Government Executive - Authors - Donald F. Kettl

Insights from the inside: How former feds think we ought to reform the federal government

Donald F. Kettl — Wed, 08 Apr 2026 07:00:00 -0400

We surely haven’t suffered from a lack of ideas for fixing what ails the federal government. There was DOGE, with its campaign to take a chainsaw to federal employment but which in the end saved little money. The conservative Cato Institute’s analysis concluded that its impact on federal spending was negligible and, in fact, that anyone looking at the long-term trend in federal outlays wouldn’t be able to tell when DOGE started.

In a January deposition, former DOGE staffer Nate Cavanaugh was asked, “Did you reduce the federal deficit?” Cavanaugh replied “No, we didn’t.” Never was so much blood shed for such little result. The Office of Management and Budget is now back with a Presidential Management Agenda to take another crack at the problem — in one page.

But despite the problems with DOGE, there’s no doubt that the federal government needs serious fixing. Last December, seven former senior feds gathered to ask, given their decades of experience, what it would take to fix it.

“Or in other words, what would we have changed if we had the power entrusted to DOGE but defined efficiency in terms of effectively delivering value?” they asked.

They marked the first anniversary of the launch of DOGE on Jan. 20, 2026, with a remarkable 39-page report, written by April Harding, entitled “We the Doers.” Their goal was to write a game plan on “how to achieve real government reform in a post-DOGE world.”

Outsiders’ perspectives on reform are easy to come by, from the right, left and center, in which I happily implicate myself. But we’ve never seen anything quite like We the Doers. As Government Executive profiled them back in January, Harding, Maureen Klovers and five other former feds felt “sort of unleashed,” now that they no longer work for the federal government, and they’ve produced a truly important report. A former IRS employee, Harding was told “as a favor” that the administration was eliminating her position. The real favor is to us.

They identify four big problems that the system must, in their view, confront:

The bottom line is undefined: We have lots of performance data, but the data we’ve got doesn’t drive meaningful results because, in their view, the leading current performance system, the Government Performance and Results Act, “has devolved into a bewildering, uncoordinated profusion of metrics, siloed by program and not validated by a uniform methodology that is understood or valued by the American public.” The key, they concluded, is to identify what citizens want and then measure how well the government delivers programs in ways that demonstrate a return on taxpayers’ investment.
There’s no feedback loop with Congress: Legislation tends to be overly prescriptive, with requirements that pile on instead of producing results. Congress has the need for effective oversight but, by the same token, it ought to listen to the doers about whether what it wants is doable. The key, they concluded, is to link what’s desired with what can be done.
The budgetary process is broken: Congress’s budgetary dysfunction, meanwhile, undermines the ability of the doers to manage contracts because the start/stop, lather/rinse/repeat saga of budget battles makes it impossible for government managers and their private partners to deliver value. The key, they concluded, is to focus on the metrics that the public wants and to focus on a return on the public’s investments.
The bureaucratic culture is built around compliance, not delivery: The report sadly concludes, “Even though the members of We the Doers were part of the federal bureaucracy, the truth is we spent much of our time battling it”. There’s grand policy, emanating from OMB and OPM, but the policy often bounces against implementation barriers that the top agencies had never considered. The result is a system increasingly trapped in rules that spin off unintended consequences and lack common sense, especially in the hiring arena, where the result too often is that no one gets hired for important positions. “Government is a gullible customer” in dealing with contractors, and too often “leadership doesn’t lead.” The key, they concluded, is to rebuild a culture based more on delivery than on compliance.

And then, toward the end of their report, the Doers point to the need to “get good at technology.” That’s a need growing exponentially, and it’s a place where the government needs to run to catch up. Government needs to smash the boundaries that too often separate the budget, people power, decisions and measurement and focus all of them toward the outcome the government wants to produce. It needs to grow past cool websites to an experience that weaves together the people with the service the government is providing. It needs to bring more of its IT work in-house, so it’s no longer as captive to private contractors, and to boost its cybersecurity. It needs to make data sensible to ordinary members of the public, to reinforce its clarity on the inside. And it needs to “declare bankruptcy” on existing, sprawling IT systems, both hardware and software.

This is a broad and ambitious agenda. Some of the points are familiar; all of them, as the Doers lay out in the report, have a real-world grounding that’s reinforced by the experience that backs them up. It’s all the more ambitious because it stretches so far what the 2027 budget proposes. When it comes to the “results” of the budget, it focuses on “Buy America.” When it comes to people, it calls for “removing poor performers and limiting hiring to essential jobs”.

For anyone truly interested in reforming what ails the federal government, this report is essential reading. It provides insights from the inside about how to make changes from the outside in. The authors certainly aren’t circling the wagons in defense of the current system. And their report shows what DOGE done right would look like.

]]>

Memo to JD Vance: Fighting the War on waste

Donald F. Kettl — Thu, 26 Feb 2026 15:51:00 -0500

During President Trump’s State of the Union address, the president said, “I am officially announcing the war on fraud to be led by our great Vice President JD Vance.” And, he boldly added, “we’re able to find enough of that fraud, we will actually have a balanced budget overnight. It’ll go very quickly.”

The pledge to balance the budget by eliminating fraud is an old one. Clinton, Reagan, Obama and Biden all said they would wring out fraud and they never got close to the pot of gold at the end of that rainbow. Moreover, the idea that we could remotely come close to balancing the budget by eliminating fraud is ludicrous. The deficit at the end of fiscal 2025 was $1.8 trillion. The Government Accountability Office estimates that the amount of fraud in the federal government stands between $233 billion and $521 billion.

There are two things we can conclude. Even at the best, eliminating every possible dollar of fraud would only cut the deficit by a little more than 25%. But $521 billion ain’t chump change. We need to wish Vice President Vance godspeed in what’s surely an important mission.

So: here’s a memo to the vice president about how to start.

Dear Vice President Vance,

You stand in the shoes of many previous generals in the war on waste. The launch of their campaigns has always created great headlines, but their results have usually been disappointing.

Let’s start with two boring but important facts. One is that no one really knows how much fraud there is. Money can leak like water from an old water main—you know how much you’re putting in, you have a fair idea about what’s coming out and you can’t always tell where the leaks are along the way.

The second is that “fraud” is a deceptive label. It’s clearly wrong to create a bogus identity to milk cash from the federal government, but it’s often very hard to separate a good program poorly managed from a bad program criminally draining cash.

In the tales of rampant fraud among Minnesota day care centers, which prompted the latest headlines, some of what is presented as fraud actually is sloppy overbilling. Many of the day care centers had state licensing violations, including keeping the facilities clean and keeping records of vaccinations, but state investigators found very few instances of outright fraud. A 2025 federal report found that there were errors in 11% of the payments made to Minnesota day care centers, so there clearly are big problems. Figuring out what’s causing them is much tougher than appears on the surface.

These twin facts often sent the generals of previous wars on fraud to give up the battle and look for other wars to fight, but you can make big progress. Here’s how:

Know the enemy. Announcing an attack on “fraud” is tempting. Delivering is a lot harder, because fraud is like fat marbled through a terrific steak. A clumsy job of trying to cut it out can miss the fat and turn the steak into expensive hamburger.

Identify the target. The place to start is with “improper payments,” which GAO defines as “payments that should not have been made or that were made in an improper amount.” It’s a clumsy term. Not all improper payments are fraud—some are legitimate but trapped in sloppy bookkeeping. Not all fraud comes from improper payments—some people commit non-financial fraud, like illegally obtaining a passport. But improper payments are the place to start.

Go where the money is. The bank robber Willie Sutton was right—the highest potential for recovering money must start by attacking the biggest targets. More than half of all improper payments—53%—come from just Medicare and Medicaid. Another 7% in SNAP. There are tax-related improper payments, especially in the earned income tax credit. But in general, the rest of the government accounts for $41 billion in improper payments. This suggests where to aim the war on fraud.

Invest in the front-line troops. This means shoring up the inspectors general, who are the government’s fraud-fighting cops. It means investing in the key federal employees, since one person’s salary can leverage enormous amounts of money. The Centers for Medicare and Medicaid Services, for example, spends $252 million for every one of its employees.

Strengthen risk management and financial control activities. Sounds boring, I know. In fact, the 1990s sitcom Seinfeld spent a whole episode getting laughs from just how boring risk management seemed, and the episode won an award from the Writers Guild of America. But the key to separating fraud from other federal activities requires knowing where to look and having the tools to poke under the blankets. Most important, it’s essential to build roadblocks to fraud before it occurs. That’s a whole lot cheaper and more effective.

Develop new tools. Artificial intelligence seems invented to fight the way on waste. In the Medicaid program, for example, the biggest sources of fraud are billing for unnecessary services or services no one actually provides, upcoding (bumping up the level of complexity of a service to make more money), card sharing and obtaining drugs for sale on the side, among others. In a program with millions of transactions, this is a true needle-in-a-haystack problem. AI can help identify the right haystacks and find the most promising needles.

Understand that some problems that look like fraud are actually the result of poor management capacity, that others actually are fraud and that it’s often hard to tell the difference. Billions of dollars of federal money go to small nonprofits without much experience in managing programs or keeping track of the money. On the other hand, bad actors are always working hard to bilk the federal government out of taxpayers’ dollars. An effective war on fraud requires devising separate strategies for each of these problems—and figuring out when to use which one. It’s a problem that’s been around as long as there have been wars on waste.

Finally, a word of warning. It can be tremendously tempting to use the war on fraud to target political opponents. That’s easy because, unfortunately, there’s fraud everywhere. The rate of improper payments isn’t a measure of fraud, but it does give us clues about where to go looking.

Minnesota, for example, has an improper payment rate of 2.2%, but that’s less than half of the national average of 5.9%. The blue-state average is 6.3 percent. In blue Delaware, it’s 19.6%. Connecticut’s rate is 19.8%. But it’s not a partisan problem. The red-state average is 5.7%. In red Wyoming, it’s 20.7%; it’s 20.5% in South Carolina; and it’s 18.7% in Idaho.

There’s big money to recapture from controlling fraud and improving management. Any serious campaign to rake the money back, however, will fail if it’s a partisan war on waste. Nobody has a monopoly the problem—or the solutions.

So, Mr. Vice President, you are heading down a road that many have traveled before. Many travelers in the past have gotten stuck in ruts and potholes. But if you want to make real progress in saving taxpayer dollars, this is the road forward.

Donald F. Kettl is Professor Emeritus and Former Dean at the University of Maryland School of Public Policy. He is a Fellow of the National Academy of Public Administration.

]]>

The tail wagging the dog: Snapshots of the public service a year into the second Trump administration

Donald F. Kettl — Thu, 05 Feb 2026 18:24:00 -0500

In the months leading up to the 2024 election, Elon Musk honed his pitch for a Department of Government Efficiency. The DOGE acronym was no accident. Musk was taken in by the tremendous success of the Dogecoin, the virtual currency that began as a joke in 2013 but quickly became wildly popular, although its founders continue to use an adorable dog as its online mascot. The dog, in fact, became so popular that it’s got its own Dogecoin swag store.

Musk had been taken by Dogecoin for years. So when he got his chance with the Trump transition, he couldn’t resist finding a way to use DOGE as an acronym for his campaign against government.

But after all the headlines and turmoil, just what has DOGE and the administration’s other strategies to slash federal employment produced? The Office of Personnel Management has just released an enormous tranche of data—not nearly as detailed and useful as FedScope, which it replaced, but useful nevertheless for creating snapshots of the changes in the federal workforce since Trump’s inauguration. (The old FedScope has vaporized, although data can be extracted through a very complex series of steps.)

The snapshots nevertheless tell a fascinating story of how the DOGE tail has wagged the federal dog.

The administration eliminated 9% of the federal workforce

The Trump administration reduced the federal civilian workforce by 209,775 employees, a 9% cut. This size of the workforce isn’t unprecedented, however. It’s almost exactly the same now as in 1994 and 1973, although of course both the country’s population and the budget were much smaller.

Every department is smaller now than before Trump’s inauguration

All cabinet departments suffered cuts, ranging from a very small cut in the Department of Homeland Security to a 69% cut in the Department of Education. The Department of Housing and Urban Development and the National Science Foundation come next, with reductions of about 40% each.

There were enormous changes in the Department of Homeland Security

Perhaps no department had wider changes than DHS. Immigration and Customs Enforcement and Customs and Border Protection got big increases in the number of employees. The Federal Emergency Management Agency suffered big cuts. And as Customs and Border Protection grew, U.S. Citizenship and Immigration Services, the branch of DHS charged with managing legal immigration, shrank.

The geographical distribution of federal employees has changed slightly but significantly

For most members of the public, a huge surprise has always been that most federal employees work outside the Washington area. That's even more so now. In fiscal year 2024, 85.9 percent of all feds worked outside the Washington area. Today, it's 86.6 percent—that is, the share of federal employees working outside Washington has increased, which is in sync with the Trump administration's commitment to reducing the power of Washington civil service professionals.

Most states saw relatively little change in total federal employment, but there have been some big winners and losers. The losers are an unusual collection of both red and blue states, with red-state losers (Wyoming, Montana and Kansas) typically smaller and more dependent on relatively small operations like the Veterans Affairs Department and national parks. The nine winners likewise are a disparate collection: six states that went for Kamala Harris (Hawaii, Connecticut, California, Minnesota, New Mexico and Virginia), and three states that went for Trump (North Carolina, Florida and Texas). Programmatic changes drove these patterns, but there’s no evidence overall of political targeting—except at the Washington area.

Employees in lower General Schedule grades took the biggest hits

One of Trump’s most powerful themes was that he was going to “drain the swamp.” The implication of the pledge was that federal employees in policy-related positions would take the biggest cuts; they, presumably, were in the strongest position to block the administration’s policies.

In fact, however, the largest cuts, in terms of percentage of the workforce, came at lower levels of the General Schedule. There were also significant cuts at the GS-7 and GS-9 levels, presumably as part of the administration’s strategy of firing probationary employees.

The Senior Executive Service was a major target

In terms of “draining the swamp,” however, the SES was a major target. These senior executives are among government’s most experienced employees, and they serve as a critical shock absorber between political appointees and the career service. There are now 9.4% fewer members of the SES than in 2024.

Public employee unions also took a hit

The share of federal employees in a bargaining unit dropped from 56.2% in November 2024 to 37.9% in November 2025.

Employee performance ratings dropped

There’s been criticism of the federal employee rating system for many years, especially because critics have believed far too many employees get the highest ratings. The administration took a big step in reversing that in 2025. The number of employees rated in 2025 was just 29% of the number in 2024. But of those employees rated, the share receiving “outstanding” and “exceeds fully successful” dropped from 64% in 2024 to 48% in 2025.

Looking for a federal job?

Not a great time, needless to say. The number of job announcements in 2025 dropped 43%, compared with 2024. But if you are looking around, the Department of Defense is the best bet, followed by the VA and the Department of Justice (especially since, as always, DOD and the VA have the largest number of employees).

Your best chances for finding a federal job are in California and Texas, followed by Virginia. The odds are slimmest in Vermont, Iowa, Kentucky, Arkansas, West Virginia, Delaware, New Jersey, Rhode Island and Connecticut.

Donald F. Kettl is Professor Emeritus and Former Dean at the University of Maryland School of Public Policy. He is a Fellow of the National Academy of Public Administration.

CORRECTION: An earlier version of this article listed Minnesota as a red state that increased its federal employment. Kamala Harris won the state in 2024.

CLARIFICATION: An earlier version of this story referred

]]>

Broken policy breaks budgeting even more

Donald F. Kettl — Wed, 28 Jan 2026 13:38:43 -0500

Furious at the killing of two American citizens in Minneapolis, Capitol Hill Democrats say they won’t give Immigration and Customs Enforcement any more money until it cleans up its act. “What’s happening in Minnesota is appalling —and unacceptable in any American city,” said Senate Democratic Leader Chuck Schumer, D-N.Y.

Sen. Jacky Rosen, D-Nev., joined the growing chorus of senators saying they would not vote to continue ICE funding past the Jan. 30 deadline. “Enough is enough. We need to rein in ICE’s out of control conduct,” she posted on X.

Behind the fusillade are two important truths. One is that the Democrats are determined to use the budget, the only big tool they have at hand, to hold ICE accountable. The other is that pretty much everyone agrees that the budgetary process is broken and that using it in the middle of a pitched battle will only break the process more.

Nobody wants to drive deeper cracks into the budgetary process. But the Democrats believe they have only one option to deal with ICE: draining the agency’s cash and shutting it down.

The Democrats are playing the hand that the Republicans inadvertently dealt them at the end of last fall’s shutdown, the longest in American history. The House passed all 12 of the appropriations bills. The Senate passed half of them and held onto the rest. The Republicans, who controlled the chamber, could load them all into a “minibus” to pass them together or feed them into the process one at a time. They went for the minibus, a name borrowed from the 1990s for party buses. But they brought no party.

