Sections

Research

How will we know if DOGE is succeeding?

June 18, 2025


  • In his zeal to make quick progress, Elon Musk created a DOGE initiative that was doomed to face early failures.
  • Early DOGE efforts were enacted on dubious legal grounds, which poses a threat to their permanence.
  • Cuts were undertaken without attention to or understanding of the government’s mission. 
Department of Government Efficiency website displayed on a laptop screen and Department of Government Efficiency logo displayed on a phone screen are seen in this illustration photo taken in Krakow, Poland on March 28, 2025. (Photo by Jakub Porzycki/NurPhoto)NO USE FRANCE
Department of Government Efficiency website displayed on a laptop screen and Department of Government Efficiency logo displayed on a phone screen on March 28, 2025. (Photo by Jakub Porzycki/NurPhoto)
Introduction

One of the most contentious but potentially transformative initiatives by the second Trump administration has been the creation of the Department of Government Efficiency (DOGE). In coordination with Elon Musk and people who worked at several of his companies, it set the task of improving efficiency by downsizing the government, cutting the federal workforce, and eliminating entire programs. Musk and his team were embedded in many departments and agencies, given highly unusual access to sensitive government data, and empowered to fire people deemed not in line with the administration’s values and priorities.

Some months have passed since DOGE started, and Musk has now left the project. How and when will we know if the Trump administration, through DOGE, has succeeded in its ambitious plans to downsize the federal government? A definitive evaluation may not be possible until a new president takes office on January 20, 2029. Yet, we can establish several numerical metrics:

  • Did the number of people working for the federal government shrink?
  • Did the number of federal contracts shrink?
  • Did the amount of federal spending shrink?
  • Did the size of the budget deficit shrink?

There are also other means of assessment. A less well-understood but equally important set of metrics revolves around government performance. Are agencies able to provide services and benefits that citizens expect? The fear of government malfunction is why the last major government reform effort undertaken by the Clinton administration’s National Performance Review, in which I was directly involved, adopted a different approach. As former Vice President Al Gore said in an interview with Bill Maher, “We used a scalpel, not a chainsaw or a butcher knife […] We cut the fat, not muscle and bone.”

Here are some examples of how cuts can hamper the provision of basic services. For instance, when we look back at Trump’s second term:

  • Was there an increase in the number of airline crashes?
  • Were there contagious disease outbreaks?
  • Were there major delays or mix-ups around the payment of social security checks or veterans’ benefits?
  • Was there a terrorist attack, military failure, cybersecurity hack, or national security disaster linked to budget or staffing cuts in agencies responsible for those areas?

By using a combination of numerical and service delivery benchmarks, we can begin to assess how well DOGE has performed, whether it has achieved its specified objectives, and whether it has improved government performance. In this report, I examine currently available data about DOGE’s efforts in several areas to evaluate its initial accomplishments.

Firstly, I find that early DOGE efforts were enacted on dubious legal grounds, which poses a threat to their permanence; secondly, that they were undertaken without attention to and understanding of the government’s mission.

An overview of DOGE efforts

When the DOGE team took over the federal government, it did so without regard to laws and norms that had been in place for decades. They were seeking ways to make as many cuts as possible, as rapidly as possible, and thought that by taking control of the Office of Personnel Management (OPM), the human resources center of the federal workforce, they could accomplish large cuts quickly. Hence, February 14, 2025, became known inside the civil service as “the Saint Valentine’s Day massacre.” On that day, OPM directed agencies to fire probationary employees. Probationary employees are those who have just been hired and do not yet have full civil service protections. The DOGE team thought that these people (around 200,000 as of that date) would be the easiest to fire; many (but not all) agencies proceeded to send out notices.

