed by peer-reviewed science. Often, OIRA officials make a proposed rule appear too costly by revising the calculation of benefits downward. As it did with the silica limits, the office can also prolong the process, holding regulations in limbo for months and sometimes years.

In an influential 2001 Harvard Law Review article, Elena Kagan, now a Supreme Court Justice, argued OIRA provides a crucial check on the actions of government bureaucrats. "From the beginning of the twentieth century onward," she wrote, "many statutes authorizing agency action included open-ended grants of power, leaving to the relevant agency's discretion major questions of public policy." Since agency heads are appointed, this could lead to administrative inefficiencies and, potentially, *****s of power. OIRA's review requires agencies to answer to the president — an elected official who, unlike agency administrators, is accountable to the people.

But in practice, OIRA operates largely in secret, exempt from most requests under the Freedom of Information Act. It routinely declines to release the changes it has proposed, the evidence it has relied upon to make them, or the identities and affiliations of White House advisers and other agencies' staff it has consulted. OIRA doesn't even disclose the names and credentials of its employees other than its two most senior officials. (Repeated requests to the office for the backgrounds of its employees drew no response.)

According to a study by the Center for Progressive Reform, a nonprofit research and educational organization critical of the office, 84 percent of the EPA's proposed rules from 2001 to 2011 featured changes suggested by OIRA as did 65 percent of other agencies' regulations. Officially, OIRA's "edits" are suggestions but they carry the weight of the White House and are typically accepted by the agency proposing the rule.

Among the rules OIRA has recently reshaped is a Federal Aviation Administration requirement on pilot rest. During its review, representatives of FedEx, UPS and other air-freight companies told OIRA that the dollar value of the pilots and aircraft that would be lost in fatigue-linked crashes would be far outweighed by the higher labor costs to the industry. The office successfully persuaded the FAA to exempt cargo pilots from the rule. (The regulation, protecting only passenger pilots, took effect this year.)

But the sphere in which OIRA's involvement has been most pronounced is environmental rules. At OIRA's prodding, the EPA removed manganese from a list of hazardous wastes and exempted certain types of engines, including motorcycles and snowmobiles, from a rule limiting emissions.

In 2008, an OIRA review by the Bush administration deleted a provision intended to protect plant life from the effects of ozone, a key component of smog. The EPA had proposed a sharp reduction in the permissible levels of ozone to protect forests and vegetation, which naturally remove carbon from the atmosphere. According to an investigation by the House Committee on Oversight and Government Reform, the White House summarily overturned the unanimous recommendation of the EPA's Clean Air Scientific Advisory Committee and an array of expert testimony.

"I don't think OIRA has been shy about meddling in any aspect of the EPA's work," said Lisa Heinzerling, senior climate-policy counsel at the EPA in 2009 and the associate administrator of the office of policy in 2009 and 2010. "They really do a line-by-line edit — sometimes that involves changing what the regulations mean,'' she said.

In 2013 the Administrative Conference, an independent federal agency that reviews government administrative processes, released a study of OIRA's effect on the application and interpretation of science the agencies gather and analyze to write rules. In examining a group of air-quality regulations, the study found that most of OIRA's suggestions involved substantive changes. The report concluded that in some instances, the office has proposed changes to the basic science underlying the rules. These included revising numbers in tables created by the EPA, altering technical discussions and recommending different standards altogether.

"OIRA review makes the science done by the agency a facade," said Wendy Wagner, author of the study and the Joe A. Worsham Centennial Professor at the University of Texas Law School. "It gives the impression that rules are much more based on science than they end up being."

The Rise of a Federal Power

Congress created OIRA in 1980 to prevent federal agencies from demanding excessive amounts of data from public and private parties. President Reagan greatly expanded its powers, signing an executive order that gave the office the authority to review all federal rules. This was an important change, since most laws say the rules are to be written by the relevant Cabinet agency, not the president and his aides. At the time, Reagan's move kicked up controversy — still alive today — about whether it was appropriate for the White House to have such a direct say in government rulemaking.

Since then, both Republican and Democratic presidents have signed executive orders enshrining OIRA's pivotal role.

OIRA gets three passes at each rule it audits. First is the informal review, which takes place before a proposed rule is officially submitted. Informal review could include memos, phone calls, in-person conversations and the swapping of early drafts. This process unfolds in secret. Documents generated as part of the executive branch's policy deliberations are exempt from the Freedom of Information Act.

The office then reviews a draft of the proposed rule and the final version.