The Democrats won the promise of a vote on health benefits and an extension of Supplemental Nutrition Assistance Program food benefits. And the deal continued the crisis already in progress, with the next phase this week.

These crises have become more regular and more damaging, not just for the politics of the system but also for the basic order of the government. We’ve been worrying about the erosion of trust in government but seem determined to make it worse.

Six of the nation’s biggest budget experts—Doug Criscitello, Ed DeSeve, Teri Gullo, Phil Joyce, Roy Meyers and Steve Redburn—have just collaborated on a manifesto, calling everyone to action for a “Radical Reform of the Federal Budget Process.”

The team writes, “Having endured the longest-ever government shutdown at the beginning of this fiscal year, even the most casual observers should see that the federal budget process is not meeting its most basic goal: funding the government. Nor does the process help leaders allocate resources to deliver reliable results for the American people.”

The budgetary process was always chock-full of politics. It lies right at the core of the system that uses government’s power to figure out how to allocate always-scarce tax dollars. It was the biggest program that Madison and his colleagues had to sort out when they drafted the Constitution in 1787.

But, as the authors of the manifesto point out, the system’s politics are playing themselves out in new ways that hyper-politicize the process and undermine any expectation for doing things in a way that makes it possible for anyone to have reasonable expectations about anything.

The executive branch is refusing to spend money that Congress appropriates and spending money that Congress hasn’t authorized. The president hasn’t made a detailed budget proposal. Congress isn’t getting its work done on time. It’s been 30 years, in fact, since it has completed its appropriations bills before the start of the fiscal year. And there’s increasingly no connection between the money coming in and the cash going out.

We don’t have to keep doing this to ourselves. The manifesto lays out 10 characteristics of an effective budget process:

Comprehensive: including all of government’s financial resources.
Legitimate: only legally responsible officials should make budgetary decisions.
Honest: decisions should rest on accurate and timely information.
Transparent: taxpayers ought to be able to understand budgetary decisions.
Timely: budgetary decisions should be completed on time.
Cooperative: the budget should complement, not compete with, other key decisions.
Perceptive: the budget ought to take a long-term view.
Judgmental: taxpayers ought to expect decisions that produce the highest value for the least cost.
Constrained: the process ought to put brakes on how much policymakers can spend
Responsive: budgetary decisions ought to match what the public wants

It’s been 50 years since we’ve rebuilt the budgetary process, with the Congressional Budget and Impoundment Control Act of 1974. The nasty headlines make clear it’s time to do it again. And the manifesto lays out a game plan for how to do it—now, and even better than the last time.

But two things we don’t want or need. One is trying to cut politics out of the process. That would be like trying to start a restaurant without ordering food.

The other is gaming the process, as happened with the current minibus, is only looking for trouble. It’s blown up in the Republicans’ faces, with just the wrong issue at the wrong time. It’s irresistible catnip for the Democrats, but playing the game only sets up the next crisis and further lowers the public’s trust.

We need a budgetary process that serves the highest purposes of government. The manifesto provides the foundation for making it happen.

Authors of the Manifesto will be discussing the project on Thursday, Jan. 29 at 2:00 pm ET, in a forum open to the public. Sign up to register.

Donald F. Kettl is Professor Emeritus and Former Dean at the University of Maryland School of Public Policy. He is a Fellow of the National Academy of Public Administration.

]]>

Why America needs the GAO: DOGE done right

Donald F. Kettl and Philip G. Joyce — Wed, 26 Nov 2025 10:00:00 -0500

Ten months after the launch of Elon Musk’s Department of Government Efficiency, one thing is clear: no one can agree on how much money it saved—or, in fact, whether it saved any money at all—and, if there were savings, what happened to them.

What if DOGE were done right? Here something else is clear: we have had the Government Accountability Office, founded as a green-eye-shade operation back in 1921 to audit every single government transaction. It’s grown into the indispensable watchdog on the federal government’s spending and operations. It’s what DOGE done right looks like.

At first, Musk promised to find $2 trillion in savings. Then the pledge dropped to $1 trillion. Then he struggled to document any savings at all.

Since 2002, GAO’s work has logged more than $1.45 trillion in savings for taxpayers, across more than 29,000 different federal operations. Last year, savings were $67.5 billion. And better than most watchdogs glad to gobble up every available can of dog food, GAO has kept watch on itself. It’s measured the return on taxpayer investment in its own operations: $123 for every budget dollar.

It's operations aren’t flashy. In fact, they’re pretty wonky, which as card-carrying wonks we greatly admire. Its staff get advanced training when they walk into the door, which never stops as employees move up through the ranks. The size of its staff, which was reduced from 5,200 staff to it current level of about 3,500 after the Republican takeover of the House in 1995, has been relatively stable over the last decade, despite expanding demands on its work.

Musk promised to trim the number of federal agencies. “There’s so many that people have never heard of. I think we should be able to get away with 99 agencies,” he said. It wasn’t clear where that number came from. It was clear from his work that DOGE didn’t know how to get there.

GAO, on the other hand, has been identifying areas of duplication and overlap since 2011, with $725 billion in savings in this area alone. As others were struggling to come up with solid savings, GAO in May issued a 154-page report detailing just how to do more.

Musk brought in IT whiz kids to comb through the federal government’s information services, but they had little to show for it except putting the security of Americans’ personal information at risk. GAO, in contrast, laid out a plan for figuring out how to save money on IT. Better proof-of-concept work in space-based laser communications could save hundreds of million dollars—and the government could pick up $157 billion by better managing the payment rates in Medicare.

In a February Oval Office meeting, Musk claimed the federal government was full of officials who had approved money for “fraudsters,” but that claim turned out to be little more than arm-waving. GAO, in contrast, has since 1990 prepared a report every two years on programs prone to waste, fraud, abuse and mismanagement. Its latest report, in February 2025, identified 38 areas needing special attention.

This “high-risk list” bears careful reading for anyone who really cares about getting into the guts of the problem. It identifies issues ranging from contract management in the Pentagon to money leaking out of federal disability programs to management of oil and gas leases to mega-challenges inside the Federal Emergency Management Agency’s disaster assistance. Even a casual look of the report—all 323 pages of it—would convince any reader that it’s a road map for which stones to unturn to get at the unholy trinity of waste, fraud and abuse.

But there’s more here. GAO keeps score so we can see who’s making progress—and who isn’t. In 10 areas, federal agencies made $84 billion in progress. Three slipped backwards—acquisition of weapons systems by the Pentagon, improving IT management throughout the government and managing the federal government’s property.

And even more. GAO doesn’t operate like a sheriff riding into town promising to clean up the saloons. It has developed a big-picture look at federal operations and identified leading practices that federal agencies can pick up. Like any oversight body, federal agencies don’t always like what they hear and often squawk. But there’s no question that we—and even they—are a whole lot better off for GAO’s work.

GAO is relatively little known outside government. Even less well known is the fact that it’s an agency that works for Congress. That often is a thorn in the side of the administration, regardless of who’s in power, because GAO often says things the administration would rather not hear. That’s especially been the case in recent months with GAO’s findings that the administration has violated the law in impounding federal money—refusing to spend money that Congress appropriated. This was at the root of Office of Management and Budget Director Russell Vought’s statement that GAO “shouldn’t exist”. The OMB Director doesn’t want anyone looking over their shoulder, just as students don’t like to take home report cards that aren’t straight-A’s or spend their time in detention.

But now that we’re closing in on the 250th anniversary of the signing of the Declaration of Independence and celebrating the Constitution that the founders crafted to govern us, we need to strengthen the balance-of-powers model that Madison and the other founders left us. They deliberately chose not to give us an easy system to use in governing ourselves. That means the three branches of government are always going to be jousting with each other—and that’s how it should be.

In the growing complexity of our society, we need governmental institutions that can keep up with that complexity. That means Congress, in turn, needs muscle to do the heavy lifting required to oversee the executive branch, whether it’s a Republican or Democratic Congress, or a Republican or Democratic President. That’s what Washington, Madison, Jefferson, Hamilton and the other founders had in mind.

And that’s why the next month is so important. The current head of GAO, Comptroller General Gene Dodaro, will be at the end of his fifteen-year term, the longest term in the federal government except for the lifetime appointment of federal judges, and the process to name a successor will shift into high gear. A special commission—consisting of congressional leaders—recommends a slate of individuals to the president. The president selects one person to nominate, and the Senate must then confirm the new head.

This appointment is a pretty big deal. It will ensure strong oversight through the 2030s into the 2040s and reinforce the balance of powers for even longer. GAO will provide the executive branch with the independent insights it needs to wring out waste, fraud and abuse. It is a tool that Congress can use to hold the executive branch accountable, essential because a weak GAO weakens the legislative branch. We all have a stake in making sure that GAO can provide the oversight the Constitution requires—and the balance of powers that they designed.

Philip G. Joyce is professor of Public Policy and Donald F. Kettl is Professor Emeritus, both at the University of Maryland School of Public Policy. They have been working with colleagues on Reform for Results, a nonpartisan coalition of public policy experts dedicated to advancing the government’s mission within the American tradition of the rule of law.

]]>

How the president expanded his power without a government

Donald F. Kettl and Philip G. Joyce — Tue, 21 Oct 2025 17:54:00 -0400

When the government shutdown ends, Donald Trump will have succeeded in staging the single biggest expansion of presidential power in American history because of the single largest shift in the constitutional balance of powers ever.

In fact, he (but more particularly, his team) has been so successful at maneuvering through this shutdown that there’s no reason for them to end it. Democrats have called on Trump to get more involved in the negotiations. They’ve gotten little more than a shrug in return. The Republicans are winning, and Vegas card players know never to leave a winning hand on the table.

The Republicans are using two of the big tools to shape governmental action: the power of the purse, which funds what government does; and the power of government bureaucrats, who make it work. The former is the engine. The latter are the wheels. And by vastly expanding the administration’s leverage over both, it—especially OMB Director Russ Vought—is in the driver’s seat.

This is vastly more important than Elon Musk’s Department of Government Efficiency, an unguided bulldozer rambling through government. Vought’s strategy is all out of a single piece of carefully woven cloth.

Start with the engine: the federal budget. Madison, in Federalist #58, made it clear why the founders, concerned with the prospect of a strong executive, vested the power of the purse with the Congress. It was, he wrote, the “most complete and effectual weapon with which any constitution can arm the immediate representatives of the people”. There are two vital components of this: the president cannot spend money for purposes not authorized in law; and that the president cannot pick which programs he wants to spend money on and on which he does not.

There are laws, which Georgetown law professor Eloise Pasachoff has referred to as the “power of the purse statutes”, that reinforce each of those constitutional imperatives. The Anti-Deficiency Act, passed in 1870, creates penalties for the unlawful spending of money that has not been appropriated, or not appropriated for that purpose. The 1974 Impoundment Control Act creates the only legal avenue for the president to refuse to spend money that has been appropriated. The Vought strategy violates both.

For the first almost eight months of the administration, the sole emphasis has been on the latter set of violations. In most cases where the administration attempted to cancel programs, it did so without following the procedures set up by the ICA. In the shutdown, the administration has crossed the ADA lines

With the deadline for the military’s payroll looming, Trump ordered the Pentagon “to use for the purpose of pay and allowances any funds appropriated by the Congress that remain available for expenditure in Fiscal Year 2026.” The administration said it was going to use funds in the Pentagon’s research and development account. DOD did indeed have unexpended R&D funds, but Congress had not authorized draining the R&D account to pay the military, and that’s a clear violation of the ADA. As Bobby Kogan at the Center for American Progress posted on Bluesky, this mechanism “is by far the most illegal budgetary action he's taken as POTUS, potentially setting the stage to break everything.”

How does it risk breaking everything? When the administration can both decide it can spend money from any budget account on anything it wants, AND that it does not have to spend money appropriated by Congress if it does not want to, there are no limits to the budgetary powers possessed by the president. This is exactly the kind of executive overreach that worried the founders.

The Trump administration, as one of us has concluded, “appears to understand that control over the federal budget is central to control over the entire federal government.” The budget maneuvers mark a huge attack on the norms that have defined the constitutional separation of powers for 238 years.

Then let’s move to the wheels: government employees. DOGE had already virtually eliminated the U.S. Agency for International Development, large parts of the Department of Education and other parts of the government deemed “woke.”

But when the shutdown began, the firings ramped up. OMB had already asked for plans from agencies about employees who should be RIFed—fired through a “reduction in force.” The firings began with a short post from Vought, a few days into the shutdown: “The RIFs have begun.” He followed that up with more than 4,000 RIFs, and it took digging into a federal court filing to find out where they were occurring, since there was no transparency from OMB on this important step.

It didn’t take long for a scramble to begin. The Department of Health and Human Services at first said that 982 people got RIF notices. Then word came that the notices went out to 1,760 employees, including 596 employees at the Centers for Disease Control, along with more workers at the Substance Abuse and Mental Health Services Administration and the Administration for Children and Families, among others.

The administration used the RIF process to accelerate its planned shutdown of the Department of Education by laying off more employees. It slashed employment at the IRS by 1,446 people, on top of the 25 percent of the workforce who were DOGEd. In just a few days, the administration eliminated many workers at agencies that were at the bull’s eye of its strategy to transform the executive branch.

With the purse nowhere to be found during the shutdown, Vought seized its power, and his decisions in the fog of the shutdown means there’s no accountability.

The country’s founders had anticipated the risk of executive power but counted on Congress’s control of the power of the purse to rein in the irresistible instincts for expanding presidential power. One of the biggest champions of the executive, Alexander Hamilton, pointed to the importance of Congress’s role. “The legislature not only commands the purse, but prescribes the rules by which the duties and rights of every citizen are to be regulated,” he wrote in Federalist 78. But that only works if Congress chooses to exercise that command—and that it is actually meets to do so.

As always, when the supporter of the Congress, Madison, and the champion of the executive, Hamilton, agree, there’s a very powerful lesson the founders teach us. They’d be aghast at the government that emerges from the shutdown. And given the way the ongoing battles are going for the administration, there’s no reason for the administration to settle the shutdown any time soon.

]]>

The ultra risks of a routine shutdown

Donald F. Kettl — Mon, 29 Sep 2025 13:35:00 -0400

It would be easy to think about the prospects of a government shutdown as the next act in a tired old play. But what’s looming this time is far, far different, with incalculable stakes. It could amount to the most dramatic shift in the separation of powers in living memory—or longer.

That might sound raving, but consider this.

Without a new budget passed on this not-so-happy-fiscal-new-year, only “excepted” parts of the government can operate. The Antideficiency Act prevents spending on anything else. Employees go on an unpaid furlough, and they’re not even allowed to use government-issued laptops, cell phones or email accounts. (None of this applies, of course, to employees funded by separate revenues, like Social Security, Medicare or the Postal Service.)

And just who is an “excepted” government employee working in an “excepted” program? OMB has a surprising amount of discretion to define this, through Circular A-11.

Here’s where a looming shutdown this time will be far, far different than the previous ones in 1995, 2013 and 2018. According to reports first published by Politico, OMB Director Rusell Vought instructed agencies to prepare RIFs for all employees without separate funding and whose work “is not consistent with the President’s priorities.” The RIFs could apply even to employees working in “excepted” programs.

Just a momentary lapse in appropriations—a shutdown that lasts minutes or hours—would give Vought the power to tell agency leaders to move past the usual furloughs—“temporary, nonduty, nonpay status”—to RIFs—being permanently fired. Vought could choose the programs that the administration has been wanting to eliminate and give a very big haircut to others.

The result would be a dramatic, instantaneous shift in the separation of powers. Let me underline this: while the purse is empty, OMB would seize the power of the purse from Congress. The administration could choose to prolong a shutdown for as long as it liked, through poison pills in the negotiations with congressional Democrats, and comb through the government to flatted programs it’s been trying to cut. With this prospect looming, the administration has little incentive to do more than shed crocodile tears over a shutdown.

And then, when the shutdown is over, whenever the administration chooses to end it, those employees who were furloughed would come back on the payroll, with back pay, through a 2019 law. The employees who were RIFed would not.

This would be an historic—and incredible—invention of a backdoor impoundment. OMB could prevent the administration of programs it opposes by gutting the government’s capacity to administer them. The Trump team could kill programs unilaterally without the inconvenience of going to Congress.

To top it all, this would all be perfectly legal.

Under the Antideficiency Act, OMB would be well within its powers to decide which programs could not continue. It could decide that programs that aren’t “excepted” had to stop. It could then RIF employees working for those non-excepted programs, and any other employees it chooses. RIFed employees would have no appeal rights and the government has no obligation to bring them back when Congress—finally—passes an appropriation, as the University of Minnesota’s Nick Bednar has concluded.