For those involved in prior attempts to cut government, as I am, the use of OPM to fire people in other agencies caused serious confusion; no one ever thought OPM had that authority. Soon, the blanket firing was reversed in a variety of ways. The first venue was the Merit Systems Protection Board (MSPB), which ruled that 6,000 workers in the Department of Agriculture had been wrongfully dismissed and should be hired back. Although the workers were dismissed again when Trump installed a new head of the MSPB, by that time, the courts had intervened. In another case, brought by labor unions and some non-profit groups, U.S. District Court Judge William Alsup (appointed by President Clinton) ruled that OPM did not have the authority to fire employees; only their home agency did. In a hearing, he stated that OPM “does not have any authority whatsoever under any statute in the history of the universe to hire and fire employees at another agency.” Simultaneously, a Federal District judge in Maryland, James K. Bredar (appointed by President Obama), ruled that 24,000 probationary employees should be reinstated.

The Trump administration has appealed all these rulings, and in some instances, the Supreme Court has lifted the stay pending a more thorough review, but without getting to the underlying issues. One thing is clear: the blanket assertion that all these employees did not live up to expectations and could therefore be fired was not in keeping with the current understanding of the laws requiring individual reviews and explanations. Hence, one easy method that DOGE relied upon to cut the workforce remains in limbo until the Supreme Court hears a case on the merits.

So far, many of the earliest DOGE actions have been reversed by lower federal courts. Then why were they carried out in the first place? The initial set of DOGE-directed firings seems aimed at testing the “unitary executive” theory. The theory, popular among conservatives since the Reagan administration, argues that the President’s control of the executive branch cannot be constrained by Congress. In practice, this means that independent agencies not controlled by the President are illegal and that the president has more power than Congress. Taken to its extreme by the DOGE team, this theory meant that the President could shut down congressionally authorized agencies by, in effect, stripping them of their ability to conduct business. In other words, rather than go to Congress and ask for passage of a bill undoing the law that created the United States Agency for International Development (USAID), the DOGE team simply fired personnel, closed off computer access, and even went so far as to take the sign down from the building. In effect, they made it impossible to do the work that Congress had ruled USAID should do since passage of the Foreign Assistance Act of 1961 (Public Law 87–195).

In late May of 2025, Judge Susan Illston, District Court judge for Northern California (appointed by President Clinton), ruled that President Trump had acted outside of his legal and constitutional powers and barred the administration from making any more widespread layoffs. As reported in Government Executive, OMB Director Russell Vought said, “And I think at the end of the day, wherever they’re at, they’re going to be successful when they get to the Supreme Court.”

Legal issues aside, a second but less obvious reason behind the early DOGE actions was to demoralize and terrify the federal workforce into leaving their jobs. Thus, the second way the Trump administration has attempted to get large reductions in the federal workforce is through an initiative known as the deferred resignation program. The program was rolled out early in Trump’s second term, but initial take-up was lower than expected, amounting to around 75,000 people. As DOGE took over more and more agencies, fears of RIFs (reductions in force) and general angst about what was happening to the government increased federal workers’ incentives to accept the administration’s offer. So, in early April, agencies offered a second opportunity, and this time the take-up seems to be more robust, prompted by fears of layoffs to follow. Some agencies didn’t wait for a second round of deferred resignations but used the buyout authority that had been passed by the Clinton administration in 1994.

Musk reduced his involvement in the government in late April and officially left in late May. By then, it had become clear that regardless of what the Supreme Court could eventually rule, it would be difficult to cut government through a centrally located entity like DOGE—even one that had close ties to the president. As Musk and Trump’s relationship deteriorated, it became clear that the former would no longer be able to persuade the latter. More importantly, Musk’s unpopularity damaged his efforts. Brandishing a chainsaw on the stage of the winter meeting of the Conservative Political Action Committee and vowing to use it to cut government was, it turns out, too much for the public to bear: while hardcore MAGA supporters loved the imagery, the broader public did not, and Musk’s approval numbers sank. Additionally, his decision to spend millions and campaign personally in a Wisconsin Supreme Court race that pitted a MAGA candidate against a traditional liberal did not help his political standing. His candidate lost by ten points, proving that whatever Musk’s advantages may have been as an advisor and funder, he was not a political asset. Meanwhile, his car company, Tesla, was suffering protests and drops in sales, and by late spring, Musk had pretty much removed himself from the leadership of DOGE.