Proponents of this process depict OIRA as a counselor, providing a neutral, fresh look at long, complex regulatory documents; double-checking cost-benefit analyses; and ensuring regulations are "consistent" with "the President's priorities," as a guiding executive order states. Proponents also point out that while OIRA's chief and deputy are presidential appointees, its staff is career civil servants, which precludes the pushing of a political agenda.

Cass R. Sunstein, a prominent legal scholar who led OIRA from 2009 to 2012, rejects many criticisms of the office, namely that it lacks transparency and that its suggested changes and delays are politically motivated. Sunstein, now a law professor at Harvard, responded neither to ProPublica's requests for an interview nor to written questions.

In a 2013 article, Sunstein said OIRA largely functions as an "aggregator" of information from the executive branch and the public on any given regulation. Sunstein acknowledged that OIRA plays a substantive role in a rule's economic aspects, but said it does not rethink scientific conclusions. "When scientific issues are engaged, there is no 'political interference with science' (in my experience),'' he wrote. "Scientific issues are explored as such by people who are competent to explore them."

In a 2010 talk, Sunstein addressed criticism of the office's transparency, pointing to its "dashboard" web page, which features basic information on rules currently under review. In other venues, he has cited OIRA's posting of meeting logs as evidence of its openness. Neither the dashboard nor the logs disclose who has proposed changes in rules or what evidence they are relying upon.

Without question, OIRA review opens the door for a long line of people to influence federal regulations: from the President's chief of staff, to a Cabinet secretary, to a presidential adviser in an executive office such as the Domestic Policy Council, to a member of Congress, to lobbyists for regulated industries.

Coal Ash Controversy

One controversial rule currently under OIRA review would create controls on the handling of coal ash, the dross from coal-fired power plants. Coal ash contains a host of hazardous substances, including arsenic, lead and mercury, and is the country's most polluting waste — more than all oil refineries, petrochemical plants, paper mills and incinerators combined — as well as being one of the country's largest sources of waste overall. Questions about its handling burst into the news earlier this year after the toxic sludge spilled into the Dan River in North Carolina.

The ash is most often stored near power plants in artificial ponds from which nearby rivers, lakes and drinking water supplies can be contaminated — even if there is no catastrophic spill.

The rules for handling this waste, known as Effluent Limitations Guidelines, have not been changed since 1982, and many experts say they are weak. The EPA leaves most standards to the discretion of the states, many of which have done little to enforce and add protections. Essentially, most coal-fired power plants have been able to dump their ash indefinitely into unmonitored, open-air pits. Not surprisingly, the utility industry is resisting stronger controls to stave off costs of adopting new waste-handling methods and more technologically advanced water-treatment equipment.

The EPA sent OIRA its proposed new rules in January 2013. The agency submitted five options from which it would choose the final rule. In its draft, the EPA indicated it would likely pick one of two options, which it listed as "preferred." Both set relatively tough standards on power companies.

In the weeks leading to OIRA's completed review of the coal ash limits, a number of utility industry lobbyists and lawyers met with the office. While OIRA makes public a list of attendees and documents given to the office's representatives at meetings, it does not disclose the substance of their discussions. Such conversations are not unusual; any member of the public can meet with OIRA to discuss a rule. But the office has become a standard stop on the lobbying circuit for industries facing tighter regulations.

A 2003 Government Accountability Office study found that "regulated parties," typically corporations or their lobbyists, frequently get what they want after meetings with OIRA. Sometimes, the language of the edited rule is similar to that proposed by the regulated parties themselves.

When the rule on coal ash effluent emerged from OIRA, three more options had been added, a diluting of the two options the EPA favored. OIRA's draft dropped the tougher of EPA's preferred rules and identified those new, less demanding options as favored.

The office also recast the EPA's scientific findings. The agency initially stated that using ponds for storing the most toxic form of coal ash, the emissions captured in the smoke stack's final filter, did "not represent the best available technology for controlling pollutants in almost all circumstances." Revisions made during OIRA review recommended eliminating this conclusion, giving no explanation why. Other changes included softening data, such as reducing coal-fired power plants' share of toxic pollutants discharged to surface water from "at least 60 percent" to "50-60 percent." The post-OIRA version also recommended reducing the projected benefits of the EPA's higher standards, which the agency did.

The EPA reached its findings for handling coal ash after almost 10 years of extensive data collection, modeling and analysis. This work required the knowledge of biologists, chemists, engineers and toxicologists working on a team that varied in size from about eight to around 30 people at various phases. The EPA lists the staff in its Office of Water responsible for the rule in the document itself, on the agency's website and in press releases.