They’d be gone, and so would the programs they manage.

It would be a bigger deal than the Musk-led DOGE at the beginning of the administration. Bigger than Schedule PC or Schedule G. It would give real and very sharp teeth to Project 2025’s ambitious goal to turn government upside down.

And it sets a trap for the Democrats. They could roll over and agree to all of the administration’s demands, as they did back in March. That would weaken the party further as leaders try to right the ship. Or they could refuse to give in, trigger a shutdown—what the Republicans are already calling a “Schumer Shutdown”—and then stand back to watch an awesome stripping away of their power.

Either way, the Democrats lose. It’s like an old western, where cowboys ride into a box canyon with no way out.

And, with the looming prospect of an enormous cut in government employees, a slashing of government functions and an historic shift in the power of the purse, this is the very embodiment of a Very Big Deal.

In Federalist 58, James Madison wrote “This power over the purse may, in fact, be regarded as the most complete and effectual weapon with which any constitution can arm the immediate representatives of the people,” the Congress. With his memo to federal agencies, Russell Vought turns Madison on his head, for he would strip Congress of the power of the purse, invent a new and legal backdoor impoundment, and disarm the representatives of the people.

And even if both sides manage an agreement, the club in that memo will always be behind the door.

]]>

Sharing services is essential for the federal government. Here’s how to make it work

Donald F. Kettl — Fri, 01 Aug 2025 09:00:00 -0400

Updated at 5:07 p.m. on Aug. 3, 2025

The inescapable reality of modern government is that no organization can control or fully manage any problem that really matters. Sharing services is the inescapable reality.

Scott Ruskan makes that point. Immediately after a river tsunami swept through the Texas Hill Country, Ruskan and the crew of a Coast Guard helicopter based in Corpus Christi, 200 miles away took off for Camp Mystic. Ruskan said that they encountered “some of the worse flying conditions we’ve been in.” They snagged five campers and then put Ruskan on the ground to coordinate the rescue of 200 more campers.

Saving lives in those awful days took the shared contributions of federal, state and local governments. In addition to the Coast Guard, help came from the federal government’s Customs and Border Protection, and FEMA, as well as state emergency rescue teams and the local first responders. The Department of Homeland Security pointed to “unprecedented action alongside Texas first responders for search and rescue operations.”

We’ve discovered, yet again, that sharing services lies at the core of so much of what government does. But this reminds me of a story that a wry senior official at the Office of Management and Budget once told me. He said that, toward the end of his time at OMB, someone came and whispered in his ear, “The answer to your problems is to collaborate!” He slapped his forehead and said, “Well, why didn’t I think of that??? !!!”

If this is such a great—and inescapable—idea, then why is it so hard?

A big answer, according to a pathbreaking University of Maryland study conducted in concert with the Shared Services Leadership Coalition, is that too often it’s too hard to develop performance metrics that point to “‘mission benefit’—how shared services contribute to agencies’ goals beyond efficiency.”

Too often, in addition, agencies can’t shift gears fast enough to focus on new problems—and “maintain long-term value and relevance.”

Solving these problems requires six “M” steps.

Motivation. If it is such a great idea, then why aren’t we doing more of it? Like the conversation that the former OMB official had, the need for collaboration and sharing seems obvious—until someone tries to do it. The underlying problem is this: What is the question to which sharing and collaboration are the answers? And the question is this: How can we make sure we can line up the government’s capacity to the problems that need to be solved?
Mismatch. The deeper problem is this: Too often, we take a vertical approach to an increasingly horizontal world. But, as I have previously written, there’s no problem that matters that any one organization can control. Our world is increasingly horizontal; vertical approaches, based on hierarchy, are bound to fail. This is a particular problem with many of the DOGE-based reforms. They’re all based on a hierarchical view of the world.
Missing link. When we do focus on a program’s results, the keystone of sharing and collaboration, too often we do not align the program’s performance standards with the incentives for individual performance. Collaboration is always riskier than staying safely inside the hierarchical cocoon. Without incentives for effective collaborative problem solving, collaboration doesn’t work—and problems don’t get solved.
Masses. Too often, the people that government tries to serve get lost in the hierarchical chair or the performance metrics. People have to come first in the quest for results.
Muscle. We have a world where, inevitably, there are fewer governmental employees, more pressure on government workers and rising expectations about government problem solving. This doesn’t mean we need to hire lots more government workers, but we do need to make sure that we have the right muscle behind our public ambition: The right workers in the right places at the right times. This is the new face of strategic planning.
Machine learning. It’s tempting to paint AI as either the grand salvation of all problems or the great evil sure to make them all worse. The reality is different: Information is the central nervous system of the increasingly horizontal world, and AI can provide the language for improving that system.

It’s tempting to conclude that all of this is very complex (which it is) to the point that it all requires new policies, regulations, executive orders and legislation (which it does not). In fact, it’s like the story of Dorothy in the Wizard of Oz.

Dorothy had a pretty rough time of it: A whirling encounter with a tornado, a challenge from a kindly witch to find the yellow brick road, a scary walk through woods and encounters with unusual creatures, a visit to a wizard who demands she bring back a wicked witch’s broomstick, all to help her get back home. It’s a charming story (although, truth be told, the flying monkeys still creep me out).

But one part of the story is the most important of all for our tale today. At the end of the movie, the Good Witch Glenda tells Dorothy, “You’ve always had the Power, my Dear. You just had to learn it for yourself.” Part of me is really annoyed for Dorothy—why did she have to go through all the nasty steps only to be told she had the power in her shoes to begin with?

Part of me, however, finds it heartening in this lesson for collaboration and sharing: We don’t need a mystical intervention. We have had the power all along—we just need to use it. The six “M’s” tell us how.

Correction: An earlier version of this report incorrectly identified the Shared Services Leadership Coalition as the Shared Services Association.

]]>

A time machine tour of civil service reform

Donald F. Kettl — Fri, 20 Jun 2025 07:00:00 -0400

In 1895, science fiction writer H.G. Wells introduced the world to the concept of time travel. Pick a time, he told his dinner guests in The Time Machine, and manipulate a lever in his machine: “this lever, being pressed over, sends the machine gliding into the future, and this other reverses the motion.”

That’s the stage we’ve set with the current debate on the merit system. The real debate between the Trump administration reformers and others in the battle is where they want to set the lever for merit system principles. If truth be told, no one—right, left, center—likes the current merit system.

So, if we know what we don’t like, where should we set our time machine setting? Here are some settings.

George Washington and the “fitness test” (1789). Washington insisted that employees of the new government must have great capability and a strong reputation for honesty. He also considered an ideological screen: were applicants supporters of the colonies before the Revolution. And, in addition, being a Federalist helped. He gave preference to former officers, but not enlisted men, in the Revolutionary Army.

Andrew Jackson and the “spoils system” (1829). After Jackson’s election, job-seekers descended on Washington like a plague of locusts. Jackson was convinced that the officials already in the federal government were a deep-rooted special interest more intent on keeping their jobs and power than doing what the president wanted. Loyalty to the president was more important than a candidate’s professional qualifications. Applicants sometimes paid a year’s salary to buy a four-year job.

The Pendleton Act and the creation of the civil service system (1883). After the assassination of President James Garfield in 1881 (it is obligatory in all accounts to refer to the shooter, Charles J. Guiteau as a “disappointed office seeker”), a bipartisan movement arose to create genuine reform. Sen. George H. Pendleton, D-Ohio, introduced a bill on civil service reform, and he got support from President Chester A. Arthur, R-N.Y. The Pendleton Act established that federal employees would be hired on the basis of merit. It also created the Civil Service Commission to manage the new system. The act covered about 10% of all federal employees. Beyond protection from being fired for political reasons, the Pendleton Act had no basic protections for federal employees, especially on removals. That was up to the president to decide.

Passage of the Lloyd-LaFollette Act and related presidential action (1912). Established protection to federal employees against dismissal except for cause and defines the procedures that supervisors must follow in doing so. Federal employees gained the right to join unions and they could not be fired for doing so, although they could not strike. The coverage of the civil service system expanded to about two-thirds of all federal employees.

World War I-era (1944-1949). Veterans preference was established, with vets receiving a points bonus in ranking applicants. Retirement provisions were improved. The Classification Act of 1949 created a much more simplified collection of grades and salaries; the GS system we know today, with standardized job titles and grades across the federal government, came from this act. By 1950, about 80% of all federal employees were covered by the merit system.

The Civil Service Reform Act and related presidential actions (1978). Moved civil rights protections to the Equal Employment Opportunity Commission. Abolished the Civil Service Commission and created the Office of Personnel Management, with authority over the examination, appointment, compensation, training, awards, and benefits for federal employees. Created the Merit Systems Protection Board to decide appeals from federal employees about possible abuses of merit system protections, and the Office of Special Counsel to deal with whistleblower cases. Established the Federal Labor Relations Authority, which took over labor relations issues that the Civil Service Commission previously had managed. Established merit system principles as well as prohibited personnel policies, which defined the basic principles that all federal personnel practices would have to meet. Created the Senior Executive Service, which took in what had been GS-16, GS-17, and GS-18 positions. Repealed the previous performance rating system and replaced it with mandatory performance management systems, created by each agency, which in turn would be used for employee rewards and decisions to demote or remove poor performing employees. Intended to make it easier to remove poor performers.

Trump administration plan (2025). The civil service is full of “deep state” denizens, devoted to keeping their jobs and undermining Trump’s policies. The most important thing is to remove as many civil service employees as possible, both by firing probationary employees and by firing or laying off very large numbers of civil service members. The administration created a new Schedule Policy/Career (or Policy P/C, for short) for 50,000 or more federal employees with a role related to policymaking or policy advice, who could be fired at will. Replacements, if any, would be hired through the merit system. Thus, the merit system would be used for firing, but its protections would not apply to dismissals. There is a proposal that, to keep civil service protections, new employees would have to pay 50% more of their income for retirement benefits. In that plan, any employee wishing to challenge an adverse personnel action would have to pay a $350 fee to file their case.

Reformers’ dream package (2035). Some reform-minded people would provide veterans’ preference but use it only for the initial hiring of new employees, which would be made without an tool for political favoritism. They would extend the probationary period for new employees, protect existing employees from at-will firings but make it much easier to dismiss poor performers, improve the performance management of federal managers by linking their results with their salaries, but prevent firing for failure to be in sync with a current administration’s policies. Most of all, they would orient the system much for the federal government’s talent needs for the future by forcing federal agencies to match their missions with the skills that they need and by focusing the federal government’s screening process on recruiting and finding those new employees.

And, to help you set your time machine lever, here’s a chart of how the percentage of federal employees covered by the federal civil service has changed over time.

Source: Donald F. Kettle. Illustration by OpenAI

Some friends of mine have said they would set it at 1911, just before the passage of the Lloyd-LaFollette Act, at a time when merit determined the hiring of federal employees but when presidents had much more power to dismiss employees. Others have said 1945, before the GS classification system emerged. One experienced expert told me it would be 1977, before the 1978 reforms that, in the expert’s opinion, introduced the pathologies that make human resources managers so unhappy today. And, of course, some of my conservative friends would choose the current Trump action.

Where would you set your time machine?

]]>

The Trump hiring plan wants to fix federal jobs, but it might just make things worse

Donald F. Kettl — Fri, 13 Jun 2025 06:00:00 -0400

If there’s anything that everyone anywhere close to the federal hiring process agrees about, it’s that the system is a mess, for applicants and agency officials alike. But the hiring plan the Trump administration launched during Memorial Day week would only make things worse—for everyone.

Its goal is to make merit matter more, by “reforming the Federal recruitment process to ensure that only the most talented, capable and patriotic Americans are hired to the Federal service.” That’s certainly a goal that everyone can get behind.

So, too, is ending self-assessments by applicants, which too often turned out to be exercises in wordy puffery, and making greater use of validated tools to assess applicants. Those are big steps forward.

But the hiring plan also misfires. First, it screens applicants for ideology more than merit. Applicants must write four essays that, among other things, requires them to explain how they’d advance the president’s priorities. Now, we expect every civil servant to use their expertise to follow the president’s directives, within the scope of the law. However, the law also sets other standards as well, including recruiting individuals “from all segments of society” and protecting them against “coercion from partisan political purposes.” The law defines 9 merit system principles, including recruiting individuals “from all segments of society” and protection against “coercion from partisan political purposes.” Creating a political standard flies directly in the face of the nonpartisan civil service that Congress created in 1883, and that has enjoyed support from presidents of both parties ever since.

Moreover, the goal of the civil service is to create an expert workforce that will endure for the long haul, not just one administration. The essays make no sense in building that workforce. And, as I know from grading college exams for 45 years, these essays would be ridiculously easy to game. Even my favorite AI app figured how to dish out the right word salad in about two minutes.

In fact, the average federal employee has a dozen years of service. We expect every federal employee will follow every legal instruction they receive, and we want them to do so as expertly as possible. That’s been the goal of the civil service system since Congress created it in 1883. Moreover—and I say this as someone who has graded college essays for 45 years—the four essays would be ridiculously easy to game. Even my favorite AI app figured that out in about two minutes.

The hiring plan also sets some very useful goals, like reducing the time to hire new employees, but it doesn’t lay out a game plan for getting there. On average, it takes 98 days to hire a new federal employee, more than twice as long as in the private sector. That’s a huge problem for government agencies and prospective employees alike, and everyone needs a detailed work plan.

In short, the administration’s hiring plan identifies many of the right problems, but it also misses several important ones, including finding the best recruits, defining the skills the government most needs for which positions, and producing a journey map, for supervisors and candidates alike, to move briskly from a vacancy to a successful new hire. That’s the biggest missing link in the administration’s hiring plan.

We Need a Third Way

Of course, no recent administration has dealt with this problem well. Nor has either end of the ideological spectrum.

The right wants to shrink government. Since it’s hard to do that—just ask Elon Musk—it seeks to cut off government’s brain and its arms and legs. There’s no reason to focus on hiring the best and the brightest when the real goal is to slash government.

The left wants to build a bigger, more robust government, but when it comes to managing the system it has focused far more on the how than on the what. Even more, the left has tended to circle the wagons around the status quo to protect the jobs of federal employees without stopping to think about what those employees do and whether they could do it better.

The right wants to cut many programs that people like and want. The left wants to create new programs without figuring out how best to make them work. It’s little wonder that government so often performs so badly. A 2013 survey found that more Americans approved of Genghis Khan than Congress.

That’s why we need a third way to improve federal hiring, a way that avoids the pathologies of the right and the left and focuses, instead, on building government’s capacity to get done what the people want to get done.

This third-way approach would put mission at the very center of government. It would build government’s means to get that mission done. And it would recreate the merit system, which has enjoyed strong bipartisan support for 140 years, to match the challenges of the twenty-first century.

Toward Effective and Responsive Hiring

Here’s what we need to build this mission-means-merit approach.

Emphasize the central importance of a nonpolitical workforce. The civil service, from its very start, has had the goal of creating a system based on merit and insulated from political pressure. The primary allegiance of federal civil servants is to the Constitution and to the body of law they are charged with implementing.
Make mission drive hiring. Federal employees play a critically important role in governance: they create the capacity to translate policy into results. However, since the 1950s, government’s culture has increasingly focused more on complying with rules instead of accomplishing the mission. Mission must be at the center of the hiring process—and personnel managers must be part of the agency management team, just as budget analysts are.
Recruit in the mode that works best for recruits. Posted job listings are as useful today as the floppy disk or computer punch card, both of which are relics now in the Smithsonian Institution. In contrast, the District of Columbia has created a wall-to-wall, job- announcement-to-onboarding process that syncs with the online instincts of job applicants—and which applicants can easily use on their smartphones. That’s an important step toward making federal employment attractive to applicants—as long as top officials do not seek to traumatize the very people they want to recruit and retain.
Qualify potential employees for broad pools, not individual jobs. The federal government, for too long, conducted its hiring one slot at a time. Recent federal hiring initiatives have empowered subject-matter experts across the government to establish the competencies required for broad categories of jobs, to pre-qualify applicants for jobs across the government, and bring candidates on board far more quickly. Finding the best match between the skills required for a position and the skills of employees is the best way to conduct hiring, as the administration’s plan suggests.
Hire for the ability to work in teams. No federal employee works in an organizational chart box. All work that matters happen in teams. The hiring process must assess a candidate’s ability to contribute to a collaborative enterprise. Ranked scoring won’t get us there.
Create a two-year probationary period. Federal supervisors need the flexibility to determine whether new employees can create the results at the core of agency missions. A two-year probationary period, which had been the standard in many agencies until recently, would provide that opportunity.
Establish a streamlined process for removing poor performers—but ensure that they are removed for legitimate reasons. There are undoubtedly poor performers in government, although the number certainly is not as high as white-hot political rhetoric often suggests. Poor performers ought to be removed briskly, but “poor performance” should not be used as an excuse for dismissing employees who refuse to carry out an illegal or unconstitutional directive or because they blew the whistle on these activities.
Ensure that veterans continue to receive consideration for employment in a manner that’s consistent with a merit-based, mission-driven hiring system. Former members of the military, who have sacrificed much in service to the country, unquestionably deserve preference in federal hiring. The current veterans’ preference system, however, effectively squeezes out many other exceptionally well-qualified applicants, hamstrings agencies in the hunt for talent, and often disserves veterans, who too often depart government sooner than other new employees. Veterans ought to be given preference in applying for their first job. They should receive a preference to get them into the pool of the top five applicants from which agency managers make their hiring decision. From that point forward, veterans should work under the same standards as all other federal employees.
Streamline the security clearance process. The administration’s hiring plan recognizes that the process of obtaining a security clearance often creates a bottleneck in hiring the best employees. In recent years, reforms have greatly reduced that bottleneck. Continued progress in reducing the security clearance backlog is essential in reducing the federal government’s time-to-hire.
Less firing requires better hiring. There is substantial debate about how to make it easier to fire poor performers in government. That, however, is a too-late issue. Firing problems are the result of hiring problems—and inadequate management of employees by their supervisors. This is a problem in all organizations, public and private. The federal government needs to step up its own efforts to solve it.
Merit must be at the center. Throughout all of these steps, the federal government must increase, not downgrade, its emphasis on merit. We cannot solve the problems of the workforce of the future by trying to move back to a time in the past when the federal personnel system seemed rosier. The current system needs fundamental reform. But so has the personnel system at every point in the nation’s history.