But even if Musk had better navigated the political environment, stage one of DOGE was doomed to fail. As the court cases began to pile up, it turned out that for both practical and legal reasons, DOGE needed to be embedded in government agencies. So, as Musk himself departed, he left behind a group of loyalists dedicated to continuing the cost-cutting efforts. Meanwhile, OMB Director Russel Vought was as intent on cutting government as Musk, and cabinet secretaries were anxious to do the same. In late February, Vought directed all executive branch agencies to develop “agency RIFs and reorganization plans” by March 13. As cabinet secretaries got confirmed, it was clear that if they wanted to run their own agencies, they needed to get out ahead and offer their own downsizing plans. In a well-publicized cabinet meeting, friction between Musk and cabinet secretaries surfaced and resulted in President Trump reinforcing that the secretaries should have the final word. The tension went so far as to result in fisticuffs between Musk and Treasury Secretary Scott Bessent.

Did the number of people working for the federal government shrink?

All across the government, in large agencies and small, a pattern is being repeated. Large layoffs are announced by DOGE or by the political appointees working with DOGE, buyouts or deferred resignation plans are offered; then, it becomes apparent that the agency is in danger of malfunctioning, and a scramble to rehire or rescind offers ensues. A recent review by the Washington Post found that, across the government, workers are being rehired.

The only agency that has not been caught up in this chaotic loop is USAID. DOGE targeted USAID early and rapidly, essentially closing it down. From their point of view, it was a wise move: the effects of shutting down U.S. food and medical aid were felt far away by very poor people in countries many Americans had never heard of. Lawsuits were brought against DOGE, arguing that shutting down USAID without congressional approval was unconstitutional. In the meantime, USAID has been folded into the State Department, and its functions—especially the popular Food for Peace and PEPFAR (the President’s Emergency Plan for Aids Relief)—may resume at some point in time.

But USAID is the exception. What has emerged from this second stage of the DOGE effort is the realization that many government agencies provide services that citizens want, but their value isn’t felt until they malfunction. Which one of these malfunctions will blow up first is anyone’s guess, but FEMA’s lack of preparedness for hurricane season is a good candidate.

Looking at the federal government overall, it is impossible to tell at this point whether there will be a shrinkage of the number of workers. It is clear that thousands of federal employees have lost their jobs, but it would not surprise me or anyone else who has followed this issue carefully if, on January 20, 2029, the government will have about the same or more civil servants than it did in early 2025. DOGE has originated hundreds of lawsuits, currently working their way through the court system. As of June 9th, 196 decisions at the district court level, and 71 decisions at the appellate level have gone against the administration. Were some of these cases to make it to the Supreme Court, a ruling in favor of the administration would allow for an expansion of executive power.

In those areas where the cuts were too deep and expertise was lost, it may be hard to hire back workers. The ugly characterizations of civil servants send a message to any would-be employee—we don’t respect who you are or what you do.

Did the number and dollar amount of grants and federal contracts shrink?

In the early stages, DOGE took aim at federal contracts. Yet, upon closer inspection, many of the savings from cutting contracts were faulty and the savings illusory. The Associated Press examined the DOGE website and determined that nearly 40% of the savings listed were bogus; some were for contracts cancelled before the Trump administration took office, others were just wrong.

The trend towards hiring contractors had been increasing long before the Trump administration, and the contract workforce has been larger than the civil service for some time now. These trends may continue for reasons unrelated to government budget and size. Aside from contracts for goods (fighter planes, office computers, etc.), it’s important to look at what the government has been contracting out in the way of services. Some contracts might be justified on the basis of flexibility or specialized skills, but a close examination of the issue might just lead a future president (Democrat or Republican) to conclude that government efficiency requires more, not fewer, civil servants. This is the intriguing determination that John Dilulio Jr. reached in his book Bring Back the Bureaucrats.