It is difficult to know the qualifications of the people at OIRA who rewrote the EPA's draft rule. The only credentials disclosed by its website are those of the current administrator, Howard Shelanksi, who has a Ph.D. in economics, and his deputy, Andrei Greenawalt, a lawyer who was a policy adviser to the chief of staff from 2011 to 2013. ProPublica's request to the OMB for a list of OIRA staff went unanswered, as did repeated requests for an interview with Shelanski. OIRA also failed to respond to written questions about its lack of transparency.

Formaldehyde: A Case Study

Prompted by evidence that formaldehyde fumes were sickening Hurricane Katrina survivors living in FEMA trailers, Congress passed a law in 2010 calling for tighter limits on the chemical. Specifically, the legislation required controls on the use of the known carcinogen in compressed wood products commonly used in building materials. The EPA, the body designated to administer the law, submitted its proposed regulation to OIRA in May 2012.

Included in the rule was a set of public health outcomes that the EPA, using peer-reviewed work by the National Academy of Sciences, judged would improve with the regulation. The agency projected benefits ranging from $91 million to $278 million, mostly from reduced medical care and hospitalizations of *****ren with asthma.

While the rule was at OIRA, industry representatives met five times with officials from the office. The United States is a major market for formaldehyde: In 2011 the American Chemistry Council estimated that U.S. companies sold more than $27.7 billion worth of formaldehyde-based building products. Some of the world's most powerful companies manufacture, sell and use formaldehyde, including Koch Industries, Exxon Mobil and Dow Chemical.

During OIRA's review, the scientific findings were challenged — it's not clear by whom or on what grounds — and were ultimately changed. In the post-OIRA version, asthma was eliminated as a quantifiable health impact of the chemical. That single change erased over 80 percent of the financial benefits from regulating formaldehyde. Meanwhile, the costs of industry compliance remained just as high, dramatically tipping the scales against stronger limits.

The explanation offered in OIRA's edited version is cryptic. It says only that the EPA "later concluded" it had insufficient information on the relationship between formaldehyde exposure and asthma to assign a dollar sum for improved health. The office presents no supporting science justifying this failure to endorse the National Academy of Sciences' findings.

Other Levers of Power

OIRA has multiple tools for slowing and even blocking rulemaking. These include ignoring an agency's submitted regulation, asking an agency to withdraw a submitted rule, and leaving a rule to languish, sometimes for years, until the agency withdraws it.

According to OIRA's governing executive order, the office is supposed to complete its review within 90 days of receiving a regulation, with the option of one 30-day extension. But rules often gather dust for far longer. As documented in an Administrative Conference analysis, delays reached an all-time peak under President Obama between 2011 and 2013.

High-level EPA staff interviewed for the analysis thought OIRA reviews took so long because of political concerns over contentious rules in the lead-up to the 2012 general election. According to the report, "The employees said their agencies were instructed that such rules were not to be issued unless deemed absolutely necessary (e.g., judicial deadline) or if it could be shown they were not controversial (e.g., clear net benefits)." Such instructions weren't issued in writing, but conveyed verbally by agency administrators who had been to meetings with key White House officials, the employees said.

The Administrative Conference report noted that in June 2013, 141 rules were under review at OIRA, of which over half had been there for more than 90 days. About half of those had been under review for more than a year, and 26 more for over two years. The rules were from a variety of agencies, including the EPA and the Departments of Energy and Labor.

One set of efficiency standards proposed by the Department of Energy has been stranded at OIRA since 2011. An EPA rule establishing a list of "chemicals of concern," those the agency flags as potentially dangerous and possibly warranting further regulatory action, sat at OIRA for over three years until the agency withdrew it in 2013. A Department of Transportation rule requiring rear-view cameras for passenger cars was marooned at OIRA for a year and a half, beginning in November 2011, before being withdrawn. During that time the DOT had data showing that, on average, "backover" car accidents injured 15,000 people annually and killed some 200 people a year, mostly *****ren and the elderly.

When OIRA buries a rule, the agency has two choices — let it disappear or pull it and try again later. Sometimes OIRA recommends to agencies that they withdraw a rule, which would leave no fingerprints since none of these types of delays are publicly documented or explained. Curtis Copeland, former specialist in American national government at the Congressional Research Service and author of the 2013 study for the Administrative Conference said, "What I heard when talking to the agencies was that OIRA often recommends agencies withdraw rules. Then it can all be done in the dark."