Accountability and Trust in Federal Hiring

The bureaucracy doesn’t exist to protect employees’ jobs or to serve as a proxy in the war against big government. Rather, its fundamental purpose is to do the people’s work, as defined in law.

We need a new, third way for leaders to effectively lead and for experts to expertly manage. We need a fresh and innovative partnership between them, with a workforce highly tuned to the collaboration—between employees, between agencies, between the public and private sectors, and even between nations—that is the emerging environment of twenty-first century government.

And we need a system that does so by nurturing trust. As the noted author Philip Howard told me, “The benefit of accountability is not mainly to purge poor performers (although important) but to instill mutual trust that everyone is pulling their share.” Neither the traditional approaches of the right nor the left will produce that trust.

Instead, we need a third way, built around the principles laid out here, with mission, means, and merit at the core.

]]>

Trump’s push for executive order loyalty risks undermining the federal workforce and the Constitution

Donald F. Kettl — Tue, 10 Jun 2025 07:00:00 -0400

Donald Trump launched his presidency with a blizzard of executive orders unlike anything in recent memory—more than three times as many as in Joe Biden’s first 100 days, seven times as many as Barack Obama, and 13 times as many as George W. Bush. And he expects federal workers to be accountable to him.

To underline the point, his Office of Personnel Management has put out a new hiring plan for federal workers and a training plan for the Senior Executive Service. Applicants for a job must demonstrate how they would “help advance the President’s Executive Orders and policy priorities.” The training plan for new SESers begins with sessions on “President Trump’s Executive Orders.”

The administration is convinced that “many career federal employees use their positions to advance their personal political or policy preferences instead of implementing the elected President's agenda,” as the proposed rule creating the new Schedule Policy/Career (formerly known as Schedule F) puts it. To the degree that happens—rarely, but certainly occasionally—it is wrong and unethical.

At the same time, however, feds need to be accountable to more than the president’s executive orders. EOs, in fact, are at the bottom of the federal legal chain, not the top.

Source: Donald F. Kettl

The Constitution is the basic law of the land. Congress passes laws to define policy and the delegation of power to execute them becomes framed in regulations, through the Administrative Procedure Act of 1946. Presidents can—and do—emphasize their own priorities through executive orders, and they always decide how much emphasis to put on which programs and regulations. If they don’t like earlier EOs, they can issue their own orders to wipe them out, and they can instruct agencies to change regulations that aren’t in sync with the administration’s priorities.

But the bottom line is that federal officials are accountable to the EOs, but also to much more, including the professional standards of their professions. Anyone who has listened to more than a few seconds of the air traffic control radio knows how much we need the controllers and how it’s not a job for amateurs.

So when Team Trump complains that federal employees don’t always follow presidential directives, it’s often because the employees are in the crosscurrents of the balance of powers. That’s a feature, not a bug of the constitutional system that the founders created. It’s a bug, not a feature, that annoys team Trump.

We can’t squash that bug without undermining our constitutional system.

And we need to remember, as well, that the public values the quality of government much more than who it’s loyal to. A poll conducted this time last year found that 95 percent of those surveyed agreed with the statement, “civil servants should be hired and promoted based on their merit, rather than their political beliefs.”

Before the creation of the politically nonpartisan merit system with the Pendleton Act of 1883, the government just wasn’t very good at doing what the people wanted done. Postal workers found an easy way to deal with the daily flow of mail—they just locked it up in a back closet. British and Prussian customshouses were four to five more times more efficient than those in the U.S. The Europeans relied on professionals; in America, these were plum political jobs.

The more the administration focuses on hiring and training employees who support executive orders, the more the president risks programmatic and, therefore, political suicide. That certainly doesn’t mean that Team Trump ought to turn the government over to the “deep state” or that feds get a free shot at kneecapping the president’s agenda.

But focusing on short-term responsiveness to executive orders would serve the president badly and the country’s long-term interests even worse. We shouldn’t have to painfully learn the lessons of the Pendleton Act all over again.

]]>

The DOGE-acolypse

Donald F. Kettl — Thu, 13 Mar 2025 08:00:00 -0400

As we re-enter the theater, the curtain on Act 1 of The DOGE-acolypse has just fallen, following a shouting match between cabinet secretaries and the head of DOGEworld, Elon Musk. It was a surprising ending to an act of our play that began, just seven weeks ago, with a fevered debate about whether President Trump would be able to win Senate confirmation of cabinet nominees like Pete Hegseth and Robert F. Kennedy Jr. as well as FBI Director Kash Patel.

Would the president possibly be able to get ALL of these controversial nominees confirmed? But as this drama played out on center stage, a surprise character entered from stage far right: Musk, leading a DOGE-acolypse. At first, the debate was what this DOGE thing was (a department? an advisory committee? a plaything of a billionaire that happened over a chance Mar-a-Lago dinner?).

But soon memos started flying, with Musk launching a three-pronged war: take over the government’s personnel system by placing loyalists inside the Office of Personnel Management; take over the government’s information systems, by pushing his team members to get control of the data that controls the government’s payments; and take over the government’s facilities, by strategizing over buildings to close and sell.

There was the surprisingly important memo on Jan. 20 that broadened, beyond anyone’s wildest dreams, the government’s power to dismiss probationary employees and put other feds on “paid leave.” That snowballed into “deferred resignations,” the creation of a deal that had never previously existed. The cabinet secretaries found that, after struggling for every last vote in the Senate, they weren’t really running their departments.

The rise of the DOGE-acolypse, however, made the headline-hungry president nervous and the cabinet furious. That culminated in the meltdown on March 6, with the curtain falling on Act 1.

That had the audience at the edge of their seats, waiting for Act 2. Who was really going to run the government? Musk and his Muskovites, maybe 150 in number, who had fanned out across the government and, in some cases, wheedled their way into departmental appointments, so they insinuated themselves in the chain of command? Or the cabinet secretaries, who thought they had come to Washington with the president’s blessing to move MAGA? Office of Management and Budget Director Russell Vought who moved into the job he coveted most, as part of the Project 2025 empire, which itself had been generated with hundreds of billions of dollars (maybe a billion dollars?) since the 2020 election—or Musk’s bulldozer, which wanted to break as much china as possible until there was no china left to break?

From the wings of our play in Act 2 comes a player that had not even appeared in Act 1: Congress, which faced the challenge of preventing the government from closing down—and, along the way, finding enough spending cuts to finance the tax cut over which they had been drooling for many months. The cabinet secretaries, Muskovites, Vought, and Republican congressional leaders all come to center stage in the middle of Act 2.

Toward the end of Act 2, however, yet another set of characters comes onto the stage: John Roberts and the Supremes, pressed to sort out delicate balance-of-power questions in a political environment as super-charged as any the Supreme Court had ever faced. The curtain falls.

That set the stage for Act 3 of our play. As the curtain rises, Chief Justice John Roberts wakes up in the middle of the night, covered in sweat, trying to figure out how to deal with a constitutional crisis he can’t escape and the fundamental legitimacy of the court, which hangs in the balance. The politicization of the Court was the nightmare he has always feared most. Cases are cascading onto his doorstep. Can the feds slash indirect cost spending in federal research grants? Can federal employees be “fired” under the pretense of “paid leave?” Was the blanket dismissal of probationary employees under the guise of poor performance, even when some “poor performance” letters went to an employee who hadn’t even started work yet? Can the president dismantle an entire agency created by Congress, like USAID? Can the president impound money appropriated by Congress?

The Supremes meet behind closed doors. Which cases should they take? When should they take them? On what grounds should they decide—small bites or the Big Bang?

As the Supremes debate these questions—and more—the spotlight goes to Supreme Court Justice Amy Coney Barrett at the corner of the table. She has the most stirring soliloquy of the entire play, weighing what “original intent” really means against the political chorus of external pressure she’s feeling.

She and Roberts try to figure out a way through. They know that what they do not only shapes the political legitimacy of the Supreme Court. It ultimately defines what the separation of powers really means as the country nears the 240th anniversary of the Constitution.

The curtain falls on Act 3, leaving everyone puzzled over how the issues will play out.

]]>

The DOGE risk to fighting waste

Donald F. Kettl — Wed, 26 Feb 2025 12:17:32 -0500

There are at least two things that everyone around the country agrees on: there’s waste, fraud, and abuse in the federal government—and that we ought to do everything we can to root it out.

And in rooting it out, there are two more things we can agree on. The war on unholy “waste, fraud, abuse” trio has been around for a very long time. Ronald Reagan, after all, launched his own proto-DOGE in 1982 with the goal of “improving management and reducing costs.” Second, there’s always been a mysticism around this trinity, which assumes it could be made to disappear like a lion in a Las Vegas show, by waiving a magic wand.

Ahh, if only it were that easy.

In fact, as the Government Accountability Office makes clear in its new high-risk report on federal programs most prone to fraud, waste, abuse, and mismanagement, we certainly could save billions of dollars. Past work on GAO’s checklist has saved $759 billion, with $84 billion recouped in just the last two years. There are billions more we could take back.

That’s what the Trump administration wants. It’s the job that DOGE has. It’s the key to reassuring taxpayers that their hard-earned tax money isn’t going to waste. But what DOGE is doing is precisely the wrong way to get there.

GAO’s new report lays out the facts in the 323 pages of its new report. Cutting waste, fraud, abuse, and mismanagement requires three “C’s”—coordination (mentioned 33 times in the report), capacity (at 68 times), and capital (human capital, that is, with 27 mentions).

Look at coordination. GAO added federal disaster assistance in this edition of the biennial report it’s been producing since 1990. In the last year, the federal government spent more than $1 billion in helping communities recover from the economic damage they suffered. But it’s tough work for the lead agency, FEMA, with its people pulled in so many directions at once, from the Franklin Fire in California to wildfires in North Dakota to Tropical Storm Ernesto in Puerto Rico. Vastly complicating FEMA’s job is the huge number of federal agencies with a hand in disaster recovery—30 in all.

Project 2025 has an answer to the problem: sending “the majority of preparedness and response costs to states,” as well as “eliminating most of DHS’s [Department of Homeland Security, which houses FEMA] grant programs.” No state, however, has pockets deep enough to fund the response when disasters strike, and no state has enough surge capacity to get the help that’s needed to the place it’s needed at the time when it’s needed. Even California called for outside aid in the midst of January’s wildfires.

So, if we’re to help people in their worst moments and save taxpayers’ money, we need to do a far better job of connecting the big federal players—FEMA, HUD, and the Transportation Department—to prevent costly duplication, on one hand, and cracks in the system through which people can slide, on the other. Then the federal system needs to connect far better with state and local agencies. If there’s anything we’ve learned about disaster response, from the recent crash of the American Airlines jet into the Potomac to Hurricane Helene in North Carolina, those connections depend on building, in advance, the relationships of trust that shape an effective response.

In short, cracking this problem requires better coordination.

Then look at capacity. Cost cutters salivate at getting their hands on improper payments in Medicare (which GAO estimates at $54.3 billion in 2024) and Medicaid (at least $31 billion). Some of this $85 billion is legitimate but isn’t backed up because of sloppy recordkeeping. Some of it is undoubtedly fraud. No one knows for sure how much fraud there is, especially because private contractors do the front-line work in managing both programs. And, of course, no one knows how much these estimates might miss (and the people committing the fraud aren’t going to tell anyone).

Running these programs, which amount to about one-fourth of all federal spending, is a tiny bureaucracy, the Centers for Medicare and Medicaid Services. It has about 6,000 employees, or 0.3% of all federal employees. For a sense of scale, compare that with the University of Montana. It has about the same number of employees, and they’re in charge of 10,800 students. Medicare enrolls 66 million Americans; for Medicaid, it’s 72 million.

What capacity does the federal government need to recoup tens of billions of fraud in programs covering these tens of billions of Americans? In Medicaid, slashing away at fraud requires strong agency leadership; CMS collaboration with states and contractors to root out problems in managed care programs, which too often feed the fraud; and better monitoring, especially to rake back money that shouldn’t have been paid out.

As a new report from the IBM Center for the Business of Government shows, this means beefing up IT systems, improving AI, and investing in government employees who know how to use these tools. It requires better capacity.

Finally, let’s look at capital—human capital, that is. GAO put 37 areas on its high-risk list. Of these, 20 struggle with a big gap between the skills needed to make the programs work better and the skills government agencies have. This problem has been on the high-risk list since 2001, and there’s no escaping the other problems on the list without first solving this one.

The list of areas where the government’s talents fall short is a long one. It doesn’t have the AI and IT experts it needs. It needs more employees with strong skills in project management, organizational performance, leadership development, and data analytics. But not only has the government’s human resourcesagency, the Office of Personnel Management, “not yet developed an action plan,” GAO explains. It has failed even to figure out which skills it most needs, where it needs them, or how to get them.

And now, in the last month, the wholesale assault on government workers has made things worse. This strategy was crafted to drive away existing workers and to discourage new employees from joining the government. For anyone trying to slash waste, fraud, and abuse, however, this is remarkably counterproductive. The entire federal payroll in 2024, counting salaries and benefits, was $337 billion. Cutting, say, one-fourth of that would save $84 billion. But it would put at grave risk the federal government’s ability to recoup the tens of billions of dollars that GAO has demonstrated that the government has saved in recent years.

Cuts of that size would even further undermine the federal government’s human capital, which already has big holes. It would weaken government’s capacity to roll out the big new AI and IT systems it needs to go after waste. And it would erode the coordination—with each other, with state and local government officials, and with private contractors—that the government’s employees have steadily developed over the years.

GAO has documented how we can make big improvements in government’s services while increasing service to taxpayers. The new report also has a game plan for reducing fraud, waste, and abuse through strategies that would far more than pay for themselves.

That’s why DOGE is the highest-risk area of all, because it puts in danger the progress that the government has made in reducing fraud, waste, abuse, and mismanagement—and it opens the door wide to even more.

]]>

What to do on the day after?

Donald F. Kettl — Fri, 07 Feb 2025 12:10:25 -0500

At some point, the Trump administration will have fired all the employees it wants to fire, cancel all the programs it wants to cancel, and cut all the programs it wants to cut. What happens on the day after?

The Trump administration is right in fingering many federal government problems. That’s a point that many of my friends on the left reject, but they need to accept the reality. However, the administration’s opening salvos aimed at reducing government employment, cutting government spending, and slashing government programs won’t solve the real problems of the U.S. government. That’s a point that many of my friends on the right will reject.

On the day after, two facts will be inescapable. One is that the people will expect that their government will work—maybe even better than before. They will want to feel safe about flying on planes without a door plug blowing out, buying eggs in the supermarket without worrying about bird flu, or working among cancer-causing chemicals at work. Cutting government is one thing. Delivering on its promise is quite another.

The second is that we’ll wake up with a government that can’t do the job we want done. Why am I so sure? Because that’s been the case for years. Seven years ago, in fact, two National Academy of Public Administration panels I chaired issued reports entitled No Time to Wait. We concluded that “the federal government faces profound problems in making government work for the American people in large part because its human capital system is fundamentally broken.” There’s even less time to wait now.