 

Did federal spending shrink? Did the deficit shrink?

The answer to this question will almost surely be no. As the DOGE team discovered after taking office and progressively shrinking their savings estimates from $2 trillion to $1 trillion to $150 billion, the real money in the federal government goes back to citizens in the form of Social Security, Medicare, Medicaid, veterans’ benefits, etc. Reform of these programs is long overdue, but it requires a political world characterized by trust and comity, not extreme polarization. As a wise old legislator once told me, the problem is not spending, it’s the “pay fors”—a legislative term referring to the need to match spending with revenue. As long as Congress and presidents continue to spend more than they take in, no amount of RIFs or buyouts or cancelled contracts will deal with the deficit problem. The sooner Americans learn this, the sooner they will stop believing that cutting waste, fraud, and abuse is a silver bullet for solving this issue.

 

Has government performance or service delivery been affected?

It is too early to say definitively how government performance or service delivery has been or will be affected.

As mentioned earlier, firings via OPM and offers of deferred resignation were two early blanket attempts to cut the workforce from the center of government. The deferred resignation program and the buyout legislation passed in 1994 under the Clinton administration might yet have some success in cutting the number of federal employees, but attempts to do across-the-board firings, such as the mass firings of probationary employees, are probably not going to be very effective in the long term. The result so far is that post-Musk, DOGE is morphing into an agency-by-agency effort—no longer run by a central executive branch office, but by DOGE recruits who have been embedded in the agencies and by political appointees, such as cabinet secretaries, who are committed to the same objectives.

As these actors continue their government reform work, and as Congress and the public focus on what is actually being cut, a new worry has emerged. Cuts could result in agency dysfunction and service delivery bottlenecks. The possibility of, say, an airline disaster being laid at the feet of DOGE, the Department of Transportation, and President Trump has tempered, in some instances, the desire to simply get rid of personnel.

The result is a somewhat erratic pattern repeated in many agencies, perhaps best described as two steps forward and one-and-a-half steps back. This pattern is the result of both court cases and worries about the effects of cuts on government functioning. Here is how it plays out:

  • DOGE or the cabinet secretary sets goals for cuts.
  • Interested parties bring those cuts to court.
  • The Trump administration loses in court and is ordered to hire back employees to preserve basic job lines and service delivery functions.
  • The agency hires people back and then prepares RIFs.
  • Inside agencies, worries about the consequences of cuts lead to dismissed employees being called back to work.
  • The agency expands options for early outs, including the deferred resignation program.

Here are a few examples of where the desire to cut personnel and dollars is having potential consequences for government performance. In one way or another, these stories can be told throughout the government.

The Department of Agriculture

At the end of February, the Department of Agriculture announced that 4,200 probationary employees had been laid off. Almost immediately, it became apparent that among those fired were people working on avian bird flu. Avian bird flu is a major cause of the high price of eggs, and the high price of eggs was a major contributor to Trump’s electoral victory in 2024. Republican members of Congress were quick to call for the reinstatement of those employees. Embarrassed and under White House pressure, the Department moved quickly to rehire these workers.

Shortly thereafter, the Merit Systems Protection Board ordered the temporary reinstatement of 6,000 employees. Most were reinstated with back pay. Then, on April 1, in a message to staff, the Department of Agriculture warned that 10,000 employees would be laid off as part of a large restructuring aimed at moving many people out of the Washington, D.C. area. But weeks later, so many people were taking advantage of the early out options and so many others refusing to move out of the D.C. area that the department found itself in the position of having to hire back retired federal employees and offer them offices and temporary remote work. As a general proposition, the administration was busy trying to shut down work-from-home options, but the Department of Agriculture had to revive that option to get its work done. In addition, as of April 15, 16,000 employees at the department had already signed up for the deferred resignation program. Former FSA Administrator Zachary Ducheneaux likened the Trump administration’s approach to a “beef cattle producer starving his cows to boost efficiency and then getting rid of them when those results, predictably, do not occur.”