All of this can have a chilling effect on agency staffers who come up with ideas for regulations and then explore, research, model and analyze them. As Copeland wrote in a 2009 Congressional Research Service survey of OIRA, "some agencies have indicated that they do not even propose certain regulatory provisions because they believe that OIRA would find them objectionable."

OIRA's motivations

While most changes made by OIRA do not recast the scope and impact of rules, many do. The GAO's 2003 report, and a follow-up in 2009, found that, of the rules examined, more than one-quarter of the regulations changed at OIRA's suggestion were substantively altered. In the earlier of the two reports, the GAO concluded that, in general, OIRA's suggested changes reduced the burden on regulated parties and, in some cases, sliced the rules' expected benefits.

John Morrall, deputy administrator of OIRA for almost 30 years until 2008, said that in high-stakes regulations, political and economic factors consistently play a role and may help account for the softening of many regulations. "Maybe on a one-by-one basis the science and cost-benefit analysis say you should regulate certain things," he said. But if OIRA approved all the scientifically valid rules agencies write, "it could destroy that sector and that could generate too much political opposition. … In particular states and congressional districts , that could be bad in an election."

Thomas Cmar, a staff attorney at Earthjustice, which sued the EPA over its delay in updating coal ash effluent limits, has, along with his organization, tracked OIRA's influence on the rule. "The regulatory process takes years, and an agency is supposed to develop reams of evidence that the actions they take are scientifically well grounded, that the rules are designed to protect people from harm," he said. Cmar believes OIRA's process is marked by favoritism for regulated industries and that it "turns our government into a completely arbitrary actor."

A Lack of Transparency?

To alleviate what some saw as OIRA's stonewalling and obfuscation, in 1993 President Bill Clinton issued Executive Order 12866. It said all federal rules must undergo a cost-benefit analysis, and it set deadlines as well as disclosure requirements on changes made during review. Clinton required that all documents exchanged between OIRA and the agency proposing a rule must be made public.

President Obama affirmed the directive, and OIRA's website now prominently quotes 12866's aim "to restore the integrity and legitimacy of regulatory review and oversight; and to make the process more accessible and open to the public."

Yet, as was true under George W. Bush and Clinton himself, the Obama administration has not fully honored 12866's disclosure guarantee, several agency officials said.

"This is being contradicted in almost every possible way," said Lisa Heinzerling, the former EPA associate administrator.

Not long after joining the office, Heinzerling, pursuant to 12866, drafted a public notice of a disagreement over a proposed rule between the EPA and OIRA. (She declined to reveal which rule.) Having reached an impasse, the parties were taking the problem up the chain of command for negotiation. However, according to Heinzerling, OIRA told her "in no uncertain terms" that she was not to make the memo public. She was surprised. "I hadn't seen it up close and hadn't known how completely it diverged from the executive order," she said. "It was unsettling even if you're a fan of having the White House in charge of regulatory matters."

Morrall, the former deputy administrator of OIRA, acknowledged the lack of transparency and defended it on the grounds that it might be politically damaging if the public learned more about the office's inner workings. "To do our job we don't need what went into our decisions out in the public domain because opponents could use that against the president."

In its 2003 report, the GAO recommended that all agencies disclose changes made at OIRA's request whenever they occur — in both informal and formal reviews. "That has never been done, under any agency," said Copeland, the expert on OIRA who is currently a special counsel at the Administrative Conference. "OIRA as an institution doesn't like that disclosure."

The GAO's 2009 follow-up report found that OIRA and the agencies had failed to adhere to seven of eight transparency recommendations made in its 2003 assessment. The one measure OIRA had taken was to disclose logs of meetings with people outside government, which it continues to do, although they are nearly impossible to search, contain minimal, sometimes-unclear information about the attendees, and have scant information about what was discussed.

While no entity has followed the transparency requirements set by the executive order, according to several former senior executives at the EPA, if the agency didn't follow OIRA's suggestions, the response would be swift. Consequences would likely be cuts to the agency's budget, a curtailing of its influence within the White House, and potentially the firing of the agency head.

Many administrative law experts, including Copeland, don't see anything inherently wrong with OIRA review, but most agree — along with the GAO and the Administrative Conference — that more transparency is needed. Still others think OIRA should be wholly revamped. Rena Steinzor, a law professor at the University of Maryland and current president of the Center for Progressive Reform, the group critical of OIRA, has called for the office to be stripped of its rulemaking powers, terminating centralized White House regulatory review. In a 2012 law review article, Steinzor wrote that the Executive would still have rulemaking power: "The President can exert sufficient control over rulemaking through the political appointees he has selected to lead the agencies."