So how, on the day after, can we pivot from making cuts to building capacity? Even those who don’t want to admit it publicly know that they need government to work for the things they care about. It’s one thing to strip our 1950s government down to the chassis. It will be another to build the vehicle we need to navigate the high speed and sharp curves of 21st century roads.

Consider eliminating waste, fraud, and abuse. To get a snapshot of the problem, we don’t need to look any farther than the federal government’s improper payments. The Government Accountability Office estimates that fraud sucks out between $231 and $521 billion every year in payments that shouldn’t have been made for Medicare, Medicaid, grantees, and contractors. That’s more than we spend on salaries and benefits for federal employees.

So let’s start with Medicare and Medicaid along with unemployment insurance. Some “improper payments” aren’t fraud at all but are caused by sloppy paperwork. Sometimes, clever operators have devised schemes to defraud the government. Elon Musk could put his army of techbros to work with their AI systems to sort one from the other and then identify the patterns that the fraudsters are using, like phantom billing (when a provider bills the programs for tests or services or equipment a patient never got) and upcoding (billing for a more expensive service than a patient received), along with blatant kickbacks and prescription drug fraud. Not chump change.

Then there’s the problem of hiring new government workers. The time to hire a new federal employee is about 119 days, or more than three times as long as in the private sector. Musk’s experts could help us break the hiring deadlock.

One of Trump’s executive orders gives a senior presidential aide until the end of April to come up with a plan to fix hiring (or, at least, cut the time-to-hire to 80 days) and bring better performance management to the federal government. But the federal government needs to learn the real lessons from the private sector. The premier private sector organization in the field, the Society for Human Resources Management, concludes that performance “reviews generally do not work.” The federal government needs to be cautious about insisting on ideas that smart private-sector managers are abandoning. That’s a project to which Musk’s techbros ought to put their shoulders to the wheel to solve.

Then there’s the talent-mismatch problem. Getting rid of thousands (or tens of thousands or hundreds of thousands) of employees is certain to produce one effect: the talent the government has won’t match the talent it needs. We can be sure of that—it’s the lesson from the buyouts during the Clinton/Gore reinventing government initiative: buyouts never leave you with the people you need. But we had a talent mismatch before Jan. 20 as well. So now we have a golden chance to fix federal hiring and its performance management system to get—and keep—the talent we need.

Finally, an agency-centered slash-and-burn strategy misses the most important opportunity for results. There’s one, simple, inescapable fact: there’s no problem that matters more than any single agency can control. Big problems that force themselves onto the agenda can only be solved by building bridges across federal agencies, across levels of government, across the sectors of society, and even across international boundaries. The fatal flaw in tackling government reform agency-by-agency is that it misses this critical point. We need instead to create a new generation of bridgebuilders.

Like the fires in LA, we have a big problem. There is the short-term focus that’s in front of us, but there is the long-term—and inevitably transformative—question about rebuilding. We have the opportunity to be the architects of the government that must rise from the ashes—and that must rise if we are to sustain the promise of American democracy.

]]>

An open letter to the public administration and public management communities

Donald F. Kettl — Wed, 11 Dec 2024 08:00:00 -0500

For those interested in the doing and the studying of government administration:

It’s time for a fundamental reexamination of the values that drive us, our work, and the governance of American political institutions.

This isn’t just about the restoration of Donald Trump to the White House and the tsunami of disruptive promises he’s making. This is a challenge that goes far deeper, not just in the U.S. and around the world.

Trust in America’s political institutions is at an historic low. In a 2023 Pew poll, 85% of those surveyed said that elected officials don’t care about what “people like them think.” When asked how they feel about politics, 65% said they were “exhausted” and 55% said they were “angry.” Less than one-third of Americans believe that “the federal government has a positive impact on people like you.” Support for existing institutions is crumbling under criticisms from the far right in Germany and Japan and France. Public administration is the action arm of the state and of the political decisions it makes; distrust of public institutions gravely weakens support for public servants as well.

For more than a century, the basic foundation of the field has been a commitment to ensuring competence in executing the public policies created by elected officials. The endless debate about the “politics-administration dichotomy” has muddied discussion about the core, but we have spent 140 years constructing a nonpartisan civil service. We have invested 90 years seeking to create the most efficient administrative structures. Since the 1970s, we have sought to improve the efficiency of government decisions through policy analysis and the implementation of policy through effective public management.

But we cannot assume that these foundations continue to serve us well. All of these traditions struggle to connect with the fundamental issues on the policy agenda today. Both elected officials and the public can scarcely hide their disdain for experts.

Moreover, we’re in a weak position to defend the traditional structural arrangements that are now under attack, because we have taken them for granted, conducting relatively little research on the basic structures and processes of government. For example, we don’t have a lot of research informing the public debate about the impact of removing protections from the civil service—and why this will matter.

It’s surely the case that not all experts will agree with these propositions. And it’s also the case that many experts see their most important role is fighting against the anti-expert propositions rolling out from populists. These battles are uncomfortable but legitimate.

In assessing the state of public administration and public management, however, we have to reckon with several fundamental propositions. Many people and elected officials believe that the experts who manage government are unaccountable—just 42% in a poll for the Partnership for Public Service think that government is “accountable.” More than half—56%—believe that government is “almost always wasteful and inefficient.” A majority of the public (58%) believes that it is too hard to rid the government of poor performers. Government employees are increasingly worried that the wrong political views will get them purged. Ten careerists in the Homeland Security Department were listed on a “watchlist” of “America’s most subversive immigration bureaucrats.” Individual employees are finding themselves called out by name in an effort to “putthem in trauma.”

In short, the fundamental arrangements about the use of power—especially about the power of experts—arrangements that have carefully developed over more than a century, are quickly disintegrating. Members of the public administration community might well want to fight a rear-guard action against this onslaught.

It’s folly to dismiss the critique as false—or that it’s temporary—or that it’s purely American. This is a deep-seated, global phenomenon with which the field has a fundamental responsibility to engage, at its most basic level.

We must reexamine the fundamental values that drive our work. We need:

A fundamental debate about the field and its goals, much like what surrounded its first years in the early 1900s, the post-World War II years, and the reexamination of the 1970s.

A focus on how best to deliver results to Americans—not outputs, outcomes, or other tech-speak metrics, but results in terms that Americans see and experience and appreciate in their lived experiences.

A reinforcement of a process that ensures freedom from political bias, since a fundamental form of distrust is that some people, from some backgrounds or regions, get different (and better) treatment.

A commitment to efficient delivery of services, so that people feel that their hard work to contribute taxes are respected.

A redefinition of accountability, to reinforce the connection of the people to the government they receive through the decisions of the public officials they elect.

These are just starters. There surely are others; some of the points I laid out might be rejected. But we need a vigorous debate about the basic principles of the field.

We also need to acknowledge that public administration can’t fix the fundamental problems with our major institutions, especially with Congress. But, within the fabric of the American constitutional system, we can recognize the role that we play. And we can focus especially on the fact that people differentiate between their anger at the system as a whole and their appreciation for services with which they engage—and which serve them well.

That, indeed, is the bedrock of public administration.

Sincerely,

Donald F. Kettl, Professor Emeritus and Former Dean, University of Maryland School of Public Policy

]]>

A Management Agenda for 2025 and beyond: Pivoting from outcomes to results

Donald F. Kettl — Fri, 06 Sep 2024 06:00:00 -0400

The new president will have a mighty big chore in the spring of 2025: devising a new management agenda.

The people just don’t think government works for them. Harris and Trump supporters agreed on one thing in a recent poll: exactly 52% of both groups said the country had big problems—and that they were unfixable.

Moreover, for the last 30 years, presidents have tried to improve the way the federal government works. They’ve reinvented government, set up traffic lights to measure success, bunched programs into categories to improve efficiency, and worked across silos.

Despite all this work, however, we haven’t boosted the public’s trust in government to do what is right. Back in 1994, that number stood at 19%. At the end of 2023, it was . . . 19%.

There’s been genuine progress under the hood of government in each of these reforms. But when it comes to creating a government that solves the country’s big problems, the people just don’t believe it’s happening.

We need something new. We need to pivot from outcomes to results.

What’s Wrong?

A big problem, of course, is that presidential management reforms are relatively small fish swimming among sharks in the depths of distrust.

Of course, the trust problem is much bigger than the work of federal administrators. In a recent Gallup survey, 60% had “very little” or “no” trust in Congress. The presidency didn’t fare much better, at 50%. Trust also depends on economic performance, and reformers don’t have any control over that.

The news, however, isn’t bad if we listen to what the people want. Trust improves when people feel they have a say in what government does; when it’s effective, competent, and accountable; when it keeps them safe and secure; when it supports struggling Americans and their families; and when government treats them with respect and appreciation.

Most of all, government builds trust when it delivers meaningful, measurable results.

That last point is where the gulf between many management agendas of the past and what people want is widest. The reforms have tended to focus on improving government outcomes and have scored their success against strategic goals. Heaven knows that’s important, but that’s an inside baseball game that few Americans see or care about.

Measures don’t communicate. Stories do. Strategic plans don’t matter. Results do. It’s impossible to convince anyone of government’s success if it doesn’t match their lived experience. That goes for scorecards, traffic lights, and websites. These tools are unquestionably valuable, but as a language to communicate within government, not to connect with the people.

We need to build government from the inside out so that it works from the outside in.

From Outcomes to Results

Not long ago, I was beginning a series of painless but lengthy medical treatments. “I hope you have a good outcome,” Frank Luntz, master pollster and communications consultant told me. But then he corrected himself. “I hope you have a good result,” he chuckled.

What’s the difference? Healthcare experts talk about outcomes in terms of the overall impact of treatments on health and the quality of life. Results go to the immediate impact of the treatment. Both matter, of course. But patients look to results—are they feeling better?—and then to outcomes—in the longer term, what’s the quality of life? The latter doesn’t matter much if the former isn’t an improvement.

For the public, wonky debates about outputs and outcomes and performance measures don’t resonate. They want to know that government, at the core, connects with them, treats them with respect, spends their money well, and provides the safety and security they need. They want results.

This isn’t an either/or in comparison with the 30-year string of reform efforts. It’s a both/and: making sure that the connections between the people and government work well.

What Government Can Do

That leads to some specific action steps for the 2025 president’s management agenda.

Make fewer mistakes. One bad experience can swamp all the effort spent in building good will. McKinsey found that people who were dissatisfied with their experience in dealing with an organization were twice as likely as satisfied individuals to complain to public representatives. Failures are the fuel of distrust. When people are dissatisfied, they’re twice as likely to publicly complain. When they’re satisfied, they’re 9 times more likely to trust the agency providing the service, even if they don’t trust government in general.

In 2024, the IRS experimented with Direct File, which allowed taxpayers in some states to submit their tax returns online. Among those who tried it, 90% rated their experience “excellent” or “above average,” and 86% of users said that the program increased their trust in the IRS.

Bounce back from failures. Nobody gets it right all the time, but resilience—learning from failures and making sure they don’t recur—reinforces trust.

FEMA’s long journey since the disaster of its response to the 2005 disaster of Hurricane Katrina shows the power of resilience. Because of strong leadership sustained across Republican and Democratic administrations, there’s been a consistent pattern in natural disasters: the dog that didn’t bark. There have been few tales of an unresponsive, inefficient FEMA. Building trust is now a central part of the FEMA training curriculum.

Make it simple and reliable. We’d expect that people want their experience with government to be fast, and that’s true. But they value simplicity and reliability even more. In fact, speed ranks only third or fourth in the public’s priorities. They want to know what they will get and when they’ll get it.

After the pandemic, there was an explosion of travel, and the processing time for passports slipped. Along the way, however, the State Department posted the estimated time. It’s now returned to pre-COVID levels, and travelers can find the current estimated time for routine and expedited processing times on the department’s website.

Share the tale. Performance wonks (with me at the head of the parade) love to wade around in the performance data. Performance.gov has become a first-class resource. Much performance data, however, is prepared by performance wonks for other performance wonks. It doesn’t create the lexicon in which reporters and members of Congress travel. Even worse, there’s a public policy version of the old “if it bleeds, it leads” approach to news coverage: bad news stories drive out the good. The statistical measures are essential for tracking the long-term progress of management initiatives, but stories are essential for communicating that progress to broader audiences.

The VA, for example, launched its VSignals program to collect feedback from vets on their care and wring out bad experiences from its programs. Asking vets what they like and what they don’t allowed the VA to increase trust from 55% in 2016 to 80% in 2024.

“I don’t have customers.” Oh yes, you do! The response by many federal managers to these customer service lessons often is that they don’t have customers, so none of this matters to them. They couldn’t be more wrong. The central purpose of the government is to deliver value to the people, and every single federal employee has a role in making that happen. (If that’s not true for any employee, then they ought to be fired.)

There’s the famous story—if it isn’t true, it should be—about the NASA administrator walking down a hallway and bumping into someone carrying a pail and a mop. “What’s your job?” the administrator asked. “Helping put a man on the moon,” the person replied. Everyone in an agency plays a role in its mission, and everyone needs to know—and perfect—that role.

Tweak the conversations. Federal leaders need to improve their conversations: not only to explain to everyone outside the agency what value they create, but also to ensure that everyone inside the agency knows the role they play in creating that value. The government reform movement has often lost the importance of this internal communication. That needs to a centerpiece of the next presidential management agenda.

As we frame employee engagement, this is what employees need to be engaged in: understanding their role in building a more effective government from the inside out so that it works better from the outside in.

We certainly don’t want to go backwards on the progress we’ve made through previous management agendas. But Americans want to know that they’re appreciated and respected for the taxes they pay, that government is devoted to connecting with them in trustworthy and transparent ways, and that government gives them what they want, need, and deserve.

This is what the pivot from a focus on outcomes to results means.

]]>

The Good Government Agenda for 2025 and beyond

Donald F. Kettl — Mon, 12 Aug 2024 06:00:00 -0400

Every new presidential administration since 1993 has come into office by launching a new management agenda. The new administration in 2025 should not do so. Brand-new research shows that the public wants instead a good government agenda.

This is a critically important distinction. Good management is an inside-looking approach focused on the government’s gears and levers. It’s important, for sure, especially to improve the way those gears and levers work. But more fundamentally we need an outside-looking approach, one that takes the essential steps to improve the public’s trust in government, which has imploded. That’s a Good Government Agenda.

A survey and focus groups conducted by analyst Frank Luntz in June found that distrust in government increased from 46% to 63%, in just the last two years. The people don’t much like government in general and truly despise the Congress. Most people have an unshakable belief that the people who work for government can never seem to get it right.

This is a crisis that demands the focused attention of the new administration. It’s even more important because distrust is highest among the youngest Americans. The biggest threat to our democracy is losing the next generation before it has a chance to lead.

The sources of distrust

Among those surveyed, 42% thought that the government does an inadequate or terrible job of serving the people. Among 16% of the survey, it’s even worse: they believe that the government is simply incapable of doing anything at all. The people tend to believe that the government is wasteful (85%), corrupt (74%), and incompetent (66%)—and all those numbers are worse than they were just two years ago.

Most taxpayers don’t see value in the taxes they pay—61% believe that they pay more in taxes than they get back in government services.

This is why we need a fundamentally new approach. The management agendas that every president since 1993 has launched have produced some good effects, but they obviously haven’t solved the key problem of connecting people with their government. In fact, that’s why, after working tirelessly on government reform for eight years while vice president, Al Gore got virtually no traction from reinventing government when he ran for president.

It’s not hopeless

There are great examples of good government. People hold Social Security, Medicare, and Medicaid in high esteem. Keeping the people safe and secure is at the top of the list of what they most want, followed by supporting struggling Americans and their families. They want a government that is as efficient and effective with their tax dollars as possible. They want a government that serves the people of the nation and delivers commonsense solutions—and real results.

In fact, the research showed an important distinction. People like the idea of strengthening communities, but they rate helping struggling families far more. The people connect with government best when government connects with people.

It’s not that the people don’t want an effective federal government: 80% believed that the federal government is an important part of a strong American democracy. That belief was strong across Republicans, Independents, and Democrats.

They just don’t think that’s what they’re now getting.

The essential principles

To build the government that Americans do want and can trust, the Good Government Agenda needs to have the following elements, the research shows.

There’s a big gap between the people’s hopes for government and what they perceive that they’re getting. The focus of government action needs to be on the people, not on agencies or programs or communities.

The people don’t think government is accountable. The focus of government action needs to be on transparency.

The people don’t feel respected and appreciated. The focus of government action needs to be on the quality of each and every interaction with the people.

The people don’t believe that the government is fair. The focus of government action needs to build on the principle of no excuses and no exceptions.