This brief chronology of the Department of Agriculture illustrates two features of the Trump administration’s cutting efforts so far. First, between the lawsuits that reverse firing decisions and the need to keep critical functions going, such as research into bird flu, it is hard to understand whether any given agency is shrinking, growing, or staying the same. Second, across-the-board cuts without attention to the critical functions of a given agency are likely not to be permanent. Already, the DOGE caucus in Congress has all but disbanded, and most conversations about putting DOGE cuts into legislation have ended.

The Central Intelligence Agency

The United States’ intelligence community consists of 18 organizations—some military and some civilian—which together seek to defend the country from attacks and protect American secrets. The most well-known is the CIA, which employs intelligence operatives and analysts here and around the world. Only a few people know how many federal employees there are at the CIA, but that has not kept its leadership from commenting on DOGE goals. In late March, Director John Ratcliffe invited Elon Musk to the Agency to discuss cuts, a move widely seen as a preemptive effort to control the cuts himself. According to reporting in the Washington Post, “This invitation to Musk is a way that Ratcliffe can stay in the driver’s seat without giving Musk the steering wheel,” said Beth Sanner, a former deputy director of national intelligence during the first Trump administration. “It’s a great offensive play. This doesn’t mean that Ratcliffe won’t cut people. But he wants to be in charge of who and why, and how.”

Meanwhile, U.S. District Judge Anthony Trenga (nominated by President George W. Bush) blocked the firing of 19 CIA employees who had been assigned to work on DEI projects. The lawyers for the fired employees argued that these intelligence officers should be reassigned instead. Their extensive (and costly) training would still be useful for the agency.

The CIA is a case study of how DOGE’s early operating procedures threaten optimal functioning. Take, for instance, DEI. For government agencies, DEI initiatives are not only important for reasons of equity and fairness in staffing—they are also central to the mission and equitable service delivery. In the case of the CIA, an ethnically, culturally, and linguistically diverse workforce ensures that the agency can better operate across domestic and foreign environments and collect intelligence in support of national security.

In addition, the intelligence community as a whole exists to give policymakers unvarnished findings regardless of what the President wants to hear. ODNI Director Tulsi Gabbard’s decision to fire two top officials who wrote an analysis contradicting President Trump’s assertion that the gang Tren de Aragua was an arm of the Venezuelan government and therefore justifying his deportation plans struck at the heart of the intelligence community’s mission.

David Ignatius of the Washington Post is an expert when it comes to the American intelligence community. In the wake of many politically motivated firings, he had this to say: “Politicization is a special poison for intelligence and national security officials. It skews judgment and encourages false or misleading reporting. It rewards those who curry favor and punishes those who tell the truth. Eventually, it leads to paralysis, as officials become fearful of taking any step that could get them into trouble.”

We will probably never know exactly how many civil servants or contractors were cut from the CIA and the remainder of the intelligence community, since those numbers are classified. But people who do know are worried. Senator Mark Warner (D-VA), ranking member on the Senate Intelligence Committee, has said the cuts “… will undoubtedly undermine our ability to detect and respond to threats and make America less safe.” Others are concerned that people dismissed from these agencies will become targets for foreign espionage by Russia, China, or other rivals of the United States.

The fact that the leaders of the intelligence community have embraced the goals of DOGE means that any malfunction will be politically catastrophic for the administration. Failure in these communities is consequential: we never hear about the successes—terrorist attacks averted, secrets stolen, anti-American operations disrupted—but everyone hears about failures. 