The people can’t cut through reformspeak to understand what government is doing. The language of “outcomes,” “performance,” and other inside-baseball jargon doesn’t work in connecting with the people. They want to focus on results and they want to see results that they are measurable and meaningful—to them, not to some reporting agency.

The people don’t think that government connects with them. They want a government that is in touch with the people.

Assessing the Left and the Right

So how close are the current proposals from the left and the right in meeting these essential principles?

First, there is no Next Big Thing that could serve as a game plan for a Good Government Agenda. There have been big ideas in previous agendas, including Osborne and Gaebler’s "reinventing government” as well as the Bush Texas reforms that he brought to the White House.

The left has the ongoing work of the Biden management agenda, which focuses on the government workforce, the customer experience, and the business of government. The initiative’s website, Performance.gov, displays sophisticated data tracking, anecdotes, and topline analysis.

The closest that the Biden agenda comes to the big principles is in its commitment to strengthening the customer experience, especially in improving services that directly affect the people, creating better links among key life experiences (like births, marriages, and deaths), and developing shared products. There is substantial data that tracks the work, but the Office of Management and Budget has frankly admitted that the federal government’s customer service lags far behind every major private sector industry. It’s no wonder that the people feel that the federal government does a poor job of serving the people when they see better service everywhere else.

If Vice President Kamala Harris wins the White House, she will have this work to build on, but there’s a big distance between the points in the Biden presidential management agenda and the essential principles that flow from the survey work. In particular, most of the Biden management agenda is under-the-hood improvements in the administration of federal programs. These are very important, of course. But the next step is ensuring that these inward-looking steps provide a stronger outward looking connection to the people.

The right has Project 2025. There has been substantial conflict, of course, over whether—and how much—the strategy in its 920-page game plan is (and should be) at the core of a Donald Trump administration. Much of the news stories amounts to public posturing. Project 2025 is essentially a “greatest hits” album of the ideas generated on the right in recent years. It’s an effort to get the old band back together. Regardless of any protests to the contrary, this is the album from which the songs of a Trump administration will emerge.

Two parts of the Project 2025 most deserve attention. One is that the large volume, and the strategies contained within it, are organized by government department and agency. The table of contents matches the federal government’s organization chart. But this is a very1950s way of thinking about a 2030s-era challenge. There is no problem that matters in the years ahead that could possibly fit inside a single agency, and an agency-based approach would doom Project 2025 to failure before it ever launches.

The attempted assassination of Donald Trump makes that point. The failures that led to Trump’s wounding were a cascade of failures that stretched across a host of federal, state, and local agencies; they are not the product of any single agency, and no change in any one of those agencies would have prevented the attempt on the former president’s life. There is nothing in Project 2025 that would lay out a game plan to fix the ills that plagued the Secret Service before the shooting—and there is nothing in it that would produce the kind of Good Government Agenda that the people clearly want.

The other part of the Project 2025 initiative that bears most on management is the proposal to bring back Schedule F, the policy announced at the end of the Trump administration that would give the president and political leaders the power to move career officials in policy-related positions into a new excepted service, where they could be removed at will. Positions in Schedule F could be filled at the discretion of these leaders, without going through the merit process. The change would make these officials directly accountable to the policy goals of the president.

The American people unquestionably want more accountability, but this is not the kind of accountability they want. They want accountability for results, not for partisanship. Six out of ten of those surveyed in June want an independent, nonpartisan civil service, as opposed to one that does the bidding of the president—and that finding is the same for Republicans, Independents, and Democrats.

When asked whether they thought having a nonpartisan civil service was important for having a strong democracy, 87% agreed—87% of Republicans, 79% of Independents, and 88% of Democrats. Should civil servants be hired and promoted based on their merit rather than on their political beliefs? Overall, 95% agreed. What about whether civil servants should serve the people more than any individual president? That’s 90%. Is the federal government less effective when decisions are driven by politics? There was agreement from 89% of the respondents.

So, in short: several elements of the work from the left would contribute to what a Good Government Agenda would look like, but it has a long way to go; most elements of the work from the right would undermine government’s effectiveness and create political influence on the bureaucracy with which most Americans disagree.

The steps to the Good Government Agenda

So how could we build a Good Government Agenda in the next administration? Several steps would lay the foundation:

Understanding that delivering results that focus on the people is far more important than improving public management practices.

Recognizing that results are far more important than process.

Focusing on the people as the center of government’s work, not an incidental shadow to the partisan wrangling that characterizes so much public policy.

Remembering that the people want accountability, but the accountability they want is transparency.

Reducing the hassle factor—the administrative burdens that Americans must carry in dealing with the government—is a fundamental first step toward these goals.

Recognizing that neither the right nor the left has a clear plan for getting there—but a clear plan is plain to see.

Beginning by asking people whether they want a government in general—and a civil service in particular—that’s based on partisan politics or proven performance. The answer would guide the steps we must take.

Donald F. Kettl is Professor Emeritus and Former Dean of the University of Maryland School of Public Policy. He is the author, with William D. Eggers, of Bridgebuilders: How Government Can Transcend Boundaries to Solve Big Problems (Harvard Business Review Press, 2023).

]]>

The positive impact of policy entrepreneurs in the public service

Donald F. Kettl and Jim Tozzi — Fri, 05 Apr 2024 06:00:00 -0400

The Senior Executive Service, the best and the brightest of the federal government’s career public servants, is full of policy entrepreneurs. But unlike entrepreneurs in the private sector, whose rewards pile up in their bank accounts, the rewards for these SESers and other top federal career officials lie in the satisfaction that comes from their contributions to the public good.

Top officials throughout the federal government do remarkable work, as the “Sammies”—the Partnership for Public Service’s “Service to America” awards—make clear, year after year. There’s Brian Key and Scott Bellamy at NASA, who led the project that successfully nudged an asteroid out of its orbit and proved we had the technology to save the earth from collision with a devastating space rock. At the Government Accountability Office, there’s Melissa Emrey-Arras, who led a crackdown on corrupt federal student loan schemes. Or consider Anne Lord Bailey and Caitlin Rawlins, whose team at the VA created a new generation of equipment and software to help VA medical centers across the country better care for the complex problems with which vets often must deal.

These are precisely the kinds of leaders that the federal government most needs to encourage and nourish.

There are enthusiastic academic articles that call for “energetic actors who engage in collaborative efforts in and around government to promote policy innovations,” as Michael Mintrom, professor of Public Policy at Monash University in Melbourne, Australia, puts it. He argues, “the need is great for such actors to step forward and catalyze change processes.”

But there’s a catch here. On one hand, the policy challenges facing government agencies—and the managers with the responsibility for solving them—has never been greater. The number of problems for which people expect government to step in, from climate change to urban homelessness, is rising. They’re also getting more complex, because any reasonable solution requires building bridges across multiple government agencies and multiple sectors of society.

On the other hand, the rewards for entrepreneurial leadership by top government careerists are low. With declining public trust and rising political attacks, the incentives for pathbreaking work might well be lower than in any previous generation. The attrition rate among members of the SES, a study by the Partnership for Public Service found, has hovered around 9.1% percent, 50% percent higher than for the federal career service as a whole.

There’s even a long history on the question of whether public servants even need to be entrepreneurial. In 1986, former Office of Personnel Management associate director Terry W. Culler argued that “most federal workers need only be competent,” and that society’s real talent ought to go to the private sector, where “wealth is produced rather than consumed.” Culler contended that “good enough” was, well, good enough.

But as companies struggling to deal with the bridge collapse in Baltimore, cancer patients hoping for help from the latest drug, California homeowners plagued by wildfires, and Native American tribes looking to crack rising crime rates would all testify, just “good enough” certainly isn’t good enough. All of them need federal entrepreneurs can help them crack their most pressing problems.

In doing so, however, creating the energy for doing so is invariably tough. In any difficult government decision, there are always winners and losers—and the losers are ready to jump into action to stop policies that threaten to leave them worse off. Nowhere is that more true than in the regulatory arena. One senior EPA official once told us that he wrote all new regulations with the expectation that they’d end up contested in court.

One of us—Jim Tozzi—has detailed the work of policy entrepreneurs from within the bureaucracy to establish and develop the work of the Office of Information and Regulatory Affairs within the Office of Management and Budget. Most accounts of OIRA’s work begin with the Paperwork Reduction Act (1980) and Executive Order 12291 (1981) and were included in the Reagan agenda.

In fact, this program of centralized regulatory review wasn’t a top-down brainchild of the Reagan team. It was, instead, the product of more than 15 years of innovation and testing, beginning in the Johnson administration—a process that grew out of the entrepreneurial thinking of top career officials in agencies from across the federal government.

Over time, this entrepreneurship has grown into a centralized process of reviewing new regulations used by both Republican and Democratic administrations. After tremendous controversy at its launch, it’s also grown into broad acceptance of the principle of applying cost-benefit analysis to new regulations. The common-sense notion—that the federal government shouldn’t impose rules where the benefits fail to exceed their costs—has become hard-wired into the rulemaking process.

This has become all the more salient with the Heritage Foundation’s Project 2025 transition plan, which calls extensively for the expansion of cost-benefit analysis across a host of government agencies. The plan is to use the analysis to ratchet back regulations to produce “the least burdensome rules possible.” At the same time, however, Project 2025 calls for bringing back the Schedule F policy of the Trump administration, which would make it easier to dismiss officials in policy-influencing positions.

What we’ve learned in the last 50 years is that pursuing the latter makes it hard to do the former. Not only did the process of applying benefit-cost analysis to federal rules emerge from policy entrepreneurs among the career service. Assessing the costs of new rules and, especially, finding ways of ensuring the most benefits for the largest number of people requires highly skilled entrepreneurs among government’s career civil servants.

In fact, a global movement to recognize public entrepreneurs is developing, with the help of the Schwab Foundation (which has created awards for social entrepreneurs), the World Economic Forum (which is focusing even more on collaboration), and the Senior Executive Service (which recognizes the role its officials play in entrepreneurial policy in the US).

Policy entrepreneurs will never get their rewards in the same way as private entrepreneurs. But creating global awards for recognizing policy entrepreneurs, not just in American government but in other countries as well, would provide the incentives that matter most for public officials.

Policy entrepreneurs are long on influence but short on affluence. Awards that recognize and strengthen the coin of the realm in which they work would help feed the incentives that matter most.

Jim Tozzi is head of the Center for Regulatory Effectiveness and former deputy administrator of OMB’s Office of Information and Regulatory Affairs.

Don Kettl is professor emeritus and former dean of the University of Maryland School of Public Policy. He is the author of many books including, most recently, Bridgebuilders: How Government Can Transcend Boundaries to Solve Big Problems (with William D. Eggers) and The Little Guide to Writing for Impact (with Katherine Barrett and Richard Greene).

]]>

The MacGuffin of Schedule F

Donald F. Kettl — Mon, 01 Apr 2024 06:00:00 -0400

The unquestioned master of suspense, filmmaker Alfred Hitchcock was especially fond of staging a MacGuffin, an element of the story that helped move the plot along but turned out to be insignificant to solving the case.

Hitchcock fans—and count me at the top of the list—always admired his ability to drag us off the scent of the main issue by the clever ways he set up the plots of his films. All the better to scare us.

And that’s just what’s happening now in the debate over Schedule F, the plan to make it easier to move large numbers of federal workers from the civil service to at-will employment. The plot is simple: increase accountability by making it possible to replace federal workers who don’t effectively implement the president’s policies. Conservative analysts are preparing to relaunch Schedule F on Day 1 of a new Trump presidency.

Schedule F is a MacGuffin. Everyone is worrying about it, but it won’t resolve the big plot problems of the civil service.

The system has three big problems: improving hiring; making it easier to deal with poor performers; and streamlining veterans’ preference. Schedule F won’t solve any of them.

First, the federal government’s hiring quagmire is simply embarrassing—and unacceptable. The average time to hire a new worker in the federal government is about 100 days, and recent efforts to shorten it haven’t made much progress. In the private sector, time-to-hire is about a third of that, at 36 days.

Schedule F would give top officials the flexibility to hire new employees without jumping through a lot of hoops. The track record for hiring new political appointees in past administrations, however, doesn’t give much hope that Schedule F would speed up the hiring process. For example, according to the Center for Presidential Transition, the Trump administration was the slowest out of the gate of any recent presidency in naming its political appointees, and it never caught up. The Biden administration was faster but, six months into the term, still had named only one-sixth of its 1,200 appointments that require Senate confirmation.

Its proponents say Schedule F could be used to make 50,000 feds—or more—into at-will appointments. That would be a staggering increase in the appointments workload for a new administration.

They could speed it up by simply rubber-stamping requests to turn career positions into Schedule F jobs. That, in fact, is just what happened to requests from the Office of Management and Budget at the end of the Trump administration. But jumping from OMB’s 500 employees to 50,000 employees across the federal government would be a staggering administrative challenge—or it would lead to a massive delegation of decisions to federal agencies. Either way, it would be virtually impossible to track what’s happening in real time. Without transparency, Schedule F wouldn’t make good on its promise of more accountability.

Second, although Schedule F promises more accountability by making it easier to fire poor performing feds, it doesn’t have a mechanism for identifying poor performers, except by whether employees follow the political will of their superiors.

For a generation, the constant refrain has been that government ought to be run more like the private sector. Schedule F isn’t how the best-managed private companies deals with employee performance.

Instead, private management consultants found that the best approach is “providing feedback, support and training to employees.” That, in turn, builds a “supportive environment.”

Schedule F turns that approach upside down. Its strategy of firing workers to improve results is more like the t-shirts for sale in Annapolis where I used to live: “The beatings will continue until morale improves.” A LinkedIn post puts the best practice like this: “Firing an employee should be a last resort after all other options for improvement have been exhausted,” it says.

The Government Accountability Office conducted extensive interviews about the implications of accountability-through-firing. Its report pointed to the risks that would flow from rolling out Schedule F again: “effects to recruiting, retaining experienced staff, and risks such as employees in Schedule F positions being subject to removal for partisan political reasons.”

The third challenge is dealing with veterans’ preference. In quiet conversations among federal leaders, there’s a powerful consensus that the current system for giving veterans’ preference in hiring and promotion is broken.

There’s overwhelming consensus on helping those who have nobly served the nation get a fast track into post-service jobs, including in the federal government. But research shows that the current veterans’ preference system for federal employees reduces the diversity of the federal workforce, leads to workers who are older and less educated, and produces employees who don’t advance as far as non-vets. Vets hired through the preference system tend to leave the government at a higher rate than non-vets, which suggests that the preference too often creates a bad match between individuals and their jobs. That, in turn, chokes off opportunities for non-veterans, who might be better qualified for particular jobs. It's also inequitable for the vets themselves, because some vets are eligible for the preference while others aren’t.

Schedule F said that “each agency shall follow the principle of veteran preference as far as administratively feasible.” That, of course, creates a mega-loophole. Just what does “as far as administratively feasible” mean? And who gets to decide? Schedule F not only fails to solve the veterans’ preference problem. It makes it far worse by making it more arbitrary.

When it came to the MacGuffin, Hitchcock might leave his audience dangling at the end of the movie, but he always wrapped up the big pieces of the plot. Schedule F leaves the workforce dangling, enmeshed in the enormous uncertainty about how it would preserve the merit system principles on which the country’s public service has relied for the last 140 years—and for which there has been bipartisan support. It also fails to solve the fundamental problems of the civil service—problems on which just about everyone agrees but which Schedule F doesn’t solve.

Leaving the audience dangling like that is something the master of suspense never would have accepted.

]]>

10 surprising things about the ‘deep state’—starting with the Roman Emperor Caligula

Donald F. Kettl — Tue, 09 Jan 2024 06:00:00 -0500

There’s been a tsunami of debate about the “deep state” and its risks to government. But behind the raucous debate is a collection of surprises that help shine a bright light on what the deep state is really all about.

In a 68-page tour through the issue since ancient times, I poke into the big and enduring themes in Experts in Government. The most important is that no government can last without expertise—but the quest for expertise inevitably brings problems of control. And, on the other hand, seeking control of experts tends to undermine their ability to bring government the expertise it needs.

There are 10 big and surprising things lurking in the shadows of this debate.

The personal loyalty of experts isn’t all it’s cracked up to be—just ask Caligula. The Roman emperor surrounded himself with a handpicked Praetorian guard to keep him safe. In 41 CE, however, elite members of the guard knifed him in a corridor of what was the White House of its day. Counting on the fealty of even the closest experts isn’t always a good bet.
Ancient Chinese emperors discovered they couldn’t rule effectively without nonpartisan experts. Long before Caligula met his end, the emperors established a civil service filled with experts who had passed an incredibly rigorous exam. Researchers have called it the “examination system from hell” which included, among other things, memorizing 400,000 characters of Confucian text. Cheating on the exam was discouraged. Anyone caught looking on the paper of the person sitting nearby was executed.
The Chinese emperor Wudi created the world’s first school of public policy. It was impossible to get the experts the empire needed without training them. In 124 BCE, Wudi created an imperial university to educate bureaucrats, which was likely the world’s first school devoted to professional training. There’s no record of whether the curriculum included employee engagement, performance measurement, and cross-organizational skills development.
The Chinese reverence for experts grew over the centuries. In fact, society held university professors in especially high regard because of their expertise. Litter-bearers carried them to meetings in sedan chairs. I’m proposing bringing back that tradition, at least for former deans.
The Romans knew how to do public law. The expertise of Roman bureaucrats endures today, from the Coliseum to aqueducts that still work. Over such a far-flung empire, however, there needed to be a system to create order, both in defining the role of the state and ways of resolving disputes among individuals. The emperor Justinian relied on his legal experts, led by the Greek Tribonian, to put together a Roman code of law in the sixth century. Modern bureaucrats didn’t invent red tape—Justinian’s code had more than a million words.
Count on the deep state to preserve the rule of law. No one really knows why, but Justinian’s massive effort disappeared for hundreds of years during the Dark Ages. But thanks to the work of monks who kept the tradition alive, the principle of the rule of law endured. In the eleventh century, Justinian’s code was rediscovered, and a monk named Gratian led an impressive effort to pull the Roman code into the not-so-Dark Ages, to weave it into the church’s canon law, and to lay the foundation for the rule of law governing modern experts.
Blame Johannes for bureaucrats’ push-back against their leaders. It’s really Johan’s fault—Johannes Gutenberg, that is. His invention of printing with movable type made it much easier to spread expertise broadly, compared with painstakingly copying how-to guides by hand, which inevitably made expertise closely held. That contraption democratized expertise and, in turn, made it much easier to contest central authority.
When it comes to the “deep state,” don’t look to America’s founders for clues. As anyone who has seen the musical Hamilton knows, the founders didn’t spend a lot of time trying to figure out how to set up a bureaucracy or how to make it accountable. That makes it a lot easier to quote the founders in saying almost anything to back up almost any claim about bureaucratic power and the deep state. That makes it easy, in turn, to pick fights about the “deep state” without having to worry about first principles in American government. Except accountability. More on that shortly.
Britain’s taste for tea is responsible for the American civil service. Thanks to the growing taste of the British for the caffeinated beverage, the English East Trading Company became a corporate behemoth whose size and power worried leaders like William Pitt the Younger. Two experts, Stafford H. Northcote and C.E. Trevelyan, wrote a report in 1854 that urged the creation of a professional civil service, based on very Chinese principles like competition for government positions through exams and appointment to positions based on merit. They proposed increasing government’s expertise so it could properly regulate the private sector, especially the trade in tea and spices. Their report laid the foundation for the creation of the American civil service system thirty years later. It did not, however, resolve the eternal British dispute about whether to add cream to the cup before or after the tea.
The rise of the American civil service and the merit system sought to end piracy in government. The American political scientist Paul van Riper hailed the rise of the merit system in forcing “an orderly retreat of parties from their prerogatives of plunder.” Swapping government officials with every election turned out to be pretty bad government. The government didn’t have the expertise it needed to govern well. It weakened its capacity to balance other major forces in society. And it fed endless temptations for corruption. The pressures from office seekers, in fact, had been so bad with every transition in power that Abraham Lincoln risked the dangers of traveling to the still-smoldering Richmond battlefield in 1865 rather than put up with the people wanting jobs in those pre-merit system days, as Tom Shoop explains.

On his way to Richmond, Lincoln told the tale of one office-seeker who asked to be named as a minister abroad. When that request failed, he countered by requesting a position as a customs officer, which was one of the most popular patronage positions available (along with local postmasters). “When he saw he could not get that,” Lincoln said, “he asked me for an old pair of trousers.”

So what does this 2,000-year tour through the role of experts in government reveal? There are basic truths that never go away:

Personal loyalty is no substitute for administrative competence in securing good government.
Accountability to the policies of individual leaders undermines the long-term competence of government, weakens the ability of leaders to secure their policy goals, and risks having power slide into the hands of private institutions that are not accountable to the people at all.
The merit system is the best safeguard for the rule of law—and vice versa.
Government without a strong merit system is messy for everyone, starting even with leaders who might have thought it was a good idea.

Donald F. Kettl is professor emeritus and former dean at the University of Maryland School of Public Policy. Experts in Government: The Deep State from Caligula to Trump and Beyond, can be downloaded for free for a limited time.

]]>

Will an attack on federal employees swing the election? Probably no

Donald F. Kettl — Fri, 13 Oct 2023 05:59:00 -0400

As the 2024 campaign has gotten rolling, a very long time before the election, there’s been an unbroken drumbeat of criticism of the “deep state” and the “unaccountability” of federal employees.

It turns out that the “deep state” gets a lot of traction among Republican voters in the swing states of Arizona and Georgia. Donald Trump lost these two states by a total of just 22,000 votes. Are Trump Republicans ready to hang with him in 2024?

In September, Democracy Corps conducted a series of focus groups in the two states. The research organization gathered four different factions of the GOP—Trump loyalists, evangelical Trump loyalists, Cheney conservatives, and Moderate Republicans. When asked about how they felt about the future of the country, they weren’t a very sanguine bunch. Words like “worried,” “concerned,” “uncertain,” and “bleak” rose to the top.

They’re especially concerned about the Constitution, which a Trump loyalist in Arizona said is “dangling by a thread” and another said is “hijacked.” An evangelical Trumper in Georgia said she thought the Constitution is “under attack.”

But it’s not criticism of federal employees that’s driving their worries. It’s prices that they view as out of control, instability at the border, the cost of the Ukraine war, and “woke” policies that they think are being forced on their kids.

When it comes to the “deep state,” a Georgia woman who’s a Trump loyalist saw a “conspiracy.” A Georgia woman who’s an evangelical Trump loyalist called it “a wolf in sheep’s clothing. Hiding right in front of our faces.” People just don’t see it because “no one cares about it. It is outside people’s everyday lives. It makes them unaccountable.”

And a Trump loyalist in Arizona said “there is a Deep State.” And, he said, there were “Marxists. Protecting their own interests.” In fact, “I believe there is a small group of people controlling world finances.”

In digging into the focus group’s views, however, sentiment was decidedly mixed. A Trump loyalist in Arizona said he would never hire federal employees in the private sector. Another Trump loyalist there said, “They don’t have a good work ethic.” Yet another: “overpaid and underqualified. Not trustworthy.” A Trump loyalist added that he believed “if you have a job for the federal government, you have a job for life. I don’t think they care about what is going on.”

Evangelical Trump loyalists in Georgia were more pointed in their views. One said she thought federal employees were “overpaid,” while another said “some of them aren’t very good.” Yet another woman added, “They seem irritated to help you.” For one Evangelical Trumper, the opinion was simple: “entitled and lazy.”

In general, however, the participants in the focus groups weren’t stirred by Trump’s promise to clean out federal employees if he’s elected. Focus group members were most concerned by their conclusion that there were just too many bureaucrats and that too many of them were lazy. On the other hand, there wasn’t general concern that federal employees were pursuing their own agenda, a core tenet of the deep state argument.

These are fascinating findings, in several respects.

First, the “deep state” critique doesn’t get a lot of traction, even among Trump supporters. From the first week of Trump’s first term, the “deep state” argument has been a steady criticism by Trump forces of the federal government, but it’s not getting far.

That certainly doesn’t mean that conservatives aren’t quietly framing an agenda around unwinding what they view as the “deep state.” That campaign, however, isn’t catching fire among even Trump’s most loyal supporters.

Second, there’s strong support for the proposition that there are just too many federal employees. However, since the mid-1970s, public opinion polls have regularly shown that Americans would rather have a smaller government and fewer services. Support for that proposition has been rising among Republicans; it’s falling among Democrats.

But when asked about which programs they’d like to cut, from education to foreign aid, there wasn’t a plurality of support for reducing spending in any of 13 different policy areas. In half of the areas, in fact, including education, veterans benefits, infrastructure, and health care, more than half of those surveyed wanted to spend even more.

Moreover, underlying the eternal debate about cutting government employees is the disconnect of that debate from the government’s mission. I dug into which federal agencies employ the most people, and it turns out that the 15 largest agencies, by employment, account for nearly 80% of all federal employees. With the exception of the IRS and the FBI, they perform functions that no one wants to cut—support for the military, Customs and Border Protection, Social Security Administration, the FAA and TSA, FEMA, federal prisons.

There’s support from Vivek Ramaswamy and some House Republicans to defund the FBI, but no one really wants to weaken the American defense against organized crime, drug cartels, or homeland security threats. And the IRS is an agency that only the U.S. Treasury truly loves, but no government in the history of the world has ever existed long without an agency to collect taxes. And no tax collection agency has ever been popular.

Third, it’s clear that many federal employees have a black eye when it comes to dealing with the people. That’s supremely ironic, because most Americans never really encounter a federal employee except when traveling through airports or violating the law. Even in the biggest entitlement programs—Social Security and Medicare—most recipients never see or talk with a fed. Banks process Social Security payments and seniors get Medicare-paid health care through their own non-governmental providers, clinics, and hospitals.

There’s a powerful case for federal employees to pay much more attention to customer service. That’s why it’s a key part of the President’s Management Agenda. But the manifest problems in government’s connections with the people are why rankings of customer experience regularly put the federal government at the very bottom of more than a dozen industries. This is certainly something the government needs to work on.

But when it comes to mobilizing worries about the “deep state” to fuel the Republican presidential campaign, as Democracy Corps found, there’s little evidence that will work.

]]>

‘Neutral competence,’ partisanship and efforts to overhaul the civil service

Donald F. Kettl — Tue, 19 Sep 2023 06:00:00 -0400

You might think that the initiative to convert career civil service employees to political appointees might have died at the end of the Trump administration. You’d be wrong.

As the 2024 presidential campaign gets into gear, conservatives are playing a short game to bring that plan back. But there’s also a long game in the works to change the rules of the civil service, which would radically transform the constitutional balance of powers. It would be a huge mistake to underestimate either gambit.

Here’s the short game. The Heritage Foundation has a long and very detailed proposal for bringing back Schedule F, the plan to turn careerists with policy roles into political appointees. James Sherk, who authored the Trump Schedule F executive order, has written a meticulous blog, “Tales from the Swamp,” which describes how he believes federal officials resisted the Trump administration’s policy ideas. Existing civil service rules, he writes “prevent meaningful accountability.”

His goal, he writes, is “returning to the original vision for the civil service: merit-based hiring and straightforward removals.” The analysis builds on a singular proposition: “Political appointees cannot simply remove intransigent employees.” Moreover, “The reformers who created the civil service wanted to avoid patronage hiring, but they also feared removal protections would entrench incompetence and insubordination.”

If the truth be told, virtually every federal manager has a tale of a poorly performing employee that they either couldn’t rid themselves of—or that they simply concluded would be too much trouble to try to dismiss. Along with many conservatives, Sherk argues that “Congress should make federal employees at-will once again,” he writes.

After all, he notes, federal employees didn’t get the right to appeal adverse decisions until the 1940s. And, in the meantime, he says that federal agencies ought to greatly expand the use of “Schedule C” authority for positions influencing policy. Schedule C employees can be dismissed at will.

But even more important is the long game the conservatives are playing. Their goal is nothing less than to establish that all federal government employees serve at the pleasure of the president—and, therefore, can be readily dismissed.

After all, the Declaration of Independence contends, in its second paragraph, that governments derive “their just powers from the consent of the governed.”

Since the passage of the Pendleton Act in 1883, which established the civil service, the system has been based on the idea of “neutral competence,” which holds that the obligation of administrators is to implement the president’s policies, using their professional skills to translate the policy of elected officials into results without a tinge of partisanship.

However, Sherk contends, “an antagonistic minority” has upset that principle. This minority, he says, works quietly behind the scenes to upset the links from voters to the president and, from there, to the work of the bureaucracy.

This two-part argument has potentially revolutionary implications, and it would radically shift the balance of power in the American political system.

In Article II of the Constitution, the founders wrote that the president “shall take Care that the laws be faithfully executed, and shall Commission all the Officers of the United States.” As the founders were debating the Constitution in 1789, moreover, James Madison said that “the lowest officers, the middle grade, and the highest, will depend, as they ought, on the president.”

The long game envisions that Congress would retain its oversight powers over the executive branch. But because Congress staggers on every substantive decision it faces, the legislative vacuum creates a huge opportunity for the expansion of the president’s power.

The federal courts would retain their role in the system. However, legislation that granted the president the expanded powers that the conservatives want would effectively deal the courts out of this balance-of-power question.

Moreover, even falling short of that big step, there are moderates like Phillip Howard who contend that public employee unions are not accountable to the president. That, in turn, interferes with the president’s “take care” powers. It therefore makes the unions unconstitutional, Howard contends. A court case advancing that argument is in the works.

Where does this all lead? The short game could be launched immediately on the swearing-in of a new Republican president in January 2025. If the Republicans also take the House and Senate, the road would be clear for the kind of radical shift of powers to the president that the conservatives envision.

If the Democrats win the presidency, the short game would be delayed, but conservative think tanks would surely continue the campaign from the trenches as they awaited their chance. The conservatives’ efforts would then move in the medium term to the courts, with an attack on public employee unions.

The right has enormous financial strength and intellectual firepower, and these forces aren’t going away. If they don’t win in the short run, they won’t let up. The 50-year campaign against Roe v. Wade demonstrates the value of this persistence.

Given the relative lack of financial support for counter-efforts in the center and on the left, and the struggle to mount both enthusiasm and ideas that contest the right’s most sweeping plans, it would be foolhardy to bet against the odds that the right, in time, would win the battle.

The result would be the biggest shift in the constitutional balance of powers since Madison crafted it back in 1789.

Donald F. Kettl is professor emeritus and former dean of the University of Maryland School of Public Policy. He is the co-author, with William D. Eggers, of Bridgebuilders: How Government Can Transcend Boundaries to Solve Big Problems.

]]>

'Long, long overdue': An oral history of the Government Performance and Results Act

Donald F. Kettl — Mon, 21 Aug 2023 06:00:00 -0400

In August 1993, President Bill Clinton signed the Government Performance and Results Act, a pioneering piece of legislation that required federal agencies to create strategic plans, set annual goals for programs, and measure actual performance against the targets. Clinton framed the law as a part of his effort to improve government performance:

It is important to restore the confidence of the American people in their government. It is important because, to the extent that our government works with greater efficiency and effectiveness and less unnecessary cost, it will strengthen the American economy as well as the bonds of our citizenship. … Everyone who has ever spent any time looking at how we do things, how decisions are made, how they tend to pile one on top of the other, year-in and year-out, without ever being examined in total or in terms of their effect, would say that this is an effort that is long, long overdue.

GPRA was a cornerstone of the Clinton administration’s Reinventing Government initiative, led by Vice President Al Gore. But over the years, it became part of not only the structure but also the culture of the federal government.

As GPRA reaches its 30th anniversary, it’s a great time to look back at how it evolved. Here, in the words of the people who were involved in the process, is the story of the birth and growth of results-oriented government.

Origin Story

The seeds of GPRA were initially bred and cultivated by a creative city manager in Sunnyvale, Calif. [John Mercer], and became the signature initiative of a fiscally conservative senator from Delaware, the late Bill Roth (he of Roth IRA fame). Until 1993, congressional Democrats consistently opposed Roth’s bill, largely driven by organized labor’s and consumer groups’ concerns that data would be used to justify de-funding worker and consumer protections. An axiom of management reform legislation is that it rarely succeeds without bipartisan support. Support by a new Democratic administration flipped the script. As I recall, the dress code for the victory celebrations was a t-shirt festooned with a cartoon image captioned “Ich bin ein Sunnyvale-er.”—Franklin Reeder, senior advisor for performance management, Office of Management and Budget

GPRA began with pilot projects in key agencies.

My recollection was that the concept for pilots was borrowed from the [Chief Financial Officers] Act, which antedated GPRA. [The CFO Act passed in 1990.]—Walter Groszyk, senior advisor for performance management, OMB

Choosing a name for the bill required some diplomacy.