The National Nuclear Security Agency

The early months of DOGE’s cuts were ordered without regard to the consequences. Whether this was intentional and part of Elon Musk’s theory that they “would make mistakes but act quickly to correct” or out of inexperience, time will tell, but certain impacts were immediately apparent. As we saw above, early cuts to people working on avian bird flu prompted a quick reversal in policy. If avian bird flu is dangerous, think of the consequences of failing to properly secure America’s nuclear weapons.

That’s what almost happened. On February 14, the National Nuclear Security Administration (NNSA), a part of the Department of Energy, rescinded termination letters sent the day before to probationary civil servant employees. Almost all but 28 of the 350 NNSA employees who were abruptly laid off had their terminations rescinded; they were part of a much larger DOE purge that targeted 2,000 employees.

The mission of the NNSA is to “ensure the United States maintains a safe, secure and reliable nuclear stockpile through the application of unparalleled science, technology, engineering and manufacturing.” Suffice it to say, there are very few people in the American workforce who have specialized experience in the American nuclear arsenal. Cutting this workforce without a thorough understanding of its role became one more glaring example that DOGE could not approach government downsizing as if it were a Twitter takeover.

The Federal Aviation Administration

The thing about the federal government is that while many Americans dislike it, they love and understand the need for individual parts of it. They don’t want faulty nuclear weapons, terrorist attacks, or eggs at $12 a dozen. Most of the time, Americans aren’t thinking about these government functions happening in the background; however, all Americans who fly know that the federal government controls air travel, and aviation has become one more flashpoint in the DOGE approach.

Back in early March, two secretaries took on Musk in front of Trump’s entire cabinet. One was the Secretary of State, Marco Rubio; the other was Secretary of Transportation Sean Duffy, who is in charge of both air traffic control and aviation safety (a bifurcation that does not exist in other developed countries.) Duffy understood that in matters of safety, accountability would be squarely on him and Trump. The administration had already dealt with a deadly crash between a military helicopter and a passenger jet near D.C.’s Reagan National Airport only nine days after inauguration. Additionally, air traffic control had been understaffed for a long time despite extra money and a government pledge in 2024 to increase the number of air traffic controllers.

But as many, including members of Congress, have pointed out, the supercharged hiring of air traffic controllers doesn’t mean that the threats to aviation are over. More than 2,700 FAA workers signed onto the deferred resignation program, among them technicians who deal with the FAA’s complicated IT systems. Air traffic controllers working with faulty radar and communications systems won’t do much to improve safety. The FAA may find itself having to rescind the deferred resignation offer or hire back people who took it.

Conclusion: The mission matters

In his zeal to make quick progress, Elon Musk created a DOGE initiative that was doomed to face early failures. Cuts were made without regard to established law and the mission of government agencies. However, we are only a few months into this administration, and there is plenty of time for the DOGE team to make substantive improvements to the government. While we don’t know as much as we should about the people Musk brought in, they appear to have a great deal of IT talent, something the federal government has been lacking since the beginning of the information age.

In the next six months, we should keep our eye on two important aspects of this continuing work. The first is its legality—has Trump succeeded in a historic expansion of presidential power? And the second is performance—has the government failed to deliver on things that the public values, such as public safety or emergency response?

All of this will be altered by Congress as it assumes its role in appropriations and encapsulates in the Big Beautiful Bill provisions that may run counter to DOGE’s goals. Another confounding factor is the recent lapses in government data collection, which will limit future ability to evaluate DOGE’s long-term impact. Finally, Trump has fired many inspectors generals and is seeking to terminate the current leadership of the Governmental Accountability Office, which is the congressional office that can undertake precious evaluation work. Weakening these institutions will complicate our ability to understand what DOGE has done.

But one thing we know. While the government in general is often unpopular, many pieces of it are not, and if they fail, as Harry Truman so famously said, “the buck stops here”: at the desk of the president.

The Brookings Institution is committed to quality, independence, and impact.
We are supported by a diverse array of funders. In line with our values and policies, each Brookings publication represents the sole views of its author(s).