John Mercer proposed “The More Bang for the Buck Act,” and I proposed “The Managing for Results Act.” I was inspired by the successful Managing for Results initiative I observed when visiting Australia to understand its performance system, as part of the background research on performance improvement initiatives that I was leading at [the General Accounting Office, later the Government Accountability Office]. The final result was “The Government Performance and Results Act”—a great compromise!—John Kamensky, deputy director of the National Partnership for Reinventing Government

President Clinton speaks at a Reinventing Government event at the State Department. (William Philpott/AFP via Getty Images)

Five-Year Timeline

Shortly after Morley Winograd arrived to replace Elaine Kamarck as Vice President Gore’s lead advisor on reinventing government in late 1997, I had the opportunity to brief him on GPRA and the timeline for its implementation. When the law was passed, it included a five-year time frame for full implementation. This meant that the first set of agency annual performance reports would be released publicly and to the Congress in the spring of 2000. Morley, looking ahead to Vice President Gore’s future presidential primary campaign, roared “who was the #!@*&% idiot who came up with that timing!” I meekly noted that I did help draft the bill. And survived his angry glare. He then buckled down to ensure the reports were well-crafted and meaningful.—John Kamensky

A number of people have asked how GPRA's five-year implementation timeline came to be. There are at least three reasons. The first was a "good government" attempt to learn to crawl before you walked. A number of other countries like Australia, New Zealand, Great Britain and others each began with some sort of experiment. It was hoped that by allowing agencies some time to experiment with performance measurement and strategic planning they could learn to understand how it could be applied to their programs.

A second reason was that then-OMB Director Dick Darman had instructed me, in no uncertain terms, that I should make sure that Bill Roth, who was Chairman of the Senate Governmental Affairs Committee, was "happy," but that I should "do nothing." I threaded that needle with the pilots, which meant that GPRA was optional for the first five years. This way we were able to keep Roth "happy" and yet ensure that the law would have no practical effect during Darman's tenure in the [George H.W.] Bush administration.

A third reason was that, at that time, all legislation needed to be "scored" for its cost. The GPRA bill was able to be scored a negligible amount because the pilots were voluntary. The first requirement for strategic plans and the use of performance measurement was pushed back five years.—Jonathan Breul, senior advisor to the deputy director for Management, OMB

Stoplight Scorecard

GPRA cleared a critical hurdle with the 2000 election. The newly elected president, George W. Bush, could have killed a program so closely associated with Gore, who he had defeated. Presidential transitions are often where a previous administration’s initiatives go to die. Instead, Bush transformed GPRA, with an emphasis on red-yellow-green color-coded scorecards grading each agency’s performance.

In the late 1990s, the National Partnership for Reinventing Government created an initiative to work closely with about 30 bureau-level agencies to improve their performance, like the Veterans Benefits Administration and the Patent and Trademark Office. We helped them develop rudimentary balanced scorecards using customer and employee survey data, and the performance data from their GPRA reporting systems. We crafted red-yellow-green stoplight scorecards for each agency based on these data and wanted to make them public, but a career OMB executive nixed the idea. Winograd said, “The president doesn’t want to be seen as choosing among his children.”

President George W. Bush's scorecard aimed to visualize agency progress on assessing performance.

Then a few years later, OMB under newly-elected President George W. Bush began publicly publishing red-yellow-green scorecards showing the progress of major agencies towards meeting the performance criteria set out in the newly created President’s Management Agenda. I was out of government by then, but in a call with a career executive in OMB, I asked about the new scorecards, and he said, “Well, this is a different president.”—John Kamensky

As the political [appointees] stuck with implementing GPRA, we faced agency attitudes that ranged from strong reluctance to put plans and objectives together publicly to passionate lip service from some who complied but were very careful about exposing themselves. I remember the first strategic plans, which were less than strategic and barely plans. The famous stoplight in George W. Bush's administration and the continuing push [for implementation] have brought us to this much better place.—G. Edward DeSeve, deputy director for management and controller, OMB

Reforming the Reforms

GPRA continued to evolve and, in 2010, Congress passed the GPRA Modernization Act, or GPRA-MA. The new legislation, as the Government Accountability Office explained it, called for:

A more coordinated and crosscutting approach to achieving meaningful results
Efforts to address weaknesses in major management functions
Ensuring performance information is both useful and used in decision-making
Sustained leadership commitment and accountability for achieving results
Engaging Congress in identifying management and performance issues to address

The act wove performance measurement far more deeply into agency operations.

OMB gets a call from Congressman Henry Cuellar’s staff. Cuellar wrote his doctoral dissertation on "A Comparative Analysis of Legislative Budget Oversight: Performance-Based Budgeting in the American States.” [Nancy] Pelosi had given Cuellar a green light to move a bill through the House. He decided to move a bill on performance management and successfully did. Amy Edwards Holmes was staffing Sen. Mark Warner, who launched Virginia Performs as governor. Amy and other key Warner staff work with the OMB team to amend GPRA 1993 and the Cuellar bill in what I continue to think are very sensible ways, informed by lessons learned on GPRA 1993 implementation along the way. Welcome, GPRA Modernization Act!—Shelley Metzenbaum, associate director for performance and personnel management, OMB

President Obama signs the GPRA Modernization Act in 2010. Behind him are OMB's Jeff Zients and Shelley Metzenbaum. (Pete Souza/White House)

My involvement with GPRA and later GPRA-MA during a 25-year career at the Department of Transportation began in 1995 when I was hired as a program analyst to coordinate the development of a strategic implementation plan for the National Highway Traffic Safety Administration. It continued until 2020, when I retired as the senior advisor for strategic management at the Federal Highway Administration.

During my time at DOT, I was involved in the development of every department strategic plan, annual performance plan and report, modal administration multi-year strategic plans, annual budget integration and implementation plans, and unit-level operating plans and reports. These efforts were met with differing levels of interest and acceptance among executives, managers and staff. The biggest challenge was overcoming the compliance mentality and ensuring that the plans were useful and relevant to issues confronting the agency on a daily basis. Our success hinged largely on the managerial style and personal preferences of senior leadership. However, I can say that GPRA and GPRA-MA were the backstops that sustained these efforts.—J. Woody Stanley, senior advisor, strategic management, Federal Highway Administration

My favorite memory of all this is meeting with leaders at the Labor Department to discuss GPRA and performance measures. The attorneys present said that there was no way to measure what they did. I said that, in that case, we could just do without them and see how it went. At which point they began to list the things they did and began to talk about how to measure the impact.—John Koskinen, deputy director for management, OMB

I remind myself of the survey of over 100 federal components that John Kamensky did while at GAO before GPRA. He asked a series of questions about the existence of plans, goals, performance measures, etc. He got back generally positive results but when he went to verify the responses, he found many were overblown. The internal GAO reaction was that the agencies lied on the survey. John said actually it is much worse—they do not know the difference between a strategic plan and an annual work plan, or the difference between an output and an outcome. This finding contributed to the decision to include pilots in GPRA and to have a long lead time for government-wide implementation—to give agencies time and experience to build the capacity to manage for results.—J. Christopher Mihm, managing director, strategic issues, GAO

The Long View

Kudos to all who were there at the origins, and those who have worked since then to make this a durable, workable framework. The ability to build from and modify that framework was a testimony to the efforts of a group of bipartisan actors in and out of government who genuinely cared about evidence and performance, could find common ground on those topics, and talk to each other over the course of many years. I worry if these sorts of networks still exist, but hope springs eternal.—Donald Moynihan, McCourt Chair, McCourt School of Public Policy, Georgetown University

Continued investment in the Performance.gov website has enabled the public to explore strategic goals and objectives not just by agency but by crosscutting themes. It also makes more data available on the Biden administration and agency priorities such as customer experience, permitting, and personnel vetting reforms.

In sum, GPRA laid the groundwork for GPRA modernization, which incorporated lessons learned from other governments’ efforts to implement performance management systems. It shifted focus from a supply-side approach of producing information towards a demand-driven model that centered on supporting leadership in identifying and accomplishing their top organizational priorities, including by aligning the strategic planning process with Presidential transitions. Furthermore, GPRA and OMB’s Circular A-11 guidance have clarified the roles and responsibilities of key actors, including agency chief operating officers, performance improvement officers and goal leaders. It has institutionalized agency routines that engage these and other officials in the use of performance and evidence to inform key program decisions.

Donald F. Kettl compiled and edited this oral history based on an email conversation among the key players in GPRA’s creation and implementation. He is the the co-author, with William D. Eggers, of Bridgebuilders: How Government Can Transcend Boundaries to Solve Big Problems.

]]>

What the United States can learn from Australia’s recent catastrophic failure to listen to career government experts

Donald F. Kettl — Mon, 31 Jul 2023 06:00:00 -0400

Debate is heating up over proposals to undermine the expertise of government career officials, with renewed criticism of the “deep state”—an issue that surfaced during the Trump administration when political appointees claimed career officials were attempting to sabotage presidential policy. Along with that comes planning by conservative Republicans to bring back Schedule F in a new Republican administration. This plan, launched in the last weeks of the Trump administration and undone by the Biden administration, would allow political appointees to replace 50,000 feds—or more—in positions that influence policy.

If rooting out the “deep state” by making government employees more politically accountable is such a good idea, why aren’t more countries doing it? In fact, why aren’t any other countries doing it?

The “deep state” complaint isn’t just an American issue. The former chief adviser to British Prime Minister Boris Johnson, Dominic Cummings, contended that “deeply entrenched institutions” control “the vast majority of government” in the United Kingdom. In Canada, political scientist Barry Cooper analyzed “the deep state in operation.” Critics of new Brazilian President Luiz Inácio Lula da Silva are convinced that his victory over Jair Bolsonaro was the product of “deep state” meddling in the electoral process.

French President Emmanuel Macron worries that “deep state” diplomats are sabotaging his foreign policy. Opponents of Indian Prime Minister Narendra Modi charged that the country’s “deep state” is “chewing” the Indian government, much like what happened in Pakistan, they said.

But an Australian crisis shows why other countries aren’t in any rush to follow the American debate about reducing the political independence of government experts. In Australia, the government set off on a radical plan to reduce overpayment of government benefits in 2016. But government ministers—overseen by Scott Morrison, who was social services minister from 2014 to 2015 and would later become prime minister—didn’t pay much attention to their career experts in creating the plan.

The Australian government had been manually searching for overpayments in programs for retirees, people with disabilities and students, among others. The 2016 program used algorithms to search out overpayments and send the bills. Christened “Robodebt,” the algorithm checked each individual’s payment against the average income of people in similar circumstances. If the algorithm determined that the person was likely overpaid by the government, it generated a bill.

Robodebt allowed the government to review 20,000 cases per week, instead of the 20,000 cases per year in the manual system it replaced. Government officials no longer had to contact employers to obtain data on employment history and payroll amounts, and the government no longer had to prove an individual had been overpaid. Instead, individuals had to prove that they had received the correct amount. If individuals didn’t pay quickly, debt collectors went to work.

The government launched Robodebt fast and claimed credit for catching recipients who had benefited from mistakes in the system. But many people receiving the notices were distraught. They often had to come up with big payments in just a few weeks. Some people had to sell their cars or take out loans, which was a huge burden on some of the country’s neediest residents. Others drained their meager savings. At least three people committed suicide, a Royal Commission found in a devastating 1,000-page report.

An investigation revealed that some repayment notices were incorrect. Some simply were false. Moreover, a 2019 court challenge found that Robodebt had violated important provisions of Australian law.

In July 2023, the Royal Commission pointed to “Robodebt’s unfairness, probable illegality, and cruelty.” When problems surfaced along the way, the commission concluded, “the path taken was to double down, to go on the attack in the media against those who complained and to maintain the falsehood that in fact the system had not changed at all” from the previous system.

How did such monumental errors happen? Australia has a long tradition of “frank and fearless advice” from career government experts to political ministers. In Robodebt, the ministers made clear that they weren’t interested in advice that complicated their speedy implementation of the policy, and they refused to listen to analysis of the escalating problems. As Andrew Podger, former head of the Australian civil service pointed out, there was excessive responsiveness of career experts to ministers and little appetite on the part of ministers to listen to the career experts.

In addition to the damage Robodebt inflicted on trust in Australia’s government, Podger also pointed to the erosion of confidence in the Australian public service. The career service, he said, “plays a critical democratic role in serving the elected government and administering its policies and programs.” Advice was less than “frank and fearless” because careerists were uncomfortable raising challenges to the government’s policy. Government officials charged ahead only to discover later that their plan wouldn’t work and was probably illegal.

The result was one of the biggest governance crises in recent Australian history. Effective democratic rule in complex programs, it turns out, requires bureaucratic expertise—and the willingness of political appointees to use that expertise in both framing and implementing policy.

Elected officials, of course, have the right—indeed, the responsibility—to pursue their promises once they’re in office. But failing to collaborate with career experts to figure out what works, what doesn’t, and what’s legal only paves the road to disaster, as the Robodebt case makes clear.

The “deep state” arguments have percolated around the globe. But if the counter to a powerful bureaucracy is to push experts aside, there can be a fearsome cost in government performance and trust.

Australia now finds itself trying to escape the crisis by rebalancing the relationship between political officials and career civil servants—between political responsiveness and expert administration.

As former British adviser Cummings contended, the fact that “brilliant 30-year-old women who no one’s heard of or who no one’s elected are running things is actually for the good.” Putting faith in that expertise, in service of the government of the day, is the key to tackling the vast complexities of 21st-century government.

In Australia, politicians didn’t want advice. Careerists were afraid to give it. Disaster ensued. And that’s why no other major democracy is racing to make it easier to fire experts in the bureaucracy.

The Down Under parable is worth remembering here in the United States. Just what role should career experts play in shaping policy? That’s always been the subject of deep debate, and will continue to be during the 2024 election cycle. But Australia’s wrenching problems serve as a warning to Americans that ignoring the advice of the career bureaucracy can pave the road to an enormous crisis that only undermines trust in the very politicians seeking big policy change.

]]>

The People Behind the Federal Spending Cut Dilemmas

Donald F. Kettl — Thu, 25 May 2023 11:00:00 -0400

Everyone knows the basic story about why it’s so tough to cut government spending. Entitlements take up nearly three-fourths of all federal spending, and squeezing cuts out of the other one-fourth of the budget creates a profound dilemma. Nothing makes that more evident than a clear-eyed look at where federal employees work.

More than two-thirds of feds do jobs that realistically can’t be cut. There’s Social Security and services for veterans, as well as homeland security functions (including Customs and Border Protection, Immigration and Customs Enforcement, Citizenship and Immigration Services, and the Coast Guard). We’re already having a hard time staffing air traffic control towers, especially going into the summer travel season, so we can’t cut the Federal Aviation Administration. And, of course, there’s the Internal Revenue Service, the agency everyone loves to hate but that collects the money that funds everything else.

That adds up to 37% percent of all federal employees. Then add to that the civilian employees who manage defense policy and the armed services. That’s another 34%. We can certainly find economies there, but not without looking carefully at defense functions and programs, and no one seems in a mood to do that. Combine these two large buckets of federal employees, and we get 71% of the federal workforce. We can’t empty the bucket for the functions we wouldn’t want to cut.

There aren’t many employees left in the rest of government, and cutting them would clobber key functions. No vacationer wants to drive up to the national park gate and find a “park closed” sign. No homeowner would want a major fire in a national forest to threaten their home. Farmers count on getting their payments from the Agriculture Department on time and coastal residents want good forecasts from the National Weather Service as they near hurricane season.

So when it comes to making cuts, most federal employees do jobs we can’t cut. Cutting the employees that budget-cutters might consider wouldn’t save much money. The same is true, of course, for the programs in the discretionary part of the government that they manage.

In fact, when pollsters start drilling down into the specifics, support for cutting spending evaporates when Americans are asked about which programs they would eliminate. A 2019 Pew Research Center survey asked about 13 different policy areas. There wasn’t support for cutting anything, from education to foreign aid. In fact, Americans favored increasing spending in every area, except for assistance to the unemployed. And that was before COVID and its effects on the economy.

There’s nothing that would undermine trust in government more than for the government to have programs the people want but that they can’t get because decisionmakers have gutted the people we need to deliver them.

Rep. Chip Roy, R-Texas, for example, in a May 24 memo to colleagues urged Republican leaders to hold the line. “We want to cut funding for the woke federal bureaucracy interfering with Americans’ ability to live free and prosper economically to pre-COVID levels,” he said. But these numbers of federal employees are the pre-COVID levels.

We can’t cut the employees who do things that no one really wants to cut. That’s a non-starter. We can cut the employees who do other things in the discretionary part of the budget, but that wouldn’t raise much money and would create problems that would be hard to defend.

Or we can cut employees in the hope of undermining programs that some officials don’t like, but that never saves much money and always annoys people who depend on those programs. We can make across-the-board cuts, but budget wonks know that there’s no such thing as an across-the-board cut, because there inevitably are programs that no one can politically touch.

There probably has never been a more treacherous policy minefield. But when it comes to finding our way through it, the worst thing we could do is to look for a shortcut by slashing federal employees without thinking about the mission we want them to perform.

Donald F. Kettl is the co-author, with William D. Eggers, of Bridgebuilders: How Government Can Transcend Boundaries to Solve Big Problems.

]]>